]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years agoAdd a missing part of jhb's previous commit. It looks like he had a
peter [Sat, 6 Mar 2004 00:44:59 +0000 (00:44 +0000)]
Add a missing part of jhb's previous commit.  It looks like he had a
patch chunk rejected that he missed.  This would manifest as a lock
assertion panic at boot (Giant not locked in kern_fork.c).

Obtained from:  jhb

20 years agoHave these instructions tell users to 'sh installworld_newk' instead
gad [Fri, 5 Mar 2004 23:10:49 +0000 (23:10 +0000)]
Have these instructions tell users to 'sh installworld_newk' instead
of './installworld_newk', because the scripts might not show up with
the +x bit set.

20 years agokthread_exit() no longer requires Giant, so don't force callers to acquire
jhb [Fri, 5 Mar 2004 22:42:17 +0000 (22:42 +0000)]
kthread_exit() no longer requires Giant, so don't force callers to acquire
Giant just to call kthread_exit().

Requested by: many

20 years agoLock Giant around the body of the adlink_loran() function used by the
jhb [Fri, 5 Mar 2004 22:41:22 +0000 (22:41 +0000)]
Lock Giant around the body of the adlink_loran() function used by the
adlink device kthreads.

20 years ago- Push down Giant in exit() and wait().
jhb [Fri, 5 Mar 2004 22:39:53 +0000 (22:39 +0000)]
- Push down Giant in exit() and wait().
- Push Giant down a bit in coredump() and call coredump() with the proc
  lock already held rather than unlocking it only to turn around and
  relock it.

Requested by: peter

20 years agoLock Giant around the single threading code in exec() to satisfy an
jhb [Fri, 5 Mar 2004 22:38:26 +0000 (22:38 +0000)]
Lock Giant around the single threading code in exec() to satisfy an
assertion in the single threading code.

20 years ago- Grab a share lock of the proctree lock while looking for a pid due to the
jhb [Fri, 5 Mar 2004 22:37:32 +0000 (22:37 +0000)]
- Grab a share lock of the proctree lock while looking for a pid due to the
  process group and session dereferences.  Also, check that p_pgrp and
  p_sesssion are NULL before dereferencing them.
- Push down Giant in fork1().

Requested by: peter

20 years agoAdd warning about loading/unloading drivers and identify
imp [Fri, 5 Mar 2004 22:21:18 +0000 (22:21 +0000)]
Add warning about loading/unloading drivers and identify

20 years agoUndo the merger of mlock()/vslock and munlock()/vsunlock() and the
truckman [Fri, 5 Mar 2004 22:03:11 +0000 (22:03 +0000)]
Undo the merger of mlock()/vslock and munlock()/vsunlock() and the
introduction of kern_mlock() and kern_munlock() in
        src/sys/kern/kern_sysctl.c      1.150
        src/sys/vm/vm_extern.h          1.69
        src/sys/vm/vm_glue.c            1.190
        src/sys/vm/vm_mmap.c            1.179
because different resource limits are appropriate for transient and
"permanent" page wiring requests.

Retain the kern_mlock() and kern_munlock() API in the revived
vslock() and vsunlock() functions.

Combine the best parts of each of the original sets of implementations
with further code cleanup.  Make the mclock() and vslock()
implementations as similar as possible.

Retain the RLIMIT_MEMLOCK check in mlock().  Move the most strigent
test, which can return EAGAIN, last so that requests that have no
hope of ever being satisfied will not be retried unnecessarily.

Disable the test that can return EAGAIN in the vslock() implementation
because it will cause the sysctl code to wedge.

Tested by: Cy Schubert <Cy.Schubert AT komquats.com>

20 years agoTeach dd(1) about parity bits.
phk [Fri, 5 Mar 2004 19:35:51 +0000 (19:35 +0000)]
Teach dd(1) about parity bits.

20 years agoAdd a test-target and reference vectors for the character converions.
phk [Fri, 5 Mar 2004 19:30:13 +0000 (19:30 +0000)]
Add a test-target and reference vectors for the character converions.

