]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years ago- If error string begins with "warning: ", don't exit, treat it as a warning
pjd [Sat, 28 Aug 2004 02:29:40 +0000 (02:29 +0000)]
- If error string begins with "warning: ", don't exit, treat it as a warning
  only.
- Use getprogname() function when informing about versions problem.

20 years agoDon't allow to insert providers, which are too small.
pjd [Sat, 28 Aug 2004 02:02:48 +0000 (02:02 +0000)]
Don't allow to insert providers, which are too small.

Reported by: Michael Handler <handler@grendel.net>

20 years agoBackout the previous backout (with scott's ok). sched_ule.c:1.122 is
peter [Sat, 28 Aug 2004 01:04:44 +0000 (01:04 +0000)]
Backout the previous backout (with scott's ok).  sched_ule.c:1.122 is
believed to fix the problem with ULE that this change triggered.

20 years agos/smp_rv_mtx/smp_ipi_mtx/g
obrien [Sat, 28 Aug 2004 00:49:55 +0000 (00:49 +0000)]
s/smp_rv_mtx/smp_ipi_mtx/g

Requested by: jhb

20 years agoCommit Jeff's suggested changes for avoiding a bug that is exposed by
peter [Sat, 28 Aug 2004 00:49:22 +0000 (00:49 +0000)]
Commit Jeff's suggested changes for avoiding a bug that is exposed by
preemption and/or the rev 1.79 kern_switch.c change that was backed out.

The thread was being assigned to a runq without adding in the load, which
would cause the counter to hit -1.

20 years ago- Auto generate device listings for the following drivers: snd_ad1816,
simon [Fri, 27 Aug 2004 22:58:29 +0000 (22:58 +0000)]
- Auto generate device listings for the following drivers: snd_ad1816,
  snd_cmi, snd_cs4281, snd_ds1, snd_emu10k1, snd_es137x, and snd_solo.

MFC after: 5 days

20 years agoremove unused prototype
sos [Fri, 27 Aug 2004 22:14:45 +0000 (22:14 +0000)]
remove unused prototype

20 years agoMake code ready to switch debug.mpsafenet to 1 since I've not able to
rik [Fri, 27 Aug 2004 22:14:26 +0000 (22:14 +0000)]
Make code ready to switch debug.mpsafenet to 1 since I've not able to
commit MPSAFE code for now it is just IFF_NEEDSGIANT.

20 years agoAdd manual pages for several sound drivers.
simon [Fri, 27 Aug 2004 22:06:56 +0000 (22:06 +0000)]
Add manual pages for several sound drivers.

Submitted by: Atte Peltomaki <koston@iki.fi>
Reviewed by: cg
MFC after: 5 days

20 years agoCatch up with the drive-by renaming of IA32 to COMPAT_IA32. Missed
marcel [Fri, 27 Aug 2004 21:57:00 +0000 (21:57 +0000)]
Catch up with the drive-by renaming of IA32 to COMPAT_IA32. Missed
11 days ago when all the other places were fixed and finally caught
by the tinderbox run...

20 years agoForced commit to add additional comment.
gibbs [Fri, 27 Aug 2004 21:33:33 +0000 (21:33 +0000)]
Forced commit to add additional comment.
 o Add MOUSE_PS2_TAP into syncmask[0] correctly when we're built with
   PSM_CONFIG_FORCETAP.  The previous code from revision 1.56 attempted
   to do this but updated the a mask variable that would be clobbered later.
o  If syncmask[1] hasn't yet been set when ioctl(MOUSE_GETMODE) is called,
   zero syncmask[0] and syncmask[1] so that sync validation is effectively
   disabled in userland applications and moused doesn't misbehave.  The
   psm driver performs sync validation so there is no loss in functionality
   in clearing the sync mask.
o  If PSM_NEED_SYNCBITS is set, setup syncmask[1], even if it's already
   correct.  This prevents the PSM_NEED_SYNCBITS bit from being left on
   after a re-initialisation and erroneously causing a subsequent
   out-of-sync packet to mis-set syncmask[1].

20 years ago- Improve handling of trailing punctuation characters for the Xr macro.
simon [Fri, 27 Aug 2004 21:32:24 +0000 (21:32 +0000)]
- Improve handling of trailing punctuation characters for the Xr macro.

