]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
11 years agoAdd flex 2.5.37 from flex.sourceforge.net to contrib.
Jung-uk Kim [Tue, 21 May 2013 19:09:30 +0000 (19:09 +0000)]
Add flex 2.5.37 from flex.sourceforge.net to contrib.

11 years agoRemove an extra semicolon from the DOT language output.
Jaakko Heinonen [Tue, 21 May 2013 18:40:54 +0000 (18:40 +0000)]
Remove an extra semicolon from the DOT language output.

PR: kern/178540
Submitted by: Trond Endrestol
MFC after: 1 week

11 years agoMention the switch to bmake by default.
Simon J. Gerraty [Tue, 21 May 2013 18:38:09 +0000 (18:38 +0000)]
Mention the switch to bmake by default.

Reviewed by: obrien

11 years agoImplement a separate hardware queue threshold for aggregate and non-aggr
Adrian Chadd [Tue, 21 May 2013 18:13:57 +0000 (18:13 +0000)]
Implement a separate hardware queue threshold for aggregate and non-aggr
traffic.

When transmitting non-aggregate traffic, we need to keep the hardware
busy whilst transmitting or small bursts in txdone/tx latency will
kill us.

This restores non-aggregate iperf performance, especially when doing
TDMA.

Tested:

* AR5416<->AR5416, TDMA
* AR5416 STA <-> AR9280 AP

11 years agoEnable the use of TDMA on an 802.11n channel (with aggregation disabled,
Adrian Chadd [Tue, 21 May 2013 18:02:54 +0000 (18:02 +0000)]
Enable the use of TDMA on an 802.11n channel (with aggregation disabled,
of course.)

There's a few things that needed to happen:

* In case someone decides to set the beacon transmission rate to be
  at an MCS rate, use the MCS-aware version of the duration calculation
  to figure out how long the received beacon frame was.

* If TxOP enforcing is available on the hardware and we're doing TDMA,
  enable it after a reset and set the TDMA guard interval to zero.
  This seems to behave fine.

TODO:

* Although I haven't yet seen packet loss, the PHY errors that would be
  triggered (specifically Transmit-Override-Receive) aren't enabled
  by the 11n HAL.  I'll have to do some work to enable these PHY errors
  for debugging.

What broke:

* My recent changes to the TX queue handling has resulted in the driver
  not keeping the hardware queue properly filled when doing non-aggregate
  traffic.  I have a patch to commit soon which fixes this situation
  (albeit by reminding me about how my ath driver locking isn't working
  out, sigh.)

  So if you want to test this without updating to the next set of patches
  that I commit, just bump the sysctl dev.ath.X.hwq_limit from 2 to 32.

Tested:

* AR5416 <-> AR5416, with ampdu disabled, HT40, 5GHz, MCS12+Short-GI.
  I saw 30mbit/sec in both directions using a bidirectional UDP test.

11 years agoFix the PowerPC Book-E register definitions used by the remote GDB
Marcel Moolenaar [Tue, 21 May 2013 18:00:47 +0000 (18:00 +0000)]
Fix the PowerPC Book-E register definitions used by the remote GDB
protocol.

Obtained from: Juniper Networks, Inc.

11 years agoAdd a makefle that recurses into the right architecture-specific
Marcel Moolenaar [Tue, 21 May 2013 17:47:53 +0000 (17:47 +0000)]
Add a makefle that recurses into the right architecture-specific
sub-directory. This to allow simpler logic outside of the csu
directory.

Obtained from: Juniper Networks, Inc.

11 years agoDon't look for headers outside of the source or object directories. In
Marcel Moolenaar [Tue, 21 May 2013 17:30:27 +0000 (17:30 +0000)]
Don't look for headers outside of the source or object directories. In
particular, don't use DESTDIR. Such creates an unnecessary dependency
on the build machine.

Obtained from:  Juniper Networks, Inc.

11 years agoDon't look for headers outside of the source or object directories. In
Marcel Moolenaar [Tue, 21 May 2013 17:28:19 +0000 (17:28 +0000)]
Don't look for headers outside of the source or object directories. In
particular, don't use DESTDIR. Such creates an unnecessary dependency
on the build machine.

