remove nqnfs.

branches: 1.3.18;
don't install nfs_var.h.

branches: 1.2.22; 1.2.34;
Remove KDIR=, since SYS_INCLUDE=symlinks and KDIR are not supported any more.

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.

Default NFS mounts to using TCP transport instead of UDP.
PR kern/53166

branches: 1.14.18;
Define filesystem attributes with vfs dependency.

branches: 1.13.20;
don't create unused fs_nfs.h

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.

branches: 1.11.2;
handle the nuidhash_max lossage differently

nuidhash_max is needed by sys_nfssvc

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

branches: 1.8.40; 1.8.44; 1.8.50; 1.8.54;
remove nqnfs.

branches: 1.7.20;
merge ktrace-lwp.

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.

branches: 1.5.4;
nuke trailing whitespace

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 ...

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.

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

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.

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).

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.

branches: 1.8.74; 1.8.84; 1.8.90;
merge ktrace-lwp.

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.

branches: 1.6.2;
keep things not needed by userland in #ifdef _KERNEL.
(e.g. prototypes for in-kernel functions)

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.

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.

Fixed RPC code to deal with RPC messages larger than one mbuf.

new RCS ID format.

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.

MBUFTRACE

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@

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.

change nfs_boot_sendrecv to take sockaddr_in * instead of mbuf *

fixes m_serv (single mbuf leak) leak in kern/subr_tftproot.c

when calling nfs_boot_sendrecv pass NULL for pointers instead of 0

m_pullup() is called in rcvproc callback functions,
so nfs_boot_sendrecv() should keep track of the head of mbuf chain.
fixes kern/48746

Fix uninitialized variable.

Found by The Brainy Code Scanner in FreeBSD.

branches: 1.37.18; 1.37.22; 1.37.38; 1.37.40;
ansify function definitions

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.

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.)

Remove all the __P() from sys (excluding sys/dist)
Diff checked with grep and MK1 eyeball.
i386 and amd64 GENERIC and sys still build.

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@.

branches: 1.32.36; 1.32.38;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.31.14;
Coverity CID 2445: Only set from_p if we succeed so that we free it on error.

branches: 1.30.4; 1.30.6; 1.30.8; 1.30.10; 1.30.12;
merge ktrace-lwp.

branches: 1.29.4;
nuke trailing whitespace

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.

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.

don't seem to need <sys/conf.h> or <net/if.h> here

add RCSIDs

branches: 1.24.2; 1.24.6;
receive, not recieve

branches: 1.23.24;
bzero->memset, bcopy->memcpy, bcmp->memcmp

Merge with Lite2 + local changes

Use functions (shared with DHCP boot) in nfs_boot.c.

Supporting changes for the new BOOTP support in nfs_mountroot.

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.

revert kprintf changes

printf -> kprintf, sprintf -> ksprintf

Another %lx -> %x for ntohl()

print result of ntohl/htonl as a long.

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.

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.

branches: 1.12.4;
Bring in a merge of Rick Macklem's NFSv3 code from Lite2

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.

Use RPCAUTH_UNIX in requests to please picky NFS servers.

Use fxdr_*() and txdr_*() macros to do byte order conversions.

Fixed RPC code to deal with RPC messages larger than one mbuf.

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.

fix typo

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.)

branches: 1.4.2;
error codes are in network order too.

New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'

New diskless boot code (uses RARP, bootparamd).

revised nfs diskless support. uses bootp+rpc to gather parameters

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).

branches: 1.80.10;
s/runable/runnable/

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.

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.

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.

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

branches: 1.75.10;
Nix LEASE_READ/LEASE_WRITE from <sys/vnode.h>.

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)

branches: 1.73.6; 1.73.10;
Retire OSI network stack. OK core@

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.

branches: 1.71.2;
remove unused r_timer member.

Change NFS to use a RB-tree for its FH->nfsnode lookups.

branches: 1.69.12; 1.69.16; 1.69.22;
merge non-intrusive nfs changes from vmlocking.

branches: 1.68.2; 1.68.4;
make NFS_ATTRTIMEO a function.

branches: 1.67.6; 1.67.8; 1.67.12;
add some #include.

it seems like a good idea to include <sys/condvar.h>, as we're using them...

use mutex and condvar.

remove R_GETONEREP.

branches: 1.63.2; 1.63.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.62.2;
remove several nqnfs definitions.

remove nqnfs.

