]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
15 years agoRemove unused variable.
fjoe [Mon, 9 Feb 2009 16:57:07 +0000 (16:57 +0000)]
Remove unused variable.

Found with: Coverity Prevent(tm)
CID: 3693

15 years agoDestroy TX tag outside of loop scope.
fjoe [Mon, 9 Feb 2009 16:53:37 +0000 (16:53 +0000)]
Destroy TX tag outside of loop scope.

Found with: Coverity Prevent(tm)
CID: 3886

15 years agoProperly implement GIO_ATTR and CONS_GETINFO.
ed [Mon, 9 Feb 2009 15:55:21 +0000 (15:55 +0000)]
Properly implement GIO_ATTR and CONS_GETINFO.

It seems I didn't implement these two ioctl()'s properly, which meant
vidcontrol couldn't properly obtain certain terminal parameters.

15 years agos/SS_FDREF/SS_NOFDREF/
mbr [Mon, 9 Feb 2009 13:29:01 +0000 (13:29 +0000)]
s/SS_FDREF/SS_NOFDREF/

15 years agoFix minor spacing problem found by s9indent from last
rrs [Mon, 9 Feb 2009 11:42:23 +0000 (11:42 +0000)]
Fix minor spacing problem found by s9indent from last
commit.

15 years agoFix INET only build breakage with SCTP - pointy hat to me :-)
rrs [Mon, 9 Feb 2009 11:41:54 +0000 (11:41 +0000)]
Fix INET only build breakage with SCTP - pointy hat to me :-)

15 years agoComment out the assertion from r188321. It is not valid for nfs.
kib [Mon, 9 Feb 2009 11:32:23 +0000 (11:32 +0000)]
Comment out the assertion from r188321. It is not valid for nfs.

Reported by: alc

15 years agoRemove a stale comment from the clists code.
ed [Mon, 9 Feb 2009 11:27:56 +0000 (11:27 +0000)]
Remove a stale comment from the clists code.

We don't support quote bits.

15 years agoAvoid some cases of unnecessary page queues locking by vm_fault's delete-
alc [Mon, 9 Feb 2009 06:23:21 +0000 (06:23 +0000)]
Avoid some cases of unnecessary page queues locking by vm_fault's delete-
behind heuristic.

15 years agoDocument loader tunable hw.re.msi_disable.
yongari [Mon, 9 Feb 2009 05:09:52 +0000 (05:09 +0000)]
Document loader tunable hw.re.msi_disable.
Bump .Dd

15 years agoReclaim transmitted frames in re_tick(). This is for PCIe
yongari [Mon, 9 Feb 2009 04:59:13 +0000 (04:59 +0000)]
Reclaim transmitted frames in re_tick(). This is for PCIe
controllers that lose Tx completion interrupts under certain
conditions. With this change it's safe to use MSI on PCIe
controllers so enable MSI on these controllers.

15 years agomark M_LASTFRAG at the last fragment.
weongyo [Mon, 9 Feb 2009 04:39:16 +0000 (04:39 +0000)]
mark M_LASTFRAG at the last fragment.

Reviewed by: sam
MFC after: 3 weeks

15 years agoAdd the EBR scheme. The EBR scheme supports the Extended Boot Records
marcel [Sun, 8 Feb 2009 23:51:44 +0000 (23:51 +0000)]
Add the EBR scheme. The EBR scheme supports the Extended Boot Records
found inside extended partitions and used to create logical partitions.
At this time write/modify support is not (yet) present.
The EBR and MBR schemes both check the parent scheme. The MBR will
back-off when nested under another MBR, whereas the EBR only nests
under a MBR.

15 years ago- Fix in_cksum for big-endian MIPS: use correct compile-time check.
gonzo [Sun, 8 Feb 2009 23:43:36 +0000 (23:43 +0000)]
- Fix in_cksum for big-endian MIPS: use correct compile-time check.

15 years agoAllow gpe_offset to be set by the scheme. When gpe_offset is zero,
marcel [Sun, 8 Feb 2009 23:39:30 +0000 (23:39 +0000)]
Allow gpe_offset to be set by the scheme. When gpe_offset is zero,
or invalid, initialize it to the start of the partition. Adjust
the mediasize when the offset lies somewhere inside the partition.

15 years agoWhen bouncing pages, allow a new option to preserve the intra-page
imp [Sun, 8 Feb 2009 22:54:58 +0000 (22:54 +0000)]
When bouncing pages, allow a new option to preserve the intra-page
offset.  This is needed for the ehci hardware buffer rings that assume
this behavior.