Obtained from: Juniper Networks, Inc.

11 years agoRestore the ability to build on systems with 32-bit compat when
Brooks Davis [Tue, 21 May 2013 16:58:23 +0000 (16:58 +0000)]
Restore the ability to build on systems with 32-bit compat when
the system compiler is not clang.  clang and gcc appear to differ
signficantly in their interpretation of -isystem and --sysroot.  Further
work is likely required to support an external gcc.

Reported by: andreast, fidaj@ukr.net, sergey.dyatko@gmail.com

11 years agoFix build break - the SetCapability calls return HAL_BOOL,
Adrian Chadd [Tue, 21 May 2013 14:28:05 +0000 (14:28 +0000)]
Fix build break - the SetCapability calls return HAL_BOOL,
not HAL_STATUS.

11 years agoRegenerate.
Konstantin Belousov [Tue, 21 May 2013 11:41:08 +0000 (11:41 +0000)]
Regenerate.

11 years agoFix the wait6(2) on 32bit architectures and for the compat32, by using
Konstantin Belousov [Tue, 21 May 2013 11:40:16 +0000 (11:40 +0000)]
Fix the wait6(2) on 32bit architectures and for the compat32, by using
the right type for the argument in syscalls.master.  Also fix the
posix_fallocate(2) and posix_fadvise(2) compat32 syscalls on the
architectures which require padding of the 64bit argument.

Noted and reviewed by: jhb
Pointy hat to: kib
MFC after: 1 week

11 years agoDo not leak the NULLV_NOUNLOCK flag from the nullfs_unlink_lowervp(),
Konstantin Belousov [Tue, 21 May 2013 11:31:56 +0000 (11:31 +0000)]
Do not leak the NULLV_NOUNLOCK flag from the nullfs_unlink_lowervp(),
for the case when the nullfs vnode is not reclaimed.  Otherwise, later
reclamation would not unlock the lower vnode.

Reported by: antoine
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week

11 years agoFix the hardware watchpoints on SMP amd64. Load the updated %dr
Konstantin Belousov [Tue, 21 May 2013 11:24:32 +0000 (11:24 +0000)]
Fix the hardware watchpoints on SMP amd64.  Load the updated %dr
registers also on other CPUs, besides the CPU which happens to execute
the ddb.  The debugging registers are stored in the pcpu area,
together with the command which is executed by the IPI stop handler
upon resume.

Reviewed by: jhb
Sponsored by: The FreeBSD Foundation
MFC after: 1 week

11 years agoAdd amd64-specific ddb command 'show phys2dmap', which calculates the
Konstantin Belousov [Tue, 21 May 2013 11:07:12 +0000 (11:07 +0000)]
Add amd64-specific ddb command 'show phys2dmap', which calculates the
address in the direct map corresponding to the given physical address.

Reviewed by: jhb
Sponsored by: The FreeBSD Foundation
MFC after: 1 week

11 years agoAdd ddb command 'show pginfo' which provides useful information about
Konstantin Belousov [Tue, 21 May 2013 11:04:00 +0000 (11:04 +0000)]
Add ddb command 'show pginfo' which provides useful information about
a vm page, denoted either by an address of the struct vm_page, or, if
the '/p' modifier is specified, by a physical address of the
corresponding frame.

Reviewed by: jhb
Sponsored by: The FreeBSD Foundation
MFC after: 1 week

11 years agoAdd new USB quirk.
Hans Petter Selasky [Tue, 21 May 2013 06:13:38 +0000 (06:13 +0000)]
Add new USB quirk.

MFC after: 1 week
PR: usb/178771

11 years agoExtend the TXOP enforce capability to support checking whether it's
Adrian Chadd [Tue, 21 May 2013 05:51:49 +0000 (05:51 +0000)]
Extend the TXOP enforce capability to support checking whether it's
supported.