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.)

nfsd: deal with variable-sized filehandles.

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!).

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

branches: 1.56.2;
integrate kauth.

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.

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.

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.

branches: 1.52.2;
merge ktrace-lwp.

Add some COMPAT_30 code to let old mountd binaries work after the NFS
exports rototill.

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.

Allow turning off the attribute cache.

branches: 1.48.10; 1.48.12;
remove #if 0'ed out definition of VA_EXCLUSIVE.

remove an unused macro, NMOD.

g/c unused NFS_*ALLOC defines.

don't do kludge for a reply to a retransmitted request
unless we actually retransmitted the request.

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.

*** empty log message ***

change n_mtime from time_t to timespec in order to improve
cache consistency.
(1 second granularity is too loose these days.)

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.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22364, verified by myself.

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.

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

- instead of scaning a list when looking up
{a idle thread, a socket with pending requests},
maintain dedicated list of them.
- add spin locks.

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.

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.

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>

Eliminate commons

use struct uucred in nfsd_svcargs so that we don't break the sys_nfssvc() ABI.

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.

branches: 1.30.2;
bound check mount args more thoroughly

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.

remove a temporary hack now that it's fixed for real. fixes PR 11731.

Set default NFS read and write sizes back to 8k, because a lot of
(old) hardware can't handle more.

Allow the default NFS_RSIZE and NFS_WSIZE to be overriden.

branches: 1.25.2;
Initial integration of the Unified Buffer Cache project.

Bump some defaults and maximums to better values.

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.

branches: 1.22.2;
Some tweaks to enable NFS over IPv6. The special-casing of AF_INET
should really be removed.

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.

branches: 1.20.10;
Clean up the NFS sysctl variables.

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.

bzero->memset, bcopy->memcpy, bcmp->memcmp

* 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.

* 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.

branches: 1.15.4; 1.15.6;
Add a few defines for WebNFS support.

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.

Allocate real malloc types for NFS, rather than using M_TEMP.

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>

Align things right in NWDELAYHASH (for the Alpha). This fixes crashes in
the server code. From John Birell.

branches: 1.10.4;
Bring in a merge of Rick Macklem's NFSv3 code from Lite2

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.

KERNEL -> _KERNEL

More LIST/CIRCLEQ migration.

Convert some more lists and queues.

Change the reply list to a TAILQ.

branches: 1.4.2;
New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'

Update to 4.4-Lite fs code, with local changes.

more rcs id adding and header cleanup. i like vi macros!

branches: 1.1.1;
Restore files lost during crash.

s/Enque/Enqueue/ in comment.

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.

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.

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@

Move proc_lock into the data segment. It was dynamically allocated because
at the time we had mutex_obj_alloc() but not __cacheline_aligned.

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.

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.

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.

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.

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.

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

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.

branches: 1.190.2;
Don't use catch as a variable name.

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())

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.

- 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.

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.

branches: 1.185.6;
Fix memory leak during some NFS writes.

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).

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.

nfs_bioread: don't truncate values in a debug printf.

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

- allocate 8 pointers on the stack to avoid stack overflow in nfs.
- make that 8 a constant
- remove bogus panic

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.

it appears the previous commit's sacrifice was "successful compilation with
NFS_V2_ONLY defined".

Another sacrifice to the stack protector gods.

don't use variable allocation on the stack.

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.

branches: 1.174.2;
ansify. from Christoph Egger.

branches: 1.173.6;
use kmem_alloc instead of malloc.

Merge vmlocking2 to head.

branches: 1.171.4;
merge non-intrusive nfs changes from vmlocking.

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

branches: 1.169.2;
make NFS_ATTRTIMEO a function.

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.

Merge brelse() changes from the vmlocking branch.

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.

push kernel_lock a little.

