]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
17 years agoAdd missing .El.
brueffer [Sat, 12 May 2007 06:33:52 +0000 (06:33 +0000)]
Add missing .El.

Found by: man2hwnotes.pl

17 years agoaxe(4) supports altq now.
brueffer [Sat, 12 May 2007 05:59:15 +0000 (05:59 +0000)]
axe(4) supports altq now.

17 years agoAdd missing curly braces.
brueffer [Sat, 12 May 2007 05:56:58 +0000 (05:56 +0000)]
Add missing curly braces.

Approved by: rwatson (mentor)

17 years agoALTQify.
brueffer [Sat, 12 May 2007 05:56:10 +0000 (05:56 +0000)]
ALTQify.

Submitted by: J.R. Oldroyd
Reviewed by: mlaier
Approved by: rwatson (mentor)
MFC after: 1 week

17 years agoRemove compat ifdefs for FreeBSD versions < 500014, makes the code
brueffer [Sat, 12 May 2007 05:53:53 +0000 (05:53 +0000)]
Remove compat ifdefs for FreeBSD versions < 500014, makes the code
quite a bit easier to read.

Approved by: rwatson (mentor)
MFC after: 1 weeks

17 years ago- Add bits for userland profiling. For sun4u this is compile-tested only.
marius [Fri, 11 May 2007 23:43:55 +0000 (23:43 +0000)]
- Add bits for userland profiling. For sun4u this is compile-tested only.
- Replace magic 14 with PIL_TICK.

17 years agoAutogenerate hardware notes for hptiop(4).
brueffer [Fri, 11 May 2007 22:58:09 +0000 (22:58 +0000)]
Autogenerate hardware notes for hptiop(4).

17 years agoConnect hptiop(4) to the build. Only install hptmv(4) on i386 and amd64.
brueffer [Fri, 11 May 2007 22:53:09 +0000 (22:53 +0000)]
Connect hptiop(4) to the build.  Only install hptmv(4) on i386 and amd64.

17 years agoManpage for the hptiop(4) driver. Could use a little more content.
brueffer [Fri, 11 May 2007 22:52:19 +0000 (22:52 +0000)]
Manpage for the hptiop(4) driver.  Could use a little more content.

17 years agoNew release notes:
brueffer [Fri, 11 May 2007 21:50:32 +0000 (21:50 +0000)]
New release notes:
- vm.kmem_size_min tunable added
- fdisk -p
- ipfw RH0 and MIPv6 RH2 filtering
- pw -M
- Camellia cipher added

Modified release notes:
- twa(4) was actually updated to version 3.70.03.006

17 years agoMistake in the logic deciding what adapters need
jfv [Fri, 11 May 2007 21:36:08 +0000 (21:36 +0000)]
Mistake in the logic deciding what adapters need
to map the IO BAR. Causing the driver to fail on
th 82542.

Reviewed by:pdeuskar
Approved by:pdeuskar

17 years agoDrop everything that doesn't belong into this new file.
andre [Fri, 11 May 2007 21:17:53 +0000 (21:17 +0000)]
Drop everything that doesn't belong into this new file.
It's neither functional not connected to the build yet.

17 years agoDrop everything that doesn't belong into this new file.
andre [Fri, 11 May 2007 21:04:57 +0000 (21:04 +0000)]
Drop everything that doesn't belong into this new file.
It's neither functional nor connected to the build yet.

17 years agoMake the TCP timer callout obtain Giant if the network stack is marked
andre [Fri, 11 May 2007 20:52:47 +0000 (20:52 +0000)]
Make the TCP timer callout obtain Giant if the network stack is marked
as non-mpsafe.

This change is to be removed when all protocols are mp-safe.

17 years agoForced commit to note repo copy from sys/netinet/tcp_input.c rev. 1.349.
andre [Fri, 11 May 2007 20:31:23 +0000 (20:31 +0000)]
Forced commit to note repo copy from sys/netinet/tcp_input.c rev. 1.349.

17 years agoForced commit to note repo copy from sys/netinet/tcp_subr.c rev. 1.280.
andre [Fri, 11 May 2007 20:30:09 +0000 (20:30 +0000)]
Forced commit to note repo copy from sys/netinet/tcp_subr.c rev. 1.280.