11 years agoAdd basic support for FDT to i386 & amd64. This change includes:
Marcel Moolenaar [Tue, 21 May 2013 03:05:49 +0000 (03:05 +0000)]
Add basic support for FDT to i386 & amd64. This change includes:
1.  Common headers for fdt.h and ofw_machdep.h under x86/include
    with indirections under i386/include and amd64/include.
2.  New modinfo for loader provided FDT blob.
3.  Common x86_init_fdt() called from hammer_time() on amd64 and
    init386() on i386.
4.  Split-off FDT specific low-level console functions from FDT
    bus methods for the uart(4) driver. The low-level console
    logic has been moved to uart_cpu_fdt.c and is used for arm,
    mips & powerpc only. The FDT bus methods are shared across
    all architectures.
5.  Add dev/fdt/fdt_x86.c to hold the fdt_fixup_table[] and the
    fdt_pic_table[] arrays. Both are empty right now.

FDT addresses are I/O ports on x86. Since the core FDT code does
not handle different address spaces, adding support for both I/O
ports and memory addresses requires some thought and discussion.
It may be better to use a compile-time option that controls this.

Obtained from: Juniper Networks, Inc.

11 years agoBMAKE is now being built by default.
Xin LI [Tue, 21 May 2013 00:41:49 +0000 (00:41 +0000)]
BMAKE is now being built by default.

11 years agoMerge ACPICA 20130517.
Jung-uk Kim [Mon, 20 May 2013 23:52:49 +0000 (23:52 +0000)]
Merge ACPICA 20130517.

11 years agoSync with upstream version (20130520) that includes same fixes made last week.
Simon J. Gerraty [Mon, 20 May 2013 22:51:11 +0000 (22:51 +0000)]
Sync with upstream version (20130520) that includes same fixes made last week.

11 years agoImport bmake-20130520 to keep us in sync.
Simon J. Gerraty [Mon, 20 May 2013 22:35:32 +0000 (22:35 +0000)]
Import bmake-20130520 to keep us in sync.

11 years ago- Prefer ACPI_COMPARE_NAME(a, b) macro over strncmp(a, b, ACPI_NAME_SIZE).
Jung-uk Kim [Mon, 20 May 2013 22:18:18 +0000 (22:18 +0000)]
- Prefer ACPI_COMPARE_NAME(a, b) macro over strncmp(a, b, ACPI_NAME_SIZE).
- Make sure the predefined name is a string type.
- Return slightly more useful errors.

11 years agoFix white spaces.
Jung-uk Kim [Mon, 20 May 2013 22:10:01 +0000 (22:10 +0000)]
Fix white spaces.

11 years agoFix distributekernel in the non NO_ROOT case.
Brooks Davis [Mon, 20 May 2013 21:16:38 +0000 (21:16 +0000)]
Fix distributekernel in the non NO_ROOT case.

PR: conf/178775
Submitted by: Garrett Cooper <yaneurabeya@gmail.com>

11 years agoUpdate the setfib man page to reflect recent changes.
Julian Elischer [Mon, 20 May 2013 20:47:40 +0000 (20:47 +0000)]
Update the setfib man page to reflect recent changes.

11 years agoCleanups and markup improvements suggested by trociny and joel.
Brooks Davis [Mon, 20 May 2013 20:32:58 +0000 (20:32 +0000)]
Cleanups and markup improvements suggested by trociny and joel.

11 years agoRemove accidentally included line..
Julian Elischer [Mon, 20 May 2013 20:14:12 +0000 (20:14 +0000)]
Remove accidentally included line..
No effect as it wasn't used but was a) wrong   b) clutter

11 years agoSpell extensions correctly.
Brooks Davis [Mon, 20 May 2013 19:41:34 +0000 (19:41 +0000)]
Spell extensions correctly.

Submitted by: dim

11 years agopopen(): Add 'e' mode character to set close-on-exec on the new fd.
Jilles Tjoelker [Mon, 20 May 2013 17:31:18 +0000 (17:31 +0000)]
popen(): Add 'e' mode character to set close-on-exec on the new fd.

If 'e' is used, the kernel must support the recently added pipe2() system
call.