20 years agoThe roundrobin callout from sched_4bsd is MPSAFE, so set up the
rwatson [Fri, 5 Mar 2004 19:27:04 +0000 (19:27 +0000)]
The roundrobin callout from sched_4bsd is MPSAFE, so set up the
callout as MPSAFE to avoid grabbing Giant.

Reviewed by: jhb

20 years agoReword two more lines to avoid wrapping.
gad [Fri, 5 Mar 2004 18:16:28 +0000 (18:16 +0000)]
Reword two more lines to avoid wrapping.

20 years agoDocument a sysctl.
njl [Fri, 5 Mar 2004 18:08:23 +0000 (18:08 +0000)]
Document a sysctl.

Submitted by: Craig Rodrigues <rodrigc@crodrigues.org>

20 years agoA user can set tz_requested via the hw.acpi.thermal.tzX.active sysctl.
njl [Fri, 5 Mar 2004 18:06:31 +0000 (18:06 +0000)]
A user can set tz_requested via the hw.acpi.thermal.tzX.active sysctl.
The previous logic meant that if a user sets it to a minimal cooling value
acpi_thermal will not use higher cooling levels.  Reverse the logic so that
the user requesting a level (say, 2) also gets 0 - 1 also.

PR: kern/61592
Submitted by: Andrew Thompson <andy@fud.org.nz>

20 years agoPut "failed to set signal flags properly for ast()" check under
rwatson [Fri, 5 Mar 2004 17:35:28 +0000 (17:35 +0000)]
Put "failed to set signal flags properly for ast()" check under
DIAGNOSTIC instead of INVARIANTS.  INVARIANTS is intended for tests
that don't substantially change code flow or behavior (passive), but
this test required locking both the proc lock and scheduler lock
in order to execute.  It also appears to be a very advisory diagnostic
as opposed to an invariant violation.

Following discussion with: bde

20 years agoThis interface doesn't return any values.
ru [Fri, 5 Mar 2004 16:38:01 +0000 (16:38 +0000)]
This interface doesn't return any values.

Submitted by: Mark Santcroos

20 years agoRemoved definition of NOPROG. It was just a style bug (a NetBSD macro
bde [Fri, 5 Mar 2004 16:13:22 +0000 (16:13 +0000)]
Removed definition of NOPROG.  It was just a style bug (a NetBSD macro
that became obsolete in NetBSD 22 months ago).

Submitted by: ru

20 years agoBack out attempt to bypass contrib's getopt.h on the way,
ache [Fri, 5 Mar 2004 16:12:31 +0000 (16:12 +0000)]
Back out attempt to bypass contrib's getopt.h on the way,
it makes cross-build fail.
Prefer invisible incorrect -current build to visible failure.

20 years agoFixed indentation of conditionals.
bde [Fri, 5 Mar 2004 15:59:33 +0000 (15:59 +0000)]
Fixed indentation of conditionals.

Submitted by: Rostislav Krasny <rosti_bsd@yahoo.com>

20 years agoFixed insertion sort errors in prototype list.
bde [Fri, 5 Mar 2004 15:30:40 +0000 (15:30 +0000)]
Fixed insertion sort errors in prototype list.

20 years agoRemoved garbage:
bde [Fri, 5 Mar 2004 15:22:05 +0000 (15:22 +0000)]
Removed garbage:
- completely unused things
- all of rev.1.102 (C++ support).  <sys/cdefs.h> is included by the
  prerequisite <sys/types.h>.  __BEGIN_DECLS/__END_DECLS has no effect
  (except possibly if undefined behaviour is invoked using a hack like
  defining away __inline) since this header doesn't really support any
  extern functions.

20 years agoMFi386: (all: keep a comment in sync with code, and don't depend on
bde [Fri, 5 Mar 2004 14:31:21 +0000 (14:31 +0000)]
MFi386: (all: keep a comment in sync with code, and don't depend on
namespace pollution).

20 years agoInclude <machine/psl.h> for the definition of PSL_I instead of depending
bde [Fri, 5 Mar 2004 14:23:13 +0000 (14:23 +0000)]
Include <machine/psl.h> for the definition of PSL_I instead of depending
on namespace pollution in <machine/cpufunc.h>.