17 years agoAdd the timestamp offset to struct tcptw so we can generate proper
andre [Fri, 11 May 2007 18:29:39 +0000 (18:29 +0000)]
Add the timestamp offset to struct tcptw so we can generate proper
ACKs in TIME_WAIT state that don't get dropped by the PAWS check
on the receiver.

17 years ago- Initialize login->id and fix problem for non-zero login id.
simokawa [Fri, 11 May 2007 14:51:13 +0000 (14:51 +0000)]
- Initialize login->id and fix problem for non-zero login id.
- Increase maxopenings for multiple lun/initiators.

MFC after: 3 days

17 years agoPerformance optimization of the "encryption without compression" case by
mav [Fri, 11 May 2007 14:36:02 +0000 (14:36 +0000)]
Performance optimization of the "encryption without compression" case by
avoiding memory allocation and data copying.
Encrypting directly at the original mbuf chain.

Approved by: glebius (mentor)

17 years agoBad merge.
mjacob [Fri, 11 May 2007 13:47:28 +0000 (13:47 +0000)]
Bad merge.

17 years agoRemove more one more stale comment regarding unpcb type-safety.
rwatson [Fri, 11 May 2007 12:28:45 +0000 (12:28 +0000)]
Remove more one more stale comment regarding unpcb type-safety.

17 years agoClarify and update quite a few comments to reflect locking optimizations,
rwatson [Fri, 11 May 2007 12:10:45 +0000 (12:10 +0000)]
Clarify and update quite a few comments to reflect locking optimizations,
the addition of unpcb refcounts, and bug fixes.  Some of these fixes are
appropriate for MFC.

MFC after: 3 days

17 years agoCoalesce two identical UCB licenses into a single license instance with
rwatson [Fri, 11 May 2007 11:21:43 +0000 (11:21 +0000)]
Coalesce two identical UCB licenses into a single license instance with
one set of copyright years.

White space and comment cleanup.

Export $FreeBSD$ via __FBSDID.

17 years agostrlcpy() may be faster than snprintf(), but it is less portable, and this
des [Fri, 11 May 2007 11:10:05 +0000 (11:10 +0000)]
strlcpy() may be faster than snprintf(), but it is less portable, and this
is not performance critical code anyway.  Also, avoid using strlen() to
obtain information which we already have.

MFC after: 3 weeks

17 years agoMinor white space and style cleanups.
rwatson [Fri, 11 May 2007 11:05:30 +0000 (11:05 +0000)]
Minor white space and style cleanups.

17 years agoWhite space and style cleanup.
rwatson [Fri, 11 May 2007 11:00:48 +0000 (11:00 +0000)]
White space and style cleanup.

17 years agoMinor white space/style normalization.
rwatson [Fri, 11 May 2007 10:50:31 +0000 (10:50 +0000)]
Minor white space/style normalization.

17 years agoNormalize style a bit: reduce pseudo-randomness of comment layout and
rwatson [Fri, 11 May 2007 10:48:30 +0000 (10:48 +0000)]
Normalize style a bit: reduce pseudo-randomness of comment layout and
white space.  Remove 'register'.

17 years agoUse ANSI C function declarations throughout netipx.
rwatson [Fri, 11 May 2007 10:38:34 +0000 (10:38 +0000)]
Use ANSI C function declarations throughout netipx.

Remove 'register' use.

17 years agoReduce network stack oddness: implement .pru_sockaddr and .pru_peeraddr
rwatson [Fri, 11 May 2007 10:20:51 +0000 (10:20 +0000)]
Reduce network stack oddness: implement .pru_sockaddr and .pru_peeraddr
protocol entry points using functions named proto_getsockaddr and
proto_getpeeraddr rather than proto_setsockaddr and proto_setpeeraddr.
While it's true that sockaddrs are allocated and set, the net effect is
to retrieve (get) the socket address or peer address from a socket, not
set it, so align names to that intent.