This is an interim solution, and a more general one is being worked
on.  This solution doesn't break anything that doesn't ask for it
directly.  The mbuf and uio variants with this flag likely don't work
and haven't been tested.

Universe builds with these changes.  I don't have a huge-memory
machine to test these changes with, but will be happy to work with
folks that do and hps if this changes turns out not to be sufficient.

Submitted by: alfred@ from Hans Peter Selasky's original

15 years agoEliminate OBJ_NEEDGIANT. After r188331, OBJ_NEEDGIANT's only use is by a
alc [Sun, 8 Feb 2009 22:17:24 +0000 (22:17 +0000)]
Eliminate OBJ_NEEDGIANT.  After r188331, OBJ_NEEDGIANT's only use is by a
redundant assertion in vm_fault().

Reviewed by: kib

15 years agoRemove an overzealous check.
trasz [Sun, 8 Feb 2009 22:08:48 +0000 (22:08 +0000)]
Remove an overzealous check.

Submitted by: das
Reviewed by: scottl
Approved by: rwatson (mentor, implicit)
Sponsored by: FreeBSD Foundation

15 years agoFix multiple -s options.
kientzle [Sun, 8 Feb 2009 22:04:17 +0000 (22:04 +0000)]
Fix multiple -s options.

MFC after: 7 days

15 years agoWhen copying file data to the archive, don't write more
kientzle [Sun, 8 Feb 2009 22:02:46 +0000 (22:02 +0000)]
When copying file data to the archive, don't write more
than was read.  This seems to have only affected the shar
writer, since other formats proactively truncate output
to the originally-advertised size.

PR: bin/131244
MFC after: 7 days

15 years agoDon't try to directly update page tables
kmacy [Sun, 8 Feb 2009 21:54:51 +0000 (21:54 +0000)]
Don't try to directly update page tables

15 years agoRemove no longer valid comment.
kib [Sun, 8 Feb 2009 21:20:13 +0000 (21:20 +0000)]
Remove no longer valid comment.

Submitted by: alc

15 years agoThe NRL value in the queue head must be zero for interrupt transfers. This was
thompsa [Sun, 8 Feb 2009 21:08:00 +0000 (21:08 +0000)]
The NRL value in the queue head must be zero for interrupt transfers. This was
diagnosed using a USB analyser obtained through a FreeBSD Foundation grant.

Thanks to: FreeBSD Foundation

15 years agoImprove comments, correct English.
kib [Sun, 8 Feb 2009 20:52:09 +0000 (20:52 +0000)]
Improve comments, correct English.

Submitted by: alc

15 years agoDo not call vm_object_deallocate() from vm_map_delete(), because we
kib [Sun, 8 Feb 2009 20:39:17 +0000 (20:39 +0000)]
Do not call vm_object_deallocate() from vm_map_delete(), because we
hold the map lock there, and might need the vnode lock for OBJT_VNODE
objects. Postpone object deallocation until caller of vm_map_delete()
drops the map lock. Link the map entries to be freed into the freelist,
that is released by the new helper function vm_map_entry_free_freelist().

Reviewed by: tegge, alc
Tested by: pho

15 years agoIn vm_map_sync(), do not call vm_object_sync() while holding map lock.
kib [Sun, 8 Feb 2009 20:30:51 +0000 (20:30 +0000)]
In vm_map_sync(), do not call vm_object_sync() while holding map lock.
Reference object, drop the map lock, and then call vm_object_sync().
The object sync might require vnode lock for OBJT_VNODE type objects.

Reviewed by: tegge
Tested by: pho

15 years agofix typo
sam [Sun, 8 Feb 2009 20:29:37 +0000 (20:29 +0000)]
fix typo

Submitted by: Christoph Mallon <christoph.mallon@gmx.de>

15 years agoDo not sleep for vnode lock while holding map lock in vm_fault. Try to
kib [Sun, 8 Feb 2009 20:23:46 +0000 (20:23 +0000)]
Do not sleep for vnode lock while holding map lock in vm_fault. Try to
acquire vnode lock for OBJT_VNODE object after map lock is dropped.
Because we have the busy page(s) in the object, sleeping there would
result in deadlock with vnode resize. Try to get lock without sleeping,
and, if the attempt failed, drop the state, lock the vnode, and restart
the fault handler from the start with already locked vnode.

Because the vnode_pager_lock() function is inlined in vm_fault(),
axe it.

