Commit Graph

149755 Commits

Author SHA1 Message Date
Dag-Erling Smørgrav
80c03b8eee As jhb@ pointed out to me, r197057 was incorrect, not least because these
are generated files.
2009-09-10 13:20:27 +00:00
John Baldwin
9425bee7a0 Don't malloc a buffer while holding the prison0 mutex. Instead, use a loop
where we figure out the hostname length under the lock, malloc the buffer
with the lock dropped, then recheck the length under the lock and loop again
if the buffer is now too small.

Tested by:	Norbert Koch  nkoch  demig de
MFC after:	3 days
2009-09-10 12:58:37 +00:00
Dag-Erling Smørgrav
edefbdd7cb If a certain feature that was present in FreeBSD 7 was removed or changed in
FreeBSD 8, the compatibility shims should be built not just when FreeBSD 7
compatibility is requested, but also when compatibility with any older
FreeBSD version where that feature was present is requested.o

Without this patch, a kernel config that sets COMPAT_FREEBSD6 but not *7
would fail to build due to inconsistencies between the declaration of the
compatibility shims and their use in the SysV code.

There are similar errors in other *proto.h headers in the tree.

MFC after:	3 weeks
2009-09-10 08:33:28 +00:00
Dag-Erling Smørgrav
767a0b2dc9 Fix comment about KERNFAST. 2009-09-10 07:37:36 +00:00
Edwin Groothuis
8ab7b839bf Mention update to tzdata2009m. 2009-09-10 01:26:45 +00:00
Warner Losh
9f85041125 The boot loader is a FreeBSD a.out binary for x86, not a VAX binary.
Rather than writing out a MID of '0', write a MID of 0x86 (aka
MID_I386) so that file gets it right.

This is a nop for boot2.  It just checks the MAGIC part of the field,
ignoring the MID.  boot2 is the only thing that loads this file, and
only on x86 so the MID_i386 is always the right value (the rest of the
code is already x86 specific).

Reviewed by:	bde@, jhb@
MFC after:	8.0 is out the door :)
2009-09-10 00:47:32 +00:00
Brooks Davis
b3bfcfd0d9 Change one no to not in r197047
Submitted by:	John Nielsen <john at jnielsen net>
2009-09-09 21:51:54 +00:00
Konstantin Belousov
b55ef216fe kern_select(9) copies fd_set in and out of userspace in quantities of
longs. Since 32bit processes longs are 4 bytes, 64bit kernel may copy in
or out 4 bytes more then the process expected.

Calculate the amount of bytes to copy taking into account size of fd_set
for the current process ABI.

Diagnosed and tested by:	Peter Jeremy <peterjeremy acm org>
Reviewed by:	jhb
MFC after:	1 week
2009-09-09 20:59:01 +00:00
Rick Macklem
8f63187ec1 Add LK_NOWITNESS to the vn_lock() calls done on newly created nfs
vnodes, since these nodes are not linked into the mount queue and,
as such, the vn_lock() cannot cause a deadlock so LORs are harmless.

Suggested by:	kib
Approved by:	kib (mentor)
MFC after:	3 days
2009-09-09 20:37:49 +00:00
Brooks Davis
469c3a5ff4 Mention that a few ed(4) devices don't emit link state change notices
and a workaround for dhclient.

Reviewed by:	 thierry
MFC after:	3 days
2009-09-09 20:28:58 +00:00
Konstantin Belousov
9411b7675e As was done in r196643 for i386 and amd64, swap the start/end virtual
addresses in pmap_invalidate_cache_range().

Reported by:	Vincent Hoffman <vince unsane co uk>
Reviewed by:	jhb
MFC after:	3 days
2009-09-09 19:40:54 +00:00
Ed Schouten
4522791bb4 Make the description of `b' a little better.
If you have a one-byte sequence, `w', `b' is the second character. Not
the third.

Submitted by:	Christoph Mallon
2009-09-09 19:38:19 +00:00
Pawel Jakub Dawidek
21038751aa Actually component with the greatest priority is used by the prefer balance
algorithm.
2009-09-09 19:06:01 +00:00
Navdeep Parhar
81af4f18d6 There is no need to log anything for a ctrlq stall or restart. These are
normal events.