MFC after: 3 days

20 years agoMove config_new_drive() to the correct place and rename it to
le [Fri, 27 Aug 2004 21:32:18 +0000 (21:32 +0000)]
Move config_new_drive() to the correct place and rename it to
gv_config_new_drive().

20 years agoFix a comment, IA32 was renamed to COMPAT_IA32
arved [Fri, 27 Aug 2004 21:29:20 +0000 (21:29 +0000)]
Fix a comment, IA32 was renamed to COMPAT_IA32

Approved by: marcel

20 years agoImprove sync recovery algorithm:
gibbs [Fri, 27 Aug 2004 21:25:16 +0000 (21:25 +0000)]
Improve sync recovery algorithm:
 o Remove PSM_SYNCERR_THRESHOLD1.  This value specified how many sync
   errors were required before the mouse is re-initialised.
   Re-initialisation is now done after (packetsize * 2) sync errors as
   things aren't likely to improve after that.

 o Reset lastinputerror when re-initialisation occurs.  We don't want
   to continue to drop data after re-initialisation.

 o Count the number of failed packets independently of the syncerrors
   statistic.  syncerrors is useful for recovering sync within a single
   packet.  pkterrors allows us to detect when the mouse changes its
   packet mode due to some external event (e.g. KVM switch).

 o Reinitialize the mouse if we see more than psmpkterrthresh errors
   during the validation period.  The validation period begins as soon
   as a sync error is detected and continues until psmerrsecs/msecs
   time has elapsed.  The defaults for these two values force a reset
   if we see two packet errors in a 2 second period.  This allows rapid
   detection of packet framing errors caused by the mouse changing packet
   modes.

 o Export psmpkterrthresh as a sysctl

 o Export psmloglevel as a sysctl.

 o Enable more debugging code to be enabled at runtime via psmloglevel.

 o Simplify verbose conditioned loging by using a VLOG macro.

 o Add several comments describing the sync recovery algorithm of
   this driver.

Large Portions by: Brian Somers <brian@Awfulhak.org>
Inspired and Frustrated by: Belkin KVMs
Reviewed by: njl, philip

20 years agoPoll() uses the array smallbits that is big enough to hold 32 struct
andre [Fri, 27 Aug 2004 21:23:50 +0000 (21:23 +0000)]
Poll() uses the array smallbits that is big enough to hold 32 struct
pollfd's to avoid calling malloc() on small numbers of fd's.  Because
smalltype's members have type char, its address might be misaligned
for a struct pollfd.  Change the array of char to an array of struct
pollfd.

PR: kern/58214
Submitted by: Stefan Farfeleder <stefan@fafoe.narf.at>
Reviewed by: bde (a long time ago)
MFC after: 3 days

20 years agoFrom __FreeBSD_version 600001 on PFIL_HOOKS is permanently in the kernel and
andre [Fri, 27 Aug 2004 20:01:08 +0000 (20:01 +0000)]
From __FreeBSD_version 600001 on PFIL_HOOKS is permanently in the kernel and
doesn't require to include opt_pfil_hooks.h.

20 years ago- Auto generate device listing for ath(4).
simon [Fri, 27 Aug 2004 19:59:29 +0000 (19:59 +0000)]
- Auto generate device listing for ath(4).

MFC after: 3 days

20 years agoopt_pfil_hooks.h is no longer required.
andre [Fri, 27 Aug 2004 19:57:33 +0000 (19:57 +0000)]
opt_pfil_hooks.h is no longer required.

20 years agoWhen detaching an interface, don't leave an obsolete pointer to the
brooks [Fri, 27 Aug 2004 19:42:40 +0000 (19:42 +0000)]
When detaching an interface, don't leave an obsolete pointer to the
soon to be deleted struct ifnet around.

PR: kern/52260
MFC After: 3 days

20 years agoMove the kernel-specific logic to adjust frompc from MI to MD. For
marcel [Fri, 27 Aug 2004 19:42:35 +0000 (19:42 +0000)]
Move the kernel-specific logic to adjust frompc from MI to MD. For
these two reasons:
1. On ia64 a function pointer does not hold the address of the first
   instruction of a functions implementation. It holds the address
   of a function descriptor. Hence the user(), btrap(), eintr() and
   bintr() prototypes are wrong for getting the actual code address.
2. The logic forces interrupt, trap and exception entry points to
   be layed-out contiguously. This can not be achieved on ia64 and is
   generally just bad programming.

