]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
11 years agoFix uplcom clear stall logic for PL2303HX.
hselasky [Wed, 21 Nov 2012 22:04:40 +0000 (22:04 +0000)]
Fix uplcom clear stall logic for PL2303HX.

Submitted by: Mark Johnston
MFC after: 1 week

11 years agoRevert r243228. This commit appears to cause more trouble than
crees [Wed, 21 Nov 2012 18:12:28 +0000 (18:12 +0000)]
Revert r243228.  This commit appears to cause more trouble than
it was designed to avoid; the issue described in the PR was no
longer an issue anyway.

11 years agoPrint correct unit number when attaching preloaded memory disks.
jh [Wed, 21 Nov 2012 17:05:57 +0000 (17:05 +0000)]
Print correct unit number when attaching preloaded memory disks.
Retire now unused mdunits variable.

11 years agoDisallow attaching preloaded memory disks via ioctl.
jh [Wed, 21 Nov 2012 16:56:47 +0000 (16:56 +0000)]
Disallow attaching preloaded memory disks via ioctl.

- The feature is dangerous because the kernel code didn't check
  validity of the memory address provided from user space.
- It seems that mdconfig(8) never really supported attaching preloaded
  memory disks.
- Preloaded memory disks are automatically attached during md(4)
  initialization. Thus there shouldn't be much use for the feature.

PR: kern/169683
Discussed on: freebsd-hackers

11 years agoSetup BAT0 and BAT1 on the Wii.
adrian [Wed, 21 Nov 2012 08:04:21 +0000 (08:04 +0000)]
Setup BAT0 and BAT1 on the Wii.

This is the missing piece for FreeBSD/Wii, but there's still a lot of
work ahead. We have to reset the MMU in locore before continuing
the boot process because we don't know how the boot loaders might
have setup the BATs. We also disable the PCI BAT because there's no PCI
bus on the Wii.

Thanks to Nathan Whitehorn and Peter Grenhan for their help.

Submitted by: Margarida Gouveia

11 years agoCorrect an error in r230623. When both VM_ALLOC_NODUMP and VM_ALLOC_ZERO
alc [Wed, 21 Nov 2012 06:26:18 +0000 (06:26 +0000)]
Correct an error in r230623.  When both VM_ALLOC_NODUMP and VM_ALLOC_ZERO
were specified to vm_page_alloc(), PG_NODUMP wasn't being set on the
allocated page when it happened to be pre-zeroed.

MFC after: 5 days

11 years agoMake sure the address starts on a cache line boundary.
cognet [Wed, 21 Nov 2012 01:38:40 +0000 (01:38 +0000)]
Make sure the address starts on a cache line boundary.

11 years agoDo not expose LIBCXXRT and LIBCPLUSPLUS in bsd.libnames.mk, if
dim [Tue, 20 Nov 2012 21:26:13 +0000 (21:26 +0000)]
Do not expose LIBCXXRT and LIBCPLUSPLUS in bsd.libnames.mk, if
WITHOUT_LIBCPLUSPLUS is specified.

Submitted by: Garrett Cooper <yanegomi@gmail.com>
MFC after: 3 days

11 years agoMerge ACPICA 20121114.
jkim [Tue, 20 Nov 2012 21:01:59 +0000 (21:01 +0000)]
Merge ACPICA 20121114.

11 years agoNon-void function should return a value.
emaste [Tue, 20 Nov 2012 19:23:44 +0000 (19:23 +0000)]
Non-void function should return a value.

Found by: clang

11 years agoSchedule garbage collection run for the in-flight rights passed over
kib [Tue, 20 Nov 2012 15:45:48 +0000 (15:45 +0000)]
Schedule garbage collection run for the in-flight rights passed over
the unix domain sockets to the next tick, coalescing the serial calls
until the collection fires.  The thought is that more work for the
collector could arise in the near time, allowing to clean more and not
spend too much CPU on repeated collection when there is no garbage.

