Cross Reference: /src/sys/nfs

History log of /src/sys/nfs
Revision	Date	Author	Comments
1.4	27-Dec-2006	yamt	remove nqnfs.
1.3	03-Jan-2006	yamt	branches: 1.3.18; don't install nfs_var.h.
1.2	26-Nov-2002	lukem	branches: 1.2.22; 1.2.34; Remove KDIR=, since SYS_INCLUDE=symlinks and KDIR are not supported any more.
1.1	12-Jun-1998	cgd	branches: 1.1.26; 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.
1.1.26.1	11-Dec-2002	thorpej	Sync with HEAD.
1.2.34.1	15-Jan-2006	yamt	sync with head.
1.2.22.2	30-Dec-2006	yamt	sync with head.
1.2.22.1	21-Jun-2006	yamt	sync with head.
1.3.18.1	12-Jan-2007	ad	Sync with head.
1.15	17-May-2018	thorpej	Default NFS mounts to using TCP transport instead of UDP. PR kern/53166
1.14	11-Oct-2014	uebayasi	branches: 1.14.18; Define filesystem attributes with vfs dependency.
1.13	02-Mar-2010	pooka	branches: 1.13.20; don't create unused fs_nfs.h
1.12	02-Mar-2010	pooka	Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.11	31-Dec-2009	christos	branches: 1.11.2; handle the nuidhash_max lossage differently
1.10	31-Dec-2009	christos	nuidhash_max is needed by sys_nfssvc
1.9	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.8	27-Dec-2006	yamt	branches: 1.8.40; 1.8.44; 1.8.50; 1.8.54; remove nqnfs.
1.7	11-Dec-2005	christos	branches: 1.7.20; merge ktrace-lwp.
1.6	23-Sep-2005	jmmv	Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.5	26-Feb-2005	perry	branches: 1.5.4; nuke trailing whitespace
1.4	12-Dec-2004	bouyer	branches: 1.4.2; 1.4.4; The macro used for static server address is NFS_BOOTSTATIC_SERVADDR, not NFS_BOOTSTATIC_SADDR. From Xen source distribution. XXX NFS_BOOTSTATIC* doesn't seem to be documented anywhere ...
1.3	11-Mar-2004	cl	branches: 1.3.6; Add static nfs boot configuration, from the kernel config file or from a driver selectable callback function. This is used in the Xen port to allow controlling the domain's network setup from the domain building environment at domain creation (vs. having to maintain/change this on a dhcp server). The Xen network driver parses a command line passed in from the domain builder.
1.2	23-Oct-2002	jdolecek	branches: 1.2.6; 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
1.1	16-Apr-2002	thorpej	branches: 1.1.6; 1.1.8; Cleanup how file system configuration information is declared, grouping related information together, with the file system code itself. This is just low-hanging fruit -- more to come.
1.1.8.3	30-Sep-2002	jdolecek	add support for kevents to NFS to detect file changes on server by other NFS clients, polling kernel thread is used to periodically check for attribute changes of watched files; the NFS server is only contacted when the vnode expires from local attrcache (which takes 5-60 seconds currently), to keep network&CPU overhead low the routine checking for remote changes is quite simplistic, but hopefully doing it's job well enough
1.1.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.8.1	16-Apr-2002	jdolecek	file files.nfs was added on branch kqueue on 2002-06-23 17:51:46 +0000
1.1.6.3	11-Nov-2002	nathanw	Catch up to -current
1.1.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.1.6.1	16-Apr-2002	nathanw	file files.nfs was added on branch nathanw_sa on 2002-06-20 03:50:00 +0000
1.2.6.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.6.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.6.4	18-Dec-2004	skrll	Sync with HEAD.
1.2.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.6.1	03-Aug-2004	skrll	Sync with HEAD
1.3.6.1	06-Apr-2005	tron	Pull up revision 1.4 (requested by bouyer in ticket #1036): The macro used for static server address is NFS_BOOTSTATIC_SERVADDR, not NFS_BOOTSTATIC_SADDR. From Xen source distribution. XXX NFS_BOOTSTATIC* doesn't seem to be documented anywhere ...
1.4.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.2.1	29-Apr-2005	kent	sync with -current
1.5.4.2	30-Dec-2006	yamt	sync with head.
1.5.4.1	21-Jun-2006	yamt	sync with head.
1.7.20.1	12-Jan-2007	ad	Sync with head.
1.8.54.1	19-Jan-2009	skrll	Sync with HEAD.
1.8.50.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.8.44.2	11-Mar-2010	yamt	sync with head
1.8.44.1	04-May-2009	yamt	sync with head.
1.8.40.1	17-Jan-2009	mjf	Sync with HEAD.
1.11.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.13.20.1	03-Dec-2017	jdolecek	update from HEAD
1.14.18.1	21-May-2018	pgoyette	Sync with HEAD
1.10	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.9	14-Mar-2009	dsl	branches: 1.9.100; Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.8	11-Dec-2005	christos	branches: 1.8.74; 1.8.84; 1.8.90; merge ktrace-lwp.
1.7	22-May-2004	jonathan	branches: 1.7.12; Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.6	05-May-2003	yamt	branches: 1.6.2; keep things not needed by userland in #ifdef _KERNEL. (e.g. prototypes for in-kernel functions)
1.5	20-Oct-1996	fvdl	Enhancements from Matthias Drochner: - Try V3 first for diskless booting. Fall back to V2 if V3 fails. - optionally (option NFS_BOOT_TCP) try a TCP mount first for diskless booting. Fall back to UDP if it fails. - Enable switching between UDP and TCP for remounts.
1.4	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.3	24-Apr-1995	gwr	Fixed RPC code to deal with RPC messages larger than one mbuf.
1.2	26-Oct-1994	cgd	new RCS ID format.
1.1	26-Sep-1994	gwr	Do the first BOOTPARAM RPC call to the broadcast address instead of using the address of the RARP server because a BOOTPARAM server might not be running on the machine that sent the RARP reply.
1.6.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.2.1	03-Aug-2004	skrll	Sync with HEAD
1.7.12.1	21-Jun-2006	yamt	sync with head.
1.8.90.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.8.84.1	28-Apr-2009	skrll	Sync with HEAD.
1.8.74.1	04-May-2009	yamt	sync with head.
1.9.100.1	02-Aug-2025	perseant	Sync with HEAD
1.44	20-Oct-2024	mlelstv	MBUFTRACE
1.43	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.42	10-Jun-2016	ozaki-r	branches: 1.42.54; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.41	21-May-2015	rtr	change nfs_boot_sendrecv to take sockaddr_in * instead of mbuf * fixes m_serv (single mbuf leak) leak in kern/subr_tftproot.c
1.40	09-May-2015	rtr	when calling nfs_boot_sendrecv pass NULL for pointers instead of 0
1.39	27-Mar-2015	hikaru	m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.38	06-Mar-2015	maxv	Fix uninitialized variable. Found by The Brainy Code Scanner in FreeBSD.
1.37	15-Mar-2009	cegger	branches: 1.37.18; 1.37.22; 1.37.38; 1.37.40; ansify function definitions
1.36	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.35	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.34	14-Mar-2009	dsl	Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.33	24-Apr-2008	ad	branches: 1.33.2; 1.33.10; 1.33.16; Merge the socket locking patch: - Socket layer becomes MP safe. - Unix protocols become MP safe. - Allows protocol processing interrupts to safely block on locks. - Fixes a number of race conditions. With much feedback from matt@ and plunky@.
1.32	04-Mar-2007	christos	branches: 1.32.36; 1.32.38; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.31	15-Apr-2006	christos	branches: 1.31.14; Coverity CID 2445: Only set from_p if we succeed so that we free it on error.
1.30	11-Dec-2005	christos	branches: 1.30.4; 1.30.6; 1.30.8; 1.30.10; 1.30.12; merge ktrace-lwp.
1.29	26-Feb-2005	perry	branches: 1.29.4; nuke trailing whitespace
1.28	22-May-2004	jonathan	branches: 1.28.4; 1.28.6; Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.27	26-Feb-2003	matt	branches: 1.27.2; Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.26	22-Sep-2002	jdolecek	don't seem to need <sys/conf.h> or <net/if.h> here
1.25	10-Nov-2001	lukem	add RCSIDs
1.24	12-Jun-2001	wiz	branches: 1.24.2; 1.24.6; receive, not recieve
1.23	09-Aug-1998	perry	branches: 1.23.24; bzero->memset, bcopy->memcpy, bcmp->memcmp
1.22	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.21	30-Sep-1997	drochner	Use functions (shared with DHCP boot) in nfs_boot.c.
1.20	29-Aug-1997	gwr	Supporting changes for the new BOOTP support in nfs_mountroot.
1.19	20-Oct-1996	fvdl	branches: 1.19.10; Enhancements from Matthias Drochner: - Try V3 first for diskless booting. Fall back to V2 if V3 fails. - optionally (option NFS_BOOT_TCP) try a TCP mount first for diskless booting. Fall back to UDP if it fails. - Enable switching between UDP and TCP for remounts.
1.18	13-Oct-1996	christos	revert kprintf changes
1.17	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.16	14-Aug-1996	thorpej	Another %lx -> %x for ntohl()
1.15	10-Jul-1996	cgd	print result of ntohl/htonl as a long.
1.14	14-Jun-1996	cgd	avoid unnecessary checks of m_get/MGET/etc.'s return values. When they're called with M_WAIT, they are defined to never return NULL.
1.13	07-Jun-1996	cgd	fix two bugs (the latter potentially fatal) in xdr_string_encode(): (1) if length needed was > MCLBYTES, an mbuf would be lost, and (2) the wrong check was being used to determine if MCLGET succeeded.
1.12	18-Feb-1996	fvdl	branches: 1.12.4; Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.11	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.10	08-Aug-1995	gwr	Use RPCAUTH_UNIX in requests to please picky NFS servers.
1.9	20-May-1995	mycroft	Use fxdr_() and txdr_() macros to do byte order conversions.
1.8	24-Apr-1995	gwr	Fixed RPC code to deal with RPC messages larger than one mbuf.
1.7	26-Sep-1994	gwr	Do the first BOOTPARAM RPC call to the broadcast address instead of using the address of the RARP server because a BOOTPARAM server might not be running on the machine that sent the RARP reply.
1.6	12-Aug-1994	cgd	fix typo
1.5	11-Aug-1994	gwr	Diskless boot will now bind the local socket to a reserved port to satisfy picky servers. Also fix some missing initializations. (Thanks to Chuck Cranor for PR#394 -- now fixed.)
1.4	30-Jun-1994	pk	branches: 1.4.2; error codes are in network order too.
1.3	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.2	13-Jun-1994	gwr	New diskless boot code (uses RARP, bootparamd).
1.1	18-Apr-1994	glass	revised nfs diskless support. uses bootp+rpc to gather parameters
1.4.2.2	12-Aug-1994	mycroft	update from trunk
1.4.2.1	11-Aug-1994	mycroft	update from trunk
1.12.4.1	07-Jun-1996	cgd	pull up from trunk: >fix two bugs (the latter potentially fatal) in xdr_string_encode(): >(1) if length needed was > MCLBYTES, an mbuf would be lost, and >(2) the wrong check was being used to determine if MCLGET succeeded.
1.19.10.2	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.19.10.1	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.23.24.3	18-Oct-2002	nathanw	Catch up to -current.
1.23.24.2	14-Nov-2001	nathanw	Catch up to -current.
1.23.24.1	21-Jun-2001	nathanw	Catch up to -current.
1.24.6.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.24.2.2	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
1.24.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.27.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.27.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.27.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.27.2.1	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
1.29.4.2	03-Sep-2007	yamt	sync with head.
1.29.4.1	21-Jun-2006	yamt	sync with head.
1.30.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.30.10.1	19-Apr-2006	elad	sync with head.
1.30.8.1	24-May-2006	yamt	sync with head.
1.30.6.1	22-Apr-2006	simonb	Sync with head.
1.30.4.1	09-Sep-2006	rpaulo	sync with head
1.31.14.1	12-Mar-2007	rmind	Sync with HEAD.
1.32.38.1	18-May-2008	yamt	sync with head.
1.32.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.33.16.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.33.10.1	28-Apr-2009	skrll	Sync with HEAD.
1.33.2.1	04-May-2009	yamt	sync with head.
1.37.40.3	09-Jul-2016	skrll	Sync with HEAD
1.37.40.2	06-Jun-2015	skrll	Sync with HEAD
1.37.40.1	06-Apr-2015	skrll	Sync with HEAD
1.37.38.2	21-Apr-2015	snj	Pull up following revision(s) (requested by maxv in ticket #713): sys/dev/ic/bwi.c: revision 1.26 sys/nfs/krpc_subr.c: revision 1.38 Fix uninitialized variable. Found by The Brainy Code Scanner in FreeBSD. -- Fix a double free. "Suggested" by Brainy. ok rjs@ riastradh@
1.37.38.1	06-Apr-2015	snj	Pull up following revision(s) (requested by hikaru in ticket #656): sys/kern/subr_tftproot.c: revision 1.14 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/nfs_bootdhcp.c: revision 1.53 sys/nfs/nfsdiskless.h: revision 1.31 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.37.22.1	03-Dec-2017	jdolecek	update from HEAD
1.37.18.1	16-Apr-2015	msaitoh	Pull up following revision(s) (requested by hikaru in ticket #1287): sys/kern/subr_tftproot.c: revision 1.14 via patch sys/nfs/nfsdiskless.h: revision 1.31 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_bootdhcp.c: revision 1.53 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.42.54.1	02-Aug-2025	perseant	Sync with HEAD
1.81	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.80	05-Dec-2021	msaitoh	branches: 1.80.10; s/runable/runnable/
1.79	04-Jun-2021	hannken	Add flag/command NFSSVC_REPLACEEXPORTSLIST to nfssvc(2) system call. Works like NFSSVC_SETEXPORTSLIST but supports "mel_nexports > 1" and will atomically update the complete exports list for a file system.
1.78	22-Aug-2018	msaitoh	branches: 1.78.16; 1.78.20; - Cleanup for dynamic sysctl: - Remove unused _NAMES macros for sysctl. - Remove unused _MAXID for sysctls. - Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and use them on all m68k machines.
1.77	25-Jan-2018	riastradh	branches: 1.77.2; 1.77.4; Use a random opaque cookie, not kva pointer, for nfssvc(2). (What were they smoking?!) I suspect most of this is actually dead code that wasn't properly amputated along with the rest of the gangrene of NFSKERB a decade ago, but I'm out of time to investigate further. If someone else wants to kill NFSSVC_AUTHIN/NFSSVC_AUTHINFAIL and the rest of the tentacular kerberosity, be my guest. Noted by Silvio Cesare of InfoSect.
1.76	21-Jan-2018	christos	PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.75	20-Apr-2015	riastradh	branches: 1.75.10; Nix LEASE_READ/LEASE_WRITE from <sys/vnode.h>.
1.74	24-Apr-2014	christos	branches: 1.74.4; PR/48426: Dimitris Karagkasidis: Convert to sized, unsigned types. Ideally we could use uint64_t, but for compatibility and performance we don't (for now)
1.73	01-Mar-2013	joerg	branches: 1.73.6; 1.73.10; Retire OSI network stack. OK core@
1.72	02-Mar-2010	pooka	branches: 1.72.10; 1.72.20; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.71	19-Jan-2010	yamt	branches: 1.71.2; remove unused r_timer member.
1.70	22-Oct-2008	matt	Change NFS to use a RB-tree for its FH->nfsnode lookups.
1.69	04-Dec-2007	yamt	branches: 1.69.12; 1.69.16; 1.69.22; merge non-intrusive nfs changes from vmlocking.
1.68	28-Oct-2007	yamt	branches: 1.68.2; 1.68.4; make NFS_ATTRTIMEO a function.
1.67	02-Jun-2007	yamt	branches: 1.67.6; 1.67.8; 1.67.12; add some #include.
1.66	01-Jun-2007	dogcow	it seems like a good idea to include <sys/condvar.h>, as we're using them...
1.65	01-Jun-2007	yamt	use mutex and condvar.
1.64	30-Apr-2007	yamt	remove R_GETONEREP.
1.63	04-Mar-2007	christos	branches: 1.63.2; 1.63.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.62	28-Dec-2006	yamt	branches: 1.62.2; remove several nqnfs definitions.
1.61	27-Dec-2006	yamt	remove nqnfs.
1.60	04-Sep-2006	yamt	branches: 1.60.2; remove (void *) cast from NFSRVFH_DATA as it sometimes discards const qualifier. pointed out by Havard Eidnes. (it wasn't detected by in-tree gcc4. seems like a compiler bug.)
1.59	02-Sep-2006	yamt	nfsd: deal with variable-sized filehandles.
1.58	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.57	07-Jun-2006	kardel	branches: 1.57.2; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.56	14-May-2006	elad	branches: 1.56.2; integrate kauth.
1.55	03-Jan-2006	yamt	branches: 1.55.2; 1.55.4; 1.55.6; 1.55.8; 1.55.10; move function prototypes from nfs.h to nfs_var.h.
1.54	03-Jan-2006	yamt	nfssvc_nfsd: reduce a chance for a slow peer to capture all our threads. instead of sleeping to wait for the socket to send our reply, just hand-off our reply to the thread which is holding the socket.
1.53	03-Jan-2006	yamt	improve nfsd locking. - don't bother to take nfs_sndlock when doing nfsrv_rcv. unlike client, we never reconnect. - nfsrv_getstream: fix the case that m_split sleeps. - free socket in nfsrv_slpderef rather than nfsrv_zapsock. fix race with nfssvc_nfsd. - while i'm here, remove NFSD_WAITING and NFSD_REQINPROG as they are redundant. - some comments and assertions.
1.52	11-Dec-2005	christos	branches: 1.52.2; merge ktrace-lwp.
1.51	25-Sep-2005	jmmv	Add some COMPAT_30 code to let old mountd binaries work after the NFS exports rototill.
1.50	23-Sep-2005	jmmv	Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.49	18-Sep-2005	christos	Allow turning off the attribute cache.
1.48	26-Oct-2004	yamt	branches: 1.48.10; 1.48.12; remove #if 0'ed out definition of VA_EXCLUSIVE.
1.47	26-Oct-2004	yamt	remove an unused macro, NMOD.
1.46	12-May-2004	yamt	g/c unused NFS_*ALLOC defines.
1.45	10-May-2004	yamt	don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.44	06-Dec-2003	jonathan	branches: 1.44.2; 1.44.4; Commit message for previous revision to sys/nfs/nfs.h: Increase NFS_MAXRAHEAD to 32. With 32k read or write requests, that amounts to 1 Mbyte of read-ahead, enough to cover about 10 ms latency at gigabit Ethernet speeds. Increase the table of nfsiod kthreads (NFS_MAXASYNCDAEMON) from 20 to 128, to match the raised value of NFS_MAXRAHEAD. (Making the limit dynamic requires replacing the compile-time array with a dynamic structure.) Add a comment explaining that each read-ahead requires an I/O thread. Wrap both parameters with an #ifdef <parameter>/#endif, to allow hand-tuned values or (later) a kernel config-file option override.
1.43	06-Dec-2003	jonathan	* empty log message *
1.42	26-Sep-2003	yamt	change n_mtime from time_t to timespec in order to improve cache consistency. (1 second granularity is too loose these days.)
1.41	16-Aug-2003	yamt	current trylater/jukebox retry delay is way too long and it has a bug in the backoff calculation. so, - clip it to 1-60 sec. (suggested by Rick Macklem) - use a constant multiplier instead of nfs_backoff, which is already exponential. - move some related constant definations to nfs.h from nqnfs.h and prefix with NFS_ instead of NQ_ because they are not nqnfs-specific.
1.40	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.39	29-Jun-2003	fvdl	branches: 1.39.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.38	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
1.37	25-Jun-2003	yamt	- instead of scaning a list when looking up {a idle thread, a socket with pending requests}, maintain dedicated list of them. - add spin locks.
1.36	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.35	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.34	01-Dec-2002	matt	Don't define VA_EXCLUSIVE if it's not defined. If we do, it'll be a different value from the one in <sys/vnode.h>
1.33	12-May-2002	matt	Eliminate commons
1.32	29-Nov-2001	christos	use struct uucred in nfsd_svcargs so that we don't break the sys_nfssvc() ABI.
1.31	15-Sep-2001	chs	add a new VFS op, vfs_reinit, which is called when desiredvnodes is adjusted via sysctl. file systems that have hash tables which are sized based on the value of this variable now resize those hash tables using the new value. the max number of FFS softdeps is also recalculated. convert various file systems to use the <sys/queue.h> macros for their hash tables.
1.30	03-Aug-2001	jdolecek	branches: 1.30.2; bound check mount args more thoroughly
1.29	01-Jul-2001	gmcgarry	branches: 1.29.2; Introduce NFS_DEFAULT_NIOTHREADS to define the default number of nfs_niothreads instead of hard-coding 4. This change has the advantage that the default can be specified at compile time. If the root filesystem is mounted over NFS we don't have an opportunity to use the syscall to limit the number of threads. Useful on small-memory machines.
1.28	03-Apr-2001	chs	remove a temporary hack now that it's fixed for real. fixes PR 11731.
1.27	02-Apr-2001	fvdl	Set default NFS read and write sizes back to 8k, because a lot of (old) hardware can't handle more.
1.26	25-Mar-2001	matt	Allow the default NFS_RSIZE and NFS_WSIZE to be overriden.
1.25	27-Nov-2000	chs	branches: 1.25.2; Initial integration of the Unified Buffer Cache project.
1.24	19-Sep-2000	fvdl	Bump some defaults and maximums to better values.
1.23	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.22	09-Jun-2000	fvdl	branches: 1.22.2; Some tweaks to enable NFS over IPv6. The special-casing of AF_INET should really be removed.
1.21	15-Apr-2000	tsarna	branches: 1.21.2; Death to nfsiod! It is replaced by kernel threads that do the same thing. The number of kernel threads used is set with the vfs.nfs.iothreads sysctl.
1.20	13-Nov-1998	thorpej	branches: 1.20.10; Clean up the NFS sysctl variables.
1.19	11-Sep-1998	mycroft	Substantial signal handling changes: * Increase the size of sigset_t to accomodate 128 signals -- adding new versions of sys_setprocmask(), sys_sigaction(), sys_sigpending() and sys_sigsuspend() to handle the changed arguments. * Abstract the guts of sys_sigaltstack(), sys_setprocmask(), sys_sigaction(), sys_sigpending() and sys_sigsuspend() into separate functions, and call them from all the emulations rather than hard-coding everything. (Avoids uses the stackgap crap for these system calls.) * Add a new flag (p_checksig) to indicate that a process may have signals pending and userret() needs to do the full (slow) check. * Eliminate SAS_ALTSTACK; it's exactly the inverse of SS_DISABLE. * Correct emulation bugs with restoring SS_ONSTACK. * Make the signal mask in the sigcontext always use the emulated mask format. * Store signals internally in sigaction structures, rather than maintaining a bunch of little sigsets for each SA_* bit. * Keep track of where we put the signal trampoline, rather than figuring it out in _sendsig(). Issue a warning when a non-emulated sigaction bit is observed. * Add missing emulated signals, and a native SIGPWR (currently not used). * Implement the `not reset when caught' semantics for relevant signals. Note: Only code touched by the i386 port has been modified. Other ports and emulations need to be updated.
1.18	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.17	19-Oct-1997	fvdl	* Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.16	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.15	24-Jun-1997	fvdl	branches: 1.15.4; 1.15.6; Add a few defines for WebNFS support.
1.14	12-May-1997	fvdl	Store RPC procnum consistently as an u_int32_t. This is as it should be, and avoid possible server crashes due to bogus comparisons. Partly from BSDI.
1.13	10-Dec-1996	mycroft	Allocate real malloc types for NFS, rather than using M_TEMP.
1.12	02-Dec-1996	thorpej	NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.11	27-May-1996	fvdl	Align things right in NWDELAYHASH (for the Alpha). This fixes crashes in the server code. From John Birell.
1.10	18-Feb-1996	fvdl	branches: 1.10.4; Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.9	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.8	26-Mar-1995	jtc	KERNEL -> _KERNEL
1.7	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.6	17-Aug-1994	mycroft	Convert some more lists and queues.
1.5	17-Aug-1994	mycroft	Change the reply list to a TAILQ.
1.4	29-Jun-1994	cgd	branches: 1.4.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.3	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.2	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.1	20-Apr-1993	mycroft	branches: 1.1.1; Restore files lost during crash.
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.4.2.1	19-Aug-1994	mycroft	update from trunk
1.10.4.2	10-Dec-1996	mycroft	From trunk: Allocate real malloc types for NFS, rather than using M_TEMP.
1.10.4.1	27-May-1996	fvdl	From trunk: Align things right in NWDELAYHASH (for the Alpha). This fixes crashes in the server code. From John Birell.
1.15.6.1	08-Sep-1997	thorpej	Significantly restructure the way signal state for a process is stored. Rather than using bitmasks to redundantly store the information kept in the process's sigacts (because the sigacts was kept in the u-area), hang sigacts directly off the process, and access it directly. Simplify signal setup code tremendously by storing information in the sigacts as an array of struct sigactions, rather than in a different format, since userspace uses sigactions. Make sigacts sharable by adding reference counting.
1.15.4.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.20.10.4	21-Apr-2001	bouyer	Sync with HEAD
1.20.10.3	27-Mar-2001	bouyer	Sync with HEAD.
1.20.10.2	08-Dec-2000	bouyer	Sync with HEAD.
1.20.10.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.21.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.22.2.4	09-Feb-2002	he	Pull up revision 1.32 (requested by he): Widen cr_ref to prevent overflow. (Missed in previous round of commits on this issue.)
1.22.2.3	16-Aug-2001	tv	Pullup [jdolecek]: sys/miscfs/umapfs/umap_vfsops.c 1.29-1.30 sys/kern/vfs_subr.c 1.156 sys/nfs/nfs.h 1.30 Bounds check mount args.
1.22.2.2	06-Apr-2001	he	Pull up revisions 1.26-1.27 (requested by fvdl): Adjust default NFS operation size back to 8KB on all systems except on i386.
1.22.2.1	14-Dec-2000	he	Pull up revision 1.24 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.25.2.6	11-Dec-2002	thorpej	Sync with HEAD.
1.25.2.5	20-Jun-2002	nathanw	Catch up to -current.
1.25.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.25.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.25.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.25.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.29.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.29.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.29.2.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.30.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.39.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.39.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.39.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.39.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.39.2.2	03-Aug-2004	skrll	Sync with HEAD
1.39.2.1	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.
1.44.4.2	27-Oct-2005	riz	Pull up following revision(s) (requested by christos in ticket #5863): sys/nfs/nfs_subs.c: revision 1.152 via patch sys/nfs/nfs.h: revision 1.49 sys/nfs/nfs_vfsops.c: revision 1.149 via patch usr.sbin/amd/include/config.h: revision 1.36 sys/nfs/nfs_vnops.c: revision 1.227 via patch sys/nfs/nfsmount.h: revision 1.34 Allow the attribute cache to be turned off, and allow amd to do it.
1.44.4.1	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.44.2.1	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.48.12.7	27-Feb-2008	yamt	revert incomplete nfs client locking for now.
1.48.12.6	15-Feb-2008	yamt	- sprinkle some locks. - disable MNT_UPDATE because it involves too much locking headache. - don't overwrite other bits in v_vflags when setting VV_ROOT.
1.48.12.5	07-Dec-2007	yamt	sync with head
1.48.12.4	15-Nov-2007	yamt	sync with head.
1.48.12.3	03-Sep-2007	yamt	sync with head.
1.48.12.2	30-Dec-2006	yamt	sync with head.
1.48.12.1	21-Jun-2006	yamt	sync with head.
1.48.10.1	26-Sep-2005	tron	Pull up following revision(s) (requested by christos in ticket #816): sys/nfs/nfs.h: revision 1.49 sys/nfs/nfsmount.h: revision 1.34 Allow turning off the attribute cache.
1.52.2.1	15-Jan-2006	yamt	sync with head.
1.55.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.55.8.3	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.
1.55.8.2	12-Mar-2006	elad	Get rid of NFSW_SAMECRED() that uses memcmp() to compare two credentials, and use a new nfsrv_samecred(), using kauth(9). Note that the NFSW_SAMECRED() macro used to check nd_flag of both descriptors for NB_KERBAUTH too; we don't do that. [documented] Based on code in FreeBSD, thanks to Jeff Roberson.
1.55.8.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.55.6.5	14-Sep-2006	yamt	sync with head.
1.55.6.4	03-Sep-2006	yamt	sync with head.
1.55.6.3	11-Aug-2006	yamt	sync with head
1.55.6.2	26-Jun-2006	yamt	sync with head.
1.55.6.1	24-May-2006	yamt	sync with head.
1.55.4.2	01-Jun-2006	kardel	Sync with head.
1.55.4.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.55.2.1	09-Sep-2006	rpaulo	sync with head
1.56.2.1	19-Jun-2006	chap	Sync with head.
1.57.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.60.2.1	12-Jan-2007	ad	Sync with head.
1.62.2.2	07-May-2007	yamt	sync with head.
1.62.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.63.4.1	11-Jul-2007	mjf	Sync with head.
1.63.2.3	26-Aug-2007	yamt	- mark nfssvc(2) MPSAFE and move the most of nfsd out of the kernel lock. - remove unused ns_solock. - remove some of KERNEL_LOCK/UNLOCK which are not necessary on this branch.
1.63.2.2	09-Jun-2007	ad	Sync with head.
1.63.2.1	08-Jun-2007	ad	Sync with head.
1.67.12.1	13-Nov-2007	bouyer	Sync with HEAD
1.67.8.2	09-Jan-2008	matt	sync with HEAD
1.67.8.1	06-Nov-2007	matt	sync with HEAD
1.67.6.2	09-Dec-2007	jmcneill	Sync with HEAD.
1.67.6.1	29-Oct-2007	joerg	Sync with HEAD.
1.68.4.2	08-Dec-2007	ad	Sync with head.
1.68.4.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.68.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.69.22.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.69.16.3	11-Mar-2010	yamt	sync with head
1.69.16.2	04-May-2009	yamt	sync with head.
1.69.16.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.69.12.1	17-Jan-2009	mjf	Sync with HEAD.
1.71.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.72.20.3	03-Dec-2017	jdolecek	update from HEAD
1.72.20.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.72.20.1	23-Jun-2013	tls	resync from head
1.72.10.1	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")
1.73.10.1	10-Aug-2014	tls	Rebase.
1.73.6.1	18-May-2014	rmind	sync with head
1.74.4.1	06-Jun-2015	skrll	Sync with HEAD
1.75.10.1	08-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #856): sys/nfs/nfs.h: revision 1.76 sys/nfs/nfs_subs.c: revision 1.230 sys/nfs/nfs_socket.c: revision 1.199 sys/nfs/nfs_clntsocket.c: revision 1.6 PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.77.4.1	10-Jun-2019	christos	Sync with HEAD
1.77.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.78.20.1	06-Jun-2021	cjep	sync with head
1.78.16.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.80.10.1	02-Aug-2025	perseant	Sync with HEAD
1.202	13-Feb-2024	andvar	s/Enque/Enqueue/ in comment.
1.201	24-Jun-2022	hannken	Remove an incorrect assertion. Just issue a readahead near the end of the vnode and enqueue an async read. Now let nfs_setattr() truncate the vnode, set its new size and nfs_vinvalbuf() waits for the pages from the readahead to become unbusy. The async read gets processed and returns with uio_resid > 0 because there is a hole and no write after the hole has been pushed yet. As the vnode size already got truncated to the new size the KASSERT() incorrectly fires.
1.200	20-Oct-2021	thorpej	Overhaul of the EVFILT_VNODE kevent(2) filter: - Centralize vnode kevent handling in the VOP_() wrappers, rather than forcing each individual file system to deal with it (except VOP_RENAME(), because VOP_RENAME() is a mess and we currently have 2 different ways of handling it; at least it's reasonably well-centralized in the "new" way). - Add support for NOTE_OPEN, NOTE_CLOSE, NOTE_CLOSE_WRITE, and NOTE_READ, compatible with the same events in FreeBSD. - Track which kevent notifications clients are interested in receiving to avoid doing work for events no one cares about (avoiding, e.g. taking locks and traversing the klist to send a NOTE_WRITE when someone is merely watching for a file to be deleted, for example). In support of the above: - Add support in vnode_if.sh for specifying PRE- and POST-op handlers, to be invoked before and after vop_pre() and vop_post(), respectively. Basic idea from FreeBSD, but implemented differently. - Add support in vnode_if.sh for specifying CONTEXT fields in the vop__args structures. These context fields are used to convey information between the file system VOP function and the VOP wrapper, but do not occupy an argument slot in the VOP_*() call itself. These context fields are initialized and subsequently interpreted by PRE- and POST-op handlers. - Version VOP_REMOVE(), uses the a context field for the file system to report back the resulting link count of the target vnode. Return this in tmpfs, udf, nfs, chfs, ext2fs, lfs, and ufs. NetBSD 9.99.92.
1.199	05-Sep-2020	riastradh	Round of uvm.h cleanup. The poorly named uvm.h is generally supposed to be for uvm-internal users only. - Narrow it to files that actually need it -- mostly files that need to query whether curlwp is the pagedaemon, which should maybe be exposed by an external header. - Use uvm_extern.h where feasible and uvm_*.h for things not exposed by it. We should split up uvm_extern.h but this will serve for now to reduce the uvm.h dependencies. - Use uvm_stat.h and #ifdef UVMHIST uvm.h for files that use UVMHIST(ubchist), since ubchist is declared in uvm.h but the reference evaporates if UVMHIST is not defined, so we reduce header file dependencies. - Make uvm_device.h and uvm_swap.h independently includable while here. ok chs@
1.198	23-May-2020	ad	Move proc_lock into the data segment. It was dynamically allocated because at the time we had mutex_obj_alloc() but not __cacheline_aligned.
1.197	17-May-2020	ad	Start trying to reduce cache misses on vm_page during fault processing. - Make PGO_LOCKED getpages imply PGO_NOBUSY and remove the latter. Mark pages busy only when there's actually I/O to do. - When doing COW on a uvm_object, don't mess with neighbouring pages. In all likelyhood they're already entered. - Don't mess with neighbouring VAs that have existing mappings as replacing those mappings with same can be quite costly. - Don't enqueue pages for neighbour faults unless not enqueued already, and don't activate centre pages unless uvmpdpol says its useful. Also: - Make PGO_LOCKED getpages on UAOs work more like vnodes: do gang lookup in the radix tree, and don't allocate new pages. - Fix many assertion failures around faults/loans with tmpfs.
1.196	23-Apr-2020	ad	PR kern/54759 (vm.ubc_direct deadlock when read()/write() into mapping of itself) - Add new flag UBC_ISMAPPED which tells ubc_uiomove() the object is mmap()ed somewhere. Use it to decide whether to do direct-mapped copy, rather than poking around directly in the vnode in ubc_uiomove(), which is ugly and doesn't work for tmpfs. It would be nicer to contain all this in UVM but the filesystem provides the needed locking here (VV_MAPPED) and to reinvent that would suck more. - Rename UBC_UNMAP_FLAG() to UBC_VNODE_FLAGS(). Pass in UBC_ISMAPPED where appropriate.
1.195	22-Mar-2020	ad	branches: 1.195.2; Process concurrent page faults on individual uvm_objects / vm_amaps in parallel, where the relevant pages are already in-core. Proposed on tech-kern. Temporarily disabled on MP architectures with __HAVE_UNLOCKED_PMAP until adjustments are made to their pmaps.
1.194	23-Feb-2020	ad	UVM locking changes, proposed on tech-kern: - Change the lock on uvm_object, vm_amap and vm_anon to be a RW lock. - Break v_interlock and vmobjlock apart. v_interlock remains a mutex. - Do partial PV list locking in the x86 pmap. Others to follow later.
1.193	15-Jan-2020	ad	Merge from yamt-pagecache (after much testing): - Reduce unnecessary page scan in putpages esp. when an object has a ton of pages cached but only a few of them are dirty. - Reduce the number of pmap operations by tracking page dirtiness more precisely in uvm layer.
1.192	13-Dec-2019	ad	branches: 1.192.2; Break the global uvm_pageqlock into a per-page identity lock and a private lock for use of the pagedaemon policy code. Discussed on tech-kern. PR kern/54209: NetBSD 8 large memory performance extremely low PR kern/54210: NetBSD-8 processes presumably not exiting PR kern/54727: writing a large file causes unreasonable system behaviour
1.191	15-Jul-2015	manu	branches: 1.191.18; Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.190	05-Sep-2014	matt	branches: 1.190.2; Don't use catch as a variable name.
1.189	12-Aug-2013	hannken	branches: 1.189.4; Function nfs_vinvalbuf() ignores errors from vinvalbuf() and therefore delayed write errors may get lost. Change nfs_vinvalbuf() to keep errors from vinvalbuf() for fsync() or close(). Presented on tech-kern@ Fix for PR kern/47980 (NFS over-quota not detected if utimes() called before fsync()/close())
1.188	27-Sep-2011	christos	branches: 1.188.2; 1.188.8; 1.188.12; 1.188.14; 1.188.16; 1.188.22; use NFS_MAXNAMLEN for all names.
1.187	19-Jun-2011	rmind	- Fix a silly bug: remove umap from uobj in ubc_release() UBC_UNMAP case. - Use UBC_WANT_UNMAP() consistently. ARM (PMAP_CACHE_VIVT case) works again.
1.186	12-Jun-2011	rmind	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.
1.185	12-Jun-2010	jakllsch	branches: 1.185.6; Fix memory leak during some NFS writes.
1.184	23-Apr-2010	pooka	Enforce RLIMIT_FSIZE before VOP_WRITE. This adds support to file system drivers where it was missing from and fixes one buggy implementation. The arguably weird semantics of the check are maintained (v_size vs. va_bytes, overwrite).
1.183	14-Mar-2009	dsl	branches: 1.183.2; 1.183.4; Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.182	13-Mar-2009	yamt	nfs_bioread: don't truncate values in a debug printf.
1.181	19-Nov-2008	ad	branches: 1.181.4; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.180	31-Oct-2008	christos	- allocate 8 pointers on the stack to avoid stack overflow in nfs. - make that 8 a constant - remove bogus panic
1.179	17-Oct-2008	christos	branches: 1.179.2; 1.179.4; Requested by yamt: - In getpages don't allocate if we are not locked - Use kmem_alloc instead of malloc and don't sleep Also provide a 64 entry stack array so we don't have to allocate in the common case.
1.178	17-Oct-2008	dogcow	it appears the previous commit's sacrifice was "successful compilation with NFS_V2_ONLY defined".
1.177	16-Oct-2008	christos	Another sacrifice to the stack protector gods.
1.176	16-Oct-2008	christos	don't use variable allocation on the stack.
1.175	24-Apr-2008	ad	branches: 1.175.2; 1.175.8; Network protocol interrupts can now block on locks, so merge the globals proclist_mutex and proclist_lock into a single adaptive mutex (proc_lock). Implications: - Inspecting process state requires thread context, so signals can no longer be sent from a hardware interrupt handler. Signal activity must be deferred to a soft interrupt or kthread. - As the proc state locking is simplified, it's now safe to take exit() and wait() out from under kernel_lock. - The system spends less time at IPL_SCHED, and there is less lock activity.
1.174	29-Mar-2008	yamt	branches: 1.174.2; ansify. from Christoph Egger.
1.173	02-Jan-2008	yamt	branches: 1.173.6; use kmem_alloc instead of malloc.
1.172	02-Jan-2008	ad	Merge vmlocking2 to head.
1.171	04-Dec-2007	yamt	branches: 1.171.4; merge non-intrusive nfs changes from vmlocking.
1.170	26-Nov-2007	pooka	branches: 1.170.2; 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
1.169	28-Oct-2007	yamt	branches: 1.169.2; make NFS_ATTRTIMEO a function.
1.168	10-Oct-2007	ad	branches: 1.168.2; Merge from vmlocking: - Split vnode::v_flag into three fields, depending on field locking. - simple_lock -> kmutex in a few places. - Fix some simple locking problems.
1.167	08-Oct-2007	ad	Merge brelse() changes from the vmlocking branch.
1.166	10-Aug-2007	yamt	branches: 1.166.2; 1.166.4; - instead of scanning an array of iods, maintain a list of idle iods. - make nfs_getset_niothreads MP friendly.
1.165	08-Aug-2007	yamt	push kernel_lock a little.
1.164	29-Jul-2007	ad	branches: 1.164.4; 1.164.6; 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.
1.163	27-Jul-2007	yamt	use ubc_uiomove for read as well.
1.162	27-Jul-2007	yamt	ubc_uiomove: add an "advice" argument rather than using UVM_ADV_RANDOM blindly.
1.161	20-Jul-2007	yamt	- fix decreasing of vfs.nfs.iothreads after the recent partial merge of vmlocking. - don't make nfsiod exit with requests left. - make NFSSVC_BIOD a dummy so that nfsiod can be simplified.
1.160	17-Jul-2007	yamt	branches: 1.160.2; remove (void)0; nonsense.
1.159	17-Jul-2007	yamt	fix a typo in a comment.
1.158	12-Jul-2007	rmind	nfs_asyncio: fix the locking in error case, problem was introduced in 1.153 revision, where ltsleep() was replaced with condvar. Problem found and fix provided by David A. Holland, PR/36610. Actually, relock is not needed here, and mutex would be unlocked only on nfs_sigintr() fail case.
1.157	09-Jul-2007	ad	Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.156	12-Jun-2007	yamt	nfs_write: - IO_SYNC: don't bother to flush dirty pages before copying data from user buffer. - IO_APPEND: don't invalidate pages blindly. PR/28472 from Brian Marcotte.
1.155	05-Jun-2007	yamt	improve post-ubc file overwrite performance in common cases. ie. when it's safe, actually overwrite blocks rather than doing read-modify-write. also fixes PR/33152 and PR/36303.
1.154	09-May-2007	yamt	nfs_write: report an error correctly in the case of IO_SYNC.
1.153	29-Apr-2007	yamt	use mutex and condver.
1.152	19-Apr-2007	yamt	hold proclist_mutex when calling psignal().
1.151	04-Mar-2007	christos	branches: 1.151.2; 1.151.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.150	27-Feb-2007	yamt	nfs_getpages: fix an inverted condition in rev.1.147.
1.149	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.148	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.147	15-Feb-2007	yamt	branches: 1.147.2; use mutex and rwlock rather than lockmgr.
1.146	27-Dec-2006	yamt	remove nqnfs.
1.145	23-Jul-2006	ad	branches: 1.145.4; Use the LWP cached credentials where sane.
1.144	30-Jun-2006	yamt	fix handling of NFSERR_NOTSUPP and NFSERR_BAD_COOKIE, which have been broken since nfs_socket.c rev.1.115.
1.143	14-May-2006	elad	branches: 1.143.4; integrate kauth.
1.142	01-Mar-2006	yamt	branches: 1.142.2; 1.142.4; 1.142.6; 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.
1.141	14-Jan-2006	yamt	branches: 1.141.2; 1.141.4; nfs_doio_read: clear uio_resid when filling a hole.
1.140	13-Dec-2005	reinoud	branches: 1.140.2; Fix of panic that was introduced since ktrace-lwp branch was merged. The shortcut to the process of the passed lwp paniced the kernel since lwp could/can be passwd as NULL in VOP_WRITE(). This was happening when ktracing to NFS. The function ktrwrite() set the uio_lwp to NULL and then calls VOP_WRITE() with this argument. nfs_write() then accessed lwp *l->l_proc wich paniced. Thanks to David Laight for his help on tracking it down.
1.139	11-Dec-2005	christos	merge ktrace-lwp.
1.138	29-Nov-2005	yamt	merge yamt-readahead branch.
1.137	04-Nov-2005	yamt	branches: 1.137.2; nfs_bioread: push delayed truncation and tweak loop accordingly. PR/31926 from Jed Davis.
1.136	06-Oct-2005	yamt	nfs_bioread: handle file truncation on the server a little more gracefully.
1.135	01-Oct-2005	jdolecek	use killproc() for killing the process due to text file modification, so that it's logged too PR: 17392 by Greg A. Woods
1.134	19-Aug-2005	yamt	fix some simple bugs in the 64bit ino_t changes. - edp -> dp - * -> +
1.133	19-Aug-2005	christos	64 bit inode changes.
1.132	21-Jul-2005	yamt	use a correct credential for readlink. discussed on source-changes@.
1.131	21-Jul-2005	yamt	nfs_doio_read: revert readlink part of 1.129 and 1.130 because they were wrong.
1.130	07-Jul-2005	christos	Back to using curproc in the VLNK case when uiop->uio_procp == NULL, and explain why we need to.
1.129	07-Jul-2005	christos	1. use p = uio->uio_procp consistently and eliminate suspicious uses of curproc (where uio->uio_procp should be used?). Don't do this for nfs_commit(), because yamt says it is possibly wrong. 2. nfs_doio() does not use struct proc; remove it and the code to compute it. 3. use copyin_proc() and copyout_proc() instead of copyin() and copyout(). 4. check return of copyout_proc(). and mark return from copyin_proc() XXX 5. Eliminate check p == curproc assertion check from nfs_write; nfs_read does not have it and we might be called in a different process context anyway (PR 20138).
1.128	26-Feb-2005	perry	branches: 1.128.2; 1.128.4; nuke trailing whitespace
1.127	27-Jan-2005	yamt	- simplify nfs_bio.c rev.1.126 - add an assertion. no functional changes.
1.126	27-Jan-2005	yamt	nfs_bioread: - if a buffer is still empty after successful nfs_doio, it implies EOF. - don't cache blocks beyond EOF.
1.125	26-Jan-2005	yamt	handle a really empty directory, which doesn't have even the dot entry.
1.124	09-Jan-2005	chs	branches: 1.124.2; 1.124.4; adjust the UBC mapping code to support non-vnode uvm_objects. this means we can no longer look at the vnode size to determine how many pages to request in a fault, which is good since for NFS the size can change out from under us on the server anyway. there's also a new flag UBC_UNMAP for ubc_release(), so that the file system code can make the decision about whether to cache mappings for files being used as executables.
1.123	14-Dec-2004	yamt	- centerize code to invalidate stale cache. - don't ignore errors when invalidating buffers in nfs_open.
1.122	26-Oct-2004	yamt	since daddr_t is 64-bit these days, simply use nfs directory cookies as buffer cache indexes. regress/sys/fs/getdents is now supposed to work. fix PR/27112.
1.121	17-Sep-2004	skrll	There's no need to pass a proc value when using UIO_SYSSPACE with vn_rdwr(9) and uiomove(9). OK'd by Jason Thorpe
1.120	15-Sep-2004	yamt	fix access-after-free bugs in dircache code by refcounting nfsdircache. PR/26864.
1.119	18-Jul-2004	yamt	nfs_doio_read: on short read, zero out the rest of the buffer unconditionally. we can't rely on n_size here because it can be changed under us.
1.118	11-Jun-2004	yamt	nfs_doio_read: use np->n_rcred instead of curproc->p_ucred for VDIR. XXX maybe it's better to use a cred passed by VOP_READDIR.
1.117	23-May-2004	christos	cut down another 7K by more NFS_V2_ONLY ifdefs.
1.116	12-Mar-2004	yamt	branches: 1.116.2; introduce a macro NFS_INVALIDATE_ATTRCACHE and use it instead of "n_attrstamp = 0".
1.115	10-Jan-2004	yamt	comments in nfs_doio_write.
1.114	07-Dec-2003	fvdl	Unix semantics dictate that access checks for files are done when it is opened. An open file can always be read from and/or written to, depending on how it was opened. Therefore, the read/write/commit RPCs should never return EACCESS, as they are only performed on files that have been successfully opened already. This change improves the current situation and works in most cases. It simply always uses the most recently known owner/group of the file, iff the authentication mechanism is AUTH_UNIX (in other cases, the creds for a succesful open are used, but note that no other cases are currently implemented). A retry mechanism can be used to catch a few more cases, but this is a good improvement for now.
1.113	17-Nov-2003	jonathan	Fix hanging-paren typo.
1.112	17-Nov-2003	jonathan	Change previous patch to have same effect as patch posted to tech-kern. Suggested reformatting inadvertently changed the meaning of the code, as noted by YAMAMOTO Takashi <yamt@mwd.biglobe.ne.jp>.
1.111	17-Nov-2003	jonathan	Commit fix for NFS write deadlock, on filesystems mounted via local-loopback (lo0). As posted for review on tech-kern 2003-18-09, with a long comment explaining (one of) the deadlock scenarios. I've used this since shortly after 2002-09-12-, without noticing performance degradataion or instability for non-loopback mounts.
1.110	26-Sep-2003	yamt	change n_mtime from time_t to timespec in order to improve cache consistency. (1 second granularity is too loose these days.)
1.109	17-Sep-2003	yamt	don't call nfs_delayedtruncate() from nfs_getpages(). it causes simplelock deadlock.
1.108	26-Aug-2003	pk	VOP_PUTPAGES() must be called with the vnode's interlock held.
1.107	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.106	03-Aug-2003	pk	Make life slightly easier for the compiler's optimisation routines.
1.105	29-Jun-2003	fvdl	branches: 1.105.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.104	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
1.103	22-May-2003	yamt	interlock for NFLUSHINPROG/NFLUSHWANT.
1.102	21-May-2003	yamt	eliminate memcpy in the common and easy case of write.
1.101	16-May-2003	yamt	correct a KASSERT.
1.100	15-May-2003	yamt	acquire vmobjlock when touch pg->flags.
1.99	07-May-2003	yamt	simple lock for nfs iod.
1.98	03-May-2003	yamt	- check page's offset in the object as well. (pointed by Chuck Silvers.) - remove false assertion.
1.97	03-May-2003	yamt	- if writerpc ends with a stable result, no need to commit them anymore. - add comments.
1.96	03-May-2003	yamt	better handling of write verifier change.
1.95	18-Apr-2003	yamt	fix a use of an uninitialized variable.
1.94	15-Apr-2003	yamt	remove line-wrapping that is no longer needed.
1.93	12-Apr-2003	yamt	fix a typo in the previous.
1.92	12-Apr-2003	yamt	set b_resid correctly.
1.91	12-Apr-2003	yamt	split nfs_doio to nfs_doio_{phys,read,write} to avoid too deep indents.
1.90	12-Apr-2003	yamt	- do FILESYNC writes if we're freeing the page or the page doesn't belong to us. otherwise, data will be lost on server crash. - use b_bcount instead of b_bufsize to determine how many pages we should deal with. based on a patch from Chuck Silvers. discussed on tech-kern.
1.89	09-Apr-2003	yamt	rename a very confusing variable name. (must_commit -> stalewriteverf)
1.88	09-Apr-2003	yamt	when commit failed and fall to write, re-set 'off' and 'cnt' because it can be changed in 'needcommit' path.
1.87	09-Apr-2003	yamt	make per-iod datas together.
1.86	18-Jan-2003	thorpej	Merge the nathanw_sa branch.
1.85	29-Oct-2002	yamt	fix panic (assertion failure) on error case. if uiomove is failed, we should clean up pages past eof. the problem reported by kay. ok'ed by Chuck Silvers.
1.84	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
1.83	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.82	01-Sep-2002	bouyer	nfs_doio(): handle the case where nfs_writerpc() returned error != 0. Fix kern/18125. OK'd by thorpej and chs.
1.81	06-May-2002	enami	branches: 1.81.4; Remove wrong assertion in previous commit.
1.80	06-May-2002	enami	The per nfsnode n_commitlock is a sleep lock, but we can't sleep if PGO_LOCKED getpages request. So, just make the lock fail and tell the caller that there is no pages available if we can't acquire it. The caller will call us again soon without PGO_LOCKED. Reviewed by chuq.
1.79	10-Apr-2002	chs	only use UBC_FAULTBUSY to access offsets past EOF, otherwise we can deadlock trying to busy the same page in uiomove().
1.78	25-Mar-2002	chs	remove PGO_WEAK, it isn't needed anymore.
1.77	23-Mar-2002	chs	only do v3 stuff for v3 filesystems.
1.76	16-Mar-2002	chs	make sure that if NMODIFIED is clear, all pages attached to the vnode are clean and without writable mappings. if we try to flush dirty pages past EOF to the server when NMODIFIED is clear, we'll update the attrcache before doing the write, which will try to free the pages past EOF and deadlock. to deal with this, we write-protect pages before we send them to the server, and restrict ourselves to creating read-only mappings if NMODIFIED isn't set. score another one for enami.
1.75	31-Jan-2002	chs	use curproc instead of b_proc for NFS. that's what we want for sync commits and it doesn't cause any problems for async commits.
1.74	26-Jan-2002	chs	re-enable NFSv3 commit RPCs by abandoning my new approach in favor of frank's scheme, with one new twist: don't wait until we've totally run out of free pages before committing, but instead notice when we've built up a largish range of uncommitted pages and commit only the older half of the range, which is likely to already be on disk on the server.
1.73	31-Dec-2001	chs	fix locking in nfs_getpages().
1.72	30-Nov-2001	chs	call VOP_PUTPAGES() directly instead of indirecting through the UVM pager op vector.
1.71	10-Nov-2001	lukem	add RCSIDs
1.70	13-Oct-2001	simonb	branches: 1.70.2; Remove so variables that are only ever set and never referenced.
1.69	15-Sep-2001	chs	a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
1.68	27-Jun-2001	thorpej	branches: 1.68.2; 1.68.4; Make sure to add NFS vnodes to the syncerd worklist.
1.67	26-May-2001	chs	replace vm_page_t with struct vm_page *.
1.66	16-Apr-2001	chs	reads at or after EOF should "succeed".
1.65	03-Apr-2001	chs	handle partially full directory buffers by only using (b_bcount - b_resid) bytes of data from the buffer.
1.64	10-Mar-2001	chs	eliminate the VM_PAGER_* error codes in favor of the traditional E* codes. the mapping is: VM_PAGER_OK 0 VM_PAGER_BAD <unused> VM_PAGER_FAIL <unused> VM_PAGER_PEND 0 (see below) VM_PAGER_ERROR EIO VM_PAGER_AGAIN EAGAIN VM_PAGER_UNLOCK EBUSY VM_PAGER_REFAULT ERESTART for async i/o requests, it used to be possible for the request to be convert to sync, and the pager would return VM_PAGER_OK or VM_PAGER_PEND to indicate whether the caller should perform post-i/o cleanup. this is no longer allowed; pagers must now return 0 to indicate that the async i/o was successfully started, and the caller never needs to worry about doing the post-i/o cleanup.
1.63	27-Feb-2001	chs	branches: 1.63.2; min() -> MIN(), max() -> MAX(). fixes more problems with file offsets > 4GB.
1.62	18-Feb-2001	chs	fix a couple more bugs: - in nfs_getpages(), unbusy any pages that we don't free in the error path. - in nfs_putpages(), only call biowait() if we actually started any i/os.
1.61	05-Feb-2001	chs	fix several bugs: - in the cases where we skip over the i/o loop, increment npages by ridx so that when the cleanup code starts processing the pgs array at index 0 it'll actually process all of the pages. - process the PG_RELEASED flag when unbusying pages. - add some missing MP locking. - use MIN() and MAX() instead of min() and max() since the latter are functions which take arguments of type "int" but we call them with values of type "off_t", so the values could be truncated.
1.60	30-Jan-2001	thorpej	Make sure bp->b_proc is initialized. Should fix a deref-garbage-pointer problem reported by msaitoh@netbsd.org. NOTE: These are marked XXXUBC since the code that allocates the bufs is new with UBC, but it may be the case that bp->b_proc needs to be intialized to curproc (it's used in a call to nfs_sigintr()).
1.59	07-Jan-2001	enami	Use uvm_aio_biodone instead of uvm_aio_aiodone for top-level buf so that uvmexp.paging is updated if this i/o was initiated by the pagedaemon.
1.58	27-Dec-2000	chs	fix several bugs: - fix math when skipping writing pages that just need a commit. - clear the needcommit stuff and PG_RDONLY flags on pages returned for overwrite requests as well as for normal write faults. - bail out of nfs_write() if we get an error. - remove a bogus attempt to clean up after failed uiomove()s. - bring over a workaround for a lock-ordering problem from the genfs code. - add some missing MP locking.
1.57	13-Dec-2000	jdolecek	<sys/trace.h> is not needed here
1.56	09-Dec-2000	chs	only zero the part of the page after EOF if we're actually initializing the page.
1.55	04-Dec-2000	fvdl	Initialize 'error' to 0, so that nfs_putpages doesn't return garbage when pages already have been committed and nothing needs to be done.
1.54	27-Nov-2000	chs	Initial integration of the Unified Buffer Cache project.
1.53	19-Sep-2000	bjh21	Extend NFS_V2_ONLY to remove NQNFS lease support as well. Saves another 10k.
1.52	19-Sep-2000	fvdl	Move handling of B_NEEDCOMMIT buffers to nfs_doio, so that bawrite() calls for them are actually done asynchronously. Idea taken from FreeBSD. Do away with nfs_writebp completely, it's not needed anymore. Keep an eye on the range of a file that needs to be committed, and do it in heaps.
1.51	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.50	27-Jun-2000	mrg	remove include of <vm/vm.h>
1.49	18-May-2000	pk	branches: 1.49.4; Fix printf() format.
1.48	30-Mar-2000	augustss	Remove register declarations.
1.47	23-Nov-1999	fvdl	Be more careful to block bio interrupts for some data structures. There were at least a few missed cases where vp->v_{clean,dirty}blkhd were unprotected since the softdep/trickle sync merge.
1.46	15-Nov-1999	fvdl	Add Kirk McKusick's soft updates code to the trunk. Not enabled by default, as the copyright on the main file (ffs_softdep.c) is such that is has been put into gnusrc. options SOFTDEP will pull this in. This code also contains the trickle syncer. Bump version number to 1.4O
1.45	24-Mar-1999	mrg	branches: 1.45.4; 1.45.8; 1.45.10; 1.45.14; completely remove Mach VM support. all that is left is the all the header files as UVM still uses (most of) these.
1.44	09-Aug-1998	perry	branches: 1.44.2; bzero->memset, bcopy->memcpy, bcmp->memcmp
1.43	21-Jun-1998	fvdl	Fix possible overflow problem in read size computation.
1.42	10-Feb-1998	mrg	- add defopt's for UVM, UVMHIST and PMAP_NEW. - remove unnecessary UVMHIST_DECL's.
1.41	05-Feb-1998	mrg	initial import of the new virtual memory system, UVM, into -current. UVM was written by chuck cranor <chuck@maria.wustl.edu>, with some minor portions derived from the old Mach code. i provided some help getting swap and paging working, and other bug fixes/ideas. chuck silvers <chuq@chuq.com> also provided some other fixes. this is the rest of the MI portion changes. this will be KNF'd shortly. :-)
1.40	23-Nov-1997	fvdl	Move the EOF check after getting a block out of the if() that determines whether we get it off the wire. An nfsiod might have been busy with it, and finished while we were waiting for it in nfs_getcacheblk, so we need to check for EOF again no matter what.
1.39	23-Oct-1997	fvdl	Oops. Fix goof in previous change.
1.38	22-Oct-1997	fvdl	Just return immediately in nfs_bioread if we got an empty buffer because of EOF on a directory.
1.37	20-Oct-1997	thorpej	branches: 1.37.2; Fix alignment problems. From Frank van der Linden <fvdl@NetBSD.ORG>.
1.36	19-Oct-1997	fvdl	Only do readaheads when reading sequential blocks; check v_lastr to achieve this. Improves performance for demand paging. From Chris Demetriou.
1.35	19-Oct-1997	fvdl	* Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.34	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.33	17-Jul-1997	fvdl	branches: 1.33.2; * Deal with servers that don't give complete FSINFO (like NT) From Olaf Seibert <rhialto@polder.ubc.kun.nl> (PR 3687) * Make an attempt to check the maximum filesize before attempting a write to the server, as write RPCs will typically happen asynchronously, and the process will not see the error. Fixes problems with unexpectly truncated files at 4G * Pass up errors in nfs_writerpc correctly
1.32	04-Jul-1997	drochner	Don't cast 64bit (off_t) file sizes to vm_offset_t (32bit on many architectures), truncate them intelligently instead. The truncation is done centralized in vnode_pager.c. This prevents from wrap-over effects when parts of large (>2^32 byte) files are mmapped. Don't allow to mmap above the numerical range of vm_offset_t. This is considered a temporary solution until the vm system handles the object sizes/offsets more cleanly.
1.31	20-Apr-1997	fvdl	Only wake up one nfsiod when there is an async write to do. (from FreeBSD).
1.30	02-Dec-1996	thorpej	NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.29	13-Oct-1996	christos	revert kprintf changes
1.28	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.27	02-Jul-1996	fvdl	Don't mistake a non-async block that needs to be commited for an interrupted write.
1.26	23-May-1996	fvdl	* Make mounts with symlinks work (needed for direct mounts with amd). PR #1917 * Never change the NQNFS flag and/or version when just doing an update mount. Fixes a problem that made diskless booting impossible under some circumstances.
1.25	29-Feb-1996	fvdl	branches: 1.25.4; Make sure to clear B_NEEDCOMMIT in the right spot. Fix 'officially blessed' by Rick Macklem. Fixes PR kern/2128.
1.24	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.23	09-Feb-1996	christos	nfs prototype changes
1.22	01-Feb-1996	jtc	Rename struct timespec fields to conform to POSIX.1b
1.21	24-Jul-1995	cgd	avoid unnecessary aging of buffers. This used to make sense, when buffer caches were much smaller, but makes little sense now, and will become more useless as RAM (and buffer cache) sizes grow. Suggested by Bob Baron.
1.20	18-Mar-1995	gwr	Make call to nfs_writerpc() consistent with others.
1.19	12-Jan-1995	mycroft	Add two missing brelse() calls. From Rick Macklem.
1.18	10-Jan-1995	mycroft	Make sure readdir requests are only truncated on block boundaries.
1.17	20-Jul-1994	mycroft	Fix a problem with write-behind causing processes to be killed occasionally. From Rick Macklem.
1.16	12-Jul-1994	cgd	minor cache consistency fix
1.15	29-Jun-1994	cgd	branches: 1.15.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.14	22-Jun-1994	pk	straighten out diskless swap code somewhat.
1.13	15-Jun-1994	mycroft	Turn P_NOSWAP and P_PHYSIO into a hold count, as suggested by a comment.
1.12	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.11	24-May-1994	cgd	MIN -> min, MAX -> max
1.10	25-Apr-1994	cgd	some prototype cleanup, eliminate/replace bogus types (e.g. quad and u_quad) -> use better types (e.g. quad_t & u_quad_t in inodes), some cleanup.
1.9	21-Apr-1994	cgd	Convert mount, vnode, and buf structs to use <sys/queue.h>. Also, some knf and structure frobbing to do along with it.
1.8	18-Dec-1993	mycroft	Canonicalize all #includes.
1.7	03-Sep-1993	jtc	branches: 1.7.2; Include systm.h to get prototypes (and possibly inlines) of *max functions.
1.6	13-Jul-1993	cgd	get rid of some more bogus changes from a week ago
1.5	13-Jul-1993	cgd	diskless changes made last time were hosed; were using NULL for "no credentials" rather than NOCRED.
1.4	07-Jul-1993	cgd	changes from ws to support diskless booting... these are "OK" on inspection and after testing... (actually, currently, none of the changed code is even used...)
1.3	30-Jun-1993	andrew	Paul Kranenburg's VM deadlock fixes. (patchkit patch 00147, part 2)
1.2	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.7.2.1	14-Nov-1993	mycroft	Canonicalize all #includes.
1.15.2.2	20-Jul-1994	cgd	from trunk, per mycroft
1.15.2.1	12-Jul-1994	cgd	consistency fix, from trunk
1.25.4.2	08-Jul-1996	jtc	Pulled up from rev 1.27 by request from Frank van der Linden
1.25.4.1	25-May-1996	fvdl	Pull in bugfixes from main branch.
1.33.2.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.37.2.3	24-Nov-1997	mellon	Pull rev 1.40 up from trunk (fvdl)
1.37.2.2	23-Oct-1997	mellon	Pull rev 1.39 up from trunk
1.37.2.1	23-Oct-1997	mellon	Pull rev 1.38 from main trunk
1.44.2.5	30-May-1999	chs	vm_page's blkno is gone.
1.44.2.4	30-Apr-1999	chs	change ubc_alloc()'s length arg to be a pointer instead of the value. the pointed-to value is the total desired length on input, and is updated to the length that will fit in the returned window. this allows callers of ubc_alloc() to be ignorant of the window size.
1.44.2.3	25-Feb-1999	chs	major overhaul of getpages and putpages functions.
1.44.2.2	16-Nov-1998	chs	set NMODIFIED in nfs_write(). putpage is now called with uobj unlocked. remove some debugging printfs.
1.44.2.1	09-Nov-1998	chs	initial snapshot. lots left to do.
1.45.14.2	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.45.14.1	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".
1.45.10.1	19-Oct-1999	fvdl	Bring in Kirk McKusick's FFS softdep code on a branch.
1.45.8.8	21-Apr-2001	bouyer	Sync with HEAD
1.45.8.7	12-Mar-2001	bouyer	Sync with HEAD.
1.45.8.6	11-Feb-2001	bouyer	Sync with HEAD.
1.45.8.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.45.8.4	05-Jan-2001	bouyer	Sync with HEAD
1.45.8.3	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.45.8.2	08-Dec-2000	bouyer	Sync with HEAD.
1.45.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.45.4.5	31-Aug-1999	perseant	Rudimentary support for LFS under UBC: - LFS-specific VOP_BALLOC and VOP_PUTPAGES vnode ops. - getblk VREG panic #ifdef'd out (can be reinstated when Ifile is internalized and Ifile can be made another type from VREG) - interface to VOP_PUTPAGES changed to pass all pager flags, not just sync. FS putpages routines must know about the pager flags. - new LFS magic disk address, -2 ("unwritten"), meaning accounted for but not assigned to a fixed disk location (since LFS does these two things separately, and the previous accounting method using buffer headers no longer will work). Changed references to (foo == (daddr_t)-1) to (foo < 0). Since disk drivers reject all addresses < 0, this should not present a problem for other FSs.
1.45.4.4	31-Jul-1999	chs	in nfs_getpages(), deal with extending writes better. also, return errnos instead of VM_PAGER_*.
1.45.4.3	11-Jul-1999	chs	remove uvm_vnp_uncache(), it's no longer needed.
1.45.4.2	04-Jul-1999	chs	update uvm_pagermapin() to match new args.
1.45.4.1	07-Jun-1999	chs	merge everything from chs-ubc branch.
1.49.4.1	14-Dec-2000	he	Pull up revision 1.52 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.63.2.19	11-Nov-2002	nathanw	Catch up to -current
1.63.2.18	22-Oct-2002	thorpej	Sync with HEAD.
1.63.2.17	17-Sep-2002	nathanw	Catch up to -current.
1.63.2.16	15-Jul-2002	nathanw	Whitespace.
1.63.2.15	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.63.2.14	24-Jun-2002	nathanw	Curproc->curlwp renaming. Change uses of "curproc->l_proc" back to "curproc", which is more like the original use. Bare uses of "curproc" are now "curlwp". "curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL) so that it is always safe to reference curproc (dereferencing curproc is another story, but that's always been true).
1.63.2.13	20-Jun-2002	nathanw	Catch up to -current.
1.63.2.12	17-Apr-2002	nathanw	Catch up to -current.
1.63.2.11	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.63.2.10	28-Feb-2002	nathanw	curproc ==> curproc->l_proc
1.63.2.9	28-Feb-2002	nathanw	Catch up to -current.
1.63.2.8	08-Jan-2002	nathanw	Catch up to -current.
1.63.2.7	14-Nov-2001	nathanw	Catch up to -current.
1.63.2.6	22-Oct-2001	nathanw	Catch up to -current.
1.63.2.5	21-Sep-2001	nathanw	Catch up to -current.
1.63.2.4	24-Aug-2001	nathanw	Catch up with -current.
1.63.2.3	21-Jun-2001	nathanw	Catch up to -current.
1.63.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.63.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.68.4.1	01-Oct-2001	fvdl	Catch up with -current.
1.68.2.5	30-Sep-2002	jdolecek	add support for kevents to NFS to detect file changes on server by other NFS clients, polling kernel thread is used to periodically check for attribute changes of watched files; the NFS server is only contacted when the vnode expires from local attrcache (which takes 5-60 seconds currently), to keep network&CPU overhead low the routine checking for remote changes is quite simplistic, but hopefully doing it's job well enough
1.68.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.68.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.68.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.68.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.70.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.81.4.1	01-Sep-2002	lukem	Pull up revision 1.82 (requested by bouyer in ticket #752): nfs_doio(): handle the case where nfs_writerpc() returned error != 0. Fix kern/18125. OK'd by thorpej and chs.
1.105.2.12	11-Dec-2005	christos	Sync with head.
1.105.2.11	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.105.2.10	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.105.2.9	04-Feb-2005	skrll	Sync with HEAD.
1.105.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.105.2.7	18-Dec-2004	skrll	Sync with HEAD.
1.105.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.105.2.5	30-Oct-2004	skrll	Correct panic message s/proc/lwp/
1.105.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.105.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.105.2.2	03-Aug-2004	skrll	Sync with HEAD
1.105.2.1	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.
1.116.2.3	01-Dec-2005	riz	Pull up following revision(s) (requested by jld in ticket #8826): sys/nfs/nfs_bio.c: revisions 1.136-1.137 via patch The problem (kern/31926): under certain conditions, which could be reliably reproduced, NFS reads would occasionally return zeroes instead of some of the file data, or fail with EINVAL.
1.116.2.2	18-Sep-2004	he	branches: 1.116.2.2.2; Pull up revision 1.120 (requested by yamt in ticket #858): Fix access-after-free bugs in dircache code by reference counting nfsdircache. Fixes PR#26864.
1.116.2.1	21-Jun-2004	tron	Pull up revision 1.118 (requested by yamt in ticket #513): nfs_doio_read: use np->n_rcred instead of curproc->p_ucred for VDIR. XXX maybe it's better to use a cred passed by VOP_READDIR.
1.116.2.2.2.2	01-Dec-2005	riz	Pull up following revision(s) (requested by jld in ticket #8826): sys/nfs/nfs_bio.c: revisions 1.136-1.137 via patch The problem (kern/31926): under certain conditions, which could be reliably reproduced, NFS reads would occasionally return zeroes instead of some of the file data, or fail with EINVAL.
1.116.2.2.2.1	30-Jan-2005	he	branches: 1.116.2.2.2.1.2; Pull up revision 1.122 (requested by yamt in ticket #968): Since daddr_t is 64-bit these days, simply use nfs directory cookies as buffer cache indexes. This should make the regress/sys/fs/getdents test work. Fixes PR#27112.
1.116.2.2.2.1.2.1	01-Dec-2005	riz	Pull up following revision(s) (requested by jld in ticket #8826): sys/nfs/nfs_bio.c: revisions 1.136-1.137 via patch The problem (kern/31926): under certain conditions, which could be reliably reproduced, NFS reads would occasionally return zeroes instead of some of the file data, or fail with EINVAL.
1.124.4.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.124.4.1	12-Feb-2005	yamt	sync with head.
1.124.2.1	29-Apr-2005	kent	sync with -current
1.128.4.8	21-Jan-2008	yamt	sync with head
1.128.4.7	07-Dec-2007	yamt	sync with head
1.128.4.6	15-Nov-2007	yamt	sync with head.
1.128.4.5	27-Oct-2007	yamt	sync with head.
1.128.4.4	03-Sep-2007	yamt	sync with head.
1.128.4.3	26-Feb-2007	yamt	sync with head.
1.128.4.2	30-Dec-2006	yamt	sync with head.
1.128.4.1	21-Jun-2006	yamt	sync with head.
1.128.2.2	21-Nov-2005	tron	Pull up following revision(s) (requested by yamt in ticket #980): sys/nfs/nfs_bio.c: revision 1.137 nfs_bioread: push delayed truncation and tweak loop accordingly. PR/31926 from Jed Davis.
1.128.2.1	21-Nov-2005	tron	Pull up following revision(s) (requested by yamt in ticket #980): sys/nfs/nfs_bio.c: revision 1.136 nfs_bioread: handle file truncation on the server a little more gracefully.
1.137.2.3	19-Nov-2005	yamt	- as read-ahead context is per-vnode now, there are less reasons to make VOP_READ call uvm_ra_request explicitly. move it to pager (uvn_get) so that it can handle accesses via mmap as well. - pass advice to pager via ubc. - tweak DPRINTF. XXX can be disturbed by PGO_LOCKED. XXX it's controversial where it should be done. (uvm_fault, uvn_get or genfs_getpages.)
1.137.2.2	18-Nov-2005	yamt	- associate read-ahead context to vnode, rather than file. - revert VOP_READ prototype.
1.137.2.1	15-Nov-2005	yamt	adapt ffs, lfs, nfs.
1.140.2.2	15-Jan-2006	yamt	sync with head.
1.140.2.1	31-Dec-2005	yamt	- adapt nfs. - nfs_doio_read: #if 0 out "killproc if text is modified" part of the code as it's broken. (a process reading the modified text is not necessarily a process which is using the file as a text.)
1.141.4.2	01-Jun-2006	kardel	Sync with head.
1.141.4.1	22-Apr-2006	simonb	Sync with head.
1.141.2.1	09-Sep-2006	rpaulo	sync with head
1.142.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.142.4.2	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.
1.142.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.142.2.2	11-Aug-2006	yamt	sync with head
1.142.2.1	24-May-2006	yamt	sync with head.
1.143.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.145.4.1	12-Jan-2007	ad	Sync with head.
1.147.2.4	17-May-2007	yamt	sync with head.
1.147.2.3	07-May-2007	yamt	sync with head.
1.147.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.147.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.151.4.1	11-Jul-2007	mjf	Sync with head.
1.151.2.12	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.
1.151.2.11	20-Aug-2007	ad	Sync with HEAD.
1.151.2.10	19-Aug-2007	ad	- Back out the biodone() changes. - Eliminate B_ERROR (from HEAD).
1.151.2.9	15-Jul-2007	ad	Sync with head.
1.151.2.8	18-Jun-2007	yamt	fix merge botches.
1.151.2.7	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.
1.151.2.6	09-Jun-2007	ad	Sync with head.
1.151.2.5	08-Jun-2007	ad	Sync with head.
1.151.2.4	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.
1.151.2.3	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.
1.151.2.2	21-Mar-2007	ad	- Replace more simple_locks, and fix up in a few places. - Use condition variables. - LOCK_ASSERT -> KASSERT.
1.151.2.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.160.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.164.6.2	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.
1.164.6.1	29-Jul-2007	ad	file nfs_bio.c was added on branch matt-mips64 on 2007-07-29 13:31:13 +0000
1.164.4.6	09-Dec-2007	jmcneill	Sync with HEAD.
1.164.4.5	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.164.4.4	29-Oct-2007	joerg	Sync with HEAD.
1.164.4.3	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.
1.164.4.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.164.4.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.166.4.1	14-Oct-2007	yamt	sync with head.
1.166.2.2	09-Jan-2008	matt	sync with HEAD
1.166.2.1	06-Nov-2007	matt	sync with HEAD
1.168.2.1	13-Nov-2007	bouyer	Sync with HEAD
1.169.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.169.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.170.2.2	08-Dec-2007	ad	Sync with head.
1.170.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.171.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.173.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.173.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.173.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.174.2.1	18-May-2008	yamt	sync with head.
1.175.8.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.175.8.1	19-Oct-2008	haad	Sync with HEAD.
1.175.2.4	10-Oct-2010	yamt	some locking changes
1.175.2.3	26-Sep-2010	yamt	locking changes
1.175.2.2	11-Aug-2010	yamt	sync with head.
1.175.2.1	04-May-2009	yamt	sync with head.
1.179.4.2	16-Jul-2010	riz	Pull up following revision(s) (requested by jakllsch in ticket #1417): sys/nfs/nfs_bio.c: revision 1.185 Fix memory leak during some NFS writes.
1.179.4.1	02-Nov-2008	snj	branches: 1.179.4.1.4; Pull up following revision(s) (requested by tron in ticket #9): sys/nfs/nfs_bio.c: revision 1.180 sys/miscfs/genfs/genfs_io.c: revision 1.14 sys/uvm/uvm_extern.h: revision 1.149 - allocate 8 pointers on the stack to avoid stack overflow in nfs. - make that 8 a constant - remove bogus panic
1.179.4.1.4.1	20-May-2011	matt	bring matt-nb5-mips64 up to date with netbsd-5-1-RELEASE (except compat).
1.179.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.179.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.181.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.183.4.3	03-Jul-2010	rmind	sync with head
1.183.4.2	30-May-2010	rmind	sync with head
1.183.4.1	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.
1.183.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.183.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.185.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.188.22.1	07-Sep-2013	bouyer	Pull up following revision(s) (requested by hannken in ticket #933): sys/nfs/nfs_bio.c: revision 1.189 Function nfs_vinvalbuf() ignores errors from vinvalbuf() and therefore delayed write errors may get lost. Change nfs_vinvalbuf() to keep errors from vinvalbuf() for fsync() or = close(). =20 Presented on tech-kern@ =20 Fix for PR kern/47980 (NFS over-quota not detected if utimes() called before fsync()/close()) =20 =20
1.188.16.1	28-Aug-2013	rmind	sync with head
1.188.14.1	07-Sep-2013	bouyer	Pull up following revision(s) (requested by hannken in ticket #933): sys/nfs/nfs_bio.c: revision 1.189 Function nfs_vinvalbuf() ignores errors from vinvalbuf() and therefore delayed write errors may get lost. Change nfs_vinvalbuf() to keep errors from vinvalbuf() for fsync() or = close(). =20 Presented on tech-kern@ =20 Fix for PR kern/47980 (NFS over-quota not detected if utimes() called before fsync()/close()) =20 =20
1.188.12.2	03-Dec-2017	jdolecek	update from HEAD
1.188.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.188.8.1	07-Sep-2013	bouyer	Pull up following revision(s) (requested by hannken in ticket #933): sys/nfs/nfs_bio.c: revision 1.189 Function nfs_vinvalbuf() ignores errors from vinvalbuf() and therefore delayed write errors may get lost. Change nfs_vinvalbuf() to keep errors from vinvalbuf() for fsync() or = close(). =20 Presented on tech-kern@ =20 Fix for PR kern/47980 (NFS over-quota not detected if utimes() called before fsync()/close()) =20 =20
1.188.2.4	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")
1.188.2.3	25-Jan-2012	yamt	uvm_loanabj: take an access pattern hint.
1.188.2.2	04-Jan-2012	yamt	enable O->A loaning read for a few filesystems.
1.188.2.1	02-Nov-2011	yamt	page cache related changes - maintain object pages in radix tree rather than rb tree. - reduce unnecessary page scan in putpages. esp. when an object has a ton of pages cached but only a few of them are dirty. - reduce the number of pmap operations by tracking page dirtiness more precisely in uvm layer. - fix nfs commit range tracking. - fix nfs write clustering. XXX hack
1.189.4.1	04-Nov-2015	riz	Pull up following revision(s) (requested by manu in ticket #882): sbin/umount/umount.c: revision 1.48 sys/nfs/nfsmount.h: revision 1.53 sys/nfs/nfs_var.h: revision 1.94 sys/nfs/nfs_iod.c: revision 1.7 sys/nfs/nfs_socket.c: revision 1.197 sys/nfs/nfs_bio.c: revision 1.191 sys/nfs/nfs_vfsops.c: revision 1.230 sys/nfs/nfs_clntsocket.c: revision 1.3 Remove useless and harmful sync(2) call in umount(8) Remove sync(2) call before unmount(2) in umount(8). This sync(2) is useless since unmount(2) will perform a VFS_SYNC anyway. But moreover, this sync(2) may be harmful, as there are some situation where it cannot return (unreachable NFS server, for instance), causing umount -f to be uneffective. Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.190.2.1	22-Sep-2015	skrll	Sync with HEAD
1.191.18.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.192.2.2	29-Feb-2020	ad	Sync with head.
1.192.2.1	17-Jan-2020	ad	Sync with head.
1.195.2.1	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.90	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.89	20-Sep-2022	knakahara	branches: 1.89.10; Remove routes on an address removal if the routes referencing to the address. Implemented by ozaki-r@n.o. A route that has a gateway is on a connected route can be invalid if the connected route is deleted, i.e., an associated address is removed. Traditionally NetBSD doesn't sweep such a route on the address removal. Sending packets over the route fails with "No route to host". Also the route holds an orphan ifaddr as rt_ifa that is destructed say by in_purgeaddr. If the same address is assgined again in such a state, there can be two different ifaddr objects with the same address. Until recently it's not a big problem because we can send packets anyway. However after MP-ification of the network stack, we can't send packets because we strictly check if rt_ifa (i.e., the (old) ifaddr) is valid. This change automatically removes such routes on a removal of an associated address to avoid keeping inconsistent routes.
1.88	17-May-2018	thorpej	Default NFS mounts to using TCP transport instead of UDP. PR kern/53166
1.87	15-Nov-2016	ozaki-r	branches: 1.87.14; Don't use rt_walktree to delete routes Some functions use rt_walktree to scan the routing table and delete matched routes. However, we shouldn't use rt_walktree to delete routes because rt_walktree is recursive to the routing table (radix tree) and isn't friendly to MP-ification. rt_walktree allows a caller to pass a callback function to delete an matched entry. The callback function is called from an API of the radix tree (rn_walktree) but also calls an API of the radix tree to delete an entry. This change adds a new API of the radix tree, rn_search_matched, which returns a matched entry that is selected by a callback function passed by a caller and the caller itself deletes the entry. By using the API, we can avoid the recursive form.
1.86	07-Jul-2016	msaitoh	branches: 1.86.2; KNF. Remove extra spaces. No functional change.