The MCOUNT_FROMPC_USER macro is used to set the frompc argument to
some kernel address which represents any frompc that falls outside
the kernel text range. The macro can expand to ~0U to bail out in
that case.
The MCOUNT_FROMPC_INTR macro is used to set the frompc argument to
some kernel address to represent a call to a trap or interrupt
handler. This to avoid that the trap or interrupt handler appear to
be called from everywhere in the call graph. The macro can expand
to ~0U to prevent adjusting frompc. Note that the argument is selfpc,
not frompc.

This commit defines the macros on all architectures equivalently to
the original code in sys/libkern/mcount.c. People can take it from
here...

Compile-tested on: alpha, amd64, i386, ia64 and sparc64
Boot-tested on: i386

20 years ago- For column lists: ignore first item when it is likely to be a header.
simon [Fri, 27 Aug 2004 19:29:21 +0000 (19:29 +0000)]
- For column lists: ignore first item when it is likely to be a header.
- Clean up and improve handling of trailing punctuation characters.
- Handle the Pa macro.
- Give a warning when ignoring unimplemented mdoc commands.

MFC after: 3 days

20 years agoBump __FreeBSD_version to indicate permanent PFIL_HOOKS.
andre [Fri, 27 Aug 2004 19:23:24 +0000 (19:23 +0000)]
Bump __FreeBSD_version to indicate permanent PFIL_HOOKS.

20 years agoProvide extern declarations for btext and etext when GPROF is defined.
marcel [Fri, 27 Aug 2004 19:20:42 +0000 (19:20 +0000)]
Provide extern declarations for btext and etext when GPROF is defined.
These are referenced in subr_prof.c when building a profiling kernel.

20 years agoHave the linker provide the btext symbol when referenced. This is needed
marcel [Fri, 27 Aug 2004 19:13:35 +0000 (19:13 +0000)]
Have the linker provide the btext symbol when referenced. This is needed
for profiling kernels. Apparently, nobody has ever tried to build one on
alpha...

20 years agoThe machine-independent parts of the virtual memory system always pass a
alc [Fri, 27 Aug 2004 19:06:17 +0000 (19:06 +0000)]
The machine-independent parts of the virtual memory system always pass a
valid pmap to the pmap functions that require one.  Remove the checks for
NULL.  (These checks have their origins in the Mach pmap.c that was
integrated into BSD.  None of the new code written specifically for
FreeBSD included them.)

20 years agoSemi-gratuitous white space synchronization with KAME tree: to reduce
rwatson [Fri, 27 Aug 2004 18:41:41 +0000 (18:41 +0000)]
Semi-gratuitous white space synchronization with KAME tree: to reduce
diffs against #ifdef'd version of IPSEC, use "struct thread *p"
rather than "struct proc *p", fix some white space, and make some
already inconsistent white space inconsiste differently.

20 years agoApply error and success logic consistently to the function netisr_queue() and
andre [Fri, 27 Aug 2004 18:33:08 +0000 (18:33 +0000)]
Apply error and success logic consistently to the function netisr_queue() and
its users.

netisr_queue() now returns (0) on success and ERRNO on failure.  At the
moment ENXIO (netisr queue not functional) and ENOBUFS (netisr queue full)
are supported.

Previously it would return (1) on success but the return value of IF_HANDOFF()
was interpreted wrongly and (0) was actually returned on success.  Due to this
schednetisr() was never called to kick the scheduling of the isr.  However this
was masked by other normal packets coming through netisr_dispatch() causing the
dequeueing of waiting packets.

PR: kern/70988
Found by: MOROHOSHI Akihiko <moro@remus.dti.ne.jp>
MFC after: 3 days

20 years agoDiff reduce against KAME: minor white space synchronization to KAME
rwatson [Fri, 27 Aug 2004 18:24:59 +0000 (18:24 +0000)]
Diff reduce against KAME: minor white space synchronization to KAME
tree.

20 years agoHide a printf under bootverbose.
phk [Fri, 27 Aug 2004 17:08:24 +0000 (17:08 +0000)]
Hide a printf under bootverbose.