17 years agoRemove unneeded wrappers for in_setsockaddr() and in_setpeeraddr(), which
rwatson [Fri, 11 May 2007 09:54:53 +0000 (09:54 +0000)]
Remove unneeded wrappers for in_setsockaddr() and in_setpeeraddr(), which
used to exist so pcbinfo locks could be acquired, but are no longer
required as a result of socket/pcb reference model refinements.

17 years agoFix pointy-hat problem with BUS_DMA_ROOTARG macro that caused problems for sparc64.
mjacob [Fri, 11 May 2007 06:28:26 +0000 (06:28 +0000)]
Fix pointy-hat problem with BUS_DMA_ROOTARG macro that caused problems for sparc64.
Candidate for immediate MFC.

Noticed by: Everyone-maxim contacted.

17 years agoDo not dereference linux_to_bsd_signal[-1] if userland has
kan [Fri, 11 May 2007 01:25:51 +0000 (01:25 +0000)]
Do not dereference linux_to_bsd_signal[-1] if userland has
passed zero as exit signal.

GCC 4.2 changes the kernel data segment layout not to have 0
in that memory location. This code ran by luck before and now
the luck has run out.

17 years agoFix an incorrect replace of a timer reference made during the TCP timer
andre [Thu, 10 May 2007 23:11:29 +0000 (23:11 +0000)]
Fix an incorrect replace of a timer reference made during the TCP timer
rewrite in rev. 1.132.  This unmasked yet another bug that causes certain
connections to get indefinately stuck in LAST_ACK state.

17 years agoFix typo.
des [Thu, 10 May 2007 18:15:30 +0000 (18:15 +0000)]
Fix typo.

Submitted by: Bård Skaflestad <bardsk@math.ntnu.no>

17 years agoMove universally to ANSI C function declarations, with relatively
rwatson [Thu, 10 May 2007 15:58:48 +0000 (15:58 +0000)]
Move universally to ANSI C function declarations, with relatively
consistent style(9)-ish layout.

17 years agoBump __FreeBSD_version for the addition of flopen(3) to libutil.
des [Thu, 10 May 2007 15:37:27 +0000 (15:37 +0000)]
Bump __FreeBSD_version for the addition of flopen(3) to libutil.

Requested by: delphij

17 years agoThe version string scheme embeds the freebsd major release version,
scottl [Thu, 10 May 2007 15:36:58 +0000 (15:36 +0000)]
The version string scheme embeds the freebsd major release version,
something that I wasn't aware of with the last import.  Adjust the
string accordingly for 7-CURRENT.

17 years agoCollapse the mfi_ld object. Add an ioctl to help management apps map
scottl [Thu, 10 May 2007 15:33:41 +0000 (15:33 +0000)]
Collapse the mfi_ld object.  Add an ioctl to help management apps map
array Id's to FreeBSD device names.

17 years agoWell gag me with a spoon... I'm so used to working at high WARNS levels
des [Thu, 10 May 2007 15:01:42 +0000 (15:01 +0000)]
Well gag me with a spoon...   I'm so used to working at high WARNS levels
that I make stupid fundamental mistakes like this when I don't.

17 years agoRemove superfluous unexpanded RCS tag.
des [Thu, 10 May 2007 14:56:07 +0000 (14:56 +0000)]
Remove superfluous unexpanded RCS tag.

17 years agoUse flopen(3).
des [Thu, 10 May 2007 14:54:53 +0000 (14:54 +0000)]
Use flopen(3).

MFC after: 3 weeks

17 years agoDTRT when O_NONBLOCK is specified.
des [Thu, 10 May 2007 14:52:57 +0000 (14:52 +0000)]
DTRT when O_NONBLOCK is specified.

MFC after: 3 weeks

17 years agoI'm tired of seeing this done incorrectly and non-portably, so add a
des [Thu, 10 May 2007 14:43:31 +0000 (14:43 +0000)]
I'm tired of seeing this done incorrectly and non-portably, so add a
flopen(3) function which reliably opens and locks a file.

MFC after: 3 weeks

17 years agozfs & zpool have been moved from /usr/sbin into /sbin. The corresponding
roberto [Thu, 10 May 2007 13:25:27 +0000 (13:25 +0000)]
zfs & zpool have been moved from /usr/sbin into /sbin.  The corresponding
libraries in /usr/lib has been replaced by symlinks into /lib already.