20 years agoUse ascii null char ('\0') instead of NULL.
nyan [Fri, 5 Mar 2004 13:58:40 +0000 (13:58 +0000)]
Use ascii null char ('\0') instead of NULL.

Submitted by: Stefan Farfeleder <stefan@fafoe.narf.at>

20 years agoFix syntax errors and wrong function prototypes in several MD header
le [Fri, 5 Mar 2004 09:19:59 +0000 (09:19 +0000)]
Fix syntax errors and wrong function prototypes in several MD header
files when using non-GNUC compilers.

PR:             kern/58515
Submitted by:   Stefan Farfeleder <stefan@fafoe.narf.at>
Approved by:    grog (mentor), obrien

20 years agoOne tab too much.
pjd [Fri, 5 Mar 2004 09:18:40 +0000 (09:18 +0000)]
One tab too much.

20 years agoTeach 'hostname' script how to act inside a jail.
pjd [Fri, 5 Mar 2004 09:17:01 +0000 (09:17 +0000)]
Teach 'hostname' script how to act inside a jail.

No objections from: mtm, arch@

20 years agoremove superfluous space before end of backtick expression
rse [Fri, 5 Mar 2004 08:49:59 +0000 (08:49 +0000)]
remove superfluous space before end of backtick expression

20 years agofix reference to sysctl variable: machdep.an_cache_mode -> hw.an.an_cache_mode
rse [Fri, 5 Mar 2004 08:42:15 +0000 (08:42 +0000)]
fix reference to sysctl variable: machdep.an_cache_mode -> hw.an.an_cache_mode

20 years agoBump __FreeBSD_version for the NULL == ((void *)0)
markm [Fri, 5 Mar 2004 08:34:59 +0000 (08:34 +0000)]
Bump __FreeBSD_version for the NULL == ((void *)0)

Requested by: kkenn

20 years agoMake NULL a (void*)0 whereever possible, and fix the warnings(-Werror)
markm [Fri, 5 Mar 2004 08:10:19 +0000 (08:10 +0000)]
Make NULL a (void*)0 whereever possible, and fix the warnings(-Werror)
that this provokes. "Wherever possible" means "In the kernel OR NOT
C++" (implying C).

There are places where (void *) pointers are not valid, such as for
function pointers, but in the special case of (void *)0, agreement
settles on it being OK.

Most of the fixes were NULL where an integer zero was needed; many
of the fixes were NULL where ascii <nul> ('\0') was needed, and a
few were just "other".

Tested on: i386 sparc64

20 years agoDocument that ENABLE_ALART controls the alarm on Intel intpm driver.
obrien [Fri, 5 Mar 2004 08:04:10 +0000 (08:04 +0000)]
Document that ENABLE_ALART controls the alarm on Intel intpm driver.

Submitted by: peter

20 years agoWhen this script included NetBSD specific logic, the NetBSD branch
mtm [Fri, 5 Mar 2004 08:03:04 +0000 (08:03 +0000)]
When this script included NetBSD specific logic, the NetBSD branch
included a start_precmd check for gated. The precommand was not
executed in the FreeBSD branch. When I did a mass removal of
NetBSD specific logic a while back this file apparently got only
a partial treatement. This bug did not have any functional consequences,
however, since the precommand was not declared to the rc.subr routines.

Noticed by: pjd

20 years agoVESA* is a mistake - it cannot exist due to requiring a 32 bit kernel for
obrien [Fri, 5 Mar 2004 07:56:23 +0000 (07:56 +0000)]
VESA* is a mistake - it cannot exist due to requiring a 32 bit kernel for
BIOS calls and/or VM86.

Submitted by: peter

20 years agoThe syslogd script should require that /var is cleaned before it runs.
mtm [Fri, 5 Mar 2004 07:55:04 +0000 (07:55 +0000)]
The syslogd script should require that /var is cleaned before it runs.
Otherwise it could be in the situation where its log socket is removed
after it has started.

Noticed by: jhay

20 years agoSync with i386/NOTES.
obrien [Fri, 5 Mar 2004 07:48:47 +0000 (07:48 +0000)]
Sync with i386/NOTES.