20 years agoProperly document and enable the IPFIREWALL_FORWARD option.
andre [Fri, 27 Aug 2004 15:46:16 +0000 (15:46 +0000)]
Properly document and enable the IPFIREWALL_FORWARD option.

20 years agoIn the case the destination of a packet was changed by the packet filter
andre [Fri, 27 Aug 2004 15:39:34 +0000 (15:39 +0000)]
In the case the destination of a packet was changed by the packet filter
to point to a local IP address; and the packet was sourced from this host
we fill in the m_pkthdr.rcvif with a pointer to the loopback interface.

Before the function ifunit("lo0") was used to obtain the ifp.  However
this is sub-optimal from a performance point of view and might be dangerous
if the loopback interface has been renamed.  Use the global variable 'loif'
instead which always points to the loopback interface.

Submitted by: brooks

20 years agoRemove a junk line left over from the recent IPFW to PFIL_HOOKS conversion.
andre [Fri, 27 Aug 2004 15:32:28 +0000 (15:32 +0000)]
Remove a junk line left over from the recent IPFW to PFIL_HOOKS conversion.

20 years agoAlways compile PFIL_HOOKS into the kernel and remove the associated kernel
andre [Fri, 27 Aug 2004 15:16:24 +0000 (15:16 +0000)]
Always compile PFIL_HOOKS into the kernel and remove the associated kernel
compile option.  All FreeBSD packet filters now use the PFIL_HOOKS API and
thus it becomes a standard part of the network stack.

If no hooks are connected the entire packet filter hooks section and related
activities are jumped over.  This removes any performance impact if no hooks
are active.

Both OpenBSD and DragonFlyBSD have integrated PFIL_HOOKS permanently as well.

20 years agoB1FLAGS -> BOOT_BOOT1_FLAGS, which is consistent with
yar [Fri, 27 Aug 2004 15:01:56 +0000 (15:01 +0000)]
B1FLAGS -> BOOT_BOOT1_FLAGS, which is consistent with
naming of other boot-related make vars.

20 years agoRemove the misleading description of the 'key' argument to
roam [Fri, 27 Aug 2004 14:51:21 +0000 (14:51 +0000)]
Remove the misleading description of the 'key' argument to
mpool_open(3) - it is *not* really used for synchronization; in fact,
it is not used at all.

PR: 70929
Submitted by: Martin Kammerhofer <dada@sbox.tugraz.at>
MFC after: 3 days

20 years agoFix the handling of "inflight" requests when doing reinit's.
sos [Fri, 27 Aug 2004 14:48:32 +0000 (14:48 +0000)]
Fix the handling of "inflight" requests when doing reinit's.
Add missing untimeout that would get lost in handling of some
error situations, and caused what looked like random timeouts
afterwards when the timeout fired.

20 years agoIntroduce g_alloc_bio() as a waiting variant of g_new_bio().
phk [Fri, 27 Aug 2004 14:43:11 +0000 (14:43 +0000)]
Introduce g_alloc_bio() as a waiting variant of g_new_bio().

Use in places where we can sleep and where we previously failed to check
for a NULL pointer.

MT5 candidate.

20 years agoTwo things:
osa [Fri, 27 Aug 2004 13:38:01 +0000 (13:38 +0000)]
Two things:
1. we are already under 6.0-CURRENT
2. add FreeBSD Design and Impl. book description
   (obtained from http://www.awprofessional.com/title/0201702452)

Approved by: wosch (maintainer)

20 years agoSet and reset the right bit for UDMA enable/disable on the ICH* chips.
sos [Fri, 27 Aug 2004 12:54:58 +0000 (12:54 +0000)]
Set and reset the right bit for UDMA enable/disable on the ICH* chips.

20 years agoFix a typo in a variable name.
yar [Fri, 27 Aug 2004 12:11:47 +0000 (12:11 +0000)]
Fix a typo in a variable name.

20 years agoRearrange the call to disk_destroy.
sos [Fri, 27 Aug 2004 12:03:18 +0000 (12:03 +0000)]
Rearrange the call to disk_destroy.

Suggested by: phk

20 years agoRemove bogus vrele() call added in previous.
tjr [Fri, 27 Aug 2004 11:24:31 +0000 (11:24 +0000)]
Remove bogus vrele() call added in previous.