Requested by:       pjd

17 years agoMake a missing or empty VPD non-fatal, as it was prior to rev 1.131.
mckay [Thu, 10 May 2007 13:10:34 +0000 (13:10 +0000)]
Make a missing or empty VPD non-fatal, as it was prior to rev 1.131.
With this change, my D-Link DGE-530T rev A1 is operational again.

17 years agoWe need four disks these days, so reflect this.
delphij [Thu, 10 May 2007 12:51:08 +0000 (12:51 +0000)]
We need four disks these days, so reflect this.

Pointed out by: phk
PR: docs/112571

17 years agoAnother merging from sbin/fdisk/fdisk.8.
nyan [Thu, 10 May 2007 12:33:00 +0000 (12:33 +0000)]
Another merging from sbin/fdisk/fdisk.8.

PR: 109929
Submitted by: Alex Kozlov

17 years agoThe amd64 monthly snapshots began overflowing disc1 (and that's without any
kensmith [Thu, 10 May 2007 12:04:29 +0000 (12:04 +0000)]
The amd64 monthly snapshots began overflowing disc1 (and that's without any
packages...).

17 years agoUpdate nologin(5) to match the modern reality of login.conf(5) and PAM.
yar [Thu, 10 May 2007 11:22:24 +0000 (11:22 +0000)]
Update nologin(5) to match the modern reality of login.conf(5) and PAM.

17 years agoHook wlan_amrr up to the build.
kevlo [Thu, 10 May 2007 08:53:57 +0000 (08:53 +0000)]
Hook wlan_amrr up to the build.

17 years agoDon't hold mutex while releasing bus resources.
ariff [Thu, 10 May 2007 01:46:51 +0000 (01:46 +0000)]
Don't hold mutex while releasing bus resources.

17 years agoAdd wlan_amrr. ural(4) uses amrr as transmit rate control.
kevlo [Thu, 10 May 2007 01:39:50 +0000 (01:39 +0000)]
Add wlan_amrr. ural(4) uses amrr as transmit rate control.

17 years agoAdd wlan_amrr. ural(4) uses amrr as transmit rate control.
kevlo [Thu, 10 May 2007 01:32:17 +0000 (01:32 +0000)]
Add wlan_amrr. ural(4) uses amrr as transmit rate control.

17 years agoNeed sys/cdevs.h for the macro FBSDID to work.
mjacob [Wed, 9 May 2007 23:19:55 +0000 (23:19 +0000)]
Need sys/cdevs.h for the macro FBSDID to work.

17 years agoUpdate stale comment on protecting UMA per-CPU caches: we now use
rwatson [Wed, 9 May 2007 22:53:34 +0000 (22:53 +0000)]
Update stale comment on protecting UMA per-CPU caches: we now use
critical sections rather than mutexes.

17 years agoIncrease debug level - this message is not that important.
pjd [Wed, 9 May 2007 22:32:49 +0000 (22:32 +0000)]
Increase debug level - this message is not that important.

17 years agoUse dummy definitions for codeset 3 and 4 like in ko_KR.eucKR.src
ache [Wed, 9 May 2007 21:34:51 +0000 (21:34 +0000)]
Use dummy definitions for codeset 3 and 4 like in ko_KR.eucKR.src

PR:             51085
Submitted by:   Yamazaki Atsushi <ats@mbi.nifty.com>

17 years agoForced commit to note that the previous revision also contained the following:
brueffer [Wed, 9 May 2007 20:24:17 +0000 (20:24 +0000)]
Forced commit to note that the previous revision also contained the following:

New release note:
- Linux 2.6.16 compat support

Submitted by: rdivacky

17 years agoNew release notes:
brueffer [Wed, 9 May 2007 20:22:10 +0000 (20:22 +0000)]
New release notes:
- mpt(4), isp(4), ciss(4) MPSAFE
- rum(4) driver added
- hptiop(4) driver added

Updated release notes:
- em(4) driver updated to 6.5.0
- small fixes in the top -j entry
- twa(4) update to 3.60.03.006

17 years agoAdd some newly supported controllers.
brueffer [Wed, 9 May 2007 20:05:43 +0000 (20:05 +0000)]
Add some newly supported controllers.