20 years agoAdd comment for 'mptable'.
obrien [Fri, 5 Mar 2004 07:44:17 +0000 (07:44 +0000)]
Add comment for 'mptable'.

Submitted by: peter

20 years agoRemove scripts we don't use from requirement lines. These were
mtm [Fri, 5 Mar 2004 07:43:38 +0000 (07:43 +0000)]
Remove scripts we don't use from requirement lines. These were
hold-overs from the initial NetBSD import.

20 years agoNote that imp is working on un-shimming this driver, afterwards it should
obrien [Fri, 5 Mar 2004 06:23:08 +0000 (06:23 +0000)]
Note that imp is working on un-shimming this driver, afterwards it should
work on AMD64.

20 years agoRemoved include of bsd.own.mk. It was just a style bug (half-baked
bde [Fri, 5 Mar 2004 05:48:58 +0000 (05:48 +0000)]
Removed include of bsd.own.mk.  It was just a style bug (half-baked
chumminess with the implementation).

20 years agoThe PECOFF support is 32-bit only.
obrien [Fri, 5 Mar 2004 05:08:22 +0000 (05:08 +0000)]
The PECOFF support is 32-bit only.

Reviewed by: peter

20 years agoIn the last revision, I introduced a physical contiguity check that is both
alc [Fri, 5 Mar 2004 04:46:32 +0000 (04:46 +0000)]
In the last revision, I introduced a physical contiguity check that is both
unnecessary and wrong.  While it is necessary to verify that the page is
still free after dropping and reacquiring the free page queue lock, the
physical contiguity of the page can not change, making this check
unnecessary.  This check was wrong in that it could cause an out-of-bounds
array access.

Tested by: rwatson

20 years agoPECOFF_* isn't useful for AMD64.
obrien [Fri, 5 Mar 2004 04:38:58 +0000 (04:38 +0000)]
PECOFF_* isn't useful for AMD64.

20 years agoNew errata: SA-04:04. Also fix a typo in the URL for SA-04:03.
bmah [Fri, 5 Mar 2004 04:19:06 +0000 (04:19 +0000)]
New errata:  SA-04:04.  Also fix a typo in the URL for SA-04:03.

20 years agoMerge the following from the English version:
rushani [Fri, 5 Mar 2004 04:14:21 +0000 (04:14 +0000)]
Merge the following from the English version:

  1.61 -> 1.65 errata/article.sgml

20 years agoFix a typo.
bms [Fri, 5 Mar 2004 02:50:28 +0000 (02:50 +0000)]
Fix a typo.

20 years agoFix style breakage.
takawata [Fri, 5 Mar 2004 02:48:22 +0000 (02:48 +0000)]
Fix style breakage.

Noticed by: njl

20 years agoForced commit to note:
wes [Fri, 5 Mar 2004 02:05:56 +0000 (02:05 +0000)]
Forced commit to note:

1) Review/prodding was provided by josef@, not joe@.

2) Josef was going to commit this and was only looking for a review.
   My bad.

Reviewed by: josef@

20 years agoUse getopt instead of hand-rolled argument parsing. Usage remains
wes [Fri, 5 Mar 2004 01:52:09 +0000 (01:52 +0000)]
Use getopt instead of hand-rolled argument parsing.  Usage remains
the same, no man page changes required.

PR: bin/48313
Submitted by: Stefan Farfeleder <stefan@fafoe.dyndns.org>
Reviewed by: joe@

20 years agoMove -I${DESTDIR}/usr/include/gnu up, before
ache [Fri, 5 Mar 2004 01:47:03 +0000 (01:47 +0000)]
Move -I${DESTDIR}/usr/include/gnu up, before
-I${DESTDIR}/usr/include to pick correct regex.h

20 years agoCorrectly read SCHILY.nlink from pax-format archives.
kientzle [Fri, 5 Mar 2004 00:09:53 +0000 (00:09 +0000)]
Correctly read SCHILY.nlink from pax-format archives.
In particular, -tv output for pax-format archives now
lists everything that ls -l does.