1.85	21-May-2015	rtr	change nfs_boot_sendrecv to take sockaddr_in * instead of mbuf * fixes m_serv (single mbuf leak) leak in kern/subr_tftproot.c
1.84	09-May-2015	rtr	change sosend() to accept sockaddr * instead of mbuf * for nam. bump to 7.99.16
1.83	03-Apr-2015	rtr	* change pr_bind to accept struct sockaddr * instead of struct mbuf * * update protocol bind implementations to use/expect sockaddr * instead of mbuf * * introduce sockaddr_big struct for storage of addr data passed via sys_bind; sockaddr_big is of sufficient size and alignment to accommodate all addr data sizes received. * modify sys_bind to allocate sockaddr_big instead of using an mbuf. * bump kernel version to 7.99.9 for change to pr_bind() parameter type. Patch posted to tech-net@ http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html The choice to use a new structure sockaddr_big has been retained since changing sockaddr_storage size would lead to unnecessary ABI change. The use of the new structure does not preclude future work that increases the size of sockaddr_storage and at that time sockaddr_big may be trivially replaced. Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
1.82	27-Mar-2015	hikaru	m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.81	25-Oct-2013	martin	branches: 1.81.4; 1.81.6; Mark a potentially unused variable
1.80	04-Oct-2010	cyber	branches: 1.80.8; 1.80.14; 1.80.18; 1.80.22; Add support to honor MTU settings from DHCP during netboot. Defines IP_MIN_MTU as 576. Glanced over quickly by martin@ and joerg@.
1.79	04-Mar-2009	nisimura	branches: 1.79.2; 1.79.4; Update comments to reflect realities; the filenames were changed in 1997 and another one was added years ago.
1.78	19-Nov-2008	ad	branches: 1.78.4; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.77	27-Oct-2008	cegger	make this build again without having NFS_BOOT_BOOTSTATIC, NFS_BOOT_BOOTP, NFS_BOOT_DHCP or NFS_BOOT_BOOTPARAM defined. I uncovered this case when compiling rump.
1.76	27-Oct-2008	cegger	change nfs boot behaviour to automatically try next boot method if boot information are incomplete to succeed. That way, it is possible combine static and dhcp boot: For example, to boot diskless you can specify the nfs-server and the rootpath statically. All other information will be taken via dhcp. Patch has been presented on port-xen, tech-kern and tech-net: http://mail-index.netbsd.org/port-xen/2008/10/24/msg004488.html http://mail-index.netbsd.org/tech-kern/2008/10/24/msg003255.html http://mail-index.netbsd.org/tech-net/2008/10/24/msg000864.html No comments, no objections.
1.75	24-Oct-2008	cegger	branches: 1.75.2; - ansify function definition - de- __P - u_int32_t -> uint32_t No functional changes.
1.74	06-Aug-2008	plunky	Convert socket options code to use a sockopt structure instead of laying everything into an mbuf. approved by core
1.73	22-May-2008	dyoung	branches: 1.73.4; Delete unnecessary cast to void *.
1.72	28-Apr-2008	martin	branches: 1.72.2; Remove clause 3 and 4 from TNF licenses
1.71	24-Apr-2008	ad	branches: 1.71.2; Merge the socket locking patch: - Socket layer becomes MP safe. - Unix protocols become MP safe. - Allows protocol processing interrupts to safely block on locks. - Fixes a number of race conditions. With much feedback from matt@ and plunky@.
1.70	05-Apr-2008	cegger	branches: 1.70.2; use aprint_*_dev and device_xname
1.69	31-Aug-2007	dyoung	branches: 1.69.20; Use sockaddr_in_init() and ifreq_setaddr() to initialize a sockaddr_in and an ifreq.ifr_addr, respectively. Get rid of an extraneous cast---down the elevator shaft! Change 'ireq' to 'ifr', which is what the kernel calls a temporary struct ifreq virtually everywhere else.
1.68	19-Jul-2007	dyoung	branches: 1.68.4; 1.68.6; 1.68.8; Take steps to hide the radix_node implementation of the forwarding table from the forwarding table's users: Introduce rt_walktree() for walking the routing table and applying a function to each rtentry. Replace most rn_walktree() calls with it. Use rt_getkey()/rt_setkey() to get/set a route's destination. Keep a pointer to the sockaddr key in the rtentry, so that rtentry users do not have to grovel in the radix_node for the key. Add a RTM_GET method to rtrequest. Use that instead of radix_node lookups in, e.g., carp(4). Add sys/net/link_proto.c, which supplies sockaddr routines for link-layer socket addresses (sockaddr_dl). Cosmetic: Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH, et cetera. Use NULL instead of 0 for null pointers. Use __arraycount(). Reduce gratuitous parenthesization. Stop using variadic arguments for rip6_output(), it is unnecessary. Remove the unnecessary rtentry member rt_genmask and the code to maintain it, since nothing actually used it. Make rt_maskedcopy() easier to read by using meaningful variable names. Extract a subroutine intern_netmask() for looking up a netmask in the masks table. Start converting backslash-ridden IPv6 macros in sys/netinet6/in6_var.h into inline subroutines that one can read without special eyeglasses. One functional change: when the kernel serves an RTM_GET, RTM_LOCK, or RTM_CHANGE request, it applies the netmask (if supplied) to a destination before searching for it in the forwarding table. I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove the unlawful radix_node knowledge. Apart from the changes to carp(4), netiso, ATM, and strip(4), I have run the changes on three nodes in my wireless routing testbed, which involves IPv4 + IPv6 dynamic routing acrobatics, and it's working beautifully so far.
1.67	09-Jun-2007	dyoung	branches: 1.67.2; Cosmetic: de-__P() et cetera.
1.66	08-May-2007	manu	Fix buid (broken by a fix introduced in the wrong file...)
1.65	08-May-2007	manu	Add the TFTPROOT kernel option for TFTP'ing root RAMdisk at root mount time. This allows working around situations where a kernel with embedded RAMdisk cannot be booted by the bootloader because the RAMdisk is too big.
1.64	04-Mar-2007	christos	branches: 1.64.2; 1.64.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.63	01-Mar-2006	yamt	branches: 1.63.18; 1.63.20; 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.
1.62	11-Dec-2005	christos	branches: 1.62.2; 1.62.4; 1.62.6; merge ktrace-lwp.
1.61	22-May-2004	jonathan	branches: 1.61.12; Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.60	11-Mar-2004	cl	Add static nfs boot configuration, from the kernel config file or from a driver selectable callback function. This is used in the Xen port to allow controlling the domain's network setup from the domain building environment at domain creation (vs. having to maintain/change this on a dhcp server). The Xen network driver parses a command line passed in from the domain builder.
1.59	29-Jun-2003	fvdl	branches: 1.59.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.58	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
1.57	10-Nov-2001	lukem	add RCSIDs
1.56	19-Jan-2001	enami	branches: 1.56.2; 1.56.4; 1.56.8; Use tsleep instead of dalay; since we're mounting root, we can sleep and no reason to use delay.
1.55	10-Dec-2000	fvdl	Make sobind() take a struct proc *. It already took curproc and passed it down to the appropriate usrreq function, and this allows usage for contexts that need to be explicitly different from curproc (like in the NFS code when binding to a reserved port).
1.54	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.53	29-Mar-2000	simonb	branches: 1.53.4; Don't need to include <sys/conf.h> here.
1.52	03-Sep-1999	drochner	branches: 1.52.2; Wait some seconds after the interface is brought up before packets are sent. Needed at least for if_ti to get the link up.
1.51	07-Jul-1999	drochner	mount diskless root with "NFSMNT_NOCONN" (which is default in "mount_nfs" for quite a while) to allow certain servers (multihomed, as our DEC NSE cluster) to be used as root filesystem without special tweaks
1.50	21-Feb-1999	drochner	branches: 1.50.2; 1.50.4; restructure the diskless NFS boot code to keep track of the used interface and the address allocated, to roll everything back if the mount fails: -put an interface pointer into "struct nfs_diskless" to have it available for cleanup, don't pass it around anymore where the "struct nfs_diskless" is already passed -add a "cleanup" function which shuts the interface down -in the protocol-specific parts, either return with "everything ready" or "completely shut down" -use common functions for interface initialization and shutdown -add a function to delete all routes associate to an interface (why is this necessary and not done by ~IFF_UP?) g/c diskless swap stuff general cleanup
1.49	13-Sep-1998	christos	Fix copyright spacing.
1.48	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.47	13-Jun-1998	drochner	Fix last change: If BOOTP/DHCP was successful, don't try RARP/BOOTPARAM.
1.46	13-Jun-1998	tv	Clean up boogered gcc warning workaround (remove goto completely) and remove a redundant `if'.
1.45	25-Apr-1998	matt	Adapt to new sosend/soreceive and upcall (now down in sowakeup)
1.44	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.43	28-Feb-1998	cgd	be a bit more clear about what protocols will be tried in the BOOTP/DHCP case.
1.42	12-Jan-1998	scottr	Consolidate NFS_BOOT_* options into opt_nfs_boot.h
1.41	11-Jan-1998	scottr	Make NFS_BOOT_DHCP work as expected.
1.40	09-Jan-1998	drochner	Use new options "NFS_BOOT_BOOTP" and "NFS_BOOT_BOOTPARAM" for parts conditional on a particular configuration method. The global flags "nfs_boot_rfc951" and "nfs_boot_bootparam" control independantly if the functions are actually called. (Previous meaning of "nfs_boot_rfc951" was "either-or".)
1.39	30-Sep-1997	drochner	Factor out some functions used by bootparam and DHCP boot.
1.38	13-Sep-1997	thorpej	Correct a comment regarding the sense of the nfs_boot_rfc951 global.
1.37	09-Sep-1997	gwr	Move the call to nfs_boot_getfh() from nfs_vfsops.c to nfs_boot.c (just for better isolation - it can now be static)
1.36	02-Sep-1997	gwr	Change test from NETHER to NARP (revarpwhoami is in if_arp.c)
1.35	29-Aug-1997	gwr	Supporting changes for the new BOOTP support in nfs_mountroot.
1.34	14-Aug-1997	drochner	1. Allow to set a netmask (option NFS_BOOT_NETMASK) for the booting interface. Without this, NFS_BOOT_NETMASK could be useless in subnetting envirinment. 2. Comment out unneeded NFS swap related stuff. Closes PR kern/3918.
1.33	27-May-1997	gwr	branches: 1.33.4; Minor reorganization of nfs_mountroot code to simplify BOOTP support. The RPC/bootparamd calls to get the root and swap paths are now done in nfs_boot_init() instead of nfs_boot_getfh(), so the latter now just does the RPC/mountd call. Also changed some panics into error returns.
1.32	17-Mar-1997	thorpej	Add some missing "\n"'s.
1.31	15-Mar-1997	is	New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.30	31-Jan-1997	thorpej	branches: 1.30.2; - Don't look for a "suitable interface"; we're now given the name of the network interface to use. - If any part of the NFS root mount process fails, don't panic. Simply return the appropriate error and let the caller recover.
1.29	20-Oct-1996	fvdl	branches: 1.29.2; Enhancements from Matthias Drochner: - Try V3 first for diskless booting. Fall back to V2 if V3 fails. - optionally (option NFS_BOOT_TCP) try a TCP mount first for diskless booting. Fall back to UDP if it fails. - Enable switching between UDP and TCP for remounts.
1.28	13-Oct-1996	christos	revert kprintf changes
1.27	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.26	07-May-1996	thorpej	Changed struct ifnet to have a pointer to the softc of the underlying device and a printable "external name" (name + unit number), thus eliminating if_name and if_unit. Updated interface to (if_watchdog)() and (if_reset)() to take a struct ifnet *, rather than a unit number.
1.25	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.24	16-Feb-1996	gwr	Add stub for nfs_boot_getfh if NETHER==0
1.23	13-Feb-1996	gwr	Do the RPC to bootparamd a little later (just before the mountd call) so that we do not ask for the "swap" path when swapping on disk.
1.22	10-Feb-1996	pk	Don't return EBADRPC if we have something else.
1.21	09-Feb-1996	christos	nfs prototype changes
1.20	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.19	12-Jun-1995	mycroft	Various cleanup, including: * Convert several data structures to use queue.h. * Split in_pcbnotify() into two parts; one for notifying a specific PCB, and one for notifying all PCBs for a particular foreign address.
1.18	23-May-1995	cgd	don't blindly set IFF_UP; or it with old flags
1.17	20-May-1995	mycroft	Use fxdr_() and txdr_() macros to do byte order conversions.
1.16	24-Apr-1995	gwr	Fixed RPC code to deal with RPC messages larger than one mbuf.
1.15	28-Mar-1995	gwr	Cosmetic changes suggested by Adam.
1.14	18-Mar-1995	gwr	Do the printf "root/swap on" elsewhere to avoid confusion.
1.13	16-Feb-1995	pk	Working "config generic" support; from Theo.
1.12	29-Oct-1994	cgd	fix a couple of obvious, painful endianness bugs introduced in last commit.
1.11	26-Sep-1994	gwr	Do the first BOOTPARAM RPC call to the broadcast address instead of using the address of the RARP server because a BOOTPARAM server might not be running on the machine that sent the RARP reply.
1.10	11-Aug-1994	mycroft	char * --> caddr_t, where appropriate.
1.9	11-Aug-1994	gwr	Diskless boot will now bind the local socket to a reserved port to satisfy picky servers. Also fix some missing initializations. (Thanks to Chuck Cranor for PR#394 -- now fixed.)
1.8	19-Jul-1994	gwr	Fix the conditionally compiled code inside #ifdef NFS_BOOT_GATEWAY and make some printf args use host byteorder.
1.7	16-Jul-1994	paulus	If we don't have ethernet, nfs_boot_init reduces to just a panic. This is so I don't get an undefined symbol compiling a kernel with NFSCLIENT but no ethernet.
1.6	29-Jun-1994	deraadt	branches: 1.6.2; knf
1.5	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.4	21-Jun-1994	pk	Construct mountpath for remote root.
1.3	13-Jun-1994	gwr	New diskless boot code (uses RARP, bootparamd).
1.2	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.
1.1	18-Apr-1994	glass	revised nfs diskless support. uses bootp+rpc to gather parameters
1.6.2.3	11-Aug-1994	mycroft	update from trunk
1.6.2.2	19-Jul-1994	cgd	from trunk, per gwr.
1.6.2.1	16-Jul-1994	cgd	update from trunk, per paulus
1.29.2.1	14-Jan-1997	thorpej	Snapshot of work-in-progress, committed to private branch. These changes implement machine-independent root device and file system selection. Notable features: - All ports behave in a consistent manner regarding root device selection. - No more "options GENERIC"; all kernels have the ability to boot with RB_ASKNAME to select root device and file system type. - Root file system type can be wildcarded; a machine-independent function will try all possible file systems for the selected root device until one succeeds. - If the root file system fails to mount, the operator will be given the chance to select a new root device and file system type, rather than having the machine simply panic. - nfs_mountroot() no longer panics if any part of the NFS mount process fails; it now returns an error, giving the operator a chance to recover. - New, more consistent, config(8) grammar. The constructs: config netbsd swap generic config netbsd root on nfs have been replaced with: config netbsd root on ? type ? config netbsd root on ? type nfs Additionally, the operator may select or wildcard root file system type in the kernel configuration file: config netbsd root on cd0a type cd9660 config(8) now requires that a "root" specification be made. "root" may be wired down or wildcarded. "swap" and "dump" specifications are optional, and follow previous semantics. - config(8) has a new "file-system" keyword, used to configure file systems into the kernel. Eventually, this will be used to generate the default vfssw[]. - "options NFSCLIENT" is obsolete, and is replaced by "file-system NFS". "options NFSSERVER" still exists, since NFS server support is independent of the NFS file system client. - sys/arch/<foo>/<foo>/swapgeneric.c is no longer used, and will be removed; all information is now generated by config(8). As of this commit, all ports except arm32 have been updated to use the new setroot(). Only SPARC, i386, and Alpha ports have been tested at this time. Port masters should test these changes on their ports, and report any problems back to me. More changes are on their way, including RB_ASKNAME support in nfs_mountroot() (to prompt for server address and path) and, potentially, the ability to select rarp/bootparam or bootp in nfs_mountroot().
1.30.2.2	10-Mar-1997	is	netinet/if_ether.h => netinet/if_inarp.h
1.30.2.1	07-Feb-1997	is	Snapshot of new ARP code. Our old ARP code was hardwired for 6-byte length medium addresses, while the protocol is designed for any size. This snapshot contains a first hack at getting rid of Ethernet specific data structures. The ep driver is updated (and tested on the PCI bus), the iy and fpa drivers have been updated, but not real life tested yet. If you want to test this with other drivers, you have to update them first yourself, and probably tag the relevant directories. Better contact me if you want to do this.
1.33.4.5	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.4.4	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.4.3	04-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.4.2	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.4.1	23-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.50.4.1	02-Aug-1999	thorpej	Update from trunk.
1.50.2.1	05-Oct-1999	he	Pull up revisions 1.51-1.52 (requested by drochner): Mount diskless root with "noconn" option to allow easier use of multi-homed servers. Wait a while between bringing up interface and using it, to allow e.g. if_ti driver to establish the link.
1.52.2.3	11-Feb-2001	bouyer	Sync with HEAD.
1.52.2.2	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.52.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.53.4.1	15-Dec-2000	he	Pull up revision 1.55 (requested by fvdl): Fix NFS+tcp client hangs on server or network outage. Again, please note that this introduces yet another kernel interface change: sobind() gains an argument.
1.56.8.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.56.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.56.2.4	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.56.2.3	24-Jun-2002	nathanw	Curproc->curlwp renaming. Change uses of "curproc->l_proc" back to "curproc", which is more like the original use. Bare uses of "curproc" are now "curlwp". "curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL) so that it is always safe to reference curproc (dereferencing curproc is another story, but that's always been true).
1.56.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.56.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.59.2.5	04-Feb-2005	skrll	Adapt to branch.
1.59.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.59.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.59.2.2	03-Aug-2004	skrll	Sync with HEAD
1.59.2.1	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.
1.61.12.2	03-Sep-2007	yamt	sync with head.
1.61.12.1	21-Jun-2006	yamt	sync with head.
1.62.6.1	22-Apr-2006	simonb	Sync with head.
1.62.4.1	09-Sep-2006	rpaulo	sync with head
1.62.2.1	31-Dec-2005	yamt	- adapt nfs. - nfs_doio_read: #if 0 out "killproc if text is modified" part of the code as it's broken. (a process reading the modified text is not necessarily a process which is using the file as a text.)
1.63.20.2	17-May-2007	yamt	sync with head.
1.63.20.1	12-Mar-2007	rmind	Sync with HEAD.
1.63.18.1	13-May-2007	jdc	Pull up revisions 1.65-1.66 (requested by manu in ticket #635). Add the TFTPROOT kernel option for TFTP'ing root RAMdisk at root mount time. This allows working around situations where a kernel with embedded RAMdisk cannot be booted by the bootloader because the RAMdisk is too big. Fix build (broken by a fix introduced in the wrong file...)
1.64.4.1	11-Jul-2007	mjf	Sync with head.
1.64.2.4	09-Oct-2007	ad	Sync with head.
1.64.2.3	20-Aug-2007	ad	Sync with HEAD.
1.64.2.2	15-Jul-2007	ad	Sync with head.
1.64.2.1	08-Jun-2007	ad	Sync with head.
1.67.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.67.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.68.8.2	19-Jul-2007	dyoung	Take steps to hide the radix_node implementation of the forwarding table from the forwarding table's users: Introduce rt_walktree() for walking the routing table and applying a function to each rtentry. Replace most rn_walktree() calls with it. Use rt_getkey()/rt_setkey() to get/set a route's destination. Keep a pointer to the sockaddr key in the rtentry, so that rtentry users do not have to grovel in the radix_node for the key. Add a RTM_GET method to rtrequest. Use that instead of radix_node lookups in, e.g., carp(4). Add sys/net/link_proto.c, which supplies sockaddr routines for link-layer socket addresses (sockaddr_dl). Cosmetic: Constify. KNF. Stop open-coding LIST_FOREACH, TAILQ_FOREACH, et cetera. Use NULL instead of 0 for null pointers. Use __arraycount(). Reduce gratuitous parenthesization. Stop using variadic arguments for rip6_output(), it is unnecessary. Remove the unnecessary rtentry member rt_genmask and the code to maintain it, since nothing actually used it. Make rt_maskedcopy() easier to read by using meaningful variable names. Extract a subroutine intern_netmask() for looking up a netmask in the masks table. Start converting backslash-ridden IPv6 macros in sys/netinet6/in6_var.h into inline subroutines that one can read without special eyeglasses. One functional change: when the kernel serves an RTM_GET, RTM_LOCK, or RTM_CHANGE request, it applies the netmask (if supplied) to a destination before searching for it in the forwarding table. I have changed sys/netinet/ip_carp.c, carp_setroute(), to remove the unlawful radix_node knowledge. Apart from the changes to carp(4), netiso, ATM, and strip(4), I have run the changes on three nodes in my wireless routing testbed, which involves IPv4 + IPv6 dynamic routing acrobatics, and it's working beautifully so far.
1.68.8.1	19-Jul-2007	dyoung	file nfs_boot.c was added on branch matt-mips64 on 2007-07-19 20:49:01 +0000
1.68.6.1	06-Nov-2007	matt	sync with HEAD
1.68.4.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.69.20.3	17-Jan-2009	mjf	Sync with HEAD.
1.69.20.2	28-Sep-2008	mjf	Sync with HEAD.
1.69.20.1	02-Jun-2008	mjf	Sync with HEAD.
1.70.2.2	04-Jun-2008	yamt	sync with head
1.70.2.1	18-May-2008	yamt	sync with head.
1.71.2.3	09-Oct-2010	yamt	sync with head
1.71.2.2	04-May-2009	yamt	sync with head.
1.71.2.1	16-May-2008	yamt	sync with head.
1.72.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.72.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.73.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.73.4.1	19-Oct-2008	haad	Sync with HEAD.
1.75.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.75.2.1	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.
1.79.4.1	05-Mar-2011	rmind	sync with head
1.79.2.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.80.22.1	18-May-2014	rmind	sync with head
1.80.18.2	03-Dec-2017	jdolecek	update from HEAD
1.80.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.80.14.1	16-Apr-2015	msaitoh	Pull up following revision(s) (requested by hikaru in ticket #1287): sys/kern/subr_tftproot.c: revision 1.14 via patch sys/nfs/nfsdiskless.h: revision 1.31 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_bootdhcp.c: revision 1.53 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.80.8.1	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")
1.81.6.3	05-Dec-2016	skrll	Sync with HEAD
1.81.6.2	06-Jun-2015	skrll	Sync with HEAD
1.81.6.1	06-Apr-2015	skrll	Sync with HEAD
1.81.4.1	06-Apr-2015	snj	Pull up following revision(s) (requested by hikaru in ticket #656): sys/kern/subr_tftproot.c: revision 1.14 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/nfs_bootdhcp.c: revision 1.53 sys/nfs/nfsdiskless.h: revision 1.31 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.86.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.87.14.1	21-May-2018	pgoyette	Sync with HEAD
1.89.10.1	02-Aug-2025	perseant	Sync with HEAD
1.60	20-Oct-2024	mlelstv	MBUFTRACE
1.59	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.58	13-May-2024	msaitoh	branches: 1.58.2; s/contigous/contiguous/ in comment.
1.57	24-Dec-2022	andvar	s/reqest/request/ in comment.
1.56	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.55	21-May-2015	rtr	change nfs_boot_sendrecv to take sockaddr_in * instead of mbuf * fixes m_serv (single mbuf leak) leak in kern/subr_tftproot.c
1.54	09-May-2015	rtr	when calling nfs_boot_sendrecv pass NULL for pointers instead of 0
1.53	27-Mar-2015	hikaru	m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.52	04-Oct-2010	cyber	branches: 1.52.14; 1.52.18; 1.52.34; 1.52.36; Add support to honor MTU settings from DHCP during netboot. Defines IP_MIN_MTU as 576. Glanced over quickly by martin@ and joerg@.
1.51	10-Jul-2009	roy	branches: 1.51.2; 1.51.4; Use a function to add extra data to the vendor area so that data added remains constant for both DISCOVER and REQUEST messages.
1.50	10-Jul-2009	roy	Protect against short IP addresses in the DHCP message.
1.49	10-Jul-2009	roy	When using DHCP, request the parameters that we need. Fixes PR kern/38830. Thanks to Tim McIntosh.
1.48	06-May-2009	cegger	correct previous: use %zu for BOOTP_SIZE_(MIN,MAX). Pointed out by David Holland
1.47	05-May-2009	cegger	buildfix: use %d for BOOTP_SIZE_(MIN,MAX). Makes i386 ALL kernel build again.
1.46	02-May-2009	manu	- Silence warning when running with debug enabled - Remind the administrator about the required DHCP option when some are missing, instead of silently failing, you stupid computer!
1.45	19-Nov-2008	ad	branches: 1.45.4; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.44	27-Oct-2008	cegger	change nfs boot behaviour to automatically try next boot method if boot information are incomplete to succeed. That way, it is possible combine static and dhcp boot: For example, to boot diskless you can specify the nfs-server and the rootpath statically. All other information will be taken via dhcp. Patch has been presented on port-xen, tech-kern and tech-net: http://mail-index.netbsd.org/port-xen/2008/10/24/msg004488.html http://mail-index.netbsd.org/tech-kern/2008/10/24/msg003255.html http://mail-index.netbsd.org/tech-net/2008/10/24/msg000864.html No comments, no objections.
1.43	24-Oct-2008	cegger	branches: 1.43.2; - ansify function definition - de- __P - u_int32_t -> uint32_t No functional changes.
1.42	06-Aug-2008	plunky	Convert socket options code to use a sockopt structure instead of laying everything into an mbuf. approved by core
1.41	20-Jul-2008	uwe	When doing pointer arithmetic to compute limit cast bootp to pointer type of correct signedness. Caught by lint.
1.40	09-May-2008	rumble	branches: 1.40.2; 1.40.4; Fix compilation with DEBUG_NFS_BOOT_DHCP and ssp.
1.39	28-Apr-2008	martin	branches: 1.39.2; Remove clause 3 and 4 from TNF licenses
1.38	24-Apr-2008	ad	branches: 1.38.2; Merge the socket locking patch: - Socket layer becomes MP safe. - Unix protocols become MP safe. - Allows protocol processing interrupts to safely block on locks. - Fixes a number of race conditions. With much feedback from matt@ and plunky@.
1.37	20-Dec-2007	dyoung	branches: 1.37.6; 1.37.8; Constify.
1.36	29-Aug-2007	dyoung	branches: 1.36.8; 1.36.12; Constify: LLADDR() -> CLLADDR().
1.35	08-May-2007	manu	branches: 1.35.2; 1.35.6; 1.35.8; Fix buid (broken by a fix introduced in the wrong file...)
1.34	08-May-2007	manu	Add the TFTPROOT kernel option for TFTP'ing root RAMdisk at root mount time. This allows working around situations where a kernel with embedded RAMdisk cannot be booted by the bootloader because the RAMdisk is too big.
1.33	04-Mar-2007	christos	branches: 1.33.2; 1.33.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.32	09-Nov-2006	yamt	branches: 1.32.2; 1.32.4; remove some __unused in function parameters.
1.31	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.30	16-Mar-2006	christos	branches: 1.30.10; 1.30.12; Don't use DEBUG, add a new DEBUG_NFS_BOOT_DHCP variable to provide more information. Print more information about what fails.
1.29	11-Dec-2005	christos	branches: 1.29.4; 1.29.6; 1.29.8; 1.29.10; merge ktrace-lwp.
1.28	26-Feb-2005	perry	branches: 1.28.4; nuke trailing whitespace
1.27	22-May-2004	jonathan	branches: 1.27.4; 1.27.6; Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.26	06-May-2004	drochner	remove duplicated snprintf(vci, ...)
1.25	21-Apr-2004	itojun	kill sprintf, use snprintf
1.24	29-Jun-2003	fvdl	branches: 1.24.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.23	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
1.22	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.21	10-Jun-2002	drochner	increase IP TTL for BOOTP/DHCP request packets to avoid dumb routers dropping the packet (seems to be a problem with Cisco and its "helper-address" feature; a Cabletron SSR I tested with didn't have this problem)
1.20	12-May-2002	simonb	branches: 1.20.2; In bootpcheck(), make sure we m_pullup() all the of bootp header that we actually examine. While here, toss out home-grown ofs() macro and use offsetof().
1.19	20-Mar-2002	thorpej	Add a NetBSD Vendor Class Identifier option as proposed on tech-net in message <20020216172527.C23901@dr-evil.shagadelic.org>.
1.18	10-Nov-2001	lukem	add RCSIDs
1.17	02-Jun-2001	kim	branches: 1.17.2; 1.17.6; It is misleading that the kernel outputs "DHCP server:" followed by the value of "next-server" from the DHCP (or BOOTP) reply. This is not the DHCP server's IP address (except by chance), so instead of "server" make it print "next-server".
1.16	05-Dec-2000	drochner	branches: 1.16.2; add a kernel configuration option to set the string passed in bp_file in diskless BOOTP/DHCP configuration - good for booting different userland versions depending on the kernel version
1.15	28-May-2000	gmcgarry	Allow nfs root over token ring. Closes PR6629.
1.14	29-Mar-2000	simonb	branches: 1.14.2; Don't need to include <sys/conf.h> here.
1.13	20-Jan-2000	enami	If server name field is overloaded for other purpose, or it just contains NULL string, don't use it as server name.
1.12	07-May-1999	drochner	branches: 1.12.2; print diskless boot related IP addresses in dot notation
1.11	21-Feb-1999	drochner	branches: 1.11.4; restructure the diskless NFS boot code to keep track of the used interface and the address allocated, to roll everything back if the mount fails: -put an interface pointer into "struct nfs_diskless" to have it available for cleanup, don't pass it around anymore where the "struct nfs_diskless" is already passed -add a "cleanup" function which shuts the interface down -in the protocol-specific parts, either return with "everything ready" or "completely shut down" -use common functions for interface initialization and shutdown -add a function to delete all routes associate to an interface (why is this necessary and not done by ~IFF_UP?) g/c diskless swap stuff general cleanup
1.10	12-Feb-1999	thorpej	Fix printf format problems on Alpha.
1.9	13-Sep-1998	christos	Fix copyright spacing.
1.8	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.7	24-Apr-1998	drochner	-catch zero and broadcast IP addresses sent by a DHCP server -warn about them (and other invalid replies) -print address of BOOTP/DHCP server for better problem tracking -KNF
1.6	01-Mar-1998	ross	Sweep up some miscellaneous leftover lite2 integration shrapnel.
1.5	12-Jan-1998	scottr	Consolidate NFS_BOOT_* options into opt_nfs_boot.h
1.4	11-Jan-1998	scottr	Make NFS_BOOT_DHCP work as expected.
1.3	09-Jan-1998	drochner	Use interface type to select "hardware type" in bootp header.
1.2	30-Sep-1997	drochner	branches: 1.2.2; Make this file deserve its name: add DHCP support, conditionalized with NFS_BOOT_DHCP. Don't increment xid between retries anymore, it is not required and it increases the response time in case of a slow server. Use common code with bootparam boot.
1.1	29-Aug-1997	gwr	branches: 1.1.2; Add support for nfs_mountroot using BOOTP based on the contributions of Tor Egge (closes PR kern/2351).
1.1.2.3	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.1.2.2	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.1.2.1	29-Aug-1997	thorpej	file nfs_bootdhcp.c was added on branch marc-pcmcia on 1997-09-01 21:02:54 +0000
1.2.2.1	08-May-1998	mycroft	Pull up 1.7, per request of drochner.
1.11.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.12.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.12.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.14.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.16.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.16.2.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.16.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.16.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.17.6.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.17.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.17.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.20.2.1	20-Jun-2002	gehenna	catch up with -current.
1.24.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.24.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.24.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.24.2.2	03-Aug-2004	skrll	Sync with HEAD
1.24.2.1	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.
1.27.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.27.4.1	29-Apr-2005	kent	sync with -current
1.28.4.4	21-Jan-2008	yamt	sync with head
1.28.4.3	03-Sep-2007	yamt	sync with head.
1.28.4.2	30-Dec-2006	yamt	sync with head.
1.28.4.1	21-Jun-2006	yamt	sync with head.
1.29.10.1	19-Apr-2006	elad	sync with head.
1.29.8.1	01-Apr-2006	yamt	sync with head.
1.29.6.1	22-Apr-2006	simonb	Sync with head.
1.29.4.1	09-Sep-2006	rpaulo	sync with head
1.30.12.2	10-Dec-2006	yamt	sync with head.
1.30.12.1	22-Oct-2006	yamt	sync with head
1.30.10.1	18-Nov-2006	ad	Sync with head.
1.32.4.2	17-May-2007	yamt	sync with head.
1.32.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.32.2.1	13-May-2007	jdc	Pull up revisions 1.34-1.35 (requested by manu in ticket #635). Add the TFTPROOT kernel option for TFTP'ing root RAMdisk at root mount time. This allows working around situations where a kernel with embedded RAMdisk cannot be booted by the bootloader because the RAMdisk is too big. Fix build (broken by a fix introduced in the wrong file...)
1.33.4.1	11-Jul-2007	mjf	Sync with head.
1.33.2.2	09-Oct-2007	ad	Sync with head.
1.33.2.1	08-Jun-2007	ad	Sync with head.
1.35.8.2	09-Jan-2008	matt	sync with HEAD
1.35.8.1	06-Nov-2007	matt	sync with HEAD
1.35.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.35.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.36.12.1	02-Jan-2008	bouyer	Sync with HEAD
1.36.8.1	26-Dec-2007	ad	Sync with head.
1.37.8.1	18-May-2008	yamt	sync with head.
1.37.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.37.6.2	28-Sep-2008	mjf	Sync with HEAD.
1.37.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.38.2.5	09-Oct-2010	yamt	sync with head
1.38.2.4	18-Jul-2009	yamt	sync with head.
1.38.2.3	16-May-2009	yamt	sync with head
1.38.2.2	04-May-2009	yamt	sync with head.
1.38.2.1	16-May-2008	yamt	sync with head.
1.39.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.39.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.40.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.40.4.1	19-Oct-2008	haad	Sync with HEAD.
1.40.2.1	28-Jul-2008	simonb	Sync with head.
1.43.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.45.4.2	23-Jul-2009	jym	Sync with HEAD.
1.45.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.51.4.1	05-Mar-2011	rmind	sync with head
1.51.2.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.52.36.3	09-Jul-2016	skrll	Sync with HEAD
1.52.36.2	06-Jun-2015	skrll	Sync with HEAD
1.52.36.1	06-Apr-2015	skrll	Sync with HEAD
1.52.34.1	06-Apr-2015	snj	Pull up following revision(s) (requested by hikaru in ticket #656): sys/kern/subr_tftproot.c: revision 1.14 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/nfs_bootdhcp.c: revision 1.53 sys/nfs/nfsdiskless.h: revision 1.31 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.52.18.1	03-Dec-2017	jdolecek	update from HEAD
1.52.14.1	16-Apr-2015	msaitoh	Pull up following revision(s) (requested by hikaru in ticket #1287): sys/kern/subr_tftproot.c: revision 1.14 via patch sys/nfs/nfsdiskless.h: revision 1.31 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_bootdhcp.c: revision 1.53 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.58.2.1	02-Aug-2025	perseant	Sync with HEAD
1.40	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.39	29-Jun-2019	kamil	branches: 1.39.34; Appease GCC and initialize arps_ip Fixes build as GCC errors with maybe-uninitialized that is a false positive.
1.38	12-Sep-2013	drochner	branches: 1.38.30; tyop in comment, from Eivind Evensen via OpenBSD
1.37	21-Mar-2010	chs	branches: 1.37.8; 1.37.18; 1.37.22; in nfs_bootparam(), set the corresponding flag for each field that we fill in.
1.36	02-Mar-2010	pooka	branches: 1.36.2; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.35	19-Nov-2008	ad	branches: 1.35.6; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.34	27-Oct-2008	cegger	change nfs boot behaviour to automatically try next boot method if boot information are incomplete to succeed. That way, it is possible combine static and dhcp boot: For example, to boot diskless you can specify the nfs-server and the rootpath statically. All other information will be taken via dhcp. Patch has been presented on port-xen, tech-kern and tech-net: http://mail-index.netbsd.org/port-xen/2008/10/24/msg004488.html http://mail-index.netbsd.org/tech-kern/2008/10/24/msg003255.html http://mail-index.netbsd.org/tech-net/2008/10/24/msg000864.html No comments, no objections.
1.33	24-Oct-2008	cegger	branches: 1.33.2; - ansify function definition - de- __P - u_int32_t -> uint32_t No functional changes.
1.32	28-Apr-2008	martin	branches: 1.32.6; Remove clause 3 and 4 from TNF licenses
1.31	02-Jan-2008	yamt	branches: 1.31.6; 1.31.8; 1.31.10; use kmem_alloc instead of malloc.
1.30	04-Mar-2007	christos	branches: 1.30.16; 1.30.22; 1.30.28; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.29	15-Apr-2006	christos	branches: 1.29.14; m_freem takes one arg.
1.28	15-Apr-2006	christos	s/mfree/m_freem/
1.27	15-Apr-2006	christos	Don't leak mbufs on error.
1.26	11-Dec-2005	christos	branches: 1.26.4; 1.26.6; 1.26.8; 1.26.10; 1.26.12; merge ktrace-lwp.
1.25	29-May-2005	christos	branches: 1.25.2; - sprinkle const - avoid shadowed variables - mark bad const use with XXXUNCONST
1.24	22-May-2004	jonathan	Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.23	29-Jun-2003	fvdl	branches: 1.23.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.22	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
1.21	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.20	22-Sep-2002	jdolecek	<sys/conf.h> and <sys/device.h> are not needed here
1.19	10-Nov-2001	lukem	add RCSIDs
1.18	13-Oct-2001	simonb	branches: 1.18.2; Remove so variables that are only ever set and never referenced.
1.17	03-Oct-2000	chs	branches: 1.17.2; 1.17.4; include opt_inet.h, needed by previous check-in.
1.16	02-Oct-2000	itojun	perform reverse ARP only if INET is compiled into the kernel
1.15	26-Jul-1999	enami	branches: 1.15.2; Don't use the result of inet_ntoa after calling the another inet_ntoa, since they share the same static storage.
1.14	07-May-1999	drochner	-print diskless boot related IP addresses in dot notation -arrange gateway code to fall back to the old method if the new "getfile" is not answered (and both are enabled -- allow to switch off the new method for symmetry) -handle error if setting the netmask fails
1.13	12-Apr-1999	ross	libkern just got an inet_addr(), but it won't compile, no prototype. Cleanup... * Add prototype to libkern.h. * Remove the almost-identical-copy from libsa/net.[ch]. * Change its type back to the (wrong, but harmless) historical one. (u_long) * Kill the XXX local prototype in nfs_bootparam.c
1.12	11-Apr-1999	gwr	Enable the code that gets our gateway+netmask from the bootparam server using the "gateway" pseudo file. (Compatible with sys/lib/libsa/dev_net.c)
1.11	21-Feb-1999	drochner	branches: 1.11.4; restructure the diskless NFS boot code to keep track of the used interface and the address allocated, to roll everything back if the mount fails: -put an interface pointer into "struct nfs_diskless" to have it available for cleanup, don't pass it around anymore where the "struct nfs_diskless" is already passed -add a "cleanup" function which shuts the interface down -in the protocol-specific parts, either return with "everything ready" or "completely shut down" -use common functions for interface initialization and shutdown -add a function to delete all routes associate to an interface (why is this necessary and not done by ~IFF_UP?) g/c diskless swap stuff general cleanup
1.10	13-Sep-1998	christos	Fix copyright spacing.
1.9	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.8	13-Jun-1998	tv	Fix boogered gcc warning workaround the right way.
1.7	01-Mar-1998	veego	Add two includes for the 'struct nfs_args' so it compiles again.
1.6	12-Jan-1998	scottr	Consolidate NFS_BOOT_* options into opt_nfs_boot.h
1.5	09-Jan-1998	drochner	Conditionalize call to RARP, check interface type. (This file can now be included even if no ARP capable interfaces are defined.)
1.4	12-Dec-1997	gwr	Temporarily disable the bootparam "gateway" support.
1.3	10-Dec-1997	gwr	Change the format of the bootparam "gateway" parameter string to gateway=server:255.255.255.0 because that is the perferred format, and the sys/libsa code already knows how to parse that format. (Copied ip_convert here from the libsa code.)
1.2	09-Sep-1997	gwr	branches: 1.2.2; Circumvent the lack of a reliable gateway/netmask value in the Sun RPC bootparam/whoami return by requesting a "pseudo file" named "gateway" and using its contents as the gateway:netmask Example /etc/bootparams line: client gateway=router:0xfffffff0
1.1	29-Aug-1997	gwr	branches: 1.1.2; Support for RARP,RPC/bootparam moved from nfs_boot.c to here.
1.1.2.3	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.1.2.2	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.1.2.1	29-Aug-1997	thorpej	file nfs_bootparam.c was added on branch marc-pcmcia on 1997-09-01 21:02:56 +0000
1.2.2.1	14-Dec-1997	mellon	Pull rev 1.3 and 1.4 up from trunk (gwr)
1.11.4.2	02-Aug-1999	thorpej	Update from trunk.
1.11.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.15.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.17.4.2	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
1.17.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.17.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.17.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.17.2.1	22-Oct-2001	nathanw	Catch up to -current.
1.18.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.23.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.23.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.23.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.23.2.2	03-Aug-2004	skrll	Sync with HEAD
1.23.2.1	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.
1.25.2.3	21-Jan-2008	yamt	sync with head
1.25.2.2	03-Sep-2007	yamt	sync with head.
1.25.2.1	21-Jun-2006	yamt	sync with head.
1.26.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.26.10.1	19-Apr-2006	elad	sync with head.
1.26.8.1	24-May-2006	yamt	sync with head.
1.26.6.1	22-Apr-2006	simonb	Sync with head.
1.26.4.1	09-Sep-2006	rpaulo	sync with head
1.29.14.1	12-Mar-2007	rmind	Sync with HEAD.
1.30.28.1	02-Jan-2008	bouyer	Sync with HEAD
1.30.22.1	18-Feb-2008	mjf	Sync with HEAD.
1.30.16.1	09-Jan-2008	matt	sync with HEAD
1.31.10.4	11-Aug-2010	yamt	sync with head.
1.31.10.3	11-Mar-2010	yamt	sync with head
1.31.10.2	04-May-2009	yamt	sync with head.
1.31.10.1	16-May-2008	yamt	sync with head.
1.31.8.1	18-May-2008	yamt	sync with head.
1.31.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.31.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.32.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.33.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.35.6.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.36.2.1	30-May-2010	rmind	sync with head
1.37.22.1	18-May-2014	rmind	sync with head
1.37.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.37.8.1	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")
1.38.30.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.39.34.1	02-Aug-2025	perseant	Sync with HEAD
1.8	23-Oct-2009	snj	Remove 3rd and 4th clauses. OK cl@ (copyright holder).
1.7	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.6	27-Oct-2008	cegger	change nfs boot behaviour to automatically try next boot method if boot information are incomplete to succeed. That way, it is possible combine static and dhcp boot: For example, to boot diskless you can specify the nfs-server and the rootpath statically. All other information will be taken via dhcp. Patch has been presented on port-xen, tech-kern and tech-net: http://mail-index.netbsd.org/port-xen/2008/10/24/msg004488.html http://mail-index.netbsd.org/tech-kern/2008/10/24/msg003255.html http://mail-index.netbsd.org/tech-net/2008/10/24/msg000864.html No comments, no objections.
1.5	08-Jul-2007	bouyer	branches: 1.5.28; 1.5.32; 1.5.38; 1.5.42; Add a new BOOTSTATIC flag, NFS_BOOTSTATIC_NOSTATIC, which causes nfs_bootstatic() to abort with EOPNOTSUPP. This allows a callback to say that there is no bootstatic config, and the next NFS boot method should be tried.
1.4	04-Mar-2007	christos	branches: 1.4.2; 1.4.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.3	11-Dec-2005	christos	branches: 1.3.26; merge ktrace-lwp.
1.2	26-Feb-2005	perry	branches: 1.2.4; nuke trailing whitespace
1.1	11-Mar-2004	cl	branches: 1.1.4; 1.1.10; 1.1.12; Add static nfs boot configuration, from the kernel config file or from a driver selectable callback function. This is used in the Xen port to allow controlling the domain's network setup from the domain building environment at domain creation (vs. having to maintain/change this on a dhcp server). The Xen network driver parses a command line passed in from the domain builder.
1.1.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.10.1	29-Apr-2005	kent	sync with -current
1.1.4.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.4.5	04-Feb-2005	skrll	Adapt to branch.
1.1.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	11-Mar-2004	skrll	file nfs_bootstatic.c was added on branch ktrace-lwp on 2004-08-03 10:56:17 +0000
1.2.4.2	03-Sep-2007	yamt	sync with head.
1.2.4.1	21-Jun-2006	yamt	sync with head.
1.3.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.4.4.1	11-Jul-2007	mjf	Sync with head.
1.4.2.1	15-Jul-2007	ad	Sync with head.
1.5.42.1	19-Jan-2009	skrll	Sync with HEAD.
1.5.38.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.5.32.2	11-Mar-2010	yamt	sync with head
1.5.32.1	04-May-2009	yamt	sync with head.
1.5.28.1	17-Jan-2009	mjf	Sync with HEAD.
1.7	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.6	21-Jan-2018	christos	branches: 1.6.40; PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.5	17-Jun-2016	christos	branches: 1.5.10; Serialize all access to the NFS request queue via splsoftnet(). Fixes random crashes. XXX: Pullup-7
1.4	13-Jun-2016	christos	Simplify, no functional change.
1.3	15-Jul-2015	manu	Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.2	05-Sep-2014	matt	branches: 1.2.2; Don't use catch as a variable name.
1.1	02-Mar-2010	pooka	branches: 1.1.2; 1.1.6; 1.1.24; 1.1.40; 1.1.42; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.1.42.1	10-Jul-2016	martin	Pull up following revision(s) (requested by christos in ticket #1184): sys/nfs/nfs_socket.c: revision 1.198 sys/nfs/nfs_clntsocket.c: revision 1.5 Serialize all access to the NFS request queue via splsoftnet(). Fixes random crashes.
1.1.40.2	10-Jul-2016	martin	Pull up following revision(s) (requested by christos in ticket #1184): sys/nfs/nfs_socket.c: revision 1.198 sys/nfs/nfs_clntsocket.c: revision 1.5 Serialize all access to the NFS request queue via splsoftnet(). Fixes random crashes. XXX: Pullup-7
1.1.40.1	04-Nov-2015	riz	Pull up following revision(s) (requested by manu in ticket #882): sbin/umount/umount.c: revision 1.48 sys/nfs/nfsmount.h: revision 1.53 sys/nfs/nfs_var.h: revision 1.94 sys/nfs/nfs_iod.c: revision 1.7 sys/nfs/nfs_socket.c: revision 1.197 sys/nfs/nfs_bio.c: revision 1.191 sys/nfs/nfs_vfsops.c: revision 1.230 sys/nfs/nfs_clntsocket.c: revision 1.3 Remove useless and harmful sync(2) call in umount(8) Remove sync(2) call before unmount(2) in umount(8). This sync(2) is useless since unmount(2) will perform a VFS_SYNC anyway. But moreover, this sync(2) may be harmful, as there are some situation where it cannot return (unreachable NFS server, for instance), causing umount -f to be uneffective. Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.1.24.1	03-Dec-2017	jdolecek	update from HEAD
1.1.6.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.6.1	02-Mar-2010	uebayasi	file nfs_clntsocket.c was added on branch uebayasi-xip on 2010-04-30 14:44:22 +0000
1.1.2.3	21-Mar-2010	yamt	fix merge botches
1.1.2.2	11-Mar-2010	yamt	sync with head
1.1.2.1	02-Mar-2010	yamt	file nfs_clntsocket.c was added on branch yamt-nfs-mp on 2010-03-11 15:04:31 +0000
1.2.2.2	09-Jul-2016	skrll	Sync with HEAD
1.2.2.1	22-Sep-2015	skrll	Sync with HEAD
1.5.10.1	08-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #856): sys/nfs/nfs.h: revision 1.76 sys/nfs/nfs_subs.c: revision 1.230 sys/nfs/nfs_socket.c: revision 1.199 sys/nfs/nfs_clntsocket.c: revision 1.6 PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.6.40.1	02-Aug-2025	perseant	Sync with HEAD
1.7	21-Mar-2023	christos	PR/57279: Izumi Tsutsui: Fix some {int,long} -> time_t. Still things will break eventually because parts of the nfs protocol assume time_t will fit in 32 bits.
1.6	28-Feb-2022	hannken	branches: 1.6.4; Revert the hack from the last commit now that VOP_UNLOCK() no longer may hold v_interlock or vmobjlock.
1.5	14-Jan-2022	christos	This is a temporary hack to avoid nfs crashes related to nfs_delaytruncate.
1.4	23-Feb-2020	ad	UVM locking changes, proposed on tech-kern: - Change the lock on uvm_object, vm_amap and vm_anon to be a RW lock. - Break v_interlock and vmobjlock apart. v_interlock remains a mutex. - Do partial PV list locking in the x86 pmap. Others to follow later.
1.3	03-Sep-2018	riastradh	branches: 1.3.6; 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!)
1.2	12-Jun-2011	rmind	branches: 1.2.52; 1.2.54; 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.
1.1	02-Mar-2010	pooka	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.12; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.1.12.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.1.6.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.6.1	02-Mar-2010	uebayasi	file nfs_clntsubs.c was added on branch uebayasi-xip on 2010-04-30 14:44:22 +0000
1.1.4.1	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.
1.1.2.4	10-Oct-2010	yamt	some locking changes
1.1.2.3	26-Sep-2010	yamt	locking changes
1.1.2.2	11-Mar-2010	yamt	sync with head
1.1.2.1	02-Mar-2010	yamt	file nfs_clntsubs.c was added on branch yamt-nfs-mp on 2010-03-11 15:04:31 +0000
1.2.54.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.2.54.1	10-Jun-2019	christos	Sync with HEAD
1.2.52.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.3.6.1	29-Feb-2020	ad	Sync with head.
1.6.4.1	20-Sep-2024	martin	Pull up following revision(s) (requested by rin in ticket #880): sys/nfs/nfs_iod.c: revision 1.9 sys/nfs/nfs_vfsops.c: revision 1.245 sys/nfs/nfs_clntsubs.c: revision 1.7 PR/57279: Izumi Tsutsui: Fix some {int,long} -> time_t. Still things will break eventually because parts of the nfs protocol assume time_t will fit in 32 bits.
1.63	04-Jun-2021	hannken	Add flag/command NFSSVC_REPLACEEXPORTSLIST to nfssvc(2) system call. Works like NFSSVC_SETEXPORTSLIST but supports "mel_nexports > 1" and will atomically update the complete exports list for a file system.
1.62	17-Jan-2020	ad	branches: 1.62.10; 1.62.14; VFS_VGET(), VFS_ROOT(), VFS_FHTOVP(): give them a "int lktype" argument, to allow us to get shared locks (or no lock) on the returned vnode. Matches FreeBSD.
1.61	22-Dec-2019	ad	branches: 1.61.2; Make mntvnode_lock per-mount, and address false sharing of struct mount.
1.60	17-Apr-2017	hannken	branches: 1.60.12; Remove unused argument "nextp" from vfs_busy() and vfs_unbusy(). Remove argument "keepref" from vfs_unbusy() and add vfs_ref() where needed.
1.59	20-Nov-2016	maxv	branches: 1.59.2; Memory leak, found by Mootja.
1.58	14-Dec-2013	christos	branches: 1.58.6; 1.58.10; don't allow the nfs server module to unload if it has exported filesystems.
1.57	23-Nov-2013	christos	convert from CIRCLEQ to TAILQ
1.56	15-Sep-2013	martin	Remove __CT_LOCAL_.. hack
1.55	14-Sep-2013	martin	Guard a function local CTASSERT with pro/epilogue
1.54	30-Aug-2013	dholland	Use __CTASSERT instead of handrolled version.
1.53	30-Aug-2013	dholland	more typos in comments
1.52	30-Aug-2013	dholland	typo in comment
1.51	27-Sep-2011	christos	branches: 1.51.2; 1.51.12; 1.51.16; use NFS_MAXNAMLEN for all names.
1.50	31-Mar-2011	dyoung	Hide the radix-trie implementation of the forwarding table so that we will have an easier time replacing it with something different, even if it is a second radix-trie implementation. sys/net/route.c and sys/net/rtsock.c no longer operate directly on radix_nodes or radix_node_heads. Hopefully this will reduce the temptation to implement multipath or source-based routing using grotty hacks to the grotty old radix-trie code, too. :-)
1.49	19-Nov-2010	dholland	branches: 1.49.2; Introduce struct pathbuf. This is an abstraction to hold a pathname and the metadata required to interpret it. Callers of namei must now create a pathbuf and pass it to NDINIT (instead of a string and a uio_seg), then destroy the pathbuf after the namei session is complete. Update all namei call sites accordingly. Add a pathbuf(9) man page and update namei(9). The pathbuf interface also now appears in a couple of related additional places that were passing string/uio_seg pairs that were later fed into NDINIT. Update other call sites accordingly.
1.48	07-Jul-2009	christos	branches: 1.48.4; The compatibility call to re-export from sys_mount() calls mountd_set_exports_list, with the mnt_updating mutex held. Account for that to avoid a locking against myself panic.
1.47	23-May-2009	ad	Broken assertion.
1.46	23-May-2009	ad	- Cosmetic change to previous. - Add a comment.
1.45	23-May-2009	ad	- Fix a race between umount()/mount() and nfssvc(). - Toss netexport state on nfsserver module unload.
1.44	17-Dec-2008	cegger	branches: 1.44.2; kill MALLOC and FREE macros.
1.43	28-Nov-2008	pooka	Use kmem instead of malloc to avoid hassle with dynamically attaching a malloc type. Makes nfsserver-as-a-module work. reported on current-users & tested by Geoff Wing
1.42	25-Nov-2008	pooka	Comment police. No functional or GENERIC size change.
1.41	25-Nov-2008	pooka	When testing if a file system handles file handles (ha ha ha), be content with just the vp-to-fh-size check. (removes a very weird error path in the code)
1.40	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.39	14-Nov-2008	ad	Remove COMPAT ifdefs that might as well be comments (i.e., they cost us almost nothing).
1.38	10-May-2008	rumble	branches: 1.38.4; 1.38.6; Convert file systems to dynamically attach with the new module interface. Make VFS hooks dynamic while we're here and say farewell to VFS_ATTACH and VFS_HOOKS_ATTACH linksets. As a consequence, most of the file systems can now be loaded as new style modules. Quick sanity check by ad@.
1.37	06-May-2008	ad	branches: 1.37.2; PR kern/38141 lookup/vfs_busy acquire rwlock recursively Simplify the mount locking. Remove all the crud to deal with recursion on the mount lock, and crud to deal with unmount as another weirdo lock. Hopefully this will once and for all fix the deadlocks with this. With this commit there are two locks on each mount: - krwlock_t mnt_unmounting. This is used to prevent unmount across critical sections like getnewvnode(). It's only ever read locked with rw_tryenter(), and is only ever write locked in dounmount(). A write hold can't be taken on this lock if the current LWP could hold a vnode lock. - kmutex_t mnt_updating. This is taken by threads updating the mount, for example when going r/o -> r/w, and is only present to serialize updates. In order to take this lock, a read hold must first be taken on mnt_unmounting, and the two need to be held across the operation. One effect of this change: previously if an unmount failed, we would make a half hearted attempt to back out of it gracefully, but that was unlikely to work in a lot of cases. Now while an unmount that will be aborted is in progress, new file operations within the mount will fail instead of being delayed. That is unlikely to be a problem though, because if the admin requests unmount of a file system then s(he) has made a decision to deny access to the resource.
1.36	30-Apr-2008	ad	PR kern/38135 vfs_busy/vfs_trybusy confusion The previous fix worked, but it opened a window where mounts could have disappeared from mountlist while the caller was traversing it using vfs_trybusy(). Fix that.
1.35	29-Apr-2008	ad	kern/38135 vfs_busy/vfs_trybusy confusion The symptom was that sometimes file systems would occasionally not appear in output from 'df' or 'mount' if the system was busy. Resolution: - Make mount locks work somewhat like vm_map locks. - vfs_trybusy() now only fails if the mount is gone, or if someone is unmounting the file system. Simple contention on mnt_lock doesn't cause it to fail. - vfs_busy() will wait even if the file system is being unmounted.
1.34	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.33	28-Feb-2008	elad	branches: 1.33.2; 1.33.4; Introduce a new kauth action, KAUTH_NETWORK_NFS, and two requests, KAUTH_REQ_NETWORK_NFS_EXPORT and KAUTH_REQ_NETWORK_NFS_SVC, and use them to replace two KAUTH_GENERIC_ISSUSER calls in the NFS code. Also replace two more with KAUTH_SYSTEM_MKNOD, where appropriate. Documetnation and examples updated. More to come.
1.32	30-Jan-2008	ad	branches: 1.32.2; 1.32.6; PR kern/37706 (forced unmount of file systems is unsafe): - Do reference counting for 'struct mount'. Each vnode associated with a mount takes a reference, and in turn the mount takes a reference to the vfsops. - Now that mounts are reference counted, replace the overcomplicated mount locking inherited from 4.4BSD with a recursable rwlock.
1.31	08-Dec-2007	pooka	Remove cn_lwp from struct componentname. curlwp should be used from on. The NDINIT() macro no longer takes the lwp parameter and associates the credentials of the calling thread with the namei structure.
1.30	12-Jul-2007	dsl	branches: 1.30.6; 1.30.8; 1.30.14; 1.30.16; Change the VFS_MOUNT() interface so that the 'data' buffer passed to the fs code is a kernel buffer, pass though the length of the buffer as well. Since the length of the userspace buffer isn'it (yet) passed through the mount system call, add a field to the vfsops structure containing the default length. Split sys_mount() for calls from compat code. Ride one of the recent kernel version changes - old fs LKMs will load, but sys_mount() will reject any attempt to use them.
1.29	09-Jul-2007	ad	Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.28	09-Jun-2007	dyoung	There is only one radix trie walker, and it is rn_walktree(), so use that instead of the indirect function call through rnh, rnh->rnh_walktree.
1.27	04-Mar-2007	christos	branches: 1.27.2; 1.27.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.26	05-Feb-2007	chs	branches: 1.26.2; add back a mistakenly removed vput().
1.25	04-Feb-2007	chs	more fixes for the new vnode locking scheme: - don't use SAVESTART in calls to relookup() from unionfs, just vref() the desired vnode when we need to. - fix locking and refcounting in the unionfs EEXIST error cases. - release any vnode locks before calling VFS_ROOT(), vfs_busy() is enough. this allows us to simplify union_root() and fix PR 3006. - union_lock() doesn't handle shared lock requests correctly, so convert them to exclusive instead. fixes PR 34775. - in relookup(), avoid reusing "dp" for different purposes, the error handling wasn't right. (actually just get rid of dp.) also, change relookup() to ignore LOCKLEAF and always return the vnode locked since the callers already expect this.
1.24	04-Jan-2007	elad	Consistent usage of KAUTH_GENERIC_ISSUSER.
1.23	09-Dec-2006	chs	a smorgasbord of improvements to vnode locking and path lookup: - LOCKPARENT is no longer relevant for lookup(), relookup() or VOP_LOOKUP(). these now always return the parent vnode locked. namei() works as before. lookup() and various other paths no longer acquire vnode locks in the wrong order via vrele(). fixes PR 32535. as a nice side effect, path lookup is also up to 25% faster. - the above allows us to get rid of PDIRUNLOCK. - also get rid of WANTPARENT (just use LOCKPARENT and unlock it). - remove an assumption in layer_node_find() that all file systems implement a recursive VOP_LOCK() (unionfs doesn't). - require that all file systems supply vfs_vptofh and vfs_fhtovp routines. fill in eopnotsupp() for file systems that don't support being exported and remove the checks for NULL. (layerfs calls these without checking.) - in union_lookup1(), don't change refcounts in the ISDOTDOT case, just adjust which vnode is locked. fixes PR 33374. - apply fixes for ufs_rename() from ufs_vnops.c rev. 1.61 to ext2fs_rename().
1.22	09-Nov-2006	yamt	branches: 1.22.2; remove some __unused in function parameters.
1.21	31-Oct-2006	mjf	Revert the changes I introduced trying to solve tmpfs' NFS export problem. Requested by yamt@
1.20	30-Oct-2006	jmmv	Fix a typo in a comment.
1.19	24-Oct-2006	mjf	Add support to allow a file system to not permit being exported over NFS. Approved by elad@ and wrstuden@
1.18	22-Oct-2006	pooka	kauth_cred_uucvt() -> kauth_uucred_to_cred(), introduce kauth_cred_to_uucred() per tech-kern proposal
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	23-Jul-2006	ad	branches: 1.16.4; 1.16.6; Use the LWP cached credentials where sane.
1.15	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.14	17-Jun-2006	yamt	branches: 1.14.2; - introduce vfs_composefh() and use it where appropriate. - fix lock/unlock mismatch in sys_getfh.
1.13	18-May-2006	yamt	branches: 1.13.2; 1.13.4; - fix some leaks in nfsd, introduced by kauth changes. - simplify code. - add some assertions. - wrap some long lines. - remove an unnecessary ";".
1.12	18-May-2006	yamt	- nfs_export_unmount: don't forget to free exports. - rename clear_exports for consistency.
1.11	14-May-2006	elad	integrate kauth.
1.10	27-Mar-2006	martin	KASSERT that the returned file id length from VPTOFH is <= the maximum allowed value (_VFS_MAXFIDSZ).
1.9	05-Jan-2006	yamt	branches: 1.9.2; 1.9.4; 1.9.6; 1.9.8; 1.9.10; mountd_set_exports_list: check if VFS_VPTOFH actually works.
1.8	05-Jan-2006	yamt	ensure the export list is not changed during nfsd operations.
1.7	15-Dec-2005	yamt	branches: 1.7.2; netcred_lookup: remove a wrong assertion and handle the case properly.
1.6	11-Dec-2005	christos	merge ktrace-lwp.
1.5	22-Nov-2005	yamt	- reduce number of linear search per rpc. - coalesce mount_netexport_pair into netexport.
1.4	20-Nov-2005	yamt	fix bugs introduced by the recent exports list rototill. - mountd_set_exports_list: don't discard error. - nfs_update_exports_30: check the correct flag.
1.3	25-Sep-2005	jmmv	branches: 1.3.6; 1.3.8; Add some COMPAT_30 code to let old mountd binaries work after the NFS exports rototill.
1.2	23-Sep-2005	jmmv	Remove the mount<->netexport entry from the map during umount... otherwise we leave a dangling pointer in the list ouch.
1.1	23-Sep-2005	jmmv	Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.3.8.1	22-Nov-2005	yamt	sync with head.
1.3.6.3	11-Dec-2005	christos	Sync with head.
1.3.6.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.3.6.1	25-Sep-2005	skrll	file nfs_export.c was added on branch ktrace-lwp on 2005-11-10 14:11:55 +0000
1.7.2.1	15-Jan-2006	yamt	sync with head.
1.9.10.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.9.10.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.9.8.6	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.
1.9.8.5	19-Apr-2006	elad	sync with head.
1.9.8.4	12-Mar-2006	elad	Rename kauth_cred_compare() to kauth_cred_uucmp(), and kauth_cred_convert() to kauth_cred_uucvt(). This makes it clearer that we're working on struct uucred. Inspired by comments from yamt@.
1.9.8.3	10-Mar-2006	elad	Some cleanup. kauth_cred_setrefcnt() was only called after kauth_cred_convert() in NFS code to convert a struct uucred to kauth_cred_t. Since there's no valid use for such a function, make kauth_cred_convert() set the reference count to 1 and eliminate the need for kauth_cred_setrefcnt() entirely. Motivated by comments from yamt@ and thorpej@.
1.9.8.2	10-Mar-2006	elad	generic_authorize() -> kauth_authorize_generic().
1.9.8.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.9.6.4	11-Aug-2006	yamt	sync with head
1.9.6.3	26-Jun-2006	yamt	sync with head.
1.9.6.2	24-May-2006	yamt	sync with head.
1.9.6.1	01-Apr-2006	yamt	sync with head.
1.9.4.2	01-Jun-2006	kardel	Sync with head.
1.9.4.1	22-Apr-2006	simonb	Sync with head.
1.9.2.1	09-Sep-2006	rpaulo	sync with head
1.13.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.13.2.1	19-Jun-2006	chap	Sync with head.
1.14.2.8	17-Mar-2008	yamt	sync with head.
1.14.2.7	04-Feb-2008	yamt	sync with head.
1.14.2.6	21-Jan-2008	yamt	sync with head
1.14.2.5	03-Sep-2007	yamt	sync with head.
1.14.2.4	26-Feb-2007	yamt	sync with head.
1.14.2.3	30-Dec-2006	yamt	sync with head.
1.14.2.2	21-Jun-2006	yamt	sync with head.
1.14.2.1	17-Jun-2006	yamt	file nfs_export.c was added on branch yamt-lazymbuf on 2006-06-21 15:11:58 +0000
1.16.6.2	10-Dec-2006	yamt	sync with head.
1.16.6.1	22-Oct-2006	yamt	sync with head
1.16.4.3	09-Feb-2007	ad	Sync with HEAD.
1.16.4.2	12-Jan-2007	ad	Sync with head.
1.16.4.1	18-Nov-2006	ad	Sync with head.
1.22.2.1	17-Feb-2007	tron	Apply patch (requested by chs in ticket #422): - Fix various deadlock problems with nullfs and unionfs. - Speed up path lookups by upto 25%.
1.26.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.27.4.1	11-Jul-2007	mjf	Sync with head.
1.27.2.2	15-Jul-2007	ad	Sync with head.
1.27.2.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.30.16.1	26-Dec-2007	ad	Sync with head.
1.30.14.1	18-Feb-2008	mjf	Sync with HEAD.
1.30.8.2	23-Mar-2008	matt	sync with HEAD
1.30.8.1	09-Jan-2008	matt	sync with HEAD
1.30.6.1	09-Dec-2007	jmcneill	Sync with HEAD.
1.32.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.32.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.32.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.32.2.1	24-Mar-2008	keiichi	sync with head.
1.33.4.4	18-Jul-2009	yamt	sync with head.
1.33.4.3	20-Jun-2009	yamt	sync with head
1.33.4.2	04-May-2009	yamt	sync with head.
1.33.4.1	16-May-2008	yamt	sync with head.
1.33.2.1	18-May-2008	yamt	sync with head.
1.37.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.38.6.1	19-Jan-2009	skrll	Sync with HEAD.
1.38.4.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.44.2.1	23-Jul-2009	jym	Sync with HEAD.
1.48.4.2	21-Apr-2011	rmind	sync with head
1.48.4.1	05-Mar-2011	rmind	sync with head
1.49.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.51.16.1	18-May-2014	rmind	sync with head
1.51.12.2	03-Dec-2017	jdolecek	update from HEAD
1.51.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.51.2.1	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")
1.58.10.2	26-Apr-2017	pgoyette	Sync with HEAD
1.58.10.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.58.6.2	28-Aug-2017	skrll	Sync with HEAD
1.58.6.1	05-Dec-2016	skrll	Sync with HEAD
1.59.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.60.12.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.61.2.1	17-Jan-2020	ad	Sync with head.
1.62.14.1	06-Jun-2021	cjep	sync with head
1.62.10.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.9	21-Mar-2023	christos	PR/57279: Izumi Tsutsui: Fix some {int,long} -> time_t. Still things will break eventually because parts of the nfs protocol assume time_t will fit in 32 bits.
1.8	03-Sep-2018	riastradh	branches: 1.8.30; 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!)
1.7	15-Jul-2015	manu	branches: 1.7.16; 1.7.18; Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.6	25-Oct-2013	martin	branches: 1.6.4; 1.6.6; Mark a diagnostic-only variable
1.5	14-Sep-2013	martin	Avoid unused variable warnings
1.4	31-Dec-2009	christos	branches: 1.4.12; 1.4.22; 1.4.26; handle the nuidhash_max lossage differently
1.3	14-Mar-2009	dsl	branches: 1.3.2; ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.2	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.1	19-Nov-2008	ad	branches: 1.1.4; 1.1.6; 1.1.8; 1.1.10; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.1.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1.8.3	28-Apr-2009	skrll	Sync with HEAD.
1.1.8.2	19-Jan-2009	skrll	Sync with HEAD.
1.1.8.1	19-Nov-2008	skrll	file nfs_iod.c was added on branch nick-hppapmap on 2009-01-19 13:20:20 +0000
1.1.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.1.6.1	19-Nov-2008	mjf	file nfs_iod.c was added on branch mjf-devfs2 on 2009-01-17 13:29:34 +0000
1.1.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.1.4.1	19-Nov-2008	haad	file nfs_iod.c was added on branch haad-dm on 2008-12-13 01:15:28 +0000
1.3.2.4	11-Mar-2010	yamt	sync with head
1.3.2.3	16-Jul-2009	yamt	fix a merge botch.
1.3.2.2	04-May-2009	yamt	sync with head.
1.3.2.1	14-Mar-2009	yamt	file nfs_iod.c was added on branch yamt-nfs-mp on 2009-05-04 08:14:22 +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	22-Sep-2015	skrll	Sync with HEAD
1.6.4.1	04-Nov-2015	riz	Pull up following revision(s) (requested by manu in ticket #882): sbin/umount/umount.c: revision 1.48 sys/nfs/nfsmount.h: revision 1.53 sys/nfs/nfs_var.h: revision 1.94 sys/nfs/nfs_iod.c: revision 1.7 sys/nfs/nfs_socket.c: revision 1.197 sys/nfs/nfs_bio.c: revision 1.191 sys/nfs/nfs_vfsops.c: revision 1.230 sys/nfs/nfs_clntsocket.c: revision 1.3 Remove useless and harmful sync(2) call in umount(8) Remove sync(2) call before unmount(2) in umount(8). This sync(2) is useless since unmount(2) will perform a VFS_SYNC anyway. But moreover, this sync(2) may be harmful, as there are some situation where it cannot return (unreachable NFS server, for instance), causing umount -f to be uneffective. Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.7.18.1	10-Jun-2019	christos	Sync with HEAD
1.7.16.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.8.30.1	20-Sep-2024	martin	Pull up following revision(s) (requested by rin in ticket #880): sys/nfs/nfs_iod.c: revision 1.9 sys/nfs/nfs_vfsops.c: revision 1.245 sys/nfs/nfs_clntsubs.c: revision 1.7 PR/57279: Izumi Tsutsui: Fix some {int,long} -> time_t. Still things will break eventually because parts of the nfs protocol assume time_t will fit in 32 bits.
1.32	20-Oct-2021	thorpej	Overhaul of the EVFILT_VNODE kevent(2) filter: - Centralize vnode kevent handling in the VOP_() wrappers, rather than forcing each individual file system to deal with it (except VOP_RENAME(), because VOP_RENAME() is a mess and we currently have 2 different ways of handling it; at least it's reasonably well-centralized in the "new" way). - Add support for NOTE_OPEN, NOTE_CLOSE, NOTE_CLOSE_WRITE, and NOTE_READ, compatible with the same events in FreeBSD. - Track which kevent notifications clients are interested in receiving to avoid doing work for events no one cares about (avoiding, e.g. taking locks and traversing the klist to send a NOTE_WRITE when someone is merely watching for a file to be deleted, for example). In support of the above: - Add support in vnode_if.sh for specifying PRE- and POST-op handlers, to be invoked before and after vop_pre() and vop_post(), respectively. Basic idea from FreeBSD, but implemented differently. - Add support in vnode_if.sh for specifying CONTEXT fields in the vop__args structures. These context fields are used to convey information between the file system VOP function and the VOP wrapper, but do not occupy an argument slot in the VOP_*() call itself. These context fields are initialized and subsequently interpreted by PRE- and POST-op handlers. - Version VOP_REMOVE(), uses the a context field for the file system to report back the resulting link count of the target vnode. Return this in tmpfs, udf, nfs, chfs, ext2fs, lfs, and ufs. NetBSD 9.99.92.
1.31	11-Oct-2021	thorpej	Mark the EVFILT_VNODE filters MP-safe.
1.30	11-Oct-2021	thorpej	Setting EV_EOF requires modifying kn->kn_flags. However, that relies on holding the kq_lock of that note's kq. Rather than exposing this directly, add new knote_set_eof() and knote_clear_eof() functions that handle the necessary locking and don't leak as many implementation details to modules. NetBSD 9.99.91
1.29	10-Oct-2021	thorpej	Must hold kn->kn_kq->kq_lock to modify kn->kn_flags.
1.28	26-Sep-2021	thorpej	Change the kqueue filterops::f_isfd field to filterops::f_flags, and define a flag FILTEROP_ISFD that has the meaning of the prior f_isfd. Field and flag name aligned with OpenBSD. This does not constitute a functional or ABI change, as the field location and size, and the value placed in that field, are the same as the previous code, but we're bumping __NetBSD_Version__ so 3rd-party module source code can adapt, as needed. NetBSD 9.99.89
1.27	05-Sep-2020	riastradh	Round of uvm.h cleanup. The poorly named uvm.h is generally supposed to be for uvm-internal users only. - Narrow it to files that actually need it -- mostly files that need to query whether curlwp is the pagedaemon, which should maybe be exposed by an external header. - Use uvm_extern.h where feasible and uvm_*.h for things not exposed by it. We should split up uvm_extern.h but this will serve for now to reduce the uvm.h dependencies. - Use uvm_stat.h and #ifdef UVMHIST uvm.h for files that use UVMHIST(ubchist), since ubchist is declared in uvm.h but the reference evaporates if UVMHIST is not defined, so we reduce header file dependencies. - Make uvm_device.h and uvm_swap.h independently includable while here. ok chs@
1.26	25-Oct-2017	maya	Use C99 initializer for filterops Mostly done with spatch with touchups for indentation @@ expression a; identifier b,c,d; identifier p; @@ const struct filterops p = - { a, b, c, d + { + .f_isfd = a, + .f_attach = b, + .f_detach = c, + .f_event = d, };
1.25	24-Oct-2011	hannken	branches: 1.25.12; VOP_GETATTR() needs a shared lock at least. As nfs_kqpoll() ignores the return value from VOP_GETATTR() initialize the attrributes to zero -- nfs_kqfilter() does the same.
1.24	12-Jun-2011	rmind	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.
1.23	19-Nov-2008	ad	branches: 1.23.8; 1.23.14; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.22	28-Apr-2008	martin	branches: 1.22.6; 1.22.8; Remove clause 3 and 4 from TNF licenses
1.21	21-Mar-2008	ad	branches: 1.21.2; 1.21.4; Catch up with descriptor handling changes. See kern_descrip.c revision 1.173 for details.
1.20	12-Feb-2008	yamt	branches: 1.20.6; nfs_kqfilter: fix lock/unlock mismatch. PR/38003 from Geoff C. Wing.
1.19	05-Feb-2008	ad	Lock v_knlist with the vnode interlock. PR kern/37881.
1.18	02-Jan-2008	yamt	use kmem_alloc instead of malloc.
1.17	05-Dec-2007	pooka	branches: 1.17.4; Do not "return 1" from kqfilter for errors. That value is passed directly to the userland caller and results in a mysterious EPERM. Instead, return EINVAL or something else sensible depending on the case.
1.16	26-Nov-2007	pooka	branches: 1.16.2; 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
1.15	09-Jul-2007	ad	branches: 1.15.6; 1.15.8; 1.15.14; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.14	29-Apr-2007	yamt	use mutex and condvar.
1.13	09-Nov-2006	yamt	branches: 1.13.4; 1.13.8; 1.13.10; remove some __unused in function parameters.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	23-Jul-2006	ad	branches: 1.11.4; 1.11.6; Use the LWP cached credentials where sane.
1.10	14-May-2006	elad	integrate kauth.
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.6; 1.9.8; 1.9.10; 1.9.12; merge ktrace-lwp.
1.8	26-Feb-2005	perry	branches: 1.8.4; nuke trailing whitespace
1.7	30-Oct-2003	simonb	branches: 1.7.8; 1.7.10; Remove some assigned-to but otherwise unused variables.
1.6	29-Jun-2003	fvdl	branches: 1.6.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.5	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
1.4	27-Mar-2003	jdolecek	nfs_kqfilter(): add the knote to v_klist only if guaranteed to return success
1.3	27-Feb-2003	jdolecek	fix typo in comment
1.2	23-Oct-2002	jdolecek	branches: 1.2.2; 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
1.1	30-Sep-2002	jdolecek	branches: 1.1.2; file nfs_kq.c was initially added on branch kqueue.
1.1.2.2	02-Oct-2002	jdolecek	knote data is now 64bit, g/c obsolete comment
1.1.2.1	30-Sep-2002	jdolecek	add support for kevents to NFS to detect file changes on server by other NFS clients, polling kernel thread is used to periodically check for attribute changes of watched files; the NFS server is only contacted when the vnode expires from local attrcache (which takes 5-60 seconds currently), to keep network&CPU overhead low the routine checking for remote changes is quite simplistic, but hopefully doing it's job well enough
1.2.2.2	11-Nov-2002	nathanw	Catch up to -current
1.2.2.1	23-Oct-2002	nathanw	file nfs_kq.c was added on branch nathanw_sa on 2002-11-11 22:16:08 +0000
1.6.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.6.2.2	03-Aug-2004	skrll	Sync with HEAD
1.6.2.1	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.
1.7.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.8.1	29-Apr-2005	kent	sync with -current
1.8.4.8	24-Mar-2008	yamt	sync with head.
1.8.4.7	27-Feb-2008	yamt	sync with head.
1.8.4.6	11-Feb-2008	yamt	sync with head.
1.8.4.5	21-Jan-2008	yamt	sync with head
1.8.4.4	07-Dec-2007	yamt	sync with head
1.8.4.3	03-Sep-2007	yamt	sync with head.
1.8.4.2	30-Dec-2006	yamt	sync with head.
1.8.4.1	21-Jun-2006	yamt	sync with head.
1.9.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.9.10.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.9.8.2	11-Aug-2006	yamt	sync with head
1.9.8.1	24-May-2006	yamt	sync with head.
1.9.6.1	01-Jun-2006	kardel	Sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.11.6.2	10-Dec-2006	yamt	sync with head.
1.11.6.1	22-Oct-2006	yamt	sync with head
1.11.4.1	18-Nov-2006	ad	Sync with head.
1.13.10.1	11-Jul-2007	mjf	Sync with head.
1.13.8.4	08-Jun-2007	ad	Sync with head.
1.13.8.3	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.
1.13.8.2	10-Apr-2007	ad	Nuke the deferred kthread creation stuff, as it's no longer needed. Pointed out by thorpej@.
1.13.8.1	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.
1.13.4.1	07-May-2007	yamt	sync with head.
1.15.14.2	18-Feb-2008	mjf	Sync with HEAD.
1.15.14.1	08-Dec-2007	mjf	Sync with HEAD.
1.15.8.2	23-Mar-2008	matt	sync with HEAD
1.15.8.1	09-Jan-2008	matt	sync with HEAD
1.15.6.2	09-Dec-2007	jmcneill	Sync with HEAD.
1.15.6.1	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.16.2.1	08-Dec-2007	ad	Sync with head.
1.17.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.20.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.20.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.20.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.21.4.5	11-Aug-2010	yamt	sync with head.
1.21.4.4	21-Mar-2010	yamt	lock vnode when calling VOP_GETATTR.
1.21.4.3	04-May-2009	yamt	sync with head.
1.21.4.2	16-May-2008	yamt	sync with head.
1.21.4.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.21.2.1	18-May-2008	yamt	sync with head.
1.22.8.1	19-Jan-2009	skrll	Sync with HEAD.
1.22.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.23.14.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.23.8.1	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.
1.25.12.1	03-Dec-2017	jdolecek	update from HEAD
1.126	01-May-2020	hannken	Resolve delayed truncation from nfs_inactive() too. Should prevent "locking against self" from nfs_unlock().
1.125	24-Feb-2020	ad	v_interlock -> vmobjlock
1.124	18-Oct-2019	msaitoh	branches: 1.124.2; s/initalize/initialize/ in comment or printf message.
1.123	28-May-2018	chs	branches: 1.123.2; add a genfs method to allow a file system to limit the range of pages that are given to a single GOP_WRITE() call. needed by ZFS.
1.122	26-May-2017	riastradh	branches: 1.122.8; Eliminate crusty debugging sludge. We have a mostly sane vnode lifecycle now. If this needs debugging, it should be done once at the call site of VOP_RECLAIM.
1.121	26-May-2017	riastradh	Make VOP_RECLAIM do the last unlock of the vnode. VOP_RECLAIM naturally has exclusive access to the vnode, so having it locked on entry is not strictly necessary -- but it means if there are any final operations that must be done on the vnode, such as ffs_update, requiring exclusive access to it, we can now kassert that the vnode is locked in those operations. We can't just have the caller release the last lock because some file systems don't use genfs_lock, and require the vnode to remain valid for VOP_UNLOCK to work, notably unionfs.
1.120	11-Apr-2017	riastradh	Make VOP_INACTIVE preserve vnode lock on return. Discussed on tech-kern: https://mail-index.netbsd.org/tech-kern/2017/04/01/msg021751.html Ride 7.99.68, a bumpy bus of incremental vfs improvements!
1.119	20-Aug-2016	hannken	branches: 1.119.2; Remove now obsolete operation vcache_remove(). Welcome to 7.99.36
1.118	30-May-2014	hannken	branches: 1.118.4; 1.118.8; Change NFS from rbtree to vcache.