17 years agoIntegrate the Camellia Block Cipher. For more information see RFC 4132
gnn [Wed, 9 May 2007 19:37:02 +0000 (19:37 +0000)]
Integrate the Camellia Block Cipher.  For more information see RFC 4132
and its bibliography.

Submitted by:   Tomoyuki Okazaki <okazaki at kick dot gr dot jp>
MFC after:      1 month

17 years agoo Teach get_mac_addr_mask() to not silently accept incorrect MAC
maxim [Wed, 9 May 2007 18:31:49 +0000 (18:31 +0000)]
o Teach get_mac_addr_mask() to not silently accept incorrect MAC
addresses.
o Swap a couple of magic 6s by ETHER_ADDR_LEN.

PR: bin/80913
Submitted by: Andrey V. Elsukov
MFC after: 1 month

17 years agoIt turns out that the hptiop driver isn't portable after all. Confine it to
scottl [Wed, 9 May 2007 15:55:45 +0000 (15:55 +0000)]
It turns out that the hptiop driver isn't portable after all.  Confine it to
amd64 and i386 for now.

17 years agoTwo major items here:
rrs [Wed, 9 May 2007 13:30:06 +0000 (13:30 +0000)]
Two major items here:
- All printf that was surrounded by #ifdef SCTP_DEBUG moves to
  a macro that does all of this. This removes all printfs from
  the code and makes the code more portable and easier to
  read.
- Static Analysis (cisco) - found a few bugs, but mostly we
  add checks for NULL pointers and such to make the tool
  happy. We now pass the Cisco SA tools checks except for
  where it does not understand tailq/lists. We still need
  to look at the coverity tools output too (this is like
  the cisco SA tool) and see if it wants us to fix any other
  items. Hopefully this will be the last major churn in the
  code other than bug fixes.

17 years agoReduce diffs against fdisk/fdisk.8.
nyan [Wed, 9 May 2007 11:46:31 +0000 (11:46 +0000)]
Reduce diffs against fdisk/fdisk.8.

17 years agoo Fix typo: comments start by "#" not "*".
maxim [Wed, 9 May 2007 11:43:04 +0000 (11:43 +0000)]
o Fix typo: comments start by "#" not "*".

17 years agoo Fix style(9) bugs introduced in the last commit.
maxim [Wed, 9 May 2007 11:39:46 +0000 (11:39 +0000)]
o Fix style(9) bugs introduced in the last commit.

Pointed out by: bde

17 years ago- A partition number is allowed to 1-8.
nyan [Wed, 9 May 2007 11:38:58 +0000 (11:38 +0000)]
- A partition number is allowed to 1-8.
- Remove non-exist -b flag.
- Add reference to boot98cfg and newfs.
- s/disklabel/bsdlabel/

17 years ago- Remove dependency on deprecated if_watchdog ABI.
kevlo [Wed, 9 May 2007 09:32:41 +0000 (09:32 +0000)]
- Remove dependency on deprecated if_watchdog ABI.
- Move amrr code into net80211 itself.
- Handle ->INIT state transition in ural_newstate() to eliminate the race.

Tested with a ASUS WL-167g.

17 years agotl(4) appears to support long frames.
yar [Wed, 9 May 2007 09:02:11 +0000 (09:02 +0000)]
tl(4) appears to support long frames.

Tested by: Peter Jeremy <peterjeremy at optushome dot com dot au>

17 years agobackout experimental adaptive spinning mutex for product use.
davidxu [Wed, 9 May 2007 08:39:33 +0000 (08:39 +0000)]
backout experimental adaptive spinning mutex for product use.

17 years agoIntroduce a driver for the Highpoint RocketRAID 3xxx series of controllers.
scottl [Wed, 9 May 2007 07:07:26 +0000 (07:07 +0000)]
Introduce a driver for the Highpoint RocketRAID 3xxx series of controllers.
The driver relies on CAM.