Based on suggestion by: alc
Reviewed by: tegge, alc
Tested by: pho

15 years agoPrefer the start and end attributes over the offset and size
marcel [Sun, 8 Feb 2009 20:19:19 +0000 (20:19 +0000)]
Prefer the start and end attributes over the offset and size
attributes. The start and end more accurately describe the
space taken by a partition. The offset and size are used to
describe the effective (usable) storage of that partition.

15 years agoo Add the "PART::scheme" attribute that returns the name of the
marcel [Sun, 8 Feb 2009 20:15:08 +0000 (20:15 +0000)]
o  Add the "PART::scheme" attribute that returns the name of the
   underlying partitioning scheme.
o  Put the start and end of the partition in the XML configuration.
   The start and end are the LBAs of the first and last sector
   (resp.) of the partition. They are currently identical to the
   offset and size attributes, which describe the partition as an
   offset and size in bytes, but may not in the future. The start
   and end will be used for the logical partition boundaries and
   may include metadata. The offset and size will always represent
   the useful storage space within the partition. Typically these
   two notions are the same, but for logical partitions in an
   extended partition, the EBR is more naturally treated as being
   part of the partition.

15 years agoo use humanize_number to print totals
sam [Sun, 8 Feb 2009 20:02:58 +0000 (20:02 +0000)]
o use humanize_number to print totals
o shrink some fields now that we don't have to worry about overflow

15 years agoAdd the comments to vm_map_simplify_entry() and vmspace_fork(),
kib [Sun, 8 Feb 2009 20:00:33 +0000 (20:00 +0000)]
Add the comments to vm_map_simplify_entry() and vmspace_fork(),
describing why several calls to vm_deallocate_object() with locked map
do not result in the acquisition of the vnode lock after map lock.

Suggested and reviewed by: tegge

15 years agoLock the new map in vmspace_fork(). The newly allocated map should not
kib [Sun, 8 Feb 2009 19:55:03 +0000 (19:55 +0000)]
Lock the new map in vmspace_fork(). The newly allocated map should not
be accessible outside vmspace_fork() yet, but locking it would satisfy
the protocol of the vm_map_entry_link() and other functions called
from vmspace_fork().

Use trylock that is supposedly cannot fail, to silence WITNESS warning
of the nested acquisition of the sx lock with the same name.

Suggested and reviewed by: tegge

15 years agoAssert that vnode is exclusively locked when its vm object is resized.
kib [Sun, 8 Feb 2009 19:44:50 +0000 (19:44 +0000)]
Assert that vnode is exclusively locked when its vm object is resized.

Reviewed by: tegge

15 years agoDo not leak the MAP_ENTRY_IN_TRANSITION flag when copying map entry
kib [Sun, 8 Feb 2009 19:41:08 +0000 (19:41 +0000)]
Do not leak the MAP_ENTRY_IN_TRANSITION flag when copying map entry
on fork. Otherwise, copied entry cannot be removed in the child map.

Reviewed by: tegge
MFC after: 2 weeks

15 years agoStyle.
kib [Sun, 8 Feb 2009 19:37:01 +0000 (19:37 +0000)]
Style.

15 years agoLookup up the directory entry for the tmpfs node that are deleted by
kib [Sun, 8 Feb 2009 19:18:33 +0000 (19:18 +0000)]
Lookup up the directory entry for the tmpfs node that are deleted by
both node pointer and name component. This does the right thing for
hardlinks to the same node in the same directory.

Submitted by: Yoshihiro Ota <ota j email ne jp>
PR: kern/131356
MFC after: 2 weeks

15 years agoShutup warning for DNAME RR.
ume [Sun, 8 Feb 2009 16:58:05 +0000 (16:58 +0000)]
Shutup warning for DNAME RR.

PR: bin/127591
Submitted by: "Eugene M. Kim" <20080111.freebsd.org__at__ab.ote.we.lv>
MFC after: 1 week

15 years agoSet the lower bound on queue size for an audit pipe to 1 instead of 0,
rwatson [Sun, 8 Feb 2009 15:38:31 +0000 (15:38 +0000)]
Set the lower bound on queue size for an audit pipe to 1 instead of 0,
as an audit pipe with a queue length of 0 is less useful.

Obtained from: TrustedBSD Project
Sponsored by: Apple, Inc.
MFC after: 1 week

15 years agoFix several filenames for "make clean" target.
wkoszek [Sun, 8 Feb 2009 14:43:20 +0000 (14:43 +0000)]
Fix several filenames for "make clean" target.