Currently the collection task is fired immediately upon unix domain
socket close if there are any rights in flight, which caused excessive
CPU usage and too long blocking of the threads waiting for
unp_list_lock and unp_link_rwlock in write mode.

Robert noted that it would be nice if we could find some heuristic by
which we decide whether to run GC a bit more quickly.  E.g., if the
number of UNIX domain sockets is close to its resource limit, but not
quite.

Reported and tested by: Markus Gebert <markus.gebert@hostpoint.ch>
Reviewed by: rwatson
MFC after: 2 weeks

11 years agoAdd a special meaning to the negative ticks argument for
kib [Tue, 20 Nov 2012 15:33:48 +0000 (15:33 +0000)]
Add a special meaning to the negative ticks argument for
taskqueue_enqueue_timeout().  Do not rearm the callout if it is
already armed and the ticks is negative.  Otherwise rearm it to fire
in abs(ticks) ticks in the future.

The intended use is to call taskqueue_enqueue_timeout() for the given
timeout_task with the same negative ticks argument.  As result, the
task is scheduled to execute not further than abs(ticks) ticks in
future, and the consequent enqueues are coalesced until the already
scheduled task is finished.

Reviewed by: rwatson
Tested by: Markus Gebert <markus.gebert@hostpoint.ch>
MFC after: 2 weeks

11 years agoRemove the check and panic for an impossible condition. The NULL
kib [Tue, 20 Nov 2012 15:25:00 +0000 (15:25 +0000)]
Remove the check and panic for an impossible condition. The NULL
lowervp vnode v_vnlock would cause panic due to NULL pointer
dereference much earlier.

MFC after: 1 week

11 years agoFix module build after r243245.
kib [Tue, 20 Nov 2012 15:23:22 +0000 (15:23 +0000)]
Fix module build after r243245.

11 years agoConnect ip6_mroute kernel module to the build.
ae [Tue, 20 Nov 2012 14:11:27 +0000 (14:11 +0000)]
Connect ip6_mroute kernel module to the build.

MFC after: 1 week

11 years agoRemove opt_inet.h, it isn't required here.
ae [Tue, 20 Nov 2012 14:09:37 +0000 (14:09 +0000)]
Remove opt_inet.h, it isn't required here.

MFC after: 1 week

11 years agoIn NIS mode first chmod(2) the temporary file and is succeed then rename(2)
bapt [Tue, 20 Nov 2012 14:05:46 +0000 (14:05 +0000)]
In NIS mode first chmod(2) the temporary file  and is succeed then rename(2)

11 years agoonly rename(2) after chmod(2) has succeed
bapt [Tue, 20 Nov 2012 14:03:09 +0000 (14:03 +0000)]
only rename(2) after chmod(2) has succeed
report error if chmod(2) fails

Reported by: jh

11 years ago- Don't pass geom and provider names as format strings.
jh [Tue, 20 Nov 2012 12:32:18 +0000 (12:32 +0000)]
- Don't pass geom and provider names as format strings.
- Add __printflike() attributes.
- Remove an extra argument for the g_new_geomf() call in swapongeom_ev().

Reviewed by: pjd

11 years agoCorrectly set the password file mode after renaming in NIS mode
bapt [Tue, 20 Nov 2012 10:59:41 +0000 (10:59 +0000)]
Correctly set the password file mode after renaming in NIS mode

11 years agochange mode the group file to 0644 after a successfull rename(2)
bapt [Tue, 20 Nov 2012 07:22:07 +0000 (07:22 +0000)]
change mode the group file to 0644 after a successfull rename(2)

11 years agoDo not put "already running" message when rc_quiet=yes.
hrs [Tue, 20 Nov 2012 04:45:04 +0000 (04:45 +0000)]
Do not put "already running" message when rc_quiet=yes.

PR: bin/165477

11 years agoUnbreak amd64 cross-build where amd64 is the target. While clang
marcel [Tue, 20 Nov 2012 03:21:26 +0000 (03:21 +0000)]
Unbreak amd64 cross-build where amd64 is the target. While clang
may be installed as cc and we don't need to build gcc as a
cross-tools, we still build gcc and thus need cc_tools built
as a build tool. Not doing this results in building gengenrtl with
the target compiler while we need to run it on the build machine.