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.

use ubc_uiomove for read as well.

ubc_uiomove: add an "advice" argument rather than using UVM_ADV_RANDOM blindly.

- 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.

branches: 1.160.2;
remove (void)0; nonsense.

fix a typo in a comment.

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.

Merge some of the less invasive changes from the vmlocking branch:

- kthread, callout, devsw API changes
- select()/poll() improvements
- miscellaneous MT safety improvements

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.

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.

nfs_write: report an error correctly in the case of IO_SYNC.

use mutex and condver.

hold proclist_mutex when calling psignal().

branches: 1.151.2; 1.151.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

nfs_getpages: fix an inverted condition in rev.1.147.

TRUE -> true, FALSE -> false

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.

branches: 1.147.2;
use mutex and rwlock rather than lockmgr.

remove nqnfs.

branches: 1.145.4;
Use the LWP cached credentials where sane.

fix handling of NFSERR_NOTSUPP and NFSERR_BAD_COOKIE,
which have been broken since nfs_socket.c rev.1.115.

branches: 1.143.4;
integrate kauth.

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.

branches: 1.141.2; 1.141.4;
nfs_doio_read: clear uio_resid when filling a hole.

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.

merge ktrace-lwp.

merge yamt-readahead branch.

branches: 1.137.2;
nfs_bioread: push delayed truncation and tweak loop accordingly.
PR/31926 from Jed Davis.

nfs_bioread: handle file truncation on the server a little more gracefully.

use killproc() for killing the process due to text file modification, so
that it's logged too

PR: 17392 by Greg A. Woods

fix some simple bugs in the 64bit ino_t changes.
- edp -> dp
- * -> +

64 bit inode changes.

use a correct credential for readlink. discussed on source-changes@.

nfs_doio_read: revert readlink part of 1.129 and 1.130 because they were wrong.

Back to using curproc in the VLNK case when uiop->uio_procp == NULL,
and explain why we need to.

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).

branches: 1.128.2; 1.128.4;
nuke trailing whitespace

- simplify nfs_bio.c rev.1.126
- add an assertion.

no functional changes.

nfs_bioread:
- if a buffer is still empty after successful nfs_doio, it implies EOF.
- don't cache blocks beyond EOF.

handle a really empty directory, which doesn't have even the dot entry.

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.

- centerize code to invalidate stale cache.
- don't ignore errors when invalidating buffers in nfs_open.

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.

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

fix access-after-free bugs in dircache code by refcounting nfsdircache.
PR/26864.

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.

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.

cut down another 7K by more NFS_V2_ONLY ifdefs.

branches: 1.116.2;
introduce a macro NFS_INVALIDATE_ATTRCACHE and use it
instead of "n_attrstamp = 0".

comments in nfs_doio_write.

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.

Fix hanging-paren typo.

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>.

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.

change n_mtime from time_t to timespec in order to improve
cache consistency.
(1 second granularity is too loose these days.)

don't call nfs_delayedtruncate() from nfs_getpages().
it causes simplelock deadlock.

VOP_PUTPAGES() must be called with the vnode's interlock held.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22364, verified by myself.

Make life slightly easier for the compiler's optimisation routines.

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.

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

interlock for NFLUSHINPROG/NFLUSHWANT.

eliminate memcpy in the common and easy case of write.

correct a KASSERT.

acquire vmobjlock when touch pg->flags.

simple lock for nfs iod.

- check page's offset in the object as well. (pointed by Chuck Silvers.)
- remove false assertion.

- if writerpc ends with a stable result, no need to commit them anymore.
- add comments.

better handling of write verifier change.

fix a use of an uninitialized variable.

remove line-wrapping that is no longer needed.

fix a typo in the previous.

set b_resid correctly.

split nfs_doio to nfs_doio_{phys,read,write} to avoid too deep indents.

- 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.

rename a very confusing variable name.
(must_commit -> stalewriteverf)

when commit failed and fall to write, re-set 'off' and 'cnt'
because it can be changed in 'needcommit' path.

make per-iod datas together.

Merge the nathanw_sa branch.

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.