Many thanks to Highpoint for providing this driver.

17 years agoo Unbreak "options TCPDEBUG" && "nooptions INET6" kernel build.
maxim [Wed, 9 May 2007 06:09:40 +0000 (06:09 +0000)]
o Unbreak "options TCPDEBUG" && "nooptions INET6" kernel build.

PR: kern/112517
Submitted by: vd

17 years agoWrite the output parameter (if present) for the add, create, delete
marcel [Wed, 9 May 2007 05:37:53 +0000 (05:37 +0000)]
Write the output parameter (if present) for the add, create, delete
destroy and modify verbs.

17 years agoo Add "line" device to a list of mixer(8) devices.
maxim [Wed, 9 May 2007 05:33:54 +0000 (05:33 +0000)]
o Add "line" device to a list of mixer(8) devices.

PR: docs/112520
Submitted by: Pietro Cerutti
MFC after: 1 week

17 years agoAdd rum(4).
brueffer [Wed, 9 May 2007 05:18:04 +0000 (05:18 +0000)]
Add rum(4).

17 years agoBump date for previous commit, also new sentence -> new line.
brueffer [Wed, 9 May 2007 05:05:50 +0000 (05:05 +0000)]
Bump date for previous commit, also new sentence -> new line.

17 years agoImport version 3.60.03.006 of the TWA driver:
scottl [Wed, 9 May 2007 04:16:32 +0000 (04:16 +0000)]
Import version 3.60.03.006 of the TWA driver:

This patch does the following:

- Remove un-necessary code that is not even compiling into the driver
 under TW_OSL_NON_DMA_MEM_ALLOC_PER_REQUEST defines.
- Remove bundled firmware image and associated "files" entry for tw_cl_fwimg.c
- Remove bundled firmware flashing routines. We now have tw_update userspace
 FreeBSD controller flash utility.
- Fix driver crash on load due to shared interrupt.
- Fix 2 lock leaks for Giant lock.
- Fix CCB leak.
- Add support for 9650SE controllers.

Many thanks to 3Ware/AMCC for continuing to support FreeBSD.

17 years agoAdd documentation for the vm.kmem_size_min and vm.kmem_size_max tunables.
sepotvin [Wed, 9 May 2007 02:37:58 +0000 (02:37 +0000)]
Add documentation for the vm.kmem_size_min and vm.kmem_size_max tunables.

Approved by: njl (mentor, blanket)

17 years agoWhen reverting the creation of a partitioning scheme on a provider,
marcel [Wed, 9 May 2007 01:46:42 +0000 (01:46 +0000)]
When reverting the creation of a partitioning scheme on a provider,
the failure to probe an existing partitioning scheme means that no
previous partitioning scheme existed. Don't error. Just destroy the
geom.

17 years agoA couple bug fixes that I've had internally at Intel. First is a long
jfv [Wed, 9 May 2007 00:41:45 +0000 (00:41 +0000)]
A couple bug fixes that I've had internally at Intel. First is a long
time workaround for problems with 82571 adapters and LAAs, one port
getting reset can cause the other to have its RAR[0] also reset,
thus overwriting an LAA. This fix works around it by also keeping
the address in the last array member.

The other bug is specific to the new 575 adapter, its transmit code
logic in handling hwassists was too crude, it broken when doing
bridges. I am much happier with the new logic,we may want to change
the legacy path at some point to something similar.

Reviewed by: pdeuskar
Approved by: pdeuskar

17 years agoHandle CPUs with APIC IDs higher than 32 (at least one IBM server uses
jhb [Tue, 8 May 2007 22:01:04 +0000 (22:01 +0000)]
Handle CPUs with APIC IDs higher than 32 (at least one IBM server uses
an APIC ID of 38 for its second CPU):
- Add a new MAX_APIC_ID constant for the highest valid APIC ID for modern
  systems.
- Size the various arrays in the MADT, MP Table, and SMP code that are
  indexed by APIC IDs to allow for up to MAX_APIC_ID.
- Explicitly go through and assign logical cpu ids to local APICs before
  starting any of the APs up rather than doing it while starting up the
  APs.  This step is now where we honor MAXCPU.