11 years agoRemove unneeded includes.
eadler [Tue, 20 Nov 2012 02:12:01 +0000 (02:12 +0000)]
Remove unneeded includes.

Tested with "make universe"; there are no conditional features.

Approved by: cperciva
MFC after: 3 daus

11 years agoAdd 'w' flag to:
eadler [Tue, 20 Nov 2012 01:57:21 +0000 (01:57 +0000)]
Add 'w' flag to:

Use whitespace (spaces and tabs) as the delimiter.
Consecutive spaces and tabs count as one single field
separator.

Reviewed by: swildner@dragonflybsd.org
Approved by: cperciva
Obtained from: DragonFlyBSD
MFC after: 1 week

11 years agoRemove unused variable.
emaste [Tue, 20 Nov 2012 01:42:18 +0000 (01:42 +0000)]
Remove unused variable.

11 years agoDon't allocate or program a key for the AR5210.
adrian [Mon, 19 Nov 2012 23:54:05 +0000 (23:54 +0000)]
Don't allocate or program a key for the AR5210.

The AR5210 doesn't support HAL_CIPHER_CLR ('clear encryption' keycache
slots), so don't bother - just map them to slot 0 and never program them.

11 years agoDisable WEP hardware encryption on the AR5210, in order to allow other
adrian [Mon, 19 Nov 2012 23:42:46 +0000 (23:42 +0000)]
Disable WEP hardware encryption on the AR5210, in order to allow other
encryption types.

The AR5210 only has four WEP key slots, in contrast to what the
later MACs have (ie, the keycache.)  So there's no way to store a "clear"
key.

Even if the driver is taught to not allocate CLR key entries for
the AR5210, the hardware will actually attempt to decode the encrypted
frames with the (likely all 0!) WEP keys.

So for now, disable the hardware encryption entirely and just so it
all in software.  That allows both WEP -and- WPA to actually work.

If someone wishes to try and make hardware WEP _but_ software WPA work,
they'll have to create a HAL capability to enable/disable hardware
encryption based on the current STA/Hostap mode. However, making
multi-vap work with one WEP and one WPA VAP will require hardware
encryption to be disabled anyway.

11 years agoUse '%zu' and '%zd' as appropriate for size_t / ssize_t.
emaste [Mon, 19 Nov 2012 23:07:38 +0000 (23:07 +0000)]
Use '%zu' and '%zd' as appropriate for size_t / ssize_t.

11 years agoClean up and update comments for CPUTYPE.
jkim [Mon, 19 Nov 2012 23:04:22 +0000 (23:04 +0000)]
Clean up and update comments for CPUTYPE.

Requested by: rdivacky

11 years agoZero the whole struct not just the size of a pointer.
emaste [Mon, 19 Nov 2012 22:56:51 +0000 (22:56 +0000)]
Zero the whole struct not just the size of a pointer.

Found by: clang

11 years agoUse '%zd' format specifier for ssize_t
emaste [Mon, 19 Nov 2012 22:53:57 +0000 (22:53 +0000)]
Use '%zd' format specifier for ssize_t

Found by: clang

11 years agoUse '%zd' printf format for ssize_t.
emaste [Mon, 19 Nov 2012 22:46:17 +0000 (22:46 +0000)]
Use '%zd' printf format for ssize_t.

11 years agor16312 is not any longer real since many years (likely since when VFS
attilio [Mon, 19 Nov 2012 22:43:45 +0000 (22:43 +0000)]
r16312 is not any longer real since many years (likely since when VFS
received granular locking) but the comment present in UFS has been
copied all over other filesystems code incorrectly for several times.

Removes comments that makes no sense now.

Reviewed by: kib
MFC after: 3 days

11 years agoAdd x86 CPUs supported by clang on head.
jkim [Mon, 19 Nov 2012 21:58:14 +0000 (21:58 +0000)]
Add x86 CPUs supported by clang on head.