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

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.

nfs_doio(): handle the case where nfs_writerpc() returned error != 0.
Fix kern/18125. OK'd by thorpej and chs.

branches: 1.81.4;
Remove wrong assertion in previous commit.

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.

only use UBC_FAULTBUSY to access offsets past EOF,
otherwise we can deadlock trying to busy the same page in uiomove().

remove PGO_WEAK, it isn't needed anymore.

only do v3 stuff for v3 filesystems.

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.

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.

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.

fix locking in nfs_getpages().

call VOP_PUTPAGES() directly instead of indirecting through
the UVM pager op vector.

add RCSIDs

branches: 1.70.2;
Remove so variables that are only ever set and never referenced.

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.

branches: 1.68.2; 1.68.4;
Make sure to add NFS vnodes to the syncerd worklist.

replace vm_page_t with struct vm_page *.

reads at or after EOF should "succeed".

handle partially full directory buffers by only using (b_bcount - b_resid)
bytes of data from the buffer.

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.

branches: 1.63.2;
min() -> MIN(), max() -> MAX().
fixes more problems with file offsets > 4GB.

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.

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.

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()).

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.

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.

<sys/trace.h> is not needed here

only zero the part of the page after EOF if we're actually
initializing the page.

Initialize 'error' to 0, so that nfs_putpages doesn't return garbage
when pages already have been committed and nothing needs to be done.

Initial integration of the Unified Buffer Cache project.

Extend NFS_V2_ONLY to remove NQNFS lease support as well. Saves another 10k.

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.

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.

remove include of <vm/vm.h>

branches: 1.49.4;
Fix printf() format.

Remove register declarations.

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.

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

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.

branches: 1.44.2;
bzero->memset, bcopy->memcpy, bcmp->memcmp

Fix possible overflow problem in read size computation.

- add defopt's for UVM, UVMHIST and PMAP_NEW.
- remove unnecessary UVMHIST_DECL's.

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. :-)

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.

Oops. Fix goof in previous change.

Just return immediately in nfs_bioread if we got an empty buffer because
of EOF on a directory.

branches: 1.37.2;
Fix alignment problems. From Frank van der Linden <fvdl@NetBSD.ORG>.

Only do readaheads when reading sequential blocks; check v_lastr to
achieve this. Improves performance for demand paging. From Chris Demetriou.

* 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.

* 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.

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

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.

Only wake up one nfsiod when there is an async write to do. (from FreeBSD).

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>

revert kprintf changes

printf -> kprintf, sprintf -> ksprintf

Don't mistake a non-async block that needs to be commited for an
interrupted write.

* 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.

branches: 1.25.4;
Make sure to clear B_NEEDCOMMIT in the right spot. Fix 'officially blessed'
by Rick Macklem. Fixes PR kern/2128.

Bring in a merge of Rick Macklem's NFSv3 code from Lite2

nfs prototype changes

Rename struct timespec fields to conform to POSIX.1b

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.

Make call to nfs_writerpc() consistent with others.

Add two missing brelse() calls. From Rick Macklem.

Make sure readdir requests are only truncated on block boundaries.

Fix a problem with write-behind causing processes to be killed occasionally.
From Rick Macklem.

minor cache consistency fix

branches: 1.15.2;
New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'

straighten out diskless swap code somewhat.

Turn P_NOSWAP and P_PHYSIO into a hold count, as suggested by a comment.

Update to 4.4-Lite fs code, with local changes.

MIN -> min, MAX -> max

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.

Convert mount, vnode, and buf structs to use <sys/queue.h>. Also,
some knf and structure frobbing to do along with it.

Canonicalize all #includes.

branches: 1.7.2;
Include systm.h to get prototypes (and possibly inlines) of *max functions.

get rid of some more bogus changes from a week ago

diskless changes made last time were hosed; were using NULL for
"no credentials" rather than NOCRED.

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...)

Paul Kranenburg's VM deadlock fixes. (patchkit patch 00147, part 2)

more rcs id adding and header cleanup. i like vi macros!

branches: 1.1.1;
Initial revision

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@

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.