MFC after: 1 week

17 years agoAdd destroyed cookie values for sx locks and rwlocks as well as extra
jhb [Tue, 8 May 2007 21:51:37 +0000 (21:51 +0000)]
Add destroyed cookie values for sx locks and rwlocks as well as extra
KASSERTs so that any lock operations on a destroyed lock will panic or
hang.

17 years agoTeach 'show lock' to properly handle a destroyed mutex.
jhb [Tue, 8 May 2007 21:50:46 +0000 (21:50 +0000)]
Teach 'show lock' to properly handle a destroyed mutex.

17 years agoFix a potential LOR with sx_sleep() and cv_wait() with sx locks by
jhb [Tue, 8 May 2007 21:49:59 +0000 (21:49 +0000)]
Fix a potential LOR with sx_sleep() and cv_wait() with sx locks by
1) adding the thread to the sleepq via sleepq_add() before dropping the
lock, and 2) dropping the sleepq lock around calls to lc_unlock() for
sleepable locks (i.e. locks that use sleepq's in their implementation).

17 years agoMinor fixes and tweaks to the x86 interrupt code:
jhb [Tue, 8 May 2007 21:29:14 +0000 (21:29 +0000)]
Minor fixes and tweaks to the x86 interrupt code:
- Split the intr_table_lock into an sx lock used for most things, and a
  spin lock to protect intrcnt_index.  Originally I had this as a spin lock
  so interrupt code could use it to lookup sources.  However, we don't
  actually do that because it would add a lot of overhead to interrupts,
  and if we ever do support removing interrupt sources, we can use other
  means to safely do so w/o locking in the interrupt handling code.
- Replace is_enabled (boolean) with is_handlers (a count of handlers) to
  determine if a source is enabled or not.  This allows us to notice when
  a source is no longer in use.  When that happens, we now invoke a new
  PIC method (pic_disable_intr()) to inform the PIC driver that the
  source is no longer in use.  The I/O APIC driver frees the APIC IDT
  vector when this happens.  The MSI driver no longer needs to have a
  hack to clear is_enabled during msi_alloc() and msix_alloc() as a result
  of this change as well.
- Add an apic_disable_vector() to reset an IDT vector back to Xrsvd to
  complement apic_enable_vector() and use it in the I/O APIC and MSI code
  when freeing an IDT vector.
- Add a new nexus hook: nexus_add_irq() to ask the nexus driver to add an
  IRQ to its irq_rman.  The MSI code uses this when it creates new
  interrupt sources to let the nexus know about newly valid IRQs.
  Previously the msi_alloc() and msix_alloc() passed some extra stuff
  back to the nexus methods which then added the IRQs.  This approach is
  a bit cleaner.
- Change the MSI sx lock to a mutex.  If we need to create new sources,
  drop the lock, create the required number of sources, then get the lock
  and try the allocation again.

17 years agoMFp4:
marcel [Tue, 8 May 2007 20:18:17 +0000 (20:18 +0000)]
MFp4:
119373: o  Remove the query verb, along with the request and response
   parameters.
o  Add the version and output parameters.
119390: [APM,GPT] Properly clear deleted entries.
119394: o  Make the alias the standard and use the '!' to prefix
   literal partition types.
o  Treat schemes and partition types as case insensitive.
119462: [GPT] Fix a page fault caused when modifying a partition entry
without a new partition type.

17 years agoReduce the default number of header options that the IPv6 protocol
gnn [Tue, 8 May 2007 20:11:36 +0000 (20:11 +0000)]
Reduce the default number of header options that the IPv6 protocol
stack will process from 50 to 15.  As this is a sysctl variable it
can be tuned up or down at the user/administrator's whim.

Submitted by: itojun
MFC after: 1 day

17 years agoHandle temporary redirects (HTTP status code 307)
des [Tue, 8 May 2007 19:28:03 +0000 (19:28 +0000)]
Handle temporary redirects (HTTP status code 307)

PR: kern/112515
Submitted by: Ryan C. Gordon <icculus@icculus.org>
MFC after: 3 weeks

17 years agoo Add AnyData ADU-E100H modem. Sort a list of ubsa(4) supported devices.
maxim [Tue, 8 May 2007 18:51:40 +0000 (18:51 +0000)]
o Add AnyData ADU-E100H modem.  Sort a list of ubsa(4) supported devices.