20 years ago- Some older Atheros drivers want KeInitializeTimer(), so implement it,
wpaul [Thu, 4 Mar 2004 23:04:02 +0000 (23:04 +0000)]
- Some older Atheros drivers want KeInitializeTimer(), so implement it,
  along with KeInitializeTimerEx(), KeSetTimer(), KeSetTimerEx(),
  KeCancelTimer(), KeReadStateTimer() and KeInitializeDpc(). I don't
  know for certain that these will make the Atheros driver happy since
  I don't have the card/driver combo needed to test it, but these are
  fairly independent so they shouldn't break anything else.

- Debugger() is present even in kernels without options DDB, so no
  conditional compilation is necessary (pointed out by bde).

- Remove the extra km_acquirecnt member that I added to struct kmutant
  and embed it within an unused portion of the structure instead, so that
  we don't make the structure larger than it's defined to be in Windows.
  I don't know what crack I was smoking when I decided it was ok to do
  this, but it's worn off now.

20 years agoHook ieee80211 manpages up to the build; fix 1 snafu which crept in.
bms [Thu, 4 Mar 2004 21:13:06 +0000 (21:13 +0000)]
Hook ieee80211 manpages up to the build; fix 1 snafu which crept in.

Noticed by: Matthias Schuendehuette
Pointy hat to: bms

20 years agoForgotten commit: Hook /etc/rc.d/ike up to the build.
bms [Thu, 4 Mar 2004 21:03:36 +0000 (21:03 +0000)]
Forgotten commit: Hook /etc/rc.d/ike up to the build.

20 years agoImplement a crude but functional usbd_ratecheck() to limit the number
phk [Thu, 4 Mar 2004 20:49:03 +0000 (20:49 +0000)]
Implement a crude but functional usbd_ratecheck() to limit the number
of "usb0: %d scheduling overruns" messages I have to contend with.

20 years agoAdd -I${DESTDIR}/usr/include before -I${SRCDIR}/include to pick right
ache [Thu, 4 Mar 2004 20:36:18 +0000 (20:36 +0000)]
Add -I${DESTDIR}/usr/include before -I${SRCDIR}/include to pick right
getopt.h (as workaround until it will be removed from contrib)

20 years agoUpdate URL of GCC status page to GCC 3.3
josef [Thu, 4 Mar 2004 20:29:27 +0000 (20:29 +0000)]
Update URL of GCC status page to GCC 3.3

Approved by:    simon(mentor)
Submitted by:   Stefan Farfeleder <stefan@fafoe.narf.at>
PR:             docs/63370

20 years agoEliminate hard sentence breaks.
trhodes [Thu, 4 Mar 2004 19:12:27 +0000 (19:12 +0000)]
Eliminate hard sentence breaks.
General markup fixes (use the .Dq macro).

20 years agoMinor stylistic improvements in the SPECIAL_INSTALLCHECKS section, mainly
gad [Thu, 4 Mar 2004 18:41:35 +0000 (18:41 +0000)]
Minor stylistic improvements in the SPECIAL_INSTALLCHECKS section, mainly
to reduce the number of wrapped lines.

Suggested by: bde

20 years agoRewrite TCP segment reassembly note to mention SA-04:04, note MFC,
bmah [Thu, 4 Mar 2004 17:06:30 +0000 (17:06 +0000)]
Rewrite TCP segment reassembly note to mention SA-04:04, note MFC,
relocate to security advisory section.

20 years agoMake unprivilaged user can see battery info.
takawata [Thu, 4 Mar 2004 17:03:49 +0000 (17:03 +0000)]
Make unprivilaged user can see battery info.

20 years agoOnly setup sii_reset on sii311[24].
sos [Thu, 4 Mar 2004 16:39:59 +0000 (16:39 +0000)]
Only setup sii_reset on sii311[24].

20 years agoForced commit to note that last commit was...
roam [Thu, 4 Mar 2004 15:54:34 +0000 (15:54 +0000)]
Forced commit to note that last commit was...

Approved by: silence on -audit

20 years agoMake rpc.lockd bind to a reserved port, since there are NFS clients
roam [Thu, 4 Mar 2004 15:52:28 +0000 (15:52 +0000)]
Make rpc.lockd bind to a reserved port, since there are NFS clients
which ignore NLM requests not coming from a reserved port.