The use of pipe2() with O_CLOEXEC also fixes race conditions between
concurrent popen() calls from different threads, even if the close-on-exec
flag on the fd of the returned FILE is later cleared (because popen() closes
all file descriptors from earlier popen() calls in the child process).
Therefore, this approach should be used in all cases when pipe2() can be
assumed present.

The old version of popen() rejects "re" and "we" but treats "r+e" like "r+".

11 years agoNetBSD 6.1 added.
Maxim Konovalov [Mon, 20 May 2013 15:11:30 +0000 (15:11 +0000)]
NetBSD 6.1 added.

11 years agoAdd a test program for popen().
Jilles Tjoelker [Mon, 20 May 2013 13:05:51 +0000 (13:05 +0000)]
Add a test program for popen().

11 years agoMake the HT rate duration calculation work for MCS rates > 15.
Adrian Chadd [Mon, 20 May 2013 07:10:43 +0000 (07:10 +0000)]
Make the HT rate duration calculation work for MCS rates > 15.

11 years agogrep: change some int types.
Pedro F. Giffuni [Mon, 20 May 2013 03:15:25 +0000 (03:15 +0000)]
grep: change some int types.

Change several int variables to size_t, ssize_t, or ptrdiff_t.
This should fix the bug described in CVE-2012-5667 when an input
line is so long that its length cannot be stored in an int
variable.

This is based on NetBSD's revision which says:

This change to NetBSD's version of GNU grep 2.5.1 (licenced under
GPLv2) was made without direct reference to any code licenced
under GPLv3.

Obtained from: NetBSD
MFC after: 3 days

11 years agoRemove a block of code that was not intended to be part of the previous
Mark Johnston [Mon, 20 May 2013 02:10:18 +0000 (02:10 +0000)]
Remove a block of code that was not intended to be part of the previous
revision.

X-MFC with: r250812

11 years agoFix vdc->Secondary_Element_Count metadata field access from 16 to 8 bit.
Alexander Motin [Mon, 20 May 2013 00:33:54 +0000 (00:33 +0000)]
Fix vdc->Secondary_Element_Count metadata field access from 16 to 8 bit.
In some cases it could cause kernel panic during failed drive replacement.

Reported by: trasz
MFC after: 1 week

11 years agoStyle nits.
Pawel Jakub Dawidek [Sun, 19 May 2013 23:30:24 +0000 (23:30 +0000)]
Style nits.

11 years agoUse SDT_PROBE1() instead of SDT_PROBE().
Pawel Jakub Dawidek [Sun, 19 May 2013 23:29:22 +0000 (23:29 +0000)]
Use SDT_PROBE1() instead of SDT_PROBE().

11 years agoProtect SDT_PROBE() with do { } while (0) loop.
Pawel Jakub Dawidek [Sun, 19 May 2013 23:28:28 +0000 (23:28 +0000)]
Protect SDT_PROBE() with do { } while (0) loop.

11 years agoReally fix netmask address family this time.
Alexander V. Chernikov [Sun, 19 May 2013 19:42:46 +0000 (19:42 +0000)]
Really fix netmask address family this time.

MFC with: r250813

11 years agoFinish r85740 : Make IPv6 netmask has address family set.
Alexander V. Chernikov [Sun, 19 May 2013 19:19:01 +0000 (19:19 +0000)]
Finish r85740 : Make IPv6 netmask has address family set.
This pleases routing daemons like bird.

MFC after: 2 weeks

11 years agoRe-introduce another part of r249367. This commit fixes a register leak in
Mark Johnston [Sun, 19 May 2013 17:14:36 +0000 (17:14 +0000)]
Re-introduce another part of r249367. This commit fixes a register leak in
dt_cg_ptrsize() and generally cleans up some of the error handling around
register allocation.

This change corresponds to part of illumos-gate commit e5803b76927480:
  3025 register leak in D code generation

Reviewed by: pfg
Obtained from: illumos
MFC after: 1 month

11 years agoAdd a comment explaining why stack_capture is empty for EABI and clang.
Andrew Turner [Sun, 19 May 2013 16:25:09 +0000 (16:25 +0000)]
Add a comment explaining why stack_capture is empty for EABI and clang.