Reviewed by: arch (silence)
X-MFC: r242624

11 years agoinsmntque() is always called with the lock held in exclusive mode,
attilio [Mon, 19 Nov 2012 20:43:19 +0000 (20:43 +0000)]
insmntque() is always called with the lock held in exclusive mode,
then:
- assume the lock is held in exclusive mode and remove a moot check
  about the lock acquisition.
- in the destructor remove !MPSAFE specific chunk.

Reviewed by: kib
MFC after: 2 weeks

11 years agoFix build after r243245.
marius [Mon, 19 Nov 2012 19:31:54 +0000 (19:31 +0000)]
Fix build after r243245.

Submitted by: trasz

11 years agoCleanup the code a bit, which improves the portability.
tuexen [Mon, 19 Nov 2012 19:26:19 +0000 (19:26 +0000)]
Cleanup the code a bit, which improves the portability.
MFC after: 1 week

11 years agoFix the handling of mapped IPv6 addresses in sctp_connectx().
tuexen [Mon, 19 Nov 2012 19:19:04 +0000 (19:19 +0000)]
Fix the handling of mapped IPv6 addresses in sctp_connectx().

MFC after: 3 days

11 years agoUse .Nm instead of a self xref
eadler [Mon, 19 Nov 2012 15:12:44 +0000 (15:12 +0000)]
Use .Nm instead of a self xref

Approved by: bcr (mentor)
MFC after: 1 week

11 years agoassert_vop_locked should treat LK_EXCLOTHER as the not locked case
avg [Mon, 19 Nov 2012 11:35:56 +0000 (11:35 +0000)]
assert_vop_locked should treat LK_EXCLOTHER as the not locked case

... from a perspective of the current thread.

Spotted by: mjg
Discussed with: kib
MFC after: 18 days

11 years agovnode_if: fix locking protocol description for lookup and cachedlookup
avg [Mon, 19 Nov 2012 11:32:56 +0000 (11:32 +0000)]
vnode_if: fix locking protocol description for lookup and cachedlookup

Also remove the checks from vop_lookup_pre and vop_lookup_post, which
are now completely redundant (before this change they were partially
redundant).

Discussed with: kib
MFC after: 10 days

11 years agozfs_remove: assert that delete_now case is never true on FreeBSD
avg [Mon, 19 Nov 2012 11:30:08 +0000 (11:30 +0000)]
zfs_remove: assert that delete_now case is never true on FreeBSD

That case is specific to Solaris VFS and it would violate pretty
fundamental contracts of FreeBSD VFS.

Discussed with: pjd
MFC after: 12 days

11 years agozfs_remove: set VV_NOSYNC flag if a node is unlinked
avg [Mon, 19 Nov 2012 11:25:20 +0000 (11:25 +0000)]
zfs_remove: set VV_NOSYNC flag if a node is unlinked

Suggested by: kib
MFC after: 12 days

11 years agoReturn port numbers for ATCA-7220 SPI interfaces in a different place for
jmallett [Mon, 19 Nov 2012 08:35:58 +0000 (08:35 +0000)]
Return port numbers for ATCA-7220 SPI interfaces in a different place for
consistency reasons, and to ensure that CRC addition is disabled on output.
With this, transmit seems to be working properly on the ATCA-7220.

11 years agoPrevent hang on ATCA-7220 when transmitting packets < 60 bytes.
jmallett [Mon, 19 Nov 2012 08:30:29 +0000 (08:30 +0000)]
Prevent hang on ATCA-7220 when transmitting packets < 60 bytes.

11 years agoRemove redundant printf of SDK version which already appears earlier in boot.
jmallett [Mon, 19 Nov 2012 08:29:53 +0000 (08:29 +0000)]
Remove redundant printf of SDK version which already appears earlier in boot.