1.117	27-Feb-2014	hannken	branches: 1.117.2; The current implementation of vn_lock() is racy. Modification of the vnode operations vector for active vnodes is unsafe because it is not known whether deadfs or the original file system will be called. - Pass down LK_RETRY to the lock operation (hint for deadfs only). - Change deadfs lock operation to return ENOENT if LK_RETRY is unset. - Change all other lock operations to check for dead vnode once the vnode is locked and unlock and return ENOENT in this case. With these changes in place vnode lock operations will never succeed after vclean() has marked the vnode as VI_XLOCK and before vclean() has changed the operations vector. Adresses PR kern/37706 (Forced unmount of file systems is unsafe) Discussed on tech-kern. Welcome to 6.99.33
1.116	12-Jun-2011	rmind	branches: 1.116.2; 1.116.12; 1.116.16; 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.
1.115	19-May-2011	rmind	branches: 1.115.2; Remove cache_purge(9) calls from reclamation routines in the file systems, as vclean(9) performs it for us since Lite2 merge.
1.114	24-Sep-2010	rmind	branches: 1.114.2; Fixes/improvements to RB-tree implementation: 1. Fix inverted node order, so that negative value from comparison operator would represent lower (left) node, and positive - higher (right) node. 2. Add an argument (i.e. "context"), passed to comparison operators. 3. Change rb_tree_insert_node() to return a node - either inserted one or already existing one. 4. Amend the interface to manipulate the actual object, instead of the rb_node (in a similar way as Patricia-tree interface does). 5. Update all RB-tree users accordingly. XXX: Perhaps rename rb.h to rbtree.h, since cleaning-up.. 1-3 address the PR/43488 by Jeremy Huddleston. Passes RB-tree regression tests. Reviewed by: matt@, christos@
1.113	21-Jul-2010	hannken	Make holding v_interlock mandatory for callers of vget(). Announced some time ago on tech-kern.
1.112	01-Jul-2010	hannken	Remove vlockmgr(). Generic vnode lock operations now use a rwlock located in the vnode. All LK_* flags move from sys/lock.h to sys/vnode.h. Calls to vlockmgr() in file systems get replaced with VOP_LOCK() or VOP_UNLOCK(). Welcome to 5.99.34. Discussed on tech-kern.
1.111	24-Jun-2010	hannken	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.
1.110	15-Mar-2009	cegger	branches: 1.110.2; 1.110.4; ansify function definitions
1.109	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.108	02-Jan-2009	ad	branches: 1.108.2; - Don't vput() a vnode that we do not hold locked. - Eliminate one of the few remaining uses of LK_CANRECURSE.
1.107	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.106	22-Oct-2008	matt	branches: 1.106.2; 1.106.4; Don't need nfs_vfs_reinit anymore since we don't resize tables anymore. Move reinit code to init case.
1.105	22-Oct-2008	matt	Change NFS to use a RB-tree for its FH->nfsnode lookups.
1.104	30-Sep-2008	pooka	Initialize nfsnode pools and malloc type dynamically in the constructor instead of depending on link sets. Consequently, rename nfs_nh{init,reinit,done} to nfs_node_{init,reinit,done}, respectively, to better convey the function.
1.103	24-May-2008	tron	branches: 1.103.4; Make sure that we flush the NFS directory cache in case of an NFS mount using the translate cookie option during unmount. This fixes PR kern/38100. Patch suggested by Michael van Elst during Hackathon 11.
1.102	05-May-2008	ad	branches: 1.102.2; - Convert hashinit() to use kmem_alloc(). The hash tables can be large and it's better to not have them in kmem_map. - Convert a couple of minor items along the way to kmem_alloc(). - Fix some memory leaks.
1.101	30-Jan-2008	ad	branches: 1.101.6; 1.101.8; 1.101.10; Replace struct lock on vnodes with a simpler lock object built on krwlock_t. This is a step towards removing lockmgr and simplifying vnode locking. Discussed on tech-kern.
1.100	26-Jan-2008	ad	- Make nfsnode hash MPSAFE. - Replace use of lockmgr().
1.99	17-Jan-2008	ad	Correct test of v_usecount.
1.98	02-Jan-2008	yamt	use kmem_alloc instead of malloc.
1.97	02-Jan-2008	ad	Merge vmlocking2 to head.
1.96	26-Nov-2007	pooka	branches: 1.96.2; 1.96.6; 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
1.95	06-Aug-2007	yamt	branches: 1.95.2; 1.95.8; 1.95.10; nfs_inactive: turn a panic into a printf for now, as it isn't critical. PR/36572 from Martin Husemann.
1.94	12-Jun-2007	yamt	branches: 1.94.2; 1.94.6; nfs_inactive: don't clear NTRUNCDELAYED erroneously. (fix cache consistency problems like NUL bytes near EOF.)
1.93	12-Mar-2007	ad	branches: 1.93.2; Pass an ipl argument to pool_init/POOL_INIT to be used when initializing the pool's lock.
1.92	21-Feb-2007	thorpej	branches: 1.92.4; Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.91	20-Feb-2007	ad	Call genfs_node_destroy() where appropriate.
1.90	15-Feb-2007	yamt	branches: 1.90.2; use mutex and rwlock rather than lockmgr.
1.89	28-Dec-2006	yamt	remove several nqnfs definitions.
1.88	27-Dec-2006	yamt	remove nqnfs.
1.87	09-Nov-2006	yamt	remove some __unused in function parameters.
1.86	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.85	23-Jul-2006	ad	branches: 1.85.4; 1.85.6; Use the LWP cached credentials where sane.
1.84	14-May-2006	elad	integrate kauth.
1.83	30-Mar-2006	yamt	some cleanups after the introduction of GOP_SIZE_MEM flag. - remove GOP_SIZE_READ/GOP_SIZE_WRITE flags. they have not been used since the change. - ufs_balloc_range: remove code which has been no-op since the change. thanks Konrad Schroder for explaining the original intention of the code. - ffs_gop_size: don't extend past eof, in the case of GOP_SIZE_MEM. otherwise genfs_getpages end up to allocate pages past eof unnecessarily.
1.82	02-Jan-2006	yamt	branches: 1.82.2; 1.82.4; 1.82.6; 1.82.8; 1.82.10; nfs_inactive: - use LK_CANRECURSE instead of LK_RECURSEFAIL. PR/32435 from Valeriy E. Ushakov. - panic explicitly if the parent directory has been revoked. add an XXX comment.
1.81	11-Dec-2005	christos	branches: 1.81.2; merge ktrace-lwp.
1.80	28-Jun-2005	yamt	branches: 1.80.2; - constify genfs_ops. - use member designators.
1.79	26-Feb-2005	perry	branches: 1.79.2; nuke trailing whitespace
1.78	27-Jan-2005	yamt	keep directory eof cache when inactivating vnode because there's no reason to throw it away. (fix an unintended side effect of nfs_subs.c rev.1.144.)
1.77	25-Apr-2004	simonb	branches: 1.77.4; 1.77.6; Initialise (most) pools from a link set instead of explicit calls to pool_init. Untouched pools are ones that either in arch-specific code, or aren't initialiased during initial system startup. Convert struct session, ucred and lockf to pools.
1.76	20-Apr-2004	yamt	nfs_inactive: inactive the vp before doing sillyrename works. vp can be reclaimed soon after it's unlocked.
1.75	05-Apr-2004	yamt	nfs_readdirplusrpc: fix a deadlock problem. don't wait for vnode lock to load attributes. otherwise, because READDIRPLUS returns DOTDOT entry as well, we violate locking order.
1.74	05-Apr-2004	yamt	don't issue VOP_GETATTR blindly in nfs_nget(). in many cases, GETATTR RPCs here is redundant because the caller has postop_attr. instead, make sure the resulted vnode have a valid attribute in nfs_lookup().
1.73	12-Mar-2004	yamt	branches: 1.73.2; shrink sizeof struct nfsnode by putting exclusive members into union.
1.72	23-Jan-2004	wrstuden	Adjust sillyrename cleanup code to deal with the parent vnode already being locked by our thread. VOP_INACTIVATE() makes no statement as to the lock state of the parent, yet this code assumed we had it unlocked. With this change, we let vn_lock() fail with EDEADLK if we already have the parent locked. We then handle the rename cleanup, and on the way out just vrele() the parent vnode, not vput() it. Fixes a case seen by Steve Woodford at Wasabisystems dot com where we'd panic while running a pkgsrc configure test that verified fork() functionality. I expect the problem is a result of the recent exit() changes and the performance of the machines he tested on. Specifically we would crash during an nfs_remove(). As best I can tell, when nfs_remove() tested to see if we should rename or we should remove, v_usecount was > 1 and vattr.va_nlink was 1. Thus we did the sillyrename in nfs_remove(). However by the time we got down to the vput(vp), v_usecount had dropped to one and thus vput() triggered the VOP_INACTIVATE() code path. nfs_inactive() tries to lock the parent to undo the sillyrename, and deadlocks as we still have it locked.
1.71	07-Dec-2003	fvdl	Unix semantics dictate that access checks for files are done when it is opened. An open file can always be read from and/or written to, depending on how it was opened. Therefore, the read/write/commit RPCs should never return EACCESS, as they are only performed on files that have been successfully opened already. This change improves the current situation and works in most cases. It simply always uses the most recently known owner/group of the file, iff the authentication mechanism is AUTH_UNIX (in other cases, the creds for a succesful open are used, but note that no other cases are currently implemented). A retry mechanism can be used to catch a few more cases, but this is a good improvement for now.
1.70	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.69	30-Jul-2003	yamt	vrecycle removed nfs vnodes. not perfect, but enough for most cases.
1.68	29-Jun-2003	fvdl	branches: 1.68.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.67	29-Jun-2003	thorpej	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.66	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
1.65	22-May-2003	yamt	avoid double free with xlatecookie.
1.64	21-May-2003	yamt	remove local definitions of TRUE and FALSE.
1.63	07-May-2003	yamt	use hashdone to free hashinit'ed memory.
1.62	02-Apr-2003	yamt	use queue manipulation macros.
1.61	17-Feb-2003	perseant	Add code to UBCify LFS. This is still behind "#ifdef LFS_UBC" for now (there are still some details to work out) but expect that to go away soon. To support these basic changes (creation of lfs_putpages, lfs_gop_write, mods to lfs_balloc) several other changes were made, to wit: * Create a writer daemon kernel thread whose purpose is to handle page writes for the pagedaemon, but which also takes over some of the functions of lfs_check(). This thread is started the first time an LFS is mounted. * Add a "flags" parameter to GOP_SIZE. Current values are GOP_SIZE_READ, meaning that the call should return the size of the in-core version of the file, and GOP_SIZE_WRITE, meaning that it should return the on-disk size. One of GOP_SIZE_READ or GOP_SIZE_WRITE must be specified. * Instead of using malloc(...M_WAITOK) for everything, reserve enough resources to get by and use malloc(...M_NOWAIT), using the reserves if necessary. Use the pool subsystem for structures small enough that this is feasible. This also obsoletes LFS_THROTTLE. And a few that are not strictly necessary: * Moves the LFS inode extensions off onto a separately allocated structure; getting closer to LFS as an LKM. "Welcome to 1.6O." * Unified GOP_ALLOC between FFS and LFS. * Update LFS copyright headers to correct values. * Actually cast to unsigned in lfs_shellsort, like the comment says. * Keep track of which segments were empty before the previous checkpoint; any segments that pass two checkpoints both dirty and empty can be summarily cleaned. Do this. Right now lfs_segclean still works, but this should be turned into an effectless compatibility syscall.
1.60	15-Feb-2003	drochner	Don't remove the nfsnode from the hash chain in nfs_inactive. It will never get back... it will not be found in nfs_nget, a new nfsnode+vnode is allocated instead, which causes a node leak, and also makes the mountpointness of the vnode to be forgotten, breaking filesystem crossing lookups through this vnode.
1.59	12-Feb-2003	fvdl	Move purging the dircache and removing a vnode from the nqnfs timer queue into nfs_inactive, this is a better place for it. This doesn't actually solve the actual problem, which appears to be a race condition with unmounting and vnode recycling somewhere, but it fixes it in the sense that nfs_reclaim will not reference a bad v_mount anymore.
1.58	10-Feb-2003	christos	move the MALLOC decl for DIROFFS to nfs_subs.c
1.57	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.56	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.55	01-Oct-2002	christos	back out previous. the problem is not the v_mount is null, but it appears corrupted.
1.54	30-Sep-2002	christos	deal with v_mount == NULL in nfs_reclaim(). We should not be touching this anyway, but nq-nfs wants us to.
1.53	16-Mar-2002	chs	branches: 1.53.6; make sure that if NMODIFIED is clear, all pages attached to the vnode are clean and without writable mappings. if we try to flush dirty pages past EOF to the server when NMODIFIED is clear, we'll update the attrcache before doing the write, which will try to free the pages past EOF and deadlock. to deal with this, we write-protect pages before we send them to the server, and restrict ourselves to creating read-only mappings if NMODIFIED isn't set. score another one for enami.
1.52	08-Mar-2002	thorpej	Pool deals fairly well with physical memory shortage, but it doesn't deal with shortages of the VM maps where the backing pages are mapped (usually kmem_map). Try to deal with this: * Group all information about the backend allocator for a pool in a separate structure. The pool references this structure, rather than the individual fields. * Change the pool_init() API accordingly, and adjust all callers. * Link all pools using the same backend allocator on a list. * The backend allocator is responsible for waiting for physical memory to become available, but will still fail if it cannot callocate KVA space for the pages. If this happens, carefully drain all pools using the same backend allocator, so that some KVA space can be freed. * Change pool_reclaim() to indicate if it actually succeeded in freeing some pages, and use that information to make draining easier and more efficient. * Get rid of PR_URGENT. There was only one use of it, and it could be dealt with by the caller. From art@openbsd.org.
1.51	26-Jan-2002	chs	re-enable NFSv3 commit RPCs by abandoning my new approach in favor of frank's scheme, with one new twist: don't wait until we've totally run out of free pages before committing, but instead notice when we've built up a largish range of uncommitted pages and commit only the older half of the range, which is likely to already be on disk on the server.
1.50	21-Jan-2002	fvdl	VOP_UNLOCK + vgone --> vput, since the vnode will already have a reference.
1.49	18-Jan-2002	fvdl	Unlock vnode before calling vgone() in case of getattr failure during nfs_nget. Fixes problem reported by Chuck Cranor.
1.48	06-Dec-2001	lukem	Replace nfs_hash() (with its extremely bad hash) with a macro to call hash32_buf() to obtain a 32 bit hash. On some tests I ran I obtained a 30x improvement in hash distribution and a 6x reduction in time spent in nfs_nget().
1.47	10-Nov-2001	lukem	add RCSIDs
1.46	15-Sep-2001	chs	branches: 1.46.2; a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
1.45	15-Sep-2001	chs	add a new VFS op, vfs_reinit, which is called when desiredvnodes is adjusted via sysctl. file systems that have hash tables which are sized based on the value of this variable now resize those hash tables using the new value. the max number of FFS softdeps is also recalculated. convert various file systems to use the <sys/queue.h> macros for their hash tables.
1.44	03-May-2001	fvdl	branches: 1.44.2; 1.44.4; Drop vnode lock before removing the sillyrename file, to avoid a lock-o-death.
1.43	20-Apr-2001	fvdl	On VOP_GETATTR failure in nfs_nget, call vgone() to get rid of the vnode that was just created. Suggested by Enami.
1.42	20-Apr-2001	fvdl	Unlock the hash lock before returning an error in nfs_nget. From IWAMOTO Toshihiro.
1.41	07-Feb-2001	tsutsui	branches: 1.41.2; Fix nested extern declaration of prtactive.
1.40	06-Feb-2001	fvdl	In nfs_inactive there's no need anymore for an extra refcount around nfs_vinvalbuf, since it has a real lock on the vnode now, so getnewvnode will not hijack it.
1.39	06-Feb-2001	fvdl	Do actual vnode locking for NFS.
1.38	27-Nov-2000	chs	Initial integration of the Unified Buffer Cache project.
1.37	08-Nov-2000	ad	Update for hashinit() change.
1.36	19-Sep-2000	fvdl	Initialize the lock needed to serialize commits for one NFS node.
1.35	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.34	03-Aug-2000	thorpej	MALLOC()/FREE() are not to be used for variable size allocations.
1.33	30-Mar-2000	augustss	branches: 1.33.4; Remove register declarations.
1.32	30-Mar-2000	simonb	Delete redundant decl of nfsv2_vnodeop_p, it's in <nfs/nfsnode.h>.
1.31	16-Mar-2000	jdolecek	Add new VFS op routine - vfs_done and call it on filesystem detach in vfs_detach(). vfs_done may free global filesystem's resources, typically those allocated in respective filesystem's init function. Needed so those filesystems which went in via LKM have a chance to clean after themselves before unloading. This fixes random panics when LKM for filesystem using pools was loaded and unloaded several times. For each leaf filesystem, add appropriate vfs_done routine.
1.30	29-Nov-1999	fvdl	Insert an extra VOP_ACCESS check in nfs_lookup, to avoid cached access mishaps for lookup and getattr. Closes PR 8884. While at it, cache access RPCs.
1.29	08-Jul-1999	wrstuden	branches: 1.29.2; 1.29.8; Modify file systems to deal with struct lock in struct vnode. All leaf fs's other than nfs use genfs_lock() for locking. Modify lookup routines to set PDIRUNLOCK when they unlock the parrent.
1.28	01-Sep-1998	thorpej	branches: 1.28.2; 1.28.6; 1.28.8; Use the pool allocator and the "nointr" pool page allocator for NFS nodes and vattr structures.
1.27	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.26	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.25	07-Feb-1998	chs	add flags arg to hashinit(), to pass to malloc().
1.24	19-Oct-1997	fvdl	* Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.23	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.22	07-Jul-1997	fvdl	branches: 1.22.2; Do locking around nfsnode hashing (perhaps even right this time!)
1.21	07-Jul-1997	fvdl	Revert until I have time to fix it today (lock applied wrongly).
1.20	06-Jul-1997	fvdl	Put lock around nfs node hashing to avoid race conditions, as MALLOC or getnewvnode may block.
1.19	22-Feb-1997	fvdl	Fixes from BSDI (thanks go to Keith Bostic). Original RCS messages: date: 1996/09/06 03:00:31; author: donn; state: Exp; lines: +1 -2 Because NFS doesn't implement vnode locking, nfs_inactive() doesn't really have the vnode locked and hence it can't reliably access the vnode after it performs a blocking operation. We remove one blocking call and push the no-op VOP_UNLOCK higher so that we don't access the vnode after we delete the sillyrename file. This should prevent crashes we've seen in which the vnode turned into a UFS vnode and caused a panic in ufs_unlock() when we tried to 'unlock' it. date: 1996/09/25 19:15:21; author: cp; state: Exp; lines: +4 -0 Kirk's change to not corrupt files after a delete. date: 1996/11/08 19:53:45; author: donn; state: Exp; lines: +16 -4 Krik's change to solve the paradox that vclean() calls nfs_inactive() with VXLOCK set on the vnode, and nfs_inactive() was calling vget() to get a reference on the vnode, which in turn hung on VXLOCK. Nfs_inactive() now checks v_usecount to make sure that the vnode is not coming from vclean() before it does a vget().
1.18	12-Feb-1997	fvdl	Don't set sillyrename field to 0 for directories, as it's in a union with the head of the cookie list. Fixes PR 3215, fix supplied by Hiroshi Tezuka <tezuka@trc.rwcp.or.jp>. Should also fix M_NFSDIROFF memory leak.
1.17	01-Sep-1996	mycroft	branches: 1.17.4; Add a set of generic file system operations that most file systems use. Also, fix some time stamp bogosities.
1.16	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.15	09-Feb-1996	christos	nfs prototype changes
1.14	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.13	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.12	29-Jun-1994	cgd	branches: 1.12.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.11	13-Jun-1994	mycroft	Undo last change.
1.10	13-Jun-1994	gwr	Fix unresolved: prtactive
1.9	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.8	25-Apr-1994	cgd	some prototype cleanup, eliminate/replace bogus types (e.g. quad and u_quad) -> use better types (e.g. quad_t & u_quad_t in inodes), some cleanup.
1.7	21-Apr-1994	cgd	blow away all vestiges of nfsnode locking. (1) it's unnecessary (2) it causes machines to hang (yup!) (3) it'd be gone in a few days anyway (it'd been yanked out of 4.4-Lite by macklem long ago) It was only there because macklem couldn't originally decide if things should be locked, or not...
1.6	01-Mar-1994	pk	Enable nfs_lock(); useful when IO_APPEND'ing.
1.5	15-Feb-1994	pk	Update {a,m}time vnode attributes on special files a la ufs_vnode.c, but make it a non-urgent operation, to leave us some performance.
1.4	18-Dec-1993	mycroft	Canonicalize all #includes.
1.3	28-Jul-1993	cgd	branches: 1.3.2; incorporate changes from 0-9-base to 0-9-ALPHA
1.2	20-May-1993	cgd	branches: 1.2.2; more rcs id adding and header cleanup. i like vi macros!
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.2.2.1	24-Jul-1993	cgd	clean the nfsnode's lockf field after getting a new vnode; this probably explains some strange NFS-related lockf crashes on pain, and UFS does it, so it can't hurt.
1.3.2.1	14-Nov-1993	mycroft	Canonicalize all #includes.
1.12.2.1	19-Aug-1994	mycroft	update from trunk
1.17.4.1	12-Mar-1997	is	Merge in changes from Trunk
1.22.2.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.28.8.2	02-Aug-1999	thorpej	Update from trunk.
1.28.8.1	07-Jun-1999	chs	merge everything from chs-ubc branch.
1.28.6.1	05-Jan-2000	he	Pull up revision 1.30 (requested by fvdl): Insert an extra VOP_ACCESS check in nfs_lookup, preventing cached access mishaps for lookup and getattr. Fixes PR#8884.
1.28.2.1	30-May-1999	chs	there's a new rule that all vnodes must call uvm_vnp_setsize() before anyone can possibly access them, so do this in nfs_nget().
1.29.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.29.2.6	23-Apr-2001	bouyer	Sync with HEAD.
1.29.2.5	21-Apr-2001	bouyer	Sync with HEAD
1.29.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.29.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.29.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.29.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.33.4.1	14-Dec-2000	he	Pull up revision 1.36 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.41.2.13	11-Dec-2002	thorpej	Sync with HEAD.
1.41.2.12	18-Oct-2002	nathanw	Catch up to -current.
1.41.2.11	15-Jul-2002	nathanw	Whitespace.
1.41.2.10	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.41.2.9	24-Jun-2002	nathanw	Curproc->curlwp renaming. Change uses of "curproc->l_proc" back to "curproc", which is more like the original use. Bare uses of "curproc" are now "curlwp". "curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL) so that it is always safe to reference curproc (dereferencing curproc is another story, but that's always been true).
1.41.2.8	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.41.2.7	28-Feb-2002	nathanw	Catch up to -current.
1.41.2.6	08-Jan-2002	nathanw	Catch up to -current.
1.41.2.5	14-Nov-2001	nathanw	Catch up to -current.
1.41.2.4	25-Sep-2001	nathanw	Fix typo in previous commit.
1.41.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.41.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.41.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.44.4.1	01-Oct-2001	fvdl	Catch up with -current.
1.44.2.5	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
1.44.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.44.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.44.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.44.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.46.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.53.6.1	28-Jul-2003	he	Apply patch (requested by christos in ticket #1171): Apply a stopgap fix preventing a panic for non-NQNFS when nfs_reclaim is called on a vnode of an unmounted NFS file system.
1.68.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.68.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.68.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.68.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.68.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.68.2.3	24-Aug-2004	skrll	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.68.2.2	03-Aug-2004	skrll	Sync with HEAD
1.68.2.1	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.
1.73.2.2	10-Jul-2004	tron	Pull up revision 1.75 (requested by tls in ticket #634): nfs_readdirplusrpc: fix a deadlock problem. don't wait for vnode lock to load attributes. otherwise, because READDIRPLUS returns DOTDOT entry as well, we violate locking order.
1.73.2.1	10-Jul-2004	tron	Pull up revision 1.74 (requested by tls in ticket #634): don't issue VOP_GETATTR blindly in nfs_nget(). in many cases, GETATTR RPCs here is redundant because the caller has postop_attr. instead, make sure the resulted vnode have a valid attribute in nfs_lookup().
1.77.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.77.6.1	12-Feb-2005	yamt	sync with head.
1.77.4.1	29-Apr-2005	kent	sync with -current
1.79.2.1	24-Aug-2005	riz	Pull up following revision(s) (requested by yamt in ticket #688): sys/miscfs/genfs/genfs_vnops.c: revision 1.98 via patch sys/ufs/ffs/ffs_vfsops.c: revision 1.165 sys/ufs/lfs/lfs_extern.h: revision 1.69 sys/fs/filecorefs/filecore_vfsops.c: revision 1.20 sys/nfs/nfs_node.c: revision 1.80 sys/fs/smbfs/smbfs_node.c: revision 1.24 sys/fs/cd9660/cd9660_vfsops.c: revision 1.24 sys/fs/msdosfs/msdosfs_denode.c: revision 1.8 sys/miscfs/genfs/genfs_node.h: revision 1.6 sys/ufs/lfs/lfs_vfsops.c: revision 1.183 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.86 sys/fs/adosfs/advfsops.c: revision 1.23 sys/fs/ntfs/ntfs_vfsops.c: revision 1.31 - constify genfs_ops. - use member designators. sys/miscfs/genfs/genfs_vnops.c: revision 1.99 via patch genfs_getpages: don't forget to put the vnode onto the syncer's work que ue even in the case of PGO_LOCKED. sys/uvm/uvm_bio.c: revision 1.40 sys/uvm/uvm_pager.h: revision 1.29 sys/miscfs/genfs/genfs_vnops.c: revision 1.100 via patch sys/ufs/ufs/ufs_inode.c: revision 1.50 - introduce PGO_NOBLOCKALLOC and use it for ubc mapping to prevent unnecessary block allocations in the case that page size > block size. - ufs_balloc_range: use VM_PROT_WRITE+PGO_NOBLOCKALLOC rather than VM_PROT_READ. sys/uvm/uvm_fault.c: revision 1.96 sys/miscfs/genfs/genfs_vnops.c: revision 1.101 via patch sys/uvm/uvm_object.h: revision 1.19 sys/miscfs/genfs/genfs_node.h: revision 1.7 ensure that vnodes with dirty pages are always on syncer's queue. - genfs_putpages: wait for i/o completion of PG_RELEASED/PG_PAGEOUT pages by setting "wasclean" false when encountering them. suggested by Stephan Uphoff in PR/24596 (1). - genfs_putpages: write protect pages when cleaning out, if we're going to take the vnode off the syncer's queue. uvm_fault: don't write-map pages unless its vnode is already on the syncer's queue. fix PR/24596 (3) but in the different way from the suggested fix. (to keep our current behaviour, ie. not to require explicit msync. discussed on tech-kern@.) - genfs_putpages: don't mistakenly take a vnode off the queue by introducing a generation number in genfs_node. genfs_getpages: increment the generation number. suggested by Stephan Uphoff in PR/24596 (2). - add some assertions. sys/miscfs/genfs/genfs_vnops.c: revision 1.102 via patch genfs_putpages: don't bother to clean the vnode unless VONWORKLST. sys/ufs/ffs/ffs_vnops.c: revision 1.71 ffs_full_fsync: because VBLK/VCHR can be mmap'ed, do VOP_PUTPAGES for them as well. sys/uvm/uvm_fault.c: revision 1.97 uvm_fault: check a correct object in the case of layered filesystems. fix PR/30811 from Jukka Salmi. sys/uvm/uvm_object.h: revision 1.20 sys/ufs/ffs/ffs_vfsops.c: revision 1.167 sys/uvm/uvm_bio.c: revision 1.41 sys/ufs/ufs/ufs_vnops.c: revision 1.129 sys/uvm/uvm_mmap.c: revision 1.92 sys/uvm/uvm_fault.c: revision 1.98 sys/kern/vfs_subr.c: revision 1.252 sys/fs/msdosfs/denode.h: revision 1.5 sys/miscfs/genfs/genfs_vnops.c: revision 1.103 via patch sys/fs/msdosfs/msdosfs_denode.c: revision 1.9 sys/sys/vnode.h: revision 1.141 sys/ufs/ufs/ufs_inode.c: revision 1.51 sys/ufs/ufs/ufs_extern.h: revision 1.45 via patch sys/miscfs/genfs/genfs_node.h: revision 1.8 sys/ufs/lfs/lfs_vfsops.c: revision 1.184 sys/uvm/uvm_pager.h: revision 1.30 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.87 update file timestamps for nfsd loaned-read and mmap. PR/25279. discussed on tech-kern@. sys/miscfs/genfs/genfs_vnops.c: revision 1.104 via patch don't write-protect wired pages. pointed by Chuck Silvers. for now, leave a vnode on the syncer's queue, as suggested by him. sys/ufs/ffs/ffs_vnops.c: revision 1.72 revert VCHR part of ffs_vnops.c 1.71. as VCHR uses the device pager, no point to call VOP_PUTPAGES here. pointed by Chuck Silvers.
1.80.2.7	04-Feb-2008	yamt	sync with head.
1.80.2.6	21-Jan-2008	yamt	sync with head
1.80.2.5	07-Dec-2007	yamt	sync with head
1.80.2.4	03-Sep-2007	yamt	sync with head.
1.80.2.3	26-Feb-2007	yamt	sync with head.
1.80.2.2	30-Dec-2006	yamt	sync with head.
1.80.2.1	21-Jun-2006	yamt	sync with head.
1.81.2.1	15-Jan-2006	yamt	sync with head.
1.82.10.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.82.10.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.82.8.3	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.
1.82.8.2	19-Apr-2006	elad	sync with head.
1.82.8.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.82.6.3	11-Aug-2006	yamt	sync with head
1.82.6.2	24-May-2006	yamt	sync with head.
1.82.6.1	01-Apr-2006	yamt	sync with head.
1.82.4.2	01-Jun-2006	kardel	Sync with head.
1.82.4.1	22-Apr-2006	simonb	Sync with head.
1.82.2.1	09-Sep-2006	rpaulo	sync with head
1.85.6.2	10-Dec-2006	yamt	sync with head.
1.85.6.1	22-Oct-2006	yamt	sync with head
1.85.4.2	12-Jan-2007	ad	Sync with head.
1.85.4.1	18-Nov-2006	ad	Sync with head.
1.90.2.2	24-Mar-2007	yamt	sync with head.
1.90.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.92.4.4	16-Sep-2007	ad	Checkpoint work in progress on the vnode lifecycle and reference counting stuff. This makes it work properly without kernel_lock and fixes a few quite old bugs. See vfs_subr.c 1.283.2.17 for details.
1.92.4.3	20-Aug-2007	ad	Sync with HEAD.
1.92.4.2	15-Jul-2007	ad	Sync with head.
1.92.4.1	13-Mar-2007	ad	Sync with head.
1.93.2.1	11-Jul-2007	mjf	Sync with head.
1.94.6.2	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.94.6.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.94.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.95.10.2	06-Aug-2007	yamt	nfs_inactive: turn a panic into a printf for now, as it isn't critical. PR/36572 from Martin Husemann.
1.95.10.1	06-Aug-2007	yamt	file nfs_node.c was added on branch matt-mips64 on 2007-08-06 11:55:09 +0000
1.95.8.2	18-Feb-2008	mjf	Sync with HEAD.
1.95.8.1	08-Dec-2007	mjf	Sync with HEAD.
1.95.2.2	23-Mar-2008	matt	sync with HEAD
1.95.2.1	09-Jan-2008	matt	sync with HEAD
1.96.6.2	19-Jan-2008	bouyer	Sync with HEAD
1.96.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.96.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.101.10.6	10-Oct-2010	yamt	some locking changes
1.101.10.5	09-Oct-2010	yamt	sync with head
1.101.10.4	11-Aug-2010	yamt	sync with head.
1.101.10.3	04-May-2009	yamt	sync with head.
1.101.10.2	16-May-2008	yamt	sync with head.
1.101.10.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.101.8.2	04-Jun-2008	yamt	sync with head
1.101.8.1	18-May-2008	yamt	sync with head.
1.101.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.101.6.2	05-Oct-2008	mjf	Sync with HEAD.
1.101.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.102.2.2	10-Oct-2008	skrll	Sync with HEAD.
1.102.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.103.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.103.4.1	19-Oct-2008	haad	Sync with HEAD.
1.106.4.1	02-Feb-2009	snj	Pull up following revision(s) (requested by ad in ticket #344): sys/nfs/nfs_node.c: revision 1.108 sys/nfs/nfsnode.h: revision 1.69 - Don't vput() a vnode that we do not hold locked. - Eliminate one of the few remaining uses of LK_CANRECURSE.
1.106.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.106.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.108.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.110.4.6	31-May-2011	rmind	sync with head
1.110.4.5	22-May-2011	rmind	nfs_gop_write: acquire the lock for pmap_page_protect() operation.
1.110.4.4	19-May-2011	rmind	Implement sharing of vnode_t::v_interlock amongst vnodes: - Lock is shared amongst UVM objects using uvm_obj_setlock() or getnewvnode(). - Adjust vnode cache to handle unsharing, add VI_LOCKSHARE flag for that. - Use sharing in tmpfs and layerfs for underlying object. - Simplify locking in ubc_fault(). - Sprinkle some asserts. Discussed with ad@.
1.110.4.3	05-Mar-2011	rmind	sync with head
1.110.4.2	03-Jul-2010	rmind	sync with head
1.110.4.1	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.
1.110.2.2	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.110.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.114.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.115.2.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.116.16.1	18-May-2014	rmind	sync with head
1.116.12.2	03-Dec-2017	jdolecek	update from HEAD
1.116.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.116.2.1	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")
1.117.2.1	10-Aug-2014	tls	Rebase.
1.118.8.1	26-Apr-2017	pgoyette	Sync with HEAD
1.118.4.2	28-Aug-2017	skrll	Sync with HEAD
1.118.4.1	05-Oct-2016	skrll	Sync with HEAD
1.119.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.122.8.1	25-Jun-2018	pgoyette	Sync with HEAD
1.123.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.123.2.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.124.2.1	29-Feb-2020	ad	Sync with head.
1.75	27-Dec-2006	yamt	remove nqnfs.
1.74	09-Nov-2006	yamt	remove some __unused in function parameters.
1.73	17-Oct-2006	dogcow	now that we have -Wno-unused-parameter, back out all the tremendously ugly code to gratuitously access said parameters.
1.72	13-Oct-2006	christos	more __unused
1.71	13-Oct-2006	dogcow	more unused variable fallout.
1.70	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.69	02-Sep-2006	yamt	branches: 1.69.2; 1.69.4; nfsd: deal with variable-sized filehandles.
1.68	02-Sep-2006	yamt	#ifdef out nqsrv_getlease and friends unless defined(NFSSERVER).
1.67	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.66	17-Jun-2006	yamt	- introduce vfs_composefh() and use it where appropriate. - fix lock/unlock mismatch in sys_getfh.
1.65	07-Jun-2006	kardel	branches: 1.65.2; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.64	20-May-2006	yamt	nqsrv_getlease: call nfs_init() to fix NFSSERVER && !NFS case.
1.63	18-May-2006	yamt	branches: 1.63.2; - fix some leaks in nfsd, introduced by kauth changes. - simplify code. - add some assertions. - wrap some long lines. - remove an unnecessary ";".
1.62	14-May-2006	elad	integrate kauth.
1.61	15-Apr-2006	christos	Coverity CID 1175: Remove dead code.
1.60	15-Apr-2006	christos	From my posting of April 3 to tech-kern: My understanding is that the CLRSIG() is supposed to clear the signal that was sent to the syncer process to prevent it from being delivered to the syncer process in case unmounting fails, so that the syncer process does not die while the filesystem is still mounted. The typical scenario is, the syncher process is tsleep()ing in the kernel, and waking up when it needs to do work. If someone sends a signal to it, eg. kill -TERM the mfs process, then the kernel will try to unmount the mfs filesystem before delivering the signal to the process. If that unmount fails, then we should not really kill the process because that will hang the mount. So we call CLRSIG() to stop the signal from being delivered. So the first call to issignal() will return the signal number that was sent to the syncer process (unless someone malicious was able to send a lower numbered signal between the time tsleep() returned and we called issignal()... something that is not really easy to do). But you are right, we should not be calling it many times as a side effect of this macro. Rewrite CLRSIG() clear all the signals and call issignal() the correct number of times.
1.59	15-Apr-2006	christos	Coverity CID 2509: Initialize cache
1.58	27-Mar-2006	martin	KASSERT that the returned file id length from VPTOFH is <= the maximum allowed value (_VFS_MAXFIDSZ).
1.57	11-Dec-2005	christos	branches: 1.57.4; 1.57.6; 1.57.8; 1.57.10; 1.57.12; merge ktrace-lwp.
1.56	22-May-2004	jonathan	branches: 1.56.12; Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.55	21-Apr-2004	christos	use VFS_MAXFIDSZ
1.54	21-Apr-2004	christos	Replace the statfs() family of system calls with statvfs(). Retain binary compatibility.
1.53	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.52	30-Jul-2003	yamt	eliminate v_id.
1.51	29-Jun-2003	fvdl	branches: 1.51.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.50	29-Jun-2003	thorpej	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.49	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
1.48	21-May-2003	yamt	remove local definitions of TRUE and FALSE.
1.47	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.46	02-Apr-2003	yamt	use queue manipulation macros.
1.45	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.44	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.43	18-Jan-2003	thorpej	Merge the nathanw_sa branch.
1.42	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.41	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.40	12-May-2002	matt	Eliminate commons
1.39	10-Nov-2001	lukem	branches: 1.39.4; add RCSIDs
1.38	16-Apr-2001	thorpej	branches: 1.38.2; 1.38.6; When unmounting a file system, acquire the syncer_lock before vfs_busy'ing just before the dounmount() call. This is to avoid sleeping with the mountlist_slock held -- but we must acquire syncer_lock before vfs_busy because the syncer itself uses syncer_lock -> vfs_busy locking order.
1.37	21-Feb-2001	jdolecek	branches: 1.37.2; make some more constant arrays 'const'
1.36	06-Feb-2001	fvdl	Do actual vnode locking for NFS.
1.35	24-Nov-2000	chs	put more ISO bits under ifdef ISO.
1.34	19-Sep-2000	bjh21	Extend NFS_V2_ONLY to remove NQNFS lease support as well. Saves another 10k.
1.33	19-Sep-2000	fvdl	Adapt for VOP_FSYNC parameter change.
1.32	09-Jun-2000	fvdl	branches: 1.32.2; Some tweaks to enable NFS over IPv6. The special-casing of AF_INET should really be removed.
1.31	30-Mar-2000	augustss	branches: 1.31.2; Remove register declarations.
1.30	10-Oct-1999	sommerfeld	branches: 1.30.2; Fix bug in error handling for NFSv3 + nqnfs. With nfsv2, the nfsm_reply() macro always causes the service routine to return if error was nonzero. With nfsv3, we can keep going after nfsm_reply() without returning, but nqnfsrv_getlease() didn't take this into account, so add a return(0) after each error-case nfsm_reply(0).
1.29	25-Mar-1999	sommerfe	branches: 1.29.2; 1.29.8; Fix crash reported in PR7116 on shutdown
1.28	06-Mar-1999	fair	Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.27	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.26	25-Jun-1998	thorpej	- Rename nqnfs_vop_lease_check() to genfs_lease_check(). If NFSSERVER is not in the kernel, genfs_lease_check() is simply a no-op. This allows LKM'd file systems to be exported (previously did not work properly due to a compile-time decision based on -DNFSSERVER). - defopt NFSSERVER
1.25	05-Jun-1998	kleink	Convert fsync vnode operator implementations and usage from the old `waitfor' argument and MNT_WAIT/MNT_NOWAIT to `flags' and FSYNC_WAIT.
1.24	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.23	19-Feb-1998	thorpej	Include the NFS option header.
1.22	19-Oct-1997	fvdl	* Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.21	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.20	24-Jun-1997	fvdl	branches: 1.20.4; Provide extra arg to nfsrv_fhtovp (just FALSE in this case), it was extended for WebNFS support.
1.19	22-Feb-1997	fvdl	Fixes from BSDI (thanks go to Keith Bostic). Original RCS messages: date: 1995/11/30 20:37:53; author: cp; state: Exp; lines: +3 -3 Change splsoftclock() to splnet(); Make nfsrv_getstream create two copies of data when splitting up an mbuf rather than two references to the same external buffer. The symptom this fixes is client hangs. date: 1996/10/16 00:06:05; author: ewv; state: Exp; lines: +5 -3 Clear pending signal when an unmount fails, this allows us another chance at the umount after a short sleep. The fixes a problem where /usr is mounted via nqnfs and the system hangs on shutdown since the umount() always fails with EBUSY (inetd is still busy on usr) and since we don't clear the signal we end up stuck looping and never give inetd a chance to catch its SIGKILL. date: 1996/10/23 18:22:14; author: donn; state: Exp; lines: +12 -7 Kirk's changes to prevent races when unmounting. (1) Unmount() and vfs_unmountall() now call vfs_busy() so that they participates in the mount structure locking scheme. Dounmount() calls vfs_unbusy() to unlock things, and makes sure to wake up waiters if there's an error. (2) The MFS and NQNFS daemons also now use vfs_busy() when unmounting filesystems. Kirk restructured the code so that a successful unmount by another process won't leave the possibility that a daemon might reference a mount structure that has been freed.
1.18	09-Feb-1997	fvdl	* Fix some bugs in NQNFS (malformed RPC requests, no directory lease eviction) * Avoid possible NULL ptr ref in nfs_reply * Don't ever try to sillyrename directories (from FreeBSD)
1.17	31-Jan-1997	thorpej	branches: 1.17.2; NFSCLIENT -> NFS.
1.16	13-Oct-1996	christos	branches: 1.16.2; revert kprintf changes
1.15	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.14	18-Feb-1996	fvdl	branches: 1.14.4; Fix a missing 'error =' before 'if (error)'.
1.13	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.12	09-Feb-1996	christos	nfs prototype changes
1.11	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.10	18-Jun-1995	cgd	don't assume the f_fsnamelen is nul-truncated or longer than MFSNAMELEN
1.9	23-May-1995	mycroft	Remove gratuitous extra indirections.
1.8	18-Jan-1995	mycroft	Turn mountlist into a CIRCLEQ, and handle setting and checking of MNT_ROOTFS differently.
1.7	13-Dec-1994	mycroft	Turn lease_check() into a vnode op, per CSRG.
1.6	21-Aug-1994	cgd	branches: 1.6.2; cleanliness; don't wrap lines.
1.5	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.4	17-Aug-1994	mycroft	Convert some more lists and queues.
1.3	17-Aug-1994	mycroft	Change the reply list to a TAILQ.
1.2	29-Jun-1994	cgd	branches: 1.2.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.1	08-Jun-1994	mycroft	branches: 1.1.1; Update to 4.4-Lite fs code, with local changes.
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.2.2.1	19-Aug-1994	mycroft	update from trunk
1.6.2.2	21-Aug-1994	cgd	cleanliness; don't wrap lines.
1.6.2.1	21-Aug-1994	cgd	file nfs_nqlease.c was added on branch netbsd-1-0 on 1994-08-21 21:07:14 +0000
1.14.4.1	04-Mar-1997	mycroft	Pull up bug fixes from -current, per fvdl.
1.16.2.1	14-Jan-1997	thorpej	Snapshot of work-in-progress, committed to private branch. These changes implement machine-independent root device and file system selection. Notable features: - All ports behave in a consistent manner regarding root device selection. - No more "options GENERIC"; all kernels have the ability to boot with RB_ASKNAME to select root device and file system type. - Root file system type can be wildcarded; a machine-independent function will try all possible file systems for the selected root device until one succeeds. - If the root file system fails to mount, the operator will be given the chance to select a new root device and file system type, rather than having the machine simply panic. - nfs_mountroot() no longer panics if any part of the NFS mount process fails; it now returns an error, giving the operator a chance to recover. - New, more consistent, config(8) grammar. The constructs: config netbsd swap generic config netbsd root on nfs have been replaced with: config netbsd root on ? type ? config netbsd root on ? type nfs Additionally, the operator may select or wildcard root file system type in the kernel configuration file: config netbsd root on cd0a type cd9660 config(8) now requires that a "root" specification be made. "root" may be wired down or wildcarded. "swap" and "dump" specifications are optional, and follow previous semantics. - config(8) has a new "file-system" keyword, used to configure file systems into the kernel. Eventually, this will be used to generate the default vfssw[]. - "options NFSCLIENT" is obsolete, and is replaced by "file-system NFS". "options NFSSERVER" still exists, since NFS server support is independent of the NFS file system client. - sys/arch/<foo>/<foo>/swapgeneric.c is no longer used, and will be removed; all information is now generated by config(8). As of this commit, all ports except arm32 have been updated to use the new setroot(). Only SPARC, i386, and Alpha ports have been tested at this time. Port masters should test these changes on their ports, and report any problems back to me. More changes are on their way, including RB_ASKNAME support in nfs_mountroot() (to prompt for server address and path) and, potentially, the ability to select rarp/bootparam or bootp in nfs_mountroot().
1.17.2.1	12-Mar-1997	is	Merge in changes from Trunk
1.20.4.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.29.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.29.2.1	10-Oct-1999	cgd	pull up rev 1.30 from trunk (requested by sommerfeld): Fix an odd corner case if you use nfsv3 and nqnfs at the same time: v3 changes the error-case behavior of the nfsm_reply macro, but the caller keeps going and in this case you end up calling vput(NULL).
1.30.2.5	21-Apr-2001	bouyer	Sync with HEAD
1.30.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.30.2.3	11-Feb-2001	bouyer	Sync with HEAD.
1.30.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.30.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.31.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.32.2.1	14-Dec-2000	he	Pull up revision 1.33 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.37.2.7	11-Dec-2002	thorpej	Sync with HEAD.
1.37.2.6	22-Oct-2002	thorpej	Sync with HEAD.
1.37.2.5	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.37.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.37.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.37.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.37.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.38.6.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.38.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.38.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.39.4.1	11-Mar-2002	thorpej	Make syncer_lock an adaptive mutex and rename it to syncer_mutex.
1.51.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.51.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.51.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.51.2.3	24-Aug-2004	skrll	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.51.2.2	03-Aug-2004	skrll	Sync with HEAD
1.51.2.1	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.
1.56.12.2	30-Dec-2006	yamt	sync with head.
1.56.12.1	21-Jun-2006	yamt	sync with head.
1.57.12.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.57.12.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.57.10.3	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.
1.57.10.2	19-Apr-2006	elad	sync with head.
1.57.10.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.57.8.5	03-Sep-2006	yamt	sync with head.
1.57.8.4	11-Aug-2006	yamt	sync with head
1.57.8.3	26-Jun-2006	yamt	sync with head.
1.57.8.2	24-May-2006	yamt	sync with head.
1.57.8.1	01-Apr-2006	yamt	sync with head.
1.57.6.3	01-Jun-2006	kardel	Sync with head.
1.57.6.2	22-Apr-2006	simonb	Sync with head.
1.57.6.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.57.4.1	09-Sep-2006	rpaulo	sync with head
1.63.2.1	19-Jun-2006	chap	Sync with head.
1.65.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.69.4.2	10-Dec-2006	yamt	sync with head.
1.69.4.1	22-Oct-2006	yamt	sync with head
1.69.2.3	12-Jan-2007	ad	Sync with head.
1.69.2.2	18-Nov-2006	ad	Sync with head.
1.69.2.1	11-Sep-2006	ad	- Convert some locks to mutexes and RW locks. - Use the proclist_lock to protect pgrps and sessions in some places.
1.184	23-Mar-2023	riastradh	nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them. XXX pullup-8 XXX pullup-9 XXX pullup-10
1.183	27-Apr-2022	hannken	branches: 1.183.4; As VOP_GETATTR() needs a shared lock at least move the preopattr lookup inside nfs_namei() where we may lock the start directory without violating the lock order.
1.182	16-Sep-2021	andvar	fix various typos, mainly in comments.
1.181	05-Sep-2020	riastradh	Round of uvm.h cleanup. The poorly named uvm.h is generally supposed to be for uvm-internal users only. - Narrow it to files that actually need it -- mostly files that need to query whether curlwp is the pagedaemon, which should maybe be exposed by an external header. - Use uvm_extern.h where feasible and uvm_*.h for things not exposed by it. We should split up uvm_extern.h but this will serve for now to reduce the uvm.h dependencies. - Use uvm_stat.h and #ifdef UVMHIST uvm.h for files that use UVMHIST(ubchist), since ubchist is declared in uvm.h but the reference evaporates if UVMHIST is not defined, so we reduce header file dependencies. - Make uvm_device.h and uvm_swap.h independently includable while here. ok chs@
1.180	04-Apr-2020	mlelstv	NFSv2 is limited to use only 32bit in metadata. Prevent that larger metadata values are simply truncated. -> clamp filesystem block counts to signed 32bit. -> clamp file sizes to signed 32bit () Some NFSv2 clients also have problems to handle buffer sizes larger than (signed) 16bit. -> clamp buffer sizes to signed 16bit for better compatibility. () This can lead to erroneous behaviour for files larger than 2GB that NFSv2 cannot handle but it is still better than before. An alternative would be to (partially) reject operations on files larger than 2GB, but which causes other problems.
1.179	17-Jan-2020	ad	VFS_VGET(), VFS_ROOT(), VFS_FHTOVP(): give them a "int lktype" argument, to allow us to get shared locks (or no lock) on the returned vnode. Matches FreeBSD.
1.178	02-Jan-2020	thorpej	branches: 1.178.2; - Eliminate the global "boottime" variable, which was being accessed without any synchronization against changes by e.g. clock_settime(). - Replace with new getbinboottime() / getnanoboottime() / getmicroboottime() functions (naming mirrors that of other time access functions in kern_tc.c). It returns the (maybe-converted) value of timebasebin, which also tracks our estimate of when the system was booted (i.e. the legacy "boottime" was redundant). XXX There needs to be a lockless synchronization mechanism for reading timebasebin, but this is a problem in kern_tc.c that pre-existed these "boottime" changes. At least now the problem is centralized in one location.
1.177	20-Feb-2019	hannken	branches: 1.177.4; Bracket do_sys_renameat() and nfsrv_rename() with fstrans. The v_mount field for vnodes on the same file system as "from" is now stable for referenced vnodes. VFS_RENAMELOCK no longer may use lock from an unreferenced and freed "struct mount".
1.176	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.175	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!)
1.174	03-May-2018	hannken	branches: 1.174.2; nfsrv_readlink: stop attaching a zero-length mbuf for zero length symlinks.
1.173	26-Apr-2017	riastradh	branches: 1.173.4; 1.173.10; Change VOP_REMOVE and VOP_RMDIR to preserve lock/ref on dvp. No change to vp -- the plan is to replace the node by the componentname in the vop parameters, and let all directory vops do lookups internally. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2017/04/17/msg021825.html
1.172	21-Apr-2015	riastradh	Cull unused INRENAME and INRELOOKUP from callers.
1.171	20-Apr-2015	riastradh	Make VOP_LINK return directory still locked and referenced. Ride 7.99.10 bump.
1.170	23-Jan-2014	hannken	branches: 1.170.6; Change vnode operations create, mknod, mkdir and symlink to return the resulting vnode *vpp unlocked. Discussed on tech-kern@ Welcome to 6.99.30
1.169	17-Jan-2014	hannken	Change vnode operations create, mknod, mkdir and symlink to keep the directory node dvp locked on return. Discussed on tech-kern@ Welcome to 6.99.29
1.168	14-Dec-2013	christos	only prevent autounload, not regular unload when we have exports
1.167	14-Dec-2013	christos	don't allow the nfs server module to unload if it has exported filesystems.
1.166	14-Sep-2013	martin	Backout wildcard pragma to kill warnings and instead sprinkle a few dozen __unused attributes. Requested by joerg@
1.165	29-Aug-2012	christos	branches: 1.165.2; 1.165.4; When unloading the nfsserver module, call nfs_fini() so that the nfsrvdescpl pool gets destroyed. Otherwise we are left with a stray pool that points to unmapped memory behind (and bad things happen). Typically you get seemingly random page faults (without printing uvm_fault) that happen in various pool operations. Most frequent one is the pool_drain() from the page daemon.
1.164	27-Aug-2012	chs	fix error handling in nfsrv_rename(): when the first nfs_namei() fails, don't try to free the resources allocated by a successful lookup.
1.163	01-Feb-2012	matt	branches: 1.163.2; 1.163.4; When using socket loaning, make sure the KVA used for the loan has the same color as the UVA being loaned.
1.162	21-Nov-2011	hannken	branches: 1.162.2; nfsrv_lookup(): Defer the postopattr lookup on dirp until the child node is unlocked. Ok: YAMAMOTO Takashi <yamt@netbsd.org>
1.161	30-Oct-2011	hannken	branches: 1.161.2; VOP_GETATTR() needs a shared lock at least.
1.160	08-Aug-2011	dholland	nfs_namei() should not return a non-null path buffer except on success, even though the callers are apparently prepared to cope. Fixes last tidyup part of PR 44625.
1.159	18-Apr-2011	dholland	Back in -r1.60 of nfs_serv.c (a long time ago) VOP_MKNOD was changed so nfsd no longer needed to do a lookup() call immediately afterwards to retrieve the newly created object. Since that change there has been no way for ISSYMLINK to be set upon return from VOP_MKNOD (or before the call to VOP_MKNOD either) so remove the test for it and associated block of dead code. (I do not understand how this code was reachable before then either. The logic in question is only reached if no object by that name existed, and there's no reasonable way that a successful call to VOP_MKNOD should ever create a symlink. The code appears to come from 4.4lite; maybe they had locking bugs?)
1.158	11-Apr-2011	dholland	Clean up. Move some more code across from nfsd's private entry points.
1.157	19-Mar-2011	dholland	Fix memory leak introduced with the struct pathbuf changes. Hi, me. Closes PR 44625.
1.156	05-Feb-2011	yamt	typo in a comment
1.155	02-Jan-2011	dholland	branches: 1.155.2; 1.155.4; Remove remaining references to SAVESTART.
1.154	02-Jan-2011	dholland	Remove the special refcount behavior (adding an extra reference to the parent dir) associated with SAVESTART in relookup(). Check all call sites to make sure that SAVESTART wasn't set while calling relookup(); if it was, adjust the refcount behavior. Remove related references to SAVESTART. The only code that was reaching the extra ref was msdosfs_rename, where the refcount behavior was already fairly broken and/or gross; repair it. Add a dummy 4th argument to relookup to make sure code that hasn't been inspected won't compile. (This will go away next time the relookup semantics change, which they will.)
1.153	02-Jan-2011	dholland	Remove unused nameidata field ni_startdir.
1.152	30-Nov-2010	dholland	Abolish struct componentname's cn_pnbuf. Use the path buffer in the pathbuf object passed to namei as work space instead. (For now a pnbuf pointer appears in struct nameidata, to support certain unclean things that haven't been fixed yet, but it will be going away in the future.) This removes the need for the SAVENAME and HASBUF namei flags.
1.151	24-Jun-2010	hannken	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.
1.150	08-Jan-2010	pooka	branches: 1.150.2; 1.150.4; The VATTR_NULL/VREF/VHOLD/HOLDRELE() macros lost their will to live years ago when the kernel was modified to not alter ABI based on DIAGNOSTIC, and now just call the respective function interfaces (in lowercase). Plenty of mix'n match upper/lowercase has creeped into the tree since then. Nuke the macros and convert all callsites to lowercase. no functional change
1.149	23-Dec-2009	pooka	Define namei flag INRENAME and set it if a lookup operation is part of rename. This helps with building better asserts for rename in the DELETE lookup ... the RENAME lookup is quite obviously a part of rename.
1.148	07-Nov-2009	cegger	Add a flags argument to pmap_kenter_pa(9). Patch showed on tech-kern@ http://mail-index.netbsd.org/tech-kern/2009/11/04/msg006434.html No objections.
1.147	27-Sep-2009	dholland	Rename lookup() to lookup_for_nfsd(), to make it clear just whose private backdoor entry point this is. Also, clone the lookup_for_nfsd() entry point as lookup_for_nfsd_index(), for use by a different call site in nfsd that does different unclean things with nameidata.
1.146	23-May-2009	ad	- Fix a race between umount()/mount() and nfssvc(). - Toss netexport state on nfsserver module unload.
1.145	23-May-2009	ad	Fix a crash when unloading nfsserver module.
1.144	10-Apr-2009	bouyer	PR kern/41158: nfs_rename() locking against myself nfsrv_rename() can exit without calling genfs_renamelock_exit() because the nfsm_reply() can do return (0) on error. Change nfsm_reply to use 'error = 0; goto nfsmout' instead. Fix a few place so it's safe to goto nfsmout from nfsm_reply, or other macros calling it. As a side effect it could fix a missing vrele(dirp) in various place where nfsm_reply could return(0).
1.143	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.142	11-Jan-2009	christos	branches: 1.142.2; merge christos-time_t
1.141	03-Dec-2008	pooka	nfsd_use_loan: int -> bool
1.140	27-Nov-2008	pooka	Use struct nfs_fattr in struct flrep instead of uint32_t array acrobatics to get rid of type punning warning.
1.139	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.138	28-Mar-2008	dholland	branches: 1.138.2; 1.138.6; 1.138.12; 1.138.14; 1.138.16; Yet another rename workaround - this time check for . and .. early because relookup() objects to being asked to handle them.
1.137	08-Mar-2008	yamt	desupport link/unlink of directories. noted by Elad Efrat.
1.136	28-Feb-2008	elad	Introduce a new kauth action, KAUTH_NETWORK_NFS, and two requests, KAUTH_REQ_NETWORK_NFS_EXPORT and KAUTH_REQ_NETWORK_NFS_SVC, and use them to replace two KAUTH_GENERIC_ISSUSER calls in the NFS code. Also replace two more with KAUTH_SYSTEM_MKNOD, where appropriate. Documetnation and examples updated. More to come.
1.135	20-Feb-2008	matt	branches: 1.135.2; 1.135.6; Fix extern declaration to match actual declaration (add const).
1.134	28-Jan-2008	dholland	Fix some race conditions in rename. Introduce a per-FS rename lock and new vfsops to manipulate it. Get this lock while renaming. Also add another relookup() in do_sys_rename, which is a hack to kludge around some of the worst deficiencies of ufs_rename. reviewed-by: pooka (and an earlier rev by ad) posted on tech-kern with no objections.
1.133	22-Dec-2007	yamt	nfsrv_create: fix a use-after-release.
1.132	04-Dec-2007	yamt	branches: 1.132.4; merge non-intrusive nfs changes from vmlocking.
1.131	26-Nov-2007	pooka	branches: 1.131.2; 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
1.130	10-Oct-2007	ad	branches: 1.130.4; Merge from vmlocking: - Split vnode::v_flag into three fields, depending on field locking. - simple_lock -> kmutex in a few places. - Fix some simple locking problems.
1.129	27-Jul-2007	yamt	branches: 1.129.4; 1.129.6; 1.129.8; 1.129.10; stop nfs tick when we have nothing to do.
1.128	06-Apr-2007	hannken	branches: 1.128.4; Remove calls to now obsolete vn_start_write() and vn_finished_write().
1.127	05-Mar-2007	yamt	branches: 1.127.2; 1.127.4; nfsrv_setattr: revive nfsm_srvsattr which was (mistakenly?) removed with caddr_t.
1.126	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.125	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.124	20-Feb-2007	pooka	after freeing cookies, set the pointer to NULL to prevent dangling reuse
1.123	04-Feb-2007	chs	branches: 1.123.2; in nfsrv_rename(), change vput(fdirp) back to vrele() since the dirp returned from nfs_namei() is not locked in this case. fixes PR 35542. also, apply the same fix here as was made in rename_files() to handle the case when dvp and vp are the same.
1.122	04-Jan-2007	elad	Consistent usage of KAUTH_GENERIC_ISSUSER.
1.121	27-Dec-2006	yamt	remove nqnfs.
1.120	09-Dec-2006	chs	a smorgasbord of improvements to vnode locking and path lookup: - LOCKPARENT is no longer relevant for lookup(), relookup() or VOP_LOOKUP(). these now always return the parent vnode locked. namei() works as before. lookup() and various other paths no longer acquire vnode locks in the wrong order via vrele(). fixes PR 32535. as a nice side effect, path lookup is also up to 25% faster. - the above allows us to get rid of PDIRUNLOCK. - also get rid of WANTPARENT (just use LOCKPARENT and unlock it). - remove an assumption in layer_node_find() that all file systems implement a recursive VOP_LOCK() (unionfs doesn't). - require that all file systems supply vfs_vptofh and vfs_fhtovp routines. fill in eopnotsupp() for file systems that don't support being exported and remove the checks for NULL. (layerfs calls these without checking.) - in union_lookup1(), don't change refcounts in the ISDOTDOT case, just adjust which vnode is locked. fixes PR 33374. - apply fixes for ufs_rename() from ufs_vnops.c rev. 1.61 to ext2fs_rename().
1.119	09-Nov-2006	yamt	branches: 1.119.2; remove some __unused in function parameters.
1.118	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.117	02-Sep-2006	yamt	branches: 1.117.2; 1.117.4; nfsd: deal with variable-sized filehandles.
1.116	02-Sep-2006	christos	fix default type decls fix incomplete initializer
1.115	20-Jul-2006	christos	When there are too many empty entries in a row, and we need to try to read the next block, free the cookie buffer before doing so to avoid a memory leak. Reported by Mark Davies.
1.114	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.113	30-Jun-2006	yamt	wrap long lines and fix indents after kauth merge.
1.112	17-Jun-2006	yamt	- introduce vfs_composefh() and use it where appropriate. - fix lock/unlock mismatch in sys_getfh.
1.111	09-Jun-2006	christos	branches: 1.111.2; stack police: don't allocate statvfs on the stack.
1.110	07-Jun-2006	kardel	merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.109	14-May-2006	elad	branches: 1.109.2; integrate kauth.
1.108	10-May-2006	mrg	quell GCC 4.1 uninitialised variable warnings. XXX: we should audit the tree for which old ones are no longer needed after getting the older compilers out of the tree..
1.107	15-Apr-2006	christos	Coverity CID 735: Remove duplicate code.
1.106	15-Apr-2006	christos	Coverity CID 736: Comment out dead code.
1.105	15-Apr-2006	christos	Coverity CID 1143: Prevent NULL deref.
1.104	15-Apr-2006	christos	Coverity CID 1144: Protect against NULL deref.
1.103	15-Apr-2006	christos	Coverity CID 2510-2514: Always initialize cache.
1.102	27-Mar-2006	martin	KASSERT that the returned file id length from VPTOFH is <= the maximum allowed value (_VFS_MAXFIDSZ).
1.101	01-Mar-2006	yamt	branches: 1.101.2; 1.101.4; 1.101.6; 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.
1.100	03-Jan-2006	yamt	branches: 1.100.2; 1.100.4; remove a few unnecessary caddr_t casts.
1.99	11-Dec-2005	christos	branches: 1.99.2; merge ktrace-lwp.
1.98	06-Oct-2005	yamt	branches: 1.98.6; - remove a ufs dependency. - bump readdir block size to 1024. (the same value as userland DIRBLKSIZ)
1.97	06-Sep-2005	jmmv	Set va_type to VLNK before calling VOP_SYMLINK to match the change in the vfs_syscalls.c file. Pointed out by yamt@.
1.96	19-Aug-2005	yamt	as we now have 64bit ino_t, no need to truncate nfsv3 fileids.
1.95	18-May-2005	yamt	branches: 1.95.2; nfsrv_mknod: reject device numbers which we can't handle.
1.94	26-Feb-2005	perry	nuke trailing whitespace
1.93	09-Dec-2004	yamt	branches: 1.93.2; 1.93.4; nfsrv_commit: make cnt unsigned so that very large commit requests can be handled properly.
1.92	09-Dec-2004	yamt	when calling create-type VOP, make sure that va_mode is set even when a client doesn't specify it. (most filesystems get confused if va_mode is VNOVAL.)
1.91	04-Dec-2004	yamt	nfsrv_read: fall back to copying when fail to loan pages. (i forgot to commit this with uvm_loan.c rev.1.51.)
1.90	17-Sep-2004	skrll	There's no need to pass a proc value when using UIO_SYSSPACE with vn_rdwr(9) and uiomove(9). OK'd by Jason Thorpe
1.89	31-May-2004	yamt	nfsrv_create: fix an LP64 problem for exclusive create.
1.88	21-Apr-2004	christos	Replace the statfs() family of system calls with statvfs(). Retain binary compatibility.
1.87	07-Jan-2004	yamt	branches: 1.87.2; - get pages to loan out in uvm_loanuobjpages() rather than having caller (nfsd, in this case) do so. - tweak locking so that nfs loaned READ works on layered filesystems.
1.86	05-Nov-2003	hannken	Clean up the usage of vn_start_write(). At least one occurence clobbered previous error conditions. If "(flags & (V_WAIT\|V_PCATCH)) == V_WAIT" the return value is always zero. Ignore the return value in these cases. From Darrin B. Jewell.
1.85	29-Oct-2003	mycroft	Back out the bogus initializer -- the compiler bug is fixed.
1.84	28-Oct-2003	cl	note 'm68k {u,}int64_t used uninitialized' bug. add reference to gcc bug report. mark all (known) occurrences.
1.83	20-Oct-2003	yamt	set READres EOF flag correctly.
1.82	15-Oct-2003	hannken	Add the gating of system calls that cause modifications to the underlying file system. The function vfs_write_suspend stops all new write operations to a file system, allows any file system modifying system calls already in progress to complete, then sync's the file system to disk and returns. The function vfs_write_resume allows the suspended write operations to complete. From FreeBSD with slight modifications. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.81	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.80	09-Jul-2003	bouyer	nfsrv_commit(): return success and don't do anything for requests which starts past the end of the file. This can happen when two clients are writting to the same file. Close PR 21696 by myself, discussed on tech-net in 2003/05 and 2003/06. Issue raised by Chuck Silvers (commit and truncate ops needs to be serialised) still unadressed.
1.79	29-Jun-2003	fvdl	branches: 1.79.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.78	29-Jun-2003	thorpej	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.77	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
1.76	09-Jun-2003	yamt	rework zero padding of rpc reply. - for READ procedure, don't send back more bytes than requested. - don't have doubtful assumptions on mbuf chain structure. - rename a function (nfsm_adj -> nfs_zeropad) to avoid confusion as the semantics of the function was changed.
1.75	29-May-2003	yamt	workaround for UBC limit. while our nfsd announces MAXBSIZE as wtmax for tcp, VOP_GETPAGES of filesystems that uses genfs_getpages can't handle >= MAX_READ_AHEAD(16) pages at once. therefore, depending on PAGE_SIZE of the machine and file offset of a read request, we can't VOP_GETPAGES the range at once.
1.74	07-May-2003	yamt	- indent. - fix a comment typo (mus -> must) - remove an unneeded caddr_t cast.
1.73	04-May-2003	yamt	fix handling of the case that readsize == 0.
1.72	03-May-2003	yamt	use uvm page loanout mechanism for nfsd READ procedure processing. reviewed by Frank van der Linden and Chuck Silvers. tested by Wojciech Puchar.
1.71	03-Apr-2003	yamt	return rtmax bytes if we get READ requests larger than rtmax.
1.70	02-Apr-2003	yamt	use queue manipulation macros.
1.69	28-Mar-2003	yamt	reply FSINFO rtmax and wtmax for DGRAM properly.
1.68	28-Mar-2003	yamt	reply ENAMETOOLONG properly instead of discarding request as BADRPC. my own PR20791.
1.67	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.66	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.65	27-Sep-2002	bouyer	In nfs_commit(), sanity-check what we get from network: if we try to fluch past end of file, or if off + cnt overflows a quad_t, flush to end of file.
1.64	26-Sep-2002	bouyer	In nfsrv_create(), kill an extra PNBUF_PUT() in the NFSv2 mknod case. The pnbuf has already been freed by VOP_MKNOD. This should have been removed in rev 1.60. Should fix PR 18013, OK'd by fvdl.
1.63	26-Sep-2002	bouyer	nfsrv_commit(): Properly handle the case cnt == 0, which means "flush to end of file". Calling VOP_FSYNC with start == end triggers a DIAGNOSTIC check. Noticed with NFSv3 Linux clients. OK'd by fvdl.
1.62	10-Nov-2001	lukem	branches: 1.62.10; add RCSIDs
1.61	23-Sep-2001	chs	branches: 1.61.2; remove SAVESTART from the symlink, mknod and create operations. it was unnecessary, and removing it also fixes a v_usecount leak that was introduced in the previous revision.
1.60	24-Jul-2001	assar	branches: 1.60.2; change vop_symlink and vop_mknod to return vpp (the created node) refed, so that the caller can actually use it. update callers and file systems that implement these vnode operations
1.59	27-Nov-2000	chs	branches: 1.59.2; 1.59.4; Initial integration of the Unified Buffer Cache project.
1.58	19-Sep-2000	fvdl	Adapt for VOP_FSYNC parameter change.
1.57	03-Aug-2000	thorpej	Convert namei pathname buffer allocation to use the pool allocator.
1.56	03-Aug-2000	thorpej	MALLOC()/FREE() are not to be used for variable size allocations.
1.55	27-Jun-2000	mrg	remove include of <vm/vm.h>
1.54	30-Mar-2000	augustss	branches: 1.54.4; Remove register declarations.
1.53	30-Mar-2000	simonb	Delete redundant decl of nfs_pub - it's in <sys/mount.h>.
1.52	05-Dec-1999	fvdl	The length check for readdirplus entries wasn't right, causing troubles with 32k readdir sizes. From FreeBSD.
1.51	04-May-1999	sommerfe	branches: 1.51.2; 1.51.8; Fix vnode lock leak in nfsrv_mknod() if to-be-created vnode already existed.
1.50	30-Mar-1999	mycroft	branches: 1.50.2; Fix two problems with NFSV3CREATE_GUARDED: * We shouldn't truncate the file. * We were leaving the vnode locked (unless the truncate happened to fail). Solaris clients may cause this under some conditions. Problem reported by chopps, analysis and fix by me.
1.49	24-Mar-1999	mrg	completely remove Mach VM support. all that is left is the all the header files as UVM still uses (most of) these.
1.48	06-Mar-1999	fair	Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.47	05-Mar-1999	mycroft	Clean up some sign extension bogosity in statfs, so negative numbers are actually negative on a LP64 client.
1.46	31-Jan-1999	mrg	non-root users can mkfifo over NFS.
1.45	18-Aug-1998	thorpej	Add some braces to make egcs happy (ambiguous else warning).
1.44	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.43	05-Jun-1998	kleink	Convert fsync vnode operator implementations and usage from the old `waitfor' argument and MNT_WAIT/MNT_NOWAIT to `flags' and FSYNC_WAIT.
1.42	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.41	10-Feb-1998	mrg	- add defopt's for UVM, UVMHIST and PMAP_NEW. - remove unnecessary UVMHIST_DECL's.
1.40	05-Feb-1998	mrg	initial import of the new virtual memory system, UVM, into -current. UVM was written by chuck cranor <chuck@maria.wustl.edu>, with some minor portions derived from the old Mach code. i provided some help getting swap and paging working, and other bug fixes/ideas. chuck silvers <chuq@chuq.com> also provided some other fixes. this is the rest of the MI portion changes. this will be KNF'd shortly. :-)
1.39	22-Dec-1997	fvdl	Check vnode for VDIR type before doing anything with it in the NFS readdir service.
1.38	10-Oct-1997	fvdl	branches: 1.38.2; * New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.37	17-Jul-1997	fvdl	branches: 1.37.2; * Deal with servers that don't give complete FSINFO (like NT) From Olaf Seibert <rhialto@polder.ubc.kun.nl> (PR 3687) * Make an attempt to check the maximum filesize before attempting a write to the server, as write RPCs will typically happen asynchronously, and the process will not see the error. Fixes problems with unexpectly truncated files at 4G * Pass up errors in nfs_writerpc correctly
1.36	15-Jul-1997	fvdl	A filesystem may not support VFS_VGET (like msdosfs). If it doesn't, the server code would always skip all dir entries for a readdirplus operation. To avoid endlessly retrying clients, try VFS_VGET first, and it it fails, return NFSERR_NOTSUPP so that client will fall back to normal readdir operations.
1.35	24-Jun-1997	fvdl	Provide the extra arg to nfsrv_fhtovp, signalling if we're dealing with a request on the public filehandle. Extend the lookup operation to support WebNFS, including index file support (URL style). Yucky, it's optional in the spec, but Solaris 2.6 will support it, so..
1.34	12-May-1997	fvdl	In nfsrvw_coalesce, make sure the coalesce list from the nfsd is moved as well. This fixes client hangs. (from Naofumi Honda <honda@Kururu.math.sci.hokudai.ac.jp> / NetBSD-pc98)
1.33	08-May-1997	mycroft	Pass the vnode type to vaccess(), and use it when checking VEXEC. Make sure that the mode bits passed to vaccess() and returned by foo_getattr() contain only permission bits.
1.32	08-May-1997	mycroft	VEXEC -> VLOOKUP, as appropriate.
1.31	22-Feb-1997	fvdl	Fixes from BSDI (thanks go to Keith Bostic). Original RCS message: date: 1996/11/20 20:02:54; author: pjd; state: Exp; lines: +7 -4 In nfsrv_access(), if VOP_ACCESS() returns an error and the error == EPERM or its not the owner doing the access, return the error.
1.30	10-Feb-1997	fvdl	Move vnode_pager_uncache to a better spot in nfsrv_remove. Also use it in nfsrv_rename, if the 2nd argument is an existing file and will thus be removed.
1.29	31-Jan-1997	fvdl	branches: 1.29.2; nfsrv_readdirplus also suffered from the off-by-one loop problem; fix it too.
1.28	31-Jan-1997	fvdl	Fix order error in loop condition which could cause a crash in nfsrv_readdir(). Fixes PR #3170
1.27	11-Dec-1996	fvdl	Give permission to the owner of the file to preserve semantics only in the relevant cases (read, write). Fixes PR 3017.
1.26	01-Jul-1996	fvdl	Always call vnode_pager_uncache when removing a file in the server (same as in sys_unlink()).
1.25	02-Mar-1996	jtk	branches: 1.25.4; Do not return whiteout directory entries in NFS readdir replies. (The NFS protocol doesn't know how to deal with them properly, yet.)
1.24	20-Feb-1996	cgd	Third argument to VOP_PATHCONF is a register_t *, and register_t may be different than 'int'. Do the right thing when declaring variables which are used this way.
1.23	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.22	09-Feb-1996	christos	nfs prototype changes
1.21	09-Feb-1996	mycroft	Fix vop_link, vop_symlink, and vop_remove semantics in several ways: * Change the argument names to vop_link so they actually make sense. * Implement vop_link and vop_symlink for all file systems, so they do proper cleanup. * Require the file system to decide whether or not linking and unlinking of directories is allowed, and disable it for all current file systems. (Also, remove the cross-device link check, that was moved into the file systems some time ago.)
1.20	01-Feb-1996	jtc	Rename struct timespec fields to conform to POSIX.1b
1.19	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.18	23-May-1995	mycroft	Remove gratuitous extra indirections.
1.17	13-Dec-1994	mycroft	Sync with CSRG.
1.16	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.15	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.14	21-Apr-1994	cgd	Convert mount, vnode, and buf structs to use <sys/queue.h>. Also, some knf and structure frobbing to do along with it.
1.13	10-Apr-1994	cgd	patchkit date deletions!
1.12	12-Mar-1994	cgd	fix rcs id
1.11	09-Mar-1994	ws	Make FFS optional
1.10	18-Dec-1993	mycroft	Canonicalize all #includes.
1.9	26-Nov-1993	ws	Bug fixes to ISOFS
1.8	07-Sep-1993	ws	branches: 1.8.2; Changes to VFS readdir semantics NFS changes for better cookie support ISOFS changes for better Rockridge support and support for generation numbers
1.7	03-Sep-1993	jtc	Include systm.h to get prototypes (and possibly inlines) of *max functions.
1.6	16-Jul-1993	cgd	ANSI mods. (originally committed by andrew on 1993/06/27 06:58:35)
1.5	16-Jul-1993	cgd	fix for macklem's bogus use of the va_flags field, supplied by John Woods, jfwfrom: @ksr.com. also, fixes the following problems: the va_gen field is in a similar position (Suns are going to be reporting the change-date microseconds as their "generation"), I've supplied my own set of diffs below for your inspection. Note these aren't even compiled, but they're pretty similar to what I had to do to our older version of OSF/1 here. (There's also an unrelated change supplied for xdr_subs.h; the pointer types supplied to the fxdr_time() and txdr_time() macros are not, in fact, both struct timevals. That turns out to be one of many tips-of-the-iceberg facing those porting the (old) Berkeley NFS code to 64-bit machines...) (originally committed by cgd on 1993/06/03 01:12:42)
1.4	16-Jul-1993	cgd	more rcs id adding and header cleanup. i like vi macros! (originally committed by cgd on 1993/05/20 03:18:44)
1.3	10-Apr-1993	glass	migrated code to make split possible
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
1.1.1.4	01-Mar-1998	fvdl	Import some files that were changed after Lite2
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.8.2.3	26-Nov-1993	mycroft	Merge changes from trunk.
1.8.2.2	14-Nov-1993	mycroft	Canonicalize all #includes.
1.8.2.1	24-Sep-1993	mycroft	Make all files using spl*() #include cpu.h. Changes from trunk. nfs_vfsops.c, nfsmount.h: Make nfs_quotactl() take an int rather than a uid_t, as it might be -1. nfs_vnops.c: va_size and va_bytes are now quads.
1.25.4.2	04-Mar-1997	mycroft	Pull up bug fixes from -current, per fvdl.
1.25.4.1	11-Dec-1996	mycroft	From trunk: Always call vnode_pager_uncache() when removing a file.
1.29.2.1	12-Mar-1997	is	Merge in changes from Trunk
1.37.2.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.38.2.2	01-Feb-1999	cgd	pull up rev 1.46 from trunk (mrg)
1.38.2.1	22-Dec-1997	perry	pullup a fix to a critical NFS bug (fvdl)
1.50.2.2	16-Dec-1999	he	Pull up revision 1.52 (requested by fvdl): Correct length check in readdirplus, making 32k readdir sizes work.
1.50.2.1	04-May-1999	perry	branches: 1.50.2.1.2; pullup 1.50->1.51 (sommerfeld)
1.50.2.1.2.2	11-Jul-1999	chs	remove uvm_vnp_uncache(), it's no longer needed.
1.50.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.51.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.51.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.51.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.54.4.1	14-Dec-2000	he	Pull up revision 1.58 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.59.4.3	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
1.59.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.59.4.1	03-Aug-2001	lukem	update to -current
1.59.2.5	11-Dec-2002	thorpej	Sync with HEAD.
1.59.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.59.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.59.2.2	26-Sep-2001	nathanw	Catch up to -current. Again.
1.59.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.60.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.61.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.62.10.5	05-Jun-2004	jdc	Pull up revision 1.89 (requested by yamt in ticket #1706). nfsrv_create: fix an LP64 problem for exclusive create.
1.62.10.4	10-Jul-2003	tron	Pull up revision 1.80 (requested by bouyer in ticket #1373): nfsrv_commit(): return success and don't do anything for requests which starts past the end of the file. This can happen when two clients are writting to the same file. Close PR 21696 by myself, discussed on tech-net in 2003/05 and 2003/06. Issue raised by Chuck Silvers (commit and truncate ops needs to be serialised) still unadressed.
1.62.10.3	30-Sep-2002	lukem	Pull up revision 1.65 (requested by bouyer in ticket #880): In nfs_commit(), sanity-check what we get from network: if we try to fluch past end of file, or if off + cnt overflows a quad_t, flush to end of file.
1.62.10.2	30-Sep-2002	lukem	Pull up revision 1.63 (requested by bouyer in ticket #880): nfsrv_commit(): Properly handle the case cnt == 0, which means "flush to end of file". Calling VOP_FSYNC with start == end triggers a DIAGNOSTIC check. Noticed with NFSv3 Linux clients. OK'd by fvdl.
1.62.10.1	30-Sep-2002	lukem	Pull up revision 1.64 (requested by bouyer in ticket #879): In nfsrv_create(), kill an extra PNBUF_PUT() in the NFSv2 mknod case. The pnbuf has already been freed by VOP_MKNOD. This should have been removed in rev 1.60. Should fix PR 18013, OK'd by fvdl.
1.79.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.79.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.79.2.6	18-Dec-2004	skrll	Sync with HEAD.
1.79.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.79.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.79.2.3	24-Aug-2004	skrll	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.79.2.2	03-Aug-2004	skrll	Sync with HEAD
1.79.2.1	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.
1.87.2.1	05-Jun-2004	jdc	Pull up revision 1.89 (requested by yamt in ticket #445). nfsrv_create: fix an LP64 problem for exclusive create.
1.93.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.93.2.1	29-Apr-2005	kent	sync with -current
1.95.2.12	17-Mar-2008	yamt	sync with head.
1.95.2.11	27-Feb-2008	yamt	drop lazy mapping of mbuf external storage for now, because: - it's currently broken wrt asm code. (cpu_in_cksum) - there are other approaches worth to consider. eg. sf_buf
1.95.2.10	27-Feb-2008	yamt	sync with head.
1.95.2.9	04-Feb-2008	yamt	sync with head.
1.95.2.8	21-Jan-2008	yamt	sync with head
1.95.2.7	07-Dec-2007	yamt	sync with head
1.95.2.6	27-Oct-2007	yamt	sync with head.