15 years agoChange various routines that are responsible for transforming audit
rwatson [Sun, 8 Feb 2009 14:39:35 +0000 (14:39 +0000)]
Change various routines that are responsible for transforming audit
event IDs based on arguments to return au_event_t rather than int.

Obtained from: TrustedBSD Project
Sponsored by: Apple, Inc.
MFC after: 1 week

15 years agoAudit AUE_MAC_EXECVE; currently just the standard AUE_EXECVE arguments
rwatson [Sun, 8 Feb 2009 14:24:35 +0000 (14:24 +0000)]
Audit AUE_MAC_EXECVE; currently just the standard AUE_EXECVE arguments
and not the label.

Obtained from: TrustedBSD Project
Sponsored by: Apple, Inc.
MFC after: 1 week

15 years agoAudit the flag argument to the nfssvc(2) system call.
rwatson [Sun, 8 Feb 2009 14:04:08 +0000 (14:04 +0000)]
Audit the flag argument to the nfssvc(2) system call.

Obtained from: TrustedBSD Project
Sponsored by: Apple, Inc.

15 years agosi(4) seems to build without a problem. However, since noone noticed lack
wkoszek [Sun, 8 Feb 2009 12:40:33 +0000 (12:40 +0000)]
si(4) seems to build without a problem. However, since noone noticed lack
of this driver, put it in a comment.

15 years agoFurther NOTES cleanup -- following drivers didn't survive TTY-ng
wkoszek [Sun, 8 Feb 2009 12:33:05 +0000 (12:33 +0000)]
Further NOTES cleanup -- following drivers didn't survive TTY-ng
and aren't included in NOTES anyway: cy(4), rc(4), rp(4).

si(4) doesn't belong to global NOTES.

15 years agoAdd missing pcfclock description.
wkoszek [Sun, 8 Feb 2009 12:12:19 +0000 (12:12 +0000)]
Add missing pcfclock description.

15 years agoBring missing comments on EPSON_BOUNCEDMA and EPSON_MEMWIN flags.
wkoszek [Sun, 8 Feb 2009 10:21:25 +0000 (10:21 +0000)]
Bring missing comments on EPSON_BOUNCEDMA and EPSON_MEMWIN flags.

Submitted by: nyan

15 years agoTry to remove/assimilate as much of formerly IPv4/6 specific
bz [Sun, 8 Feb 2009 09:27:07 +0000 (09:27 +0000)]
Try to remove/assimilate as much of formerly IPv4/6 specific
(duplicate) code in sys/netipsec/ipsec.c and fold it into
common, INET/6 independent functions.

The file local functions ipsec4_setspidx_inpcb() and
ipsec6_setspidx_inpcb() were 1:1 identical after the change
in r186528. Rename to ipsec_setspidx_inpcb() and remove the
duplicate.

Public functions ipsec[46]_get_policy() were 1:1 identical.
Remove one copy and merge in the factored out code from
ipsec_get_policy() into the other. The public function left
is now called ipsec_get_policy() and callers were adapted.

Public functions ipsec[46]_set_policy() were 1:1 identical.
Rename file local ipsec_set_policy() function to
ipsec_set_policy_internal().
Remove one copy of the public functions, rename the other
to ipsec_set_policy() and adapt callers.

Public functions ipsec[46]_hdrsiz() were logically identical
(ignoring one questionable assert in the v6 version).
Rename the file local ipsec_hdrsiz() to ipsec_hdrsiz_internal(),
the public function to ipsec_hdrsiz(), remove the duplicate
copy and adapt the callers.
The v6 version had been unused anyway. Cleanup comments.

Public functions ipsec[46]_in_reject() were logically identical
apart from statistics. Move the common code into a file local
ipsec46_in_reject() leaving vimage+statistics in small AF specific
wrapper functions. Note: unfortunately we already have a public
ipsec_in_reject().

Reviewed by: sam
Discussed with: rwatson (renaming to *_internal)
MFC after: 26 days
X-MFC: keep wrapper functions for public symbols?

15 years agoFix language on atol(3) manpage. Add a COMPATIBILITY section
keramida [Sun, 8 Feb 2009 08:26:58 +0000 (08:26 +0000)]
Fix language on atol(3) manpage.  Add a COMPATIBILITY section
stating that in FreeBSD the atol() and atoll() functions affect
errno in the same way as strtol() and stroll().