Default NFS mounts to using TCP transport instead of UDP.
PR kern/53166

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.

branches: 1.86.2;
KNF. Remove extra spaces. No functional change.

change nfs_boot_sendrecv to take sockaddr_in * instead of mbuf *

fixes m_serv (single mbuf leak) leak in kern/subr_tftproot.c

change sosend() to accept sockaddr * instead of mbuf * for nam.

bump to 7.99.16

* 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@

m_pullup() is called in rcvproc callback functions,
so nfs_boot_sendrecv() should keep track of the head of mbuf chain.
fixes kern/48746

branches: 1.81.4; 1.81.6;
Mark a potentially unused variable

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@.

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.

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

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.

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.

branches: 1.75.2;
- ansify function definition
- de- __P
- u_int32_t -> uint32_t

No functional changes.

Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.

approved by core

branches: 1.73.4;
Delete unnecessary cast to void *.

branches: 1.72.2;
Remove clause 3 and 4 from TNF licenses

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@.

branches: 1.70.2;
use aprint_*_dev and device_xname

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.

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.

branches: 1.67.2;
Cosmetic: de-__P() et cetera.

Fix buid (broken by a fix introduced in the wrong file...)

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.

branches: 1.64.2; 1.64.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

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.

branches: 1.62.2; 1.62.4; 1.62.6;
merge ktrace-lwp.

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.

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.

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.

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

add RCSIDs

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.

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).

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.

branches: 1.53.4;
Don't need to include <sys/conf.h> here.

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.

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

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

Fix copyright spacing.

bzero->memset, bcopy->memcpy, bcmp->memcmp

Fix last change: If BOOTP/DHCP was successful, don't try RARP/BOOTPARAM.

Clean up boogered gcc warning workaround (remove goto completely) and remove
a redundant `if'.

Adapt to new sosend/soreceive and upcall (now down in sowakeup)

Merge with Lite2 + local changes

be a bit more clear about what protocols will be tried in the
BOOTP/DHCP case.

Consolidate NFS_BOOT_* options into opt_nfs_boot.h

Make NFS_BOOT_DHCP work as expected.

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".)

Factor out some functions used by bootparam and DHCP boot.

Correct a comment regarding the sense of the nfs_boot_rfc951 global.

Move the call to nfs_boot_getfh() from nfs_vfsops.c to nfs_boot.c
(just for better isolation - it can now be static)

Change test from NETHER to NARP (revarpwhoami is in if_arp.c)

Supporting changes for the new BOOTP support in nfs_mountroot.

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.

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.

Add some missing "\n"'s.

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.

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.

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.

revert kprintf changes

printf -> kprintf, sprintf -> ksprintf

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.

Bring in a merge of Rick Macklem's NFSv3 code from Lite2

Add stub for nfs_boot_getfh if NETHER==0

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.

Don't return EBADRPC if we have something else.

nfs prototype changes

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.

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.

don't blindly set IFF_UP; or it with old flags

Use fxdr_*() and txdr_*() macros to do byte order conversions.

Fixed RPC code to deal with RPC messages larger than one mbuf.

Cosmetic changes suggested by Adam.

Do the printf "root/swap on" elsewhere to avoid confusion.

Working "config generic" support; from Theo.

fix a couple of obvious, painful endianness bugs introduced in last commit.

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.

char * --> caddr_t, where appropriate.

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.)

Fix the conditionally compiled code inside #ifdef NFS_BOOT_GATEWAY
and make some printf args use host byteorder.

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.

branches: 1.6.2;
knf

New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'

Construct mountpath for remote root.

New diskless boot code (uses RARP, bootparamd).

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.

revised nfs diskless support. uses bootp+rpc to gather parameters

MBUFTRACE

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@

branches: 1.58.2;
s/contigous/contiguous/ in comment.

s/reqest/request/ in comment.

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.

change nfs_boot_sendrecv to take sockaddr_in * instead of mbuf *

fixes m_serv (single mbuf leak) leak in kern/subr_tftproot.c

when calling nfs_boot_sendrecv pass NULL for pointers instead of 0

m_pullup() is called in rcvproc callback functions,
so nfs_boot_sendrecv() should keep track of the head of mbuf chain.
fixes kern/48746

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@.