20 years agoUse a temporary void * variable to work around a strict aliasing
iedowse [Fri, 27 Aug 2004 08:54:40 +0000 (08:54 +0000)]
Use a temporary void * variable to work around a strict aliasing
warning that gcc generates at -O2 and higher.

20 years agoDocument the -O option's new behavior in the presence of -t.
kientzle [Fri, 27 Aug 2004 04:19:28 +0000 (04:19 +0000)]
Document the -O option's new behavior in the presence of -t.

MFC after: 3 days

20 years agoSupport -t -O as in gtar. Perversely enough, -O means "send to stderr"
kientzle [Fri, 27 Aug 2004 04:13:15 +0000 (04:13 +0000)]
Support -t -O as in gtar.  Perversely enough, -O means "send to stderr"
when used with -t, "send to stdout" when used with -x.

Thanks to: Ryan Hamilton for pointing out this odd beast
MFC after: 3 days

20 years agoCorrect va_end usage.
kientzle [Fri, 27 Aug 2004 03:52:29 +0000 (03:52 +0000)]
Correct va_end usage.

Thanks to: Sean McNeil
MFC after: 3 days

20 years agoCache uid-from-uname and gid-from-gname lookups during extraction.
kientzle [Fri, 27 Aug 2004 03:40:48 +0000 (03:40 +0000)]
Cache uid-from-uname and gid-from-gname lookups during extraction.
This should provide a big performance boost for folks using NIS or LDAP.

MFC after: 3 days
Thanks to: Jun Kuriyama (for reminding me that this was still on my TODO list)

20 years agoReintroduce slightly modified patch from kern/69964. Check for
kan [Fri, 27 Aug 2004 01:41:28 +0000 (01:41 +0000)]
Reintroduce slightly modified patch from kern/69964. Check for
LK_HAVE_EXL in both acquire invocations.

MFC after: 5 days

20 years agoWhen trying each linker class in turn with a preloaded module, exit
iedowse [Fri, 27 Aug 2004 01:20:26 +0000 (01:20 +0000)]
When trying each linker class in turn with a preloaded module, exit
the loop if the preload was successful. Previously a successful
preload was ignored if the linker class was not the last in the
list.

20 years agoAdd a new module information type MODINFOMD_SHDR that will be used
iedowse [Fri, 27 Aug 2004 01:10:16 +0000 (01:10 +0000)]
Add a new module information type MODINFOMD_SHDR that will be used
by the loader to pass the section header table from preloaded ELF
relocatable modules into the kernel.

20 years agoAdd MD relocation support for amd64 and i386 platforms. The no-op
iedowse [Fri, 27 Aug 2004 01:06:57 +0000 (01:06 +0000)]
Add MD relocation support for amd64 and i386 platforms. The no-op
relocation is not sufficient for ELF relocatable object format
modules, since accessing the module metadata involves following
pointers between different ELF sections.

This allows kldxref to correctly build linker.hints on the amd64
platform.

20 years agoAdd support for reading ELF relocatable object file format modules.
iedowse [Fri, 27 Aug 2004 00:58:40 +0000 (00:58 +0000)]
Add support for reading ELF relocatable object file format modules.

20 years agoAlways set the status and move a printf under bootverbose.
njl [Fri, 27 Aug 2004 00:53:11 +0000 (00:53 +0000)]
Always set the status and move a printf under bootverbose.

Tested by: gj

20 years agoCall the file format specific code through a table of function
iedowse [Fri, 27 Aug 2004 00:51:21 +0000 (00:51 +0000)]
Call the file format specific code through a table of function
pointers and remove knowledge of the file format from kldxref.c.
This will make it possible to support more than one file format.

20 years agoIn conjunction with having boot0cfg(8) default boot0 to packet mode.
obrien [Fri, 27 Aug 2004 00:18:03 +0000 (00:18 +0000)]
In conjunction with having boot0cfg(8) default boot0 to packet mode.

Submitted by: Tor.Egge@cvsup.no.freebsd.org

20 years agoWhen attaching a consumer from a volume to a plex, check if the
le [Thu, 26 Aug 2004 21:04:41 +0000 (21:04 +0000)]
When attaching a consumer from a volume to a plex, check if the
volume already has a plex attached and adjust the access counts
of the new consumer accordingly.