11 years agoUse the correct size when allocating the cmdbuf string.
rpaulo [Mon, 19 Nov 2012 08:03:40 +0000 (08:03 +0000)]
Use the correct size when allocating the cmdbuf string.

cmdlengthdelta is the size of the header and we were using it to
allocate a buffer to store the command line. This would mean that
the cmdbuf could be too short. In practice this was never noticed unless
you usually run top -a. On a stock FreeBSD system you can see the
problem by running sendmail and then running top -a on a big terminal
window. In practice this doubles to size available to cmdbuf since the
header is around 65-68 bytes.

Reviewed by: adrian

11 years agoFix packet receive on the ATCA-7220 by disabling FCS-related checks, since the
jmallett [Mon, 19 Nov 2012 05:24:33 +0000 (05:24 +0000)]
Fix packet receive on the ATCA-7220 by disabling FCS-related checks, since the
FCS is stripped by the underlying hardware before it reaches the Octeon.

11 years agoAdd basic support for the Radisys-specific PCI console mechanism found on the
jmallett [Mon, 19 Nov 2012 01:58:20 +0000 (01:58 +0000)]
Add basic support for the Radisys-specific PCI console mechanism found on the
Radisys ATCA-7220.

11 years agoFix build on powerpc.
trasz [Mon, 19 Nov 2012 01:15:32 +0000 (01:15 +0000)]
Fix build on powerpc.

Reviewed by: nwhitehorn

11 years agoo) Do boot descriptor parsing before console setup so that we can use a console
jmallett [Mon, 19 Nov 2012 00:19:27 +0000 (00:19 +0000)]
o) Do boot descriptor parsing before console setup so that we can use a console
   other than UART 0 from the outset.
o) Print board information from sysinfo after consoles have been initialized
   rather than doing it during boot descriptor parsing.
o) Use cvmx_safe_printf and platform_reset rather than panic when doing very
   early boot descriptor parsing before the console is set up.
o) Get rid of the global octeon_bootinfo.

11 years agosh: Apply rlimits to parser/alias10.0 so it fails fast.
jilles [Sun, 18 Nov 2012 23:15:22 +0000 (23:15 +0000)]
sh: Apply rlimits to parser/alias10.0 so it fails fast.

Requested by: uqs

11 years agoImprove grammar.
bjk [Sun, 18 Nov 2012 22:52:17 +0000 (22:52 +0000)]
Improve grammar.

Approved by: hrs (mentor)

11 years agoFix build of kdump(1).
trasz [Sun, 18 Nov 2012 22:03:31 +0000 (22:03 +0000)]
Fix build of kdump(1).

11 years agoRemove this include, it isn't needed.
adrian [Sun, 18 Nov 2012 20:41:46 +0000 (20:41 +0000)]
Remove this include, it isn't needed.

11 years agoRemove superfluous paragraph macro.
joel [Sun, 18 Nov 2012 19:39:42 +0000 (19:39 +0000)]
Remove superfluous paragraph macro.

11 years agoAdd change missed in 243245.
trasz [Sun, 18 Nov 2012 19:16:10 +0000 (19:16 +0000)]
Add change missed in 243245.

11 years agoMake it possible to resize filesystems mounted read-write, using newly
trasz [Sun, 18 Nov 2012 19:01:00 +0000 (19:01 +0000)]
Make it possible to resize filesystems mounted read-write, using newly
introduced UFS write suspension mechanism.

Reviewed by: kib, mckusick
Sponsored by: FreeBSD Foundation

11 years agoAdd UFS writesuspension mechanism, designed to allow userland processes
trasz [Sun, 18 Nov 2012 18:57:19 +0000 (18:57 +0000)]
Add UFS writesuspension mechanism, designed to allow userland processes
to modify on-disk metadata for filesystems mounted for write.

Reviewed by: kib, mckusick
Sponsored by: FreeBSD Foundation

11 years agoMinor mdoc fix.
joel [Sun, 18 Nov 2012 16:58:08 +0000 (16:58 +0000)]
Minor mdoc fix.

11 years agoModernize parts of the ports.7 manual page.
eadler [Sun, 18 Nov 2012 16:34:09 +0000 (16:34 +0000)]
Modernize parts of the ports.7 manual page.