1.95.2.5	03-Sep-2007	yamt	sync with head.
1.95.2.4	26-Feb-2007	yamt	sync with head.
1.95.2.3	30-Dec-2006	yamt	sync with head.
1.95.2.2	21-Jun-2006	yamt	sync with head.
1.95.2.1	07-Jul-2005	yamt	nfsrv_read: defer mbuf mapping.
1.98.6.2	18-Nov-2005	yamt	- associate read-ahead context to vnode, rather than file. - revert VOP_READ prototype.
1.98.6.1	15-Nov-2005	yamt	adapt ffs, lfs, nfs.
1.99.2.2	15-Jan-2006	yamt	sync with head.
1.99.2.1	31-Dec-2005	yamt	- adapt nfs. - nfs_doio_read: #if 0 out "killproc if text is modified" part of the code as it's broken. (a process reading the modified text is not necessarily a process which is using the file as a text.)
1.100.4.3	01-Jun-2006	kardel	Sync with head.
1.100.4.2	22-Apr-2006	simonb	Sync with head.
1.100.4.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.100.2.1	09-Sep-2006	rpaulo	sync with head
1.101.6.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.101.6.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.101.4.6	11-May-2006	elad	sync with head
1.101.4.5	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.
1.101.4.4	19-Apr-2006	elad	sync with head.
1.101.4.3	12-Mar-2006	elad	Get rid of NFSW_SAMECRED() that uses memcmp() to compare two credentials, and use a new nfsrv_samecred(), using kauth(9). Note that the NFSW_SAMECRED() macro used to check nd_flag of both descriptors for NB_KERBAUTH too; we don't do that. [documented] Based on code in FreeBSD, thanks to Jeff Roberson.
1.101.4.2	10-Mar-2006	elad	generic_authorize() -> kauth_authorize_generic().
1.101.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.101.2.5	03-Sep-2006	yamt	sync with head.
1.101.2.4	11-Aug-2006	yamt	sync with head
1.101.2.3	26-Jun-2006	yamt	sync with head.
1.101.2.2	24-May-2006	yamt	sync with head.
1.101.2.1	01-Apr-2006	yamt	sync with head.
1.109.2.1	19-Jun-2006	chap	Sync with head.
1.111.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.117.4.2	10-Dec-2006	yamt	sync with head.
1.117.4.1	22-Oct-2006	yamt	sync with head
1.117.2.3	09-Feb-2007	ad	Sync with HEAD.
1.117.2.2	12-Jan-2007	ad	Sync with head.
1.117.2.1	18-Nov-2006	ad	Sync with head.
1.119.2.2	10-Mar-2007	bouyer	Pull up following revision(s) (requested by chs in ticket #506): sys/nfs/nfs_serv.c: revision 1.124 after freeing cookies, set the pointer to NULL to prevent dangling reuse
1.119.2.1	17-Feb-2007	tron	Apply patch (requested by chs in ticket #422): - Fix various deadlock problems with nullfs and unionfs. - Speed up path lookups by upto 25%.
1.123.2.3	15-Apr-2007	yamt	sync with head.
1.123.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.123.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.127.4.1	11-Jul-2007	mjf	Sync with head.
1.127.2.4	26-Aug-2007	yamt	- mark nfssvc(2) MPSAFE and move the most of nfsd out of the kernel lock. - remove unused ns_solock. - remove some of KERNEL_LOCK/UNLOCK which are not necessary on this branch.
1.127.2.3	20-Aug-2007	ad	Sync with HEAD.
1.127.2.2	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.
1.127.2.1	10-Apr-2007	ad	Sync with head.
1.128.4.1	15-Aug-2007	skrll	Sync with HEAD.
1.129.10.2	27-Jul-2007	yamt	stop nfs tick when we have nothing to do.
1.129.10.1	27-Jul-2007	yamt	file nfs_serv.c was added on branch matt-mips64 on 2007-07-27 10:03:59 +0000
1.129.8.1	14-Oct-2007	yamt	sync with head.
1.129.6.3	23-Mar-2008	matt	sync with HEAD
1.129.6.2	09-Jan-2008	matt	sync with HEAD
1.129.6.1	06-Nov-2007	matt	sync with HEAD
1.129.4.3	09-Dec-2007	jmcneill	Sync with HEAD.
1.129.4.2	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.129.4.1	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.
1.130.4.3	18-Feb-2008	mjf	Sync with HEAD.
1.130.4.2	27-Dec-2007	mjf	Sync with HEAD.
1.130.4.1	08-Dec-2007	mjf	Sync with HEAD.
1.131.2.3	26-Dec-2007	ad	Sync with head.
1.131.2.2	08-Dec-2007	ad	Sync with head.
1.131.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.132.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.135.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.135.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.135.2.1	24-Mar-2008	keiichi	sync with head.
1.138.16.3	30-Sep-2012	bouyer	Pull up following revision(s) (requested by chs in ticket #1794): sys/nfs/nfs_serv.c: revision 1.164 via patch fix error handling in nfsrv_rename(): when the first nfs_namei() fails, don't try to free the resources allocated by a successful lookup.
1.138.16.2	14-Feb-2010	bouyer	Pull up following revision(s) (requested by pooka in ticket #1289): sys/sys/namei.src: revision 1.14 sys/kern/vfs_syscalls.c: revision 1.401 sys/nfs/nfs_serv.c: revision 1.149 sys/sys/namei.h: regen Define namei flag INRENAME and set it if a lookup operation is part of rename. This helps with building better asserts for rename in the DELETE lookup ... the RENAME lookup is quite obviously a part of rename.
1.138.16.1	13-Apr-2009	snj	branches: 1.138.16.1.4; Pull up following revision(s) (requested by ad in ticket #700): sys/nfs/nfs_serv.c: revision 1.144 sys/nfs/nfsm_subs.h: revision 1.51 PR kern/41158: nfs_rename() locking against myself nfsrv_rename() can exit without calling genfs_renamelock_exit() because the nfsm_reply() can do return (0) on error. Change nfsm_reply to use 'error = 0; goto nfsmout' instead. Fix a few place so it's safe to goto nfsmout from nfsm_reply, or other macros calling it. As a side effect it could fix a missing vrele(dirp) in various place where nfsm_reply could return(0).
1.138.16.1.4.2	24-Dec-2011	matt	Fix call to sokvaalloc (now takes 3 arguments)
1.138.16.1.4.1	20-Apr-2010	matt	Pullin some NFS fixes from netbsd-5.
1.138.14.2	28-Apr-2009	skrll	Sync with HEAD.
1.138.14.1	19-Jan-2009	skrll	Sync with HEAD.
1.138.12.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.138.6.4	11-Aug-2010	yamt	sync with head.
1.138.6.3	11-Mar-2010	yamt	sync with head
1.138.6.2	20-Jun-2009	yamt	sync with head
1.138.6.1	04-May-2009	yamt	sync with head.
1.138.2.3	27-Dec-2008	christos	merge with head.
1.138.2.2	20-Nov-2008	christos	merge with head.
1.138.2.1	29-Mar-2008	christos	Welcome to the time_t=long long dev_t=uint64_t branch.
1.142.2.2	23-Jul-2009	jym	Sync with HEAD.
1.142.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.150.4.3	21-Apr-2011	rmind	sync with head
1.150.4.2	05-Mar-2011	rmind	sync with head
1.150.4.1	03-Jul-2010	rmind	sync with head
1.150.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.155.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.155.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.161.2.3	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")
1.161.2.2	30-Oct-2012	yamt	sync with head
1.161.2.1	17-Apr-2012	yamt	sync with head
1.162.2.1	18-Feb-2012	mrg	merge to -current.
1.163.4.1	01-Nov-2012	matt	sync with netbsd-6-0-RELEASE.
1.163.2.2	03-Sep-2012	riz	Pull up following revision(s) (requested by christos in ticket #537): sys/nfs/nfs_serv.c: revision 1.165 When unloading the nfsserver module, call nfs_fini() so that the nfsrvdescpl pool gets destroyed. Otherwise we are left with a stray pool that points to unmapped memory behind (and bad things happen). Typically you get seemingly random page faults (without printing uvm_fault) that happen in various pool operations. Most frequent one is the pool_drain() from the page daemon.
1.163.2.1	03-Sep-2012	riz	Pull up following revision(s) (requested by chs in ticket #530): sys/nfs/nfs_serv.c: revision 1.164 fix error handling in nfsrv_rename(): when the first nfs_namei() fails, don't try to free the resources allocated by a successful lookup.
1.165.4.1	18-May-2014	rmind	sync with head
1.165.2.2	03-Dec-2017	jdolecek	update from HEAD
1.165.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.170.6.2	28-Aug-2017	skrll	Sync with HEAD
1.170.6.1	06-Jun-2015	skrll	Sync with HEAD
1.173.10.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.173.10.1	21-May-2018	pgoyette	Sync with HEAD
1.173.4.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #1810): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.174.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.174.2.1	10-Jun-2019	christos	Sync with HEAD
1.177.4.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #1617): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.178.2.1	17-Jan-2020	ad	Sync with head.
1.183.4.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #134): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.203	22-Feb-2025	mlelstv	Poll for interrupted NFS operations if the waiting process hasn't issued the operation and then won't get the interrupt signal itself.
1.202	05-Feb-2024	andvar	branches: 1.202.2; fix various typos in comments.
1.201	09-Apr-2023	riastradh	nfs: Simplify assertion. No functional change intended.
1.200	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!)
1.199	21-Jan-2018	christos	branches: 1.199.2; 1.199.4; PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.198	17-Jun-2016	christos	branches: 1.198.10; Serialize all access to the NFS request queue via splsoftnet(). Fixes random crashes. XXX: Pullup-7
1.197	15-Jul-2015	manu	Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.196	09-May-2015	rtr	change sosend() to accept sockaddr * instead of mbuf * for nam. bump to 7.99.16
1.195	02-May-2015	rtr	make connect syscall use sockaddr_big and modify pr_{send,connect} nam parameter type from buf * to sockaddr . final commit for parameter type changes to protocol user requests bump kernel version to 7.99.15 for parameter type changes to pr_{send,connect}
1.194	03-Apr-2015	rtr	* change pr_bind to accept struct sockaddr * instead of struct mbuf * * update protocol bind implementations to use/expect sockaddr * instead of mbuf * * introduce sockaddr_big struct for storage of addr data passed via sys_bind; sockaddr_big is of sufficient size and alignment to accommodate all addr data sizes received. * modify sys_bind to allocate sockaddr_big instead of using an mbuf. * bump kernel version to 7.99.9 for change to pr_bind() parameter type. Patch posted to tech-net@ http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html The choice to use a new structure sockaddr_big has been retained since changing sockaddr_storage size would lead to unnecessary ABI change. The use of the new structure does not preclude future work that increases the size of sockaddr_storage and at that time sockaddr_big may be trivially replaced. Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
1.193	05-Sep-2014	matt	branches: 1.193.2; Don't use catch as a variable name.
1.192	05-Aug-2014	rtr	branches: 1.192.2; 1.192.4; split PRU_SEND function out of pr_generic() usrreq switches and put into separate functions xxx_send(struct socket , struct mbuf , struct mbuf , struct mbuf , struct lwp *) - always KASSERT(solocked(so)) even if not implemented - replace calls to pr_generic() with req = PRU_SEND with calls to pr_send() rename existing functions that operate on PCB for consistency (and to free up their names for xxx_send() PRUs - l2cap_send() -> l2cap_send_pcb() - sco_send() -> sco_send_pcb() - rfcomm_send() -> rfcomm_send_pcb() patch reviewed by rmind
1.191	18-May-2014	rmind	Add struct pr_usrreqs with a pr_generic function and prepare for the dismantling of pr_usrreq in the protocols; no functional change intended. PRU_ATTACH/PRU_DETACH changes will follow soon. Bump for struct protosw. Welcome to 6.99.62!
1.190	14-Sep-2013	martin	branches: 1.190.2; Backout wildcard pragma to kill warnings and instead sprinkle a few dozen __unused attributes. Requested by joerg@
1.189	23-Mar-2011	tls	branches: 1.189.4; 1.189.14; 1.189.18; As suggested by matt@: change socket buffer reservations for NFS send/receive to 3 times max RPC size rather than 2 times. Avoids nasty TCP stalls observed at Panix. Will require increase to sbmax via sysctl for those running really huge NFS rsize/wsize (>64K).
1.188	17-Dec-2010	yamt	branches: 1.188.2; nfs_rcvunlock: don't wake up all waiters.
1.187	02-Mar-2010	pooka	branches: 1.187.2; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.186	13-Feb-2010	yamt	nfs_msg: #if 0 out tprintf for now and comment why.
1.185	19-Jan-2010	yamt	branches: 1.185.2; nfs_request: fix races which break congestion window and make nfs client stuck.
1.184	31-Dec-2009	christos	appease gcc.
1.183	06-Dec-2009	dyoung	For readability's sake, write NULL instead of (type *)0.
1.182	05-Nov-2009	bouyer	Handle EWOULDBLOCK the same way as EPIPE. It seems the TCP socket layer can return EWOULDBLOCK on some occasion when the connection is broken.
1.181	16-Oct-2009	pooka	If send fails with EMSGSIZE for whatever reason, it's unlikely to succeed no matter how hard we retry. So just fail the request.
1.180	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.179	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.178	04-Feb-2009	ad	branches: 1.178.2; PR kern/40491 5.0: nfs timer can crash/break on smp Hack around it by acquiring softnet_lock around the client-side timer loop.
1.177	21-Jan-2009	yamt	restore the pre socket locking patch signal behaviour. this fixes a busy-loop in nfs_connect.
1.176	18-Jan-2009	mrg	Actually enforce the maximum timeout (60s by default) rather than backing off to 256*SRTT. This is why it sometimes could take hours for a NFS mount to come back when the server returned. contributed anonymously.
1.175	23-Nov-2008	mrg	avoid noisy nfs_timer/nfs_reply DEBUG output that occurs when the NFS server goes away. use ratelimit(9) and only print the console error once every 10 seconds. PR#31562.
1.174	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.173	07-Oct-2008	pooka	branches: 1.173.2; 1.173.4; nuke outdated comment
1.172	30-Sep-2008	pooka	Apply #ifdef modern art to make NFSSERVER-without-NFS possible.
1.171	06-Aug-2008	plunky	Convert socket options code to use a sockopt structure instead of laying everything into an mbuf. approved by core
1.170	24-Apr-2008	ad	branches: 1.170.2; 1.170.4; 1.170.8; Merge the socket locking patch: - Socket layer becomes MP safe. - Unix protocols become MP safe. - Allows protocol processing interrupts to safely block on locks. - Fixes a number of race conditions. With much feedback from matt@ and plunky@.
1.169	10-Apr-2008	yamt	branches: 1.169.2; - make nfs_receive and nfs_reply static. - ansify.
1.168	28-Mar-2008	yamt	whitespace.
1.167	02-Jan-2008	yamt	branches: 1.167.6; use kmem_alloc instead of malloc.
1.166	02-Jan-2008	ad	Merge vmlocking2 to head.
1.165	04-Dec-2007	yamt	branches: 1.165.4; merge non-intrusive nfs changes from vmlocking.
1.164	21-Oct-2007	yamt	branches: 1.164.2; 1.164.4; remove lwp argument from nfs_reconnect and always use &lwp0 because who triggers a reconnect doesn't really matter here. PR/37145.
1.163	05-Aug-2007	yamt	branches: 1.163.2; 1.163.6; 1.163.8; use kpause rather than lbolt.
1.162	02-Aug-2007	yamt	branches: 1.162.2; nfsdsock_unlock: add an assertion.
1.161	27-Jul-2007	yamt	stop nfs tick when we have nothing to do.
1.160	09-Jul-2007	ad	branches: 1.160.2; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.159	30-Jun-2007	dsl	Updates for changes prototype of kauth_cred_set/getgroups().
1.158	22-Jun-2007	yamt	- nfsrv_slpderef: fix a locking botch. (should fix "slp->ns_sref == 0" assertion failures in nfsrv_init) - add some related assertions.
1.157	01-Jun-2007	yamt	use mutex and condvar.
1.156	01-Jun-2007	yamt	nfsdsock_lock: fix an inverted check of SLP_VALID.
1.155	28-May-2007	yamt	- remove nfs_exit exit hook. ok'ed by christos@. - as far as i understand the code, it shouldn't be necessary because nfs_request can't return without removing its request and r->r_lwp is either curlwp or NULL. - even if it's necessary, leaking requests is not the correct way to recover from the condition. - nfs_request: add a related assertion.
1.154	02-May-2007	yamt	nfs_rcvlock: fix NFSMNT_INT check, which has been broken since rev.1.39.
1.153	02-May-2007	yamt	- nfs_reply: keep rcvlock longer so that lwp which already have its reply received won't be stuck in nfs_receive. - nfs_rcvlock: check exceptions before sleeping on the lock. - nfs_rcvunlock: use cv_broadcast rather than cv_signal to ensure that lwps which received its reply get woken up.
1.152	30-Apr-2007	yamt	remove R_GETONEREP.
1.151	29-Apr-2007	yamt	use condvar.
1.150	29-Apr-2007	yamt	use mutex and condver.
1.149	12-Mar-2007	ad	branches: 1.149.2; Pass an ipl argument to pool_init/POOL_INIT to be used when initializing the pool's lock.
1.148	04-Mar-2007	christos	branches: 1.148.2; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.147	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.146	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.145	09-Feb-2007	ad	branches: 1.145.2; Merge newlock2 to head.
1.144	27-Dec-2006	yamt	remove nqnfs.
1.143	06-Dec-2006	yamt	nfs_disconnect: 2 -> SHUT_RDWR. no functional change.
1.142	06-Dec-2006	yamt	nfsrv_rcv: claim ownership of received mbufs.
1.141	09-Nov-2006	yamt	remove some __unused in function parameters.
1.140	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.139	10-Oct-2006	dogcow	change the MOWNER_INIT define to take two args; fix extant struct mowner decls to use it. Makes options MBUFTRACE compile again and not whinge about missing structure declarations. (Also makes initialization consistent.)
1.138	02-Sep-2006	yamt	branches: 1.138.2; 1.138.4; nfsdreq_free: remove an assertion which is not true.
1.137	15-Jul-2006	yamt	nfs_getreq: fix a kauth fallout. pointed by nanashi-san. http://pc8.2ch.net/test/read.cgi/unix/1145181361/786
1.136	30-Jun-2006	yamt	nfs_request: don't bother to handle NFSERR_STALEWRITEVERF because it isn't a real nfs error value.
1.135	07-Jun-2006	kardel	branches: 1.135.2; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.134	28-May-2006	blymn	Clean up bogus whitespace
1.133	28-May-2006	yamt	nfs_request: use kauth_cred_free rather than kauth_cred_destroy.
1.132	19-May-2006	yamt	branches: 1.132.2; - fix compilation problem for !NFSSERVER && NFS. pointed by Tom Spindler on source-changes@. - make nfs_srvdesc_pool static.
1.131	18-May-2006	yamt	- fix some leaks in nfsd, introduced by kauth changes. - simplify code. - add some assertions. - wrap some long lines. - remove an unnecessary ";".
1.130	14-May-2006	elad	integrate kauth.
1.129	10-May-2006	mrg	quell GCC 4.1 uninitialised variable warnings. XXX: we should audit the tree for which old ones are no longer needed after getting the older compilers out of the tree..
1.128	15-Apr-2006	dogcow	#if -> #ifdef
1.127	15-Apr-2006	christos	Coverity CID 734: Define NFS_TEST_HEAVY for testing nfsds, and use this to ifdef out dead code. XXX: Why is this turned on by default?
1.126	01-Mar-2006	rpaulo	branches: 1.126.2; 1.126.4; 1.126.6; Back out revision 1.125 and 1.124. The code for checking if slp->ns_reclen == 0, was already there since "Linux sometimes generates 0-lenght records.". Bad Rui...
1.125	01-Mar-2006	rpaulo	In nfsrv_getstream(), ns_reclen will never be negative due to the previous assignment (recmark & ~0x80000000). Pointed out by Christos.
1.124	01-Mar-2006	rpaulo	From FreeBSD SA-06:10 Correct a remote kernel panic when processing zero-length RPC records via TCP.
1.123	01-Mar-2006	yamt	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.
1.122	03-Jan-2006	yamt	branches: 1.122.2; 1.122.4; nfssvc_nfsd: reduce a chance for a slow peer to capture all our threads. instead of sleeping to wait for the socket to send our reply, just hand-off our reply to the thread which is holding the socket.
1.121	03-Jan-2006	yamt	improve nfsd locking. - don't bother to take nfs_sndlock when doing nfsrv_rcv. unlike client, we never reconnect. - nfsrv_getstream: fix the case that m_split sleeps. - free socket in nfsrv_slpderef rather than nfsrv_zapsock. fix race with nfssvc_nfsd. - while i'm here, remove NFSD_WAITING and NFSD_REQINPROG as they are redundant. - some comments and assertions.
1.120	30-Dec-2005	jmmv	branches: 1.120.2; Avoid dereferencing the lwp parameter in nfs_receive, as it is always NULL. Solves a crash when mounting NFS shares. (The proc parameter used before the conversion to lwp's was NULL too, so the addition of 'l->l_proc' in the code was extra.)
1.119	11-Dec-2005	christos	merge ktrace-lwp.
1.118	21-Nov-2005	yamt	use c99 initializers for proct.
1.117	25-Sep-2005	tron	branches: 1.117.6; Correct typo in last commit to fix compilation error.
1.116	25-Sep-2005	christos	Add missing TIMEDOUT and IO errors.
1.115	25-Sep-2005	christos	Convert from nfs error values to regular errno's. Although most values of nfs errors are chosen to be the same as errno, some of them are not and it is better for portability to do the conversion anyway. Also a server can return a bad error number that can cause the server to crash, because it can have the high bits that are used internally set. This was the case with amd. Finally nfs_request() should return a valid errno, because we can return a bogus value to userland. Thanks to rpaulo for debugging this.
1.114	29-May-2005	christos	branches: 1.114.2; - sprinkle const - avoid shadowed variables - mark bad const use with XXXUNCONST
1.113	29-Mar-2005	yamt	nfsrv_rcv: don't do so_receive from socket upcall context. while there's little benefits, it complicates locking and confuses flow control.
1.112	26-Feb-2005	perry	branches: 1.112.2; nuke trailing whitespace
1.111	17-Sep-2004	skrll	branches: 1.111.4; 1.111.6; There's no need to pass a proc value when using UIO_SYSSPACE with vn_rdwr(9) and uiomove(9). OK'd by Jason Thorpe
1.110	24-Aug-2004	yamt	nfs_request: a workaround for servers doing "maproot". for i/o requests which are expected not to fail due to permission to mimic unix file open semantics (READ, WRITE, COMMIT), try two credentials. namely, the file owner's one and open time one. remember which credential worked in per-file basis and try it first next time to minimize number of retries. ideas from Chuck Silvers. PR/23716 and PR/24987.
1.109	18-Aug-2004	yamt	remove a "proc botch" debug printf. ok'ed by Jonathan Stone.
1.108	24-Jun-2004	jonathan	Rename MBUFTRACE helper function m_claim() to m_claimm(), for consistency with M_FREE() and m_freem(). Affected files: sys/mbuf.h kern/uipc_socket2.c kern/uipc_mbuf.c net/if_ethersubr.c netatalk/ddp_input.c nfs/nfs_socket.c
1.107	24-May-2004	jonathan	Change DIAGNOSTIC warning in nfs_send() about NULL rep->r_procp: the warning is triggered pervasively, so print it only once per boot. (The callers who pass NULL r_procps should soon be fixed to pass a valid struct proc* ).
1.106	23-May-2004	yamt	- for tcp, use SO_RCVTIMEO to recover from server crash. otherwise we can be stuck in soreceive forever. the problem is pointed by Minoura Makoto. PR/25662 - clear r_rexmit on reconnect and clear r_rtt and R_TIMING on retransmit so that the above (and soft mounts) happy.
1.105	22-May-2004	jonathan	Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.104	10-May-2004	yamt	don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.103	21-Apr-2004	christos	Replace the statfs() family of system calls with statvfs(). Retain binary compatibility.
1.102	17-Mar-2004	yamt	branches: 1.102.2; nfs_sndlock: fix nfsd null dereference.
1.101	10-Mar-2004	matt	Don't report EPIPE errors on nfs sockets. These can be due to idle tcp mounts which will be closed by netapp, solaris, etc. if left idle too long.
1.100	07-Dec-2003	fvdl	Unix semantics dictate that access checks for files are done when it is opened. An open file can always be read from and/or written to, depending on how it was opened. Therefore, the read/write/commit RPCs should never return EACCESS, as they are only performed on files that have been successfully opened already. This change improves the current situation and works in most cases. It simply always uses the most recently known owner/group of the file, iff the authentication mechanism is AUTH_UNIX (in other cases, the creds for a succesful open are used, but note that no other cases are currently implemented). A retry mechanism can be used to catch a few more cases, but this is a good improvement for now.
1.99	09-Oct-2003	yamt	for nfs_timer_ch, use callout_schedule rather than callout_reset as the former is a little more efficient.
1.98	16-Aug-2003	yamt	use sizeof() instead of a hardcorded constant.
1.97	16-Aug-2003	yamt	current trylater/jukebox retry delay is way too long and it has a bug in the backoff calculation. so, - clip it to 1-60 sec. (suggested by Rick Macklem) - use a constant multiplier instead of nfs_backoff, which is already exponential. - move some related constant definations to nfs.h from nqnfs.h and prefix with NFS_ instead of NQ_ because they are not nqnfs-specific.
1.96	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.95	23-Jul-2003	yamt	when rexmitting a request due to NFSERR_JUKEBOX, use a new xid as RFC1813 says.
1.94	23-Jul-2003	yamt	fix parenthesis mismatch in rev.1.93.
1.93	23-Jul-2003	yamt	use sizeof() instead of hardcoding the size of the array.
1.92	29-Jun-2003	fvdl	branches: 1.92.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.91	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
1.90	25-Jun-2003	yamt	- instead of scaning a list when looking up {a idle thread, a socket with pending requests}, maintain dedicated list of them. - add spin locks.
1.89	23-Jun-2003	martin	Make sure to include opt_foo.h if a defflag option FOO is used.
1.88	22-May-2003	yamt	poolify nfsrv_descript.
1.87	22-May-2003	yamt	interlock for nfs_rcvlock.
1.86	21-May-2003	yamt	- use FREE not free for MALLOC'ed memory. - remove unneeded caddr_t casts.
1.85	21-May-2003	yamt	remove local definitions of TRUE and FALSE.
1.84	21-May-2003	yamt	indent
1.83	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.82	15-Apr-2003	yamt	fix indent.
1.81	03-Apr-2003	yamt	use m_copydata and m_split instead of similar inlined ones.
1.80	02-Apr-2003	yamt	use queue manipulation macros.
1.79	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.78	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.77	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.76	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.75	02-Aug-2002	fvdl	Initialize recm to NULL inside the loop, so that a record length of NULL will not accidentallly append bogus data (the previous record). Derived from a fix by Matt Dillon in FreeBSD.
1.74	12-May-2002	matt	branches: 1.74.2; 1.74.4; Eliminate commons
1.73	17-Mar-2002	christos	use the exithook mechanism to remove the exiting process from the list of processes to be signalled in a soft mount.
1.72	27-Feb-2002	lukem	nfs_connect(): if NFSMNT_RESVPORT is set, set IP_PORTRANGE_LOW on the socket rather than using home-grown code to find a free reserved socket. this also results in nfs pcb's having the INP_ANONPORT and INP_LOWPORT flags set, which is useful for netstat(1) to know.
1.71	22-Jan-2002	minoura	Back out the previous. It was my misreading from the lack of mbuf usage... Sorry for the mess.
1.70	21-Jan-2002	minoura	Correctly write back the updated value of the local variable to the struct nfssvc_sock. Affected only when a recordmark of RPC over TCP is fragmented to multiple mbufs. I do not know whether this code has ever been executed :)
1.69	10-Nov-2001	lukem	add RCSIDs
1.68	13-Oct-2001	simonb	branches: 1.68.2; Don't initialise the 5th element of some 4 element arrays.
1.67	09-May-2001	fvdl	branches: 1.67.2; Suppress another case of a potentially noisy error message which isn't fatal.
1.66	21-Feb-2001	jdolecek	branches: 1.66.2; make some more constant arrays 'const'
1.65	27-Dec-2000	jdolecek	update commented out code to recent changes of signal structures
1.64	27-Dec-2000	bjh21	Extra diagnostic assertion: subtle pmap bugs can ultimately lead to trying to use NULL credentials for NFS ops, so spot them before we dereference them.
1.63	10-Dec-2000	fvdl	Make sobind() take a struct proc *. It already took curproc and passed it down to the appropriate usrreq function, and this allows usage for contexts that need to be explicitly different from curproc (like in the NFS code when binding to a reserved port).
1.62	27-Sep-2000	fvdl	Avoid unused variables for V2_ONLY case.
1.61	19-Sep-2000	bjh21	Extend NFS_V2_ONLY to remove NQNFS lease support as well. Saves another 10k.
1.60	19-Sep-2000	fvdl	"ENOBUF" on socket writes isn't really fatal; we may just be too fast for the driver. Don't log the error, just try again. Could try to be smart and do a backoff, but it's probably not worth the trouble.
1.59	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.58	27-Jun-2000	mrg	remove include of <vm/vm.h>
1.57	09-Jun-2000	fvdl	branches: 1.57.2; Some tweaks to enable NFS over IPv6. The special-casing of AF_INET should really be removed.
1.56	27-May-2000	thorpej	branches: 1.56.2; sleep() -> tsleep()
1.55	30-Mar-2000	augustss	Remove register declarations.
1.54	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.
1.53	29-Aug-1999	sommerfeld	branches: 1.53.2; Fix overzealous DIAGNOSTIC check in nfs_disconnect() (fix pr8249, 8288)
1.52	30-Jul-1999	fvdl	Don't try to copy an mbuf that may have been freed in case of an error.
1.51	04-Jul-1999	sommerfeld	kern/5591: Fix race in the NFS socket code during umount -f and system shutdown: During an unmount, wake up all the processes which are waiting to lock the socket for receive, and wait for them (and the process blocked in soreceive, if any) to go away before blowing away the socket and the mount structure.
1.50	06-Mar-1999	fair	branches: 1.50.2; 1.50.4; Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.49	12-Feb-1999	thorpej	Fix printf format warnings on Alpha.
1.48	12-Nov-1998	fvdl	Use different names for the "nfscon" label to tsleep(), so that it can be seen in which one a process is sleeping.
1.47	11-Sep-1998	mycroft	Substantial signal handling changes: * Increase the size of sigset_t to accomodate 128 signals -- adding new versions of sys_setprocmask(), sys_sigaction(), sys_sigpending() and sys_sigsuspend() to handle the changed arguments. * Abstract the guts of sys_sigaltstack(), sys_setprocmask(), sys_sigaction(), sys_sigpending() and sys_sigsuspend() into separate functions, and call them from all the emulations rather than hard-coding everything. (Avoids uses the stackgap crap for these system calls.) * Add a new flag (p_checksig) to indicate that a process may have signals pending and userret() needs to do the full (slow) check. * Eliminate SAS_ALTSTACK; it's exactly the inverse of SS_DISABLE. * Correct emulation bugs with restoring SS_ONSTACK. * Make the signal mask in the sigcontext always use the emulated mask format. * Store signals internally in sigaction structures, rather than maintaining a bunch of little sigsets for each SA_* bit. * Keep track of where we put the signal trampoline, rather than figuring it out in _sendsig(). Issue a warning when a non-emulated sigaction bit is observed. * Add missing emulated signals, and a native SIGPWR (currently not used). * Implement the `not reset when caught' semantics for relevant signals. Note: Only code touched by the i386 port has been modified. Other ports and emulations need to be updated.
1.46	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.45	20-Jul-1998	fvdl	Avoid possibly overflowing an mbuf. From Dan S. Decasper, via Chuck Cranor.
1.44	25-Jun-1998	thorpej	defopt NFSSERVER
1.43	25-Apr-1998	matt	Adapt to new sosend/soreceive and upcall (now down in sowakeup)
1.42	19-Feb-1998	thorpej	Include the NFS option header.
1.41	30-Jan-1998	fvdl	Only take the receive lock before disconnecting when doing it from nfs_decode_args. Otherwise we might just end up locking against ourselves. XXX workaround, will do ok for now. Proper fix forthcoming.
1.40	16-Nov-1997	fvdl	Make sure the receive lock is taken when disconnecting a socket. Also change a check for a 'connected' socket to use the socket rather than the mount flags. From Matthias Drochner.
1.39	10-Oct-1997	fvdl	branches: 1.39.2; * New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.38	22-May-1997	gwr	branches: 1.38.4; 1.38.6; Temporary work-around for PR kern/3579 (from Jonathan Stone).
1.37	12-May-1997	fvdl	* If nfs_reconnect fails, be sure to release the sndlock, otherwise no other requests will get through and the mount point will be effectively dead. This could happen for mounts using TCP and -i and/or -s. * Reserve enough space for UDP sockets. Fixes PR 3008, from Naofumi Honda.
1.36	08-Apr-1997	fvdl	Avoid nfsiods acquiring/releasing a lock, then acquiring it again, before anyone else can get to it, by checking if a reply was received, and it has thus become unnecessary to take the lock. From FreeBSD. XXX I don't really like this, "locks" potentially suffer from the same problem throughout the whole kernel; they should probably be FIFO everywhere.
1.35	22-Feb-1997	fvdl	Fixes from BSDI (thanks go to Keith Bostic). Original RCS messages: date: 1995/11/30 20:37:03; author: cp; state: Exp; lines: +25 -14 Change splsoftclock() to splnet(); Make nfsrv_getstream create two copies of data when splitting up an mbuf rather than two references to the same external buffer. The symptom this fixes is client hangs. date: 1997/02/10 18:41:13; author: cp; state: Exp; lines: +4 -1 Make nfs_realign go away on sparc and add functionality to nfsm_disct.
1.34	09-Feb-1997	fvdl	* Fix some bugs in NQNFS (malformed RPC requests, no directory lease eviction) * Avoid possible NULL ptr ref in nfs_reply * Don't ever try to sillyrename directories (from FreeBSD)
1.33	04-Feb-1997	fvdl	branches: 1.33.2; * Make sure a new socket is created when switching to/from NOCONN with a mount * Add extra printf statements to hopefully get some more info on lockups, specifically when a send error is ignored.
1.32	31-Jan-1997	thorpej	NFSCLIENT -> NFS.
1.31	13-Oct-1996	christos	branches: 1.31.2; revert kprintf changes
1.30	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.29	02-Jul-1996	fvdl	Remove bogus check on record length < NFS_MINPACKET. (From Guy Harris via Rick Macklem).
1.28	22-May-1996	mycroft	Pass a proc pointer down to the usrreq and pcbbind functions for PRU_ATTACH, PRU_BIND and PRU_CONTROL. The usrreq interface really needs to be split up, but this will have to wait. Remove SS_PRIV completely.
1.27	15-Apr-1996	thorpej	branches: 1.27.4; Make this compile again on a SPARC if NFSCLIENT is defined without NFSSERVER. (-Wall unused variable lossage)
1.26	25-Feb-1996	fvdl	Oops. Do previous fix on the right line this time.. (thanks Charles)
1.25	25-Feb-1996	fvdl	Call soreserve() with the right size for receives (from pk).
1.24	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.23	09-Feb-1996	christos	nfs prototype changes
1.22	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.21	13-Aug-1995	mycroft	splnet --> splsoftnet
1.20	02-Jun-1995	mycroft	Fix more off by one errors.
1.19	02-Jun-1995	mycroft	Fix another off by one error.
1.18	02-Jun-1995	mycroft	Imported group list now starts at offset 0, not 1.
1.17	17-Aug-1994	mycroft	Convert some more lists and queues.
1.16	17-Aug-1994	mycroft	Change the reply list to a TAILQ.
1.15	29-Jun-1994	cgd	branches: 1.15.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.14	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.13	24-May-1994	cgd	MIN -> min, MAX -> max
1.12	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.
1.11	10-Apr-1994	cgd	patchkit date deletions!
1.10	22-Dec-1993	cgd	minor cleanup
1.9	18-Dec-1993	mycroft	Canonicalize all #includes.
1.8	07-Sep-1993	ws	branches: 1.8.2; Changes to VFS readdir semantics NFS changes for better cookie support ISOFS changes for better Rockridge support and support for generation numbers
1.7	06-Sep-1993	mycroft	Make nfs_timer() return void.
1.6	03-Sep-1993	jtc	Include systm.h to get prototypes (and possibly inlines) of *max functions.
1.5	22-May-1993	cgd	add include of select.h if necessary for protos, or delete if extraneous
1.4	18-May-1993	cgd	make kernel select interface be one-stop shopping & clean it all up.
1.3	10-Apr-1993	glass	migrated code to make split possible
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
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.8.2.2	14-Nov-1993	mycroft	Canonicalize all #includes.
1.8.2.1	24-Sep-1993	mycroft	Make all files using spl*() #include cpu.h. Changes from trunk. nfs_vfsops.c, nfsmount.h: Make nfs_quotactl() take an int rather than a uid_t, as it might be -1. nfs_vnops.c: va_size and va_bytes are now quads.
1.15.2.1	19-Aug-1994	mycroft	update from trunk
1.27.4.4	04-Mar-1997	mycroft	Pull up bug fixes from -current, per fvdl.
1.27.4.3	11-Dec-1996	mycroft	From trunk: Eliminate SS_PRIV; instead, pass down a proc pointer to the usrreq methods that need it. Fix numerous memory leaks and bogus return values.
1.27.4.2	10-Jul-1996	jtc	Patch from frank needed to compile without corresponding network changes
1.27.4.1	08-Jul-1996	jtc	Pulled up from rev 1.29 by request from Frank van der Linden
1.31.2.1	14-Jan-1997	thorpej	Snapshot of work-in-progress, committed to private branch. These changes implement machine-independent root device and file system selection. Notable features: - All ports behave in a consistent manner regarding root device selection. - No more "options GENERIC"; all kernels have the ability to boot with RB_ASKNAME to select root device and file system type. - Root file system type can be wildcarded; a machine-independent function will try all possible file systems for the selected root device until one succeeds. - If the root file system fails to mount, the operator will be given the chance to select a new root device and file system type, rather than having the machine simply panic. - nfs_mountroot() no longer panics if any part of the NFS mount process fails; it now returns an error, giving the operator a chance to recover. - New, more consistent, config(8) grammar. The constructs: config netbsd swap generic config netbsd root on nfs have been replaced with: config netbsd root on ? type ? config netbsd root on ? type nfs Additionally, the operator may select or wildcard root file system type in the kernel configuration file: config netbsd root on cd0a type cd9660 config(8) now requires that a "root" specification be made. "root" may be wired down or wildcarded. "swap" and "dump" specifications are optional, and follow previous semantics. - config(8) has a new "file-system" keyword, used to configure file systems into the kernel. Eventually, this will be used to generate the default vfssw[]. - "options NFSCLIENT" is obsolete, and is replaced by "file-system NFS". "options NFSSERVER" still exists, since NFS server support is independent of the NFS file system client. - sys/arch/<foo>/<foo>/swapgeneric.c is no longer used, and will be removed; all information is now generated by config(8). As of this commit, all ports except arm32 have been updated to use the new setroot(). Only SPARC, i386, and Alpha ports have been tested at this time. Port masters should test these changes on their ports, and report any problems back to me. More changes are on their way, including RB_ASKNAME support in nfs_mountroot() (to prompt for server address and path) and, potentially, the ability to select rarp/bootparam or bootp in nfs_mountroot().
1.33.2.1	12-Mar-1997	is	Merge in changes from Trunk
1.38.6.1	08-Sep-1997	thorpej	Significantly restructure the way signal state for a process is stored. Rather than using bitmasks to redundantly store the information kept in the process's sigacts (because the sigacts was kept in the u-area), hang sigacts directly off the process, and access it directly. Simplify signal setup code tremendously by storing information in the sigacts as an array of struct sigactions, rather than in a different format, since userspace uses sigactions. Make sigacts sharable by adding reference counting.
1.38.4.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.39.2.2	07-Feb-1998	mellon	Pull up 1.41 (fvdl)
1.39.2.1	17-Nov-1997	thorpej	Sync w/ trunk (fvdl)
1.50.4.1	02-Aug-1999	thorpej	Update from trunk.
1.50.2.2	09-Nov-1999	he	Pull up revisions 1.52-1.53 (requested by fvdl): Fix overzealous DIAGNOSTIC check in nfs_disconnect() and don't try to copy a possibly freed mbuf. Fixes PR#8249, PR#8288 and PR#8766.
1.50.2.1	05-Nov-1999	cgd	pull up rev 1.51 from trunk (requested by fvdl): Avoid a panic when forcibly unmounting a hung NFS mount, e.g. at reboot.
1.53.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.53.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.53.2.2	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.53.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.56.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.57.2.3	04-Sep-2002	itojun	pullup sys/nfs/nfs_socket.c 1.75 (fvdl) Initialize recm to NULL inside the loop, so that a record length of NULL will not accidentallly append bogus data (the previous record). Derived from a fix by Matt Dillon in FreeBSD.
1.57.2.2	15-Dec-2000	he	Pull up revision 1.63 (requested by fvdl): Fix NFS+tcp client hangs on server or network outage. Again, please note that this introduces yet another kernel interface change: sobind() gains an argument.
1.57.2.1	14-Dec-2000	he	Pull up revision 1.60 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.66.2.12	11-Dec-2002	thorpej	Sync with HEAD.
1.66.2.11	18-Oct-2002	nathanw	Catch up to -current.
1.66.2.10	13-Aug-2002	nathanw	Catch up to -current.
1.66.2.9	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.66.2.8	24-Jun-2002	nathanw	Curproc->curlwp renaming. Change uses of "curproc->l_proc" back to "curproc", which is more like the original use. Bare uses of "curproc" are now "curlwp". "curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL) so that it is always safe to reference curproc (dereferencing curproc is another story, but that's always been true).
1.66.2.7	20-Jun-2002	nathanw	Catch up to -current.
1.66.2.6	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.66.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.66.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.66.2.3	22-Oct-2001	nathanw	Catch up to -current.
1.66.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.66.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.67.2.6	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
1.67.2.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.67.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.67.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.67.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.67.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.68.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.74.4.1	02-Aug-2002	lukem	Pull up revision 1.75 (requested by fvdl in ticket #604): Initialize recm to NULL inside the loop, so that a record length of NULL will not accidentallly append bogus data (the previous record). Derived from a fix by Matt Dillon in FreeBSD.
1.74.2.1	29-Aug-2002	gehenna	catch up with -current.
1.92.2.12	11-Dec-2005	christos	Sync with head.
1.92.2.11	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.92.2.10	01-Apr-2005	skrll	Sync with HEAD.
1.92.2.9	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.92.2.8	07-Feb-2005	skrll	Don't deref a NULL struct lwp *.
1.92.2.7	21-Sep-2004	skrll	Fix the sync with head I botched.
1.92.2.6	18-Sep-2004	skrll	Sync with HEAD.
1.92.2.5	03-Sep-2004	skrll	Sync with HEAD
1.92.2.4	25-Aug-2004	skrll	Sync with HEAD.
1.92.2.3	18-Aug-2004	skrll	Revert to passing struct proc for {exit,exec}hook.
1.92.2.2	03-Aug-2004	skrll	Sync with HEAD
1.92.2.1	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.
1.102.2.4	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.102.2.3	30-Aug-2004	tron	branches: 1.102.2.3.2; Pull up revision 1.110 via patch (requested by yamt in ticket #803): nfs_request: a workaround for servers doing "maproot". for i/o requests which are expected not to fail due to permission to mimic unix file open semantics (READ, WRITE, COMMIT), try two credentials. namely, the file owner's one and open time one. remember which credential worked in per-file basis and try it first next time to minimize number of retries. ideas from Chuck Silvers. PR/23716 and PR/24987.
1.102.2.2	14-Jul-2004	tron	Pull up revision 1.108 (requested by jonathan in ticket #648): Rename MBUFTRACE helper function m_claim() to m_claimm(), for consistency with M_FREE() and m_freem(). Affected files: sys/mbuf.h kern/uipc_socket2.c kern/uipc_mbuf.c net/if_ethersubr.c netatalk/ddp_input.c nfs/nfs_socket.c
1.102.2.1	10-Jul-2004	tron	Pull up revision 1.106 via patch (requested by yamt in ticket #617): - for tcp, use SO_RCVTIMEO to recover from server crash. otherwise we can be stuck in soreceive forever. the problem is pointed by Minoura Makoto. PR/25662 - clear r_rexmit on reconnect and clear r_rtt and R_TIMING on retransmit so that the above (and soft mounts) happy.
1.102.2.3.2.1	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.111.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.111.4.1	29-Apr-2005	kent	sync with -current
1.112.2.2	15-Dec-2005	tron	Pull up following revision(s) (requested by christos in ticket #1055): sys/nfs/nfs_socket.c: revision 1.115 via patch Convert from nfs error values to regular errno's. Although most values of nfs errors are chosen to be the same as errno, some of them are not and it is better for portability to do the conversion anyway. Also a server can return a bad error number that can cause the server to crash, because it can have the high bits that are used internally set. This was the case with amd. Finally nfs_request() should return a valid errno, because we can return a bogus value to userland. Thanks to rpaulo for debugging this.
1.112.2.1	04-Apr-2005	tron	Pull up revision 1.113 (requested by yamt in ticket #89): nfsrv_rcv: don't do so_receive from socket upcall context. while there's little benefits, it complicates locking and confuses flow control.
1.114.2.9	27-Feb-2008	yamt	revert incomplete nfs client locking for now.
1.114.2.8	15-Feb-2008	yamt	- sprinkle some locks. - disable MNT_UPDATE because it involves too much locking headache. - don't overwrite other bits in v_vflags when setting VV_ROOT.
1.114.2.7	21-Jan-2008	yamt	sync with head
1.114.2.6	07-Dec-2007	yamt	sync with head
1.114.2.5	27-Oct-2007	yamt	sync with head.
1.114.2.4	03-Sep-2007	yamt	sync with head.
1.114.2.3	26-Feb-2007	yamt	sync with head.
1.114.2.2	30-Dec-2006	yamt	sync with head.
1.114.2.1	21-Jun-2006	yamt	sync with head.
1.117.6.1	22-Nov-2005	yamt	sync with head.
1.120.2.2	15-Jan-2006	yamt	sync with head.
1.120.2.1	31-Dec-2005	yamt	- adapt nfs. - nfs_doio_read: #if 0 out "killproc if text is modified" part of the code as it's broken. (a process reading the modified text is not necessarily a process which is using the file as a text.)
1.122.4.3	01-Jun-2006	kardel	Sync with head.
1.122.4.2	22-Apr-2006	simonb	Sync with head.
1.122.4.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.122.2.1	09-Sep-2006	rpaulo	sync with head
1.126.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.126.4.12	11-May-2006	elad	sync with head
1.126.4.11	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.
1.126.4.10	19-Apr-2006	elad	sync with head.
1.126.4.9	14-Apr-2006	elad	Plug some possible leaks of kauth_cred_t.
1.126.4.8	13-Apr-2006	elad	Add a missing kauth_cred_alloc() and remove some misleading code and comment.
1.126.4.7	31-Mar-2006	elad	oops. fix function names... pointed out by yamt@, thanks!
1.126.4.6	31-Mar-2006	elad	fill up real/saved [ug]ids too.
1.126.4.5	14-Mar-2006	elad	Use kauth_cred_[sg]etgroups() where appropriate.
1.126.4.4	10-Mar-2006	elad	Cleanup more interface abuse. Make nfsrv_setcred() take a kauth_cred_t * as outcred. The original code just modified it directly; we can't do that, nor do we want to. Get rid of another case of kauth_cred_zero() followed by kauth_cred_hold() and use kauth_cred_clone() to make sure we don't leave out important members. Add another DIAGNOSTIC check for reference count of above one. Again, this should be tested.
1.126.4.3	10-Mar-2006	elad	Okay, what I've done here is pretty bogus, and trying to use kauth_cred_t as something it's not. Then again, it was part of a fast sweep so I have good excuses. :) DON'T kauth_cred_zero() and then kauth_cred_hold(); that's guaranteed to trip over trying to lock an uninitialized lock. Also, kauth_cred_t now contains more than just a struct ucred, so treat it properly. Use a call to kauth_cred_copy() to ensure we have reference count of one, and sprinkle some DIAGNOSTIC check to let us know if we ever leak memory. Also, don't forget to set proper values for the real/saved user- and group-ids. This should be tested at some point...
1.126.4.2	10-Mar-2006	elad	Remove some #if 0'd code. There's no need to call kauth_cred_setngroups() here because right above we use kauth_cred_addgroup() that does the management of ngroups for us. Also, for the same reason, there's no need to call nfsrvw_sort(), because we are guaranteed to have the group list sorted at all times.
1.126.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.126.2.4	03-Sep-2006	yamt	sync with head.
1.126.2.3	11-Aug-2006	yamt	sync with head
1.126.2.2	26-Jun-2006	yamt	sync with head.
1.126.2.1	24-May-2006	yamt	sync with head.
1.132.2.2	19-Jun-2006	chap	Sync with head.
1.132.2.1	19-May-2006	chap	file nfs_socket.c was added on branch chap-midi on 2006-06-19 04:10:37 +0000
1.135.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.138.4.2	10-Dec-2006	yamt	sync with head.
1.138.4.1	22-Oct-2006	yamt	sync with head
1.138.2.3	12-Jan-2007	ad	Sync with head.
1.138.2.2	18-Nov-2006	ad	Sync with head.
1.138.2.1	21-Oct-2006	ad	Update for sigpending1() change.
1.145.2.4	07-May-2007	yamt	sync with head.
1.145.2.3	24-Mar-2007	yamt	sync with head.
1.145.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.145.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.148.2.10	01-Nov-2007	yamt	fix a compilation problem w/o NFSSERVER. reported by Juan RP via Andrew Doran.
1.148.2.9	27-Aug-2007	yamt	- fix/add assertions. - fix numnfsrvcache.
1.148.2.8	26-Aug-2007	yamt	- mark nfssvc(2) MPSAFE and move the most of nfsd out of the kernel lock. - remove unused ns_solock. - remove some of KERNEL_LOCK/UNLOCK which are not necessary on this branch.
1.148.2.7	20-Aug-2007	ad	Sync with HEAD.
1.148.2.6	15-Jul-2007	ad	Sync with head.
1.148.2.5	01-Jul-2007	ad	Adapt to callout API change.
1.148.2.4	09-Jun-2007	ad	Sync with head.
1.148.2.3	08-Jun-2007	ad	Sync with head.
1.148.2.2	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.148.2.1	13-Mar-2007	ad	Sync with head.
1.149.2.1	11-Jul-2007	mjf	Sync with head.
1.160.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.162.2.3	09-Dec-2007	jmcneill	Sync with HEAD.
1.162.2.2	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.
1.162.2.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.163.8.2	05-Aug-2007	yamt	use kpause rather than lbolt.
1.163.8.1	05-Aug-2007	yamt	file nfs_socket.c was added on branch matt-mips64 on 2007-08-05 09:40:40 +0000
1.163.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.163.2.2	09-Jan-2008	matt	sync with HEAD
1.163.2.1	06-Nov-2007	matt	sync with HEAD
1.164.4.4	29-Dec-2007	yamt	to prepare merge, put nfsd back under kernel_lock for now.
1.164.4.3	08-Dec-2007	ad	Sync with head.
1.164.4.2	04-Dec-2007	yamt	apply the following change, which seems to get lost during vmlocking -> vmlocking2 transition. Module Name: src Committed By: yamt Date: Sun Oct 21 08:23:20 UTC 2007 Modified Files: src/sys/nfs: nfs_socket.c nfs_var.h Log Message: remove lwp argument from nfs_reconnect and always use &lwp0 because who triggers a reconnect doesn't really matter here. PR/37145. To generate a diff of this commit: cvs rdiff -r1.163 -r1.164 src/sys/nfs/nfs_socket.c cvs rdiff -r1.72 -r1.73 src/sys/nfs/nfs_var.h Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
1.164.4.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.164.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.164.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.165.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.167.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.167.6.4	05-Oct-2008	mjf	Sync with HEAD.
1.167.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.167.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.167.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.169.2.1	18-May-2008	yamt	sync with head.
1.170.8.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.170.8.1	19-Oct-2008	haad	Sync with HEAD.
1.170.4.2	10-Oct-2008	skrll	Sync with HEAD.
1.170.4.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.170.2.7	11-Mar-2010	yamt	sync with head
1.170.2.6	19-Jul-2009	yamt	debug printf and comments. no functional changes.
1.170.2.5	16-Jul-2009	yamt	remove sndlock. it's superseded by nm_solock. suggested by Andrew Doran.
1.170.2.4	04-May-2009	yamt	fix a merge botch.
1.170.2.3	04-May-2009	yamt	fix merge botches.
1.170.2.2	04-May-2009	yamt	sync with head.
1.170.2.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.173.4.9	24-Apr-2011	riz	Pull up following revision(s) (requested by tls in ticket #1600): sys/nfs/nfs_socket.c: revision 1.189 As suggested by matt@: change socket buffer reservations for NFS send/receive to 3 times max RPC size rather than 2 times. Avoids nasty TCP stalls observed at Panix. Will require increase to sbmax via sysctl for those running really huge NFS rsize/wsize (>64K).
1.173.4.8	29-Mar-2011	riz	Pull up following revision(s) (requested by tls in ticket #1583): sys/nfs/nfs_socket.c: revision 1.186 nfs_msg: #if 0 out tprintf for now and comment why.
1.173.4.7	29-Mar-2011	riz	Pull up following revision(s) (requested by tls in ticket #1582): sys/nfs/nfs_socket.c: revision 1.185 nfs_request: fix races which break congestion window and make nfs client stuck.
1.173.4.6	29-Mar-2011	riz	Pull up following revision(s) (requested by tls in ticket #1581): sys/nfs/nfs_socket.c: revision 1.181 If send fails with EMSGSIZE for whatever reason, it's unlikely to succeed no matter how hard we retry. So just fail the request.
1.173.4.5	29-Mar-2011	riz	Pull up following revision(s) (requested by tls in ticket #1580): sys/nfs/nfs_socket.c: revision 1.175 avoid noisy nfs_timer/nfs_reply DEBUG output that occurs when the NFS server goes away. use ratelimit(9) and only print the console error once every 10 seconds. PR#31562.
1.173.4.4	13-Nov-2009	sborrill	Pull up the following revisions(s) (requested by bouyer in ticket #1128): sys/nfs/nfs_socket.c: revision 1.182 Handle EWOULDBLOCK the same way as EPIPE. It seems the TCP socket layer can return EWOULDBLOCK on some occasion when the connection is broken.
1.173.4.3	06-Feb-2009	snj	branches: 1.173.4.3.4; Pull up following revision(s) (requested by ad in ticket #412): sys/nfs/nfs_socket.c: revision 1.178 PR kern/40491 5.0: nfs timer can crash/break on smp Hack around it by acquiring softnet_lock around the client-side timer loop.
1.173.4.2	02-Feb-2009	snj	Pull up following revision(s) (requested by yamt in ticket #393): sys/kern/uipc_socket.c: revision 1.185 sys/kern/uipc_socket2.c: revision 1.101 sys/kern/uipc_syscalls.c: revision 1.135 sys/miscfs/portal/portal_vnops.c: revision 1.81 sys/netsmb/smb_trantcp.c: revision 1.40 sys/nfs/nfs_socket.c: revision 1.177 sys/sys/socketvar.h: revision 1.118 restore the pre socket locking patch signal behaviour. this fixes a busy-loop in nfs_connect.
1.173.4.1	02-Feb-2009	snj	Pull up following revision(s) (requested by mrg in ticket #390): sys/nfs/nfs_socket.c: revision 1.176 Actually enforce the maximum timeout (60s by default) rather than backing off to 256*SRTT. This is why it sometimes could take hours for a NFS mount to come back when the server returned. contributed anonymously.
1.173.4.3.4.1	20-Apr-2010	matt	Pullin some NFS fixes from netbsd-5.
1.173.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.173.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.173.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.178.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.185.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.187.2.2	21-Apr-2011	rmind	sync with head
1.187.2.1	05-Mar-2011	rmind	sync with head
1.188.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.189.18.2	18-May-2014	rmind	sync with head
1.189.18.1	28-Aug-2013	rmind	Checkpoint work in progress: - Initial split of the protocol user-request method into the following methods: pr_attach, pr_detach and pr_generic for old the pr_usrreq. - Adjust socreate(9) and sonewconn(9) to call pr_attach without the socket lock held (as a preparation for the locking scheme adjustment). - Adjust all pr_attach routines to assert that PCB is not set. - Sprinkle various comments, document some routines and their locking. - Remove M_PCB, replace with kmem(9). - Fix few bugs spotted on the way.
1.189.14.2	03-Dec-2017	jdolecek	update from HEAD
1.189.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.189.4.1	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")
1.190.2.1	10-Aug-2014	tls	Rebase.
1.192.4.1	10-Jul-2016	martin	Pull up following revision(s) (requested by christos in ticket #1184): sys/nfs/nfs_socket.c: revision 1.198 sys/nfs/nfs_clntsocket.c: revision 1.5 Serialize all access to the NFS request queue via splsoftnet(). Fixes random crashes.
1.192.2.2	10-Jul-2016	martin	Pull up following revision(s) (requested by christos in ticket #1184): sys/nfs/nfs_socket.c: revision 1.198 sys/nfs/nfs_clntsocket.c: revision 1.5 Serialize all access to the NFS request queue via splsoftnet(). Fixes random crashes. XXX: Pullup-7
1.192.2.1	04-Nov-2015	riz	Pull up following revision(s) (requested by manu in ticket #882): sbin/umount/umount.c: revision 1.48 sys/nfs/nfsmount.h: revision 1.53 sys/nfs/nfs_var.h: revision 1.94 sys/nfs/nfs_iod.c: revision 1.7 sys/nfs/nfs_socket.c: revision 1.197 sys/nfs/nfs_bio.c: revision 1.191 sys/nfs/nfs_vfsops.c: revision 1.230 sys/nfs/nfs_clntsocket.c: revision 1.3 Remove useless and harmful sync(2) call in umount(8) Remove sync(2) call before unmount(2) in umount(8). This sync(2) is useless since unmount(2) will perform a VFS_SYNC anyway. But moreover, this sync(2) may be harmful, as there are some situation where it cannot return (unreachable NFS server, for instance), causing umount -f to be uneffective. Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.193.2.4	09-Jul-2016	skrll	Sync with HEAD
1.193.2.3	22-Sep-2015	skrll	Sync with HEAD
1.193.2.2	06-Jun-2015	skrll	Sync with HEAD
1.193.2.1	06-Apr-2015	skrll	Sync with HEAD
1.198.10.1	08-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #856): sys/nfs/nfs.h: revision 1.76 sys/nfs/nfs_subs.c: revision 1.230 sys/nfs/nfs_socket.c: revision 1.199 sys/nfs/nfs_clntsocket.c: revision 1.6 PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.199.4.1	10-Jun-2019	christos	Sync with HEAD
1.199.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.202.2.1	02-Aug-2025	perseant	Sync with HEAD
1.45	15-Mar-2009	cegger	ansify function definitions
1.44	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.43	19-Nov-2008	ad	branches: 1.43.4; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.42	05-May-2008	ad	branches: 1.42.6; 1.42.8; - Convert hashinit() to use kmem_alloc(). The hash tables can be large and it's better to not have them in kmem_map. - Convert a couple of minor items along the way to kmem_alloc(). - Fix some memory leaks.
1.41	04-Dec-2007	yamt	branches: 1.41.12; 1.41.14; 1.41.16; merge non-intrusive nfs changes from vmlocking.
1.40	01-Jun-2007	yamt	branches: 1.40.6; 1.40.8; 1.40.14; 1.40.16; use mutex and condvar.
1.39	12-Mar-2007	ad	branches: 1.39.2; Pass an ipl argument to pool_init/POOL_INIT to be used when initializing the pool's lock.
1.38	04-Mar-2007	christos	branches: 1.38.2; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.37	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.36	05-Feb-2007	yamt	branches: 1.36.2; have a mowner dedicated to nfs reply cache.
1.35	05-Feb-2007	yamt	nfsrv_updatecache: actually use reply caches for connection oriented protocols.
1.34	17-Jan-2007	yamt	plug mbuf leaks.
1.33	27-Dec-2006	yamt	remove nqnfs.
1.32	11-Dec-2005	christos	branches: 1.32.20; 1.32.24; 1.32.26; merge ktrace-lwp.
1.31	21-May-2004	yamt	branches: 1.31.12; enable reply cache for connection oriented protocols as well. linux retransmits rpcs even when using tcp.
1.30	20-Nov-2003	yamt	branches: 1.30.2; comments.
1.29	20-Nov-2003	yamt	fix a race case of nfsrv_getcache.
1.28	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.27	21-May-2003	yamt	branches: 1.27.2; simplelock for nfsd request cache.
1.26	21-May-2003	yamt	- KNF. - remove unneeded casts.
1.25	21-May-2003	yamt	poolify nfsd request cache.
1.24	21-May-2003	yamt	indent.
1.23	21-May-2003	yamt	remove local definitions of TRUE and FALSE.
1.22	02-Apr-2003	yamt	use queue manipulation macros.
1.21	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.20	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.19	10-Nov-2001	lukem	branches: 1.19.10; add RCSIDs
1.18	21-Feb-2001	jdolecek	branches: 1.18.2; 1.18.4; 1.18.8; make some more constant arrays 'const'
1.17	08-Nov-2000	ad	Update for hashinit() change.
1.16	30-Mar-2000	augustss	Remove register declarations.
1.15	09-Aug-1998	perry	branches: 1.15.12; bzero->memset, bcopy->memcpy, bcmp->memcmp
1.14	05-Jul-1998	jonathan	defopt ISO TPIP.
1.13	07-Feb-1998	chs	add flags arg to hashinit(), to pass to malloc().
1.12	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.11	09-Feb-1996	christos	nfs prototype changes
1.10	13-Dec-1994	mycroft	Sync with CSRG.
1.9	17-Aug-1994	mycroft	Use LIST and TAILQ for hash chain and LRU chain, respectively.
1.8	29-Jun-1994	cgd	branches: 1.8.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.7	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.6	18-Dec-1993	mycroft	Canonicalize all #includes.
1.5	07-Sep-1993	ws	branches: 1.5.2; Changes to VFS readdir semantics NFS changes for better cookie support ISOFS changes for better Rockridge support and support for generation numbers
1.4	22-May-1993	cgd	add include of select.h if necessary for protos, or delete if extraneous
1.3	21-May-1993	cgd	add rcsid again; fix RCS+crash fuckup
1.2	10-Apr-1993	glass	migrated code to make split possible
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.5.2.1	14-Nov-1993	mycroft	Canonicalize all #includes.
1.8.2.1	19-Aug-1994	mycroft	update from trunk
1.15.12.3	12-Mar-2001	bouyer	Sync with HEAD.
1.15.12.2	22-Nov-2000	bouyer	Sync with HEAD.
1.15.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.18.8.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.18.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.18.2.2	11-Dec-2002	thorpej	Sync with HEAD.
1.18.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.19.10.1	05-Jun-2004	jdc	Pull up revision 1.31 (via patch) (requested by yamt in ticket #1705). enable reply cache for connection oriented protocols as well. linux retransmits rpcs even when using tcp.
1.27.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.27.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.27.2.1	03-Aug-2004	skrll	Sync with HEAD
1.30.2.1	05-Jun-2004	jdc	Pull up revision 1.31 (requested by yamt in ticket #444). enable reply cache for connection oriented protocols as well. linux retransmits rpcs even when using tcp.
1.31.12.4	07-Dec-2007	yamt	sync with head
1.31.12.3	03-Sep-2007	yamt	sync with head.
1.31.12.2	26-Feb-2007	yamt	sync with head.
1.31.12.1	30-Dec-2006	yamt	sync with head.
1.32.26.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.32.24.2	05-Jun-2007	bouyer	Pull up following revision(s) (requested by yamt in ticket #707): sys/nfs/nfs_srvcache.c: revision 1.35 nfsrv_updatecache: actually use reply caches for connection oriented protocols.
1.32.24.1	05-Jun-2007	bouyer	Pull up following revision(s) (requested by yamt in ticket #704): sys/nfs/nfs_srvcache.c: revision 1.34 plug mbuf leaks.
1.32.20.3	09-Feb-2007	ad	Sync with HEAD.
1.32.20.2	01-Feb-2007	ad	Sync with head.
1.32.20.1	12-Jan-2007	ad	Sync with head.
1.36.2.3	24-Mar-2007	yamt	sync with head.
1.36.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.36.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.38.2.5	27-Aug-2007	yamt	- fix/add assertions. - fix numnfsrvcache.
1.38.2.4	26-Aug-2007	yamt	- mark nfssvc(2) MPSAFE and move the most of nfsd out of the kernel lock. - remove unused ns_solock. - remove some of KERNEL_LOCK/UNLOCK which are not necessary on this branch.
1.38.2.3	09-Jun-2007	ad	Sync with head.
1.38.2.2	05-Apr-2007	ad	Compile fixes.
1.38.2.1	13-Mar-2007	ad	Sync with head.
1.39.2.1	11-Jul-2007	mjf	Sync with head.
1.40.16.2	08-Dec-2007	ad	Sync with head.
1.40.16.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.40.14.1	08-Dec-2007	mjf	Sync with HEAD.
1.40.8.1	09-Jan-2008	matt	sync with HEAD
1.40.6.1	09-Dec-2007	jmcneill	Sync with HEAD.
1.41.16.2	04-May-2009	yamt	sync with head.
1.41.16.1	16-May-2008	yamt	sync with head.
1.41.14.1	18-May-2008	yamt	sync with head.
1.41.12.2	17-Jan-2009	mjf	Sync with HEAD.
1.41.12.1	02-Jun-2008	mjf	Sync with HEAD.
1.42.8.2	28-Apr-2009	skrll	Sync with HEAD.
1.42.8.1	19-Jan-2009	skrll	Sync with HEAD.
1.42.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.43.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.6	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.5	20-Dec-2022	hannken	branches: 1.5.6; When partitioning a mbuf chain with m_split() the last mbuf of the returned tail chain is not necessarily the same as the last mbuf of the initial chain. Always set "slp->ns_rawend" to the last mbuf of the tail chain to prevent mbuf leaks and corruption.
1.4	03-Sep-2009	tls	branches: 1.4.68; 1.4.94; Missed this file in previous commit, accidentally checked in fix to local repository copy! Sorry about that, folks.
1.3	14-Mar-2009	dsl	branches: 1.3.2; Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.2	14-Mar-2009	dsl	Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.1	19-Nov-2008	ad	branches: 1.1.4; 1.1.6; 1.1.8; 1.1.10; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.1.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1.8.3	28-Apr-2009	skrll	Sync with HEAD.
1.1.8.2	19-Jan-2009	skrll	Sync with HEAD.
1.1.8.1	19-Nov-2008	skrll	file nfs_srvsocket.c was added on branch nick-hppapmap on 2009-01-19 13:20:20 +0000
1.1.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.1.6.1	19-Nov-2008	mjf	file nfs_srvsocket.c was added on branch mjf-devfs2 on 2009-01-17 13:29:34 +0000
1.1.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.1.4.1	19-Nov-2008	haad	file nfs_srvsocket.c was added on branch haad-dm on 2008-12-13 01:15:28 +0000
1.3.2.4	16-Sep-2009	yamt	sync with head
1.3.2.3	04-May-2009	yamt	fix merge botches.
1.3.2.2	04-May-2009	yamt	sync with head.
1.3.2.1	14-Mar-2009	yamt	file nfs_srvsocket.c was added on branch yamt-nfs-mp on 2009-05-04 08:14:22 +0000
1.4.94.1	20-Dec-2022	martin	Pull up following revision(s) (requested by hannken in ticket #12): sys/nfs/nfs_srvsocket.c: revision 1.5 When partitioning a mbuf chain with m_split() the last mbuf of the returned tail chain is not necessarily the same as the last mbuf of the initial chain. Always set "slp->ns_rawend" to the last mbuf of the tail chain to prevent mbuf leaks and corruption.
1.4.68.1	20-Dec-2022	martin	Pull up following revision(s) (requested by hannken in ticket #1555): sys/nfs/nfs_srvsocket.c: revision 1.5 When partitioning a mbuf chain with m_split() the last mbuf of the returned tail chain is not necessarily the same as the last mbuf of the initial chain. Always set "slp->ns_rawend" to the last mbuf of the tail chain to prevent mbuf leaks and corruption.
1.5.6.1	02-Aug-2025	perseant	Sync with HEAD
1.17	23-Mar-2023	riastradh	nfs: Avoid integer overflow in nfs_namei bounds check. XXX pullup-8 XXX pullup-9 XXX pullup-10
1.16	27-Apr-2022	hannken	branches: 1.16.4; As VOP_GETATTR() needs a shared lock at least move the preopattr lookup inside nfs_namei() where we may lock the start directory without violating the lock order.
1.15	17-Jan-2020	ad	VFS_VGET(), VFS_ROOT(), VFS_FHTOVP(): give them a "int lktype" argument, to allow us to get shared locks (or no lock) on the returned vnode. Matches FreeBSD.
1.14	05-Nov-2012	dholland	branches: 1.14.30; 1.14.38; 1.14.42; 1.14.44; Rename the new ni_startdir (the slot used to hold the starting point for openat() and friends) to ni_atdir to avoid confusion with a previously existing (and, alas, still documented) ni_startdir field that meant something else entirely.
1.13	13-Oct-2012	dholland	Replace hack implementation of NDAT() for "nameiat" with a proper one. (This change requires a kernel bump.)
1.12	27-Sep-2011	christos	branches: 1.12.2; 1.12.12; use NFS_MAXPATHLEN instead of MAXPATHLEN
1.11	08-Aug-2011	dholland	nfs_namei() should not return a non-null path buffer except on success, even though the callers are apparently prepared to cope. Fixes last tidyup part of PR 44625.
1.10	11-Apr-2011	dholland	Clean up. Move some more code across from nfsd's private entry points.
1.9	19-Mar-2011	dholland	Fix memory leak introduced with the struct pathbuf changes. Hi, me. Closes PR 44625.
1.8	30-Nov-2010	dholland	branches: 1.8.2; Abolish struct componentname's cn_pnbuf. Use the path buffer in the pathbuf object passed to namei as work space instead. (For now a pnbuf pointer appears in struct nameidata, to support certain unclean things that haven't been fixed yet, but it will be going away in the future.) This removes the need for the SAVENAME and HASBUF namei flags.
1.7	19-Nov-2010	dholland	Introduce struct pathbuf. This is an abstraction to hold a pathname and the metadata required to interpret it. Callers of namei must now create a pathbuf and pass it to NDINIT (instead of a string and a uio_seg), then destroy the pathbuf after the namei session is complete. Update all namei call sites accordingly. Add a pathbuf(9) man page and update namei(9). The pathbuf interface also now appears in a couple of related additional places that were passing string/uio_seg pairs that were later fed into NDINIT. Update other call sites accordingly.
1.6	24-Jun-2010	hannken	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.
1.5	27-Sep-2009	dholland	branches: 1.5.2; 1.5.4; Move a big wodge of symlink-following code from nfsd to inside lookup_for_nfsd(). This code is, or at least should be, the same as the regular symlink-following code plus an extra flag nfsd needs. The two lots of code can/will be merged in the future.
1.4	27-Sep-2009	dholland	Rename lookup() to lookup_for_nfsd(), to make it clear just whose private backdoor entry point this is. Also, clone the lookup_for_nfsd() entry point as lookup_for_nfsd_index(), for use by a different call site in nfsd that does different unclean things with nameidata.
1.3	04-May-2009	yamt	branches: 1.3.2; when freeing cn_pnbuf, make it NULL if DIAGNOSTIC.
1.2	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.1	19-Nov-2008	ad	branches: 1.1.4; 1.1.6; 1.1.8; 1.1.10; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.1.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1.8.3	28-Apr-2009	skrll	Sync with HEAD.
1.1.8.2	19-Jan-2009	skrll	Sync with HEAD.
1.1.8.1	19-Nov-2008	skrll	file nfs_srvsubs.c was added on branch nick-hppapmap on 2009-01-19 13:20:20 +0000
1.1.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.1.6.1	19-Nov-2008	mjf	file nfs_srvsubs.c was added on branch mjf-devfs2 on 2009-01-17 13:29:34 +0000
1.1.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.1.4.1	19-Nov-2008	haad	file nfs_srvsubs.c was added on branch haad-dm on 2008-12-13 01:15:28 +0000
1.3.2.4	11-Aug-2010	yamt	sync with head.
1.3.2.3	11-Mar-2010	yamt	sync with head
1.3.2.2	04-May-2009	yamt	sync with head.
1.3.2.1	04-May-2009	yamt	file nfs_srvsubs.c was added on branch yamt-nfs-mp on 2009-05-04 08:14:22 +0000
1.5.4.3	21-Apr-2011	rmind	sync with head
1.5.4.2	05-Mar-2011	rmind	sync with head
1.5.4.1	03-Jul-2010	rmind	sync with head
1.5.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.8.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.12.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.12.2.2	16-Jan-2013	yamt	sync with (a bit old) head
1.12.2.1	30-Oct-2012	yamt	sync with head
1.14.44.1	17-Jan-2020	ad	Sync with head.
1.14.42.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #1617): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.14.38.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.14.30.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #1810): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.16.4.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #134): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.242	09-Feb-2022	andvar	s/ony/only/
1.241	05-Sep-2020	riastradh	Round of uvm.h cleanup. The poorly named uvm.h is generally supposed to be for uvm-internal users only. - Narrow it to files that actually need it -- mostly files that need to query whether curlwp is the pagedaemon, which should maybe be exposed by an external header. - Use uvm_extern.h where feasible and uvm_*.h for things not exposed by it. We should split up uvm_extern.h but this will serve for now to reduce the uvm.h dependencies. - Use uvm_stat.h and #ifdef UVMHIST uvm.h for files that use UVMHIST(ubchist), since ubchist is declared in uvm.h but the reference evaporates if UVMHIST is not defined, so we reduce header file dependencies. - Make uvm_device.h and uvm_swap.h independently includable while here. ok chs@
1.240	25-May-2020	ad	- Alter the convention for uvm_page_array slightly, so the basic search parameters can't change part way through a search: move the "uobj" and "flags" arguments over to uvm_page_array_init() and store those with the array. - With that, detect when it's not possible to find any more pages in the tree with the given search parameters, and avoid repeated tree lookups if the caller loops over uvm_page_array_fill_and_peek().
1.239	04-Apr-2020	mlelstv	NFSv2 is limited to use only 32bit in metadata. Prevent that larger metadata values are simply truncated. -> clamp filesystem block counts to signed 32bit. -> clamp file sizes to signed 32bit () Some NFSv2 clients also have problems to handle buffer sizes larger than (signed) 16bit. -> clamp buffer sizes to signed 16bit for better compatibility. () This can lead to erroneous behaviour for files larger than 2GB that NFSv2 cannot handle but it is still better than before. An alternative would be to (partially) reject operations on files larger than 2GB, but which causes other problems.
1.238	08-Mar-2020	mgorny	Update NFS errno mapping and add assert for correctness Add the mapping for errno values missing in nfsrv_v2errmap[]. While at it, add a compile-time assert to make sure that the array does not become out-of-date again.
1.237	24-Feb-2020	ad	v_interlock -> vmobjlock
1.236	15-Dec-2019	ad	branches: 1.236.2; Merge from yamt-pagecache: - do gang lookup of pages using radixtree. - remove now unused uvm_object::uo_memq and vm_page::listq.queue.
1.235	22-Dec-2018	maxv	Replace M_ALIGN and MH_ALIGN by m_align.
1.234	22-Dec-2018	maxv	Replace: M_MOVE_PKTHDR -> m_move_pkthdr. No functional change, since the former is a macro to the latter.
1.233	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!)
1.232	08-May-2018	maxv	branches: 1.232.2; Use M_MOVE_PKTHDR.
1.231	26-Apr-2018	maxv	Hum. This should be M_READONLY, not M_ROMAP. M_ROMAP tells us whether the mbuf storage is mapped on a read-only page. But an mbuf can still be read-only in the sense that the storage is shared with other mbufs.
1.230	21-Jan-2018	christos	branches: 1.230.2; PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.229	01-Apr-2017	riastradh	branches: 1.229.6; KASSERT(mutex_owned(vp->v_interlock)) in vnode iterator selector.
1.228	10-Jun-2016	ozaki-r	branches: 1.228.2; 1.228.4; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.227	10-Aug-2014	tls	branches: 1.227.4; Merge tls-earlyentropy branch into HEAD.
1.226	24-May-2014	christos	Introduce a selector function to the vfs vnode iterator so that we don't need to vget() vnodes that we are not interested at, and optimize locking a bit. Iterator changes reviewed by Hannken (thanks), the rest of the bugs are mine.
1.225	17-Mar-2014	hannken	branches: 1.225.2; Change nfs_clearcommit() to use vfs_vnode_iterator.
1.224	18-Sep-2013	pgoyette	knf (blank line even if there are no local declarations)
1.223	18-Sep-2013	christos	Use reference counting to keep track of construction and destruction of the structures used by both the nfs server and client code. Tested by pgoyette@ 1. mount remote fs via nfs (my /home directory), which autoloads nfs module 2. manually modload nfsserver 3. wait a bit 4. manually modunload nfsserver 5. wait a couple minutes 6. verify that client access still works (/bin/ls ~paul home dir) 7. manually modload nfsserver again 8. start an nfsd process 9. wait a bit 10. kill nfsd process 11. wait 12. manually modunload nfsserver again 13. verify continued client access XXX: Note that nfs_vfs_init() calls nfs_init(), but nfs_vfs_done() does not call nfs_fini(). Also note that the destruction order is wrong in it, but probably does not matter. "someone" (!= me) should fix it :-) and run the above tests.
1.222	19-Nov-2011	tls	branches: 1.222.8; 1.222.12; 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.
1.221	12-Jun-2011	rmind	branches: 1.221.2; 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.
1.220	06-Nov-2010	uebayasi	branches: 1.220.6; Include uvm/uvm.h to use UVM internal type (struct vm_page).
1.219	02-Mar-2010	pooka	branches: 1.219.2; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.218	31-Dec-2009	christos	branches: 1.218.2; put nuidhash_max in a file that is shared between server and client code.
1.217	14-May-2009	yamt	nfs_clearcommit: fix a race with vnode cleaning.
1.216	15-Mar-2009	cegger	ansify function definitions
1.215	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.214	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.213	14-Mar-2009	dsl	Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.212	17-Dec-2008	cegger	branches: 1.212.2; kill MALLOC and FREE macros.
1.211	28-Nov-2008	pooka	g/c unused malloc types
1.210	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.209	22-Oct-2008	matt	branches: 1.209.2; Don't need nfs_vfs_reinit anymore since we don't resize tables anymore. Move reinit code to init case.
1.208	22-Oct-2008	matt	Change NFS to use a RB-tree for its FH->nfsnode lookups.
1.207	09-Oct-2008	pooka	Use atomic op to get next xid. Initialize value with arc4random() at nfs init time instead system time based trickery intermingled with the runtime code. le bouef: kills last simple_lock from nfs
1.206	30-Sep-2008	pooka	Initialize nfsnode pools and malloc type dynamically in the constructor instead of depending on link sets. Consequently, rename nfs_nh{init,reinit,done} to nfs_node_{init,reinit,done}, respectively, to better convey the function.
1.205	15-Jul-2008	christos	explicitly set birthtime to VNOVAL, since there is no such thing in nfsv{2,3}
1.204	04-Jun-2008	ad	branches: 1.204.2; 1.204.4; vm_page: put TAILQ_ENTRY into a union with LIST_ENTRY, so we can use both.
1.203	10-May-2008	rumble	Convert file systems to dynamically attach with the new module interface. Make VFS hooks dynamic while we're here and say farewell to VFS_ATTACH and VFS_HOOKS_ATTACH linksets. As a consequence, most of the file systems can now be loaded as new style modules. Quick sanity check by ad@.
1.202	05-May-2008	ad	branches: 1.202.2; - Convert hashinit() to use kmem_alloc(). The hash tables can be large and it's better to not have them in kmem_map. - Convert a couple of minor items along the way to kmem_alloc(). - Fix some memory leaks.
1.201	24-Mar-2008	yamt	branches: 1.201.2; 1.201.4; merge yamt-lazymbuf branch.
1.200	05-Mar-2008	elad	Nuke a KAUTH_GENERIC_ISSUSER, this time in favor of an euid == 0, as the traditional NFS maproot functionality goes. Put in a note explaining why and who, also mark for future greps. Okay yamt@.
1.199	13-Feb-2008	yamt	branches: 1.199.2; 1.199.6; reject files larger than nm_maxfilesize.
1.198	28-Jan-2008	yamt	nfs_check_wccdata: unifdef wcc kludge messages.
1.197	24-Jan-2008	ad	specfs changes for PR kern/37717 (raidclose() is no longer called on shutdown). There are still problems with device access and a PR will be filed. - Kill checkalias(). Allow multiple vnodes to reference a single device. - Don't play dangerous tricks with block vnodes to ensure that only one vnode can describe a block device. Instead, prohibit concurrent opens of block devices. As a bonus remove the unreliable code that prevents multiple file system mounts on the same device. It's no longer needed. - Track opens by vnode and by device. Issue cdev_close() when the last open goes away, instead of abusing vnode::v_usecount to tell if the device is open.
1.196	02-Jan-2008	yamt	use kmem_alloc instead of malloc.