20 years agoCorrect a denial-of-service vulnerability in zlib.
nectar [Thu, 26 Aug 2004 19:37:06 +0000 (19:37 +0000)]
Correct a denial-of-service vulnerability in zlib.
For reference, this bug was first reported at
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=252253

Submitted by: "Dmitry V. Levin" <ldv@altlinux.org>

20 years agoFixed the shell arithmetic to work with 4.x /bin/sh.
ru [Thu, 26 Aug 2004 19:15:20 +0000 (19:15 +0000)]
Fixed the shell arithmetic to work with 4.x /bin/sh.

Tested by: make release of HEAD on 4.10-STABLE

20 years agoIf getting the current setting after modifying the link failed, we assume
njl [Thu, 26 Aug 2004 17:14:36 +0000 (17:14 +0000)]
If getting the current setting after modifying the link failed, we assume
it succeeded.  However, we also need to set the status to AE_OK.

20 years agoCorrect capitalization of DeLorme Publishing.
des [Thu, 26 Aug 2004 16:59:50 +0000 (16:59 +0000)]
Correct capitalization of DeLorme Publishing.

20 years agoThe check for r_flag was accidentally removed in the previous commit.
des [Thu, 26 Aug 2004 15:51:10 +0000 (15:51 +0000)]
The check for r_flag was accidentally removed in the previous commit.

Submitted by: SANETO Takanori <sanewo@ba2.so-net.ne.jp>
MFC after: 3 days

20 years agoDocument how a standalone module can be built with divert(4) support.
ru [Thu, 26 Aug 2004 14:35:05 +0000 (14:35 +0000)]
Document how a standalone module can be built with divert(4) support.

20 years agoRevert the last change to sys/modules/ipfw/Makefile and fix a
ru [Thu, 26 Aug 2004 14:18:30 +0000 (14:18 +0000)]
Revert the last change to sys/modules/ipfw/Makefile and fix a
standalone module build in a better way.

Silence from: andre
MFC after: 3 days

20 years agoImprove the robustness of MSDOSFSMNT_KICONV handling:
tjr [Thu, 26 Aug 2004 13:16:44 +0000 (13:16 +0000)]
Improve the robustness of MSDOSFSMNT_KICONV handling:
- Use copyinstr() to read cs_win, cs_dos, cs_local strings from the
  mount argument structure instead of reading through user-space pointers(!).
- When mounting a filesystem, or updating an existing mount, only try to
  update the iconv handles from the information in the mount argument
  structure if the structure itself has the MSDOSFSMNT_KICONV flag set.
- Attempt to handle failure of update_mp() in the MNT_UPDATE case.

20 years agoSkip providers with not defined sector size.
pjd [Thu, 26 Aug 2004 12:42:47 +0000 (12:42 +0000)]
Skip providers with not defined sector size.

Reported by: kuriyama

20 years agoBackout the CVSTAG variable, it could potentially be dangerous if
paul [Thu, 26 Aug 2004 10:24:25 +0000 (10:24 +0000)]
Backout the CVSTAG variable, it could potentially be dangerous if
you track multiple releases in different trees.

Leave the CVSOPTIONS variable there since it could be useful.

20 years agos/_/./g for an entity name.
hrs [Thu, 26 Aug 2004 08:52:46 +0000 (08:52 +0000)]
s/_/./g for an entity name.

Reviewed by:    simon

20 years agoAutogenerate device listings for cue(4) and kue(4)
brueffer [Thu, 26 Aug 2004 08:37:40 +0000 (08:37 +0000)]
Autogenerate device listings for cue(4) and kue(4)

20 years agoAdd cue(4) and kue(4)
brueffer [Thu, 26 Aug 2004 08:36:09 +0000 (08:36 +0000)]
Add cue(4) and kue(4)

20 years ago- move the list of supported adapters into a HARDWARE section
brueffer [Thu, 26 Aug 2004 08:34:09 +0000 (08:34 +0000)]
- move the list of supported adapters into a HARDWARE section
- properly capitalize Ethernet
- replace a misleading 'adapter' with 'chipset'
- bump document date

MFC after: 3 days

20 years ago- move the list of supported adapters into a HARDWARE section
brueffer [Thu, 26 Aug 2004 08:33:03 +0000 (08:33 +0000)]
- move the list of supported adapters into a HARDWARE section
- sort the list of adapters
- add the 3Com 3c460 HomeConnect Ethernet USB Adapter [1]
- properly capitalize Ethernet
- replace a misleading 'adapter' with 'chipset'
- bump document date

Obtained from: NetBSD [1]
MFC after: 3 days

20 years agoRe-generate from sl_SI.ISO8859-2.src rev. 1.8 (lowercase month/weekday names).
tjr [Thu, 26 Aug 2004 08:18:56 +0000 (08:18 +0000)]
Re-generate from sl_SI.ISO8859-2.src rev. 1.8 (lowercase month/weekday names).