Approved by: bcr (mentor)
MFC after: 3 days

11 years agoUse the macro for standard error return values.
eadler [Sun, 18 Nov 2012 16:34:06 +0000 (16:34 +0000)]
Use the macro for standard error return values.

.Dd not bumped because there is no important content change.

Approved by: bcr (mentor)
MFC after: 3 days
Obtained from: DragonflyBSD

11 years agoMake it clear that amin and friends take + and - options.
eadler [Sun, 18 Nov 2012 16:34:03 +0000 (16:34 +0000)]
Make it clear that amin and friends take + and - options.

PR: docs/173265
Submitted by: Anton Shterenlikht <mexas@bristol.ac.uk>
Approved by: bcr (mentor)
MFC after: 3 days

11 years agoStandardize EXIT STATUS instructions in man pages when possible.
eadler [Sun, 18 Nov 2012 16:33:51 +0000 (16:33 +0000)]
Standardize EXIT STATUS instructions in man pages when possible.

Approved by: bcr (mentor)
MFC after: 3 days

11 years agoCheck if an extracted zoneid is equal to the non-zero sin6_scope_id only when
hrs [Sun, 18 Nov 2012 16:06:51 +0000 (16:06 +0000)]
Check if an extracted zoneid is equal to the non-zero sin6_scope_id only when
it is link-local or MC interface-local.

11 years agoFix possible fp reference leak in posix_openpt
mjg [Sun, 18 Nov 2012 15:48:34 +0000 (15:48 +0000)]
Fix possible fp reference leak in posix_openpt

Reviewed by: ed
Approved by: trasz (mentor)
MFC after: 3 days

11 years ago- Increase the number of retry for NET_RT_DUMP from 5 to 15.
hrs [Sun, 18 Nov 2012 15:48:02 +0000 (15:48 +0000)]
- Increase the number of retry for NET_RT_DUMP from 5 to 15.
- Use 2001:db8:: as an example instead of deprecated 3ffe:: address block.
- Add check for connected routes.
- Add support of RTM_IFANNOUNCE for dyanmically-added/removed interfaces.
- Add support of *, ?, and [ in the interface list.
- Add -P number to specify route flag which will never expire.
- Add -Q number to specify route flag which route6d will add to routes via RIP.
- Add -p pidfile to specify the process ID file.

11 years agoOverhaul of route6d(8):
hrs [Sun, 18 Nov 2012 15:37:27 +0000 (15:37 +0000)]
Overhaul of route6d(8):

- Use queue(3) for linked-list.
- Use a consistent naming scheme for struct members.
- Use ANSI C style function declaration.
- Add check of RTM_VERSION mismatch.

There is no functional change.

11 years agoUse sin6_scope_id instead of KAME-specific embedded scope id.
hrs [Sun, 18 Nov 2012 15:11:47 +0000 (15:11 +0000)]
Use sin6_scope_id instead of KAME-specific embedded scope id.

11 years agocp -R misses out dotfiles; use pax instead to copy file hierarchies
crees [Sun, 18 Nov 2012 14:21:05 +0000 (14:21 +0000)]
cp -R misses out dotfiles; use pax instead to copy file hierarchies

PR: conf/99721 (based on)
Submitted by: Florian Zavatzki <f_zavatzki@blue-network.org>
Approved by: hrs
MFC after: 1 month

11 years agoCorrectly use spaces here.
adrian [Sun, 18 Nov 2012 14:05:28 +0000 (14:05 +0000)]
Correctly use spaces here.

Pointed out by: pjd

11 years agospa_import_rootpool: fall back to use configuration from zpool.cache...
avg [Sun, 18 Nov 2012 11:47:25 +0000 (11:47 +0000)]
spa_import_rootpool: fall back to use configuration from zpool.cache...

if we fail to generate a proper root pool configuration based on disk
probing.  Currently we can not properly generate the configuration for
multi-vdev pools.  Make that explicit.