PR: usb/95803
Submitted by: Jindrich Fucik
Tested by: R.Mahmatkhanov
MFC after: 2 weeks

17 years ago- Add handling of MXGEFW_CMD_UNKNOWN in mxge_send_cmd().
gallatin [Tue, 8 May 2007 18:45:43 +0000 (18:45 +0000)]
- Add handling of MXGEFW_CMD_UNKNOWN in mxge_send_cmd().
- Convert mxge_send_cmd result handling to a switch rather
  than adding a new elseif for MXGEFW_CMD_UNKNOWN

Sponsored by: Myricom Inc.

17 years ago- Copyright change, cisco's silly tool wants it to say:
rrs [Tue, 8 May 2007 17:01:12 +0000 (17:01 +0000)]
-  Copyright change, cisco's silly tool wants it to say:
       "Copyright (c) 2001-2007, by Cisco Systems,"
   instead of
       *Copyright (c) 2001-2007, Cisco Systems,"

-  Also fix a few straglers that were still in 2006.

17 years ago- Get rid of the sctp_inpcb_free() "magic numbers", now they
rrs [Tue, 8 May 2007 15:53:03 +0000 (15:53 +0000)]
-  Get rid of the sctp_inpcb_free() "magic numbers", now they
   are sensible defines that tell what you are directing
   the function to do.

17 years agoSet the program return value non-zero if we enounter
kientzle [Tue, 8 May 2007 15:22:21 +0000 (15:22 +0000)]
Set the program return value non-zero if we enounter
an error while skipping the body of an entry during
"tar -t" processing.

Thanks to: Spencer Minear
PR: misc/112521

17 years ago- Static analyisis fixes for cisco's commit (this is equivilant
rrs [Tue, 8 May 2007 14:32:53 +0000 (14:32 +0000)]
-  Static analyisis fixes for cisco's commit (this is equivilant
   to the coverity tool.. may even be the same one.. not sure).
-  A bug in the way sctp_abort() and friends were
   setting the IP_CLOSE flag.. and NOT passing the
   last argument as a (,1)... so that things would
   get freed..

17 years agoSet the debug.acpi.acpi_ca_version sysctl even if ACPI support is not
marks [Tue, 8 May 2007 14:31:49 +0000 (14:31 +0000)]
Set the debug.acpi.acpi_ca_version sysctl even if ACPI support is not
available.

17 years agoFirmware update & improvements to firmware selection:
gallatin [Tue, 8 May 2007 14:19:43 +0000 (14:19 +0000)]
Firmware update & improvements to firmware selection:

- Update to latest (1.4.17) firmware.

- Use the new MXGEFW_CMD_UNALIGNED_TEST (added in firmare 1.4.16) to
  have the firmware tell us if the PCIe chipset supports aligned PCIe
  completions.

- Hard to maintain, and frequently out of date whitelist of PCIe
  chipsets known to produce aligned completions removed, as it has been
  replaced in its role of selecting the correct firmware to run by the
  use of MXGEFW_CMD_UNALIGNED_TEST.

- Break the dma test out of mxge_reset() and into its own function
  (mxge_dma_test()) so it can be used by both the normal DMA test, and
  to run the unaligned test.

- Improved support for enabling ECRCs

Sponsored by: Myricom Inc.

17 years agoAdd missing socket buffer unlock before returning to userland.
yongari [Tue, 8 May 2007 12:34:14 +0000 (12:34 +0000)]
Add missing socket buffer unlock before returning to userland.

Reviewed by:    rwatson

17 years agoAdd .Li mark-up to Hesiod, NIS, and nsswitch keywords so that they stand out
yar [Tue, 8 May 2007 11:00:07 +0000 (11:00 +0000)]
Add .Li mark-up to Hesiod, NIS, and nsswitch keywords so that they stand out
when the manpage is read via an advanced output device.