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.

Protect against short IP addresses in the DHCP message.

When using DHCP, request the parameters that we need. Fixes PR kern/38830.
Thanks to Tim McIntosh.

correct previous: use %zu for BOOTP_SIZE_(MIN,MAX).
Pointed out by David Holland

buildfix: use %d for BOOTP_SIZE_(MIN,MAX).
Makes i386 ALL kernel build again.

- 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!

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

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.

branches: 1.43.2;
- ansify function definition
- de- __P
- u_int32_t -> uint32_t

No functional changes.

Convert socket options code to use a sockopt structure
instead of laying everything into an mbuf.

approved by core

When doing pointer arithmetic to compute limit cast bootp to pointer
type of correct signedness. Caught by lint.

branches: 1.40.2; 1.40.4;
Fix compilation with DEBUG_NFS_BOOT_DHCP and ssp.

branches: 1.39.2;
Remove clause 3 and 4 from TNF licenses

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@.

branches: 1.37.6; 1.37.8;
Constify.

branches: 1.36.8; 1.36.12;
Constify: LLADDR() -> CLLADDR().

branches: 1.35.2; 1.35.6; 1.35.8;
Fix buid (broken by a fix introduced in the wrong file...)

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.

branches: 1.33.2; 1.33.4;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.32.2; 1.32.4;
remove some __unused in function parameters.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

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.

branches: 1.29.4; 1.29.6; 1.29.8; 1.29.10;
merge ktrace-lwp.

branches: 1.28.4;
nuke trailing whitespace

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.

remove duplicated snprintf(vci, ...)

kill sprintf, use snprintf

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.

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

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.

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)

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().

Add a NetBSD Vendor Class Identifier option as proposed on tech-net in
message <20020216172527.C23901@dr-evil.shagadelic.org>.

add RCSIDs

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".

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

Allow nfs root over token ring. Closes PR6629.

branches: 1.14.2;
Don't need to include <sys/conf.h> here.

If server name field is overloaded for other purpose, or it just contains
NULL string, don't use it as server name.

branches: 1.12.2;
print diskless boot related IP addresses in dot notation

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

Fix printf format problems on Alpha.

Fix copyright spacing.

bzero->memset, bcopy->memcpy, bcmp->memcmp

-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

Sweep up some miscellaneous leftover lite2 integration shrapnel.

Consolidate NFS_BOOT_* options into opt_nfs_boot.h

Make NFS_BOOT_DHCP work as expected.

Use interface type to select "hardware type" in bootp header.

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.

branches: 1.1.2;
Add support for nfs_mountroot using BOOTP based on the contributions
of Tor Egge (closes PR kern/2351).

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@

branches: 1.39.34;
Appease GCC and initialize arps_ip

Fixes build as GCC errors with maybe-uninitialized that is a false
positive.

branches: 1.38.30;
tyop in comment, from Eivind Evensen via OpenBSD

branches: 1.37.8; 1.37.18; 1.37.22;
in nfs_bootparam(), set the corresponding flag for each field that we fill in.

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.

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

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.

branches: 1.33.2;
- ansify function definition
- de- __P
- u_int32_t -> uint32_t

No functional changes.

branches: 1.32.6;
Remove clause 3 and 4 from TNF licenses

branches: 1.31.6; 1.31.8; 1.31.10;
use kmem_alloc instead of malloc.

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.

branches: 1.29.14;
m_freem takes one arg.

s/mfree/m_freem/

Don't leak mbufs on error.

branches: 1.26.4; 1.26.6; 1.26.8; 1.26.10; 1.26.12;
merge ktrace-lwp.

branches: 1.25.2;
- sprinkle const
- avoid shadowed variables
- mark bad const use with XXXUNCONST

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.

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.

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

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.

<sys/conf.h> and <sys/device.h> are not needed here

add RCSIDs

branches: 1.18.2;
Remove so variables that are only ever set and never referenced.

branches: 1.17.2; 1.17.4;
include opt_inet.h, needed by previous check-in.

perform reverse ARP only if INET is compiled into the kernel

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.

-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