Reported by: madpilot, Bartosz Stec <bartosz.stec@it4pro.pl>
Tested by: madpilot, Bartosz Stec <bartosz.stec@it4pro.pl>
MFC after: 4 days

11 years agoFix condition to check if the maximum number of FIBs is greater than 0 or not.
hrs [Sun, 18 Nov 2012 11:22:15 +0000 (11:22 +0000)]
Fix condition to check if the maximum number of FIBs is greater than 0 or not.

Spotted by: zont

11 years agoFix up a compile time warning if INET6 isn't defined.
adrian [Sun, 18 Nov 2012 04:51:46 +0000 (04:51 +0000)]
Fix up a compile time warning if INET6 isn't defined.

11 years agoRemove one wholly-unused and buggy routine and some nearby alternative symbols.
jmallett [Sat, 17 Nov 2012 23:53:12 +0000 (23:53 +0000)]
Remove one wholly-unused and buggy routine and some nearby alternative symbols.

While here, also correct a comment that seems to imply that this file is
NetBSD's all-singing, all-dancing locore.S, rather than our conservative set of
assembly support routines.

11 years agoSort options.
grog [Sat, 17 Nov 2012 23:52:38 +0000 (23:52 +0000)]
Sort options.
Add comment on standards conformity.

MFC after: 14 days

11 years agoCorrectly handle keywords without options.
grog [Sat, 17 Nov 2012 23:49:20 +0000 (23:49 +0000)]
Correctly handle keywords without options.

Reported by: swills@
MFC after: 14 days

11 years agoClean up the two i80321 copies of initarm to be closer to one another.
andrew [Sat, 17 Nov 2012 23:06:00 +0000 (23:06 +0000)]
Clean up the two i80321 copies of initarm to be closer to one another.

11 years agoJust compile the whole ath chipset support in.
adrian [Sat, 17 Nov 2012 21:55:49 +0000 (21:55 +0000)]
Just compile the whole ath chipset support in.

11 years agoUse -fib N modifier to add/delete a route to/from multiple FIBs.
hrs [Sat, 17 Nov 2012 21:44:02 +0000 (21:44 +0000)]
Use -fib N modifier to add/delete a route to/from multiple FIBs.

11 years agoFill sin6_scope_id in sockaddr_in6 before passing it from the kernel to
hrs [Sat, 17 Nov 2012 20:19:00 +0000 (20:19 +0000)]
Fill sin6_scope_id in sockaddr_in6 before passing it from the kernel to
userland via routing socket or sysctl.  This eliminates the following
KAME-specific sin6_scope_id handling routine from each userland utility:

 sin6.sin6_scope_id = ntohs(*(u_int16_t *)&sin6.sin6_addr.s6_addr[2]);

This behavior can be controlled by net.inet6.ip6.deembed_scopeid.  This is
set to 1 by default (sin6_scope_id will be filled in the kernel).

Reviewed by: bz

11 years agoAdd support for SCTP/UDP/IPV6.
tuexen [Sat, 17 Nov 2012 20:04:04 +0000 (20:04 +0000)]
Add support for SCTP/UDP/IPV6.
This completes the support of
http://tools.ietf.org/html/draft-ietf-tsvwg-sctp-udp-encaps

MFC after: 1 week

11 years agoAdd -fib modifier to specify FIB number. The FIB number can be in a
hrs [Sat, 17 Nov 2012 19:54:23 +0000 (19:54 +0000)]
Add -fib modifier to specify FIB number.  The FIB number can be in a
comma-separated list and/or range specification:

 # route add -inet 192.0.2.0/24 198.51.100.1 -fib 1,3-5,6

Although all of the subcommands supports the modifier, "monitor" does not
support the list or range specification at this moment.

Reviewed by: bz

11 years agoAdd check_namevarlist() to check if ${name}_var is reserved in rc.subr or not.
hrs [Sat, 17 Nov 2012 18:11:10 +0000 (18:11 +0000)]
Add check_namevarlist() to check if ${name}_var is reserved in rc.subr or not.