1.195	02-Jan-2008	ad	Merge vmlocking2 to head.
1.194	08-Dec-2007	pooka	branches: 1.194.4; Remove cn_lwp from struct componentname. curlwp should be used from on. The NDINIT() macro no longer takes the lwp parameter and associates the credentials of the calling thread with the namei structure.
1.193	26-Nov-2007	pooka	branches: 1.193.2; 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
1.192	28-Oct-2007	yamt	branches: 1.192.2; make NFS_ATTRTIMEO a function.
1.191	27-Jul-2007	yamt	branches: 1.191.4; 1.191.6; 1.191.10; 1.191.12; stop nfs tick when we have nothing to do.
1.190	09-Jul-2007	ad	branches: 1.190.2; Fix build when not !NFSSERVER.
1.189	09-Jul-2007	ad	Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.188	06-Jun-2007	yamt	nfs_getattrcache: simplify. no functional changes.
1.187	28-May-2007	yamt	- remove nfs_exit exit hook. ok'ed by christos@. - as far as i understand the code, it shouldn't be necessary because nfs_request can't return without removing its request and r->r_lwp is either curlwp or NULL. - even if it's necessary, leaking requests is not the correct way to recover from the condition. - nfs_request: add a related assertion.
1.186	29-Apr-2007	yamt	use mutex and condver.
1.185	22-Apr-2007	dsl	Change the way that emulations locate files within the emulation root to avoid having to allocate space in the 'stackgap' - which is very LWP unfriendly. The additional code for non-emulation namei() is trivial, the reduction for the emulations is massive. The vnode for a processes emulation root is saved in the cwdi structure during process exec. If the emulation root the TRYEMULROOT flag are set, namei() will do an initial search for absolute pathnames in the emulation root, if that fails it will retry from the normal root. ".." at the emulation root will always go to the real root, even in the middle of paths and when expanding symlinks. Absolute symlinks found using absolute paths in the emulation root will be relative to the emulation root (so /usr/lib/xxx.so -> /lib/xxx.so links inside the emulation root don't need changing). If the root of the emulation would be returned (for an emulation lookup), then the real root is returned instead (matching the behaviour of emul_lookup, but being a cheap comparison here) so that programs that scan "../.." looking for the root dircetory don't loop forever. The target for symbolic links is no longer mangled (it used to get the CHECK_ALT_xxx() treatment, so could get /emul/xxx prepended). CHECK_ALT_xxx() are no more. Most of the change is deleting them, and adding TRYEMULROOT to the flags to NDINIT(). A lot of the emulation system call stubs could now be deleted.
1.184	09-Mar-2007	yamt	branches: 1.184.2; 1.184.4; nfs_check_wccdata: print timestamps.
1.183	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.182	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.181	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.180	15-Feb-2007	yamt	branches: 1.180.2; use mutex and rwlock rather than lockmgr.
1.179	27-Dec-2006	yamt	remove nqnfs.
1.178	09-Dec-2006	chs	a smorgasbord of improvements to vnode locking and path lookup: - LOCKPARENT is no longer relevant for lookup(), relookup() or VOP_LOOKUP(). these now always return the parent vnode locked. namei() works as before. lookup() and various other paths no longer acquire vnode locks in the wrong order via vrele(). fixes PR 32535. as a nice side effect, path lookup is also up to 25% faster. - the above allows us to get rid of PDIRUNLOCK. - also get rid of WANTPARENT (just use LOCKPARENT and unlock it). - remove an assumption in layer_node_find() that all file systems implement a recursive VOP_LOCK() (unionfs doesn't). - require that all file systems supply vfs_vptofh and vfs_fhtovp routines. fill in eopnotsupp() for file systems that don't support being exported and remove the checks for NULL. (layerfs calls these without checking.) - in union_lookup1(), don't change refcounts in the ISDOTDOT case, just adjust which vnode is locked. fixes PR 33374. - apply fixes for ufs_rename() from ufs_vnops.c rev. 1.61 to ext2fs_rename().
1.177	09-Nov-2006	yamt	branches: 1.177.2; remove some __unused in function parameters.
1.176	20-Oct-2006	reinoud	Replace the LIST structure mp->mnt_vnodelist to a TAILQ structure since all vnodes were synced and processed backwards. This meant that the last accessed node was processed first and the earlierst last. An extra benefit is the removal of the ugly hack from the Berkly days on LFS. In the proces, i've also replaced the various variations hand written loops by the TAILQ_FOREACH() macro's.
1.175	17-Oct-2006	dogcow	now that we have -Wno-unused-parameter, back out all the tremendously ugly code to gratuitously access said parameters.
1.174	14-Oct-2006	yamt	grab glock when calling uvm_unp_setsize, so that it doesn't interfere mmap'ed accesses. this fixes an assertion failure in in nfs_doio_read. ("vp->v_size >= uiop->uio_offset + uiop->uio_resid")
1.173	13-Oct-2006	christos	more __unused
1.172	13-Oct-2006	dogcow	more unused variable fallout.
1.171	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.170	04-Sep-2006	yamt	branches: 1.170.2; 1.170.4; nfs_ispublicfh: fix another instance of cast-qual.
1.169	04-Sep-2006	yamt	remove (void *) cast from NFSRVFH_DATA as it sometimes discards const qualifier. pointed out by Havard Eidnes. (it wasn't detected by in-tree gcc4. seems like a compiler bug.)
1.168	02-Sep-2006	yamt	nfsd: deal with variable-sized filehandles.
1.167	02-Sep-2006	christos	fix default type decls fix incomplete initializer
1.166	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.165	07-Jun-2006	kardel	branches: 1.165.2; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.164	19-May-2006	yamt	branches: 1.164.2; - fix compilation problem for !NFSSERVER && NFS. pointed by Tom Spindler on source-changes@. - make nfs_srvdesc_pool static.
1.163	18-May-2006	yamt	- fix some leaks in nfsd, introduced by kauth changes. - simplify code. - add some assertions. - wrap some long lines. - remove an unnecessary ";".
1.162	14-May-2006	elad	integrate kauth.
1.161	14-May-2006	christos	XXX: GCC uninitialized
1.160	15-Apr-2006	christos	Coverity CID 1141: Add a KASSERT before deref.
1.159	15-Apr-2006	christos	Coverity CID 1142: Add a KASSERT before deref.
1.158	01-Mar-2006	yamt	branches: 1.158.2; 1.158.4; 1.158.6; 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.
1.157	16-Jan-2006	yamt	branches: 1.157.2; 1.157.4; - tweak RUN_ONCE api to allow init_func returns an error. - physio: handle failure of workqueue_create.
1.156	11-Dec-2005	christos	branches: 1.156.2; merge ktrace-lwp.
1.155	25-Nov-2005	thorpej	Use a once control to initialize the NFS server / client shared data from nfs_vfs_init() or sys_nfssvc(). Remove the nfs_init() call from main().
1.154	22-Nov-2005	yamt	- reduce number of linear search per rpc. - coalesce mount_netexport_pair into netexport.
1.153	23-Sep-2005	jmmv	branches: 1.153.6; Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.152	19-Sep-2005	christos	ATTRTIMEO takes 2 args.
1.151	19-Aug-2005	yamt	as we now have 64bit ino_t, no need to truncate nfsv3 fileids.
1.150	07-Jul-2005	christos	1. use p = uio->uio_procp consistently and eliminate suspicious uses of curproc (where uio->uio_procp should be used?). Don't do this for nfs_commit(), because yamt says it is possibly wrong. 2. nfs_doio() does not use struct proc; remove it and the code to compute it. 3. use copyin_proc() and copyout_proc() instead of copyin() and copyout(). 4. check return of copyout_proc(). and mark return from copyin_proc() XXX 5. Eliminate check p == curproc assertion check from nfs_write; nfs_read does not have it and we might be called in a different process context anyway (PR 20138).
1.149	29-May-2005	christos	branches: 1.149.2; - sprinkle const - avoid shadowed variables - mark bad const use with XXXUNCONST
1.148	26-Feb-2005	perry	branches: 1.148.2; nuke trailing whitespace
1.147	28-Jan-2005	yamt	nfs_namei: return EACCES for empty filenames as rfc1813 says.
1.146	28-Jan-2005	yamt	nfs_clearcommit: don't attempt to clear commit info (n_pushlo, etc) unless the vnode is of VREG. union members used to keep commit info are used for other purposes in the case of !VREG.
1.145	27-Jan-2005	yamt	keep directory eof cache when inactivating vnode because there's no reason to throw it away. (fix an unintended side effect of nfs_subs.c rev.1.144.)
1.144	26-Jan-2005	yamt	handle a really empty directory, which doesn't have even the dot entry.
1.143	25-Jan-2005	yamt	branches: 1.143.2; nfs_check_wccdata: comment.
1.142	21-Jan-2005	yamt	s/time/mono_time/ for n_attrstamp and n_accstamp. (parts of) PR/25641.
1.141	19-Jan-2005	yamt	implement inaccurate mtime/ctime detection. namely, if mtime or ctime are same between pre_op_attr and post_op_attr when we expected them to be changed, don't trust the server.
1.140	09-Jan-2005	yamt	branches: 1.140.2; invalidate cache if filesize is changed besides our activity because it means that we're out of sync with the server.
1.139	06-Jan-2005	yamt	nfs_loadattrcache: invalidate access cache when ctime is changed.
1.138	26-Oct-2004	yamt	since daddr_t is 64-bit these days, simply use nfs directory cookies as buffer cache indexes. regress/sys/fs/getdents is now supposed to work. fix PR/27112.
1.137	03-Oct-2004	yamt	nfs_enterdircache: initialize dc_flags of a newly allocated dircache entry. provided by Greg Oster.
1.136	17-Sep-2004	skrll	There's no need to pass a proc value when using UIO_SYSSPACE with vn_rdwr(9) and uiomove(9). OK'd by Jason Thorpe
1.135	15-Sep-2004	yamt	fix access-after-free bugs in dircache code by refcounting nfsdircache. PR/26864.
1.134	14-Jun-2004	yamt	nfs_searchdircache: fix a null dereference in the case that offset!=0 and dircache hasn't been initialized yet.
1.133	21-Apr-2004	christos	Replace the statfs() family of system calls with statvfs(). Retain binary compatibility.
1.132	19-Mar-2004	yamt	branches: 1.132.2; nfs_getattrcache: deal with timer wraparound.
1.131	12-Mar-2004	yamt	shrink sizeof struct nfsnode by putting exclusive members into union.
1.130	29-Nov-2003	yamt	nfs_zeropad: remove an unneeded substitution (and clean up a little.)
1.129	02-Oct-2003	itojun	plug mbuf leak due to manual mbuf handling. PR kern/13807. (martti confirmed that it stabilizes the situation described in kren/13807)
1.128	26-Sep-2003	yamt	change n_mtime from time_t to timespec in order to improve cache consistency. (1 second granularity is too loose these days.)
1.127	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.126	23-Jul-2003	yamt	when rexmitting a request due to NFSERR_JUKEBOX, use a new xid as RFC1813 says.
1.125	29-Jun-2003	fvdl	branches: 1.125.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.124	29-Jun-2003	thorpej	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.123	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
1.122	09-Jun-2003	yamt	rework zero padding of rpc reply. - for READ procedure, don't send back more bytes than requested. - don't have doubtful assumptions on mbuf chain structure. - rename a function (nfsm_adj -> nfs_zeropad) to avoid confusion as the semantics of the function was changed.
1.121	22-May-2003	yamt	poolify nfsrv_descript.
1.120	22-May-2003	yamt	avoid double free with xlatecookie.
1.119	07-May-2003	yamt	simple lock for nfs iod.
1.118	03-May-2003	yamt	tweak nfsm_adj to pay attention to read only mbufs.
1.117	03-May-2003	yamt	more comment.
1.116	03-May-2003	yamt	better handling of write verifier change.
1.115	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.114	16-Apr-2003	yamt	sync a comment with reality.
1.113	02-Apr-2003	yamt	use queue manipulation macros.
1.112	01-Apr-2003	yamt	add an assertion.
1.111	31-Mar-2003	yamt	rename fvdl_debug to NFS_DEBUG_COMMIT. ok'ed by fvdl.
1.110	28-Mar-2003	yamt	reply ENAMETOOLONG properly instead of discarding request as BADRPC. my own PR20791.
1.109	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.108	10-Feb-2003	christos	move the MALLOC decl for DIROFFS to nfs_subs.c
1.107	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.106	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
1.105	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.104	23-Aug-2002	enami	s/FREE/PNBUF_PUT/
1.103	17-Mar-2002	christos	branches: 1.103.4; 1.103.6; use the exithook mechanism to remove the exiting process from the list of processes to be signalled in a soft mount.
1.102	11-Mar-2002	jdolecek	Cosmetic change for nfs_enterdircache() - since 'blkno' is last arg, define it's type last, too. Noted in kern/14742 by John Franklin.
1.101	28-Feb-2002	fvdl	Invalidate the access cache when loading a new set of attributes into the atribute cache. Fixes access cache problem seen by Nathan Funk of the UofS, relayed by Greg Oster.
1.100	26-Jan-2002	chs	re-enable NFSv3 commit RPCs by abandoning my new approach in favor of frank's scheme, with one new twist: don't wait until we've totally run out of free pages before committing, but instead notice when we've built up a largish range of uncommitted pages and commit only the older half of the range, which is likely to already be on disk on the server.
1.99	10-Nov-2001	lukem	add RCSIDs
1.98	27-Sep-2001	fvdl	branches: 1.98.2; Always initialize ni_rootdir in nfs_namei. From Andrei Petrov.
1.97	15-Sep-2001	chs	a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
1.96	15-Sep-2001	chs	add a new VFS op, vfs_reinit, which is called when desiredvnodes is adjusted via sysctl. file systems that have hash tables which are sized based on the value of this variable now resize those hash tables using the new value. the max number of FFS softdeps is also recalculated. convert various file systems to use the <sys/queue.h> macros for their hash tables.
1.95	07-Jun-2001	lukem	branches: 1.95.2; 1.95.4; delint lvalue cast abuse
1.94	21-Apr-2001	bjh21	In nfs_loadattrcache(), if checkalias() gives us a new vnode, lock it. This prevents us losing the locked state of the old vnode. fvdl thinks the old vnode is certain to be locked at this point. I've put in a KASSERT to be on the safe side. This seems to fix PR kern/12661.
1.93	23-Mar-2001	fvdl	Same change as in the UFS code: unlock vnode before setting v_op to spec_vnode_ops. From Bill Studenmund.
1.92	21-Feb-2001	jdolecek	branches: 1.92.2; make some more constant arrays 'const'
1.91	14-Feb-2001	fvdl	Fix some possible locking errors in nfs_namei (XXX this function should die)
1.90	18-Jan-2001	jdolecek	constify
1.89	27-Nov-2000	chs	Initial integration of the Unified Buffer Cache project.
1.88	08-Nov-2000	ad	Update for hashinit() change.
1.87	24-Oct-2000	fvdl	Do not accept vnode type changes to an active node. This may wreak havoc if the server erroneously uses the same filehandle for different files. This changes back revision 1.28; the PR that that revision fixed doesn't apply anymore, it has been verified not to be a problem with this change.
1.86	27-Sep-2000	fvdl	Avoid unused variables for V2_ONLY case.
1.85	24-Sep-2000	enami	Don't bother to clear commit information for the vnode of type VNON. It is not necessary since it is a vnode being initialized and it shouldn't be done since filesystem private data may not be assigned yet.
1.84	19-Sep-2000	bjh21	Extend NFS_V2_ONLY to remove NQNFS lease support as well. Saves another 10k.
1.83	19-Sep-2000	fvdl	Add functions to deal with keeping track of commit ranges.
1.82	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.81	03-Aug-2000	thorpej	Convert namei pathname buffer allocation to use the pool allocator.
1.80	03-Aug-2000	thorpej	MALLOC()/FREE() are not to be used for variable size allocations.
1.79	28-Jun-2000	mrg	remove include of <vm/vm.h>
1.78	27-Jun-2000	mrg	remove include of <vm/vm.h>
1.77	20-Jun-2000	mrg	branches: 1.77.2; disable the bloated NFS structure check on 64bit sparc64.
1.76	09-Jun-2000	fvdl	Some tweaks to enable NFS over IPv6. The special-casing of AF_INET should really be removed.
1.75	30-Mar-2000	augustss	branches: 1.75.2; Remove register declarations.
1.74	30-Mar-2000	simonb	Delete redundant decl of nfs_pub - it's in <sys/mount.h>. Delete redundant decls of nfsv{2,3}_type - they're in <nfs/nfsproto.h>.
1.73	16-Mar-2000	jdolecek	Add new VFS op routine - vfs_done and call it on filesystem detach in vfs_detach(). vfs_done may free global filesystem's resources, typically those allocated in respective filesystem's init function. Needed so those filesystems which went in via LKM have a chance to clean after themselves before unloading. This fixes random panics when LKM for filesystem using pools was loaded and unloaded several times. For each leaf filesystem, add appropriate vfs_done routine.
1.72	01-Nov-1999	fvdl	Stuff values in va_blocksize that are closer to reality.
1.71	06-Sep-1999	is	branches: 1.71.2; 1.71.4; 1.71.6; Don't truncate minor numbers >= 256. Problem reported by Saitoh Masanobu, fix by Frank van der Linden.
1.70	08-Jul-1999	wrstuden	Modify file systems to deal with struct lock in struct vnode. All leaf fs's other than nfs use genfs_lock() for locking. Modify lookup routines to set PDIRUNLOCK when they unlock the parrent.
1.69	24-Mar-1999	mrg	branches: 1.69.2; 1.69.4; completely remove Mach VM support. all that is left is the all the header files as UVM still uses (most of) these.
1.68	16-Mar-1999	fvdl	..JUKEBOX can happen on writes too.
1.67	16-Mar-1999	fvdl	The JUKEBOX error may be returned by the read operation, so don't filter it out.
1.66	06-Mar-1999	fair	Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.65	27-Feb-1999	wrstuden	Rationalize the vfs_checkexp macro to be VFS_CHECKEXP.
1.64	26-Feb-1999	wrstuden	Modify vfsops to seperate vfs_fhtovp() into two routines. vfs_fhtovp() now only handles the file handle to vnode conversion, and a new call, vfs_checkexp(), performs the export verification.
1.63	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.62	05-Jul-1998	jonathan	defopt ISO TPIP.
1.61	25-Jun-1998	thorpej	defopt NFSSERVER
1.60	24-Jun-1998	sommerfe	Always include fifos; "not an option any more".
1.59	22-Jun-1998	sommerfe	defopt for options FIFO
1.58	08-May-1998	kleink	Fix some arithmetics lossage on typeless pointers.
1.57	03-Mar-1998	fvdl	Only free cookies on error when they were actually allocated by the readdir vop.
1.56	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.55	19-Feb-1998	thorpej	Include the NFS option header.
1.54	10-Feb-1998	mrg	- add defopt's for UVM, UVMHIST and PMAP_NEW. - remove unnecessary UVMHIST_DECL's.
1.53	07-Feb-1998	chs	add flags arg to hashinit(), to pass to malloc().
1.52	06-Feb-1998	mikel	ELAST incremented, update nfsrv_v2errmap[] initialization
1.51	05-Feb-1998	mrg	initial import of the new virtual memory system, UVM, into -current. UVM was written by chuck cranor <chuck@maria.wustl.edu>, with some minor portions derived from the old Mach code. i provided some help getting swap and paging working, and other bug fixes/ideas. chuck silvers <chuq@chuq.com> also provided some other fixes. this is the rest of the MI portion changes. this will be KNF'd shortly. :-)
1.50	22-Jan-1998	fvdl	Refuse to create entries in the dir cache for offset 0. This is a special case anyway, and amd(8) erroneously returns some entries with cookie 0. Fixes PR 4844
1.49	19-Oct-1997	fvdl	branches: 1.49.2; * Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.48	11-Oct-1997	fvdl	Move cookie heuristic function inside ifdef NFS, to make a kernel with server code but without client code link again. From Erik Bertelsen, PR 4259
1.47	10-Oct-1997	fvdl	Fix unitialized var warning (did not appear on i386, but did on sparc).
1.46	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.45	14-Jul-1997	fvdl	branches: 1.45.2; Don't assume that pointers into mbuf data remain valid across nfsm_dissect. In readdirplus, don't keep such pointers but store the file attributes in a variable instead until they are needed. Change nfsm_loadattr* a bit so it can accept a direct pointer to an nfs_fattr structure.
1.44	04-Jul-1997	drochner	Don't cast 64bit (off_t) file sizes to vm_offset_t (32bit on many architectures), truncate them intelligently instead. The truncation is done centralized in vnode_pager.c. This prevents from wrap-over effects when parts of large (>2^32 byte) files are mmapped. Don't allow to mmap above the numerical range of vm_offset_t. This is considered a temporary solution until the vm system handles the object sizes/offsets more cleanly.
1.43	24-Jun-1997	fvdl	Extend lookup handling for WebNFS. This means that nfs_namei deals with full pathname lookups if a public filehandle is used, and that it translates the '%' escapes (URL-style) in the same case. Also, make nfsrv_fhtovp convert the public filehandle to the vp of the publicly exported filesystem, as stored in the nfs_pub structure.
1.42	08-May-1997	mycroft	Pass the vnode type to vaccess(), and use it when checking VEXEC. Make sure that the mode bits passed to vaccess() and returned by foo_getattr() contain only permission bits.
1.41	27-Mar-1997	thorpej	Update for new mbuf code.
1.40	23-Mar-1997	fvdl	Check for the use of reserved ports on a per-request basis, unless MNT_EXNORESPORT is specified. The check is cheap and doesn't impose any extra overhead.
1.39	22-Feb-1997	fvdl	Fixes from BSDI (thanks go to Keith Bostic). Original RCS messages: date: 1997/02/10 18:41:14; author: cp; state: Exp; lines: +110 -46 Make nfs_realign go away on sparc and add functionality to nfsm_disct.
1.38	31-Jan-1997	thorpej	branches: 1.38.2; NFSCLIENT -> NFS.
1.37	09-Dec-1996	fvdl	branches: 1.37.2; Comment change in previous made for some bad english..
1.36	09-Dec-1996	fvdl	Move '#ifdef NFSSERVER' back to the right spot; NQNFS datastructures need to be initialized on both the client and the server side. Remove misleading comment about this being just server stuff.
1.35	03-Dec-1996	thorpej	Make NFSSERVER work without NFSCLIENT. This is achieved by splitting the client and server/shared data initialization into separate functions, and calling the server/shared initialization directly from main(). Problem noted in PR #1308 (Kenneth Stailey) and PR #1780 (Chris Demetriou). Fix suggested in PR #1780 by Chris Demetriou, and munged a bit by me, and OK'd by Frank van der Linden <fvdl@netbsd.org>.
1.34	02-Dec-1996	thorpej	NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.33	25-Oct-1996	cgd	make the namei struct members ni_dirp and ni_next, and the componentname struct member cn_nameptr 'const', since they should never be used to modify the path name. (Only the pathname buffer, cn_pnbuf, should be modified.) Propagate the const poisoning to code that uses the namei and componentname structs.
1.32	13-Oct-1996	christos	revert kprintf changes
1.31	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.30	07-Jul-1996	fvdl	Start XIDs at a value based on the current time, not 0. This avoids nasty XID confusions with servers that cache them over a long period and with clients that reboot quickly. Problems: because of the sanity check that is done by comparing the system time with filesystem time, XIDs will start at 0 until root is mounted, which means it isn't completely safe for diskless setups. But it's clearly better than it was. It would also be cleaner if all XID handling (more generally, all RPC handling) within the kernel went through the same functions.
1.29	01-Jul-1996	fvdl	We're only handling uio with iovcnt == 1, so don't ever attempt to increment uio_iov, this will get us into nasty trouble. (Thanks to Matthias Drochner for tracking this down).
1.28	23-May-1996	fvdl	* Make mounts with symlinks work (needed for direct mounts with amd). PR #1917 * Never change the NQNFS flag and/or version when just doing an update mount. Fixes a problem that made diskless booting impossible under some circumstances.
1.27	03-Apr-1996	thorpej	branches: 1.27.4; Make these link in the absense of "options FIFO".
1.26	13-Mar-1996	fvdl	Disable invalidating of directory offsets cookies. Should fix one or two directory problems. XXX There is no clean solution to the cookie/cookieverifier validity mess. Together with the disabled strict cookie check, this puts us back at what v2 did in this case. Slightly better solution possible by consequently storing 64bit cookies in other places too.
1.25	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.24	09-Feb-1996	christos	nfs prototype changes
1.23	01-Feb-1996	jtc	Rename struct timespec fields to conform to POSIX.1b
1.22	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.21	08-Sep-1995	ws	No point in computing cn_hash here, as it gets immediately recomputed in lookup
1.20	02-Jun-1995	mycroft	Exported group list now starts at offset 0, not 1.
1.19	01-Jun-1995	jtc	Moved egid credential from cr_groups[0] to new field cr_gid. POSIX.1 requires that sgid executables and the setuid() syscall not change the supplemental group list.
1.18	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.17	17-Aug-1994	mycroft	Convert some more lists and queues.
1.16	17-Aug-1994	mycroft	Change the reply list to a TAILQ.
1.15	22-Jul-1994	mycroft	Set the group list length when copying credentials.
1.14	29-Jun-1994	cgd	branches: 1.14.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.13	13-Jun-1994	mycroft	Move a misplaced #endif.
1.12	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.11	09-Mar-1994	ws	Make FFS optional
1.10	06-Feb-1994	mycroft	Eliminate some more uses of b_actl.
1.9	18-Dec-1993	mycroft	Canonicalize all #includes.
1.8	07-Sep-1993	ws	branches: 1.8.2; Changes to VFS readdir semantics NFS changes for better cookie support ISOFS changes for better Rockridge support and support for generation numbers
1.7	02-Aug-1993	mycroft	Make bpos arg to nfsm_reqh a caddr_t, not a caddr_t*, as that's what it is actually passed.
1.6	13-Jul-1993	cgd	bpos is really a caddr_t **. doesn't really make a diff to the code generated...
1.5	13-Jul-1993	cgd	diskless changes made last time were hosed; were using NULL for "no credentials" rather than NOCRED.
1.4	07-Jul-1993	cgd	changes from ws to support diskless booting... these are "OK" on inspection and after testing... (actually, currently, none of the changed code is even used...)
1.3	21-May-1993	cgd	add rcsid again; fix RCS+crash fuckup
1.2	10-Apr-1993	glass	migrated code to make split possible
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.8.2.2	14-Nov-1993	mycroft	Canonicalize all #includes.
1.8.2.1	24-Sep-1993	mycroft	Make all files using spl*() #include cpu.h. Changes from trunk. nfs_vfsops.c, nfsmount.h: Make nfs_quotactl() take an int rather than a uid_t, as it might be -1. nfs_vnops.c: va_size and va_bytes are now quads.
1.14.2.2	19-Aug-1994	mycroft	update from trunk
1.14.2.1	24-Jul-1994	cgd	from trunk.
1.27.4.3	08-Jul-1996	jtc	Pulled up from rev 1.30 by request from Frank van der Linden
1.27.4.2	02-Jul-1996	jtc	Pulled up from rev 1.29 by request from fvdl
1.27.4.1	25-May-1996	fvdl	Pull in bugfixes from main branch.
1.37.2.1	14-Jan-1997	thorpej	Snapshot of work-in-progress, committed to private branch. These changes implement machine-independent root device and file system selection. Notable features: - All ports behave in a consistent manner regarding root device selection. - No more "options GENERIC"; all kernels have the ability to boot with RB_ASKNAME to select root device and file system type. - Root file system type can be wildcarded; a machine-independent function will try all possible file systems for the selected root device until one succeeds. - If the root file system fails to mount, the operator will be given the chance to select a new root device and file system type, rather than having the machine simply panic. - nfs_mountroot() no longer panics if any part of the NFS mount process fails; it now returns an error, giving the operator a chance to recover. - New, more consistent, config(8) grammar. The constructs: config netbsd swap generic config netbsd root on nfs have been replaced with: config netbsd root on ? type ? config netbsd root on ? type nfs Additionally, the operator may select or wildcard root file system type in the kernel configuration file: config netbsd root on cd0a type cd9660 config(8) now requires that a "root" specification be made. "root" may be wired down or wildcarded. "swap" and "dump" specifications are optional, and follow previous semantics. - config(8) has a new "file-system" keyword, used to configure file systems into the kernel. Eventually, this will be used to generate the default vfssw[]. - "options NFSCLIENT" is obsolete, and is replaced by "file-system NFS". "options NFSSERVER" still exists, since NFS server support is independent of the NFS file system client. - sys/arch/<foo>/<foo>/swapgeneric.c is no longer used, and will be removed; all information is now generated by config(8). As of this commit, all ports except arm32 have been updated to use the new setroot(). Only SPARC, i386, and Alpha ports have been tested at this time. Port masters should test these changes on their ports, and report any problems back to me. More changes are on their way, including RB_ASKNAME support in nfs_mountroot() (to prompt for server address and path) and, potentially, the ability to select rarp/bootparam or bootp in nfs_mountroot().
1.38.2.1	12-Mar-1997	is	Merge in changes from Trunk
1.45.2.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.49.2.1	29-Jan-1998	mellon	Pull up 1.50 (fvdl)
1.69.4.1	02-Aug-1999	thorpej	Update from trunk.
1.69.2.1	10-Sep-1999	he	Pull up revision 1.71: Don't truncate minor numbers > 255 on a NFS client. (is)
1.71.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.71.4.1	15-Nov-1999	fvdl	Sync with -current
1.71.2.7	23-Apr-2001	bouyer	Sync with HEAD.
1.71.2.6	27-Mar-2001	bouyer	Sync with HEAD.
1.71.2.5	12-Mar-2001	bouyer	Sync with HEAD.
1.71.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.71.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.71.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.71.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.75.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.77.2.5	15-Nov-2001	he	Pull up revision 1.98 (requested by fvdl): Always initialize ni_rootdir in nfs_namei(). This fixes a problem where use of ``..'' would instead use the information for ``.'', when using NetBSD/alpha NFS servers, and fixes PR#11618 ands PR#12953.
1.77.2.4	06-Apr-2001	he	Pull up revision 1.93 (requested by wrstuden): Explicitly VOP_UNLOCK before setting v_op to spec_vnode_ops_p. Works around a lock leak and eventual kernel panic.
1.77.2.3	14-Dec-2000	he	Pull up revisions 1.85,1.87 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.77.2.2	30-Oct-2000	tv	Pullup 1.87 [fvdl]: Do not accept vnode type changes to an active node. This may wreak havoc if the server erroneously uses the same filehandle for different files. This changes back revision 1.28; the PR that that revision fixed doesn't apply anymore, it has been verified not to be a problem with this change.
1.77.2.1	20-Jun-2000	tv	file nfs_subs.c was added on branch netbsd-1-5 on 2000-10-30 22:22:58 +0000
1.92.2.11	11-Dec-2002	thorpej	Sync with HEAD.
1.92.2.10	11-Nov-2002	nathanw	Catch up to -current
1.92.2.9	22-Oct-2002	thorpej	Sync with HEAD.
1.92.2.8	27-Aug-2002	nathanw	Catch up to -current.
1.92.2.7	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.92.2.6	28-Feb-2002	nathanw	Catch up to -current.
1.92.2.5	14-Nov-2001	nathanw	Catch up to -current.
1.92.2.4	08-Oct-2001	nathanw	Catch up to -current.
1.92.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.92.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.92.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.95.4.1	01-Oct-2001	fvdl	Catch up with -current.
1.95.2.6	30-Sep-2002	jdolecek	add support for kevents to NFS to detect file changes on server by other NFS clients, polling kernel thread is used to periodically check for attribute changes of watched files; the NFS server is only contacted when the vnode expires from local attrcache (which takes 5-60 seconds currently), to keep network&CPU overhead low the routine checking for remote changes is quite simplistic, but hopefully doing it's job well enough
1.95.2.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.95.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.95.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.95.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.95.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.98.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.103.6.1	04-Oct-2003	tron	Pull up revision 1.129 (requested by martti in ticket #1506): plug mbuf leak due to manual mbuf handling. PR kern/13807. (martti confirmed that it stabilizes the situation described in kren/13807)
1.103.4.1	29-Aug-2002	gehenna	catch up with -current.
1.125.2.13	11-Dec-2005	christos	Sync with head.
1.125.2.12	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.125.2.11	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.125.2.10	04-Feb-2005	skrll	Sync with HEAD.
1.125.2.9	24-Jan-2005	skrll	Sync with HEAD.
1.125.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.125.2.7	02-Nov-2004	skrll	Sync with HEAD.
1.125.2.6	19-Oct-2004	skrll	Sync with HEAD
1.125.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.125.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.125.2.3	24-Aug-2004	skrll	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.125.2.2	03-Aug-2004	skrll	Sync with HEAD
1.125.2.1	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.
1.132.2.5	16-Mar-2005	tron	Pull up revision 1.146 (requested by yamt in ticket #1148): nfs_clearcommit: don't attempt to clear commit info (n_pushlo, etc) unless the vnode is of VREG. union members used to keep commit info are used for other purposes in the case of !VREG.
1.132.2.4	11-Jan-2005	jmc	Pullup rev 1.140 (requested by yamt in ticket #1079) Invalidate cache if filesize is changed besides our activity because it means that were out of sync with the server.
1.132.2.3	04-Oct-2004	jmc	branches: 1.132.2.3.2; Pullup rev 1.137 (requested by yamt in ticket #889) nfs_enterdircache: initialize dc_flags of a newly allocated dircache entry.
1.132.2.2	18-Sep-2004	he	Pull up revision 1.135 (requested by yamt in ticket #858): Fix access-after-free bugs in dircache code by reference counting nfsdircache. Fixes PR#26864.
1.132.2.1	21-Jun-2004	tron	Pull up revision 1.134 (requested by yamt in ticket #515): nfs_searchdircache: fix a null dereference in the case that offset!=0 and dircache hasn't been initialized yet.
1.132.2.3.2.4	27-Oct-2005	riz	Pull up following revision(s) (requested by christos in ticket #5863): sys/nfs/nfs_subs.c: revision 1.152 via patch sys/nfs/nfs.h: revision 1.49 sys/nfs/nfs_vfsops.c: revision 1.149 via patch usr.sbin/amd/include/config.h: revision 1.36 sys/nfs/nfs_vnops.c: revision 1.227 via patch sys/nfs/nfsmount.h: revision 1.34 Allow the attribute cache to be turned off, and allow amd to do it.
1.132.2.3.2.3	16-Mar-2005	tron	Pull up revision 1.146 (requested by yamt in ticket #1148): nfs_clearcommit: don't attempt to clear commit info (n_pushlo, etc) unless the vnode is of VREG. union members used to keep commit info are used for other purposes in the case of !VREG.
1.132.2.3.2.2	30-Jan-2005	he	Pull up revision 1.138 (requested by yamt in ticket #968): Since daddr_t is 64-bit these days, simply use nfs directory cookies as buffer cache indexes. This should make the regress/sys/fs/getdents test work. Fixes PR#27112.
1.132.2.3.2.1	11-Jan-2005	jmc	Pullup rev 1.140 (requested by yamt in ticket #1079) Invalidate cache if filesize is changed besides our activity because it means that were out of sync with the server.
1.140.2.1	29-Apr-2005	kent	sync with -current
1.143.2.3	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.143.2.2	12-Feb-2005	yamt	sync with head.
1.143.2.1	25-Jan-2005	yamt	file nfs_subs.c was added on branch yamt-km on 2005-02-12 18:17:55 +0000
1.148.2.1	27-Sep-2005	tron	Pull up following revision(s) (requested by christos in ticket #816): sys/nfs/nfs_subs.c: revision 1.152 ATTRTIMEO takes 2 args.
1.149.2.13	17-Mar-2008	yamt	sync with head.
1.149.2.12	27-Feb-2008	yamt	revert incomplete nfs client locking for now.
1.149.2.11	27-Feb-2008	yamt	sync with head.
1.149.2.10	15-Feb-2008	yamt	- sprinkle some locks. - disable MNT_UPDATE because it involves too much locking headache. - don't overwrite other bits in v_vflags when setting VV_ROOT.
1.149.2.9	04-Feb-2008	yamt	sync with head.
1.149.2.8	21-Jan-2008	yamt	sync with head
1.149.2.7	07-Dec-2007	yamt	sync with head
1.149.2.6	15-Nov-2007	yamt	sync with head.
1.149.2.5	03-Sep-2007	yamt	sync with head.
1.149.2.4	26-Feb-2007	yamt	sync with head.
1.149.2.3	30-Dec-2006	yamt	sync with head.
1.149.2.2	21-Jun-2006	yamt	sync with head.
1.149.2.1	07-Jul-2005	yamt	adapt to mbuf.h changes.
1.153.6.2	29-Nov-2005	yamt	sync with head.
1.153.6.1	22-Nov-2005	yamt	sync with head.
1.156.2.2	01-Feb-2006	yamt	sync with head.
1.156.2.1	31-Dec-2005	yamt	- adapt nfs. - nfs_doio_read: #if 0 out "killproc if text is modified" part of the code as it's broken. (a process reading the modified text is not necessarily a process which is using the file as a text.)
1.157.4.3	01-Jun-2006	kardel	Sync with head.
1.157.4.2	22-Apr-2006	simonb	Sync with head.
1.157.4.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.157.2.1	09-Sep-2006	rpaulo	sync with head
1.158.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.158.4.9	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.
1.158.4.8	19-Apr-2006	elad	sync with head.
1.158.4.7	14-Mar-2006	elad	Use kauth_cred_clone() where appropriate.
1.158.4.6	13-Mar-2006	elad	kauth_cred_clone() takes care of the groups for us, remove redundant code.
1.158.4.5	12-Mar-2006	elad	Fix group cleaning loop, as pointed out by yamt@. While I'm here, add an XXX near the second loop that watches NGROUPS; this should be internal to kauth(9).
1.158.4.4	11-Mar-2006	elad	Replace check for euid == 0 with kauth_authorize_generic().
1.158.4.3	10-Mar-2006	elad	Cleanup more interface abuse. Make nfsrv_setcred() take a kauth_cred_t * as outcred. The original code just modified it directly; we can't do that, nor do we want to. Get rid of another case of kauth_cred_zero() followed by kauth_cred_hold() and use kauth_cred_clone() to make sure we don't leave out important members. Add another DIAGNOSTIC check for reference count of above one. Again, this should be tested.
1.158.4.2	10-Mar-2006	elad	Remove some more no longer needed calls to kauth_cred_setngroups() and nfsrvw_sort().
1.158.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.158.2.5	14-Sep-2006	yamt	sync with head.
1.158.2.4	03-Sep-2006	yamt	sync with head.
1.158.2.3	11-Aug-2006	yamt	sync with head
1.158.2.2	26-Jun-2006	yamt	sync with head.
1.158.2.1	24-May-2006	yamt	sync with head.
1.164.2.2	19-Jun-2006	chap	Sync with head.
1.164.2.1	19-May-2006	chap	file nfs_subs.c was added on branch chap-midi on 2006-06-19 04:10:37 +0000
1.165.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.170.4.2	10-Dec-2006	yamt	sync with head.
1.170.4.1	22-Oct-2006	yamt	sync with head
1.170.2.2	12-Jan-2007	ad	Sync with head.
1.170.2.1	18-Nov-2006	ad	Sync with head.
1.177.2.1	17-Feb-2007	tron	Apply patch (requested by chs in ticket #422): - Fix various deadlock problems with nullfs and unionfs. - Speed up path lookups by upto 25%.
1.180.2.3	07-May-2007	yamt	sync with head.
1.180.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.180.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.184.4.1	11-Jul-2007	mjf	Sync with head.
1.184.2.9	24-Oct-2007	ad	Do locking / use marker vnodes when traversing mountpoint vnode lists.
1.184.2.8	16-Sep-2007	ad	Checkpoint work in progress on the vnode lifecycle and reference counting stuff. This makes it work properly without kernel_lock and fixes a few quite old bugs. See vfs_subr.c 1.283.2.17 for details.
1.184.2.7	20-Aug-2007	ad	Sync with HEAD.
1.184.2.6	15-Jul-2007	ad	Sync with head.
1.184.2.5	15-Jul-2007	ad	Sync with head.
1.184.2.4	01-Jul-2007	ad	Adapt to callout API change.
1.184.2.3	18-Jun-2007	yamt	fix a merge botch.
1.184.2.2	09-Jun-2007	ad	Sync with head.
1.184.2.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.190.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.191.12.2	27-Jul-2007	yamt	stop nfs tick when we have nothing to do.
1.191.12.1	27-Jul-2007	yamt	file nfs_subs.c was added on branch matt-mips64 on 2007-07-27 10:03:59 +0000
1.191.10.1	13-Nov-2007	bouyer	Sync with HEAD
1.191.6.3	23-Mar-2008	matt	sync with HEAD
1.191.6.2	09-Jan-2008	matt	sync with HEAD
1.191.6.1	06-Nov-2007	matt	sync with HEAD
1.191.4.3	09-Dec-2007	jmcneill	Sync with HEAD.
1.191.4.2	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.191.4.1	29-Oct-2007	joerg	Sync with HEAD.
1.192.2.3	18-Feb-2008	mjf	Sync with HEAD.
1.192.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.192.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.193.2.2	26-Dec-2007	ad	Sync with head.
1.193.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.194.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.199.6.6	17-Jan-2009	mjf	Sync with HEAD.
1.199.6.5	05-Oct-2008	mjf	Sync with HEAD.
1.199.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.199.6.3	05-Jun-2008	mjf	Sync with HEAD. Also fix build.
1.199.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.199.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.199.2.1	24-Mar-2008	keiichi	sync with head.
1.201.4.7	10-Oct-2010	yamt	some locking changes
1.201.4.6	26-Sep-2010	yamt	locking changes
1.201.4.5	11-Mar-2010	yamt	sync with head
1.201.4.4	16-May-2009	yamt	sync with head
1.201.4.3	04-May-2009	yamt	sync with head.
1.201.4.2	16-May-2008	yamt	sync with head.
1.201.4.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.201.2.2	17-Jun-2008	yamt	sync with head.
1.201.2.1	18-May-2008	yamt	sync with head.
1.202.2.3	10-Oct-2008	skrll	Sync with HEAD.
1.202.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.202.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.204.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.204.4.1	19-Oct-2008	haad	Sync with HEAD.
1.204.2.1	18-Jul-2008	simonb	Sync with head.
1.209.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.209.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.212.2.2	23-Jul-2009	jym	Sync with HEAD.
1.212.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.218.2.2	27-May-2010	uebayasi	Include uvm/uvm.h, because this touches uvm internal.
1.218.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.219.2.2	05-Mar-2011	rmind	sync with head
1.219.2.1	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.
1.220.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.221.2.6	22-May-2014	yamt	fix a merge botch
1.221.2.5	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")
1.221.2.4	17-Apr-2012	yamt	sync with head
1.221.2.3	26-Nov-2011	yamt	update after radixtree.h api changes
1.221.2.2	06-Nov-2011	yamt	remove pg->listq and uobj->memq
1.221.2.1	02-Nov-2011	yamt	page cache related changes - maintain object pages in radix tree rather than rb tree. - reduce unnecessary page scan in putpages. esp. when an object has a ton of pages cached but only a few of them are dirty. - reduce the number of pmap operations by tracking page dirtiness more precisely in uvm layer. - fix nfs commit range tracking. - fix nfs write clustering. XXX hack
1.222.12.1	18-May-2014	rmind	sync with head
1.222.8.2	03-Dec-2017	jdolecek	update from HEAD
1.222.8.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.225.2.2	10-Aug-2014	tls	Rebase.
1.225.2.1	17-Jul-2014	tls	Adjustments to the "earlyentropy" branch in response to the various discussions beginning with my initial proposal http://mail-index.netbsd.org/tech-kern/2014/04/08/msg016876.html and particularly the long discussion of cprng_fast() performance (e.g. https://mail-index.netbsd.org/tech-crypto/2014/04/21/msg000642.html). In particular: * Per-CPU, lockless cprng_fast replacement using Dennis Ferguson's "ccrand" implementation of ChaCha8. * libkern arc4random() is gone, gone, gone. * Entropy estimator reverted to 32-bit recordkeeping and timestamps per Dennis' comments and analysis. * LZF entropy estimator removed: it required a great deal of state, and rejected only truly pathological input. I have not yet reverted the changes that provide LZF in the kernel as generic functionality; I will likely revert those changes prior to any merge of this branch to HEAD.
1.227.4.2	28-Aug-2017	skrll	Sync with HEAD
1.227.4.1	09-Jul-2016	skrll	Sync with HEAD
1.228.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.228.2.1	26-Apr-2017	pgoyette	Sync with HEAD
1.229.6.1	08-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #856): sys/nfs/nfs.h: revision 1.76 sys/nfs/nfs_subs.c: revision 1.230 sys/nfs/nfs_socket.c: revision 1.199 sys/nfs/nfs_clntsocket.c: revision 1.6 PR/40491: From Tobias Ulmer in tech-kern@: 1. Protect the nfs request queue with its own mutex 2. make the nfs_receive queue check for signals so that intr mounts can be interrupted. XXX: pullup-8
1.230.2.4	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.230.2.3	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.230.2.2	21-May-2018	pgoyette	Sync with HEAD
1.230.2.1	02-May-2018	pgoyette	Synch with HEAD
1.232.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.232.2.1	10-Jun-2019	christos	Sync with HEAD
1.236.2.1	29-Feb-2020	ad	Sync with head.
1.164	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.163	04-Jun-2021	hannken	branches: 1.163.16; Add flag/command NFSSVC_REPLACEEXPORTSLIST to nfssvc(2) system call. Works like NFSSVC_SETEXPORTSLIST but supports "mel_nexports > 1" and will atomically update the complete exports list for a file system.
1.162	14-Mar-2020	ad	branches: 1.162.8; 1.162.12; - Hide the details of SPCF_SHOULDYIELD and related behind a couple of small functions: preempt_point() and preempt_needed(). - preempt(): if the LWP has exceeded its timeslice in kernel, strip it of any priority boost gained earlier from blocking.
1.161	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.160	16-Mar-2018	christos	branches: 1.160.2; PR/53103: Timo Buhrmester: linux emulation of sendto(2) broken The sockargs refactoring broke it, because sockargs only works with a user address. Added an argument to sockargs to indicate where the address is coming from. Welcome to 8.99.14.
1.159	25-Jan-2018	riastradh	branches: 1.159.2; Use a random opaque cookie, not kva pointer, for nfssvc(2). (What were they smoking?!) I suspect most of this is actually dead code that wasn't properly amputated along with the rest of the gangrene of NFSKERB a decade ago, but I'm out of time to investigate further. If someone else wants to kill NFSSVC_AUTHIN/NFSSVC_AUTHINFAIL and the rest of the tentacular kerberosity, be my guest. Noted by Silvio Cesare of InfoSect.
1.158	12-Feb-2017	maxv	Memory leak, found by Mootja; not tested, but obvious enough.
1.157	10-Jun-2016	ozaki-r	branches: 1.157.2; 1.157.4; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.156	22-Jun-2015	mrg	add netbsd32 support for nfssvc(2). we do this by defining 5 copyin/out functions that do all the ugly work, are just plain copyin/out for the native system calls, and do the necessary translations for netbsd32. with this i'm able to run 32 bit nfsd and mountd on 64 bit kernel and mount the file systems remotely.
1.155	05-Sep-2014	matt	branches: 1.155.2; Try not to use f_data, use f_{vnode,socket,pipe,mqueue,kqueue,ksem} to get a correctly typed pointer.
1.154	27-Nov-2013	christos	branches: 1.154.4; CID 271162: NULL deref check
1.153	31-Dec-2009	christos	branches: 1.153.12; 1.153.22; 1.153.26; put nuidhash_max in a file that is shared between server and client code.
1.152	31-Dec-2009	christos	handle the nuidhash_max lossage differently
1.151	21-Oct-2009	rmind	Remove uarea swap-out functionality: - Addresses the issue described in PR/38828. - Some simplification in threading and sleepq subsystems. - Eliminates pmap_collect() and, as a side note, allows pmap optimisations. - Eliminates XS_CTL_DATA_ONSTACK in scsipi code. - Avoids few scans on LWP list and thus potentially long holds of proc_lock. - Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k. - Removes __SWAP_BROKEN cases. Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on acorn26 (thanks to <bjh21>). Discussed on <tech-kern>, reviewed by <ad>.
1.150	14-Sep-2009	pooka	Remove stale comment about super user. no functional change
1.149	07-Jul-2009	christos	The compatibility call to re-export from sys_mount() calls mountd_set_exports_list, with the mnt_updating mutex held. Account for that to avoid a locking against myself panic.
1.148	23-May-2009	ad	Remove pointless error check.
1.147	10-Apr-2009	bouyer	PR kern/41154: possible races in NFS server code Fix some of the races (but probably not all of them) in the NFS server code. nfssvc_nfsd(): change a splsoftclock()/spx() to mutex_enter/exit(&nfsd_lock) (I guess it was forgotten when the nfsd code was made SMP safe) m_freem(nd_nam) in nfsrv_slpderef() instead of nfsrv_zapsock() to avoid possible use after free in nfssvc_nfsd() Fix nfsrv_slpderef() to not release nfsd_lock before testing SLP_VALID and reaquiring it just after. This could cause a use after free of the slp if one thread is in nfsrv_slpderef() and the other one grabs slp from nfssvc_sockpending and zap it.
1.146	15-Mar-2009	cegger	ansify function definitions
1.145	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.144	14-Mar-2009	dsl	Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.143	28-Nov-2008	pooka	branches: 1.143.4; g/c unused malloc types
1.142	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.141	14-Nov-2008	ad	Remove COMPAT ifdefs that might as well be comments (i.e., they cost us almost nothing).
1.140	09-Oct-2008	christos	branches: 1.140.2; 1.140.4; do the proper ifdef dance for non-inet families
1.139	28-Sep-2008	pooka	Don't free nd_mrep in case of no reply. It is (at least in one case) freed already within the rpc handler. XXX: this line and another was originally committed with "don't leak mbufs", but given that currently it can double-free an mbuf and essentially crash the system, I'll opt for the leak. Needless to say, this needs revisiting, but that requires a large scale campaign due to the sticky nature of nfsm love.
1.138	06-Aug-2008	plunky	Convert socket options code to use a sockopt structure instead of laying everything into an mbuf. approved by core
1.137	24-Jun-2008	ad	branches: 1.137.2; Replace references to getsock/getvnode.
1.136	20-May-2008	ad	branches: 1.136.2; Make it compile.
1.135	28-Apr-2008	yamt	branches: 1.135.2; as softint network processing is now safe to block, make some mutexes adaptive.
1.134	24-Apr-2008	ad	branches: 1.134.2; Merge the socket locking patch: - Socket layer becomes MP safe. - Unix protocols become MP safe. - Allows protocol processing interrupts to safely block on locks. - Fixes a number of race conditions. With much feedback from matt@ and plunky@.
1.133	23-Mar-2008	rmind	branches: 1.133.2; G/C l->l_locks. OK by <ad>.
1.132	21-Mar-2008	ad	Catch up with descriptor handling changes. See kern_descrip.c revision 1.173 for details.
1.131	28-Feb-2008	elad	Introduce a new kauth action, KAUTH_NETWORK_NFS, and two requests, KAUTH_REQ_NETWORK_NFS_EXPORT and KAUTH_REQ_NETWORK_NFS_SVC, and use them to replace two KAUTH_GENERIC_ISSUSER calls in the NFS code. Also replace two more with KAUTH_SYSTEM_MKNOD, where appropriate. Documetnation and examples updated. More to come.
1.130	02-Jan-2008	yamt	branches: 1.130.2; 1.130.6; use kmem_alloc instead of malloc.
1.129	02-Jan-2008	ad	Merge vmlocking2 to head.
1.128	20-Dec-2007	dsl	Convert all the system call entry points from: int foo(struct lwp l, void v, register_t retval) to: int foo(struct lwp l, const struct foo_args uap, register_t retval) Fixup compat code to not write into 'uap' and (in some cases) to actually pass a correctly formatted 'uap' structure with the right name to the next routine. A few 'compat' routines that just call standard ones have been deleted. All the 'compat' code compiles (along with the kernels required to test build it). 98% done by automated scripts.
1.127	04-Dec-2007	yamt	branches: 1.127.4; merge non-intrusive nfs changes from vmlocking.
1.126	22-Nov-2007	yamt	branches: 1.126.2; nfssvc_nfsd: remove a wrong assertion.
1.125	08-Oct-2007	ad	branches: 1.125.2; 1.125.4; Merge file descriptor locking, cwdi locking and cross-call changes from the vmlocking branch.
1.124	10-Aug-2007	yamt	branches: 1.124.2; 1.124.4; - instead of scanning an array of iods, maintain a list of idle iods. - make nfs_getset_niothreads MP friendly.
1.123	08-Aug-2007	yamt	push kernel_lock a little.
1.122	02-Aug-2007	yamt	branches: 1.122.2; 1.122.4; nfsrv_slpderef: add an assertion.
1.121	02-Aug-2007	yamt	nfssvc_nfsd: fix a wrong assertion. PR/36710 from Tobias Nygren.
1.120	02-Aug-2007	yamt	nfsrv_zapsock: update SLP_DOREC for consistency.
1.119	02-Aug-2007	yamt	nfssvc_nfsd: don't leave sockets with SLP_DISCONN.
1.118	20-Jul-2007	yamt	- fix decreasing of vfs.nfs.iothreads after the recent partial merge of vmlocking. - don't make nfsiod exit with requests left. - make NFSSVC_BIOD a dummy so that nfsiod can be simplified.
1.117	09-Jul-2007	ad	branches: 1.117.2; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.116	22-Jun-2007	yamt	- nfsrv_slpderef: fix a locking botch. (should fix "slp->ns_sref == 0" assertion failures in nfsrv_init) - add some related assertions.
1.115	19-Jun-2007	yamt	nfssvc_iod: fix nm_bufqiods accounting on exit.
1.114	12-Jun-2007	yamt	- nfssvc_nfsd: clear nfsd_slp when exiting. (fix an assertion failure in rev.1.112.) - nfsrv_init: add assertions.
1.113	01-Jun-2007	yamt	nfssvc_nfsd: check SPCF_SHOULDYIELD and yield cpu.
1.112	01-Jun-2007	yamt	nfssvc_nfsd: add assertions.
1.111	01-Jun-2007	yamt	use mutex and condvar.
1.110	30-Apr-2007	yamt	fix a lock leak in rev.1.109. pointed by Mindaugas R.
1.109	29-Apr-2007	yamt	use mutex and condver.
1.108	19-Apr-2007	yamt	hold proclist_mutex when calling psignal().
1.107	04-Mar-2007	christos	branches: 1.107.2; 1.107.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.106	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.105	09-Feb-2007	ad	branches: 1.105.2; Merge newlock2 to head.
1.104	17-Jan-2007	yamt	plug a mbuf leak.
1.103	04-Jan-2007	elad	Consistent usage of KAUTH_GENERIC_ISSUSER.
1.102	28-Dec-2006	yamt	remove several nqnfs definitions.
1.101	27-Dec-2006	yamt	- remove the rest of nqnfs. - reject NFSMNT_MNTD and NFSMNT_KERB. (no users in tree.)
1.100	27-Dec-2006	yamt	remove nqnfs.
1.99	09-Nov-2006	yamt	branches: 1.99.2; 1.99.4; remove some __unused in function parameters.
1.98	22-Oct-2006	pooka	kauth_cred_uucvt() -> kauth_uucred_to_cred(), introduce kauth_cred_to_uucred() per tech-kern proposal
1.97	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.96	23-Jul-2006	ad	branches: 1.96.4; 1.96.6; Use the LWP cached credentials where sane.
1.95	07-Jun-2006	kardel	merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.94	19-May-2006	yamt	branches: 1.94.2; - fix compilation problem for !NFSSERVER && NFS. pointed by Tom Spindler on source-changes@. - make nfs_srvdesc_pool static.
1.93	18-May-2006	yamt	- fix some leaks in nfsd, introduced by kauth changes. - simplify code. - add some assertions. - wrap some long lines. - remove an unnecessary ";".
1.92	14-May-2006	elad	integrate kauth.
1.91	10-May-2006	mrg	quell GCC 4.1 uninitialised variable warnings. XXX: we should audit the tree for which old ones are no longer needed after getting the older compilers out of the tree..
1.90	15-Apr-2006	christos	Coverity CID 1162: Prevent NULL deref.
1.89	15-Apr-2006	christos	Coverity CID 1165: Cannot have nfsiod without an lwp, so remove the superfluous test.
1.88	05-Jan-2006	yamt	branches: 1.88.2; 1.88.4; 1.88.6; 1.88.8; 1.88.10; ensure the export list is not changed during nfsd operations.
1.87	03-Jan-2006	yamt	fix a deadlock due to a spl problem.
1.86	03-Jan-2006	yamt	nfssvc_nfsd: reduce a chance for a slow peer to capture all our threads. instead of sleeping to wait for the socket to send our reply, just hand-off our reply to the thread which is holding the socket.
1.85	03-Jan-2006	yamt	improve nfsd locking. - don't bother to take nfs_sndlock when doing nfsrv_rcv. unlike client, we never reconnect. - nfsrv_getstream: fix the case that m_split sleeps. - free socket in nfsrv_slpderef rather than nfsrv_zapsock. fix race with nfssvc_nfsd. - while i'm here, remove NFSD_WAITING and NFSD_REQINPROG as they are redundant. - some comments and assertions.
1.84	11-Dec-2005	christos	branches: 1.84.2; merge ktrace-lwp.
1.83	25-Nov-2005	thorpej	Use a once control to initialize the NFS server / client shared data from nfs_vfs_init() or sys_nfssvc(). Remove the nfs_init() call from main().
1.82	23-Sep-2005	jmmv	branches: 1.82.6; Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.81	11-Sep-2005	rpaulo	Wrap a multiple line comment so that it doesn't go beyond 80 columns.
1.80	03-Aug-2005	onoe	Fix mbuf leak in nfssvc_nfsd().
1.79	07-Jul-2005	christos	1. use p = uio->uio_procp consistently and eliminate suspicious uses of curproc (where uio->uio_procp should be used?). Don't do this for nfs_commit(), because yamt says it is possibly wrong. 2. nfs_doio() does not use struct proc; remove it and the code to compute it. 3. use copyin_proc() and copyout_proc() instead of copyin() and copyout(). 4. check return of copyout_proc(). and mark return from copyin_proc() XXX 5. Eliminate check p == curproc assertion check from nfs_write; nfs_read does not have it and we might be called in a different process context anyway (PR 20138).
1.78	26-Feb-2005	perry	branches: 1.78.4; nuke trailing whitespace
1.77	10-Jun-2004	yamt	branches: 1.77.4; 1.77.6; make sure that nfssvc sockets are zapped before being freed.
1.76	10-Jun-2004	yamt	nfsrv_zapsock: fix an inverted condition in nfs_syscall.c rev.1.74.
1.75	22-May-2004	jonathan	Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.74	17-Mar-2004	yamt	branches: 1.74.2; nfsrv_zapsock: zap an nfsd socket only if it's valid.
1.73	17-Mar-2004	yamt	nfsrv_zapsock: remove slp from nfssvc_sockpending before zapping.
1.72	17-Mar-2004	yamt	SHUT_RDWR rather than bare 2.
1.71	07-Dec-2003	thorpej	Fix a couple of small whitespace errors.
1.70	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.69	29-Jun-2003	fvdl	branches: 1.69.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.68	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
1.67	26-Jun-2003	yamt	add appropriate #ifdef's. pointed by Simon Burge on source-changes and by some people in private mail.
1.66	25-Jun-2003	yamt	- instead of scaning a list when looking up {a idle thread, a socket with pending requests}, maintain dedicated list of them. - add spin locks.
1.65	26-May-2003	yamt	make duplicated codes to a function, nfsrv_sockalloc.
1.64	22-May-2003	yamt	poolify nfsrv_descript.
1.63	22-May-2003	yamt	indent (nfssvc_nfsd)
1.62	21-May-2003	yamt	- use FREE not free for MALLOC'ed memory. - remove unneeded caddr_t casts.
1.61	21-May-2003	yamt	remove local definitions of TRUE and FALSE.
1.60	07-May-2003	yamt	simple lock for nfs iod.
1.59	07-May-2003	yamt	indent.
1.58	09-Apr-2003	yamt	make per-iod datas together.
1.57	02-Apr-2003	yamt	use queue manipulation macros.
1.56	31-Mar-2003	yamt	adapt to file interlock.
1.55	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.54	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.53	18-Jan-2003	thorpej	Merge the nathanw_sa branch.
1.52	14-Sep-2002	chs	pick up a fix from openbsd: revision 1.33 date: 2002/07/24 23:32:11; author: nordin; state: Exp; lines: +3 -3 Use sizeof(array) instead of sizeof(array *) for bcopy length. ok deraadt@
1.51	12-May-2002	matt	Eliminate commons
1.50	29-Nov-2001	christos	sprinkle crcvt()
1.49	10-Nov-2001	lukem	add RCSIDs
1.48	27-Nov-2000	chs	branches: 1.48.2; 1.48.4; 1.48.8; Initial integration of the Unified Buffer Cache project.
1.47	24-Nov-2000	chs	put more ISO bits under ifdef ISO.
1.46	24-Oct-2000	matt	Change a DIAGNOSTIC panic slightly to print the locked vnodes and to just print a diagnostic but not panic.
1.45	23-Oct-2000	chs	fix nfs iod management so we don't lose i/os when iods die.
1.44	19-Sep-2000	fvdl	Don't do write gathering for v3; it makes no sense. Unless the client is broken and does sync writes all the time, but that's the client's fault.
1.43	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.42	23-Aug-2000	nathanw	Fix typo in comment.
1.41	09-Jun-2000	fvdl	branches: 1.41.2; Some tweaks to enable NFS over IPv6. The special-casing of AF_INET should really be removed.
1.40	07-May-2000	tsarna	branches: 1.40.2; Auto-adjusting vfs.nfs.iothreads: when mounting the first nfs filesystem, if the number of threads is "-1", meaning it's never been set, then set it to 4. You can override by setting this to some other number (including 0) before or after mounting, of course. Thanks to whoever it was that suggested this on ICB... sorry I don't remember who.
1.39	15-Apr-2000	tsarna	Death to nfsiod! It is replaced by kernel threads that do the same thing. The number of kernel threads used is set with the vfs.nfs.iothreads sysctl.
1.38	30-Mar-2000	augustss	Remove register declarations.
1.37	30-Mar-2000	simonb	Delete redundant decl of nfs_pub - it's in <sys/mount.h>. Delete redundant decl of nfsrv_zapsock() - it's in <nfs/nfs_var.h>.
1.36	29-Jun-1999	wrstuden	branches: 1.36.2; Add fhopen, fhstat, fhstatfs syscalls. Also move getfh in from the nfs syscall code.
1.35	05-May-1999	thorpej	Add "use counting" to file entries. When closing a file, and it's reference count is 0, wait for use count to drain before finishing the close. This is necessary in order for multiple processes to safely share file descriptor tables.
1.34	04-May-1999	sommerfe	Include checks (under DIAGNOSTIC) to catch vnode lock leaks soon after they happen (while we still know which remote op is to blame for it), instead of later when we trip over the already-locked vnode.
1.33	08-Nov-1998	mycroft	branches: 1.33.6; Do not permit the u area for nfsd or nfsiod to be swapped out.
1.32	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.31	05-Jul-1998	jonathan	defopt ISO TPIP.
1.30	25-Jun-1998	thorpej	defopt NFSSERVER
1.29	25-Apr-1998	matt	Adapt to new sosend/soreceive and upcall (now down in sowakeup)
1.28	19-Feb-1998	thorpej	Include the NFS option header.
1.27	10-Oct-1997	fvdl	branches: 1.27.2; * New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.26	24-Jun-1997	fvdl	branches: 1.26.4; Invalidate nfs_pub info when reinitting the NFS server.
1.25	24-Mar-1997	mycroft	KNF police.
1.24	31-Jan-1997	thorpej	NFSCLIENT -> NFS.
1.23	03-Dec-1996	thorpej	branches: 1.23.2; Make NFSSERVER work without NFSCLIENT. This is achieved by splitting the client and server/shared data initialization into separate functions, and calling the server/shared initialization directly from main(). Problem noted in PR #1308 (Kenneth Stailey) and PR #1780 (Chris Demetriou). Fix suggested in PR #1780 by Chris Demetriou, and munged a bit by me, and OK'd by Frank van der Linden <fvdl@netbsd.org>.
1.22	02-Dec-1996	thorpej	NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.21	13-Oct-1996	christos	revert kprintf changes
1.20	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.19	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.18	09-Feb-1996	christos	nfs prototype changes
1.17	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.16	07-Oct-1995	mycroft	Prefix names of system call implementation functions with `sys_'.
1.15	19-Sep-1995	thorpej	Make system calls conform to a standard prototype and bring those prototypes into scope.
1.14	13-Aug-1995	mycroft	splnet --> splsoftnet
1.13	20-Oct-1994	cgd	update for new syscall args description mechanism
1.12	17-Aug-1994	mycroft	Convert some more lists and queues.
1.11	29-Jun-1994	cgd	branches: 1.11.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.10	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.9	14-Feb-1994	cgd	be more intelligent with credentials, so nfsd's don't inherit strange credentials. This doesn't actually have any effect on performance, because the remote cred is used for all operations, anyway. however, it makes "ps" et al. look normal, because the proc's ucred is no longer clobbered.
1.8	06-Feb-1994	mycroft	Eliminate some more uses of b_actl.
1.7	18-Dec-1993	mycroft	Canonicalize all #includes.
1.6	17-Jul-1993	mycroft	branches: 1.6.4; Finish moving struct definitions outside of function declarations.
1.5	22-May-1993	cgd	add include of select.h if necessary for protos, or delete if extraneous
1.4	18-May-1993	cgd	make kernel select interface be one-stop shopping & clean it all up.
1.3	10-Apr-1993	glass	migrated code to make split possible
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
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.6.4.1	14-Nov-1993	mycroft	Canonicalize all #includes.
1.11.2.1	19-Aug-1994	mycroft	update from trunk
1.23.2.1	14-Jan-1997	thorpej	Snapshot of work-in-progress, committed to private branch. These changes implement machine-independent root device and file system selection. Notable features: - All ports behave in a consistent manner regarding root device selection. - No more "options GENERIC"; all kernels have the ability to boot with RB_ASKNAME to select root device and file system type. - Root file system type can be wildcarded; a machine-independent function will try all possible file systems for the selected root device until one succeeds. - If the root file system fails to mount, the operator will be given the chance to select a new root device and file system type, rather than having the machine simply panic. - nfs_mountroot() no longer panics if any part of the NFS mount process fails; it now returns an error, giving the operator a chance to recover. - New, more consistent, config(8) grammar. The constructs: config netbsd swap generic config netbsd root on nfs have been replaced with: config netbsd root on ? type ? config netbsd root on ? type nfs Additionally, the operator may select or wildcard root file system type in the kernel configuration file: config netbsd root on cd0a type cd9660 config(8) now requires that a "root" specification be made. "root" may be wired down or wildcarded. "swap" and "dump" specifications are optional, and follow previous semantics. - config(8) has a new "file-system" keyword, used to configure file systems into the kernel. Eventually, this will be used to generate the default vfssw[]. - "options NFSCLIENT" is obsolete, and is replaced by "file-system NFS". "options NFSSERVER" still exists, since NFS server support is independent of the NFS file system client. - sys/arch/<foo>/<foo>/swapgeneric.c is no longer used, and will be removed; all information is now generated by config(8). As of this commit, all ports except arm32 have been updated to use the new setroot(). Only SPARC, i386, and Alpha ports have been tested at this time. Port masters should test these changes on their ports, and report any problems back to me. More changes are on their way, including RB_ASKNAME support in nfs_mountroot() (to prompt for server address and path) and, potentially, the ability to select rarp/bootparam or bootp in nfs_mountroot().
1.26.4.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.27.2.1	08-Nov-1998	cgd	pull up rev 1.33 from trunk (mycroft)
1.33.6.2	25-Oct-2000	he	Pull up revision 1.45 (via patch, requested by chs): Fix a bug where NFS async I/O requests will be lost if the number of I/O daemons is ever reduced to a smaller but non-zero number.
1.33.6.1	04-May-1999	perry	branches: 1.33.6.1.2; pullup 1.33->1.34 (sommerfeld)
1.33.6.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.33.6.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.36.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.36.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.40.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.41.2.3	09-Feb-2002	he	Pull up revision 1.50 (requested by christos): Widen cr_ref to prevent overflow.
1.41.2.2	14-Dec-2000	he	Pull up revisions 1.44-1.45 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.41.2.1	24-Oct-2000	tv	Pullup 1.45 [chs]: fix nfs iod management so we don't lose i/os when iods die.
1.48.8.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.48.4.3	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
1.48.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.48.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.48.2.9	17-Sep-2002	nathanw	Catch up to -current.
1.48.2.8	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.48.2.7	24-Jun-2002	nathanw	Curproc->curlwp renaming. Change uses of "curproc->l_proc" back to "curproc", which is more like the original use. Bare uses of "curproc" are now "curlwp". "curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL) so that it is always safe to reference curproc (dereferencing curproc is another story, but that's always been true).
1.48.2.6	20-Jun-2002	nathanw	Catch up to -current.
1.48.2.5	29-May-2002	nathanw	#include <sys/sa.h> before <sys/syscallargs.h>, to provide sa_upcall_t now that <sys/param.h> doesn't include <sys/sa.h>. (Behold the Power of Ed)
1.48.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.48.2.3	27-Nov-2001	thorpej	Make lockmgr() lwp-aware: - Locks are counted against LWPs, not procs. - When we record the lockholder in the lock structure, we need to also record the lwpid. - When we are checking who holds the lock, also consider lwpid. Fixes a "locking against myself" panic reported by Allen Briggs that could be easily triggered by redirecting the output of an LWP-using program to a file.
1.48.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.48.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.69.2.7	11-Dec-2005	christos	Sync with head.
1.69.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.69.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.69.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.69.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.69.2.2	03-Aug-2004	skrll	Sync with HEAD
1.69.2.1	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.
1.74.2.1	14-Jun-2004	jmc	Pullup rev 1.76 (requested by yamt in ticket #466) nfsrv_zapsock: fix an inverted condition in nfs_syscall.c rev.1.74.
1.77.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.77.4.1	29-Apr-2005	kent	sync with -current
1.78.4.11	24-Mar-2008	yamt	sync with head.
1.78.4.10	17-Mar-2008	yamt	sync with head.
1.78.4.9	27-Feb-2008	yamt	revert incomplete nfs client locking for now.
1.78.4.8	15-Feb-2008	yamt	- sprinkle some locks. - disable MNT_UPDATE because it involves too much locking headache. - don't overwrite other bits in v_vflags when setting VV_ROOT.
1.78.4.7	21-Jan-2008	yamt	sync with head
1.78.4.6	07-Dec-2007	yamt	sync with head
1.78.4.5	27-Oct-2007	yamt	sync with head.
1.78.4.4	03-Sep-2007	yamt	sync with head.
1.78.4.3	26-Feb-2007	yamt	sync with head.
1.78.4.2	30-Dec-2006	yamt	sync with head.
1.78.4.1	21-Jun-2006	yamt	sync with head.
1.82.6.1	29-Nov-2005	yamt	sync with head.
1.84.2.1	15-Jan-2006	yamt	sync with head.
1.88.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.88.8.9	11-May-2006	elad	sync with head
1.88.8.8	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.
1.88.8.7	19-Apr-2006	elad	sync with head.
1.88.8.6	14-Apr-2006	elad	Plug some possible leaks of kauth_cred_t.
1.88.8.5	12-Mar-2006	elad	Rename kauth_cred_compare() to kauth_cred_uucmp(), and kauth_cred_convert() to kauth_cred_uucvt(). This makes it clearer that we're working on struct uucred. Inspired by comments from yamt@.
1.88.8.4	10-Mar-2006	elad	Cleanup more interface abuse. Make nfsrv_setcred() take a kauth_cred_t * as outcred. The original code just modified it directly; we can't do that, nor do we want to. Get rid of another case of kauth_cred_zero() followed by kauth_cred_hold() and use kauth_cred_clone() to make sure we don't leave out important members. Add another DIAGNOSTIC check for reference count of above one. Again, this should be tested.
1.88.8.3	10-Mar-2006	elad	Some cleanup. kauth_cred_setrefcnt() was only called after kauth_cred_convert() in NFS code to convert a struct uucred to kauth_cred_t. Since there's no valid use for such a function, make kauth_cred_convert() set the reference count to 1 and eliminate the need for kauth_cred_setrefcnt() entirely. Motivated by comments from yamt@ and thorpej@.
1.88.8.2	10-Mar-2006	elad	generic_authorize() -> kauth_authorize_generic().
1.88.8.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.88.6.3	11-Aug-2006	yamt	sync with head
1.88.6.2	26-Jun-2006	yamt	sync with head.
1.88.6.1	24-May-2006	yamt	sync with head.
1.88.4.3	01-Jun-2006	kardel	Sync with head.
1.88.4.2	22-Apr-2006	simonb	Sync with head.
1.88.4.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.88.2.1	09-Sep-2006	rpaulo	sync with head
1.94.2.2	19-Jun-2006	chap	Sync with head.
1.94.2.1	19-May-2006	chap	file nfs_syscalls.c was added on branch chap-midi on 2006-06-19 04:10:37 +0000
1.96.6.2	10-Dec-2006	yamt	sync with head.
1.96.6.1	22-Oct-2006	yamt	sync with head
1.96.4.4	01-Feb-2007	ad	Sync with head.
1.96.4.3	30-Jan-2007	ad	Remove support for SA. Ok core@.
1.96.4.2	12-Jan-2007	ad	Sync with head.
1.96.4.1	18-Nov-2006	ad	Sync with head.
1.99.4.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.99.2.1	05-Jun-2007	bouyer	Pull up following revision(s) (requested by yamt in ticket #705): sys/nfs/nfs_syscalls.c: revision 1.104 plug a mbuf leak.
1.105.2.3	07-May-2007	yamt	sync with head.
1.105.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.105.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.107.4.1	11-Jul-2007	mjf	Sync with head.
1.107.2.13	27-Oct-2007	yamt	call soshutdown with kernel_lock held.
1.107.2.12	26-Aug-2007	yamt	- mark nfssvc(2) MPSAFE and move the most of nfsd out of the kernel lock. - remove unused ns_solock. - remove some of KERNEL_LOCK/UNLOCK which are not necessary on this branch.
1.107.2.11	20-Aug-2007	ad	Sync with HEAD.
1.107.2.10	15-Jul-2007	ad	Sync with head.
1.107.2.9	18-Jun-2007	yamt	fix a merge botch.
1.107.2.8	09-Jun-2007	ad	Sync with head.
1.107.2.7	08-Jun-2007	ad	Sync with head.
1.107.2.6	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.
1.107.2.5	10-Apr-2007	ad	Nuke the deferred kthread creation stuff, as it's no longer needed. Pointed out by thorpej@.
1.107.2.4	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.
1.107.2.3	05-Apr-2007	ad	Compile fixes.
1.107.2.2	21-Mar-2007	ad	- Put a lock around the proc's CWD info (work in progress). - Replace some more simplelocks. - Make lbolt a condvar.
1.107.2.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.117.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.122.4.2	02-Aug-2007	yamt	nfsrv_slpderef: add an assertion.
1.122.4.1	02-Aug-2007	yamt	file nfs_syscalls.c was added on branch matt-mips64 on 2007-08-02 12:46:04 +0000
1.122.2.5	09-Dec-2007	jmcneill	Sync with HEAD.
1.122.2.4	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.122.2.3	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.
1.122.2.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.122.2.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.124.4.1	14-Oct-2007	yamt	sync with head.
1.124.2.3	23-Mar-2008	matt	sync with HEAD
1.124.2.2	09-Jan-2008	matt	sync with HEAD
1.124.2.1	06-Nov-2007	matt	sync with HEAD
1.125.4.3	18-Feb-2008	mjf	Sync with HEAD.
1.125.4.2	27-Dec-2007	mjf	Sync with HEAD.
1.125.4.1	08-Dec-2007	mjf	Sync with HEAD.
1.125.2.1	22-Nov-2007	bouyer	Sync with HEAD
1.126.2.4	29-Dec-2007	yamt	to prepare merge, put nfsd back under kernel_lock for now.
1.126.2.3	26-Dec-2007	ad	Sync with head.
1.126.2.2	08-Dec-2007	ad	Sync with head.
1.126.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.127.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.130.6.6	17-Jan-2009	mjf	Sync with HEAD.
1.130.6.5	05-Oct-2008	mjf	Sync with HEAD.
1.130.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.130.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.130.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.130.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.130.2.1	24-Mar-2008	keiichi	sync with head.
1.133.2.2	04-Jun-2008	yamt	sync with head
1.133.2.1	18-May-2008	yamt	sync with head.
1.134.2.7	11-Mar-2010	yamt	sync with head
1.134.2.6	16-Sep-2009	yamt	sync with head
1.134.2.5	18-Jul-2009	yamt	sync with head.
1.134.2.4	20-Jun-2009	yamt	sync with head
1.134.2.3	04-May-2009	yamt	sync with head.
1.134.2.2	16-May-2008	yamt	sync with head.
1.134.2.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.135.2.5	10-Oct-2008	skrll	Sync with HEAD.
1.135.2.4	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.135.2.3	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.135.2.2	14-May-2008	wrstuden	Per discussion with ad, remove most of the #include <sys/sa.h> lines as they were including sa.h just for the type(s) needed for syscallargs.h. Instead, create a new file, sys/satypes.h, which contains just the types needed for syscallargs.h. Yes, there's only one now, but that may change and it's probably more likely to change if it'd be difficult to handle. :-) Per discussion with matt at n dot o, add an include of satypes.h to sigtypes.h. Upcall handlers are kinda signal handlers, and signalling is the header file that's already included for syscallargs.h that closest matches SA. This shaves about 3000 lines off of the diff of the branch relative to the base. That also represents about 18% of the total before this checkin. I think this reduction is very good thing.
1.135.2.1	10-May-2008	wrstuden	Initial checkin of re-adding SA. Everything except kern_sa.c compiles in GENERIC for i386. This is still a work-in-progress, but this checkin covers most of the mechanical work (changing signalling to be able to accomidate SA's process-wide signalling and re-adding includes of sys/sa.h and savar.h). Subsequent changes will be much more interesting. Also, kern_sa.c has received partial cleanup. There's still more to do, though.
1.136.2.1	27-Jun-2008	simonb	Sync with head.
1.137.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.137.2.1	19-Oct-2008	haad	Sync with HEAD.
1.140.4.1	13-Apr-2009	snj	Pull up following revision(s) (requested by ad in ticket #701): sys/nfs/nfs_syscalls.c: revision 1.147 PR kern/41154: possible races in NFS server code Fix some of the races (but probably not all of them) in the NFS server code. nfssvc_nfsd(): change a splsoftclock()/spx() to mutex_enter/exit(&nfsd_lock) (I guess it was forgotten when the nfsd code was made SMP safe) m_freem(nd_nam) in nfsrv_slpderef() instead of nfsrv_zapsock() to avoid possible use after free in nfssvc_nfsd() Fix nfsrv_slpderef() to not release nfsd_lock before testing SLP_VALID and reaquiring it just after. This could cause a use after free of the slp if one thread is in nfsrv_slpderef() and the other one grabs slp from nfssvc_sockpending and zap it.
1.140.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.140.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.143.4.2	23-Jul-2009	jym	Sync with HEAD.
1.143.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.153.26.1	18-May-2014	rmind	sync with head
1.153.22.2	03-Dec-2017	jdolecek	update from HEAD
1.153.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.153.12.1	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")
1.154.4.1	04-Nov-2015	riz	Pull up following revision(s) (requested by mrg in ticket #956): sys/compat/netbsd32/files.netbsd32: revision 1.36 sys/compat/netbsd32/netbsd32_sysent.c: revision 1.115 sys/compat/netbsd32/netbsd32_syscallargs.h: revision 1.116 sys/nfs/nfs_var.h: revision 1.93 sys/compat/netbsd32/netbsd32_conv.h: revision 1.30 sys/compat/netbsd32/netbsd32_syscall.h: revision 1.116 sys/compat/netbsd32/netbsd32_syscalls.c: revision 1.115 sys/compat/netbsd32/netbsd32_nfssvc.c: revision 1.1 sys/compat/netbsd32/netbsd32_nfssvc.c: revision 1.3 sys/nfs/nfs_syscalls.c: revision 1.156 sys/compat/netbsd32/syscalls.master: revision 1.108 sys/compat/netbsd32/netbsd32.h: revision 1.107 add netbsd32 support for nfssvc(2). we do this by defining 5 copyin/out functions that do all the ugly work, are just plain copyin/out for the native system calls, and do the necessary translations for netbsd32. with this i'm able to run 32 bit nfsd and mountd on 64 bit kernel and mount the file systems remotely. don't copy the first netbsd32_export_args nexports times, but actually advance the userland pointer each entry through the loop. oops.
1.155.2.3	28-Aug-2017	skrll	Sync with HEAD
1.155.2.2	09-Jul-2016	skrll	Sync with HEAD
1.155.2.1	22-Sep-2015	skrll	Sync with HEAD
1.157.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.157.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.159.2.1	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.160.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.160.2.1	10-Jun-2019	christos	Sync with HEAD
1.162.12.1	06-Jun-2021	cjep	sync with head
1.162.8.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.163.16.1	02-Aug-2025	perseant	Sync with HEAD
1.97	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.96	27-Apr-2022	hannken	branches: 1.96.10; As VOP_GETATTR() needs a shared lock at least move the preopattr lookup inside nfs_namei() where we may lock the start directory without violating the lock order.