PR: docs/126487
Submitted by: edwin
Reviewed by: trhodes, gabor
MFC after: 1 week

15 years agoRetire NO_DMA completely.
imp [Sun, 8 Feb 2009 08:13:36 +0000 (08:13 +0000)]
Retire NO_DMA completely.

15 years agoFix g_part_*dumpconf to return void to match kobj definition.
imp [Sun, 8 Feb 2009 07:05:23 +0000 (07:05 +0000)]
Fix g_part_*dumpconf to return void to match kobj definition.
Fix g_part_*name to return a const char * rather than a char *.

15 years agoCompanion for r188301: fix the prototypes.
imp [Sun, 8 Feb 2009 07:03:34 +0000 (07:03 +0000)]
Companion for r188301: fix the prototypes.

15 years agoCorrect parameter types for pcib_{read,write}_config by fixing the
imp [Sun, 8 Feb 2009 07:02:42 +0000 (07:02 +0000)]
Correct parameter types for pcib_{read,write}_config by fixing the
protptyoes for the legacy_* impelemtnations of these kobj methods.

15 years agoFix leak of kqueue() file descriptors when linked with static
deischen [Sun, 8 Feb 2009 06:44:58 +0000 (06:44 +0000)]
Fix leak of kqueue() file descriptors when linked with static
libc_r.

PR: 58687
Submitted by: Jonathon Lennox <lennox at cs dot columbia dot edu>

15 years agoSilent LINT: add 2 stubs (update_crc32 and sctp_finalize_crc32) to fix LIBALIAS ...
piso [Sun, 8 Feb 2009 03:03:55 +0000 (03:03 +0000)]
Silent LINT: add 2 stubs (update_crc32 and sctp_finalize_crc32) to fix LIBALIAS + SCTP_NO_CSUM case.

15 years agoFix LIBALIAS option for a static kernel.
piso [Sun, 8 Feb 2009 03:02:06 +0000 (03:02 +0000)]
Fix LIBALIAS option for a static kernel.

15 years agoResort NOTES a bit to easily distinguish, which comments are actual and
wkoszek [Sun, 8 Feb 2009 00:16:24 +0000 (00:16 +0000)]
Resort NOTES a bit to easily distinguish, which comments are actual and
refer to used options, and which comments are obseleted.

Reviewed by: imp

15 years agoMake sure the comparison is done with an unsigned char.
imp [Sat, 7 Feb 2009 19:34:44 +0000 (19:34 +0000)]
Make sure the comparison is done with an unsigned char.

15 years agoAdd SCTP NAT support.
piso [Sat, 7 Feb 2009 18:49:42 +0000 (18:49 +0000)]
Add SCTP NAT support.

Submitted by: CAIA (http://caia.swin.edu.au)

15 years agoNamed normally cannot be started chrooted inside a jail. Thus treat
bz [Sat, 7 Feb 2009 16:37:02 +0000 (16:37 +0000)]
Named normally cannot be started chrooted inside a jail. Thus treat
the jail case specifically. In case we find a proper pre-seeded
devfs in the chroot path (mounted from the base system) permit
starting chrooted else give proper warn/error messages.

PR: conf/103489
Reviewed by: dougb
MFC after: 5 days

15 years agoFix build, sigh.
thompsa [Sat, 7 Feb 2009 15:51:32 +0000 (15:51 +0000)]
Fix build, sigh.

15 years agoMove the comment to it's correct place.
wkoszek [Sat, 7 Feb 2009 11:40:47 +0000 (11:40 +0000)]
Move the comment to it's correct place.

15 years agoRemove leftover of alpha support for config(8)--we have MAP_FAILED
wkoszek [Sat, 7 Feb 2009 11:12:30 +0000 (11:12 +0000)]
Remove leftover of alpha support for config(8)--we have MAP_FAILED
globally defined.

15 years agoDon't forget to create opt_agp.h on ia64, which also uses agp(4).
wkoszek [Sat, 7 Feb 2009 09:57:14 +0000 (09:57 +0000)]
Don't forget to create opt_agp.h on ia64, which also uses agp(4).

15 years agoDont hold the lock over the controller init, we are still attaching.
thompsa [Sat, 7 Feb 2009 06:27:16 +0000 (06:27 +0000)]
Dont hold the lock over the controller init, we are still attaching.

15 years agoC99 TC2 now wants FP_FAST_FMA* to be defined to 1, if the macros are
das [Sat, 7 Feb 2009 05:41:24 +0000 (05:41 +0000)]
C99 TC2 now wants FP_FAST_FMA* to be defined to 1, if the macros are
defined at all. See also: defect report #223.