Approved by:	gnn (mentor)
MFC after:	1 month
2009-09-09 18:55:18 +00:00
Bjoern A. Zeeb
8a2044b6e0 Remove dpms.4 missed in r197025. 2009-09-09 14:17:07 +00:00
Pawel Jakub Dawidek
fdc63f4908 Correct typo after manual patching.
Noticed by:	b. f.
2009-09-09 13:23:26 +00:00
Pawel Jakub Dawidek
1e8a101ef2 Fix usecount leak in mknod(2) on file system exported over NFS.
While I'm here, correct typo in comment.

Reviewed by:	kan, kib
MFC after:	3 days
2009-09-09 12:56:05 +00:00
Konstantin Belousov
1ef6ea9b60 Unlock the image vnode around the call of pmc PMC_FN_PROCESS_EXEC hook.
The hook calls vn_fullpath(9), that should not be executed with a vnode
lock held.

Reported by:	Bruce Cran <bruce cran org uk>
Tested by:	pho
MFC after:	3 days
2009-09-09 10:52:36 +00:00
Konstantin Belousov
427992ecdb In vfs_mark_atime(9), be resistent against reclaimed vnodes.
Assert that neccessary locks are taken, since vop might not be called.

Tested by:	pho
MFC after:	3 days
2009-09-09 10:51:50 +00:00
Poul-Henning Kamp
aa3e6183ce Revert r196976, now that <machine/param.h> behaves predictably. 2009-09-09 10:31:20 +00:00
Yoshihiro Takahashi
e89ed83af9 MFi386: Remove old file. 2009-09-09 10:05:50 +00:00
Xin LI
ee5e90dab2 - Teach vesa(4) and dpms(4) about x86emu. [1]
- Add vesa kernel options for amd64.
 - Connect libvgl library and splash kernel modules to amd64 build.
 - Connect manual page dpms(4) to amd64 build.
 - Remove old vesa/dpms files.

Submitted by:	paradox <ddkprog yahoo com> [1], swell k at gmail.com
		(with some minor tweaks)
2009-09-09 09:50:31 +00:00
Xin LI
f45c063aa7 Copy the following files to new places, a subsequent commit would remove
them from the old place.  This commit necessary so that the tree would not
enter a broken state.

	sys/i386/isa/vesa.c -> dev/fb/vesa.c
	sys/i386/include/pc/vesa.h -> dev/fb/vesa.h
	sys/i386/isa/dpms.c -> dev/dpms/dpms.c
2009-09-09 07:45:08 +00:00
Warner Losh
fedf3e9ed3 kern_execve.c hasn't been around in ages, so update the file(s) where
a_magic is used instead of the a_midmag....

# maybe we can retire this hack soon...
2009-09-09 06:49:49 +00:00
Xin LI
efba048eb5 - Port x86emu to FreeBSD.
- Connect x86emu to build.

Tested with:	make universe
Submitted by:	swell.k at gmail com
2009-09-09 05:53:26 +00:00
Alexander Motin
f66de33a6a Add NVidia MCP89 HDA controller IDs. 2009-09-09 04:48:41 +00:00
Alexander Motin
f147bd6d3d Add Intel 82801JD (one more ICH10) HDA controller ID.
Submitted by:	yongari
2009-09-09 04:36:56 +00:00
Ed Maste
7771d6bf70 Whitespace tweak.
Submitted by:	phk, via accidental commit in r196969.
2009-09-09 04:04:09 +00:00
Ed Maste
4a42b06106 Increase AAC_CMD_TIMEOUT from 30s to 120s to help avoid spurious
"COMMAND 0x........ TIMEOUT AFTER .. SECONDS" messages.  Any commands
that get truly stuck will still trigger the warning and the hardware
health check, just a little bit later.
2009-09-09 03:41:40 +00:00
Ed Maste
1bdc73d337 Compare pointer with NULL, not 0. 2009-09-09 03:36:43 +00:00
Xin LI
884e6d6087 Import x86emu from OpenBSD as of OPENBSD_4_6. 2009-09-09 01:19:26 +00:00
Xin LI
1d1f4d6590 Import x86emu from OpenBSD (src/sys/dev/x86emu).
x86emu is a software emulator for the real mode x86 processor, originally
written by SciTech Software, Inc. for XFree86 and has been later ported
to NetBSD then OpenBSD.
2009-09-09 01:14:25 +00:00
Edwin Groothuis
a97a1c9e47 MFV of r196998.
Import of tzdata2008m

Samoa will go in DST on 4 October 2009 till 28 March 2010
Palestine will go back from DST on 4 September 2009
2009-09-09 00:07:05 +00:00
Edwin Groothuis
190e419c74 Vendor import of tzdata2008m:
Samoa will go in DST on 4 October 2009 till 28 March 2010
Palestine will go back from DST on 4 September 2009

Obtained from:	ftp://elsie.nci.nih.gov/pub/
2009-09-09 00:01:09 +00:00
Navdeep Parhar
9a31144537 Add arp_update_event. This replaces route_arp_update_event, which
has not worked since the arp-v2 rewrite.