20 years agoBegin month and weekday names with a lowercase letter. This is consistent
tjr [Thu, 26 Aug 2004 08:15:48 +0000 (08:15 +0000)]
Begin month and weekday names with a lowercase letter. This is consistent
with (at least) Solaris, ICU and glibc.

PR: 65317
Submitted by: Uros
MFC after: 1 week

20 years agoPermit -P to be combined with -t. (It's a no-op then, just as in gtar.)
kientzle [Thu, 26 Aug 2004 06:28:39 +0000 (06:28 +0000)]
Permit -P to be combined with -t.  (It's a no-op then, just as in gtar.)

20 years agoBackout recent -j changes, the flags is deprecated.
maxim [Thu, 26 Aug 2004 06:28:05 +0000 (06:28 +0000)]
Backout recent -j changes, the flags is deprecated.

Requested by: tjr

20 years agoDon't forget to va_end() the va_list we get from va_copy().
des [Thu, 26 Aug 2004 06:25:28 +0000 (06:25 +0000)]
Don't forget to va_end() the va_list we get from va_copy().

Submitted by: Sean McNeil <sean@mcneil.com>
MFC after: 3 days

20 years agoRemove unnecessary check for curthread == NULL.
alc [Thu, 26 Aug 2004 04:34:39 +0000 (04:34 +0000)]
Remove unnecessary check for curthread == NULL.

20 years agoAdd reference to vslock(9) for locking user address pages into core.
imp [Thu, 26 Aug 2004 04:16:55 +0000 (04:16 +0000)]
Add reference to vslock(9) for locking user address pages into core.

Suggested by: alc

20 years agoAdd pmap locking to many of the functions.
alc [Thu, 26 Aug 2004 04:15:36 +0000 (04:15 +0000)]
Add pmap locking to many of the functions.

Many thanks to Andrew Gallatin for resolving a powerpc-specific
initialization problem in my original patch.

Tested by: gallatin@

20 years agoDon't edit permissions of pre-existing directories during extract.
kientzle [Thu, 26 Aug 2004 03:53:43 +0000 (03:53 +0000)]
Don't edit permissions of pre-existing directories during extract.

This closes a security hole.  Otherwise, libarchive will happily
extract into directories to which it lacks write permissions by
resetting the permissions during the extract.

Thanks to: Kris Kennaway

20 years agoIt is not legal to re-use a va_list variable. This caused
kientzle [Thu, 26 Aug 2004 03:33:53 +0000 (03:33 +0000)]
It is not legal to re-use a va_list variable.  This caused
a crash on amd64, in particular.

Thanks to: Sean McNeil

20 years agoAdd missing brackets. It was committed from wrong tree.
davidxu [Thu, 26 Aug 2004 02:41:01 +0000 (02:41 +0000)]
Add missing brackets. It was committed from wrong tree.

20 years agoTell getopt to accept the (already-implemented) -I option.
kientzle [Thu, 26 Aug 2004 01:58:14 +0000 (01:58 +0000)]
Tell getopt to accept the (already-implemented) -I option.

20 years agoBump document date for last commit.
simon [Thu, 26 Aug 2004 00:26:52 +0000 (00:26 +0000)]
Bump document date for last commit.

20 years ago- Add a HARDWARE section which lists supported devices.
simon [Thu, 26 Aug 2004 00:19:54 +0000 (00:19 +0000)]
- Add a HARDWARE section which lists supported devices.
- Mark up each item in the device list with .It, as per mdoc(7) and to
  make it simpler for auto generated Hardware Notes to parse the manual
  page.