15 years agocount stuck beacon events
sam [Sat, 7 Feb 2009 05:34:41 +0000 (05:34 +0000)]
count stuck beacon events

15 years agofix building w/o CFI_ARMEDANDDANGEROUS
sam [Sat, 7 Feb 2009 05:32:19 +0000 (05:32 +0000)]
fix building w/o CFI_ARMEDANDDANGEROUS

15 years agoexpand CFI_ARMEDANDDANGEROUS to include writing the user segment
sam [Sat, 7 Feb 2009 05:03:25 +0000 (05:03 +0000)]
expand CFI_ARMEDANDDANGEROUS to include writing the user segment
of the PR; this register is actually write-once so deserves the
safety-belt as much as the PLR

15 years agosi_cmdname() is defined only when SI_DEBUG is enabled, thus
wkoszek [Sat, 7 Feb 2009 02:14:27 +0000 (02:14 +0000)]
si_cmdname() is defined only when SI_DEBUG is enabled, thus
using this function in simple printf() causes compile-time
problems. Use conditionally enabled DPRINT() macro instead.

15 years agofix 11n channel construction
sam [Sat, 7 Feb 2009 01:54:57 +0000 (01:54 +0000)]
fix 11n channel construction

15 years agoadd macro for future regulatory mods
sam [Sat, 7 Feb 2009 01:53:44 +0000 (01:53 +0000)]
add macro for future regulatory mods

15 years agoRearrange this code slightly to pass softcs around instead of device_t,
nwhitehorn [Sat, 7 Feb 2009 01:15:13 +0000 (01:15 +0000)]
Rearrange this code slightly to pass softcs around instead of device_t,
solving a possible panic when snd_ai2s is loaded at boot time. Also change
the device setup to indicate to the pcm layer that the device is MPSAFE.

Submitted by: Marco Trillo
Suggestions by: Ariff Abdullah

15 years agoRegulatory fixups:
sam [Sat, 7 Feb 2009 01:12:51 +0000 (01:12 +0000)]
Regulatory fixups:
o add missing channel flags for ECM, indoor, and outdoor constraints
o use HT capabilities to short-circuit HT20/HT40 channel construction
o rewrite 1/2 and 1/4 width channel handling yet again; previously
  we assumed there was a full-width version of the channel in the
  calibration table but that's not always true (e.g. for the Public
  Safety Band), now we first check the calibration table for the
  exact channel we want then fall back to the heuristics we used before
o fix HT channel construction; wasn't adjusting band edges for HT40
  channel bandwidth requirements

15 years agoTidy NOTES a bit:
wkoszek [Sat, 7 Feb 2009 00:15:30 +0000 (00:15 +0000)]
Tidy NOTES a bit:
- ath(4) is the last listed device, so make it's comment last as well
- since we have hints for le(4), bring it back by inserting commented
  out line until I check, if it can be safely enabled.
- bring snc(4) explanation
- put pmtimer comment together with other drivers' comments in a block
- bring comments for canbus, canbepm, pmc

olpt comment has been left blank, since I don't know how does this
driver differ from other printer drivers.

15 years agoTidy NOTES a bit:
wkoszek [Sat, 7 Feb 2009 00:06:13 +0000 (00:06 +0000)]
Tidy NOTES a bit:
- leave pmtimer comment that is common to other architectures.
- bring pbio explanation to the block comment relating to other
  drivers in the same block.

15 years agoTidy NOTES a bit:
wkoszek [Sat, 7 Feb 2009 00:01:10 +0000 (00:01 +0000)]
Tidy NOTES a bit:
- remove misleading nve/nfe comments, which make it hard to
  distinguish those two at a first glance
- bring pbio documentation to the block comment together with
  other drivers

I also brought commented out line responsible for si(4), since it
seems to compile and already has respective comment in this file.

15 years agoAdd rudimentary support for symbolic links on UDF. Links are stored as a
jhb [Fri, 6 Feb 2009 22:24:03 +0000 (22:24 +0000)]
Add rudimentary support for symbolic links on UDF.  Links are stored as a
sequence of pathname components.  We walk the list building a string in
the caller's passed in buffer.  Currently this only handles path names
in CS8 (character set 8) as that is what mkisofs generates for UDF images.