1.95	04-Jun-2021	hannken	Add flag/command NFSSVC_REPLACEEXPORTSLIST to nfssvc(2) system call. Works like NFSSVC_SETEXPORTSLIST but supports "mel_nexports > 1" and will atomically update the complete exports list for a file system.
1.94	15-Jul-2015	manu	branches: 1.94.34; 1.94.38; Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.93	22-Jun-2015	mrg	add netbsd32 support for nfssvc(2). we do this by defining 5 copyin/out functions that do all the ugly work, are just plain copyin/out for the native system calls, and do the necessary translations for netbsd32. with this i'm able to run 32 bit nfsd and mountd on 64 bit kernel and mount the file systems remotely.
1.92	30-May-2014	hannken	branches: 1.92.2; 1.92.4; Change NFS from rbtree to vcache.
1.91	14-Dec-2013	christos	branches: 1.91.2; don't allow the nfs server module to unload if it has exported filesystems.
1.90	02-Mar-2010	pooka	branches: 1.90.10; 1.90.20; 1.90.24; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.89	03-Sep-2009	tls	branches: 1.89.2; ...and one more missed in the earlier commit (sigh). Kernels should build again now.
1.88	07-Jul-2009	christos	The compatibility call to re-export from sys_mount() calls mountd_set_exports_list, with the mnt_updating mutex held. Account for that to avoid a locking against myself panic.
1.87	23-May-2009	ad	- Fix a race between umount()/mount() and nfssvc(). - Toss netexport state on nfsserver module unload.
1.86	14-Mar-2009	dsl	Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.85	28-Nov-2008	pooka	branches: 1.85.4; g/c unused malloc types
1.84	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.83	14-Nov-2008	ad	Remove COMPAT ifdefs that might as well be comments (i.e., they cost us almost nothing).
1.82	22-Oct-2008	matt	branches: 1.82.2; Don't need nfs_vfs_reinit anymore since we don't resize tables anymore. Move reinit code to init case.
1.81	22-Oct-2008	matt	Change NFS to use a RB-tree for its FH->nfsnode lookups.
1.80	30-Sep-2008	pooka	Initialize nfsnode pools and malloc type dynamically in the constructor instead of depending on link sets. Consequently, rename nfs_nh{init,reinit,done} to nfs_node_{init,reinit,done}, respectively, to better convey the function.
1.79	28-Apr-2008	martin	branches: 1.79.2; 1.79.6; Remove clause 3 and 4 from TNF licenses
1.78	10-Apr-2008	yamt	branches: 1.78.2; 1.78.4; - make nfs_receive and nfs_reply static. - ansify.
1.77	02-Jan-2008	yamt	branches: 1.77.6; use kmem_alloc instead of malloc.
1.76	20-Dec-2007	dsl	Convert all the system call entry points from: int foo(struct lwp l, void v, register_t retval) to: int foo(struct lwp l, const struct foo_args uap, register_t retval) Fixup compat code to not write into 'uap' and (in some cases) to actually pass a correctly formatted 'uap' structure with the right name to the next routine. A few 'compat' routines that just call standard ones have been deleted. All the 'compat' code compiles (along with the kernels required to test build it). 98% done by automated scripts.
1.75	04-Dec-2007	yamt	branches: 1.75.4; merge non-intrusive nfs changes from vmlocking.
1.74	28-Oct-2007	yamt	branches: 1.74.2; 1.74.4; make NFS_ATTRTIMEO a function.
1.73	21-Oct-2007	yamt	remove lwp argument from nfs_reconnect and always use &lwp0 because who triggers a reconnect doesn't really matter here. PR/37145.
1.72	10-Aug-2007	yamt	branches: 1.72.2; 1.72.6; - instead of scanning an array of iods, maintain a list of idle iods. - make nfs_getset_niothreads MP friendly.
1.71	27-Jul-2007	yamt	branches: 1.71.4; 1.71.6; stop nfs tick when we have nothing to do.
1.70	20-Jul-2007	yamt	- fix decreasing of vfs.nfs.iothreads after the recent partial merge of vmlocking. - don't make nfsiod exit with requests left. - make NFSSVC_BIOD a dummy so that nfsiod can be simplified.
1.69	12-Jul-2007	dsl	branches: 1.69.2; Change the VFS_MOUNT() interface so that the 'data' buffer passed to the fs code is a kernel buffer, pass though the length of the buffer as well. Since the length of the userspace buffer isn'it (yet) passed through the mount system call, add a field to the vfsops structure containing the default length. Split sys_mount() for calls from compat code. Ride one of the recent kernel version changes - old fs LKMs will load, but sys_mount() will reject any attempt to use them.
1.68	28-May-2007	yamt	- remove nfs_exit exit hook. ok'ed by christos@. - as far as i understand the code, it shouldn't be necessary because nfs_request can't return without removing its request and r->r_lwp is either curlwp or NULL. - even if it's necessary, leaking requests is not the correct way to recover from the condition. - nfs_request: add a related assertion.
1.67	29-Apr-2007	yamt	use mutex and condver.
1.66	04-Mar-2007	christos	branches: 1.66.2; 1.66.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.65	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.64	27-Dec-2006	yamt	branches: 1.64.2; remove nqnfs.
1.63	02-Sep-2006	yamt	branches: 1.63.2; nfsd: deal with variable-sized filehandles.
1.62	01-Jul-2006	yamt	if a file is sillyrename'ed because it's a destination of rename, make sillyrename (try to) use LINK operation rather than RENAME. PR/33861 from Jed Davis. he provided the almost same patch. according to him, it also happen to be what opensolaris does in this case. from the PR: > In nfs_rename(), if the destination appears to exist and is "in use" > (this check is apparently satisfied even if the file isn't in use by > anything except the rename itself), it will sillyrename it, then delete > the sillyrenamed file even if the rename fails -- for instance, because > the "from" file no longer exists on the server. > mkdir a b; touch a/x; perl -e 'fork(); rename("a/x","b/x") or die "$!\n"' > > Afterwards, neither a/x nor b/x will exist. > 1) Lookup of b/x; fails with NOENT. > 2) Rename from a/x to b/x; succeeds. > 3) Lookup of b/x; fails with NOENT. > 4) Rename from b/x to b/.nfsA23a3; succeeds. > 5) Rename from a/x to b/x; fails with NOENT. > 6) Remove of b/.nfsA23a3; succeeds.
1.61	19-May-2006	yamt	branches: 1.61.2; 1.61.4; - fix compilation problem for !NFSSERVER && NFS. pointed by Tom Spindler on source-changes@. - make nfs_srvdesc_pool static.
1.60	18-May-2006	yamt	- fix some leaks in nfsd, introduced by kauth changes. - simplify code. - add some assertions. - wrap some long lines. - remove an unnecessary ";".
1.59	14-May-2006	elad	integrate kauth.
1.58	05-Jan-2006	yamt	branches: 1.58.2; 1.58.4; 1.58.6; 1.58.8; 1.58.10; ensure the export list is not changed during nfsd operations.
1.57	03-Jan-2006	yamt	de-__P.
1.56	03-Jan-2006	yamt	move function prototypes from nfs.h to nfs_var.h.
1.55	11-Dec-2005	christos	branches: 1.55.2; merge ktrace-lwp.
1.54	22-Nov-2005	yamt	- reduce number of linear search per rpc. - coalesce mount_netexport_pair into netexport.
1.53	25-Sep-2005	jmmv	branches: 1.53.6; Add some COMPAT_30 code to let old mountd binaries work after the NFS exports rototill.
1.52	23-Sep-2005	jmmv	Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.51	07-Jul-2005	christos	1. use p = uio->uio_procp consistently and eliminate suspicious uses of curproc (where uio->uio_procp should be used?). Don't do this for nfs_commit(), because yamt says it is possibly wrong. 2. nfs_doio() does not use struct proc; remove it and the code to compute it. 3. use copyin_proc() and copyout_proc() instead of copyin() and copyout(). 4. check return of copyout_proc(). and mark return from copyin_proc() XXX 5. Eliminate check p == curproc assertion check from nfs_write; nfs_read does not have it and we might be called in a different process context anyway (PR 20138).
1.50	29-May-2005	christos	branches: 1.50.2; - sprinkle const - avoid shadowed variables - mark bad const use with XXXUNCONST
1.49	27-Jan-2005	yamt	branches: 1.49.4; 1.49.6; keep directory eof cache when inactivating vnode because there's no reason to throw it away. (fix an unintended side effect of nfs_subs.c rev.1.144.)
1.48	19-Jan-2005	yamt	branches: 1.48.2; implement inaccurate mtime/ctime detection. namely, if mtime or ctime are same between pre_op_attr and post_op_attr when we expected them to be changed, don't trust the server.
1.47	14-Dec-2004	yamt	branches: 1.47.2; - centerize code to invalidate stale cache. - don't ignore errors when invalidating buffers in nfs_open.
1.46	15-Sep-2004	yamt	fix access-after-free bugs in dircache code by refcounting nfsdircache. PR/26864.
1.45	22-May-2004	jonathan	Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.44	10-May-2004	yamt	don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.43	05-Apr-2004	yamt	nfs_readdirplusrpc: fix a deadlock problem. don't wait for vnode lock to load attributes. otherwise, because READDIRPLUS returns DOTDOT entry as well, we violate locking order.
1.42	23-Jul-2003	yamt	branches: 1.42.2; when rexmitting a request due to NFSERR_JUKEBOX, use a new xid as RFC1813 says.
1.41	29-Jun-2003	fvdl	branches: 1.41.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.40	29-Jun-2003	thorpej	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.39	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
1.38	09-Jun-2003	yamt	rework zero padding of rpc reply. - for READ procedure, don't send back more bytes than requested. - don't have doubtful assumptions on mbuf chain structure. - rename a function (nfsm_adj -> nfs_zeropad) to avoid confusion as the semantics of the function was changed.
1.37	22-May-2003	yamt	poolify nfsrv_descript.
1.36	22-May-2003	yamt	avoid double free with xlatecookie.
1.35	22-May-2003	yamt	interlock for nfs_rcvlock.
1.34	21-May-2003	yamt	eliminate memcpy in the common and easy case of write.
1.33	07-May-2003	yamt	simple lock for nfs iod.
1.32	05-May-2003	yamt	keep things not needed by userland in #ifdef _KERNEL. (e.g. prototypes for in-kernel functions)
1.31	03-May-2003	yamt	better handling of write verifier change.
1.30	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.29	28-Mar-2003	yamt	i forgot to check this in with the previous (reply ENAMETOOLONG properly).
1.28	02-Feb-2003	christos	protect <sys/mallocvar.h> ifdef _KERNEL
1.27	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.26	18-Jan-2003	thorpej	Merge the nathanw_sa branch.
1.25	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
1.24	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.23	17-Mar-2002	christos	use the exithook mechanism to remove the exiting process from the list of processes to be signalled in a soft mount.
1.22	11-Mar-2002	jdolecek	nfs_enterdircache() had last two parameter types swapped Noted in kern/14742 by John Franklin.
1.21	05-Dec-2001	lukem	don't need nfs_hash prototype here
1.20	15-Sep-2001	chs	add a new VFS op, vfs_reinit, which is called when desiredvnodes is adjusted via sysctl. file systems that have hash tables which are sized based on the value of this variable now resize those hash tables using the new value. the max number of FFS softdeps is also recalculated. convert various file systems to use the <sys/queue.h> macros for their hash tables.
1.19	27-Nov-2000	chs	branches: 1.19.2; 1.19.4; 1.19.6; Initial integration of the Unified Buffer Cache project.
1.18	19-Sep-2000	fvdl	Add prototypes for commitrange functions.
1.17	15-Apr-2000	tsarna	branches: 1.17.4; Death to nfsiod! It is replaced by kernel threads that do the same thing. The number of kernel threads used is set with the vfs.nfs.iothreads sysctl.
1.16	16-Mar-2000	jdolecek	Add new VFS op routine - vfs_done and call it on filesystem detach in vfs_detach(). vfs_done may free global filesystem's resources, typically those allocated in respective filesystem's init function. Needed so those filesystems which went in via LKM have a chance to clean after themselves before unloading. This fixes random panics when LKM for filesystem using pools was loaded and unloaded several times. For each leaf filesystem, add appropriate vfs_done routine.
1.15	05-Sep-1998	christos	branches: 1.15.12; Assign copyright to TNF.
1.14	25-Jun-1998	thorpej	- Rename nqnfs_vop_lease_check() to genfs_lease_check(). If NFSSERVER is not in the kernel, genfs_lease_check() is simply a no-op. This allows LKM'd file systems to be exported (previously did not work properly due to a compile-time decision based on -DNFSSERVER). - defopt NFSSERVER
1.13	29-Mar-1998	mrg	add forward decl for union nethostaddr.
1.12	30-Jan-1998	fvdl	Only take the receive lock before disconnecting when doing it from nfs_decode_args. Otherwise we might just end up locking against ourselves. XXX workaround, will do ok for now. Proper fix forthcoming.
1.11	19-Oct-1997	fvdl	branches: 1.11.2; * Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.10	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.9	29-Aug-1997	gwr	The nfs_boot_xxx functions are declared in nfsdiskless.h so remove the duplicate declarations here.
1.8	14-Jul-1997	fvdl	branches: 1.8.2; Don't assume that pointers into mbuf data remain valid across nfsm_dissect. In readdirplus, don't keep such pointers but store the file attributes in a variable instead until they are needed. Change nfsm_loadattr* a bit so it can accept a direct pointer to an nfs_fattr structure.
1.7	24-Jun-1997	fvdl	Add prototype for nfs_ispublicfh, change the ones for nfs_namei and nfsrv_fhtovp.
1.6	11-Dec-1996	fvdl	Give permission to the owner of the file to preserve semantics only in the relevant cases (read, write). Fixes PR 3017.
1.5	25-Oct-1996	cgd	make the namei struct members ni_dirp and ni_next, and the componentname struct member cn_nameptr 'const', since they should never be used to modify the path name. (Only the pathname buffer, cn_pnbuf, should be modified.) Propagate the const poisoning to code that uses the namei and componentname structs.
1.4	01-Sep-1996	mycroft	Add a set of generic file system operations that most file systems use. Also, fix some time stamp bogosities.
1.3	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.2	13-Feb-1996	christos	add 2 missing fwd struct declarations
1.1	09-Feb-1996	christos	nfs prototype changes
1.8.2.2	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.8.2.1	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.11.2.1	07-Feb-1998	mellon	Pull up 1.12
1.15.12.2	08-Dec-2000	bouyer	Sync with HEAD.
1.15.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.17.4.1	14-Dec-2000	he	Pull up revision 1.18 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.19.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.19.4.4	30-Sep-2002	jdolecek	add support for kevents to NFS to detect file changes on server by other NFS clients, polling kernel thread is used to periodically check for attribute changes of watched files; the NFS server is only contacted when the vnode expires from local attrcache (which takes 5-60 seconds currently), to keep network&CPU overhead low the routine checking for remote changes is quite simplistic, but hopefully doing it's job well enough
1.19.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.19.4.2	16-Mar-2002	jdolecek	Catch up with -current.
1.19.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.19.2.6	11-Nov-2002	nathanw	Catch up to -current
1.19.2.5	22-Oct-2002	thorpej	Sync with HEAD.
1.19.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.19.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.19.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.19.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.41.2.11	11-Dec-2005	christos	Sync with head.
1.41.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.41.2.9	04-Feb-2005	skrll	Sync with HEAD.
1.41.2.8	24-Jan-2005	skrll	Sync with HEAD.
1.41.2.7	18-Dec-2004	skrll	Sync with HEAD.
1.41.2.6	21-Sep-2004	skrll	Fix the sync with head I botched.
1.41.2.5	18-Sep-2004	skrll	Sync with HEAD.
1.41.2.4	24-Aug-2004	skrll	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.41.2.3	18-Aug-2004	skrll	Revert to passing struct proc for {exit,exec}hook.
1.41.2.2	03-Aug-2004	skrll	Sync with HEAD
1.41.2.1	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.
1.42.2.3	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.42.2.2	18-Sep-2004	he	branches: 1.42.2.2.2; Pull up revision 1.46 (requested by yamt in ticket #858): Fix access-after-free bugs in dircache code by reference counting nfsdircache. Fixes PR#26864.
1.42.2.1	10-Jul-2004	tron	Pull up revision 1.43 (requested by tls in ticket #634): nfs_readdirplusrpc: fix a deadlock problem. don't wait for vnode lock to load attributes. otherwise, because READDIRPLUS returns DOTDOT entry as well, we violate locking order.
1.42.2.2.2.1	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.47.2.1	29-Apr-2005	kent	sync with -current
1.48.2.1	12-Feb-2005	yamt	sync with head.
1.49.6.1	16-Jul-2006	ghen	Pull up following revision(s) (requested by jld in ticket #1424): sys/nfs/nfs_vnops.c: revision 1.240 via patch sys/nfs/nfs_var.h: revision 1.62 via patch Fix race condition in NFS renaming that could cause the renamed file to be deleted (PR/33861).
1.49.4.1	16-Jul-2006	ghen	Pull up following revision(s) (requested by jld in ticket #1424): sys/nfs/nfs_vnops.c: revision 1.240 via patch sys/nfs/nfs_var.h: revision 1.62 via patch Fix race condition in NFS renaming that could cause the renamed file to be deleted (PR/33861).
1.50.2.8	21-Jan-2008	yamt	sync with head
1.50.2.7	07-Dec-2007	yamt	sync with head
1.50.2.6	15-Nov-2007	yamt	sync with head.
1.50.2.5	27-Oct-2007	yamt	sync with head.
1.50.2.4	03-Sep-2007	yamt	sync with head.
1.50.2.3	26-Feb-2007	yamt	sync with head.
1.50.2.2	30-Dec-2006	yamt	sync with head.
1.50.2.1	21-Jun-2006	yamt	sync with head.
1.53.6.4	22-Nov-2005	yamt	sync with head.
1.53.6.3	22-Nov-2005	yamt	remove uvm_ractx forward decl. which is no longer used.
1.53.6.2	18-Nov-2005	yamt	- associate read-ahead context to vnode, rather than file. - revert VOP_READ prototype.
1.53.6.1	15-Nov-2005	yamt	adapt ffs, lfs, nfs.
1.55.2.1	15-Jan-2006	yamt	sync with head.
1.58.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.58.8.3	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.
1.58.8.2	10-Mar-2006	elad	Cleanup more interface abuse. Make nfsrv_setcred() take a kauth_cred_t * as outcred. The original code just modified it directly; we can't do that, nor do we want to. Get rid of another case of kauth_cred_zero() followed by kauth_cred_hold() and use kauth_cred_clone() to make sure we don't leave out important members. Add another DIAGNOSTIC check for reference count of above one. Again, this should be tested.
1.58.8.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.58.6.3	03-Sep-2006	yamt	sync with head.
1.58.6.2	11-Aug-2006	yamt	sync with head
1.58.6.1	24-May-2006	yamt	sync with head.
1.58.4.1	01-Jun-2006	kardel	Sync with head.
1.58.2.1	09-Sep-2006	rpaulo	sync with head
1.61.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.61.2.2	19-May-2006	yamt	- fix compilation problem for !NFSSERVER && NFS. pointed by Tom Spindler on source-changes@. - make nfs_srvdesc_pool static.
1.61.2.1	19-May-2006	yamt	file nfs_var.h was added on branch chap-midi on 2006-05-19 13:53:12 +0000
1.63.2.1	12-Jan-2007	ad	Sync with head.
1.64.2.3	07-May-2007	yamt	sync with head.
1.64.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.64.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.66.4.1	11-Jul-2007	mjf	Sync with head.
1.66.2.5	26-Aug-2007	yamt	- mark nfssvc(2) MPSAFE and move the most of nfsd out of the kernel lock. - remove unused ns_solock. - remove some of KERNEL_LOCK/UNLOCK which are not necessary on this branch.
1.66.2.4	20-Aug-2007	ad	Sync with HEAD.
1.66.2.3	15-Jul-2007	ad	Sync with head.
1.66.2.2	09-Jun-2007	ad	Sync with head.
1.66.2.1	08-Jun-2007	ad	Sync with head.
1.69.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.71.6.2	27-Jul-2007	yamt	stop nfs tick when we have nothing to do.
1.71.6.1	27-Jul-2007	yamt	file nfs_var.h was added on branch matt-mips64 on 2007-07-27 10:03:59 +0000
1.71.4.4	09-Dec-2007	jmcneill	Sync with HEAD.
1.71.4.3	29-Oct-2007	joerg	Sync with HEAD.
1.71.4.2	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.
1.71.4.1	16-Aug-2007	jmcneill	Sync with HEAD.
1.72.6.2	13-Nov-2007	bouyer	Sync with HEAD
1.72.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.72.2.2	09-Jan-2008	matt	sync with HEAD
1.72.2.1	06-Nov-2007	matt	sync with HEAD
1.74.4.4	26-Dec-2007	ad	Sync with head.
1.74.4.3	08-Dec-2007	ad	Sync with head.
1.74.4.2	04-Dec-2007	yamt	apply the following change, which seems to get lost during vmlocking -> vmlocking2 transition. Module Name: src Committed By: yamt Date: Sun Oct 21 08:23:20 UTC 2007 Modified Files: src/sys/nfs: nfs_socket.c nfs_var.h Log Message: remove lwp argument from nfs_reconnect and always use &lwp0 because who triggers a reconnect doesn't really matter here. PR/37145. To generate a diff of this commit: cvs rdiff -r1.163 -r1.164 src/sys/nfs/nfs_socket.c cvs rdiff -r1.72 -r1.73 src/sys/nfs/nfs_var.h Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files.
1.74.4.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.74.2.3	18-Feb-2008	mjf	Sync with HEAD.
1.74.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.74.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.75.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.77.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.77.6.2	05-Oct-2008	mjf	Sync with HEAD.
1.77.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.78.4.8	11-Mar-2010	yamt	sync with head
1.78.4.7	16-Sep-2009	yamt	sync with head
1.78.4.6	18-Jul-2009	yamt	sync with head.
1.78.4.5	20-Jun-2009	yamt	sync with head
1.78.4.4	04-May-2009	yamt	fix merge botches.
1.78.4.3	04-May-2009	yamt	sync with head.
1.78.4.2	16-May-2008	yamt	sync with head.
1.78.4.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.78.2.1	18-May-2008	yamt	sync with head.
1.79.6.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.79.6.1	19-Oct-2008	haad	Sync with HEAD.
1.79.2.1	10-Oct-2008	skrll	Sync with HEAD.
1.82.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.82.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.85.4.2	23-Jul-2009	jym	Sync with HEAD.
1.85.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.89.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.90.24.1	18-May-2014	rmind	sync with head
1.90.20.2	03-Dec-2017	jdolecek	update from HEAD
1.90.20.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.90.10.1	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")
1.91.2.1	10-Aug-2014	tls	Rebase.
1.92.4.1	22-Sep-2015	skrll	Sync with HEAD
1.92.2.2	04-Nov-2015	riz	Pull up following revision(s) (requested by mrg in ticket #956): sys/compat/netbsd32/files.netbsd32: revision 1.36 sys/compat/netbsd32/netbsd32_sysent.c: revision 1.115 sys/compat/netbsd32/netbsd32_syscallargs.h: revision 1.116 sys/nfs/nfs_var.h: revision 1.93 sys/compat/netbsd32/netbsd32_conv.h: revision 1.30 sys/compat/netbsd32/netbsd32_syscall.h: revision 1.116 sys/compat/netbsd32/netbsd32_syscalls.c: revision 1.115 sys/compat/netbsd32/netbsd32_nfssvc.c: revision 1.1 sys/compat/netbsd32/netbsd32_nfssvc.c: revision 1.3 sys/nfs/nfs_syscalls.c: revision 1.156 sys/compat/netbsd32/syscalls.master: revision 1.108 sys/compat/netbsd32/netbsd32.h: revision 1.107 add netbsd32 support for nfssvc(2). we do this by defining 5 copyin/out functions that do all the ugly work, are just plain copyin/out for the native system calls, and do the necessary translations for netbsd32. with this i'm able to run 32 bit nfsd and mountd on 64 bit kernel and mount the file systems remotely. don't copy the first netbsd32_export_args nexports times, but actually advance the userland pointer each entry through the loop. oops.
1.92.2.1	04-Nov-2015	riz	Pull up following revision(s) (requested by manu in ticket #882): sbin/umount/umount.c: revision 1.48 sys/nfs/nfsmount.h: revision 1.53 sys/nfs/nfs_var.h: revision 1.94 sys/nfs/nfs_iod.c: revision 1.7 sys/nfs/nfs_socket.c: revision 1.197 sys/nfs/nfs_bio.c: revision 1.191 sys/nfs/nfs_vfsops.c: revision 1.230 sys/nfs/nfs_clntsocket.c: revision 1.3 Remove useless and harmful sync(2) call in umount(8) Remove sync(2) call before unmount(2) in umount(8). This sync(2) is useless since unmount(2) will perform a VFS_SYNC anyway. But moreover, this sync(2) may be harmful, as there are some situation where it cannot return (unreachable NFS server, for instance), causing umount -f to be uneffective. Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.94.38.1	06-Jun-2021	cjep	sync with head
1.94.34.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.96.10.1	02-Aug-2025	perseant	Sync with HEAD
1.246	13-May-2024	msaitoh	ficticious -> fictitious in comment.
1.245	21-Mar-2023	christos	PR/57279: Izumi Tsutsui: Fix some {int,long} -> time_t. Still things will break eventually because parts of the nfs protocol assume time_t will fit in 32 bits.
1.244	17-Mar-2023	mlelstv	Avoid overflow of nfs_commitsize on machines with > 32GB RAM.
1.243	13-Jun-2021	mlelstv	branches: 1.243.10; Don't pretend that files are limited to 1TB on NFSv3.
1.242	02-Apr-2021	christos	branches: 1.242.2; Set f_namemax during mount time like all the other filesystems so that it does gets the right data in copy_statvfs_info(). Otherwise f_namemax can end up being 0. To reproduce: unmount the remote filesystem, remount it, and kill -HUP mountd to refresh exports.
1.241	13-Apr-2020	ad	branches: 1.241.2; 1.241.4; Replace most uses of vp->v_usecount with a call to vrefcnt(vp), a function that hides the details and does atomic_load_relaxed(). Signature matches FreeBSD.
1.240	16-Mar-2020	pgoyette	branches: 1.240.2; Use the module subsystem's ability to process SYSCTL_SETUP() entries to automate installation of sysctl nodes. Note that there are still a number of device and pseudo-device modules that create entries tied to individual device units, rather than to the module itself. These are not changed.
1.239	27-Feb-2020	ad	Tighten up the locking around vp->v_iflag a little more after the recent split of vmobjlock & v_interlock.
1.238	17-Jan-2020	ad	VFS_VGET(), VFS_ROOT(), VFS_FHTOVP(): give them a "int lktype" argument, to allow us to get shared locks (or no lock) on the returned vnode. Matches FreeBSD.
1.237	03-Sep-2018	riastradh	branches: 1.237.4; 1.237.6; 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!)
1.236	16-Mar-2018	christos	branches: 1.236.2; PR/53103: Timo Buhrmester: linux emulation of sendto(2) broken The sockargs refactoring broke it, because sockargs only works with a user address. Added an argument to sockargs to indicate where the address is coming from. Welcome to 8.99.14.
1.235	17-Apr-2017	hannken	branches: 1.235.10; Remove unused argument "nextp" from vfs_busy() and vfs_unbusy(). Remove argument "keepref" from vfs_unbusy() and add vfs_ref() where needed.
1.234	17-Apr-2017	hannken	Add vfs_ref(mp) and vfs_rele(mp) to add or remove a reference to struct mount. Rename vfs_destroy(mp) to vfs_rele(mp) and replace incrementing mp->mnt_refcnt with vfs_ref(mp).
1.233	01-Apr-2017	riastradh	KASSERT(mutex_owned(vp->v_interlock)) in vnode iterator selector.
1.232	17-Feb-2017	hannken	Add generic genfs_suspendctl() and use it for all file systems. Layered file systems need work.
1.231	02-Nov-2015	pgoyette	branches: 1.231.2; 1.231.4; Don't forget to call nfs_fini() when we're finished. Without this, we leave a dangling pool nfsrvdescpl around.
1.230	15-Jul-2015	manu	Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.229	30-May-2014	hannken	branches: 1.229.2; 1.229.4; Change NFS from rbtree to vcache.
1.228	24-May-2014	christos	Introduce a selector function to the vfs vnode iterator so that we don't need to vget() vnodes that we are not interested at, and optimize locking a bit. Iterator changes reviewed by Hannken (thanks), the rest of the bugs are mine.
1.227	16-Apr-2014	maxv	An (un)privileged user can easily make the kernel dereference a NULL pointer. The kernel allows 'data' to be NULL; it's the fs's responsibility to ensure that it isn't NULL (if the fs actually needs data). ok christos@
1.226	23-Mar-2014	hannken	branches: 1.226.2; Change all vfsops to use C99 designated initializers. No functional changes intended.
1.225	17-Mar-2014	hannken	Change nfs_sync() to use vfs_vnode_iterator.
1.224	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.223	23-Nov-2013	christos	change the mountlist CIRCLEQ into a TAILQ
1.222	14-Sep-2013	martin	Remove unused variable
1.221	22-Jan-2013	dholland	branches: 1.221.2; Stuff UFS_ in front of a few of ufs's symbols to reduce namespace pollution. Specifically: ROOTINO -> UFS_ROOTINO WINO -> UFS_WINO NXADDR -> UFS_NXADDR NDADDR -> UFS_NDADDR NIADDR -> UFS_NIADDR MAXSYMLINKLEN -> UFS_MAXSYMLINKLEN MAXSYMLINKLEN_UFS[12] -> UFS[12]_MAXSYMLINKLEN (for consistency) Sort out ext2fs's misuse of NDADDR and NIADDR; fortunately, these have the same values in ext2fs and ffs. No functional change intended.
1.220	24-Oct-2011	hannken	branches: 1.220.2; 1.220.8; 1.220.12; 1.220.14; 1.220.16; VOP_GETATTR() needs a shared lock at least. As nfs_kqpoll() ignores the return value from VOP_GETATTR() initialize the attrributes to zero -- nfs_kqfilter() does the same.
1.219	07-Oct-2011	hannken	As vnalloc() always allocates with PR_WAITOK there is no longer the need to test its result for NULL.
1.218	12-Jun-2011	rmind	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.
1.217	12-Aug-2010	pooka	branches: 1.217.6; Do not return a garbage vnode in vpp if fhtovp fails. Fixes PR kern/43745 for nfs.
1.216	21-Jul-2010	hannken	Make holding v_interlock mandatory for callers of vget(). Announced some time ago on tech-kern.
1.215	09-Jul-2010	hannken	nfs_unmount(): No need to take a second reference for the root node. nfs_root(): Replace vget() with vref()/vn_lock(), this node already has a reference.
1.214	24-Jun-2010	hannken	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.
1.213	24-Jun-2010	hannken	Clean up vnode lock operations: - VOP_LOCK(vp, flags): Limit the set of allowed flags to LK_EXCLUSIVE, LK_SHARED and LK_NOWAIT. LK_INTERLOCK is no longer allowed as it makes no sense here. - VOP_ISLOCKED(vp): Remove the for some time unused return value LK_EXCLOTHER. Mark this operation as "diagnostic only". Making a lock decision based on this operation is no longer allowed. Discussed on tech-kern.
1.212	15-May-2010	dholland	nfs_statvfs should return NFS_MAXNAMLEN, not MAXNAMLEN. (Compile-tested only, but that should be ok)
1.211	02-Mar-2010	pooka	branches: 1.211.2; Get rid of dependency on fs_nfs.h, i.e. source modules with conditional content depending on if the NFS client is wanted or not. The server can now be made an independent module not depending on the nfs client. Tested with rump_nfs (standalone client), rump_nfsd (standalone nfsd) and a qemu installation with both the client and the server.
1.210	15-Mar-2009	cegger	branches: 1.210.2; ansify function definitions
1.209	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.208	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.207	14-Mar-2009	dsl	Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.206	17-Dec-2008	cegger	branches: 1.206.2; kill MALLOC and FREE macros.
1.205	19-Nov-2008	ad	Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.204	14-Nov-2008	ad	Remove COMPAT ifdefs that might as well be comments (i.e., they cost us almost nothing).
1.203	22-Oct-2008	matt	branches: 1.203.2; 1.203.4; 1.203.10; 1.203.14; Don't need nfs_vfs_reinit anymore since we don't resize tables anymore. Move reinit code to init case.
1.202	22-Oct-2008	matt	Change NFS to use a RB-tree for its FH->nfsnode lookups.
1.201	30-Sep-2008	pooka	Since the nfs root vnode is eternally constant, fully initialize it in mountfs instead of deferring part of the initialization to VFS_ROOT(). Fixes theoretical future bugs for nfs roots.
1.200	10-May-2008	rumble	branches: 1.200.4; Convert file systems to dynamically attach with the new module interface. Make VFS hooks dynamic while we're here and say farewell to VFS_ATTACH and VFS_HOOKS_ATTACH linksets. As a consequence, most of the file systems can now be loaded as new style modules. Quick sanity check by ad@.
1.199	06-May-2008	ad	branches: 1.199.2; PR kern/38141 lookup/vfs_busy acquire rwlock recursively Simplify the mount locking. Remove all the crud to deal with recursion on the mount lock, and crud to deal with unmount as another weirdo lock. Hopefully this will once and for all fix the deadlocks with this. With this commit there are two locks on each mount: - krwlock_t mnt_unmounting. This is used to prevent unmount across critical sections like getnewvnode(). It's only ever read locked with rw_tryenter(), and is only ever write locked in dounmount(). A write hold can't be taken on this lock if the current LWP could hold a vnode lock. - kmutex_t mnt_updating. This is taken by threads updating the mount, for example when going r/o -> r/w, and is only present to serialize updates. In order to take this lock, a read hold must first be taken on mnt_unmounting, and the two need to be held across the operation. One effect of this change: previously if an unmount failed, we would make a half hearted attempt to back out of it gracefully, but that was unlikely to work in a lot of cases. Now while an unmount that will be aborted is in progress, new file operations within the mount will fail instead of being delayed. That is unlikely to be a problem though, because if the admin requests unmount of a file system then s(he) has made a decision to deny access to the resource.
1.198	30-Apr-2008	ad	PR kern/38135 vfs_busy/vfs_trybusy confusion The previous fix worked, but it opened a window where mounts could have disappeared from mountlist while the caller was traversing it using vfs_trybusy(). Fix that.
1.197	29-Apr-2008	ad	PR kern/38057 ffs makes assuptions about devvp file system PR kern/33406 softdeps get stuck in endless loop Introduce VFS_FSYNC() and call it when syncing a block device, if it has a mounted file system.
1.196	13-Feb-2008	yamt	branches: 1.196.6; 1.196.8; 1.196.10; reject files larger than nm_maxfilesize.
1.195	13-Feb-2008	yamt	nfs_mountroot: kmem_alloc+memset -> kmem_zalloc
1.194	30-Jan-2008	ad	PR kern/37706 (forced unmount of file systems is unsafe): - Do reference counting for 'struct mount'. Each vnode associated with a mount takes a reference, and in turn the mount takes a reference to the vfsops. - Now that mounts are reference counted, replace the overcomplicated mount locking inherited from 4.4BSD with a recursable rwlock.
1.193	28-Jan-2008	dholland	Fix some race conditions in rename. Introduce a per-FS rename lock and new vfsops to manipulate it. Get this lock while renaming. Also add another relookup() in do_sys_rename, which is a hack to kludge around some of the worst deficiencies of ufs_rename. reviewed-by: pooka (and an earlier rev by ad) posted on tech-kern with no objections.
1.192	20-Jan-2008	joerg	Now that __HAVE_TIMECOUNTER and __HAVE_GENERIC_TODR are invariants, remove the conditionals and the code associated with the undef case.
1.191	03-Jan-2008	pooka	valloc -> vnalloc, vfree -> vnfree Avoids collision with userland valloc(3). no functional change ad ok
1.190	02-Jan-2008	yamt	use kmem_alloc instead of malloc.
1.189	02-Jan-2008	ad	Merge vmlocking2 to head.
1.188	26-Nov-2007	pooka	branches: 1.188.2; 1.188.6; 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
1.187	28-Oct-2007	yamt	branches: 1.187.2; make NFS_ATTRTIMEO a function.
1.186	10-Oct-2007	ad	branches: 1.186.2; Merge from vmlocking: - Split vnode::v_flag into three fields, depending on field locking. - simple_lock -> kmutex in a few places. - Fix some simple locking problems.
1.185	06-Sep-2007	rmind	branches: 1.185.2; nfs_mount: Plug a possible leaks. Invented in 1.114 rev. From CID: 4534
1.184	10-Aug-2007	yamt	branches: 1.184.2; - instead of scanning an array of iods, maintain a list of idle iods. - make nfs_getset_niothreads MP friendly.
1.183	05-Aug-2007	yamt	branches: 1.183.2; use kpause rather than lbolt.
1.182	31-Jul-2007	pooka	branches: 1.182.2; * nuke the nameidata parameter from VFS_MOUNT(). Nobody on tech-kern knew what it was supposed to be used for and wrstuden gave a go-ahead * while rototilling, convert file systems which went easily to use VFS_PROTOS() instead of manually prototyping the methods
1.181	26-Jul-2007	pooka	Use eopnotsupp() instead of vfs_stdsuspendctl() and retire the latter.
1.180	20-Jul-2007	pooka	In sync, skip over vnodes based on if they are clean rather than if they have pages.
1.179	17-Jul-2007	pooka	branches: 1.179.2; Make set_statvfs_info() take a parameter for the vfs name instead of always retrieving it from mp->mnt_op->vfs_name christos ok
1.178	12-Jul-2007	dsl	Change the VFS_MOUNT() interface so that the 'data' buffer passed to the fs code is a kernel buffer, pass though the length of the buffer as well. Since the length of the userspace buffer isn'it (yet) passed through the mount system call, add a field to the vfsops structure containing the default length. Split sys_mount() for calls from compat code. Ride one of the recent kernel version changes - old fs LKMs will load, but sys_mount() will reject any attempt to use them.
1.177	29-Apr-2007	yamt	don't forget to destroy mutex and condvar.
1.176	29-Apr-2007	yamt	use condvar.
1.175	29-Apr-2007	yamt	use mutex and condver.
1.174	04-Mar-2007	christos	branches: 1.174.2; 1.174.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.173	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.172	15-Feb-2007	yamt	branches: 1.172.2; use mutex and rwlock rather than lockmgr.
1.171	19-Jan-2007	hannken	New file system suspension API to replace vn_start_write and vn_finished_write. The suspension helpers are now put into file system specific operations. This means every file system not supporting these helpers cannot be suspended and therefore snapshots are no longer possible. Implemented for file systems of type ffs. The new API is enabled on a kernel option NEWVNGATE. This option is not enabled by default in any kernel config. Presented and discussed on tech-kern with much input from Bill Studenmund <wrstuden@netbsd.org> and YAMAMOTO Takashi <yamt@netbsd.org>. Welcome to 4.99.9 (new vfs op vfs_suspendctl).
1.170	27-Dec-2006	yamt	- remove the rest of nqnfs. - reject NFSMNT_MNTD and NFSMNT_KERB. (no users in tree.)
1.169	27-Dec-2006	yamt	remove nqnfs.
1.168	09-Nov-2006	yamt	remove some __unused in function parameters.
1.167	25-Oct-2006	reinoud	Revisit mnt_vnodelist TAILQ patch. Remove all suspicious TAILQ_FOREACH() loops where vnodes can get removed or added during the loops. This could lead to panic's on unmount since nodes are skipped or otherwise TAILQ_NEXT(0xdeadbeef, ...) was dereferenced.
1.166	20-Oct-2006	reinoud	Replace the LIST structure mp->mnt_vnodelist to a TAILQ structure since all vnodes were synced and processed backwards. This meant that the last accessed node was processed first and the earlierst last. An extra benefit is the removal of the ugly hack from the Berkly days on LFS. In the proces, i've also replaced the various variations hand written loops by the TAILQ_FOREACH() macro's.
1.165	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.164	02-Sep-2006	yamt	branches: 1.164.2; 1.164.4; nfs_fhtovp: try to detect stale or invalid handles by issuing VOP_GETATTR.
1.163	02-Sep-2006	yamt	implement vptofh and fhtovp for nfs.
1.162	02-Sep-2006	christos	fix default type decls fix incomplete initializer
1.161	24-Aug-2006	christos	Don't free what we did not allocate.
1.160	23-Aug-2006	christos	Change iostat_alloc() to take the parent pointer and the name directly, so that callers are not responsible for initializing the fields. Store the name inside the struct instead of maintaining a pointer to external storage, or leaked memory (nfs case).
1.159	23-Jul-2006	ad	Use the LWP cached credentials where sane.
1.158	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.157	07-Jun-2006	kardel	branches: 1.157.2; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.156	20-May-2006	yamt	mountnfs: reject wrongly-sized filehandle for nfsv2.
1.155	14-May-2006	elad	branches: 1.155.2; integrate kauth.
1.154	20-Apr-2006	blymn	Prefix iostat structure elements with io_
1.153	14-Apr-2006	blymn	Make i/o statistics collection more generic, include tape drives and nfs mounts in the set of devices that statistics will be reported on.
1.152	21-Feb-2006	thorpej	branches: 1.152.2; 1.152.4; 1.152.6; Use device_class() instead of accessing dv_class directly.
1.151	11-Dec-2005	christos	branches: 1.151.2; 1.151.4; 1.151.6; merge ktrace-lwp.
1.150	23-Sep-2005	jmmv	Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.149	19-Sep-2005	christos	ATTRTIMEO takes 2 args.
1.148	09-Jun-2005	atatat	branches: 1.148.2; Properly fix the constipated lossage wrt -Wcast-qual and the sysctl code. I know it's not the prettiest code, but it seems to work rather well in spite of itself.
1.147	29-May-2005	christos	- sprinkle const - avoid shadowed variables - mark bad const use with XXXUNCONST
1.146	29-Mar-2005	thorpej	- Define a VFS_ATTACH() macro that places a reference to a vfsops structure into the "vfsops" link set. - Use VFS_ATTACH() where vfsops are declared for individual file systems. - In vfsinit(), traverse the "vfsops" link set, rather than vfs_list_initial[].
1.145	26-Feb-2005	perry	branches: 1.145.2; nuke trailing whitespace
1.144	02-Jan-2005	thorpej	branches: 1.144.2; 1.144.4; Add the system call and VFS infrastructure for file system extended attributes. From FreeBSD.
1.143	15-Aug-2004	mycroft	Fixing age old cruft: * Rather than using mnt_maxsymlinklen to indicate that a file systems returns d_type fields(!), add a new internal flag, IMNT_DTYPE. Add 3 new elements to ufsmount: * um_maxsymlinklen, replaces mnt_maxsymlinklen (which never should have existed in the first place). * um_dirblksiz, which tracks the current directory block size, eliminating the FS-specific checks littered throughout the code. This may be used later to make the block size variable. * um_maxfilesize, which is the maximum file size, possibly adjusted lower due to implementation issues. Sync some bug fixes from FFS into ext2fs, particularly: * ffs_lookup.c 1.21, 1.28, 1.33, 1.48 * ffs_inode.c 1.43, 1.44, 1.45, 1.66, 1.67 * ffs_vnops.c 1.84, 1.85, 1.86 Clean up some crappy pointer frobnication.
1.142	12-Jul-2004	yamt	nfs_fsinfo: when changing rsize/wsize, keep mnt_fs_bshift in-sync. otherwise genfs_getpages behaves badly.
1.141	05-Jul-2004	pk	Call inittodr() from main(). Let file system code set the recorded `last update' time (if any) through the new function setrootfstime().
1.140	25-May-2004	hannken	Add ffs internal snapshots. Written by Marshall Kirk McKusick for FreeBSD. - Not enabled by default. Needs kernel option FFS_SNAPSHOT. - Change parameters of ffs_blkfree. - Let the copy-on-write functions return an error so spec_strategy may fail if the copy-on-write fails. - Change genfs_lock() to use vp->v_vnlock instead of &vp->v_lock. - Add flag B_METAONLY to VOP_BALLOC to return indirect block buffer. - Add a function ffs_checkfreefile needed for snapshot creation. - Add special handling of snapshot files: Snapshots may not be opened for writing and the attributes are read-only. Use the mtime as the time this snapshot was taken. Deny mtime updates for snapshot files. - Add function transferlockers to transfer any waiting processes from one lock to another. - Add vfsop VFS_SNAPSHOT to take a snapshot and make it accessible through a vnode. - Add snapshot support to ls, fsck_ffs and dump. Welcome to 2.0F. Approved by: Jason R. Thorpe <thorpej@netbsd.org>
1.139	25-May-2004	atatat	Sysctl descriptions under vfs subtree
1.138	22-May-2004	jonathan	Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.137	27-Apr-2004	jrf	First pass for some caddr_t removal and changes to get rid of it where we no longer use and/or need it - removed casts from unionfs, deadfs and fdesc (there are more to hunt down still) - changed vfs_quotactl args argumet from caddr_t to void * - changed vfs_quotactl structures/callers to reflect the api change Compiled fine and ran for about a day. Approved/reviewed by christos@netbsd.org and gimpy@netbsd.org.
1.136	21-Apr-2004	christos	Replace the statfs() family of system calls with statvfs(). Retain binary compatibility.
1.135	24-Mar-2004	atatat	branches: 1.135.2; Tango on sysctl_createv() and flags. The flags have all been renamed, and sysctl_createv() now uses more arguments.
1.134	04-Dec-2003	atatat	Dynamic sysctl. Gone are the old kern_sysctl(), cpu_sysctl(), hw_sysctl(), vfs_sysctl(), etc, routines, along with sysctl_int() et al. Now all nodes are registered with the tree, and nodes can be added (or removed) easily, and I/O to and from the tree is handled generically. Since the nodes are registered with the tree, the mapping from name to number (and back again) can now be discovered, instead of having to be hard coded. Adding new nodes to the tree is likewise much simpler -- the new infrastructure handles almost all the work for simple types, and just about anything else can be done with a small helper function. All existing nodes are where they were before (numerically speaking), so all existing consumers of sysctl information should notice no difference. PS - I'm sorry, but there's a distinct lack of documentation at the moment. I'm working on sysctl(3/8/9) right now, and I promise to watch out for buses.
1.133	02-Oct-2003	itojun	plug mbuf leak due to manual mbuf handling. PR kern/13807. (martti confirmed that it stabilizes the situation described in kren/13807)
1.132	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.131	29-Jun-2003	fvdl	branches: 1.131.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.130	29-Jun-2003	thorpej	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.129	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
1.128	21-May-2003	yamt	remove local definitions of TRUE and FALSE.
1.127	03-May-2003	yamt	better handling of write verifier change.
1.126	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.125	16-Apr-2003	christos	PR/1796: John Kohl: statfs misbehaves under chrooted environments. - Under chroot it displays only the visible filesystems with appropriate paths. - The statfs f_mntonname gets adjusted to contain the real path from root. - While was there, fixed a bug in ext2fs, locking problems with vfs_getfsstat(), and factored out some of the vfsop statfs() code to copy_statfs_info(). This fixes the problem where some filesystems forgot to set fsid. - Made coda look more like a normal fs.
1.124	02-Apr-2003	yamt	use queue manipulation macros.
1.123	28-Mar-2003	yamt	if rsize was explicitly specified by mount_nfs, prefer it to rtpref from nfsd. the same for wsize and wtpref. ok'ed by fvdl.
1.122	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.121	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.120	24-Nov-2002	scw	Fix an uninitialised variable warning.
1.119	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.118	21-Oct-2002	enami	When printing filesystem specific parameters, also print the address and port of server numerically.
1.117	01-Oct-2002	christos	forgot to set deadthresh; thanks to YAMAMOTO Takashi.
1.116	21-Sep-2002	christos	MNT_GETARGS support
1.115	30-Jul-2002	soren	Die, qaddr_t, die! - mnt_data in struct mount is already effectively a void *, so stop pretending otherwise.
1.114	26-Jul-2002	enami	Synchronize code and comment again to prevent mbuf leak. Sprinkle some KNF while I'm here.
1.113	25-Jul-2002	jdolecek	Reduce stack usage on the NFS mount code path. This fixes kernel stack overflow when using IPsec on vax, as reported by Olaf Seibert on current-users@.
1.112	04-Dec-2001	christos	branches: 1.112.8; 1.112.10; PR/14817: Gregory McGarry: NFS_V2_ONLY doesn't seem to work.
1.111	10-Nov-2001	lukem	add RCSIDs
1.110	08-Oct-2001	chs	branches: 1.110.2; revert a change that I accidentally included with ubcperf.
1.109	20-Sep-2001	chs	fix nfs_bmap() so that it works for both genfs_{get,put}pages() and swap/vnd.
1.108	15-Sep-2001	chs	a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
1.107	15-Sep-2001	chs	add a new VFS op, vfs_reinit, which is called when desiredvnodes is adjusted via sysctl. file systems that have hash tables which are sized based on the value of this variable now resize those hash tables using the new value. the max number of FFS softdeps is also recalculated. convert various file systems to use the <sys/queue.h> macros for their hash tables.
1.106	30-Jul-2001	jdolecek	branches: 1.106.2; Check the passed file handle length _before_, not _after_ copyin()
1.105	30-Jul-2001	fvdl	Check the length of a passed in filehandle to the mount call before doing a copyin. From Ken Ashcraft @ Stanford via Constantine Sapuntzakis.
1.104	01-Jul-2001	gmcgarry	branches: 1.104.2; Introduce NFS_DEFAULT_NIOTHREADS to define the default number of nfs_niothreads instead of hard-coding 4. This change has the advantage that the default can be specified at compile time. If the root filesystem is mounted over NFS we don't have an opportunity to use the syscall to limit the number of threads. Useful on small-memory machines.
1.103	30-May-2001	mrg	use _KERNEL_OPT
1.102	28-Apr-2001	bjh21	When NFS_V2_ONLY is defined, refuse to mount NFSv3 and NQNFS filesystems, rather than pretending they're NFSv2 and hoping for the best. Fix based on that supplied by Christian Groessler.
1.101	12-Feb-2001	fvdl	branches: 1.101.2; Instead of storing the filehandle in the mount structure, store the vnode pointer. This avoids a locking problem with nfs_nget, and can be done because we always have a reference on the root vnode of the filesystem.
1.100	06-Feb-2001	fvdl	Do actual vnode locking for NFS.
1.99	22-Jan-2001	jdolecek	make filesystem vnodeop, specop, fifoop and vnodeopv_* arrays const
1.98	10-Dec-2000	chs	in *_sync(), don't skip vnodes which have (potentially dirty) pages.
1.97	27-Nov-2000	chs	Initial integration of the Unified Buffer Cache project.
1.96	19-Sep-2000	fvdl	Update for VOP_FSYNC parameter change.
1.95	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.94	23-Aug-2000	enami	Update nfs mount flags correctly. Fixes a bug introduced in rev. 1.65.
1.93	30-Jul-2000	simonb	Remove inclusion of <uvm/uvm_extern.h> that was there only to keep <sys/sysctl.h> happy.
1.92	27-Jun-2000	mrg	remove include of <vm/vm.h>
1.91	10-Jun-2000	assar	branches: 1.91.2; make vfs_getnewfsid only take one argument and fetch the name of the filesystem from the supplied mount argument. also make makefstype take a const parameter. update all the callers.
1.90	07-May-2000	tsarna	branches: 1.90.2; Auto-adjusting vfs.nfs.iothreads: when mounting the first nfs filesystem, if the number of threads is "-1", meaning it's never been set, then set it to 4. You can override by setting this to some other number (including 0) before or after mounting, of course. Thanks to whoever it was that suggested this on ICB... sorry I don't remember who.
1.89	15-Apr-2000	tsarna	Death to nfsiod! It is replaced by kernel threads that do the same thing. The number of kernel threads used is set with the vfs.nfs.iothreads sysctl.
1.88	30-Mar-2000	augustss	Remove register declarations.
1.87	29-Mar-2000	simonb	Don't need to include <sys/conf.h> here.
1.86	16-Mar-2000	jdolecek	Add new VFS op routine - vfs_done and call it on filesystem detach in vfs_detach(). vfs_done may free global filesystem's resources, typically those allocated in respective filesystem's init function. Needed so those filesystems which went in via LKM have a chance to clean after themselves before unloading. This fixes random panics when LKM for filesystem using pools was loaded and unloaded several times. For each leaf filesystem, add appropriate vfs_done routine.
1.85	15-Nov-1999	fvdl	Add Kirk McKusick's soft updates code to the trunk. Not enabled by default, as the copyright on the main file (ffs_softdep.c) is such that is has been put into gnusrc. options SOFTDEP will pull this in. This code also contains the trickle syncer. Bump version number to 1.4O
1.84	29-Aug-1999	sommerfeld	branches: 1.84.2; 1.84.4; 1.84.8; Once the mount structure is definitely doomed, always set the NFSMNT_DISMNT bit in it so that any waiters can go away cleanly. (formerly, we did this only in the NQNFS/KERB cases).
1.83	06-Mar-1999	fair	branches: 1.83.2; 1.83.4; Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.82	05-Mar-1999	mycroft	Clean up some sign extension bogosity in statfs, so negative numbers are actually negative on a LP64 client.
1.81	26-Feb-1999	wrstuden	Modify vfsops to seperate vfs_fhtovp() into two routines. vfs_fhtovp() now only handles the file handle to vnode conversion, and a new call, vfs_checkexp(), performs the export verification.
1.80	21-Feb-1999	drochner	-call nfs_boot_cleanup() if mount failed -g/c diskless swap initialization
1.79	12-Nov-1998	fvdl	Use different names for the "nfscon" label to tsleep(), so that it can be seen in which one a process is sleeping.
1.78	28-Sep-1998	drochner	Use the "atime" instead of "mtime" of the remote root directory as base for inittodr() - it is closer to the current time.
1.77	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.76	05-Jul-1998	jonathan	* defopt COMPAT_{09,10,11,12,13} and COMPAT_NOMID. TODO: revisit interaction between native compat and emul compat usage.
1.75	24-Jun-1998	sommerfe	Always include fifos; "not an option any more".
1.74	22-Jun-1998	sommerfe	defopt for options FIFO
1.73	05-Jun-1998	kleink	Convert fsync vnode operator implementations and usage from the old `waitfor' argument and MNT_WAIT/MNT_NOWAIT to `flags' and FSYNC_WAIT.
1.72	24-Mar-1998	fvdl	Re-instate call to "safe" disconnect function that got lost during the Lite2 merge.
1.71	03-Mar-1998	thorpej	Historical practice assumes that NFS root mounts are initially read/write.
1.70	03-Mar-1998	fvdl	Don't try to apply the cookie endian heuristic on a mounted file (e.g. a swapfile). From Matthias Drochner.
1.69	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.68	18-Feb-1998	thorpej	Place a pointer to an array of our vnodeopv_desc *'s in our vfsops structure, for use by vfs_attach().
1.67	30-Jan-1998	fvdl	Only take the receive lock before disconnecting when doing it from nfs_decode_args. Otherwise we might just end up locking against ourselves. XXX workaround, will do ok for now. Proper fix forthcoming.
1.66	19-Oct-1997	fvdl	branches: 1.66.2; * Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.65	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.64	09-Sep-1997	gwr	Move the call to nfs_boot_getfh() from nfs_vfsops.c to nfs_boot.c (just for better isolation - it can now be static)
1.63	29-Aug-1997	gwr	Supporting changes for the new BOOTP support in nfs_mountroot.
1.62	18-Jul-1997	christos	branches: 1.62.2; Fix reversed test for version 3 that broke nfs version 2 mounts.
1.61	17-Jul-1997	fvdl	* Deal with servers that don't give complete FSINFO (like NT) From Olaf Seibert <rhialto@polder.ubc.kun.nl> (PR 3687) * Make an attempt to check the maximum filesize before attempting a write to the server, as write RPCs will typically happen asynchronously, and the process will not see the error. Fixes problems with unexpectly truncated files at 4G * Pass up errors in nfs_writerpc correctly
1.60	12-Jun-1997	mrg	remove swap configuration.
1.59	27-May-1997	gwr	Minor reorganization of nfs_mountroot code to simplify BOOTP support. The RPC/bootparamd calls to get the root and swap paths are now done in nfs_boot_init() instead of nfs_boot_getfh(), so the latter now just does the RPC/mountd call. Also changed some panics into error returns.
1.58	22-Feb-1997	fvdl	Silently clear NFSMNT_NOCONN if it's a TCP mount.
1.57	04-Feb-1997	fvdl	branches: 1.57.2; 1.57.4; * Make sure a new socket is created when switching to/from NOCONN with a mount * Add extra printf statements to hopefully get some more info on lockups, specifically when a send error is ignored.
1.56	31-Jan-1997	thorpej	- Add nfs_mountroot to nfs_vfsops. - Only attempt to mount NFS root on a DV_IFNET class device. - If nfs_boot_init() failes, return the error code to the caller.
1.55	22-Dec-1996	cgd	branches: 1.55.2; Change the second and third args to struct vfsops' (vfs_mount)() to 'const char ', and 'void ', respectively. The second arg is taken directly from user arguments, and is const there, so must be const in the prototypes and functions. The third arg is also taken directly from user arguments. It doesn't have to be changed, but since it's cleaner to keep the type the same as the user arg's type, and I'm already making the 'const char ' change...
1.54	03-Dec-1996	thorpej	Make NFSSERVER work without NFSCLIENT. This is achieved by splitting the client and server/shared data initialization into separate functions, and calling the server/shared initialization directly from main(). Problem noted in PR #1308 (Kenneth Stailey) and PR #1780 (Chris Demetriou). Fix suggested in PR #1780 by Chris Demetriou, and munged a bit by me, and OK'd by Frank van der Linden <fvdl@netbsd.org>.
1.53	02-Dec-1996	thorpej	NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.52	20-Oct-1996	fvdl	Enhancements from Matthias Drochner: - Try V3 first for diskless booting. Fall back to V2 if V3 fails. - optionally (option NFS_BOOT_TCP) try a TCP mount first for diskless booting. Fall back to UDP if it fails. - Enable switching between UDP and TCP for remounts.
1.51	13-Oct-1996	christos	revert kprintf changes
1.50	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.49	24-Jun-1996	pk	Ignore the mountpoint's `v_usecount' in nfs_unmount() if MNT_FORCE is on. This takes care of two related problems: - `umount -f' wouldn't work if someone's working directory is the filesystem root. - vfs_unmountall() would complain about a busy `/' on a diskless setup.
1.48	14-Jun-1996	cgd	avoid unnecessary checks of m_get/MGET/etc.'s return values. When they're called with M_WAIT, they are defined to never return NULL.
1.47	23-May-1996	fvdl	* Make mounts with symlinks work (needed for direct mounts with amd). PR #1917 * Never change the NQNFS flag and/or version when just doing an update mount. Fixes a problem that made diskless booting impossible under some circumstances.
1.46	24-Mar-1996	fvdl	branches: 1.46.4; Return earlier on error in nfs_statfs. Should fix problem reported by both mrg and cgd.
1.45	17-Mar-1996	christos	Fix printf format strings.
1.44	13-Mar-1996	fvdl	Make readdirsize default to rsize if rsize is explicitly specified, and readdirsize isn't.
1.43	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.42	13-Feb-1996	gwr	Do the RPC to bootparamd a little later (just before the mountd call) so that we do not ask for the "swap" path when swapping on disk.
1.41	09-Feb-1996	christos	nfs prototype changes
1.40	01-Feb-1996	jtc	Rename struct timespec fields to conform to POSIX.1b
1.39	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.38	13-Aug-1995	mycroft	splnet --> splsoftnet
1.37	18-Jun-1995	cgd	don't assume the f_fsnamelen is nul-truncated or longer than MFSNAMELEN
1.36	02-Jun-1995	mycroft	Fix more off by one errors.
1.35	18-Mar-1995	gwr	Print the "root/swap on ..." messages here. Add NFS_BOOT_OPTIONS for things like NFSMNT_NOCONN.
1.34	09-Mar-1995	mycroft	copy*str() should use size_t.
1.33	18-Jan-1995	mycroft	Turn mountlist into a CIRCLEQ, and handle setting and checking of MNT_ROOTFS differently.
1.32	23-Aug-1994	pk	branches: 1.32.2; When updating an NFS mountpoint, we cannot just increase `rsize' or `wsize' without also adjusting the corresponding socket buffers. We could probably call sbrelease/sbreserve/soreserve ourselves without much harm, but we'd have to duplicate much of the logic in nfs_connect(). In stead, blow the socket away entirely and let nfs_connect() do its job again.
1.31	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.30	14-Aug-1994	gwr	Add the option NFS_BOOT_RWSIZE to allow diskless boot configuration to start with a reduced NFS read and write size (need for wd8003).
1.29	12-Aug-1994	cgd	kill two errant spaces.
1.28	11-Aug-1994	gwr	Diskless boot will now bind the local socket to a reserved port to satisfy picky servers. Also fix some missing initializations. (Thanks to Chuck Cranor for PR#394 -- now fixed.)
1.27	03-Jul-1994	mycroft	branches: 1.27.2; Save FS type at mount time for some later tests.
1.26	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.25	28-Jun-1994	gwr	Minor nits: replace ... with ... p->p_cred->pc_ucred p->p_ucred x / DEV_BSIZE x >> DEV_BSHIFT
1.24	22-Jun-1994	pk	straighten out diskless swap code somewhat.
1.23	14-Jun-1994	gwr	Fix false "hits" in the attribute cache when booting diskless. (Yet another thing that breaks when time.tv_sec is near zero...)
1.22	13-Jun-1994	gwr	New diskless boot code (uses RARP, bootparamd).
1.21	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.20	18-May-1994	cgd	put sync printing in one place
1.19	13-May-1994	mycroft	Trivial function name change.
1.18	11-May-1994	mycroft	Cast some args to caddr_t.
1.17	23-Apr-1994	cgd	make fs types consistent over new kernels. also, some proto foo.
1.16	21-Apr-1994	cgd	Convert mount, vnode, and buf structs to use <sys/queue.h>. Also, some knf and structure frobbing to do along with it.
1.15	18-Apr-1994	glass	revised nfs diskless support. uses bootp+rpc to gather parameters
1.14	14-Apr-1994	cgd	fs types are names now.
1.13	10-Apr-1994	cgd	make damn sure nothing's holding on the the mount point vnode
1.12	31-Mar-1994	glass	make panic string unique
1.11	21-Dec-1993	cgd	oops; fix last
1.10	21-Dec-1993	cgd	from jsp: Changed to get attributes of root node and generate correct type, rather than assuming it's a directory. This allows Amd direct mounts to work correctly.
1.9	18-Dec-1993	mycroft	Canonicalize all #includes.
1.8	07-Dec-1993	pk	Exclusive access when manipulating flag field in mount structure.
1.7	07-Dec-1993	pk	Don't allow the NFS_LOCKBITS to be set or reset from user land. Allow other flags (SOFT,HARD,SPONGY, etc) to be altered by `mount -u'.
1.6	06-Dec-1993	pk	Allow changing of various NFS parameters by using `mount -u ...'.
1.5	19-Nov-1993	cgd	patch from Ukai Fumitoshi <ukai@kmc.kyoto-u.ac.jp> to do the right thing with NFS fsid's and getnewfsid()
1.4	13-Jul-1993	cgd	branches: 1.4.4; diskless changes made last time were hosed; were using NULL for "no credentials" rather than NOCRED.
1.3	07-Jul-1993	cgd	changes from ws to support diskless booting... these are "OK" on inspection and after testing... (actually, currently, none of the changed code is even used...)
1.2	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.4.4.5	21-Dec-1993	cgd	update from trunk
1.4.4.4	21-Dec-1993	cgd	update from trunk
1.4.4.3	20-Nov-1993	cgd	update from trunk
1.4.4.2	14-Nov-1993	mycroft	Canonicalize all #includes.
1.4.4.1	24-Sep-1993	mycroft	Make all files using spl*() #include cpu.h. Changes from trunk. nfs_vfsops.c, nfsmount.h: Make nfs_quotactl() take an int rather than a uid_t, as it might be -1. nfs_vnops.c: va_size and va_bytes are now quads.
1.27.2.4	19-Aug-1994	mycroft	update from trunk
1.27.2.3	14-Aug-1994	mycroft	update from trunk
1.27.2.2	12-Aug-1994	mycroft	update from trunk
1.27.2.1	11-Aug-1994	mycroft	update from trunk
1.32.2.2	23-Aug-1994	pk	When updating an NFS mountpoint, we cannot just increase `rsize' or `wsize' without also adjusting the corresponding socket buffers. We could probably call sbrelease/sbreserve/soreserve ourselves without much harm, but we'd have to duplicate much of the logic in nfs_connect(). In stead, blow the socket away entirely and let nfs_connect() do its job again.
1.32.2.1	23-Aug-1994	pk	file nfs_vfsops.c was added on branch netbsd-1-0 on 1994-08-23 09:31:01 +0000
1.46.4.2	11-Dec-1996	mycroft	From trunk: Ignore reference count when using MNT_FORCE.
1.46.4.1	25-May-1996	fvdl	Pull in bugfixes from main branch.
1.55.2.1	14-Jan-1997	thorpej	Snapshot of work-in-progress, committed to private branch. These changes implement machine-independent root device and file system selection. Notable features: - All ports behave in a consistent manner regarding root device selection. - No more "options GENERIC"; all kernels have the ability to boot with RB_ASKNAME to select root device and file system type. - Root file system type can be wildcarded; a machine-independent function will try all possible file systems for the selected root device until one succeeds. - If the root file system fails to mount, the operator will be given the chance to select a new root device and file system type, rather than having the machine simply panic. - nfs_mountroot() no longer panics if any part of the NFS mount process fails; it now returns an error, giving the operator a chance to recover. - New, more consistent, config(8) grammar. The constructs: config netbsd swap generic config netbsd root on nfs have been replaced with: config netbsd root on ? type ? config netbsd root on ? type nfs Additionally, the operator may select or wildcard root file system type in the kernel configuration file: config netbsd root on cd0a type cd9660 config(8) now requires that a "root" specification be made. "root" may be wired down or wildcarded. "swap" and "dump" specifications are optional, and follow previous semantics. - config(8) has a new "file-system" keyword, used to configure file systems into the kernel. Eventually, this will be used to generate the default vfssw[]. - "options NFSCLIENT" is obsolete, and is replaced by "file-system NFS". "options NFSSERVER" still exists, since NFS server support is independent of the NFS file system client. - sys/arch/<foo>/<foo>/swapgeneric.c is no longer used, and will be removed; all information is now generated by config(8). As of this commit, all ports except arm32 have been updated to use the new setroot(). Only SPARC, i386, and Alpha ports have been tested at this time. Port masters should test these changes on their ports, and report any problems back to me. More changes are on their way, including RB_ASKNAME support in nfs_mountroot() (to prompt for server address and path) and, potentially, the ability to select rarp/bootparam or bootp in nfs_mountroot().
1.57.4.1	02-Mar-1997	mrg	swap configuration is no longer done at boot time.
1.57.2.1	12-Mar-1997	is	Merge in changes from Trunk
1.62.2.3	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.62.2.2	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.62.2.1	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.66.2.1	07-Feb-1998	mellon	Pull up 1.67 (fvdl)
1.83.4.1	04-Jul-1999	chs	initialize new struct mount fields in nfs_mountfs().
1.83.2.1	05-Nov-1999	cgd	pull up rev 1.84 from trunk (requested by fvdl): Avoid a panic when forcibly unmounting a hung NFS mount, e.g. at reboot.
1.84.8.2	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.84.8.1	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".
1.84.4.1	19-Oct-1999	fvdl	Bring in Kirk McKusick's FFS softdep code on a branch.
1.84.2.5	12-Mar-2001	bouyer	Sync with HEAD.
1.84.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.84.2.3	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.84.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.84.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.90.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.91.2.1	14-Dec-2000	he	Pull up revision 1.96 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.101.2.13	11-Dec-2002	thorpej	Sync with HEAD.
1.101.2.12	22-Oct-2002	thorpej	Sync with HEAD.
1.101.2.11	18-Oct-2002	nathanw	Catch up to -current.
1.101.2.10	01-Aug-2002	nathanw	Catch up to -current.
1.101.2.9	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.101.2.8	24-Jun-2002	nathanw	Curproc->curlwp renaming. Change uses of "curproc->l_proc" back to "curproc", which is more like the original use. Bare uses of "curproc" are now "curlwp". "curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL) so that it is always safe to reference curproc (dereferencing curproc is another story, but that's always been true).
1.101.2.7	08-Jan-2002	nathanw	Catch up to -current.
1.101.2.6	14-Nov-2001	nathanw	Catch up to -current.
1.101.2.5	22-Oct-2001	nathanw	Catch up to -current.
1.101.2.4	21-Sep-2001	nathanw	Catch up to -current.
1.101.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.101.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.101.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.104.2.4	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
1.104.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.104.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.104.2.1	03-Aug-2001	lukem	update to -current
1.106.2.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.106.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.110.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.112.10.3	04-Oct-2003	tron	Pull up revision 1.133 (requested by martti in ticket #1506): plug mbuf leak due to manual mbuf handling. PR kern/13807. (martti confirmed that it stabilizes the situation described in kren/13807)
1.112.10.2	29-Jul-2002	lukem	Pull up revision 1.114 (requested by enami in ticket #555): Synchronize code and comment again to prevent mbuf leak. Sprinkle some KNF while I'm here.
1.112.10.1	29-Jul-2002	lukem	Pull up revision 1.113 (requested by jaromir in ticket #555): Reduce stack usage on the NFS mount code path. This fixes kernel stack overflow when using IPsec on vax, as reported by Olaf Seibert on current-users@.
1.112.8.1	29-Aug-2002	gehenna	catch up with -current.
1.131.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.131.2.9	01-Apr-2005	skrll	Sync with HEAD.
1.131.2.8	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.131.2.7	17-Jan-2005	skrll	Sync with HEAD.
1.131.2.6	21-Sep-2004	skrll	Fix the sync with head I botched.
1.131.2.5	18-Sep-2004	skrll	Sync with HEAD.
1.131.2.4	25-Aug-2004	skrll	Sync with HEAD.
1.131.2.3	24-Aug-2004	skrll	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.131.2.2	03-Aug-2004	skrll	Sync with HEAD
1.131.2.1	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.
1.135.2.1	29-May-2004	tron	branches: 1.135.2.1.2; Pull up revision 1.139 (requested by atatat in ticket #393): Sysctl descriptions under vfs subtree
1.135.2.1.2.1	27-Oct-2005	riz	Pull up following revision(s) (requested by christos in ticket #5863): sys/nfs/nfs_subs.c: revision 1.152 via patch sys/nfs/nfs.h: revision 1.49 sys/nfs/nfs_vfsops.c: revision 1.149 via patch usr.sbin/amd/include/config.h: revision 1.36 sys/nfs/nfs_vnops.c: revision 1.227 via patch sys/nfs/nfsmount.h: revision 1.34 Allow the attribute cache to be turned off, and allow amd to do it.
1.144.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.144.2.1	29-Apr-2005	kent	sync with -current
1.145.2.1	27-Sep-2005	tron	Pull up following revision(s) (requested by christos in ticket #816): sys/nfs/nfs_vfsops.c: revision 1.149 sys/nfs/nfs_vnops.c: revision 1.227 ATTRTIMEO takes 2 args.
1.148.2.12	27-Feb-2008	yamt	revert incomplete nfs client locking for now.
1.148.2.11	27-Feb-2008	yamt	sync with head.
1.148.2.10	15-Feb-2008	yamt	- sprinkle some locks. - disable MNT_UPDATE because it involves too much locking headache. - don't overwrite other bits in v_vflags when setting VV_ROOT.
1.148.2.9	04-Feb-2008	yamt	sync with head.
1.148.2.8	21-Jan-2008	yamt	sync with head
1.148.2.7	07-Dec-2007	yamt	sync with head
1.148.2.6	15-Nov-2007	yamt	sync with head.
1.148.2.5	27-Oct-2007	yamt	sync with head.
1.148.2.4	03-Sep-2007	yamt	sync with head.
1.148.2.3	26-Feb-2007	yamt	sync with head.
1.148.2.2	30-Dec-2006	yamt	sync with head.
1.148.2.1	21-Jun-2006	yamt	sync with head.
1.151.6.3	01-Jun-2006	kardel	Sync with head.
1.151.6.2	22-Apr-2006	simonb	Sync with head.
1.151.6.1	04-Feb-2006	simonb	In the timecounter case, call tc_setclock() instead of setting time.tv_sec/tv_nsec directly.
1.151.4.1	09-Sep-2006	rpaulo	sync with head
1.151.2.1	01-Mar-2006	yamt	sync with head.
1.152.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.152.4.5	11-May-2006	elad	sync with head
1.152.4.4	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.
1.152.4.3	19-Apr-2006	elad	sync with head.
1.152.4.2	14-Apr-2006	elad	Store real/saved user/group ids too.
1.152.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.152.2.4	03-Sep-2006	yamt	sync with head.
1.152.2.3	11-Aug-2006	yamt	sync with head
1.152.2.2	26-Jun-2006	yamt	sync with head.
1.152.2.1	24-May-2006	yamt	sync with head.
1.155.2.1	19-Jun-2006	chap	Sync with head.
1.157.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.164.4.2	10-Dec-2006	yamt	sync with head.
1.164.4.1	22-Oct-2006	yamt	sync with head
1.164.2.3	01-Feb-2007	ad	Sync with head.
1.164.2.2	12-Jan-2007	ad	Sync with head.
1.164.2.1	18-Nov-2006	ad	Sync with head.
1.172.2.3	07-May-2007	yamt	sync with head.
1.172.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.172.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.174.4.1	11-Jul-2007	mjf	Sync with head.
1.174.2.10	25-Oct-2007	ad	Fix up mnt_vnodelist handling.
1.174.2.9	24-Oct-2007	ad	Do locking / use marker vnodes when traversing mountpoint vnode lists.
1.174.2.8	09-Oct-2007	ad	Sync with head.
1.174.2.7	16-Sep-2007	ad	Checkpoint work in progress on the vnode lifecycle and reference counting stuff. This makes it work properly without kernel_lock and fixes a few quite old bugs. See vfs_subr.c 1.283.2.17 for details.
1.174.2.6	20-Aug-2007	ad	Sync with HEAD.
1.174.2.5	15-Jul-2007	ad	Sync with head.
1.174.2.4	18-Jun-2007	yamt	fix merge botches.
1.174.2.3	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.
1.174.2.2	08-Jun-2007	ad	Sync with head.
1.174.2.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.179.2.2	10-Sep-2007	skrll	Sync with HEAD.
1.179.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.182.2.6	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.182.2.5	29-Oct-2007	joerg	Sync with HEAD.
1.182.2.4	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.
1.182.2.3	02-Oct-2007	joerg	Sync with HEAD.
1.182.2.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.182.2.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.183.2.2	05-Aug-2007	yamt	use kpause rather than lbolt.
1.183.2.1	05-Aug-2007	yamt	file nfs_vfsops.c was added on branch matt-mips64 on 2007-08-05 09:40:41 +0000
1.184.2.3	23-Mar-2008	matt	sync with HEAD
1.184.2.2	09-Jan-2008	matt	sync with HEAD
1.184.2.1	06-Nov-2007	matt	sync with HEAD
1.185.2.1	14-Oct-2007	yamt	sync with head.
1.186.2.1	13-Nov-2007	bouyer	Sync with HEAD
1.187.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.187.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.188.6.3	23-Jan-2008	bouyer	Sync with HEAD.
1.188.6.2	08-Jan-2008	bouyer	Sync with HEAD
1.188.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.188.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.196.10.8	09-Oct-2010	yamt	sync with head
1.196.10.7	11-Aug-2010	yamt	sync with head.
1.196.10.6	11-Mar-2010	yamt	sync with head
1.196.10.5	24-Jun-2009	yamt	lock vnode when calling VOP_GETATTR because there's no reasonable way for an implementation of VOP_GETATTR to prevent the vnode from being revoked.
1.196.10.4	24-Jun-2009	yamt	nfs_mount: re-enable MNT_UPDATE. it's broken as it is in trunk.
1.196.10.3	04-May-2009	yamt	sync with head.
1.196.10.2	16-May-2008	yamt	sync with head.
1.196.10.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.196.8.1	18-May-2008	yamt	sync with head.
1.196.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.196.6.2	05-Oct-2008	mjf	Sync with HEAD.
1.196.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.199.2.2	10-Oct-2008	skrll	Sync with HEAD.
1.199.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.200.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.200.4.1	19-Oct-2008	haad	Sync with HEAD.
1.203.14.1	28-Apr-2014	sborrill	Pull up the following revisions(s) (requested by maxv in ticket #1901): sys/kern/vfs_syscalls.c: revision 1.478, 1.480 via patch sys/coda/coda_vfsops.c: revision 1.81 sys/fs/adosfs/advfsops.c: revision 1.70 sys/fs/cd9660/cd9660_vfsops.c: revision 1.84 sys/fs/efs/efs_vfsops.c: revision 1.25 sys/fs/filecorefs/filecore_vfsops.c: revision 1.76 sys/fs/hfs/hfs_vfsops.c: revision 1.31 sys/fs/msdosfs/msdosfs_vfsops.c: revision 1.107 sys/fs/ntfs/ntfs_vfsops.c: revision 1.94 sys/fs/ptyfs/ptyfs_vfsops.c: revision 1.50 via patch sys/fs/puffs/puffs_vfsops.c: revision 1.110 via patch sys/fs/smbfs/smbfs_vfsops.c: revision 1.100 sys/fs/sysvbfs/sysvbfs_vfsops.c: revision 1.43 sys/fs/tmpfs/tmpfs_vfsops.c: revision 1.59 via patch sys/fs/udf/udf_vfsops.c: revision 1.67 sys/fs/union/union_vfsops.c: revision 1.72 sys/fs/unionfs/unionfs_vfsops.c: revision 1.13 sys/kern/vfs_syscalls.c: revision 1.479 sys/miscfs/nullfs/null_vfsops.c: revision 1.88 via patch sys/miscfs/overlay/overlay_vfsops.c: revision 1.61 sys/miscfs/procfs/procfs_vfsops.c: revision 1.91 sys/miscfs/umapfs/umap_vfsops.c: revision 1.92 sys/nfs/nfs_vfsops.c: revision 1.227 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.180 sys/ufs/ffs/ffs_vfsops.c: revision 1.297 sys/ufs/lfs/lfs_vfsops.c: revision 1.321 sys/ufs/mfs/mfs_vfsops.c: revision 1.107 Due to missing checks in the mount syscall, and a wrong assumption on the file systems side, the kernel could allocate an unbounded or zero-sized memory buffer, and could dereference a NULL pointer when particular arguments are given by a user.
1.203.10.1	28-Apr-2014	sborrill	Pull up the following revisions(s) (requested by maxv in ticket #1901): sys/kern/vfs_syscalls.c: revision 1.478, 1.480 via patch sys/coda/coda_vfsops.c: revision 1.81 sys/fs/adosfs/advfsops.c: revision 1.70 sys/fs/cd9660/cd9660_vfsops.c: revision 1.84 sys/fs/efs/efs_vfsops.c: revision 1.25 sys/fs/filecorefs/filecore_vfsops.c: revision 1.76 sys/fs/hfs/hfs_vfsops.c: revision 1.31 sys/fs/msdosfs/msdosfs_vfsops.c: revision 1.107 sys/fs/ntfs/ntfs_vfsops.c: revision 1.94 sys/fs/ptyfs/ptyfs_vfsops.c: revision 1.50 via patch sys/fs/puffs/puffs_vfsops.c: revision 1.110 via patch sys/fs/smbfs/smbfs_vfsops.c: revision 1.100 sys/fs/sysvbfs/sysvbfs_vfsops.c: revision 1.43 sys/fs/tmpfs/tmpfs_vfsops.c: revision 1.59 via patch sys/fs/udf/udf_vfsops.c: revision 1.67 sys/fs/union/union_vfsops.c: revision 1.72 sys/fs/unionfs/unionfs_vfsops.c: revision 1.13 sys/kern/vfs_syscalls.c: revision 1.479 sys/miscfs/nullfs/null_vfsops.c: revision 1.88 via patch sys/miscfs/overlay/overlay_vfsops.c: revision 1.61 sys/miscfs/procfs/procfs_vfsops.c: revision 1.91 sys/miscfs/umapfs/umap_vfsops.c: revision 1.92 sys/nfs/nfs_vfsops.c: revision 1.227 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.180 sys/ufs/ffs/ffs_vfsops.c: revision 1.297 sys/ufs/lfs/lfs_vfsops.c: revision 1.321 sys/ufs/mfs/mfs_vfsops.c: revision 1.107 Due to missing checks in the mount syscall, and a wrong assumption on the file systems side, the kernel could allocate an unbounded or zero-sized memory buffer, and could dereference a NULL pointer when particular arguments are given by a user.