PR: 56500
Submitted by: Jonathan Lennox <lennox@cs.columbia.edu>
MFC after: 1 week

20 years agoRev. 1.32 moved a comment to the wrong line. The hack refered to
mtm [Thu, 4 Mar 2004 15:46:14 +0000 (15:46 +0000)]
Rev. 1.32 moved a comment to the wrong line. The hack refered to
in the comment applies to a decision that needs to be made in relation
to the year 2000.

In fact, that statement probably should be changed to be
more generic (getting the year from the current time perhaps). Otherwise,
starting in 2069 two digit year conversions in date(1) will start assuming
1900 instead of 2000. hehe.

20 years agoUse auto-sense data provided by the lowlevel ATA code.
thomas [Thu, 4 Mar 2004 15:37:39 +0000 (15:37 +0000)]
Use auto-sense data provided by the lowlevel ATA code.

20 years agomove in6_addmulti()/in6_delmulti() into mld6.c
ume [Thu, 4 Mar 2004 15:07:42 +0000 (15:07 +0000)]
move in6_addmulti()/in6_delmulti() into mld6.c

Obtained from: KAME

20 years agoNursemaid: Fix tinderbox builds by removing the shadowing of the global
bms [Thu, 4 Mar 2004 14:16:12 +0000 (14:16 +0000)]
Nursemaid: Fix tinderbox builds by removing the shadowing of the global
preprocessor macro DEBUG. DEBUG() -> CTAU_DEBUG().

20 years agoJust because the timecounter reads the same value on two samples
phk [Thu, 4 Mar 2004 14:14:23 +0000 (14:14 +0000)]
Just because the timecounter reads the same value on two samples
after each other doesn't mean that nothing happened.

20 years ago&release.current; security fix branch does not exist, it should be
rushani [Thu, 4 Mar 2004 13:59:45 +0000 (13:59 +0000)]
&release.current; security fix branch does not exist, it should be
&release.prev; security fix branch.

20 years agoAdd an MLINKS entry for radiotap(9).
bms [Thu, 4 Mar 2004 12:54:22 +0000 (12:54 +0000)]
Add an MLINKS entry for radiotap(9).

20 years agoInitial import of documentation for net80211 interface code.
bms [Thu, 4 Mar 2004 12:52:10 +0000 (12:52 +0000)]
Initial import of documentation for net80211 interface code.
Of particular interest here is the ieee80211_radiotap(9) man page.

TODO: Flesh out ieee80211_node.9, ieee80211_proto.9.

Submitted by: bms (with help from Darron Broad)

20 years agomissing splx().
ume [Thu, 4 Mar 2004 12:08:25 +0000 (12:08 +0000)]
missing splx().

Obtained from: KAME
MFC after: 3 days

20 years agoFixed the XXX'ed namespace pollution in rev.1.54 by using
bde [Thu, 4 Mar 2004 11:35:30 +0000 (11:35 +0000)]
Fixed the XXX'ed namespace pollution in rev.1.54 by using
<machine/limits.h> and __CHAR_BIT instead of <sys/limits.h> and CHAR_BIT.
some reason I didn't use the BSD spelling NBBY.

20 years agoDon't manually optimize for 20 year old compilers by casting to u_int
bde [Thu, 4 Mar 2004 11:20:02 +0000 (11:20 +0000)]
Don't manually optimize for 20 year old compilers by casting to u_int
to get a free check for negative ints.  Rev.1.35 got my request to
remove the cast mostly backwards.

20 years agoFixed some style bugs (mainly misformatting, and pointers to `struct thread'
bde [Thu, 4 Mar 2004 11:11:59 +0000 (11:11 +0000)]
Fixed some style bugs (mainly misformatting, and pointers to `struct thread'
named p).

20 years agoFixed insertion sort errors in includes and prototypes. This was more
bde [Thu, 4 Mar 2004 10:56:29 +0000 (10:56 +0000)]
Fixed insertion sort errors in includes and prototypes.  This was more
than a style bug for the includes -- queue.h is a prerequisite for
_lock.h and _mutex.h but was included after them.

Removed bogus prototype for fget_locked().  The prototype was originally
needed to support K&R but was bogotified by converting the function header
to new-style.