The event handler will be called with the llentry write-locked and
can examine la_flags to determine whether the entry is being added
or removed.

Reviewed by:	gnn, kmacy
Approved by:	gnn (mentor)
MFC after:	1 month
2009-09-08 21:17:17 +00:00
Poul-Henning Kamp
a254d1f16d Get rid of the _NO_NAMESPACE_POLLUTION kludge by creating an
architecture specific include file containing the _ALIGN*
stuff which <sys/socket.h> needs.
2009-09-08 20:45:40 +00:00
Nathan Whitehorn
57cf46fea3 Remove some debugging (KTR_VERBOSE) that crept into ppc GENERIC long ago
and is present on no other architectures by default.

MFC after:	4 days
2009-09-08 20:12:26 +00:00
Pawel Jakub Dawidek
bf55edc7c8 Implement __assert() for Solaris-specific code. Until now Solaris code was
using Solaris prototype for __assert(), but FreeBSD's implementation.
Both take different arguments, so we were either core-dumping in assert()
or printing garbage.

Reported by:	avg
MFC after:	1 week
2009-09-08 20:06:35 +00:00
Brooks Davis
d534b0c238 cr_groups is no longer embedded in struct ucred and is instead stored
in a seperate array.  As such we need to use kvm_read rather than bcopy
to populate the ki_groups field.

This fixes a crash when running ps -ax on a coredump.

Reported by:	brucec
Tested by:	brucec
MFC after:	3 days
2009-09-08 19:37:59 +00:00
Edward Tomasz Napierala
2ff8f63aa6 Remove useless variable assignment. 2009-09-08 17:23:32 +00:00
Edward Tomasz Napierala
bb3fd7ff4f Remove unused variable. 2009-09-08 17:20:17 +00:00
Pawel Jakub Dawidek
3770996142 Only log successful commands! Without this fix we log even unsuccessful
commands executed by unprivileged users. Action is not really taken, but it is
logged to pool history, which might be confusing.

Reported by:	Denis Ahrens <denis@h3q.com>
MFC after:	3 days
2009-09-08 16:40:08 +00:00
Warner Losh
d2111f569c Cleanup tortured logic a big. Noticed by FlexLint, per phk.
# My reading of the docs suggests this can only happen on 10Base5 and
# 10Base2 setups (and maybe only the former), which I can't test
# adequately since I have nothing but 10BaseT, etc here.
2009-09-08 16:37:18 +00:00
Scott Long
216e03861c Free the correct buffer in an error case.
Submitted by:	phk
2009-09-08 16:09:28 +00:00
Pawel Jakub Dawidek
d6b8039292 We don't export individual snapshots, so mnt_export field in snapshot's
mount point is NULL. That's why when we try to access snapshots over NFS
use mnt_export field from the parent file system.

MFC after:	1 week
2009-09-08 15:57:03 +00:00
Roman Divacky
d77fa17750 Add C/c/f/p/v switches plus a bunch of minor fixes and cleanups.
Obtained from:	NetBSD
Approved by:	des (maintainer)
Approved by:	ed (mentor, implicit)
2009-09-08 15:55:13 +00:00
Pawel Jakub Dawidek
f148fd9a4a When we automatically mount snapshot we want to return vnode of the mount point
from the lookup and not covered vnode. This is one of the fixes for using .zfs/
over NFS.

MFC after:	1 week
2009-09-08 15:51:40 +00:00
Pawel Jakub Dawidek
2391003912 On FreeBSD we don't have to look for snapshot's mount point,
because fhtovp method is already called with proper mount point.

MFC after:	1 week
2009-09-08 15:42:55 +00:00
Pawel Jakub Dawidek
6f8e88e1da Call ZFS_EXIT() after locking the vnode.
MFC after:	1 week
2009-09-08 15:37:01 +00:00
Poul-Henning Kamp
0901106030 Add necessary include 2009-09-08 15:18:25 +00:00