1.203.4.1	25-Apr-2014	sborrill	Pull up the following revisions(s) (requested by maxv in ticket #1901): sys/kern/vfs_syscalls.c: revision 1.478, 1.480 via patch sys/coda/coda_vfsops.c: revision 1.81 sys/fs/adosfs/advfsops.c: revision 1.70 sys/fs/cd9660/cd9660_vfsops.c: revision 1.84 sys/fs/efs/efs_vfsops.c: revision 1.25 sys/fs/filecorefs/filecore_vfsops.c: revision 1.76 sys/fs/hfs/hfs_vfsops.c: revision 1.31 sys/fs/msdosfs/msdosfs_vfsops.c: revision 1.107 sys/fs/ntfs/ntfs_vfsops.c: revision 1.94 sys/fs/ptyfs/ptyfs_vfsops.c: revision 1.50 via patch sys/fs/puffs/puffs_vfsops.c: revision 1.110 via patch sys/fs/smbfs/smbfs_vfsops.c: revision 1.100 sys/fs/sysvbfs/sysvbfs_vfsops.c: revision 1.43 sys/fs/tmpfs/tmpfs_vfsops.c: revision 1.59 via patch sys/fs/udf/udf_vfsops.c: revision 1.67 sys/fs/union/union_vfsops.c: revision 1.72 sys/fs/unionfs/unionfs_vfsops.c: revision 1.13 sys/kern/vfs_syscalls.c: revision 1.479 sys/miscfs/nullfs/null_vfsops.c: revision 1.88 via patch sys/miscfs/overlay/overlay_vfsops.c: revision 1.61 sys/miscfs/procfs/procfs_vfsops.c: revision 1.91 sys/miscfs/umapfs/umap_vfsops.c: revision 1.92 sys/nfs/nfs_vfsops.c: revision 1.227 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.180 sys/ufs/ffs/ffs_vfsops.c: revision 1.297 sys/ufs/lfs/lfs_vfsops.c: revision 1.321 sys/ufs/mfs/mfs_vfsops.c: revision 1.107 Due to missing checks in the mount syscall, and a wrong assumption on the file systems side, the kernel could allocate an unbounded or zero-sized memory buffer, and could dereference a NULL pointer when particular arguments are given by a user.
1.203.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.203.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.206.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.210.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.210.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.211.2.4	05-Mar-2011	rmind	sync with head
1.211.2.3	03-Jul-2010	rmind	sync with head
1.211.2.2	30-May-2010	rmind	sync with head
1.211.2.1	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.
1.217.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.220.16.1	21-Apr-2014	bouyer	Pull up following revision(s) (requested by maxv in ticket #1050): sys/ufs/chfs/chfs_vfsops.c: revision 1.11 sys/fs/unionfs/unionfs_vfsops.c: revision 1.13 sys/fs/nilfs/nilfs_vfsops.c: revision 1.16 sys/ufs/mfs/mfs_vfsops.c: revision 1.107 sys/fs/sysvbfs/sysvbfs_vfsops.c: revision 1.43 sys/ufs/ffs/ffs_vfsops.c: revision 1.297 sys/kern/vfs_syscalls.c: revision 1.478 sys/kern/vfs_syscalls.c: revision 1.479 sys/fs/puffs/puffs_vfsops.c: revision 1.110 sys/fs/cd9660/cd9660_vfsops.c: revision 1.84 sys/nfs/nfs_vfsops.c: revision 1.227 sys/fs/v7fs/v7fs_vfsops.c: revision 1.10 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.180 sys/miscfs/umapfs/umap_vfsops.c: revision 1.92 sys/fs/filecorefs/filecore_vfsops.c: revision 1.76 sys/miscfs/nullfs/null_vfsops.c: revision 1.88 sys/fs/ptyfs/ptyfs_vfsops.c: revision 1.50 sys/coda/coda_vfsops.c: revision 1.81 sys/ufs/lfs/lfs_vfsops.c: revision 1.321 sys/fs/tmpfs/tmpfs_vfsops.c: revision 1.59 sys/fs/hfs/hfs_vfsops.c: revision 1.31 sys/miscfs/overlay/overlay_vfsops.c: revision 1.61 sys/fs/union/union_vfsops.c: revision 1.72 sys/fs/ntfs/ntfs_vfsops.c: revision 1.94 sys/kern/vfs_syscalls.c: revision 1.480 sys/fs/efs/efs_vfsops.c: revision 1.25 sys/kern/vfs_syscalls.c: revision 1.482 sys/fs/msdosfs/msdosfs_vfsops.c: revision 1.107 external/cddl/osnet/dist/uts/common/fs/zfs/zfs_vfsops.c: revision 1.12 sys/miscfs/procfs/procfs_vfsops.c: revision 1.91 sys/fs/smbfs/smbfs_vfsops.c: revision 1.100 sys/fs/adosfs/advfsops.c: revision 1.70 sys/fs/udf/udf_vfsops.c: revision 1.67 Limit check for 'data_len'. Otherwise a (un)privileged user can easily panic the system by passing a huge size. ok christos@ An (un)privileged user can easily make the kernel dereference a NULL pointer. The kernel allows 'data' to be NULL; it's the fs's responsibility to ensure that it isn't NULL (if the fs actually needs data). ok christos@ Some fs's - like kernfs - set their vfs_min_mount_data to zero. Add a check to prevent an (un)privileged user from requesting a zero-sized allocation (and thus a panic). This thing is totally buggy: 'data_len' is modified by the fs, so calling kmem_free with it while its value has changed since the kmem_alloc is far from being a good idea. If the kernel figures out that something mismatches, it will panic (typically with kernfs).
1.220.14.1	21-Apr-2014	bouyer	Pull up following revision(s) (requested by maxv in ticket #1050): sys/ufs/chfs/chfs_vfsops.c: revision 1.11 sys/fs/unionfs/unionfs_vfsops.c: revision 1.13 sys/fs/nilfs/nilfs_vfsops.c: revision 1.16 sys/ufs/mfs/mfs_vfsops.c: revision 1.107 sys/fs/sysvbfs/sysvbfs_vfsops.c: revision 1.43 sys/ufs/ffs/ffs_vfsops.c: revision 1.297 sys/kern/vfs_syscalls.c: revision 1.478 sys/kern/vfs_syscalls.c: revision 1.479 sys/fs/puffs/puffs_vfsops.c: revision 1.110 sys/fs/cd9660/cd9660_vfsops.c: revision 1.84 sys/nfs/nfs_vfsops.c: revision 1.227 sys/fs/v7fs/v7fs_vfsops.c: revision 1.10 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.180 sys/miscfs/umapfs/umap_vfsops.c: revision 1.92 sys/fs/filecorefs/filecore_vfsops.c: revision 1.76 sys/miscfs/nullfs/null_vfsops.c: revision 1.88 sys/fs/ptyfs/ptyfs_vfsops.c: revision 1.50 sys/coda/coda_vfsops.c: revision 1.81 sys/ufs/lfs/lfs_vfsops.c: revision 1.321 sys/fs/tmpfs/tmpfs_vfsops.c: revision 1.59 sys/fs/hfs/hfs_vfsops.c: revision 1.31 sys/miscfs/overlay/overlay_vfsops.c: revision 1.61 sys/fs/union/union_vfsops.c: revision 1.72 sys/fs/ntfs/ntfs_vfsops.c: revision 1.94 sys/kern/vfs_syscalls.c: revision 1.480 sys/fs/efs/efs_vfsops.c: revision 1.25 sys/kern/vfs_syscalls.c: revision 1.482 sys/fs/msdosfs/msdosfs_vfsops.c: revision 1.107 external/cddl/osnet/dist/uts/common/fs/zfs/zfs_vfsops.c: revision 1.12 sys/miscfs/procfs/procfs_vfsops.c: revision 1.91 sys/fs/smbfs/smbfs_vfsops.c: revision 1.100 sys/fs/adosfs/advfsops.c: revision 1.70 sys/fs/udf/udf_vfsops.c: revision 1.67 Limit check for 'data_len'. Otherwise a (un)privileged user can easily panic the system by passing a huge size. ok christos@ An (un)privileged user can easily make the kernel dereference a NULL pointer. The kernel allows 'data' to be NULL; it's the fs's responsibility to ensure that it isn't NULL (if the fs actually needs data). ok christos@ Some fs's - like kernfs - set their vfs_min_mount_data to zero. Add a check to prevent an (un)privileged user from requesting a zero-sized allocation (and thus a panic). This thing is totally buggy: 'data_len' is modified by the fs, so calling kmem_free with it while its value has changed since the kmem_alloc is far from being a good idea. If the kernel figures out that something mismatches, it will panic (typically with kernfs).
1.220.12.3	03-Dec-2017	jdolecek	update from HEAD
1.220.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.220.12.1	25-Feb-2013	tls	resync with head
1.220.8.1	21-Apr-2014	bouyer	Pull up following revision(s) (requested by maxv in ticket #1050): sys/ufs/chfs/chfs_vfsops.c: revision 1.11 sys/fs/unionfs/unionfs_vfsops.c: revision 1.13 sys/fs/nilfs/nilfs_vfsops.c: revision 1.16 sys/ufs/mfs/mfs_vfsops.c: revision 1.107 sys/fs/sysvbfs/sysvbfs_vfsops.c: revision 1.43 sys/ufs/ffs/ffs_vfsops.c: revision 1.297 sys/kern/vfs_syscalls.c: revision 1.478 sys/kern/vfs_syscalls.c: revision 1.479 sys/fs/puffs/puffs_vfsops.c: revision 1.110 sys/fs/cd9660/cd9660_vfsops.c: revision 1.84 sys/nfs/nfs_vfsops.c: revision 1.227 sys/fs/v7fs/v7fs_vfsops.c: revision 1.10 sys/ufs/ext2fs/ext2fs_vfsops.c: revision 1.180 sys/miscfs/umapfs/umap_vfsops.c: revision 1.92 sys/fs/filecorefs/filecore_vfsops.c: revision 1.76 sys/miscfs/nullfs/null_vfsops.c: revision 1.88 sys/fs/ptyfs/ptyfs_vfsops.c: revision 1.50 sys/coda/coda_vfsops.c: revision 1.81 sys/ufs/lfs/lfs_vfsops.c: revision 1.321 sys/fs/tmpfs/tmpfs_vfsops.c: revision 1.59 sys/fs/hfs/hfs_vfsops.c: revision 1.31 sys/miscfs/overlay/overlay_vfsops.c: revision 1.61 sys/fs/union/union_vfsops.c: revision 1.72 sys/fs/ntfs/ntfs_vfsops.c: revision 1.94 sys/kern/vfs_syscalls.c: revision 1.480 sys/fs/efs/efs_vfsops.c: revision 1.25 sys/kern/vfs_syscalls.c: revision 1.482 sys/fs/msdosfs/msdosfs_vfsops.c: revision 1.107 external/cddl/osnet/dist/uts/common/fs/zfs/zfs_vfsops.c: revision 1.12 sys/miscfs/procfs/procfs_vfsops.c: revision 1.91 sys/fs/smbfs/smbfs_vfsops.c: revision 1.100 sys/fs/adosfs/advfsops.c: revision 1.70 sys/fs/udf/udf_vfsops.c: revision 1.67 Limit check for 'data_len'. Otherwise a (un)privileged user can easily panic the system by passing a huge size. ok christos@ An (un)privileged user can easily make the kernel dereference a NULL pointer. The kernel allows 'data' to be NULL; it's the fs's responsibility to ensure that it isn't NULL (if the fs actually needs data). ok christos@ Some fs's - like kernfs - set their vfs_min_mount_data to zero. Add a check to prevent an (un)privileged user from requesting a zero-sized allocation (and thus a panic). This thing is totally buggy: 'data_len' is modified by the fs, so calling kmem_free with it while its value has changed since the kmem_alloc is far from being a good idea. If the kernel figures out that something mismatches, it will panic (typically with kernfs).
1.220.2.2	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")
1.220.2.1	23-Jan-2013	yamt	sync with head
1.221.2.1	18-May-2014	rmind	sync with head
1.226.2.1	10-Aug-2014	tls	Rebase.
1.229.4.3	28-Aug-2017	skrll	Sync with HEAD
1.229.4.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.229.4.1	22-Sep-2015	skrll	Sync with HEAD
1.229.2.2	08-Nov-2015	riz	Pull up following revision(s) (requested by pgoyette in ticket #1021): sys/nfs/nfs_vfsops.c: revision 1.231 Don't forget to call nfs_fini() when we're finished. Without this, we leave a dangling pool nfsrvdescpl around.
1.229.2.1	04-Nov-2015	riz	Pull up following revision(s) (requested by manu in ticket #882): sbin/umount/umount.c: revision 1.48 sys/nfs/nfsmount.h: revision 1.53 sys/nfs/nfs_var.h: revision 1.94 sys/nfs/nfs_iod.c: revision 1.7 sys/nfs/nfs_socket.c: revision 1.197 sys/nfs/nfs_bio.c: revision 1.191 sys/nfs/nfs_vfsops.c: revision 1.230 sys/nfs/nfs_clntsocket.c: revision 1.3 Remove useless and harmful sync(2) call in umount(8) Remove sync(2) call before unmount(2) in umount(8). This sync(2) is useless since unmount(2) will perform a VFS_SYNC anyway. But moreover, this sync(2) may be harmful, as there are some situation where it cannot return (unreachable NFS server, for instance), causing umount -f to be uneffective. Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.231.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.231.2.2	26-Apr-2017	pgoyette	Sync with HEAD
1.231.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.235.10.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.235.10.1	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.236.2.3	21-Apr-2020	martin	Sync with HEAD
1.236.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.236.2.1	10-Jun-2019	christos	Sync with HEAD
1.237.6.2	29-Feb-2020	ad	Sync with head.
1.237.6.1	17-Jan-2020	ad	Sync with head.
1.237.4.1	04-May-2022	martin	Pull up following revision(s) (requested by gavan in ticket #1441): sys/nfs/nfs_vfsops.c: revision 1.243 Don't pretend that files are limited to 1TB on NFSv3.
1.240.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.241.4.1	03-Apr-2021	thorpej	Sync with HEAD.
1.241.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.242.2.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.243.10.2	20-Sep-2024	martin	Pull up following revision(s) (requested by rin in ticket #880): sys/nfs/nfs_iod.c: revision 1.9 sys/nfs/nfs_vfsops.c: revision 1.245 sys/nfs/nfs_clntsubs.c: revision 1.7 PR/57279: Izumi Tsutsui: Fix some {int,long} -> time_t. Still things will break eventually because parts of the nfs protocol assume time_t will fit in 32 bits.
1.243.10.1	20-Sep-2024	martin	Pull up following revision(s) (requested by rin in ticket #879): sys/nfs/nfs_vfsops.c: revision 1.244 Avoid overflow of nfs_commitsize on machines with > 32GB RAM.
1.325	10-Dec-2023	schmonz	NFS client: fix interop with macOS 14 servers. Symptom: a bunch of "Cannot open `.' (Invalid argument)". thorpej@ analysis and fix: on the first request to read a given directory, make sure READDIR and READDIRPLUS cookie verifiers are being set to 0. This is in RFC1813 and macOS must have gotten stricter about it. Verified on 10.0_RC1/aarch64 to fix the reproducers in PR kern/57691 as well as the original use case in which I met the bug: pkg_rr once again runs to completion.
1.324	24-May-2022	andvar	branches: 1.324.4; fix various typos in comments, docs and log messages.
1.323	30-Mar-2022	christos	restructure so we abort/unlock properly on failure.
1.322	27-Mar-2022	christos	add a kauth vnode check for creating links
1.321	20-Oct-2021	thorpej	Overhaul of the EVFILT_VNODE kevent(2) filter: - Centralize vnode kevent handling in the VOP_() wrappers, rather than forcing each individual file system to deal with it (except VOP_RENAME(), because VOP_RENAME() is a mess and we currently have 2 different ways of handling it; at least it's reasonably well-centralized in the "new" way). - Add support for NOTE_OPEN, NOTE_CLOSE, NOTE_CLOSE_WRITE, and NOTE_READ, compatible with the same events in FreeBSD. - Track which kevent notifications clients are interested in receiving to avoid doing work for events no one cares about (avoiding, e.g. taking locks and traversing the klist to send a NOTE_WRITE when someone is merely watching for a file to be deleted, for example). In support of the above: - Add support in vnode_if.sh for specifying PRE- and POST-op handlers, to be invoked before and after vop_pre() and vop_post(), respectively. Basic idea from FreeBSD, but implemented differently. - Add support in vnode_if.sh for specifying CONTEXT fields in the vop__args structures. These context fields are used to convey information between the file system VOP function and the VOP wrapper, but do not occupy an argument slot in the VOP_*() call itself. These context fields are initialized and subsequently interpreted by PRE- and POST-op handlers. - Version VOP_REMOVE(), uses the a context field for the file system to report back the resulting link count of the target vnode. Return this in tmpfs, udf, nfs, chfs, ext2fs, lfs, and ufs. NetBSD 9.99.92.
1.320	18-Jul-2021	dholland	Abolish all the silly indirection macros for initializing vnode ops tables. These are things of the form #define foofs_op genfs_op, or #define foofs_op genfs_eopnotsupp, or similar. They serve no purpose besides obfuscation, and have gotten cutpasted all over everywhere.
1.319	18-Jul-2021	dholland	Use macros for the canned parts of device and fifo vnode op tables. Add GENFS_SPECOP_ENTRIES and GENFS_FIFOOP_ENTRIES macros that contain the portion of the vnode ops table declaration that is (conservatively) the same in every fs. Use these in every fs that supports devices and/or fifos with separate ops tables. Note that ptyfs works differently (it has one type of vnode with open-coded dispatch to the specfs code, which I haven't changed in this commit) and rump/librump/rumpvfs/rumpfs.c has an indirect dynamic dispatch that already does more or less the same thing, which I also haven't changed. Also note that this anticipates a few bits in the next changeset here and there, and adds missing but unreachable calls in some cases (e.g. most fses weren't defining whiteout on devices and fifos, but it isn't reachable there), and it changes parsepath on devices and fifos to genfs_badop from genfs_parsepath (but it's not reachable there either). It appears that devices in kernfs were missing kqfilter, so it's possible that if you try to use kqueue on /kern/rootdev that it'll explode. And finally note that the ops declaration tables aren't order-dependent. (Other than vop_default_desc has to come first.) Otherwise this wouldn't work.
1.318	29-Jun-2021	dholland	- Add a new vnode op: VOP_PARSEPATH. - Move namei_getcomponent to genfs_vnops.c and call it genfs_parsepath. - Add a parsepath entry to every vnode ops table. VOP_PARSEPATH takes a directory vnode to be searched and a complete following path and chooses how much of that path to consume. To begin with, all parsepath calls are genfs_parsepath, which locates the first '/' as always. Note that the call doesn't take the whole struct componentname, only the string. The other bits of struct componentname should not be needed and there's no reason to cause potential complications by exposing them.
1.317	05-Sep-2020	riastradh	branches: 1.317.6; Round of uvm.h cleanup. The poorly named uvm.h is generally supposed to be for uvm-internal users only. - Narrow it to files that actually need it -- mostly files that need to query whether curlwp is the pagedaemon, which should maybe be exposed by an external header. - Use uvm_extern.h where feasible and uvm_*.h for things not exposed by it. We should split up uvm_extern.h but this will serve for now to reduce the uvm.h dependencies. - Use uvm_stat.h and #ifdef UVMHIST uvm.h for files that use UVMHIST(ubchist), since ubchist is declared in uvm.h but the reference evaporates if UVMHIST is not defined, so we reduce header file dependencies. - Make uvm_device.h and uvm_swap.h independently includable while here. ok chs@
1.316	27-Jun-2020	christos	Introduce genfs_pathconf() and use it for the default case in all filesystems.
1.315	16-May-2020	christos	Add ACL support for FFS. From FreeBSD.
1.314	13-Apr-2020	ad	Replace most uses of vp->v_usecount with a call to vrefcnt(vp), a function that hides the details and does atomic_load_relaxed(). Signature matches FreeBSD.
1.313	23-Feb-2020	ad	branches: 1.313.4; UVM locking changes, proposed on tech-kern: - Change the lock on uvm_object, vm_amap and vm_anon to be a RW lock. - Break v_interlock and vmobjlock apart. v_interlock remains a mutex. - Do partial PV list locking in the x86 pmap. Others to follow later.
1.312	10-Sep-2019	christos	branches: 1.312.2; remove NCHNAMLEN optimization
1.311	03-Sep-2018	riastradh	branches: 1.311.4; 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!)
1.310	26-Apr-2017	riastradh	branches: 1.310.4; 1.310.10; 1.310.12; Change VOP_REMOVE and VOP_RMDIR to preserve lock/ref on dvp. No change to vp -- the plan is to replace the node by the componentname in the vop parameters, and let all directory vops do lookups internally. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2017/04/17/msg021825.html
1.309	19-Jan-2016	hannken	Return an error if NFSPROC_LOOKUP returns the file handle of the current directory. Treating it as DOT lookup would put garbage into the name cache and could panic on future lookups. Seen with ZFS file system exported from OmniOS, an OpenSolaris derivative. Fixes PR kern/50664 "cd .." over NFS/ZFS can panic kernel
1.308	14-May-2015	chs	in nfs_writerpc(), avoid a signed/unsigned problem in computing the number of bytes to back up in the uio when we need to resend a write RPC (eg. after a server crash) on a 64-bit platform. should fix PR 35448.
1.307	20-Apr-2015	riastradh	Make VOP_LINK return directory still locked and referenced. Ride 7.99.10 bump.
1.306	25-Jul-2014	dholland	branches: 1.306.2; 1.306.4; Add VOP_FALLOCATE and VOP_FDISCARD to every vnode ops table I can find. The filesystem ones all call genfs_eopnotsupp - right now I am only implementing the plumbing and we can implement fallocate and/or fdiscard for files later. The device ones call spec_fallocate (which is also genfs_eopnotsupp) and spec_fdiscard, which dispatches to the device-level op. The fifo ones all call vn_fifo_bypass, which also ends up being EOPNOTSUPP.
1.305	05-Jul-2014	hannken	Use vcache_rekey_* for nfs_lookitup() in the "*npp != NULL" case.
1.304	07-Feb-2014	hannken	branches: 1.304.2; Change vnode operation lookup to return the resulting vnode *vpp unlocked. Change cache_lookup() to return an unlocked vnode. Discussed on tech-kern@ Welcome to 6.99.31
1.303	23-Jan-2014	hannken	Change vnode operations create, mknod, mkdir and symlink to return the resulting vnode *vpp unlocked. Discussed on tech-kern@ Welcome to 6.99.30
1.302	17-Jan-2014	hannken	Change vnode operations create, mknod, mkdir and symlink to keep the directory node dvp locked on return. Discussed on tech-kern@ Welcome to 6.99.29
1.301	15-Nov-2013	nisimura	add one more __unused attribute to shut gcc4.8 off.
1.300	14-Sep-2013	martin	Backout wildcard pragma to kill warnings and instead sprinkle a few dozen __unused attributes. Requested by joerg@
1.299	18-Mar-2013	plunky	branches: 1.299.6; C99 section 6.7.2.3 (Tags) Note 3 states that: A type specifier of the form enum identifier without an enumerator list shall only appear after the type it specifies is complete. which means that we cannot pass an "enum vtype" argument to kauth_access_action() without fully specifying the type first. Unfortunately there is a complicated include file loop which makes that difficult, so convert this minimal function into a macro (and capitalize it). (ok elad@)
1.298	07-Nov-2012	macallan	fix crash in nfs client lookups, dholland says 'my fault'
1.297	05-Nov-2012	dholland	Excise struct componentname from the namecache. This uglifies the interface, because several operations need to be passed the namei flags and cache_lookup also needs for the time being to be passed cnp->cn_nameiop. Nonetheless, it's a net benefit. The glop should be able to go away eventually but requires structural cleanup elsewhere first. This change requires a kernel bump.
1.296	05-Nov-2012	dholland	Disentangle the namecache from the internals of namei. - Move the namecache's hash computation to inside the namecache code, instead of being spread out all over the place. Remove cn_hash from struct componentname and delete all uses of it. - It is no longer necessary (if it ever was) for cache_lookup and cache_lookup_raw to clear MAKEENTRY from cnp->cn_flags for the cases that cache_enter already checks for. - Rearrange the interface of cache_lookup (and cache_lookup_raw) to make it somewhat simpler, to exclude certain nonexistent error conditions, and (most importantly) to make it not require write access to cnp->cn_flags. This change requires a kernel bump.
1.295	22-Jul-2012	rmind	branches: 1.295.2; Move some the test for MAKEENTRY into the cache_enter(9). Make some variables in vfs_cache.c static, __read_mostly, etc. No objection on tech-kern@.
1.294	27-Apr-2012	drochner	fix access permission check which got broken by some kauth rework in March, affected mostly systems with NFS root fs
1.293	28-Nov-2011	tls	branches: 1.293.2; 1.293.4; Remove arc4random() and arc4randbytes() from the kernel API. Replace arc4random() hacks in rump with stubs that call the host arc4random() to get numbers that are hopefully actually random (arc4random() keyed with stack junk is not). This should fix some of the currently failing anita tests -- we should no longer generate duplicate "random" MAC addresses in the test environment.
1.292	27-Sep-2011	christos	branches: 1.292.2; use NFS_MAXPATHLEN instead of MAXPATHLEN
1.291	12-Jun-2011	rmind	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.
1.290	24-Apr-2011	rmind	branches: 1.290.2; sys_link: prevent hard links on directories (cross-mount operations are already prevented). File systems are no longer responsible to check this. Clean up and add asserts (note that dvp == vp cannot happen in vop_link). OK dholland@
1.289	14-Dec-2010	cegger	branches: 1.289.2; Initialize mutex and cv after sanity checks
1.288	14-Dec-2010	cegger	back out rev. 1.285. The problem I try to hunt down in PR 42455 is not in the network stack as shown by PR 44206.
1.287	30-Nov-2010	dholland	Abolish the SAVENAME and HASBUF flags. There is now always a buffer, so the path in a struct componentname is now always valid during VOP calls.
1.286	30-Nov-2010	dholland	Abolish struct componentname's cn_pnbuf. Use the path buffer in the pathbuf object passed to namei as work space instead. (For now a pnbuf pointer appears in struct nameidata, to support certain unclean things that haven't been fixed yet, but it will be going away in the future.) This removes the need for the SAVENAME and HASBUF namei flags.
1.285	26-Oct-2010	cegger	Add diagnostic check which hits when PR 42455 is reproduced. Idea from hans@
1.284	24-Jun-2010	hannken	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.
1.283	29-Mar-2010	pooka	Stop exposing fifofs internals and leave only fifo_vnodeop_p visible.
1.282	08-Jan-2010	pooka	branches: 1.282.2; 1.282.4; The VATTR_NULL/VREF/VHOLD/HOLDRELE() macros lost their will to live years ago when the kernel was modified to not alter ABI based on DIAGNOSTIC, and now just call the respective function interfaces (in lowercase). Plenty of mix'n match upper/lowercase has creeped into the tree since then. Nuke the macros and convert all callsites to lowercase. no functional change
1.281	21-Oct-2009	rmind	Remove uarea swap-out functionality: - Addresses the issue described in PR/38828. - Some simplification in threading and sleepq subsystems. - Eliminates pmap_collect() and, as a side note, allows pmap optimisations. - Eliminates XS_CTL_DATA_ONSTACK in scsipi code. - Avoids few scans on LWP list and thus potentially long holds of proc_lock. - Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k. - Removes __SWAP_BROKEN cases. Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on acorn26 (thanks to <bjh21>). Discussed on <tech-kern>, reviewed by <ad>.
1.280	14-Jul-2009	apb	Use pid_t, not short, for a pid. Part of PR 41255 from Kurt Lidl.
1.279	23-Jun-2009	elad	Move the implementation of vaccess() to genfs_can_access(), in line with the other routines of the same spirit. Adjust file-system code to use it. Keep vaccess() for KPI compatibility and to keep element of least surprise. A "diagnostic" message warning that vaccess() is deprecated will be printed when it's used (obviously, only in DIAGNOSTIC kernels). No objections on tech-kern@: http://mail-index.netbsd.org/tech-kern/2009/06/21/msg005310.html
1.278	10-May-2009	yamt	nfs_lookup: vn_lock the vnode returned by cache_lookup_raw before feeding it to VOP_GETATTR. it's necessary because the vnode might be being cleaned by getcleanvnode. it's an instance of more general races between vnode reclaim and unlocked VOPs. however, this one happens somewhat often because it can be triggered by getnewvnode rather than revoke.
1.277	10-May-2009	yamt	restore lines, esp. a vrele() call, which i mistakenly removed in the previous. (rev.1.276)
1.276	04-May-2009	yamt	nfs_lookup: handle the case where the vnode returned cache_lookup_raw is being reclaimed by another thread. after recent changes in cache_lookup_raw, there's a race between cache_lookup_raw/vtryget and getcleanvnode/vclean. PR/41028.
1.275	04-May-2009	yamt	nfs_lookup: add an assertion.
1.274	04-May-2009	yamt	nfs_lookup: comments. no functional changes.
1.273	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.272	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.271	14-Mar-2009	dsl	Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.270	13-Mar-2009	yamt	nfs_lookup: fix a comment.
1.269	11-Jan-2009	christos	branches: 1.269.2; merge christos-time_t
1.268	19-Nov-2008	ad	branches: 1.268.4; Make the emulations, exec formats, coredump, NFS, and the NFS server into modules. By and large this commit: - shuffles header files and ifdefs - splits code out where necessary to be modular - adds module glue for each of the components - adds/replaces hooks for things that can be installed at runtime
1.267	15-Oct-2008	pooka	branches: 1.267.2; For NFSV3CREATE_EXCLUSIVE verifier, just use arc4random() instead of the first inet address on INET systems (which is likely to be localhost).
1.266	13-Feb-2008	yamt	branches: 1.266.6; 1.266.10; 1.266.16; reject files larger than nm_maxfilesize.
1.265	25-Jan-2008	ad	Remove VOP_LEASE. Discussed on tech-kern.
1.264	02-Jan-2008	yamt	use kmem_alloc instead of malloc.
1.263	02-Jan-2008	ad	Merge vmlocking2 to head.
1.262	17-Dec-2007	yamt	nfs_create: try GUARDED if EXCLUSIVE is NOTSUPP.
1.261	08-Dec-2007	pooka	branches: 1.261.4; Remove cn_lwp from struct componentname. curlwp should be used from on. The NDINIT() macro no longer takes the lwp parameter and associates the credentials of the calling thread with the namei structure.
1.260	26-Nov-2007	pooka	branches: 1.260.2; 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
1.259	13-Nov-2007	yamt	nfs_lookup: fix indent.
1.258	07-Nov-2007	ad	Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.257	28-Oct-2007	yamt	branches: 1.257.2; make NFS_ATTRTIMEO a function.
1.256	09-Jul-2007	ad	branches: 1.256.6; 1.256.8; 1.256.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.255	29-Apr-2007	yamt	don't forget to destroy mutex and condvar.
1.254	29-Apr-2007	yamt	use mutex and condver.
1.253	29-Apr-2007	yamt	use mutex and condvar.
1.252	04-Mar-2007	christos	branches: 1.252.2; 1.252.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.251	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.250	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.249	24-Jan-2007	hubertf	branches: 1.249.2; Remove duplicate #includes, patch contributed in private mail by Slava Semushin <slava.semushin@gmail.com>. To verify that no nasty side effects of duplicate includes (or their removal) have an effect here, I've compiled an i386/ALL kernel with and without the patch, and the only difference in the resulting .o files was in shifted line numbers in some assert() calls. The comparison of the .o files was based on the output of "objdump -D". Thanks to martin@ for the input on testing.
1.248	27-Dec-2006	yamt	- remove the rest of nqnfs. - reject NFSMNT_MNTD and NFSMNT_KERB. (no users in tree.)
1.247	27-Dec-2006	yamt	remove nqnfs.
1.246	09-Dec-2006	chs	a smorgasbord of improvements to vnode locking and path lookup: - LOCKPARENT is no longer relevant for lookup(), relookup() or VOP_LOOKUP(). these now always return the parent vnode locked. namei() works as before. lookup() and various other paths no longer acquire vnode locks in the wrong order via vrele(). fixes PR 32535. as a nice side effect, path lookup is also up to 25% faster. - the above allows us to get rid of PDIRUNLOCK. - also get rid of WANTPARENT (just use LOCKPARENT and unlock it). - remove an assumption in layer_node_find() that all file systems implement a recursive VOP_LOCK() (unionfs doesn't). - require that all file systems supply vfs_vptofh and vfs_fhtovp routines. fill in eopnotsupp() for file systems that don't support being exported and remove the checks for NULL. (layerfs calls these without checking.) - in union_lookup1(), don't change refcounts in the ISDOTDOT case, just adjust which vnode is locked. fixes PR 33374. - apply fixes for ufs_rename() from ufs_vnops.c rev. 1.61 to ext2fs_rename().
1.245	09-Nov-2006	yamt	branches: 1.245.2; remove some __unused in function parameters.
1.244	14-Oct-2006	yamt	grab glock when calling uvm_unp_setsize, so that it doesn't interfere mmap'ed accesses. this fixes an assertion failure in in nfs_doio_read. ("vp->v_size >= uiop->uio_offset + uiop->uio_resid")
1.243	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.242	29-Sep-2006	drochner	Flush regular files before setattr also if the mode bits are going to be set. Linux NFS servers (at least) reset suid/sgid bits if a write happens afterwards. Add a comment why this is done. This fixes system builds on diskless systems for me where suid bits were missing after install(1). Approved by yamt.
1.241	23-Jul-2006	ad	branches: 1.241.4; 1.241.6; Use the LWP cached credentials where sane.
1.240	01-Jul-2006	yamt	some comments taken from Jed Davis's patch.
1.239	01-Jul-2006	yamt	if a file is sillyrename'ed because it's a destination of rename, make sillyrename (try to) use LINK operation rather than RENAME. PR/33861 from Jed Davis. he provided the almost same patch. according to him, it also happen to be what opensolaris does in this case. from the PR: > In nfs_rename(), if the destination appears to exist and is "in use" > (this check is apparently satisfied even if the file isn't in use by > anything except the rename itself), it will sillyrename it, then delete > the sillyrenamed file even if the rename fails -- for instance, because > the "from" file no longer exists on the server. > mkdir a b; touch a/x; perl -e 'fork(); rename("a/x","b/x") or die "$!\n"' > > Afterwards, neither a/x nor b/x will exist. > 1) Lookup of b/x; fails with NOENT. > 2) Rename from a/x to b/x; succeeds. > 3) Lookup of b/x; fails with NOENT. > 4) Rename from b/x to b/.nfsA23a3; succeeds. > 5) Rename from a/x to b/x; fails with NOENT. > 6) Remove of b/.nfsA23a3; succeeds.
1.238	30-Jun-2006	yamt	fix handling of NFSERR_NOTSUPP and NFSERR_BAD_COOKIE, which have been broken since nfs_socket.c rev.1.115.
1.237	07-Jun-2006	kardel	branches: 1.237.2; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.236	14-May-2006	elad	branches: 1.236.2; integrate kauth.
1.235	15-Apr-2006	christos	Coverity CID 744: Conditionally define out dead code (only if it is dead)
1.234	15-Apr-2006	christos	Coverity CID 2515-2519: Initialize rexmit on error path.
1.233	15-Apr-2006	christos	Coverity CID 2520: rexmit can be uninitialized on error path.
1.232	14-Apr-2006	blymn	Make i/o statistics collection more generic, include tape drives and nfs mounts in the set of devices that statistics will be reported on.
1.231	01-Mar-2006	yamt	branches: 1.231.2; 1.231.4; 1.231.6; 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.
1.230	11-Dec-2005	christos	branches: 1.230.2; 1.230.4; 1.230.6; merge ktrace-lwp.
1.229	05-Nov-2005	yamt	branches: 1.229.2; nfs_read: return EISDIR rather than EPERM for !VREG files.
1.228	02-Nov-2005	yamt	merge yamt-vop branch. remove following VOPs. VOP_BLKATOFF VOP_VALLOC VOP_BALLOC VOP_REALLOCBLKS VOP_VFREE VOP_TRUNCATE VOP_UPDATE
1.227	19-Sep-2005	christos	branches: 1.227.2; ATTRTIMEO takes 2 args.
1.226	19-Aug-2005	yamt	fix some simple bugs in the 64bit ino_t changes. - edp -> dp - * -> +
1.225	19-Aug-2005	christos	64 bit inode changes.
1.224	21-Jul-2005	yamt	use a correct credential for readlink. discussed on source-changes@.
1.223	07-Jul-2005	christos	1. use p = uio->uio_procp consistently and eliminate suspicious uses of curproc (where uio->uio_procp should be used?). Don't do this for nfs_commit(), because yamt says it is possibly wrong. 2. nfs_doio() does not use struct proc; remove it and the code to compute it. 3. use copyin_proc() and copyout_proc() instead of copyin() and copyout(). 4. check return of copyout_proc(). and mark return from copyin_proc() XXX 5. Eliminate check p == curproc assertion check from nfs_write; nfs_read does not have it and we might be called in a different process context anyway (PR 20138).
1.222	29-May-2005	christos	branches: 1.222.2; - sprinkle const - avoid shadowed variables - mark bad const use with XXXUNCONST
1.221	17-May-2005	christos	Yes, it was a cool trick >20 years ago to use "0123456789abcdef"[a] to implement, xtoa(), but I think defining the samestring 50 times is a bit too much. Defined HEXDIGITS and hexdigits in subr_prf.c and use it...
1.220	26-Feb-2005	perry	branches: 1.220.2; nuke trailing whitespace
1.219	26-Jan-2005	yamt	nfs_readdirrpc, nfs_readdirplusrpc: avoid infinite loops when getting readdir response without any entries or eof. PR/28971.
1.218	26-Jan-2005	yamt	handle a really empty directory, which doesn't have even the dot entry.
1.217	21-Jan-2005	yamt	branches: 1.217.2; s/time/mono_time/ for n_attrstamp and n_accstamp. (parts of) PR/25641.
1.216	19-Jan-2005	yamt	implement inaccurate mtime/ctime detection. namely, if mtime or ctime are same between pre_op_attr and post_op_attr when we expected them to be changed, don't trust the server.
1.215	08-Jan-2005	yamt	branches: 1.215.2; nfs_lookup: check n_nctime for positive entries as well to improve cache consistency.
1.214	17-Dec-2004	yamt	revive spec vop_bwrite as it's needed for block devices. PR/28684 from Jukka Salmi.
1.213	14-Dec-2004	yamt	redirect some VOPs which shouldn't be used for nfs to genfs_badop (ie. panic).
1.212	14-Dec-2004	yamt	- centerize code to invalidate stale cache. - don't ignore errors when invalidating buffers in nfs_open.
1.211	03-Oct-2004	yamt	nfs_readdirrpc, nfs_readdirplusrpc: don't expose kernel garbage data to userland.
1.210	01-Oct-2004	yamt	nfs_writerpc: fix PHOLD leak on error.
1.209	23-Sep-2004	yamt	nfs_readdirplusrpc: fix spurious EBUSYs.
1.208	20-Jul-2004	yamt	nfs_readdirplusrpc: fix a very long-standing cache corruption bug. in the case of !bigenough, don't fill d_type or dnlc with bogus data.
1.207	20-Jul-2004	yamt	revert nfs_vnops.c rev.1.189. it's no longer needed because cache_enter() has been changed to handle duplicated entries by itsself.
1.206	18-Jul-2004	yamt	nfs_commit: use NAC_NOTRUNC when loading an attribute as we're called holding pages locked.
1.205	08-Jul-2004	yamt	- include opt_inet.h for INET. - catchup to in_ifaddr -> in_ifaddrhead rename. XXX the address on the top of in_ifaddrhead is likely 127.0.0.1. using it to construct the verifier doesn't make much sense. maybe it's better to use some uuid or ip_randomid-like method.
1.204	08-Jul-2004	yamt	nfs_create: after an exclusive create rpc, make sure to update timestamps, which were likely used to store the verifier. reported by Mark Davies. PR/26200
1.203	27-Jun-2004	yamt	nfs_lookup: use cache_lookup_raw() so that: - "intrusive" dirops now have more chances to get benefits from dnlc. - fixes a deadlock due to vnode locking order inversion. nfs_create and others: purge stale dnlc entries as nfs_lookup() no longer does it automatically.
1.202	16-Jun-2004	yamt	nfs_lookup: maintain PDIRUNLOCK even in the case of success to make layered filesystems happy.
1.201	27-May-2004	yamt	remove an unused instance of VOP_UPDATE.
1.200	23-May-2004	christos	cut down another 7K by more NFS_V2_ONLY ifdefs.
1.199	17-May-2004	yamt	#if 0 out CREATE optimization for now because it has a problem in the case of O_CREAT\|O_TRUNC, which is hard to be fixed without changing upper layer.
1.198	10-May-2004	yamt	nfs_lookup: handle "." by ourselves as RFC1813 3.2 says.
1.197	10-May-2004	yamt	don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.196	08-May-2004	yamt	nfs_lookup: avoid CREATE optimization for DOTDOT. creating a DOTDOT entry has no sense and will fail anyway.
1.195	08-May-2004	yamt	nfs_mkdir: handle the "." case.
1.194	08-May-2004	yamt	nfs_lookitup: handle "." correctly rather than returning garbage on the stack.
1.193	07-May-2004	yamt	check read only mount appropriately. (fix a bug of nfs_vnops.c rev.1.192.) pointed by Rob Quinn on current-users@.
1.192	06-May-2004	yamt	because nfsv3 has the same CREATE semantics as ours, we don't have to issue LOOKUP RPCs beforehand.
1.191	05-Apr-2004	yamt	nfs_readdirplusrpc: fix a deadlock problem. don't wait for vnode lock to load attributes. otherwise, because READDIRPLUS returns DOTDOT entry as well, we violate locking order.
1.190	05-Apr-2004	yamt	don't issue VOP_GETATTR blindly in nfs_nget(). in many cases, GETATTR RPCs here is redundant because the caller has postop_attr. instead, make sure the resulted vnode have a valid attribute in nfs_lookup().
1.189	05-Apr-2004	yamt	nfs_readdirplusrpc: purge existing namecache entry before entering a new one. otherwise we'll get duplicated entries.
1.188	05-Apr-2004	yamt	when entering a namecache entry for nfs, ensure to update the appropriate timestamp in the nfsnode so that we don't get namecache-miss when looking up the node we just created.
1.187	05-Apr-2004	yamt	avoid unnecessary namecache purges in some places.
1.186	12-Mar-2004	yamt	branches: 1.186.2; shrink sizeof struct nfsnode by putting exclusive members into union.
1.185	12-Mar-2004	yamt	introduce a macro NFS_INVALIDATE_ATTRCACHE and use it instead of "n_attrstamp = 0".
1.184	07-Dec-2003	fvdl	Unix semantics dictate that access checks for files are done when it is opened. An open file can always be read from and/or written to, depending on how it was opened. Therefore, the read/write/commit RPCs should never return EACCESS, as they are only performed on files that have been successfully opened already. This change improves the current situation and works in most cases. It simply always uses the most recently known owner/group of the file, iff the authentication mechanism is AUTH_UNIX (in other cases, the creds for a succesful open are used, but note that no other cases are currently implemented). A retry mechanism can be used to catch a few more cases, but this is a good improvement for now.
1.183	29-Nov-2003	yamt	pad requests correctly in the zerocopy case of write rpc.
1.182	25-Oct-2003	christos	fix uninitialized variable
1.181	26-Sep-2003	yamt	do delayed truncation in nfs_getattr.
1.180	26-Sep-2003	yamt	change n_mtime from time_t to timespec in order to improve cache consistency. (1 second granularity is too loose these days.)
1.179	25-Sep-2003	enami	Make negative name cache works again.
1.178	17-Sep-2003	yamt	change nctime to timespec from time_t. there can be too many activities in a second.
1.177	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.176	30-Jul-2003	yamt	vrecycle removed nfs vnodes. not perfect, but enough for most cases.
1.175	29-Jun-2003	fvdl	branches: 1.175.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.174	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
1.173	27-Jun-2003	yamt	if nfs_writerpc() splitted an unstable write into some rpcs and write verifier was changed, we should restart from the first.
1.172	27-Jun-2003	yamt	indent.
1.171	03-Jun-2003	yamt	fix a problem in 'protected' case of writerpc. retransmitted mbufs can survive even after requests themselves finished. so, before unbusy pages, make sure that mbufs referring them go away. pointed by enami tsugutomo on port-mips.
1.170	27-May-2003	yamt	fix a memory leak bug that i introduced in rev.1.167. patch provided by enami tsugutomo on current-users.
1.169	26-May-2003	yamt	when a result of NFSv3 READLINK is too long for us, return ENAMETOOLONG rather than EBADRPC. (it's our implementation limit, not protocol limit.)
1.168	21-May-2003	yamt	remove local definitions of TRUE and FALSE.
1.167	21-May-2003	yamt	eliminate memcpy in the common and easy case of write.
1.166	03-May-2003	yamt	better handling of write verifier change.
1.165	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.164	09-Apr-2003	yamt	rename a very confusing variable name. (must_commit -> stalewriteverf)
1.163	09-Apr-2003	yamt	make per-iod datas together.
1.162	09-Apr-2003	yamt	rename nm_verf to nm_writeverf because it's confusing with nm_verf{str,len}.
1.161	02-Apr-2003	yamt	use queue manipulation macros.
1.160	31-Mar-2003	yamt	rename fvdl_debug to NFS_DEBUG_COMMIT. ok'ed by fvdl.
1.159	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.158	18-Feb-2003	jdolecek	add missing dot in comment
1.157	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.156	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
1.155	22-Oct-2002	yamt	fix panic introduced by my previous commit. for device special files, VOP_UNLOCK is called by nfs_loadattrcache with v_data == 0. reported and tested by Matthias Drochner.
1.154	22-Oct-2002	simonb	"nmp" in nfs_lookup() is set but not used, remove it.
1.153	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.152	18-Oct-2002	thorpej	nfs_remove(): Don't vput() the vnode twice if vp == dvp, vrele() and vput() instead.
1.151	19-May-2002	tls	branches: 1.151.2; Fix client-side lockmgr: locking against myself panic immediately upon an attempt to NFS-mount a filesystem with the -l (use ReaddirPlus RPC) option. Fix from Bill Sommerfeld.
1.150	12-May-2002	matt	branches: 1.150.2; Eliminate commons
1.149	28-Feb-2002	fvdl	Invalidate the access cache when loading a new set of attributes into the atribute cache. Fixes access cache problem seen by Nathan Funk of the UofS, relayed by Greg Oster.
1.148	15-Dec-2001	fvdl	Set np->n_size before calling nfs_vinvalbuf, to avoid recursion and confusion about the actual filesize. From Matt Dillon's similar change in FreeBSD. XXX n_size is really redundant in -current and must die. This commit XXX is more of a placeholder for a pullup into the 1.5 branch.
1.147	08-Dec-2001	lukem	- Implement uint32_t namei_hash(const char p, const char ep) which determines the equivalent MI hash32_str() hash for p. If ep != NULL, calculate the hash to the character before ep. If ep == NULL, calculate the has to the first / or NUL found, and point ep to that location. - Use namei_hash() to calculate cn_hash in lookup() and relookup(). Hash distribution goes from 35-40% to 55-70%, with similar profiled time spent in cache_lookup() and cache_enter() on my P3-600. - Use namei_hash() to calculate cn_hash in nfs_readdirplusrpc(), insetad of homegrown code (that differed from that in lookup() !) namei_hash() has better spread and is faster than previous code (which used a non-constant multiplication).
1.146	04-Dec-2001	christos	PR/14817: Gregory McGarry: NFS_V2_ONLY doesn't seem to work.
1.145	30-Nov-2001	chs	call VOP_PUTPAGES() directly instead of indirecting through the UVM pager op vector.
1.144	29-Nov-2001	christos	PR/14776: Emmanuel Dreyfus: cross device hard link causes panic. Call VOP_ABORTOP on the right vnode damnit!
1.143	10-Nov-2001	lukem	add RCSIDs
1.142	07-Nov-2001	bjh21	Diagnostic panics should be enabled when DIAGNOSTIC is defined, not undefined.
1.141	13-Oct-2001	simonb	branches: 1.141.2; Remove so variables that are only ever set and never referenced.
1.140	22-Sep-2001	sommerfeld	Add fifo_putpages() placebo so that the vnode's uobj is unlocked.
1.139	20-Sep-2001	chs	fix nfs_bmap() so that it works for both genfs_{get,put}pages() and swap/vnd.
1.138	15-Sep-2001	chs	a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
1.137	17-Aug-2001	chs	branches: 1.137.2; add getpages/putpages entries for spec vnodes.
1.136	14-Aug-2001	itojun	don't panic on mknod(2) over NFS. PR 13705.
1.135	24-Jul-2001	assar	change vop_symlink and vop_mknod to return vpp (the created node) refed, so that the caller can actually use it. update callers and file systems that implement these vnode operations
1.134	07-Jun-2001	lukem	branches: 1.134.2; delint lvalue cast abuse
1.133	28-May-2001	chs	add a genfs_mmap() and change all of the disk-based filesystems to implement VOP_MMAP() with the genfs version, in preparation for actually using this VOP.
1.132	14-May-2001	fvdl	Lock vp in nfs_link while we're busy with it (doing VOP_FSYNC, etc).
1.131	20-Apr-2001	fvdl	Don't forget to unlock the vnode returned by cache_lookup if the subsequent access check fails. Don't overwrite the error code returned by cache_lookup. Remove a piece of redundant code. Should fix kern/12680.
1.130	11-Feb-2001	enami	branches: 1.130.2; Unlock the rename target vnode after sillyrename'ing it.
1.129	06-Feb-2001	fvdl	Get locking in rmdir right. Don't unlock a vnode when passing its associated nfsnode to nfs_lookitup, it is not needed, and fixes nfs_remove.
1.128	06-Feb-2001	fvdl	Do actual vnode locking for NFS.
1.127	22-Jan-2001	jdolecek	make filesystem vnodeop, specop, fifoop and vnodeopv_* arrays const
1.126	12-Dec-2000	chs	initialize read creds in nfs_open() too.
1.125	30-Nov-2000	chs	in nfs_open(), initialize the write creds if we're opening for writing. otherwise we would never set them if we only modify the file via mmap().
1.124	27-Nov-2000	chs	Initial integration of the Unified Buffer Cache project.
1.123	08-Nov-2000	chs	in nfs_flush(), only play games with B_NOCACHE for VREG vnodes. if we do this for VBLK vnodes which are in use by softdep mounts, brelse() will mark the buffer B_INVAL as well, which makes the softdep code very unhappy.
1.122	02-Oct-2000	itojun	check in_ifaddr only if INET is compiled
1.121	19-Sep-2000	bjh21	Extend NFS_V2_ONLY to remove NQNFS lease support as well. Saves another 10k.
1.120	19-Sep-2000	fvdl	Update for VOP_FSYNC parameter change. Simplify nfs_flush.
1.119	19-Sep-2000	bjh21	New kernel option, NFS_V2_ONLY, which aims to reduce the NFS client to just that required to support NFSv2 mounts. Not finished yet, but already provides some 44k of saving in code size on arm26. More savings, and some documentation, are still to come.
1.118	19-Sep-2000	fvdl	Fix bug in access cache that might result in permission being denied needlessly. From Matthias Drochner.
1.117	03-Aug-2000	thorpej	Convert namei pathname buffer allocation to use the pool allocator.
1.116	03-Aug-2000	thorpej	MALLOC()/FREE() are not to be used for variable size allocations.
1.115	22-Jul-2000	jdolecek	change the lf_advlock() arguments from int lf_advlock __P((struct lockf *, off_t, caddr_t, int, struct flock , int)); to int lf_advlock __P((struct vop_advlock_args , struct lockf *, off_t)); This matches common usage and is also compatible with similar change in FreeBSD (though they use u_quad_t as last arg).
1.114	27-Jun-2000	mrg	remove include of <vm/vm.h>
1.113	26-May-2000	enami	branches: 1.113.4; - Try to commit another buffer even if previous commit failed except the case that write verf is changed. Suggested by mycroft@netbsd.org. - Reset wcred to NULL (i.e., write credential isn't decieded) everytime before gathering buffer for new commit, so that there is a chance to the commit request is merged.
1.112	25-May-2000	enami	In nfs_flush, if the previous commit succeeded and we may have more uncommitted dirty buffer, attempt to commit them.
1.111	30-Mar-2000	augustss	Remove more register declarations.
1.110	30-Mar-2000	augustss	Remove register declarations.
1.109	30-Mar-2000	simonb	Delete redundant decls of fifo_vnodeop_p - it's in <miscfs/fifofs/fifo.h>. Don't need <sys/conf.h> here.
1.108	29-Nov-1999	fvdl	Insert an extra VOP_ACCESS check in nfs_lookup, to avoid cached access mishaps for lookup and getattr. Closes PR 8884. While at it, cache access RPCs.
1.107	15-Nov-1999	fvdl	Add Kirk McKusick's soft updates code to the trunk. Not enabled by default, as the copyright on the main file (ffs_softdep.c) is such that is has been put into gnusrc. options SOFTDEP will pull this in. This code also contains the trickle syncer. Bump version number to 1.4O
1.106	05-Sep-1999	jdolecek	branches: 1.106.2; 1.106.4; 1.106.8; Adapt to cache_lookup() changes. XXX I had no chance to actually test the changes for nfs, but hopefully I got it right. Tested by: jdolecek Rewieved by: wrstuden
1.105	03-Aug-1999	wrstuden	Add support for fcntl(2) to generate VOP_FCNTL calls. Any fcntl call with F_FSCTL set and F_SETFL calls generate calls to a new fileop fo_fcntl. Add genfs_fcntl() and soo_fcntl() which return 0 for F_SETFL and EOPNOTSUPP otherwise. Have all leaf filesystems use genfs_fcntl(). Reviewed by: thorpej Tested by: wrstuden
1.104	02-Aug-1999	wrstuden	Teach nfs_lookup() to set PDIRUNLOCK when appropriate. Should resolve PR 8051 by Konrad Schroder.
1.103	29-Jul-1999	thorpej	In nfs_create(), make sure error is reset to 0 if we restart the operation.
1.102	08-Jul-1999	wrstuden	Teach nfs_lookup to clear PDIRUNLOCK.
1.101	29-May-1999	fvdl	Be more correct with attribute structures for setattr RPCs and friends, so that picky servers (e.g. Solaris 7) don't refuse our requests. Move some code into a macro, and a bit of KNF. From OpenBSD.
1.100	24-Mar-1999	mrg	branches: 1.100.2; 1.100.4; 1.100.6; completely remove Mach VM support. all that is left is the all the header files as UVM still uses (most of) these.
1.99	22-Mar-1999	kleink	Add _PC_FILESIZEBITS to pathconf vnop.
1.98	06-Mar-1999	fair	Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.97	09-Aug-1998	perry	branches: 1.97.2; bzero->memset, bcopy->memcpy, bcmp->memcmp
1.96	08-Aug-1998	kleink	Pathconf: for V2 mounts, revert back to failing with EINVAL if an RPC would be necessary to obtain the information, as this fits the pathconf semantics of `no association supported' better than `no limit available.'
1.95	07-Aug-1998	kleink	Add client pathconf support.
1.94	24-Jun-1998	sommerfe	Always include fifos; "not an option any more".
1.93	22-Jun-1998	sommerfe	defopt for options FIFO
1.92	05-Jun-1998	kleink	Convert fsync vnode operator implementations and usage from the old `waitfor' argument and MNT_WAIT/MNT_NOWAIT to `flags' and FSYNC_WAIT.
1.91	08-May-1998	kleink	Fix some arithmetics lossage on typeless pointers.
1.90	03-Mar-1998	fvdl	Fix cookie handling I messed up totally when doing the Lite2 thing. (Hello McFly? Anybody home?)
1.89	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.88	10-Feb-1998	mrg	- add defopt's for UVM, UVMHIST and PMAP_NEW. - remove unnecessary UVMHIST_DECL's.
1.87	05-Feb-1998	mrg	initial import of the new virtual memory system, UVM, into -current. UVM was written by chuck cranor <chuck@maria.wustl.edu>, with some minor portions derived from the old Mach code. i provided some help getting swap and paging working, and other bug fixes/ideas. chuck silvers <chuq@chuq.com> also provided some other fixes. this is the rest of the MI portion changes. this will be KNF'd shortly. :-)
1.86	20-Oct-1997	thorpej	Fix alignment problems. From Frank van der Linden <fvdl@NetBSD.ORG>.
1.85	19-Oct-1997	fvdl	* Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.84	17-Oct-1997	christos	u_short -> mode_t
1.83	13-Oct-1997	fvdl	Get rid of some MARKCACHED calls I thought better of. Make sure d_reclen is aligned for off_t access, or things will break on the Alpha.
1.82	12-Oct-1997	fvdl	Do negative lookup caching. Use a timestamp of the oldest negative cache entry, so it can be checked against directory modification time for validity.
1.81	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.80	17-Jul-1997	fvdl	branches: 1.80.2; * Deal with servers that don't give complete FSINFO (like NT) From Olaf Seibert <rhialto@polder.ubc.kun.nl> (PR 3687) * Make an attempt to check the maximum filesize before attempting a write to the server, as write RPCs will typically happen asynchronously, and the process will not see the error. Fixes problems with unexpectly truncated files at 4G * Pass up errors in nfs_writerpc correctly
1.79	14-Jul-1997	fvdl	Don't assume that pointers into mbuf data remain valid across nfsm_dissect. In readdirplus, don't keep such pointers but store the file attributes in a variable instead until they are needed. Change nfsm_loadattr* a bit so it can accept a direct pointer to an nfs_fattr structure.
1.78	04-Jul-1997	drochner	Don't cast 64bit (off_t) file sizes to vm_offset_t (32bit on many architectures), truncate them intelligently instead. The truncation is done centralized in vnode_pager.c. This prevents from wrap-over effects when parts of large (>2^32 byte) files are mmapped. Don't allow to mmap above the numerical range of vm_offset_t. This is considered a temporary solution until the vm system handles the object sizes/offsets more cleanly.
1.77	30-Jun-1997	fvdl	Immediately return EPERM for a VOP_REMOVE on a directory.
1.76	12-May-1997	fvdl	clear B_AGE for non-flush writes, buffers seem to be reused too quickly, disturbing NFS performance (XXXX needs further analysis and a _real_ fix)
1.75	08-May-1997	mycroft	Need stat.h.
1.74	08-May-1997	mycroft	Pass the vnode type to vaccess(), and use it when checking VEXEC. Make sure that the mode bits passed to vaccess() and returned by foo_getattr() contain only permission bits.
1.73	08-May-1997	mycroft	VEXEC -> VLOOKUP, as appropriate.
1.72	05-Mar-1997	mycroft	In nfs_link(), check for a cross-device mount before looking in the v_data field.
1.71	22-Feb-1997	fvdl	Fixes from BSDI (thanks go to Keith Bostic). Original RCS messages: date: 1996/07/23 17:14:46; author: donn; state: Exp; lines: +6 -4 Be sure to push out the last page of the file before truncating it. date: 1996/10/14 22:41:20; author: donn; state: Exp; lines: +2 -2 From Chris: Nfs_link() called vput() on the wrong vnode when aborting from a cross-device link, which could (and did) lead to crashes. date: 1996/10/24 16:43:43; author: pjd; state: Exp; lines: +6 -2 Return EOPNOTSUPP when trying to do a setattr with flags. === Also (from BSDI too, but the RCS message did not quite describe the change to this particular file well): move the EROFS a bit further down to let VOP_ACCESS do it's work and return an 'expected' error value to a possible layered filesystem.
1.70	09-Feb-1997	fvdl	* Fix some bugs in NQNFS (malformed RPC requests, no directory lease eviction) * Avoid possible NULL ptr ref in nfs_reply * Don't ever try to sillyrename directories (from FreeBSD)
1.69	02-Dec-1996	thorpej	branches: 1.69.4; NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.68	25-Oct-1996	cgd	make the namei struct members ni_dirp and ni_next, and the componentname struct member cn_nameptr 'const', since they should never be used to modify the path name. (Only the pathname buffer, cn_pnbuf, should be modified.) Propagate the const poisoning to code that uses the namei and componentname structs.
1.67	13-Oct-1996	christos	revert kprintf changes
1.66	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.65	07-Sep-1996	mycroft	Implement poll(2).
1.64	01-Sep-1996	mycroft	Add a set of generic file system operations that most file systems use. Also, fix some time stamp bogosities.
1.63	07-Jul-1996	fvdl	Use the right time for v3 setattr operation.
1.62	11-May-1996	mycroft	branches: 1.62.4; Change VOP_UPDATE() semantics: * Make 2nd and 3rd args timespecs, not timevals. * Consistently pass a Boolean as the 4th arg (except in LFS). Also, fix ffs_update() and lfs_update() to actually change the nsec fields.
1.61	03-Apr-1996	thorpej	Make these link in the absense of "options FIFO".
1.60	05-Mar-1996	jtk	fix panic "leaf should be empty" on diagnostic kernels when unlinking on a read-only file system.
1.59	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.58	09-Feb-1996	christos	nfs prototype changes
1.57	09-Feb-1996	mycroft	Fix vop_link, vop_symlink, and vop_remove semantics in several ways: * Change the argument names to vop_link so they actually make sense. * Implement vop_link and vop_symlink for all file systems, so they do proper cleanup. * Require the file system to decide whether or not linking and unlinking of directories is allowed, and disable it for all current file systems.
1.56	01-Feb-1996	jtc	Rename struct timespec fields to conform to POSIX.1b
1.55	31-Jan-1996	mycroft	Don't specify a uid or gid in create operations; let the server fill it in.
1.54	31-Jan-1996	mycroft	Correct some uses of -1 and VNOVAL.
1.53	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.52	14-Oct-1995	ghudson	Add cookie support. Stash cookies in the word prior to the end of each entry, and read them out in nfs_readdir(). Caveat: our current caching method for directory blocks uses the server offset of the first directory entry as an identifier, so a Linux emulation getdirentries() will wind up retrieving one block from the NFS server for each directory entry, unnecessarily thrashing the cache. The situation isn't as bad for other emulations. Instead of getblk(), we need to write a routine to scan each cache block associated with vp to find a cookie that matches at some directory entry. Some later time.
1.51	09-Oct-1995	mycroft	branches: 1.51.2; For now, return EINVAL if the client needs cookies.
1.50	18-Mar-1995	gwr	Initialize fields in the RPC data where we were sending garbage.
1.49	10-Jan-1995	mycroft	Make sure readdir requests are only truncated on block boundaries.
1.48	29-Dec-1994	mycroft	Minor consistency nits.
1.47	29-Dec-1994	mycroft	Remove a bit of redundant code.
1.46	27-Dec-1994	mycroft	Format police.
1.45	27-Dec-1994	mycroft	Fix typos in last change.
1.44	24-Dec-1994	ws	Implement and use a common access checking routine
1.43	13-Dec-1994	mycroft	Turn lease_check() into a vnode op, per CSRG.
1.42	13-Dec-1994	mycroft	Remove an old `#ifdef notyet'.
1.41	20-Oct-1994	cgd	update for new syscall args description mechanism
1.40	30-Aug-1994	pk	mknod() must release its new vnode.
1.39	21-Aug-1994	mycroft	Don't attempt to use IO_APPEND for NQNFS, as suggested by Rick Macklem.
1.38	13-Aug-1994	pk	Files with > 1 links can always be removed on the server, even if a "silly name" exists.
1.37	08-Aug-1994	deraadt	delete unused extern decl
1.36	12-Jul-1994	mycroft	Bug fix from Rick Macklem for a problem with linking to an open file.
1.35	03-Jul-1994	mycroft	branches: 1.35.2; Fix problem with O_TRUNC and NFS device nodes.
1.34	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.33	22-Jun-1994	pk	straighten out diskless swap code somewhat.
1.32	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.31	19-May-1994	cgd	stupidity for prototypes...
1.30	25-Apr-1994	cgd	some prototype cleanup, eliminate/replace bogus types (e.g. quad and u_quad) -> use better types (e.g. quad_t & u_quad_t in inodes), some cleanup.
1.29	21-Apr-1994	cgd	blow away all vestiges of nfsnode locking. (1) it's unnecessary (2) it causes machines to hang (yup!) (3) it'd be gone in a few days anyway (it'd been yanked out of 4.4-Lite by macklem long ago) It was only there because macklem couldn't originally decide if things should be locked, or not...
1.28	21-Apr-1994	cgd	Convert mount, vnode, and buf structs to use <sys/queue.h>. Also, some knf and structure frobbing to do along with it.
1.27	14-Apr-1994	pk	Remove bogus type translation; in stead, use IFTOVT again to go from `NFS mode bits' to `vnode type'. Use aliased vnode consistently.
1.26	27-Mar-1994	cgd	expand uid_t/gid_t/off_t
1.25	09-Mar-1994	ws	Make FFS optional
1.24	15-Feb-1994	mycroft	Macros bite again.
1.23	15-Feb-1994	mycroft	Format police.
1.22	15-Feb-1994	pk	Update {a,m}time vnode attributes on special files a la ufs_vnode.c, but make it a non-urgent operation, to leave us some performance.
1.21	06-Feb-1994	mycroft	Eliminate some more uses of b_actl.
1.20	10-Jan-1994	pk	reparations...
1.19	10-Jan-1994	pk	Don't deny unlink()s of files with the "silly" bit on, but still have > 1 links, but avoid doing gratuitous (possibly expensive) get_attr() calls.
1.18	04-Jan-1994	cgd	add support for union and loopback mounts, from jsp
1.17	22-Dec-1993	cgd	fix nfs_print, add cross-device link checking (From jsp)
1.16	18-Dec-1993	mycroft	Canonicalize all #includes.
1.15	16-Dec-1993	pk	Avoid dereferencing NULL pointer in nfs_doio() when B_PHYS is on. Remove comment talking about nfsiomaps that we don't have. Always use credentials that are in the buffer header, in stead of trying to get them from pageproc, which may once have been necessary to push pages to swap (cannot imaging anyone having exercised this over NFS though).
1.14	07-Dec-1993	cgd	fix a goof that i made; return before nfs_lock() is called...
1.13	20-Nov-1993	cgd	do something better with lookup return values; suggested by BSDI's msdosfs mod
1.12	12-Nov-1993	cgd	new specfs.h and fifo.h locations
1.11	07-Sep-1993	ws	branches: 1.11.2; Changes to VFS readdir semantics NFS changes for better cookie support ISOFS changes for better Rockridge support and support for generation numbers
1.10	02-Aug-1993	mycroft	Make return type of nfs_print be a void, not an int.
1.9	13-Jul-1993	cgd	get rid of some more bogus changes from a week ago
1.8	13-Jul-1993	cgd	diskless changes made last time were hosed; were using NULL for "no credentials" rather than NOCRED.
1.7	07-Jul-1993	cgd	changes from ws to support diskless booting... these are "OK" on inspection and after testing... (actually, currently, none of the changed code is even used...)
1.6	03-Jun-1993	cgd	fix for macklem's bogus use of the va_flags field, supplied by John Woods, jfwfrom: @ksr.com. also, fixes the following problems: the va_gen field is in a similar position (Suns are going to be reporting the change-date microseconds as their "generation"), I've supplied my own set of diffs below for your inspection. Note these aren't even compiled, but they're pretty similar to what I had to do to our older version of OSF/1 here. (There's also an unrelated change supplied for xdr_subs.h; the pointer types supplied to the fxdr_time() and txdr_time() macros are not, in fact, both struct timevals. That turns out to be one of many tips-of-the-iceberg facing those porting the (old) Berkeley NFS code to 64-bit machines...)
1.5	22-May-1993	cgd	add Yuval Yarom's changes (originally for BSD/386) for advisory record locking on NFS files. Note that this DOES NOT support network locking, only local advisory locks.
1.4	21-May-1993	cgd	add rcsid again; fix RCS+crash fuckup
1.3	10-Apr-1993	glass	migrated code to make split possible
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
1.1.1.4	01-Mar-1998	fvdl	Import some files that were changed after Lite2
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.11.2.5	07-Dec-1993	cgd	update from trunk
1.11.2.4	20-Nov-1993	cgd	do something better with lookup return values; suggested by BSDI's msdosfs mod
1.11.2.3	12-Nov-1993	cgd	new specfs.h and fifo.h locations, and include file syntax updates
1.11.2.2	16-Oct-1993	mycroft	Nuke #ifndef hp300 & i386 block of variable definitions that were never used.
1.11.2.1	24-Sep-1993	mycroft	Make all files using spl*() #include cpu.h. Changes from trunk. nfs_vfsops.c, nfsmount.h: Make nfs_quotactl() take an int rather than a uid_t, as it might be -1. nfs_vnops.c: va_size and va_bytes are now quads.
1.35.2.4	06-Oct-1994	mycroft	Update from trunk.
1.35.2.3	21-Aug-1994	mycroft	update from trunk
1.35.2.2	09-Aug-1994	mycroft	update from trunk
1.35.2.1	12-Jul-1994	cgd	linking to open file fix, from trunk.
1.51.2.2	02-Feb-1996	mycroft	Bring in changes for mondo patch 2.
1.51.2.1	17-Oct-1995	ghudson	Update from main branch to get cookie support into the NFS readdir vnode operation.
1.62.4.3	05-Mar-1997	mycroft	Pull up nfs_link() fix.
1.62.4.2	04-Mar-1997	mycroft	Pull up bug fixes from -current, per fvdl.
1.62.4.1	08-Jul-1996	jtc	Pulled up from rev 1.63 by request from Frank van der Linden
1.69.4.1	12-Mar-1997	is	Merge in changes from Trunk
1.80.2.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.97.2.1	09-Nov-1998	chs	initial snapshot. lots left to do.
1.100.6.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.100.4.5	02-Aug-1999	thorpej	Update from trunk.
1.100.4.4	11-Jul-1999	chs	remove uvm_vnp_uncache(), it's no longer needed.
1.100.4.3	04-Jul-1999	chs	support VOP_BALLOC().
1.100.4.2	21-Jun-1999	thorpej	Sync w/ -current.
1.100.4.1	07-Jun-1999	chs	merge everything from chs-ubc branch.
1.100.2.3	15-Aug-2000	he	Apply patch (requested by fvdl): Be careful about data consistency across operations which may block. Should fix some reported nfs_lookup panics.
1.100.2.2	05-Jan-2000	he	Pull up revision 1.108 (via patch, requested by fvdl): Insert an extra VOP_ACCESS check in nfs_lookup, preventing cached access mishaps for lookup and getattr. Fixes PR#8884.
1.100.2.1	22-Jun-1999	perry	pullup 1.100->1.101 (fvdl): fix file creation with a Solaris 7 server
1.106.8.2	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.106.8.1	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".
1.106.4.1	19-Oct-1999	fvdl	Bring in Kirk McKusick's FFS softdep code on a branch.
1.106.2.6	23-Apr-2001	bouyer	Sync with HEAD.
1.106.2.5	11-Feb-2001	bouyer	Sync with HEAD.
1.106.2.4	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.106.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.106.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.106.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.113.4.4	29-Jan-2002	he	Pull up revision 1.148 (requested by fvdl): Set np->n_size before calling nfs_vinvalbuf. Avoids confusion as to the actual file size.
1.113.4.3	14-Dec-2000	he	Pull up revisions 1.120,1.123 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.113.4.2	19-Sep-2000	fvdl	Revision 1.118: Fix bug in access cache that might result in permission being denied needlessly. From Matthias Drochner. (approved by thorpej)
1.113.4.1	30-Jul-2000	jdolecek	Pullup from trunk (approved by thorpej): Change lf_advlock() to: int lf_advlock (struct vop_advlock_args , struct lockf *, off_t) This matches it's usage. Change inspired by FreeBSD, though we use off_t instead u_quad_t as the last argument. sys/lockf.h rev. 1.9 msdosfs/msdosfs_vnops.c rev. 1.99 kern/vfs_lockf.c rev. 1.17 miscfs/specfs/spec_vnops.c rev. 1.49 nfs/nfs_vnops.c rev. 1.115 ufs/ext2fs/ext2fs_vnops.c rev. 1.28 ufs/ufs/ufs_vnops.c rev. 1.72
1.130.2.17	11-Dec-2002	thorpej	Sync with HEAD.
1.130.2.16	11-Nov-2002	nathanw	Catch up to -current
1.130.2.15	22-Oct-2002	thorpej	Sync with HEAD.
1.130.2.14	18-Oct-2002	thorpej	Sync with HEAD.
1.130.2.13	15-Jul-2002	nathanw	Whitespace.
1.130.2.12	12-Jul-2002	nathanw	No longer need to pull in lwp.h; proc.h pulls it in for us.
1.130.2.11	24-Jun-2002	nathanw	Curproc->curlwp renaming. Change uses of "curproc->l_proc" back to "curproc", which is more like the original use. Bare uses of "curproc" are now "curlwp". "curproc" is now #defined in proc.h as ((curlwp) ? (curlwp)->l_proc) : NULL) so that it is always safe to reference curproc (dereferencing curproc is another story, but that's always been true).
1.130.2.10	20-Jun-2002	nathanw	Catch up to -current.
1.130.2.9	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.130.2.8	08-Jan-2002	nathanw	Catch up to -current.
1.130.2.7	14-Nov-2001	nathanw	Catch up to -current.
1.130.2.6	22-Oct-2001	nathanw	Catch up to -current.
1.130.2.5	26-Sep-2001	nathanw	Catch up to -current. Again.
1.130.2.4	21-Sep-2001	nathanw	Catch up to -current.
1.130.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.130.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.130.2.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.134.2.8	30-Sep-2002	jdolecek	add support for kevents to NFS to detect file changes on server by other NFS clients, polling kernel thread is used to periodically check for attribute changes of watched files; the NFS server is only contacted when the vnode expires from local attrcache (which takes 5-60 seconds currently), to keep network&CPU overhead low the routine checking for remote changes is quite simplistic, but hopefully doing it's job well enough
1.134.2.7	23-Sep-2002	jdolecek	add spec kqfilter vnode op
1.134.2.6	22-Sep-2002	jdolecek	add fifo_kqfilter() to fifo ops, to switch on support for kevents
1.134.2.5	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.134.2.4	16-Mar-2002	jdolecek	Catch up with -current.
1.134.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.134.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.134.2.1	03-Aug-2001	lukem	update to -current
1.137.2.2	01-Oct-2001	fvdl	Catch up with -current.
1.137.2.1	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).
1.141.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.150.2.1	30-May-2002	gehenna	Catch up with -current.
1.151.2.1	18-Oct-2002	thorpej	Pullup revision 1.152. Original log message: nfs_remove(): Don't vput() the vnode twice if vp == dvp, vrele() and vput() instead.
1.175.2.12	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.175.2.11	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.175.2.10	04-Feb-2005	skrll	Sync with HEAD.
1.175.2.9	24-Jan-2005	skrll	Sync with HEAD.
1.175.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.175.2.7	18-Dec-2004	skrll	Sync with HEAD.
1.175.2.6	19-Oct-2004	skrll	Sync with HEAD
1.175.2.5	24-Sep-2004	skrll	Sync with HEAD.
1.175.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.175.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.175.2.2	03-Aug-2004	skrll	Sync with HEAD
1.175.2.1	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.
1.186.2.10	16-Mar-2005	tron	Pull up revision 1.219 (requested by yamt in ticket #1208): nfs_readdirrpc, nfs_readdirplusrpc: avoid infinite loops when getting readdir response without any entries or eof. PR/28971.
1.186.2.9	16-Mar-2005	tron	Pull up revision 1.206 (requested by yamt in ticket #1134): nfs_commit: use NAC_NOTRUNC when loading an attribute as we're called holding pages locked.
1.186.2.8	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.186.2.7	11-Jan-2005	jmc	Pullup patch (requested by yamt in ticket #1077) nfs_lookup: check n_nctime for positive entries as well to improve cache consistency.
1.186.2.6	10-Jul-2004	tron	branches: 1.186.2.6.2; Pull up revision 1.204 via patch (requested by yamt in ticket #636): nfs_create: after an exclusive create rpc, make sure to update timestamps, which were likely used to store the verifier. reported by Mark Davies. PR/26200
1.186.2.5	10-Jul-2004	tron	Pull up revision 1.191 (requested by tls in ticket #634): nfs_readdirplusrpc: fix a deadlock problem. don't wait for vnode lock to load attributes. otherwise, because READDIRPLUS returns DOTDOT entry as well, we violate locking order.
1.186.2.4	10-Jul-2004	tron	Pull up revision 1.190 (requested by tls in ticket #634): don't issue VOP_GETATTR blindly in nfs_nget(). in many cases, GETATTR RPCs here is redundant because the caller has postop_attr. instead, make sure the resulted vnode have a valid attribute in nfs_lookup().
1.186.2.3	10-Jul-2004	tron	Pull up revision 1.189 (requested by tls in ticket #634): nfs_readdirplusrpc: purge existing namecache entry before entering a new one. otherwise we'll get duplicated entries.
1.186.2.2	10-Jul-2004	tron	Pull up revision 1.188 (requested by tls in ticket #634): when entering a namecache entry for nfs, ensure to update the appropriate timestamp in the nfsnode so that we don't get namecache-miss when looking up the node we just created.
1.186.2.1	10-Jul-2004	tron	Pull up revision 1.187 (requested by tls in ticket #634): avoid unnecessary namecache purges in some places.
1.186.2.6.2.5	27-Oct-2005	riz	Pull up following revision(s) (requested by christos in ticket #5863): sys/nfs/nfs_subs.c: revision 1.152 via patch sys/nfs/nfs.h: revision 1.49 sys/nfs/nfs_vfsops.c: revision 1.149 via patch usr.sbin/amd/include/config.h: revision 1.36 sys/nfs/nfs_vnops.c: revision 1.227 via patch sys/nfs/nfsmount.h: revision 1.34 Allow the attribute cache to be turned off, and allow amd to do it.
1.186.2.6.2.4	16-Mar-2005	tron	Pull up revision 1.219 (requested by yamt in ticket #1208): nfs_readdirrpc, nfs_readdirplusrpc: avoid infinite loops when getting readdir response without any entries or eof. PR/28971.
1.186.2.6.2.3	16-Mar-2005	tron	Pull up revision 1.206 (requested by yamt in ticket #1134): nfs_commit: use NAC_NOTRUNC when loading an attribute as we're called holding pages locked.
1.186.2.6.2.2	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.186.2.6.2.1	11-Jan-2005	jmc	Pullup patch (requested by yamt in ticket #1077) nfs_lookup: check n_nctime for positive entries as well to improve cache consistency.
1.215.2.1	29-Apr-2005	kent	sync with -current
1.217.2.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.217.2.1	12-Feb-2005	yamt	sync with head.
1.220.2.2	16-Jul-2006	ghen	Pull up following revision(s) (requested by jld in ticket #1424): sys/nfs/nfs_vnops.c: revision 1.240 via patch sys/nfs/nfs_var.h: revision 1.62 via patch Fix race condition in NFS renaming that could cause the renamed file to be deleted (PR/33861).
1.220.2.1	27-Sep-2005	tron	branches: 1.220.2.1.2; Pull up following revision(s) (requested by christos in ticket #816): sys/nfs/nfs_vfsops.c: revision 1.149 sys/nfs/nfs_vnops.c: revision 1.227 ATTRTIMEO takes 2 args.
1.220.2.1.2.1	16-Jul-2006	ghen	Pull up following revision(s) (requested by jld in ticket #1424): sys/nfs/nfs_vnops.c: revision 1.240 via patch sys/nfs/nfs_var.h: revision 1.62 via patch Fix race condition in NFS renaming that could cause the renamed file to be deleted (PR/33861).
1.222.2.11	27-Feb-2008	yamt	revert incomplete nfs client locking for now.
1.222.2.10	27-Feb-2008	yamt	sync with head.
1.222.2.9	15-Feb-2008	yamt	- sprinkle some locks. - disable MNT_UPDATE because it involves too much locking headache. - don't overwrite other bits in v_vflags when setting VV_ROOT.
1.222.2.8	04-Feb-2008	yamt	sync with head.
1.222.2.7	21-Jan-2008	yamt	sync with head
1.222.2.6	07-Dec-2007	yamt	sync with head
1.222.2.5	15-Nov-2007	yamt	sync with head.
1.222.2.4	03-Sep-2007	yamt	sync with head.
1.222.2.3	26-Feb-2007	yamt	sync with head.
1.222.2.2	30-Dec-2006	yamt	sync with head.
1.222.2.1	21-Jun-2006	yamt	sync with head.
1.227.2.1	20-Oct-2005	yamt	adapt nfs.
1.229.2.3	22-Nov-2005	yamt	remove a whitespace change which is not related to this branch.
1.229.2.2	18-Nov-2005	yamt	- associate read-ahead context to vnode, rather than file. - revert VOP_READ prototype.
1.229.2.1	15-Nov-2005	yamt	adapt ffs, lfs, nfs.
1.230.6.3	01-Jun-2006	kardel	Sync with head.
1.230.6.2	22-Apr-2006	simonb	Sync with head.
1.230.6.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.230.4.1	09-Sep-2006	rpaulo	sync with head
1.230.2.2	15-Jan-2006	yamt	rename VMSPACE_IS_KERNEL to VMSPACE_IS_KERNEL_P. ("predicate") suggested by Matt Thomas.
1.230.2.1	31-Dec-2005	yamt	- adapt nfs. - nfs_doio_read: #if 0 out "killproc if text is modified" part of the code as it's broken. (a process reading the modified text is not necessarily a process which is using the file as a text.)
1.231.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.231.4.3	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.
1.231.4.2	19-Apr-2006	elad	sync with head.
1.231.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.231.2.3	11-Aug-2006	yamt	sync with head
1.231.2.2	26-Jun-2006	yamt	sync with head.
1.231.2.1	24-May-2006	yamt	sync with head.
1.236.2.1	19-Jun-2006	chap	Sync with head.
1.237.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.241.6.2	10-Dec-2006	yamt	sync with head.
1.241.6.1	22-Oct-2006	yamt	sync with head
1.241.4.3	01-Feb-2007	ad	Sync with head.
1.241.4.2	12-Jan-2007	ad	Sync with head.
1.241.4.1	18-Nov-2006	ad	Sync with head.
1.245.2.1	17-Feb-2007	tron	Apply patch (requested by chs in ticket #422): - Fix various deadlock problems with nullfs and unionfs. - Speed up path lookups by upto 25%.
1.249.2.3	07-May-2007	yamt	sync with head.