20 years agoCorrect year in copyrights.
pjd [Thu, 4 Mar 2004 10:22:42 +0000 (10:22 +0000)]
Correct year in copyrights.

20 years agoRecord exactly where this file was copied from. It wasn't repo-copied so
bde [Thu, 4 Mar 2004 10:18:17 +0000 (10:18 +0000)]
Record exactly where this file was copied from.  It wasn't repo-copied so
this is not very obvious.

Fixed some style bugs (mainly missing parentheses around return values).

20 years agoFixed some style bugs (mainly English usage errors in comments).
bde [Thu, 4 Mar 2004 09:56:29 +0000 (09:56 +0000)]
Fixed some style bugs (mainly English usage errors in comments).

20 years agoFixed some style bugs (mainly misplaced comments, and totally disordered
bde [Thu, 4 Mar 2004 09:47:09 +0000 (09:47 +0000)]
Fixed some style bugs (mainly misplaced comments, and totally disordered
declarations in acct_process()).

20 years agoMinor style fixes. In vm_daemon(), don't fetch the rss limit long before
bde [Thu, 4 Mar 2004 09:36:46 +0000 (09:36 +0000)]
Minor style fixes.  In vm_daemon(), don't fetch the rss limit long before
it is needed.

20 years agoImproved the description of the installkernel targets.
bde [Thu, 4 Mar 2004 09:14:30 +0000 (09:14 +0000)]
Improved the description of the installkernel targets.

Don't hide what we do to force failures for the installkernel targets
or other targets.

20 years agoDocument the virecover_enable knob.
mtm [Thu, 4 Mar 2004 08:25:53 +0000 (08:25 +0000)]
Document the virecover_enable knob.
From the PR:
Certain MTA configurations mean that the notifications from
virecover keep bouncing; so here's a patch to allow administrators
to turn them off.

PR: conf/54910
Reminded by: ru

20 years agoSync to 1.165 of usbdevs
sanpei [Thu, 4 Mar 2004 07:22:30 +0000 (07:22 +0000)]
Sync to 1.165 of usbdevs

20 years agoAdd support SimpleTech UCF-100 USB CompactFlash reader(OnSpec Electronic, Inc.)
sanpei [Thu, 4 Mar 2004 07:20:48 +0000 (07:20 +0000)]
Add support SimpleTech UCF-100 USB CompactFlash reader(OnSpec Electronic, Inc.)

PR: kern/63619
Submitted by: Greg Rivers <gcr@sa.fedex.com>
MFC after: 1 week

20 years agoAdd support PLANEX COMMUNICATIONS Inc. GN-1200TC (8169S)
sanpei [Thu, 4 Mar 2004 06:42:46 +0000 (06:42 +0000)]
Add support PLANEX COMMUNICATIONS Inc. GN-1200TC (8169S)

20 years agoFix an off-by-one error and rework our EC space handler. Writing to address
njl [Thu, 4 Mar 2004 05:58:50 +0000 (05:58 +0000)]
Fix an off-by-one error and rework our EC space handler.  Writing to address
0xFF would fail previously as AE_BAD_PARAMETER.  It's unknown if this caused
any actual problems.

20 years agoPart 2 of Project Evil: Pretend to be Windows 2000 for buggy ASL that
njl [Thu, 4 Mar 2004 05:57:41 +0000 (05:57 +0000)]
Part 2 of Project Evil:  Pretend to be Windows 2000 for buggy ASL that
always expects to be running on some MS OS.  A survey of ASL shows that
this is the 2nd most common expected OS value.  (1st is Win98 and we don't
emulate its buggy ACPI support.)  Our ACPI support is similar to Win2k,
also.  Put this behavior under ACPICA_PEDANTIC so we can get back to our
previous behavior for OSV testing.

20 years agoAdd NetBSD 1.6.2.
maxim [Thu, 4 Mar 2004 05:34:29 +0000 (05:34 +0000)]
Add NetBSD 1.6.2.