MFC after: 1 month

15 years agoComment about ural(4) isn't approprate here, since the driver is present in
wkoszek [Fri, 6 Feb 2009 22:22:08 +0000 (22:22 +0000)]
Comment about ural(4) isn't approprate here, since the driver is present in
global NOTES file.

cx(4) driver isn't present in this file, though it could be. However, cx(4)
seems to be more or less dead -- it hasn't been linked to the modules build,
and after TTY-ng transformations it doesn't compile.

Remove it until cx(4) is broken.

15 years agoural(4) is already present in global NOTES, thus there is no
wkoszek [Fri, 6 Feb 2009 21:56:55 +0000 (21:56 +0000)]
ural(4) is already present in global NOTES, thus there is no
need to explicitly list it here once again. This removes:

WARNING: duplicate option `DEV_URAL' encountered.
WARNING: duplicate device `ural' encountered.

Warnings when compiling LINT on amd64.

15 years agoFix AGP debugging code:
wkoszek [Fri, 6 Feb 2009 20:57:10 +0000 (20:57 +0000)]
Fix AGP debugging code:
- correct format strings
- fill opt_agp.h if AGP_DEBUG is defined
- bring AGP_DEBUG to LINT by mentioning it in NOTES

This should hopefully fix a warning that was...

Found by: Coverity Prevent(tm)
CID: 3676
Tested on: amd64, i386

15 years agoAdd support for fifos to UDF:
jhb [Fri, 6 Feb 2009 20:09:14 +0000 (20:09 +0000)]
Add support for fifos to UDF:
- Add a separate set of vnode operations that inherits from the fifo ops
  and use it for fifo nodes.
- Add a VOP_SETATTR() method that allows setting the size (by silently
  ignoring the requests) of fifos.  This is to allow O_TRUNC opens of
  fifo devices (e.g. I/O redirection in shells using ">").
- Add a VOP_PRINT() handler while I'm here.

15 years agoTweak the output of VOP_PRINT/vn_printf() some.
jhb [Fri, 6 Feb 2009 20:06:48 +0000 (20:06 +0000)]
Tweak the output of VOP_PRINT/vn_printf() some.
- Align the fifo output in fifo_print() with other vn_printf() output.
- Remove the leading space from lockmgr_printinfo() so its output lines up
  in vn_printf().
- lockmgr_printinfo() now ends with a newline, so remove an extra newline
  from vn_printf().

15 years agoAdd KASSERTs to make it easier to debug problems like the one fixed
trasz [Fri, 6 Feb 2009 18:16:01 +0000 (18:16 +0000)]
Add KASSERTs to make it easier to debug problems like the one fixed
in r188141.

Reviewed by: kib,attilio
Approved by: rwatson (mentor)
Tested by: pho
Sponsored by: FreeBSD Foundation

15 years agoWhen a device containing mounted UFS filesystem disappears, the type
trasz [Fri, 6 Feb 2009 17:14:07 +0000 (17:14 +0000)]
When a device containing mounted UFS filesystem disappears, the type
of devvp becomes VBAD, which UFS incorrectly interprets as snapshot
vnode, which in turns causes panic.  Fix it by replacing '!= VCHR'
with '== VREG'.

With this fix in place, you should no longer be able to panic the system
by removing a device with an UFS filesystem mounted from it - assuming
you don't use softupdates.

Reviewed by: kib
Tested by: pho
Approved by: rwatson (mentor)
Sponsored by: FreeBSD Foundation

15 years agoIn urtw_init() call urtw_stop(ifp, 0) rather than urtw_stop(ifp, 1)
kevlo [Fri, 6 Feb 2009 15:27:40 +0000 (15:27 +0000)]
In urtw_init() call urtw_stop(ifp, 0) rather than urtw_stop(ifp, 1)
to stop the device.

15 years agoNote that the slave address passed to smb(4) commands must be left-justified
jhb [Fri, 6 Feb 2009 15:09:31 +0000 (15:09 +0000)]
Note that the slave address passed to smb(4) commands must be left-justified
(LSB is 0).  The iic(4) manpage probably needs similar language to describe
the format it expects.

15 years agoRegen
kevlo [Fri, 6 Feb 2009 15:04:06 +0000 (15:04 +0000)]
Regen

15 years agoAdd the Buffalo WLI-U2-SG54HG
kevlo [Fri, 6 Feb 2009 15:03:17 +0000 (15:03 +0000)]
Add the Buffalo WLI-U2-SG54HG

15 years agoAdd a note to document that ichsmb(4) now uses left-justified SMBus slave
jhb [Fri, 6 Feb 2009 15:03:14 +0000 (15:03 +0000)]
Add a note to document that ichsmb(4) now uses left-justified SMBus slave
addresses.