1.249.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.249.2.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.252.4.1	11-Jul-2007	mjf	Sync with head.
1.252.2.4	01-Sep-2007	ad	Update for pool_cache API changes.
1.252.2.3	08-Jun-2007	ad	Sync with head.
1.252.2.2	05-Apr-2007	ad	Compile fixes.
1.252.2.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.256.12.1	13-Nov-2007	bouyer	Sync with HEAD
1.256.8.4	23-Mar-2008	matt	sync with HEAD
1.256.8.3	09-Jan-2008	matt	sync with HEAD
1.256.8.2	08-Nov-2007	matt	sync with -HEAD
1.256.8.1	06-Nov-2007	matt	sync with HEAD
1.256.6.5	09-Dec-2007	jmcneill	Sync with HEAD.
1.256.6.4	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.256.6.3	14-Nov-2007	joerg	Sync with HEAD.
1.256.6.2	11-Nov-2007	joerg	Sync with HEAD.
1.256.6.1	29-Oct-2007	joerg	Sync with HEAD.
1.257.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.257.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.257.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.257.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.260.2.2	26-Dec-2007	ad	Sync with head.
1.260.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.261.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.266.16.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.266.16.1	19-Oct-2008	haad	Sync with HEAD.
1.266.10.9	10-Oct-2010	yamt	some locking changes
1.266.10.8	26-Sep-2010	yamt	locking changes
1.266.10.7	11-Aug-2010	yamt	sync with head.
1.266.10.6	11-Mar-2010	yamt	sync with head
1.266.10.5	18-Jul-2009	yamt	sync with head.
1.266.10.4	24-Jun-2009	yamt	lock vnode when calling VOP_GETATTR because there's no reasonable way for an implementation of VOP_GETATTR to prevent the vnode from being revoked.
1.266.10.3	16-May-2009	yamt	sync with head
1.266.10.2	04-May-2009	yamt	sync with head.
1.266.10.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.266.6.1	17-Jan-2009	mjf	Sync with HEAD.
1.267.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.267.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.268.4.2	29-Dec-2008	christos	fix printf format.
1.268.4.1	19-Nov-2008	christos	file nfs_vnops.c was added on branch christos-time_t on 2008-12-29 00:14:38 +0000
1.269.2.2	23-Jul-2009	jym	Sync with HEAD.
1.269.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.282.4.5	31-May-2011	rmind	sync with head
1.282.4.4	05-Mar-2011	rmind	sync with head
1.282.4.3	03-Jul-2010	rmind	sync with head
1.282.4.2	30-May-2010	rmind	sync with head
1.282.4.1	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.
1.282.2.3	06-Nov-2010	uebayasi	Sync with HEAD.
1.282.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.282.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.289.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.290.2.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.292.2.5	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")
1.292.2.4	16-Jan-2013	yamt	sync with (a bit old) head
1.292.2.3	30-Oct-2012	yamt	sync with head
1.292.2.2	23-May-2012	yamt	sync with head.
1.292.2.1	17-Apr-2012	yamt	sync with head
1.293.4.2	14-Jul-2016	snj	Pull up following revision(s) (requested by hannken in ticket #1363): sys/nfs/nfs_vnops.c: revision 1.309 Return an error if NFSPROC_LOOKUP returns the file handle of the current directory. Treating it as DOT lookup would put garbage into the name cache and could panic on future lookups. Seen with ZFS file system exported from OmniOS, an OpenSolaris derivative. Fixes PR kern/50664 "cd .." over NFS/ZFS can panic kernel
1.293.4.1	12-Aug-2012	martin	branches: 1.293.4.1.4; 1.293.4.1.6; Pull up following revision(s) (requested by manu in ticket #484): sys/fs/nilfs/nilfs_vnops.c: revision 1.18 sys/ufs/ufs/ufs_lookup.c: revision 1.117 sys/nfs/nfs_vnops.c: revision 1.295 sys/ufs/chfs/chfs_vnops.c: revision 1.8 sys/ufs/ext2fs/ext2fs_lookup.c: revision 1.70 sys/fs/unionfs/unionfs_vnops.c: revision 1.6 sys/kern/vfs_cache.c: revision 1.89 sys/fs/efs/efs_vnops.c: revision 1.26 sys/fs/hfs/hfs_vnops.c: revision 1.26 sys/fs/adosfs/adlookup.c: revision 1.16 sys/fs/puffs/puffs_vnops.c: revision 1.168 sys/fs/tmpfs/tmpfs_vnops.c: revision 1.98 sys/fs/ntfs/ntfs_vnops.c: revision 1.52 sys/fs/cd9660/cd9660_lookup.c: revision 1.20 sys/fs/msdosfs/msdosfs_lookup.c: revision 1.24 sys/fs/smbfs/smbfs_vnops.c: revision 1.80 sys/fs/udf/udf_vnops.c: revision 1.72 sys/fs/filecorefs/filecore_lookup.c: revision 1.14 sys/fs/puffs/puffs_node.c: revision 1.25 Move some the test for MAKEENTRY into the cache_enter(9). Make some variables in vfs_cache.c static, __read_mostly, etc. No objection on tech-kern@.
1.293.4.1.6.1	14-Jul-2016	snj	Pull up following revision(s) (requested by hannken in ticket #1363): sys/nfs/nfs_vnops.c: revision 1.309 Return an error if NFSPROC_LOOKUP returns the file handle of the current directory. Treating it as DOT lookup would put garbage into the name cache and could panic on future lookups. Seen with ZFS file system exported from OmniOS, an OpenSolaris derivative. Fixes PR kern/50664 "cd .." over NFS/ZFS can panic kernel
1.293.4.1.4.1	14-Jul-2016	snj	Pull up following revision(s) (requested by hannken in ticket #1363): sys/nfs/nfs_vnops.c: revision 1.309 Return an error if NFSPROC_LOOKUP returns the file handle of the current directory. Treating it as DOT lookup would put garbage into the name cache and could panic on future lookups. Seen with ZFS file system exported from OmniOS, an OpenSolaris derivative. Fixes PR kern/50664 "cd .." over NFS/ZFS can panic kernel
1.293.2.1	29-Apr-2012	mrg	sync to latest -current.
1.295.2.4	03-Dec-2017	jdolecek	update from HEAD
1.295.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.295.2.2	23-Jun-2013	tls	resync from head
1.295.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.299.6.1	18-May-2014	rmind	sync with head
1.304.2.1	10-Aug-2014	tls	Rebase.
1.306.4.3	28-Aug-2017	skrll	Sync with HEAD
1.306.4.2	19-Mar-2016	skrll	Sync with HEAD
1.306.4.1	06-Jun-2015	skrll	Sync with HEAD
1.306.2.2	06-Feb-2016	snj	Pull up following revision(s) (requested by hannken in ticket #1094): sys/nfs/nfs_vnops.c: revision 1.309 Return an error if NFSPROC_LOOKUP returns the file handle of the current directory. Treating it as DOT lookup would put garbage into the name cache and could panic on future lookups. Seen with ZFS file system exported from OmniOS, an OpenSolaris derivative. Fixes PR kern/50664 "cd .." over NFS/ZFS can panic kernel
1.306.2.1	19-May-2015	snj	branches: 1.306.2.1.2; Pull up following revision(s) (requested by chs in ticket #769): sys/nfs/nfs_vnops.c: revision 1.308 in nfs_writerpc(), avoid a signed/unsigned problem in computing the number of bytes to back up in the uio when we need to resend a write RPC (eg. after a server crash) on a 64-bit platform. should fix PR 35448.
1.306.2.1.2.1	06-Feb-2016	snj	Pull up following revision(s) (requested by hannken in ticket #1094): sys/nfs/nfs_vnops.c: revision 1.309 Return an error if NFSPROC_LOOKUP returns the file handle of the current directory. Treating it as DOT lookup would put garbage into the name cache and could panic on future lookups. Seen with ZFS file system exported from OmniOS, an OpenSolaris derivative. Fixes PR kern/50664 "cd .." over NFS/ZFS can panic kernel
1.310.12.4	21-Apr-2020	martin	Sync with HEAD
1.310.12.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.310.12.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.310.12.1	10-Jun-2019	christos	Sync with HEAD
1.310.10.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.310.4.1	12-Dec-2023	martin	Pull up following revision(s) (requested by schmonz in ticket #1927): sys/nfs/nfs_vnops.c: revision 1.325 NFS client: fix interop with macOS 14 servers. Symptom: a bunch of "Cannot open `.' (Invalid argument)". thorpej@ analysis and fix: on the first request to read a given directory, make sure READDIR and READDIRPLUS cookie verifiers are being set to 0. This is in RFC1813 and macOS must have gotten stricter about it. Verified on 10.0_RC1/aarch64 to fix the reproducers in PR kern/57691 as well as the original use case in which I met the bug: pkg_rr once again runs to completion.
1.311.4.1	11-Dec-2023	martin	Pull up following revision(s) (requested by schmonz in ticket #1778): sys/nfs/nfs_vnops.c: revision 1.325 NFS client: fix interop with macOS 14 servers. Symptom: a bunch of "Cannot open `.' (Invalid argument)". thorpej@ analysis and fix: on the first request to read a given directory, make sure READDIR and READDIRPLUS cookie verifiers are being set to 0. This is in RFC1813 and macOS must have gotten stricter about it. Verified on 10.0_RC1/aarch64 to fix the reproducers in PR kern/57691 as well as the original use case in which I met the bug: pkg_rr once again runs to completion.
1.312.2.1	29-Feb-2020	ad	Sync with head.
1.313.4.1	20-Apr-2020	bouyer	Sync with HEAD
1.317.6.1	01-Aug-2021	thorpej	Sync with HEAD.
1.324.4.1	11-Dec-2023	martin	Pull up following revision(s) (requested by schmonz in ticket #490): sys/nfs/nfs_vnops.c: revision 1.325 NFS client: fix interop with macOS 14 servers. Symptom: a bunch of "Cannot open `.' (Invalid argument)". thorpej@ analysis and fix: on the first request to read a given directory, make sure READDIR and READDIRPLUS cookie verifiers are being set to 0. This is in RFC1813 and macOS must have gotten stricter about it. Verified on 10.0_RC1/aarch64 to fix the reproducers in PR kern/57691 as well as the original use case in which I met the bug: pkg_rr once again runs to completion.
1.4	08-Jun-1994	mycroft	Clean up deleted files.
1.3	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.2	20-Apr-1993	mycroft	Add consistent multiple-inclusion protection (repeat).
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
1.33	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.32	21-May-2015	rtr	branches: 1.32.54; change nfs_boot_sendrecv to take sockaddr_in * instead of mbuf * fixes m_serv (single mbuf leak) leak in kern/subr_tftproot.c
1.31	27-Mar-2015	hikaru	m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.30	04-Oct-2010	cyber	branches: 1.30.14; 1.30.18; 1.30.34; 1.30.36; Add support to honor MTU settings from DHCP during netboot. Defines IP_MIN_MTU as 576. Glanced over quickly by martin@ and joerg@.
1.29	27-Oct-2008	cegger	branches: 1.29.12; 1.29.14; change nfs boot behaviour to automatically try next boot method if boot information are incomplete to succeed. That way, it is possible combine static and dhcp boot: For example, to boot diskless you can specify the nfs-server and the rootpath statically. All other information will be taken via dhcp. Patch has been presented on port-xen, tech-kern and tech-net: http://mail-index.netbsd.org/port-xen/2008/10/24/msg004488.html http://mail-index.netbsd.org/tech-kern/2008/10/24/msg003255.html http://mail-index.netbsd.org/tech-net/2008/10/24/msg000864.html No comments, no objections.
1.28	24-Oct-2008	cegger	branches: 1.28.2; - ansify function definition - de- __P - u_int32_t -> uint32_t No functional changes.
1.27	28-Apr-2008	martin	branches: 1.27.6; Remove clause 3 and 4 from TNF licenses
1.26	08-Jul-2007	bouyer	branches: 1.26.28; 1.26.30; 1.26.32; Add a new BOOTSTATIC flag, NFS_BOOTSTATIC_NOSTATIC, which causes nfs_bootstatic() to abort with EOPNOTSUPP. This allows a callback to say that there is no bootstatic config, and the next NFS boot method should be tried.
1.25	08-May-2007	manu	Add the TFTPROOT kernel option for TFTP'ing root RAMdisk at root mount time. This allows working around situations where a kernel with embedded RAMdisk cannot be booted by the bootloader because the RAMdisk is too big.
1.24	11-Dec-2005	christos	branches: 1.24.24; 1.24.26; 1.24.30; 1.24.32; merge ktrace-lwp.
1.23	22-May-2004	jonathan	branches: 1.23.12; Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.22	01-May-2004	matt	Commons are not allowed in header files. extern it.
1.21	11-Mar-2004	cl	Add static nfs boot configuration, from the kernel config file or from a driver selectable callback function. This is used in the Xen port to allow controlling the domain's network setup from the domain building environment at domain creation (vs. having to maintain/change this on a dhcp server). The Xen network driver parses a command line passed in from the domain builder.
1.20	29-Jun-2003	fvdl	branches: 1.20.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.19	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
1.18	05-May-2003	yamt	keep things not needed by userland in #ifdef _KERNEL. (e.g. prototypes for in-kernel functions)
1.17	01-Dec-2002	matt	Add multiple inclusion protection.
1.16	21-Feb-1999	drochner	branches: 1.16.20; restructure the diskless NFS boot code to keep track of the used interface and the address allocated, to roll everything back if the mount fails: -put an interface pointer into "struct nfs_diskless" to have it available for cleanup, don't pass it around anymore where the "struct nfs_diskless" is already passed -add a "cleanup" function which shuts the interface down -in the protocol-specific parts, either return with "everything ready" or "completely shut down" -use common functions for interface initialization and shutdown -add a function to delete all routes associate to an interface (why is this necessary and not done by ~IFF_UP?) g/c diskless swap stuff general cleanup
1.15	30-Sep-1997	drochner	Factor out some functions used by bootparam and DHCP boot.
1.14	09-Sep-1997	gwr	Move the call to nfs_boot_getfh() from nfs_vfsops.c to nfs_boot.c (just for better isolation - it can now be static)
1.13	29-Aug-1997	gwr	Supporting changes for the new BOOTP support in nfs_mountroot.
1.12	14-Aug-1997	drochner	1. Allow to set a netmask (option NFS_BOOT_NETMASK) for the booting interface. Without this, NFS_BOOT_NETMASK could be useless in subnetting envirinment. 2. Comment out unneeded NFS swap related stuff. Closes PR kern/3918.
1.11	27-May-1997	gwr	branches: 1.11.4; Minor reorganization of nfs_mountroot code to simplify BOOTP support. The RPC/bootparamd calls to get the root and swap paths are now done in nfs_boot_init() instead of nfs_boot_getfh(), so the latter now just does the RPC/mountd call. Also changed some panics into error returns.
1.10	20-Oct-1996	fvdl	Enhancements from Matthias Drochner: - Try V3 first for diskless booting. Fall back to V2 if V3 fails. - optionally (option NFS_BOOT_TCP) try a TCP mount first for diskless booting. Fall back to UDP if it fails. - Enable switching between UDP and TCP for remounts.
1.9	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.8	13-Feb-1996	gwr	Do the RPC to bootparamd a little later (just before the mountd call) so that we do not ask for the "swap" path when swapping on disk.
1.7	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.6	13-Jun-1994	gwr	New diskless boot code (uses RARP, bootparamd).
1.5	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.4	07-Jul-1993	cgd	changes from ws to support diskless booting... these are "OK" on inspection and after testing... (actually, currently, none of the changed code is even used...)
1.3	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.2	20-Apr-1993	cgd	re-merged include file changes which got eaten by crash
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.11.4.4	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.11.4.3	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.11.4.2	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.11.4.1	23-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.16.20.1	11-Dec-2002	thorpej	Sync with HEAD.
1.20.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.20.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.20.2.2	03-Aug-2004	skrll	Sync with HEAD
1.20.2.1	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.
1.23.12.2	03-Sep-2007	yamt	sync with head.
1.23.12.1	21-Jun-2006	yamt	sync with head.
1.24.32.1	11-Jul-2007	mjf	Sync with head.
1.24.30.2	15-Jul-2007	ad	Sync with head.
1.24.30.1	08-Jun-2007	ad	Sync with head.
1.24.26.1	17-May-2007	yamt	sync with head.
1.24.24.1	13-May-2007	jdc	Pull up revision 1.25 (requested by manu in ticket #635). Add the TFTPROOT kernel option for TFTP'ing root RAMdisk at root mount time. This allows working around situations where a kernel with embedded RAMdisk cannot be booted by the bootloader because the RAMdisk is too big.
1.26.32.3	09-Oct-2010	yamt	sync with head
1.26.32.2	04-May-2009	yamt	sync with head.
1.26.32.1	16-May-2008	yamt	sync with head.
1.26.30.1	18-May-2008	yamt	sync with head.
1.26.28.2	17-Jan-2009	mjf	Sync with HEAD.
1.26.28.1	02-Jun-2008	mjf	Sync with HEAD.
1.27.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.28.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.29.14.1	05-Mar-2011	rmind	sync with head
1.29.12.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.30.36.2	06-Jun-2015	skrll	Sync with HEAD
1.30.36.1	06-Apr-2015	skrll	Sync with HEAD
1.30.34.1	06-Apr-2015	snj	Pull up following revision(s) (requested by hikaru in ticket #656): sys/kern/subr_tftproot.c: revision 1.14 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/nfs_bootdhcp.c: revision 1.53 sys/nfs/nfsdiskless.h: revision 1.31 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.30.18.1	03-Dec-2017	jdolecek	update from HEAD
1.30.14.1	16-Apr-2015	msaitoh	Pull up following revision(s) (requested by hikaru in ticket #1287): sys/kern/subr_tftproot.c: revision 1.14 via patch sys/nfs/nfsdiskless.h: revision 1.31 sys/nfs/nfs_boot.c: revision 1.82 sys/nfs/krpc_subr.c: revision 1.39 sys/nfs/nfs_bootdhcp.c: revision 1.53 m_pullup() is called in rcvproc callback functions, so nfs_boot_sendrecv() should keep track of the head of mbuf chain. fixes kern/48746
1.32.54.1	02-Aug-2025	perseant	Sync with HEAD
1.4	08-Jun-1994	mycroft	Clean up deleted files.
1.3	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.2	20-Apr-1993	mycroft	Add consistent multiple-inclusion protection (repeat).
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
1.59	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.58	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.57	23-Mar-2023	riastradh	branches: 1.57.6; nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. XXX pullup-8 XXX pullup-9 XXX pullup-10
1.56	23-Mar-2023	riastradh	nfs: Use unsigned fhlen so we don't trip over negative values. XXX pullup-8 XXX pullup-9 XXX pullup-10
1.55	12-Aug-2021	andvar	branches: 1.55.4; s/directry/directory/
1.54	04-Apr-2020	mlelstv	NFSv2 is limited to use only 32bit in metadata. Prevent that larger metadata values are simply truncated. -> clamp filesystem block counts to signed 32bit. -> clamp file sizes to signed 32bit () Some NFSv2 clients also have problems to handle buffer sizes larger than (signed) 16bit. -> clamp buffer sizes to signed 16bit for better compatibility. () This can lead to erroneous behaviour for files larger than 2GB that NFSv2 cannot handle but it is still better than before. An alternative would be to (partially) reject operations on files larger than 2GB, but which causes other problems.
1.53	14-Sep-2013	martin	branches: 1.53.22; 1.53.30; 1.53.34; Backout wildcard pragma to kill warnings and instead sprinkle a few dozen __unused attributes. Requested by joerg@
1.52	14-Sep-2013	martin	Silence gcc 4.8.1 warnings
1.51	10-Apr-2009	bouyer	PR kern/41158: nfs_rename() locking against myself nfsrv_rename() can exit without calling genfs_renamelock_exit() because the nfsm_reply() can do return (0) on error. Change nfsm_reply to use 'error = 0; goto nfsmout' instead. Fix a few place so it's safe to goto nfsmout from nfsm_reply, or other macros calling it. As a side effect it could fix a missing vrele(dirp) in various place where nfsm_reply could return(0).
1.50	04-Mar-2007	christos	branches: 1.50.40; 1.50.50; 1.50.52; 1.50.56; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.49	22-Feb-2007	thorpej	TRUE -> true, FALSE -> false
1.48	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.47	02-Sep-2006	yamt	branches: 1.47.8; nfsd: deal with variable-sized filehandles.
1.46	08-Aug-2006	yamt	nfsm_srvfhtom: ensure that padding bytes in nfsv2 file handles are zero.
1.45	07-Jun-2006	kardel	branches: 1.45.4; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.44	11-Dec-2005	christos	branches: 1.44.4; 1.44.6; 1.44.8; 1.44.14; merge ktrace-lwp.
1.43	31-Oct-2005	thorpej	Fix paste-o in the NFSV3SATTRTIME_TOSERVER case of mtime handing (need to set va_mtime, not va_atime).
1.42	01-Oct-2005	yamt	branches: 1.42.2; nfsm_srvsattr: use nanotime(9) rather than time(9) for NFSV3SATTRTIME_TOSERVER.
1.41	29-May-2005	christos	branches: 1.41.2; - sprinkle const - avoid shadowed variables - mark bad const use with XXXUNCONST
1.40	26-Feb-2005	perry	nuke trailing whitespace
1.39	19-Jan-2005	yamt	branches: 1.39.2; implement inaccurate mtime/ctime detection. namely, if mtime or ctime are same between pre_op_attr and post_op_attr when we expected them to be changed, don't trust the server.
1.38	29-Sep-2004	yamt	branches: 1.38.4; g/c NFSMINOFF, which is unused and identical with MRESETDATA.
1.37	10-May-2004	yamt	don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.36	05-Apr-2004	yamt	nfsm_mtofh: handle the case that filehandle is exist but fattr is not.
1.35	05-Apr-2004	yamt	nfsm_wcc_data: update n_ctime and n_nctime if no one other than us changed the file in the meantime so that we won't invalidate caches unnecessarily due to our own activities.
1.34	19-Mar-2004	yamt	branches: 1.34.2; comments on some nfsm_ macros.
1.33	15-Mar-2004	yamt	some comments on cryptic nfsm_ macros.
1.32	26-Sep-2003	yamt	change n_mtime from time_t to timespec in order to improve cache consistency. (1 second granularity is too loose these days.)
1.31	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.30	29-Jun-2003	fvdl	branches: 1.30.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.29	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
1.28	09-Jun-2003	yamt	rework zero padding of rpc reply. - for READ procedure, don't send back more bytes than requested. - don't have doubtful assumptions on mbuf chain structure. - rename a function (nfsm_adj -> nfs_zeropad) to avoid confusion as the semantics of the function was changed.
1.27	06-May-2003	yamt	remove nfsm_srvstrsiz as it's no longer used.
1.26	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.25	28-Mar-2003	yamt	reply ENAMETOOLONG properly instead of discarding request as BADRPC. my own PR20791.
1.24	26-Feb-2003	matt	Fix typo.
1.23	26-Feb-2003	matt	Add MBUFTRACE kernel option. Do a little mbuf rework while here. Change all uses of MGET(, M_WAIT, ) to m_get(M_WAIT, *). These are not performance critical and making them call m_get saves considerable space. Add m_clget analogue of MCLGET and make corresponding change for M_WAIT uses. Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE. Begin to change netstat to use sysctl.
1.22	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.21	03-Apr-2002	wrstuden	In the SETATTR code, if the changes to a & m time are exclusively set via NFSV3SATTRTIME_TOSERVER and not NFSV3SATTRTIME_TOCLIENT, add VA_UTIMES_NULL to the va_vflags. This reflects our policy where we're much more liberal about who can set a & m times to 'now' than we are about who can set them to a specific time. Should close PR 15597 from Martin Husemann. Patch is based on the one Matthias Drochner gave in the PR.
1.20	29-May-1999	fvdl	branches: 1.20.14; 1.20.16; Be more correct with attribute structures for setattr RPCs and friends, so that picky servers (e.g. Solaris 7) don't refuse our requests. Move some code into a macro, and a bit of KNF. From OpenBSD.
1.19	06-Mar-1999	fair	branches: 1.19.2; 1.19.4; 1.19.6; Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.18	09-Aug-1998	perry	bzero->memset, bcopy->memcpy, bcmp->memcmp
1.17	14-Jul-1997	fvdl	Don't assume that pointers into mbuf data remain valid across nfsm_dissect. In readdirplus, don't keep such pointers but store the file attributes in a variable instead until they are needed. Change nfsm_loadattr* a bit so it can accept a direct pointer to an nfs_fattr structure.
1.16	24-Jun-1997	fvdl	Let nfsm_srvmtofh deal with the public filehandle, convert to all zeroes for both v2 and v3 internally.
1.15	27-Mar-1997	thorpej	Don't assume mbuf external storage is MCLBYTES.
1.14	24-Feb-1997	fvdl	Use ALIGNED_POINTER to see whether mbuf data needs to be realigned.
1.13	22-Feb-1997	fvdl	Cast pointer to u_long, not int, when doing the alignment check. Fixes warnings on the Alpha. Needs a better solution.
1.12	22-Feb-1997	fvdl	Fixes from BSDI (thanks go to Keith Bostic). Original RCS message: date: 1997/02/10 18:41:15; author: cp; state: Exp; lines: +8 -2 Make nfs_realign go away on sparc and add functionality to nfsm_disct. === [XXX this introduces an ifdef __i386__, see the comment. Should be changed]
1.11	25-Oct-1996	cgd	branches: 1.11.4; make the namei struct members ni_dirp and ni_next, and the componentname struct member cn_nameptr 'const', since they should never be used to modify the path name. (Only the pathname buffer, cn_pnbuf, should be modified.) Propagate the const poisoning to code that uses the namei and componentname structs.
1.10	20-Mar-1996	fvdl	Make sure not to free the reply mbuf twice. Should fix PR #2240
1.9	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.8	09-Feb-1996	christos	nfs prototype changes
1.7	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.6	23-May-1995	mycroft	Remove gratuitous extra indirections.
1.5	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.4	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.3	03-Jun-1993	cgd	fix for macklem's bogus use of the va_flags field, supplied by John Woods, jfwfrom: @ksr.com. also, fixes the following problems: the va_gen field is in a similar position (Suns are going to be reporting the change-date microseconds as their "generation"), I've supplied my own set of diffs below for your inspection. Note these aren't even compiled, but they're pretty similar to what I had to do to our older version of OSF/1 here. (There's also an unrelated change supplied for xdr_subs.h; the pointer types supplied to the fxdr_time() and txdr_time() macros are not, in fact, both struct timevals. That turns out to be one of many tips-of-the-iceberg facing those porting the (old) Berkeley NFS code to 64-bit machines...)
1.2	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.1	20-Apr-1993	mycroft	branches: 1.1.1; Restore files lost during crash.
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.11.4.1	12-Mar-1997	is	Merge in changes from Trunk
1.19.6.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.19.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.19.2.1	22-Jun-1999	perry	pullup 1.19->1.20 (fvdl): fix file creation with a Solaris 7 server
1.20.16.1	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.20.14.2	22-Oct-2002	thorpej	Sync with HEAD.
1.20.14.1	17-Apr-2002	nathanw	Catch up to -current.
1.30.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.30.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.30.2.6	24-Jan-2005	skrll	Sync with HEAD.
1.30.2.5	19-Oct-2004	skrll	Sync with HEAD
1.30.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.30.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.30.2.2	03-Aug-2004	skrll	Sync with HEAD
1.30.2.1	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.
1.34.2.3	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.34.2.2	10-Jul-2004	tron	branches: 1.34.2.2.2; Pull up revision 1.36 (requested by tls in ticket #634): nfsm_mtofh: handle the case that filehandle is exist but fattr is not.
1.34.2.1	10-Jul-2004	tron	Pull up revision 1.35 (requested by tls in ticket #634): nfsm_wcc_data: update n_ctime and n_nctime if no one other than us changed the file in the meantime so that we won't invalidate caches unnecessarily due to our own activities.
1.34.2.2.2.1	11-Jan-2005	jmc	Pullup patch (requested by yamy in ticket #1078) Don't do kludge for a reply to a retransmitted request unless we actually retransmitted the request.
1.38.4.1	29-Apr-2005	kent	sync with -current
1.39.2.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.41.2.4	03-Sep-2007	yamt	sync with head.
1.41.2.3	26-Feb-2007	yamt	sync with head.
1.41.2.2	30-Dec-2006	yamt	sync with head.
1.41.2.1	21-Jun-2006	yamt	sync with head.
1.42.2.1	02-Nov-2005	yamt	sync with head.
1.44.14.1	19-Jun-2006	chap	Sync with head.
1.44.8.3	03-Sep-2006	yamt	sync with head.
1.44.8.2	11-Aug-2006	yamt	sync with head
1.44.8.1	26-Jun-2006	yamt	sync with head.
1.44.6.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.44.4.1	09-Sep-2006	rpaulo	sync with head
1.45.4.1	16-Aug-2006	tron	Pull up following revision(s) (requested by yamt in ticket #24): sys/nfs/nfsm_subs.h: revision 1.46 nfsm_srvfhtom: ensure that padding bytes in nfsv2 file handles are zero.
1.47.8.2	12-Mar-2007	rmind	Sync with HEAD.
1.47.8.1	28-Feb-2007	yamt	sync with head. (somehow missed in the previous)
1.50.56.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.52.1	13-Apr-2009	snj	Pull up following revision(s) (requested by ad in ticket #700): sys/nfs/nfs_serv.c: revision 1.144 sys/nfs/nfsm_subs.h: revision 1.51 PR kern/41158: nfs_rename() locking against myself nfsrv_rename() can exit without calling genfs_renamelock_exit() because the nfsm_reply() can do return (0) on error. Change nfsm_reply to use 'error = 0; goto nfsmout' instead. Fix a few place so it's safe to goto nfsmout from nfsm_reply, or other macros calling it. As a side effect it could fix a missing vrele(dirp) in various place where nfsm_reply could return(0).
1.50.50.1	28-Apr-2009	skrll	Sync with HEAD.
1.50.40.1	04-May-2009	yamt	sync with head.
1.53.34.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #1617): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.53.30.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.53.22.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #1810): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.55.4.1	30-Mar-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #134): sys/nfs/nfs_serv.c: revision 1.184 sys/nfs/nfs_srvsubs.c: revision 1.17 sys/nfs/nfsm_subs.h: revision 1.56 sys/nfs/nfsm_subs.h: revision 1.57 nfs: Use unsigned fhlen so we don't trip over negative values. nfs: Avoid integer overflow in nfs_namei bounds check. nfs: Use unsigned name lengths so we don't trip over negative ones. - nfsm_strsiz is only used with uint32_t in callers, but let's not leave it as a rake to step on. - nfsm_srvnamesiz is abused with signed s. The internal conversion to unsigned serves to reject both negative and too-large values in such callers. XXX Should make all callers use unsigned, rather than flipping back and forth between signed and unsigned for name lengths. nfs: Avoid free of uninitialized on bad name size in create, mknod. XXX These error branches are a nightmare and need to be more systematically cleaned up. Even if they are correct now, they are impossible to audit and extremely fragile in case anyone ever needs to make other changes to them.
1.57.6.1	02-Aug-2025	perseant	Sync with HEAD
1.54	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.53	15-Jul-2015	manu	branches: 1.53.54; Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.52	30-May-2014	hannken	branches: 1.52.2; 1.52.4; Change NFS from rbtree to vcache.
1.51	22-Jan-2011	matt	branches: 1.51.14; 1.51.28; Add the ability to mount NFS filesystems in COMPAT_NETBSD32 If in the kernel and NFS_ARGS_ONLY, just export struct nfs_args and its flags.
1.50	25-Sep-2010	matt	branches: 1.50.2; 1.50.4; Rename rb.h to rbtree.h, as it is more appropriate (c.f. ptree.h). Also helps find code that hasn't been updated to use the new rbtree API.
1.49	14-Mar-2009	dsl	branches: 1.49.2; 1.49.4; Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.48	22-Oct-2008	matt	branches: 1.48.2; 1.48.8; Don't need nfs_vfs_reinit anymore since we don't resize tables anymore. Move reinit code to init case.
1.47	22-Oct-2008	matt	Change NFS to use a RB-tree for its FH->nfsnode lookups.
1.46	31-Jul-2007	pooka	branches: 1.46.24; 1.46.28; 1.46.34; 1.46.36; * nuke the nameidata parameter from VFS_MOUNT(). Nobody on tech-kern knew what it was supposed to be used for and wrstuden gave a go-ahead * while rototilling, convert file systems which went easily to use VFS_PROTOS() instead of manually prototyping the methods
1.45	12-Jul-2007	dsl	branches: 1.45.2; Change the VFS_MOUNT() interface so that the 'data' buffer passed to the fs code is a kernel buffer, pass though the length of the buffer as well. Since the length of the userspace buffer isn'it (yet) passed through the mount system call, add a field to the vfsops structure containing the default length. Split sys_mount() for calls from compat code. Ride one of the recent kernel version changes - old fs LKMs will load, but sys_mount() will reject any attempt to use them.
1.44	29-Apr-2007	yamt	use condvar.
1.43	29-Apr-2007	yamt	use mutex and condver.
1.42	15-Feb-2007	yamt	branches: 1.42.2; 1.42.6; 1.42.8; use mutex and rwlock rather than lockmgr.
1.41	27-Dec-2006	yamt	remove nqnfs.
1.40	13-Jul-2006	martin	branches: 1.40.4; Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.39	14-May-2006	elad	branches: 1.39.4; integrate kauth.
1.38	14-Apr-2006	blymn	Make i/o statistics collection more generic, include tape drives and nfs mounts in the set of devices that statistics will be reported on.
1.37	11-Dec-2005	christos	branches: 1.37.4; 1.37.6; 1.37.8; 1.37.10; 1.37.12; merge ktrace-lwp.
1.36	25-Nov-2005	thorpej	Use a once control to initialize the NFS server / client shared data from nfs_vfs_init() or sys_nfssvc(). Remove the nfs_init() call from main().
1.35	23-Sep-2005	jmmv	branches: 1.35.6; Apply the NFS exports list rototill patch: - Remove all NFS related stuff from file system specific code. - Drop the vfs_checkexp hook and generalize it in the new nfs_check_export function, thus removing redundancy from all file systems. - Move all NFS export-related stuff from kern/vfs_subr.c to the new file sys/nfs/nfs_export.c. The former was becoming large and its code is always compiled, regardless of the build options. Using the latter, the code is only compiled in when NFSSERVER is enabled. While doing this, also make some functions in nfs_subs.c conditional to NFSSERVER. - Add a new command in nfssvc(2), called NFSSVC_SETEXPORTSLIST, that takes a path and a set of export entries. At the moment it can only clear the exports list or append entries, one by one, but it is done in a way that allows setting the whole set of entries atomically in the future (see the comment in mountd_set_exports_list or in doc/TODO). - Change mountd(8) to use the nfssvc(2) system call instead of mount(2) so that it becomes file system agnostic. In fact, all this whole thing was done to remove a 'XXX' block from this utility! - Change the mount, newfs and fsck userland utilities to not deal with NFS exports initialization; done internally by the kernel when initializing the NFS support for each file system. - Implement an interface for VFS (called VFS hooks) so that several kernel subsystems can run arbitrary code upon receipt of specific VFS events. At the moment, this only provides support for unmount and is used to destroy NFS exports lists from the file systems being unmounted, though it has room for extension. Thanks go to yamt@, chs@, thorpej@, wrstuden@ and others for their comments and advice in the development of this patch.
1.34	18-Sep-2005	christos	Allow turning off the attribute cache.
1.33	19-Jan-2005	yamt	branches: 1.33.6; 1.33.8; implement inaccurate mtime/ctime detection. namely, if mtime or ctime are same between pre_op_attr and post_op_attr when we expected them to be changed, don't trust the server.
1.32	22-May-2004	jonathan	branches: 1.32.4; Eliminate several uses of `curproc' from the socket-layer code and from NFS. Add a new explicit `struct proc p' argument to socreate(), sosend(). Use that argument instead of curproc. Follow-on changes to pass that argument to socreate(), sosend(), and (so->so_send)() calls. These changes reviewed and independently recoded by Matt Thomas. Changes to soreceive() and (dom->dom_exernalize() from Matt Thomas: pass soreceive()'s struct uio uio->uio_procp to unp_externalize(). Eliminate curproc from unp_externalize. Also, now soreceive() uses its uio->uio_procp value, pass that same value downward to ((pr->pru_usrreq)() calls for consistency, instead of (struct proc * )0. Similar changes in sys/nfs to eliminate (most) uses of curproc, either via the req-> r_procp field of a struct nfsreq req argument, or by passing down new explicit struct proc arguments. Reviewed by: Matt Thomas, posted to tech-kern. NB: The (*pr->pru_usrreq)() change should be tested on more (all!) protocols.
1.31	27-Apr-2004	jrf	First pass for some caddr_t removal and changes to get rid of it where we no longer use and/or need it - removed casts from unionfs, deadfs and fdesc (there are more to hunt down still) - changed vfs_quotactl args argumet from caddr_t to void * - changed vfs_quotactl structures/callers to reflect the api change Compiled fine and ran for about a day. Approved/reviewed by christos@netbsd.org and gimpy@netbsd.org.
1.30	21-Apr-2004	christos	Replace the statfs() family of system calls with statvfs(). Retain binary compatibility.
1.29	03-Oct-2003	yamt	branches: 1.29.4; terminate snprintb 'new' format strings correctly. (fixes overrun in mount_*)
1.28	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.27	29-Jun-2003	fvdl	branches: 1.27.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.26	29-Jun-2003	thorpej	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.25	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
1.24	03-May-2003	yamt	better handling of write verifier change.
1.23	09-Apr-2003	yamt	rename nm_verf to nm_writeverf because it's confusing with nm_verf{str,len}.
1.22	21-Sep-2002	christos	MNT_GETARGS support
1.21	15-Sep-2001	chs	add a new VFS op, vfs_reinit, which is called when desiredvnodes is adjusted via sysctl. file systems that have hash tables which are sized based on the value of this variable now resize those hash tables using the new value. the max number of FFS softdeps is also recalculated. convert various file systems to use the <sys/queue.h> macros for their hash tables.
1.20	12-Feb-2001	fvdl	branches: 1.20.2; 1.20.4; 1.20.6; Instead of storing the filehandle in the mount structure, store the vnode pointer. This avoids a locking problem with nfs_nget, and can be done because we always have a reference on the root vnode of the filesystem.
1.19	16-Mar-2000	jdolecek	Add new VFS op routine - vfs_done and call it on filesystem detach in vfs_detach(). vfs_done may free global filesystem's resources, typically those allocated in respective filesystem's init function. Needed so those filesystems which went in via LKM have a chance to clean after themselves before unloading. This fixes random panics when LKM for filesystem using pools was loaded and unloaded several times. For each leaf filesystem, add appropriate vfs_done routine.
1.18	04-Jul-1999	sommerfeld	branches: 1.18.2; kern/5591: Fix race in the NFS socket code during umount -f and system shutdown: During an unmount, wake up all the processes which are waiting to lock the socket for receive, and wait for them (and the process blocked in soreceive, if any) to go away before blowing away the socket and the mount structure.
1.17	26-Feb-1999	wrstuden	branches: 1.17.2; 1.17.4; Modify vfsops to seperate vfs_fhtovp() into two routines. vfs_fhtovp() now only handles the file handle to vnode conversion, and a new call, vfs_checkexp(), performs the export verification.
1.16	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.15	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.14	17-Jul-1997	fvdl	branches: 1.14.2; * Deal with servers that don't give complete FSINFO (like NT) From Olaf Seibert <rhialto@polder.ubc.kun.nl> (PR 3687) * Make an attempt to check the maximum filesize before attempting a write to the server, as write RPCs will typically happen asynchronously, and the process will not see the error. Fixes problems with unexpectly truncated files at 4G * Pass up errors in nfs_writerpc correctly
1.13	22-Dec-1996	cgd	Change the second and third args to struct vfsops' (vfs_mount)() to 'const char ', and 'void ', respectively. The second arg is taken directly from user arguments, and is const there, so must be const in the prototypes and functions. The third arg is also taken directly from user arguments. It doesn't have to be changed, but since it's cleaner to keep the type the same as the user arg's type, and I'm already making the 'const char ' change...
1.12	03-Dec-1996	thorpej	Make NFSSERVER work without NFSCLIENT. This is achieved by splitting the client and server/shared data initialization into separate functions, and calling the server/shared initialization directly from main(). Problem noted in PR #1308 (Kenneth Stailey) and PR #1780 (Chris Demetriou). Fix suggested in PR #1780 by Chris Demetriou, and munged a bit by me, and OK'd by Frank van der Linden <fvdl@netbsd.org>.
1.11	02-Dec-1996	thorpej	NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.10	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.9	09-Feb-1996	christos	nfs prototype changes
1.8	26-Mar-1995	jtc	KERNEL -> _KERNEL
1.7	13-Dec-1994	mycroft	Sync with CSRG.
1.6	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.5	29-Jun-1994	cgd	branches: 1.5.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.4	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.3	27-Mar-1994	cgd	expand uid_t/gid_t/off_t
1.2	20-May-1993	cgd	branches: 1.2.4; more rcs id adding and header cleanup. i like vi macros!
1.1	20-Apr-1993	mycroft	branches: 1.1.1; Restore files lost during crash.
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.2.4.1	24-Sep-1993	mycroft	Make all files using spl*() #include cpu.h. Changes from trunk. nfs_vfsops.c, nfsmount.h: Make nfs_quotactl() take an int rather than a uid_t, as it might be -1. nfs_vnops.c: va_size and va_bytes are now quads.
1.5.2.1	19-Aug-1994	mycroft	update from trunk
1.14.2.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.17.4.1	02-Aug-1999	thorpej	Update from trunk.
1.17.2.1	05-Nov-1999	cgd	pull up rev 1.18 from trunk (requested by fvdl): Avoid a panic when forcibly unmounting a hung NFS mount, e.g. at reboot.
1.18.2.2	12-Mar-2001	bouyer	Sync with HEAD.
1.18.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.20.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.20.4.2	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
1.20.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.20.2.2	06-Oct-2002	thorpej	Sync with HEAD.
1.20.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.27.2.9	11-Dec-2005	christos	Sync with head.
1.27.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.27.2.7	24-Jan-2005	skrll	Sync with HEAD.
1.27.2.6	30-Oct-2004	skrll	s/p/l/ for the struct lwp * arg.
1.27.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.27.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.27.2.3	24-Aug-2004	skrll	Undo part of the ktrace/lwp changes. In particular: * Remove the "lwp " argument that was added to vget(). Turns out that nothing actually used it! Remove the "lwp " arguments that were added to VFS_ROOT(), VFS_VGET(), and VFS_FHTOVP(); all they did was pass it to vget() (which, as noted above, didn't use it). Remove all of the "lwp *" arguments to internal functions that were added just to appease the above.
1.27.2.2	03-Aug-2004	skrll	Sync with HEAD
1.27.2.1	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.
1.29.4.1	27-Oct-2005	riz	Pull up following revision(s) (requested by christos in ticket #5863): sys/nfs/nfs_subs.c: revision 1.152 via patch sys/nfs/nfs.h: revision 1.49 sys/nfs/nfs_vfsops.c: revision 1.149 via patch usr.sbin/amd/include/config.h: revision 1.36 sys/nfs/nfs_vnops.c: revision 1.227 via patch sys/nfs/nfsmount.h: revision 1.34 Allow the attribute cache to be turned off, and allow amd to do it.
1.32.4.1	29-Apr-2005	kent	sync with -current
1.33.8.4	03-Sep-2007	yamt	sync with head.
1.33.8.3	26-Feb-2007	yamt	sync with head.
1.33.8.2	30-Dec-2006	yamt	sync with head.
1.33.8.1	21-Jun-2006	yamt	sync with head.
1.33.6.1	26-Sep-2005	tron	Pull up following revision(s) (requested by christos in ticket #816): sys/nfs/nfs.h: revision 1.49 sys/nfs/nfsmount.h: revision 1.34 Allow turning off the attribute cache.
1.35.6.1	29-Nov-2005	yamt	sync with head.
1.37.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.37.10.2	19-Apr-2006	elad	sync with head.
1.37.10.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.37.8.2	11-Aug-2006	yamt	sync with head
1.37.8.1	24-May-2006	yamt	sync with head.
1.37.6.2	01-Jun-2006	kardel	Sync with head.
1.37.6.1	22-Apr-2006	simonb	Sync with head.
1.37.4.1	09-Sep-2006	rpaulo	sync with head
1.39.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.40.4.1	12-Jan-2007	ad	Sync with head.
1.42.8.1	11-Jul-2007	mjf	Sync with head.
1.42.6.4	20-Aug-2007	ad	Sync with HEAD.
1.42.6.3	15-Jul-2007	ad	Sync with head.
1.42.6.2	08-Jun-2007	ad	Sync with head.
1.42.6.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.42.2.1	07-May-2007	yamt	sync with head.
1.45.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.46.36.2	31-Jul-2007	pooka	* nuke the nameidata parameter from VFS_MOUNT(). Nobody on tech-kern knew what it was supposed to be used for and wrstuden gave a go-ahead * while rototilling, convert file systems which went easily to use VFS_PROTOS() instead of manually prototyping the methods
1.46.36.1	31-Jul-2007	pooka	file nfsmount.h was added on branch matt-mips64 on 2007-07-31 21:14:20 +0000
1.46.34.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.46.28.4	09-Oct-2010	yamt	sync with head
1.46.28.3	16-Jul-2009	yamt	remove sndlock. it's superseded by nm_solock. suggested by Andrew Doran.
1.46.28.2	04-May-2009	yamt	sync with head.
1.46.28.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.46.24.1	17-Jan-2009	mjf	Sync with HEAD.
1.48.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.48.2.1	28-Apr-2009	skrll	Sync with HEAD.
1.49.4.1	05-Mar-2011	rmind	sync with head
1.49.2.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.50.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.50.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.51.28.1	10-Aug-2014	tls	Rebase.
1.51.14.2	03-Dec-2017	jdolecek	update from HEAD
1.51.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.52.4.1	22-Sep-2015	skrll	Sync with HEAD
1.52.2.1	04-Nov-2015	riz	Pull up following revision(s) (requested by manu in ticket #882): sbin/umount/umount.c: revision 1.48 sys/nfs/nfsmount.h: revision 1.53 sys/nfs/nfs_var.h: revision 1.94 sys/nfs/nfs_iod.c: revision 1.7 sys/nfs/nfs_socket.c: revision 1.197 sys/nfs/nfs_bio.c: revision 1.191 sys/nfs/nfs_vfsops.c: revision 1.230 sys/nfs/nfs_clntsocket.c: revision 1.3 Remove useless and harmful sync(2) call in umount(8) Remove sync(2) call before unmount(2) in umount(8). This sync(2) is useless since unmount(2) will perform a VFS_SYNC anyway. But moreover, this sync(2) may be harmful, as there are some situation where it cannot return (unreachable NFS server, for instance), causing umount -f to be uneffective. Fix soft NFS force unmount For many reasons, forcibly unmounting a soft NFS mount could hang forever. Here are the fixes: - Introduce decents timeouts in operation that awaited NFS server reply. - On timeout, fails operations on soft mounts with EIO. - Introduce NFSMNT_DISMNTFORCE to let the filesystem know that a force unmount is ongoing. This causes timeouts to be reduced and prevents the NFS client to attempt reconnecting to the NFS server. Also fix a race condition where some asynchronous I/O could reference destroyed mount structures. We fix this by awaiting asynchronous I/O to drain before proceeding. Reviewed by Chuck Silvers.
1.53.54.1	02-Aug-2025	perseant	Sync with HEAD
1.77	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.76	21-Oct-2021	andvar	branches: 1.76.10; fix various typos, mainly in comments, but also in man pages and log messages.
1.75	18-Jul-2021	dholland	Abolish all the silly indirection macros for initializing vnode ops tables. These are things of the form #define foofs_op genfs_op, or #define foofs_op genfs_eopnotsupp, or similar. They serve no purpose besides obfuscation, and have gotten cutpasted all over everywhere.
1.74	27-May-2021	simonb	Remove nfs_putpages() prototype; it's not defined anywhere.
1.73	30-May-2014	hannken	branches: 1.73.44; 1.73.46; Change NFS from rbtree to vcache.
1.72	25-Sep-2010	matt	branches: 1.72.18; 1.72.32; Rename rb.h to rbtree.h, as it is more appropriate (c.f. ptree.h). Also helps find code that hasn't been updated to use the new rbtree API.
1.71	14-Mar-2009	dsl	branches: 1.71.2; 1.71.4; Remove all the __P() from sys (excluding sys/dist) Diff checked with grep and MK1 eyeball. i386 and amd64 GENERIC and sys still build.
1.70	02-Jan-2009	christos	branches: 1.70.2; protect sillyrename with _KERNEL
1.69	02-Jan-2009	ad	- Don't vput() a vnode that we do not hold locked. - Eliminate one of the few remaining uses of LK_CANRECURSE.
1.68	22-Oct-2008	matt	branches: 1.68.2; 1.68.4; Change NFS to use a RB-tree for its FH->nfsnode lookups.
1.67	25-Jan-2008	ad	branches: 1.67.6; 1.67.10; 1.67.16; Remove VOP_LEASE. Discussed on tech-kern.
1.66	10-Aug-2007	yamt	branches: 1.66.2; 1.66.8; - instead of scanning an array of iods, maintain a list of idle iods. - make nfs_getset_niothreads MP friendly.
1.65	08-Aug-2007	yamt	push kernel_lock a little.
1.64	20-Jul-2007	yamt	branches: 1.64.4; 1.64.6; - fix decreasing of vfs.nfs.iothreads after the recent partial merge of vmlocking. - don't make nfsiod exit with requests left. - make NFSSVC_BIOD a dummy so that nfsiod can be simplified.
1.63	29-Apr-2007	yamt	branches: 1.63.2; include condvar.h. pointed by Kurt Schreiner.
1.62	29-Apr-2007	yamt	use mutex and condver.
1.61	15-Feb-2007	yamt	branches: 1.61.2; 1.61.6; 1.61.8; use mutex and rwlock rather than lockmgr.
1.60	28-Dec-2006	yamt	remove several nqnfs definitions.
1.59	27-Dec-2006	yamt	remove nqnfs.
1.58	17-Oct-2006	christos	another variable should have been _KERNEL only.
1.57	17-Oct-2006	christos	don't expose kernel variables to userland.
1.56	14-May-2006	elad	branches: 1.56.8; 1.56.10; integrate kauth.
1.55	11-Dec-2005	christos	branches: 1.55.4; 1.55.6; 1.55.8; 1.55.10; 1.55.12; merge ktrace-lwp.
1.54	26-Jan-2005	yamt	branches: 1.54.6; handle a really empty directory, which doesn't have even the dot entry.
1.53	09-Jan-2005	yamt	branches: 1.53.2; 1.53.4; invalidate cache if filesize is changed besides our activity because it means that we're out of sync with the server.
1.52	08-Jan-2005	yamt	nfs_lookup: check n_nctime for positive entries as well to improve cache consistency.
1.51	14-Dec-2004	yamt	redirect some VOPs which shouldn't be used for nfs to genfs_badop (ie. panic).
1.50	26-Oct-2004	yamt	since daddr_t is 64-bit these days, simply use nfs directory cookies as buffer cache indexes. regress/sys/fs/getdents is now supposed to work. fix PR/27112.
1.49	15-Sep-2004	yamt	fix access-after-free bugs in dircache code by refcounting nfsdircache. PR/26864.
1.48	24-Aug-2004	yamt	nfs_request: a workaround for servers doing "maproot". for i/o requests which are expected not to fail due to permission to mimic unix file open semantics (READ, WRITE, COMMIT), try two credentials. namely, the file owner's one and open time one. remember which credential worked in per-file basis and try it first next time to minimize number of retries. ideas from Chuck Silvers. PR/23716 and PR/24987.
1.47	27-May-2004	yamt	remove an unused instance of VOP_UPDATE.
1.46	12-Mar-2004	yamt	branches: 1.46.2; shrink sizeof struct nfsnode by putting exclusive members into union.
1.45	12-Mar-2004	yamt	introduce a macro NFS_INVALIDATE_ATTRCACHE and use it instead of "n_attrstamp = 0".
1.44	26-Sep-2003	yamt	change n_mtime from time_t to timespec in order to improve cache consistency. (1 second granularity is too loose these days.)
1.43	17-Sep-2003	yamt	change nctime to timespec from time_t. there can be too many activities in a second.
1.42	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.41	30-Jul-2003	yamt	vrecycle removed nfs vnodes. not perfect, but enough for most cases.
1.40	07-May-2003	yamt	branches: 1.40.2; simple lock for nfs iod.
1.39	09-Apr-2003	yamt	update a comment to follow the previous change.
1.38	09-Apr-2003	yamt	make per-iod datas together.
1.37	01-Dec-2002	matt	Make sure these all agree on the same definitons of various variables.
1.36	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
1.35	21-Oct-2002	yamt	fix a page locking deadlock problem for nfs. add a flag that specify if the file can be truncated safely or not to nfsm_loadattr and friends. when it isn't safe, just mark the nfsnode as "should be truncated later". ok'ed by Frank van der Linden and Chuck Silvers. close kern/18036.
1.34	15-Sep-2001	chs	a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
1.33	28-May-2001	chs	branches: 1.33.2; 1.33.4; add a genfs_mmap() and change all of the disk-based filesystems to implement VOP_MMAP() with the genfs version, in preparation for actually using this VOP.
1.32	06-Feb-2001	fvdl	branches: 1.32.2; Do actual vnode locking for NFS.
1.31	27-Nov-2000	chs	Initial integration of the Unified Buffer Cache project.
1.30	19-Sep-2000	fvdl	Add fields to deal with commit ranges.
1.29	30-Mar-2000	simonb	branches: 1.29.4; Delete redundant decl of nfs_vget() - it's in <nfs/nfsmount.h>.
1.28	29-Nov-1999	fvdl	Insert an extra VOP_ACCESS check in nfs_lookup, to avoid cached access mishaps for lookup and getattr. Closes PR 8884. While at it, cache access RPCs.
1.27	10-Aug-1998	matthias	branches: 1.27.2; 1.27.6; 1.27.8; 1.27.12; 1.27.18; create miscfs/genfs/genfs_vnops.c:genfs_enoioctl and make all the other filesystems use it instead of a private version.
1.26	25-Jun-1998	thorpej	- Rename nqnfs_vop_lease_check() to genfs_lease_check(). If NFSSERVER is not in the kernel, genfs_lease_check() is simply a no-op. This allows LKM'd file systems to be exported (previously did not work properly due to a compile-time decision based on -DNFSSERVER). - defopt NFSSERVER
1.25	01-Mar-1998	fvdl	Merge with Lite2 + local changes
1.24	19-Oct-1997	fvdl	* Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.23	16-Oct-1997	christos	Fix the location of the NFS_SMALLFH
1.22	12-Oct-1997	fvdl	Do negative lookup caching. Use a timestamp of the oldest negative cache entry, so it can be checked against directory modification time for validity.
1.21	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.20	11-Apr-1997	kleink	branches: 1.20.4; Implement a POSIX compliant genfs VOP_SEEK() and use it in the appropriate places; by Chris G. Demetriou and myself.
1.19	02-Dec-1996	thorpej	NFS performance improvement from Doug Rabson/FreeBSD: Improve the queuing algorithms used by NFS' asynchronous i/o. The existing mechanism uses a global queue for some buffers and the vp->b_dirtyblkhd queue for others. This turns sequential writes into randomly ordered writes to the server, affecting both read and write performance. The existing mechanism also copes badly with hung servers, tending to block accesses to other servers when all the iods are waiting for a hung server. The new mechanism uses a queue for each mount point. All asynchronous i/o goes through this queue which preserves the ordering of requests. A simple mechanism ensures that the iods are shared out fairly between active mount points. Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1.18	07-Sep-1996	mycroft	Implement poll(2).
1.17	01-Sep-1996	mycroft	Add a set of generic file system operations that most file systems use. Also, fix some time stamp bogosities.
1.16	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.15	09-Feb-1996	christos	nfs prototype changes
1.14	26-Mar-1995	jtc	KERNEL -> _KERNEL
1.13	13-Dec-1994	mycroft	Sync with CSRG.
1.12	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.11	29-Jun-1994	cgd	branches: 1.11.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.10	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.9	25-Apr-1994	cgd	some prototype cleanup, eliminate/replace bogus types (e.g. quad and u_quad) -> use better types (e.g. quad_t & u_quad_t in inodes), some cleanup.
1.8	21-Apr-1994	cgd	blow away all vestiges of nfsnode locking. (1) it's unnecessary (2) it causes machines to hang (yup!) (3) it'd be gone in a few days anyway (it'd been yanked out of 4.4-Lite by macklem long ago) It was only there because macklem couldn't originally decide if things should be locked, or not...
1.7	15-Feb-1994	pk	Update {a,m}time vnode attributes on special files a la ufs_vnode.c, but make it a non-urgent operation, to leave us some performance.
1.6	22-Dec-1993	cgd	change return type of nfs_print back to int
1.5	07-Sep-1993	ws	branches: 1.5.2; Changes to VFS readdir semantics NFS changes for better cookie support ISOFS changes for better Rockridge support and support for generation numbers
1.4	02-Aug-1993	mycroft	Make return type of nfs_print be a void, not an int.
1.3	22-May-1993	cgd	add Yuval Yarom's changes (originally for BSD/386) for advisory record locking on NFS files. Note that this DOES NOT support network locking, only local advisory locks.
1.2	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.1	20-Apr-1993	mycroft	branches: 1.1.1; Restore files lost during crash.
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.5.2.2	19-Dec-1993	pk	Undo misguided attempt to use quad_t type for n_size field.
1.5.2.1	16-Dec-1993	pk	Use u_quad for file size and dir offset a la ufs_inode.h. Must re-address this later as straight u_quad_t doesn't work somehow (make a Sparc crash with an "alignment fault").
1.11.2.1	19-Aug-1994	mycroft	update from trunk
1.20.4.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.27.18.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.27.12.3	11-Feb-2001	bouyer	Sync with HEAD.
1.27.12.2	08-Dec-2000	bouyer	Sync with HEAD.
1.27.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.27.8.2	04-Jul-1999	chs	support VOP_BALLOC().
1.27.8.1	07-Jun-1999	chs	merge everything from chs-ubc branch.
1.27.6.1	05-Jan-2000	he	Pull up revision 1.28 (requested by fvdl): Insert an extra VOP_ACCESS check in nfs_lookup, preventing cached access mishaps for lookup and getattr. Fixes PR#8884.
1.27.2.1	09-Nov-1998	chs	initial snapshot. lots left to do.
1.29.4.1	14-Dec-2000	he	Pull up revision 1.30 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.32.2.5	11-Dec-2002	thorpej	Sync with HEAD.
1.32.2.4	11-Nov-2002	nathanw	Catch up to -current
1.32.2.3	22-Oct-2002	thorpej	Sync with HEAD.
1.32.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.32.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.33.4.1	01-Oct-2001	fvdl	Catch up with -current.
1.33.2.2	30-Sep-2002	jdolecek	add support for kevents to NFS to detect file changes on server by other NFS clients, polling kernel thread is used to periodically check for attribute changes of watched files; the NFS server is only contacted when the vnode expires from local attrcache (which takes 5-60 seconds currently), to keep network&CPU overhead low the routine checking for remote changes is quite simplistic, but hopefully doing it's job well enough
1.33.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.40.2.8	04-Feb-2005	skrll	Sync with HEAD.
1.40.2.7	17-Jan-2005	skrll	Sync with HEAD.
1.40.2.6	18-Dec-2004	skrll	Sync with HEAD.
1.40.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.40.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.40.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.40.2.2	03-Sep-2004	skrll	Sync with HEAD
1.40.2.1	03-Aug-2004	skrll	Sync with HEAD
1.46.2.4	11-Jan-2005	jmc	Pullup rev 1.53 (requested by yamt in ticket #1079) Invalidate cache if filesize is changed besides our activity because it means that were out of sync with the server.
1.46.2.3	11-Jan-2005	jmc	Pullup patch (requested by yamt in ticket #1077) nfs_lookup: check n_nctime for positive entries as well to improve cache consistency.
1.46.2.2	18-Sep-2004	he	branches: 1.46.2.2.2; Pull up revision 1.49 (requested by yamt in ticket #858): Fix access-after-free bugs in dircache code by reference counting nfsdircache. Fixes PR#26864.
1.46.2.1	30-Aug-2004	tron	Pull up revision 1.48 (requested by yamt in ticket #803): nfs_request: a workaround for servers doing "maproot". for i/o requests which are expected not to fail due to permission to mimic unix file open semantics (READ, WRITE, COMMIT), try two credentials. namely, the file owner's one and open time one. remember which credential worked in per-file basis and try it first next time to minimize number of retries. ideas from Chuck Silvers. PR/23716 and PR/24987.
1.46.2.2.2.3	30-Jan-2005	he	Pull up revision 1.50 (requested by yamt in ticket #968): Since daddr_t is 64-bit these days, simply use nfs directory cookies as buffer cache indexes. This should make the regress/sys/fs/getdents test work. Fixes PR#27112.
1.46.2.2.2.2	11-Jan-2005	jmc	Pullup rev 1.53 (requested by yamt in ticket #1079) Invalidate cache if filesize is changed besides our activity because it means that were out of sync with the server.
1.46.2.2.2.1	11-Jan-2005	jmc	Pullup patch (requested by yamt in ticket #1077) nfs_lookup: check n_nctime for positive entries as well to improve cache consistency.
1.53.4.1	12-Feb-2005	yamt	sync with head.
1.53.2.1	29-Apr-2005	kent	sync with -current
1.54.6.5	04-Feb-2008	yamt	sync with head.
1.54.6.4	03-Sep-2007	yamt	sync with head.
1.54.6.3	26-Feb-2007	yamt	sync with head.
1.54.6.2	30-Dec-2006	yamt	sync with head.
1.54.6.1	21-Jun-2006	yamt	sync with head.
1.55.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.55.10.2	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.
1.55.10.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.55.8.1	24-May-2006	yamt	sync with head.
1.55.6.1	01-Jun-2006	kardel	Sync with head.
1.55.4.1	09-Sep-2006	rpaulo	sync with head
1.56.10.1	22-Oct-2006	yamt	sync with head
1.56.8.2	12-Jan-2007	ad	Sync with head.
1.56.8.1	18-Nov-2006	ad	Sync with head.
1.61.8.1	11-Jul-2007	mjf	Sync with head.
1.61.6.3	20-Aug-2007	ad	Sync with HEAD.
1.61.6.2	08-Jun-2007	ad	Sync with head.
1.61.6.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.61.2.1	07-May-2007	yamt	sync with head.
1.63.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.64.6.2	20-Jul-2007	yamt	- fix decreasing of vfs.nfs.iothreads after the recent partial merge of vmlocking. - don't make nfsiod exit with requests left. - make NFSSVC_BIOD a dummy so that nfsiod can be simplified.
1.64.6.1	20-Jul-2007	yamt	file nfsnode.h was added on branch matt-mips64 on 2007-07-20 15:36:43 +0000
1.64.4.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.64.4.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.66.8.1	18-Feb-2008	mjf	Sync with HEAD.
1.66.2.1	23-Mar-2008	matt	sync with HEAD
1.67.16.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.67.10.5	10-Oct-2010	yamt	some locking changes
1.67.10.4	09-Oct-2010	yamt	sync with head
1.67.10.3	26-Sep-2010	yamt	locking changes
1.67.10.2	04-May-2009	yamt	sync with head.
1.67.10.1	27-Apr-2008	yamt	commit some work-in-progress changes to make nfs client mp-safe to a branch, so that they won't get lost. - sprinkle some locking - mark the filesystem, nfstimer callout, and kq kthread mp-safe - add assertions and comments - disable upgrade mount for now - some unrelated cosmetic changes
1.67.6.1	17-Jan-2009	mjf	Sync with HEAD.
1.68.4.2	02-Feb-2009	snj	Pull up following revision(s) (requested by ad in ticket #344): sys/nfs/nfsnode.h: revision 1.70 protect sillyrename with _KERNEL
1.68.4.1	02-Feb-2009	snj	Pull up following revision(s) (requested by ad in ticket #344): sys/nfs/nfs_node.c: revision 1.108 sys/nfs/nfsnode.h: revision 1.69 - Don't vput() a vnode that we do not hold locked. - Eliminate one of the few remaining uses of LK_CANRECURSE.
1.68.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.68.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.70.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.71.4.1	05-Mar-2011	rmind	sync with head
1.71.2.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.72.32.1	10-Aug-2014	tls	Rebase.
1.72.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.73.46.1	31-May-2021	cjep	sync with head
1.73.44.2	01-Aug-2021	thorpej	Sync with HEAD.
1.73.44.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.76.10.1	02-Aug-2025	perseant	Sync with HEAD
1.18	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.17	27-Dec-2006	yamt	branches: 1.17.170; remove nqnfs.
1.16	02-Sep-2006	yamt	branches: 1.16.2; nfsd: deal with variable-sized filehandles.
1.15	31-Jul-2006	martin	Make filehandles opaque to userland
1.14	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.13	14-Mar-2006	yamt	branches: 1.13.6; bump NFS_MAXDGRAMDATA from 32k to 60k. (ie. near the protocol limit of udp.) - it can help performance for some environments. - administrators should be free to do silly things. :-)
1.12	11-Dec-2005	christos	branches: 1.12.4; 1.12.6; 1.12.8; 1.12.10; merge ktrace-lwp.
1.11	25-Sep-2005	christos	Add missing TIMEDOUT and IO errors.
1.10	07-Aug-2003	agc	branches: 1.10.14; 1.10.16; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.9	19-Sep-2000	fvdl	branches: 1.9.24; Adapt some defaults/max values to be more realistic.
1.8	06-Aug-1998	kleink	branches: 1.8.12; 1.8.18; 1.8.22; Like for NFSv2, add a pointer to the NFSv3 RFC, too.
1.7	19-Oct-1997	fvdl	* Implement optional 32 <-> 64 bit directory cookie translation. This uses the directory cache as translation table. See nfs_subs.c for comments. Makes the code a bit more complex to look at than I would have liked, but doesn't affect the speed of the default behavior. * Optimize caching behavior a bit when buffers are invalidated. * Save some RPCs in readdir operations by not bothering if there is a small amount left to do to fill the buffer. It'll be done in the next RPC with a larger chunk anyway. Wastes a bit of buffer space but is faster. * Make n_vattr an allocated vattr struct. This avoids nfsnode bloat, and is friendlier to the malloc routines.
1.6	17-Oct-1997	fvdl	NFS_SMALLFH should be a multiple of 4; since the nfsnode struct may grow one more member, just make it the minimum (32) now.
1.5	17-Oct-1997	christos	nfstov_mode converts 32 bits now. change NFS_SMALLFH from 44 to 38 to accommodate the mode_t and nlink_t changes.
1.4	12-Oct-1997	fvdl	Do negative lookup caching. Use a timestamp of the oldest negative cache entry, so it can be checked against directory modification time for validity.
1.3	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.2	08-May-1997	mycroft	branches: 1.2.4; Pass the vnode type to vaccess(), and use it when checking VEXEC. Make sure that the mode bits passed to vaccess() and returned by foo_getattr() contain only permission bits.
1.1	18-Feb-1996	fvdl	branches: 1.1.1; Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.2.4.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.8.22.1	14-Dec-2000	he	Pull up revision 1.9 (requested by fvdl): Improve NFS performance, possibly with as much as 100% in throughput. Please note: this implies a kernel interface change, VOP_FSYNC gains two arguments.
1.8.18.1	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".
1.8.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.9.24.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.9.24.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.9.24.2	18-Sep-2004	skrll	Sync with HEAD.
1.9.24.1	03-Aug-2004	skrll	Sync with HEAD
1.10.16.2	30-Dec-2006	yamt	sync with head.
1.10.16.1	21-Jun-2006	yamt	sync with head.
1.10.14.1	16-Dec-2005	tron	Pull up following revision(s) (requested by christos in ticket #1055): sys/nfs/nfsproto.h: revision 1.11 Add missing TIMEDOUT and IO errors.
1.12.10.1	19-Apr-2006	elad	sync with head.
1.12.8.3	03-Sep-2006	yamt	sync with head.
1.12.8.2	11-Aug-2006	yamt	sync with head
1.12.8.1	01-Apr-2006	yamt	sync with head.
1.12.6.1	22-Apr-2006	simonb	Sync with head.
1.12.4.1	09-Sep-2006	rpaulo	sync with head
1.13.6.1	13-Jul-2006	gdamore	Merge from HEAD.
1.16.2.1	12-Jan-2007	ad	Sync with head.
1.17.170.1	02-Aug-2025	perseant	Sync with HEAD
1.11	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.10	05-Sep-2014	matt	branches: 1.10.56; Don't nest structure definitions.
1.9	28-Dec-2006	yamt	branches: 1.9.90; remove several nqnfs definitions.
1.8	11-Dec-2005	christos	branches: 1.8.20; merge ktrace-lwp.
1.7	07-Aug-2003	agc	branches: 1.7.16; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.6	12-May-2002	matt	branches: 1.6.10; Eliminate commons
1.5	12-May-1997	fvdl	branches: 1.5.34; 1.5.36; Store RPC procnum consistently as an u_int32_t. This is as it should be, and avoid possible server crashes due to bogus comparisons. Partly from BSDI.
1.4	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.3	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.2	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.1	08-Jun-1994	mycroft	branches: 1.1.1; Update to 4.4-Lite fs code, with local changes.
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.5.36.1	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.5.34.1	20-Jun-2002	nathanw	Catch up to -current.
1.6.10.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.10.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.10.1	03-Aug-2004	skrll	Sync with HEAD
1.7.16.1	30-Dec-2006	yamt	sync with head.
1.8.20.1	12-Jan-2007	ad	Sync with head.
1.9.90.1	03-Dec-2017	jdolecek	update from HEAD
1.10.56.1	02-Aug-2025	perseant	Sync with HEAD
1.17	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.16	04-Dec-2007	yamt	branches: 1.16.140; merge non-intrusive nfs changes from vmlocking.
1.15	01-Jun-2007	yamt	branches: 1.15.6; 1.15.8; 1.15.14; 1.15.16; use mutex and condvar.
1.14	28-Dec-2006	yamt	branches: 1.14.6; 1.14.8; remove several nqnfs definitions.
1.13	11-Dec-2005	christos	branches: 1.13.20; merge ktrace-lwp.
1.12	07-Aug-2003	agc	branches: 1.12.16; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.11	12-May-1997	fvdl	branches: 1.11.56; Store RPC procnum consistently as an u_int32_t. This is as it should be, and avoid possible server crashes due to bogus comparisons. Partly from BSDI.
1.10	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.9	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.8	13-Dec-1994	mycroft	Sync with CSRG.
1.7	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.6	17-Aug-1994	mycroft	Use LIST and TAILQ for hash chain and LRU chain, respectively.
1.5	29-Jun-1994	cgd	branches: 1.5.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.4	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.3	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.2	20-Apr-1993	mycroft	Add consistent multiple-inclusion protection (repeat).
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.5.2.1	19-Aug-1994	mycroft	update from trunk
1.11.56.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.11.56.2	18-Sep-2004	skrll	Sync with HEAD.
1.11.56.1	03-Aug-2004	skrll	Sync with HEAD
1.12.16.3	07-Dec-2007	yamt	sync with head
1.12.16.2	03-Sep-2007	yamt	sync with head.
1.12.16.1	30-Dec-2006	yamt	sync with head.
1.13.20.1	12-Jan-2007	ad	Sync with head.
1.14.8.1	11-Jul-2007	mjf	Sync with head.
1.14.6.3	27-Aug-2007	yamt	- fix/add assertions. - fix numnfsrvcache.
1.14.6.2	26-Aug-2007	yamt	- mark nfssvc(2) MPSAFE and move the most of nfsd out of the kernel lock. - remove unused ns_solock. - remove some of KERNEL_LOCK/UNLOCK which are not necessary on this branch.
1.14.6.1	09-Jun-2007	ad	Sync with head.
1.15.16.2	08-Dec-2007	ad	Sync with head.
1.15.16.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.15.14.1	08-Dec-2007	mjf	Sync with HEAD.
1.15.8.1	09-Jan-2008	matt	sync with HEAD
1.15.6.1	09-Dec-2007	jmcneill	Sync with HEAD.
1.16.140.1	02-Aug-2025	perseant	Sync with HEAD
1.11	18-Feb-1996	fvdl	Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.10	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.9	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.8	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.7	10-Apr-1994	cgd	patchkit date deletions!
1.6	10-Mar-1994	ws	Oops. Bug fix for nfs server. Reported by Theo.
1.5	09-Mar-1994	ws	Make FFS optional
1.4	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.3	20-Apr-1993	mycroft	Add consistent multiple-inclusion protection (repeat).
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
1.22	27-Dec-2006	yamt	remove nqnfs.
1.21	02-Sep-2006	yamt	branches: 1.21.2; nfsd: deal with variable-sized filehandles.
1.20	13-Jul-2006	martin	Fix alignement problems for fhandle_t, exposed by gcc4.1. While touching all vptofh/fhtovp functions, get rid of VFS_MAXFIDSIZ, version the getfh(2) syscall and explicitly pass the size available in the filehandle from userland. Discussed on tech-kern, with lots of help from yamt (thanks!).
1.19	07-Jun-2006	kardel	branches: 1.19.2; merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.18	14-May-2006	elad	branches: 1.18.2; integrate kauth.
1.17	11-Dec-2005	christos	branches: 1.17.4; 1.17.6; 1.17.8; 1.17.10; 1.17.12; merge ktrace-lwp.
1.16	21-Apr-2004	christos	branches: 1.16.12; use VFS_MAXFIDSZ
1.15	16-Aug-2003	yamt	current trylater/jukebox retry delay is way too long and it has a bug in the backoff calculation. so, - clip it to 1-60 sec. (suggested by Rick Macklem) - use a constant multiplier instead of nfs_backoff, which is already exponential. - move some related constant definations to nfs.h from nqnfs.h and prefix with NFS_ instead of NQ_ because they are not nqnfs-specific.
1.14	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.13	29-Jun-2003	fvdl	branches: 1.13.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.12	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
1.11	05-May-2003	yamt	keep things not needed by userland in #ifdef _KERNEL. (e.g. prototypes for in-kernel functions)
1.10	24-Apr-2003	drochner	Change some subordinate functions to take a "struct nfsnode" argument instead of "struct vnode". This saves a number of pointer dereferences; it sums up to about half a kB for me. And it paves the way for future fixes. While cleaning up, eliminate a write-only member of "struct nfsreq" and a pointless assignment in the NFS_V2_ONLY case.
1.9	18-Jan-2003	thorpej	Merge the nathanw_sa branch.
1.8	12-May-2002	matt	Eliminate commons
1.7	09-Jun-2000	fvdl	branches: 1.7.4; 1.7.6; Some tweaks to enable NFS over IPv6. The special-casing of AF_INET should really be removed.
1.6	18-Feb-1996	fvdl	branches: 1.6.30; 1.6.38; Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.5	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.4	13-Dec-1994	mycroft	Sync with CSRG.
1.3	18-Aug-1994	mycroft	More LIST/CIRCLEQ migration.
1.2	29-Jun-1994	cgd	branches: 1.2.2; New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.1	08-Jun-1994	mycroft	branches: 1.1.1; Update to 4.4-Lite fs code, with local changes.
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.1	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.2.2.1	19-Aug-1994	mycroft	update from trunk
1.6.38.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.6.30.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago
1.7.6.1	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.7.4.2	20-Jun-2002	nathanw	Catch up to -current.
1.7.4.1	05-Mar-2001	nathanw	Initial commit of scheduler activations and lightweight process support.
1.13.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.13.2.2	03-Aug-2004	skrll	Sync with HEAD
1.13.2.1	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.
1.16.12.2	30-Dec-2006	yamt	sync with head.
1.16.12.1	21-Jun-2006	yamt	sync with head.
1.17.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.17.10.1	08-Mar-2006	elad	Adapt to kernel authorization KPI. This could use some testing...
1.17.8.4	03-Sep-2006	yamt	sync with head.
1.17.8.3	11-Aug-2006	yamt	sync with head
1.17.8.2	26-Jun-2006	yamt	sync with head.
1.17.8.1	24-May-2006	yamt	sync with head.
1.17.6.2	01-Jun-2006	kardel	Sync with head.
1.17.6.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.17.4.1	09-Sep-2006	rpaulo	sync with head
1.18.2.1	19-Jun-2006	chap	Sync with head.
1.19.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.21.2.1	12-Jan-2007	ad	Sync with head.
1.13	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.12	28-Dec-2006	yamt	branches: 1.12.170; remove several nqnfs definitions.
1.11	11-Dec-2005	christos	branches: 1.11.20; merge ktrace-lwp.
1.10	26-Feb-2005	perry	branches: 1.10.4; nuke trailing whitespace
1.9	07-Aug-2003	agc	branches: 1.9.8; 1.9.10; Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.8	18-Feb-1996	fvdl	branches: 1.8.64; Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.7	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.6	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.5	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.4	18-Apr-1994	glass	revised nfs diskless support. uses bootp+rpc to gather parameters
1.3	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.2	20-Apr-1993	cgd	re-merged include file changes which got eaten by crash
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.8.64.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.8.64.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.8.64.2	18-Sep-2004	skrll	Sync with HEAD.
1.8.64.1	03-Aug-2004	skrll	Sync with HEAD
1.9.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.9.8.1	29-Apr-2005	kent	sync with -current
1.10.4.1	30-Dec-2006	yamt	sync with head.
1.11.20.1	12-Jan-2007	ad	Sync with head.
1.12.170.1	02-Aug-2025	perseant	Sync with HEAD
1.10	31-Jan-1997	thorpej	This file is now obsolete.
1.9	30-Apr-1995	cgd	kill unnecessary blank line at end of file
1.8	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.7	29-Apr-1994	glass	i really wish i knew what was wrong
1.6	18-Apr-1994	glass	revised nfs diskless support. uses bootp+rpc to gather parameters
1.5	01-Mar-1994	glass	remove some warnings
1.4	18-Jan-1994	brezak	Include nfs_hack_mountroot() in NFSDISKLESS_HARDWIRE
1.3	18-Dec-1993	mycroft	Canonicalize all #includes.
1.2	14-Oct-1993	glass	this is the disgusting temporary hack to assist people booting over nfs via hacked structures until netboot works. the word "abortion" comes to mind.
1.1	07-Jul-1993	cgd	branches: 1.1.4; changes from ws to support diskless booting... these are "OK" on inspection and after testing... (actually, currently, none of the changed code is even used...)
1.1.4.1	14-Nov-1993	mycroft	Canonicalize all #includes.
1.16	07-Dec-2024	riastradh	sys/nfs/nfs: Add some missing includes and include guards. Fix up some minor KNF issues while here. No functional change intended (except to enable things to build that might not have built before because of previously required #include ordering).
1.15	11-Dec-2005	christos	branches: 1.15.200; merge ktrace-lwp.
1.14	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.13	06-Mar-1999	fair	branches: 1.13.42; Snatch a patch from OpenBSD to fix PRs 6529 and 7074. Adjust fxdr_hyper() and txdr_hyper() macros.
1.12	10-Oct-1997	fvdl	* New directory entry caching system. Provides full caching of any directory cookie that may be thrown back at us from userspace, up to a size limit. Fixes double entry problem. * Split flags for internal and external use in the NFS mount structure. * Fix some buffer structure fields that weren're being used correctly. * Fix missing directory cache inval call in nfs_open. * Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable value of 8k. * Various other things that I forget, all related to the dir caching somehow, though.
1.11	18-Feb-1996	fvdl	branches: 1.11.12; Bring in a merge of Rick Macklem's NFSv3 code from Lite2
1.10	01-Feb-1996	jtc	Rename struct timespec fields to conform to POSIX.1b
1.9	19-Dec-1995	cgd	changes to make this work on systems where pointers & longs are 64 bits. This is mostly just changes to make the stuff that goes over the wire use fixed-size types.
1.8	13-Jan-1995	mycroft	Convert unspecified usec value to 0, per discussion with Rick.
1.7	29-Jun-1994	cgd	New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1.6	08-Jun-1994	mycroft	Update to 4.4-Lite fs code, with local changes.
1.5	25-Apr-1994	cgd	some prototype cleanup, eliminate/replace bogus types (e.g. quad and u_quad) -> use better types (e.g. quad_t & u_quad_t in inodes), some cleanup.
1.4	03-Jun-1993	cgd	fix for macklem's bogus use of the va_flags field, supplied by John Woods, jfwfrom: @ksr.com. also, fixes the following problems: the va_gen field is in a similar position (Suns are going to be reporting the change-date microseconds as their "generation"), I've supplied my own set of diffs below for your inspection. Note these aren't even compiled, but they're pretty similar to what I had to do to our older version of OSF/1 here. (There's also an unrelated change supplied for xdr_subs.h; the pointer types supplied to the fxdr_time() and txdr_time() macros are not, in fact, both struct timevals. That turns out to be one of many tips-of-the-iceberg facing those porting the (old) Berkeley NFS code to 64-bit machines...)
1.3	20-May-1993	cgd	more rcs id adding and header cleanup. i like vi macros!
1.2	20-Apr-1993	mycroft	Add consistent multiple-inclusion protection (repeat).
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.3	01-Mar-1998	fvdl	Import 4.4BSD-Lite2
1.1.1.2	01-Mar-1998	fvdl	Import 4.4BSD-Lite for reference
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.11.12.1	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.13.42.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.42.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.42.1	03-Aug-2004	skrll	Sync with HEAD
1.15.200.1	02-Aug-2025	perseant	Sync with HEAD

OpenGrok