20 years agoDon't disable Cx support and throttling on machines with a P_BLK_LEN != 6
njl [Thu, 4 Mar 2004 05:17:52 +0000 (05:17 +0000)]
Don't disable Cx support and throttling on machines with a P_BLK_LEN != 6
even though the spec mandates this.  Some have a value of 5 to indicate
throttling + C2 and some have 7 to indicate an extra C3 state.  Support
throttling if the value is >= 4, C2 for >= 5, and C3 for >= 6.

20 years agoAdd a "quirks" value to disable quirks handling for a given boot.
njl [Thu, 4 Mar 2004 04:42:59 +0000 (04:42 +0000)]
Add a "quirks" value to disable quirks handling for a given boot.
Also, disable quirks if booting with a custom DSDT.  Add a quirk
to disable loading ACPI so known bad systems can be completely
blacklisted.

20 years agoAdd a new option to mountd(8), -p <port>. This allows the user to specify
bms [Thu, 4 Mar 2004 04:42:52 +0000 (04:42 +0000)]
Add a new option to mountd(8), -p <port>. This allows the user to specify
a known port for use in firewall rulesets; otherwise the port is chosen
at run-time by bindresvport().

MFC after: 1 week

20 years agoRemove some long unused definitions.
alc [Thu, 4 Mar 2004 04:26:14 +0000 (04:26 +0000)]
Remove some long unused definitions.

20 years agoRemove unneeded label 'done2' from socket(). We now grab Giant
rwatson [Thu, 4 Mar 2004 01:57:48 +0000 (01:57 +0000)]
Remove unneeded label 'done2' from socket().  We now grab Giant
only around socreate(), and don't need it for file descriptor
accesses.

Submitted by: sam

20 years agos/considred/considered/
brueffer [Thu, 4 Mar 2004 00:52:16 +0000 (00:52 +0000)]
s/considred/considered/

20 years agoAdd sanity checks to the ndis_packet and ndis_buffer pool handling
wpaul [Thu, 4 Mar 2004 00:17:14 +0000 (00:17 +0000)]
Add sanity checks to the ndis_packet and ndis_buffer pool handling
routines to guard against problems caused by (possibly) buggy drivers.

The RealTek 8180 wireless driver calls NdisFreeBuffer() to release
some of its buffers _after_ it's already called NdisFreeBufferPool()
to destroy the pool to which the buffers belong. In our implementation,
this error causes NdisFreeBuffer() to touch stale heap memory.

If you are running a release kernel, and hence have INVARIANTS et al
turned off, it turns out nothing happens. But if you're using a
development kernel config with INVARIANTS on, the malloc()/free()
sanity checks will scribble over the pool memory with 0xdeadc0de
once it's released so that any attempts to touch it will cause a
trap, and indeed this is what happens. It happens that I run 5.2-RELEASE
on my laptop, so when I tested the rtl8180.sys driver, it worked fine
for me, but people trying to run it with development systems checked
out or cvsupped from -current would get a page fault on driver load.

I can't find any reason why the NDISulator would cause the RealTek
driver to do the NdisFreeBufferPool() prematurely, and the same driver
obviously works with Windows -- or at least, it doesn't cause a crash:
the Microsoft documentation for NdisFreeBufferPool() says that failing
to return all buffers to the pool before calling  NdisFreeBufferPool()
causes a memory leak.

I've written to my contacts at RealTek asking them to check if this
is indeed a bug in their driver. In the meantime, these new sanity checks
will catch this problem and issue a warning rather than causing a trap.
The trick is to keep a count of outstanding buffers for each buffer pool,
and if the driver tries to call NdisFreeBufferPool() while there are still
buffers outstanding, we mark the pool for deletion and then defer
destroying it until after the last buffer has been reclaimed.

20 years agoUse different dummy wait channels to avoid panic in msleep().
des [Wed, 3 Mar 2004 23:03:18 +0000 (23:03 +0000)]
Use different dummy wait channels to avoid panic in msleep().

Reviewed by: jhb

20 years ago[this is just a forced commit to say:] The time_t-specific safety measure
gad [Wed, 3 Mar 2004 22:56:41 +0000 (22:56 +0000)]
[this is just a forced commit to say:]  The time_t-specific safety measure
added by the sparc64_installcheck target is mostly from Marcel, although
it includes some adjustments of my own...