MFC after: 3 days

20 years ago- use .Nm
brueffer [Wed, 25 Aug 2004 23:43:55 +0000 (23:43 +0000)]
- use .Nm
- fix some grammar and spelling mistakes

MFC after: 3 days

20 years agogcc -O2 cleanup. tested for a long time.
davidxu [Wed, 25 Aug 2004 23:42:40 +0000 (23:42 +0000)]
gcc -O2 cleanup. tested for a long time.

Reviewed by: deischen

20 years agoSupport column lists, by only extracting the first column.
simon [Wed, 25 Aug 2004 23:32:26 +0000 (23:32 +0000)]
Support column lists, by only extracting the first column.

MFC after: 3 days

20 years agoAutogenerate device listings for axe(4) and rue(4)
brueffer [Wed, 25 Aug 2004 22:57:07 +0000 (22:57 +0000)]
Autogenerate device listings for axe(4) and rue(4)

20 years agoAdd axe(4) and rue(4)
brueffer [Wed, 25 Aug 2004 22:55:02 +0000 (22:55 +0000)]
Add axe(4) and rue(4)

20 years ago- move list of supported adapters into a HARDWARE section
brueffer [Wed, 25 Aug 2004 22:46:36 +0000 (22:46 +0000)]
- move list of supported adapters into a HARDWARE section
- fix capitalization of a "to"
- bump document date

MFC after: 3 days

20 years ago- put the list of supported adapters into a HARDWARE section
brueffer [Wed, 25 Aug 2004 22:39:05 +0000 (22:39 +0000)]
- put the list of supported adapters into a HARDWARE section
- make the Buffalo entry look better
- properly capitalize Ethernet and Fast Ethernet
- bump document date

MFC after: 3 days

20 years agoFix grammar nit.
simon [Wed, 25 Aug 2004 22:36:47 +0000 (22:36 +0000)]
Fix grammar nit.

Submitted by: brueffer
MFC after: 2 days

20 years agoDon't call uart_bus_probe() for non-matching PnP-devices. Trying to probe
marius [Wed, 25 Aug 2004 22:15:33 +0000 (22:15 +0000)]
Don't call uart_bus_probe() for non-matching PnP-devices. Trying to probe
the keyboard controller with uart_bus_probe() caused a hang here on an i386
machine.

Approved by: marcel

20 years agoFix "make world DESTDIR=/mnt" to work again. A recent change
ru [Wed, 25 Aug 2004 22:06:29 +0000 (22:06 +0000)]
Fix "make world DESTDIR=/mnt" to work again.  A recent change
to make(1) that causes command-line variables to be passed as
command-line variables to sub-processes that make(1) executes
broke it.  By changing the type of all DESTDIR variables used
internally in Makefile.inc1, from environment to command-line
variables of the highest priority, I was able to "make world"
with success, with the command-line variable DESTDIR set.

20 years agoDon't hold the UNIX domain socket subsystem lock over the body of the
rwatson [Wed, 25 Aug 2004 21:24:36 +0000 (21:24 +0000)]
Don't hold the UNIX domain socket subsystem lock over the body of the
UNIX domain socket garbage collection implementation, as that risks
holding the mutex over potentially sleeping operations (as well as
introducing some nasty lock order issues, etc).  unp_gc() will hold
the lock long enough to do necessary deferal checks and set that it's
running, but then release it until it needs to reset the gc state.

RELENG_5 candidate.

Discussed with: alfred

20 years agoRemove code to initialize the lid state at boot. It interfered with lid
njl [Wed, 25 Aug 2004 20:09:15 +0000 (20:09 +0000)]
Remove code to initialize the lid state at boot.  It interfered with lid
operation for some users with pure GPE lid switches (vs. embedded
controller.)

Tested by: Anish Mistry <mistry.7_at_osu.edu>
MFC after: 3 days

20 years agoRecommend using make kernel for current -> current upgrades. No reason
imp [Wed, 25 Aug 2004 19:39:13 +0000 (19:39 +0000)]
Recommend using make kernel for current -> current upgrades.  No reason
to have multiple commands for that.
Use relative paths into the src tree consistantly in the instructions.
Fix minor nits that have crept into things.

# is preseedrandom still necessary?

20 years agoLog verification errors at level 1.
pjd [Wed, 25 Aug 2004 19:18:07 +0000 (19:18 +0000)]
Log verification errors at level 1.