11 years agoFix whitespace.
ed [Sat, 17 Nov 2012 16:47:05 +0000 (16:47 +0000)]
Fix whitespace.

MFC after: 1 week

11 years agoSummarize information about connected analog pins and EDID-like data received
mav [Sat, 17 Nov 2012 06:33:13 +0000 (06:33 +0000)]
Summarize information about connected analog pins and EDID-like data received
from HDMI/DisplayPort devices in form of general connection status and
sound(4)-style channel matrix.  Now that information is only reported in
readable form to verbose logs, but potentially could be used by sound(4)
to correctly choose default devices and configure vchans.

Fix rear and side channels swap on analog 7.1 outputs.  As soon as there is
a huge mess in industry about naming and using of these channels, duplicate
rear channels of 4 and 5.1 streams to both read and side speakers.

11 years agoEnsure hwpmc support is correctly included.
adrian [Sat, 17 Nov 2012 04:11:57 +0000 (04:11 +0000)]
Ensure hwpmc support is correctly included.

11 years agoMake MIPS24k PMC optional on "hwpmc_mips24k."
adrian [Sat, 17 Nov 2012 04:10:42 +0000 (04:10 +0000)]
Make MIPS24k PMC optional on "hwpmc_mips24k."

Requested by: juli

11 years agomake the MIPS24k HWPMC support require hwpmc_mips24k as well as hwpmc.
adrian [Sat, 17 Nov 2012 04:09:28 +0000 (04:09 +0000)]
make the MIPS24k HWPMC support require hwpmc_mips24k as well as hwpmc.

This way non-MIPS24k platforms (eg Octeon) can have separate PMC support.

Requested by: juli

11 years agoMigrate the AR71xx UART (an 8250 derivative) to hide behind uart_ar71xx.
adrian [Sat, 17 Nov 2012 04:05:46 +0000 (04:05 +0000)]
Migrate the AR71xx UART (an 8250 derivative) to hide behind uart_ar71xx.

The AR9330/AR9331 UART is a totally different thing, so having it included
with 'uart' is not going to work out.

11 years agoUpdate a comment to reflect the elimination of the hold queue in r242300.
alc [Sat, 17 Nov 2012 04:00:19 +0000 (04:00 +0000)]
Update a comment to reflect the elimination of the hold queue in r242300.

11 years agoCorrectly populate the RTS field.
adrian [Sat, 17 Nov 2012 02:39:37 +0000 (02:39 +0000)]
Correctly populate the RTS field.

Tested:
* AR5210, STA mode, RTS enabled

11 years ago* Remove ah_desc.h, it's not needed
adrian [Sat, 17 Nov 2012 02:39:09 +0000 (02:39 +0000)]
* Remove ah_desc.h, it's not needed
* Add some shifts that I'm using in userspace (athalq.)

However, this exposes a fun little bug..

11 years agoMake this a flag.
adrian [Sat, 17 Nov 2012 02:38:36 +0000 (02:38 +0000)]
Make this a flag.

11 years agoImplement AR5210 descriptor decoding.
adrian [Sat, 17 Nov 2012 02:37:25 +0000 (02:37 +0000)]
Implement AR5210 descriptor decoding.

11 years agoAdd AR5211 descriptor decoding support.
adrian [Sat, 17 Nov 2012 02:14:50 +0000 (02:14 +0000)]
Add AR5211 descriptor decoding support.

11 years ago.. include ah_desc.h here now.
adrian [Sat, 17 Nov 2012 02:02:36 +0000 (02:02 +0000)]
.. include ah_desc.h here now.

11 years agoRemove the ah_desc.h reference; it's not needed.
adrian [Sat, 17 Nov 2012 02:00:33 +0000 (02:00 +0000)]
Remove the ah_desc.h reference; it's not needed.

I'm using these descriptor header files in userland and I'm trying to
avoid populating a compatibility ah_desc.h file.

11 years agoAdd AR5212 frame decoding support.
adrian [Sat, 17 Nov 2012 01:57:45 +0000 (01:57 +0000)]
Add AR5212 frame decoding support.