15 years agoExpand the scope of the sysctllock sx lock to protect the sysctl tree itself.
jhb [Fri, 6 Feb 2009 14:51:32 +0000 (14:51 +0000)]
Expand the scope of the sysctllock sx lock to protect the sysctl tree itself.
Back in 1.1 of kern_sysctl.c the sysctl() routine wired the "old" userland
buffer for most sysctls (everything except kern.vnode.*).  I think to prevent
issues with wiring too much memory it used a 'memlock' to serialize all
sysctl(2) invocations, meaning that only one user buffer could be wired at
a time.  In 5.0 the 'memlock' was converted to an sx lock and renamed to
'sysctl lock'.  However, it still only served the purpose of serializing
sysctls to avoid wiring too much memory and didn't actually protect the
sysctl tree as its name suggested.  These changes expand the lock to actually
protect the tree.

Later on in 5.0, sysctl was changed to not wire buffers for requests by
default (sysctl_handle_opaque() will still wire buffers larger than a single
page, however).  As a result, user buffers are no longer wired as often.
However, many sysctl handlers still wire user buffers, so it is still
desirable to serialize userland sysctl requests.  Kernel sysctl requests
are allowed to run in parallel, however.

- Expose sysctl_lock()/sysctl_unlock() routines to exclusively lock the
  sysctl tree for a few places outside of kern_sysctl.c that manipulate
  the sysctl tree directly including the kernel linker and vfs_register().
- sysctl_register() and sysctl_unregister() require the caller to lock
  the sysctl lock using sysctl_lock() and sysctl_unlock().  The rest of
  the public sysctl API manage the locking internally.
- Add a locked variant of sysctl_remove_oid() for internal use so that
  external uses of the API do not need to be aware of locking requirements.
- The kernel linker no longer needs Giant when manipulating the sysctl
  tree.
- Add a missing break to the loop in vfs_register() so that we stop looking
  at the sysctl MIB once we have changed it.

MFC after: 1 month

15 years agoConsistently use <TAB> instead of spaces as option name and file
wkoszek [Fri, 6 Feb 2009 10:30:46 +0000 (10:30 +0000)]
Consistently use <TAB> instead of spaces as option name and file
separator.

15 years agofix spelling error
imp [Fri, 6 Feb 2009 09:34:17 +0000 (09:34 +0000)]
fix spelling error

Submitted by: trasz

15 years agoStore the entire funce for disk type functions (eg CF cards and the
imp [Fri, 6 Feb 2009 07:49:03 +0000 (07:49 +0000)]
Store the entire funce for disk type functions (eg CF cards and the
like).

15 years agoSet NFSMNT_ACDIRMAX flag in fallback_mount() function.
rodrigc [Fri, 6 Feb 2009 07:47:53 +0000 (07:47 +0000)]
Set NFSMNT_ACDIRMAX flag in fallback_mount() function.

15 years agoSet NFSMNT_ACREGMIN, NFSMNT_ACREGMAX, and NFSMNT_ACDIRMIN flags
rodrigc [Fri, 6 Feb 2009 07:42:21 +0000 (07:42 +0000)]
Set NFSMNT_ACREGMIN, NFSMNT_ACREGMAX, and NFSMNT_ACDIRMIN flags
in fallback_mount() function.
Add a comment to indicate that the fallback_mount() function
should eventually go away.

Submitted by: Jaakko Heinonen <jh saunalahti fi>

15 years agoRemove unused variable.
wkoszek [Fri, 6 Feb 2009 00:55:19 +0000 (00:55 +0000)]
Remove unused variable.

Found with: Coverity Prevent(tm)
CID: 4138

15 years agoMake config -x <kernel> only return non-zero characters,
wkoszek [Fri, 6 Feb 2009 00:50:21 +0000 (00:50 +0000)]
Make config -x <kernel> only return non-zero characters,
so that:

config -x <kernel> | grep <something>

just works.

Reported by: Danny Braniss <danny@cs.huji.ac.il>

15 years agoadd PSB channels to the calibration list
sam [Fri, 6 Feb 2009 00:48:56 +0000 (00:48 +0000)]
add PSB channels to the calibration list

15 years agoBring consistent debugging output for all values that are supposed
wkoszek [Thu, 5 Feb 2009 23:51:11 +0000 (23:51 +0000)]
Bring consistent debugging output for all values that are supposed
to be printed in a hexadecimal format. Otherwise, '270' doesn't say
much.

Reviewed by: imp