While here add a comment pointing out that, while r11 is not the frame
pointer on EABI as there is no frame pointer, it's value is unused so is
safe.

11 years agoInitialize the fibnum for outgoing packets to 0. This avoids
Michael Tuexen [Sun, 19 May 2013 16:06:43 +0000 (16:06 +0000)]
Initialize the fibnum for outgoing packets to 0. This avoids
crashing due to the usage of uninitialized fibnum.
This bugs became visiable after
http://svnweb.freebsd.org/changeset/base/250700

MFC after: 2 weeks

11 years agoRemove lint case for _Thread_local.
Ed Schouten [Sun, 19 May 2013 07:44:01 +0000 (07:44 +0000)]
Remove lint case for _Thread_local.

I added this block, knowing that lint does not support _Thread_local.
When linting, we could argue that we don't care about TLS (yet). It
seems, however, that external pieces of software also sometimes do a
-Dlint, regex the output and compile it again.

Reported by: swills

11 years agoFix build after doc/projects/xml-tools integration.
Hiroki Sato [Sun, 19 May 2013 05:49:21 +0000 (05:49 +0000)]
Fix build after doc/projects/xml-tools integration.

11 years agoRefine the "nojail" rc keyword, adding "nojailvnet" for files that don't
Jamie Gritton [Sun, 19 May 2013 04:10:34 +0000 (04:10 +0000)]
Refine the "nojail" rc keyword, adding "nojailvnet" for files that don't
apply to most jails but do apply to vnet jails.  This includes adding
a new sysctl "security.jail.vnet" to identify vnet jails.

PR: conf/149050
Submitted by: mdodd
MFC after: 3 days

11 years agoUpdate snd quirks for T520, T420, X220.
Sean Bruno [Sun, 19 May 2013 01:45:29 +0000 (01:45 +0000)]
Update snd quirks for T520, T420, X220.

Group onboard mic and headphone mic jack together.  Creates association that
will switch between microphone inputs depending on the state of the headphone
jack being connected to a live mic.

Fixes onboard mic not working at all on T520.

Tested on T520, T420.
Suspect X220 needs this too, untested on.

MFC after: 1 month

11 years agoMore non-ATH_DEBUG build fixes.
Adrian Chadd [Sun, 19 May 2013 01:33:17 +0000 (01:33 +0000)]
More non-ATH_DEBUG build fixes.

11 years agoSince we're now using the ah pointer, always declare it.
Adrian Chadd [Sun, 19 May 2013 00:53:06 +0000 (00:53 +0000)]
Since we're now using the ah pointer, always declare it.

This fixes non-DEBUG builds.

11 years agoAdded output of device QUIRKS for CAM and AHCI devices during boot.
Steven Hartland [Sat, 18 May 2013 23:36:21 +0000 (23:36 +0000)]
Added output of device QUIRKS for CAM and AHCI devices during boot.

Reviewed by: mav
Approved by: pjd (mentor)
MFC after: 2 weeks

11 years agoBack out r250768 until I can further investigate why it might
Tim Kientzle [Sat, 18 May 2013 22:42:21 +0000 (22:42 +0000)]
Back out r250768 until I can further investigate why it might
be causing problems with the BeagleBone Black boot.

11 years agoAdd support for the second GPIO pin bank on the Wii and add support for
Rui Paulo [Sat, 18 May 2013 19:30:42 +0000 (19:30 +0000)]
Add support for the second GPIO pin bank on the Wii and add support for
shutting down the system.

11 years agoUpdate location to ports INDEX file.
Bryan Drewery [Sat, 18 May 2013 19:18:03 +0000 (19:18 +0000)]
Update location to ports INDEX file.

PR: bin/164561
Approved by: bapt

11 years agoAdd another obsolete file: nfs fha moved
Antoine Brodin [Sat, 18 May 2013 18:28:19 +0000 (18:28 +0000)]
Add another obsolete file: nfs fha moved

11 years agoBe (very) careful about how to add more TX DMA work.
Adrian Chadd [Sat, 18 May 2013 18:27:53 +0000 (18:27 +0000)]
Be (very) careful about how to add more TX DMA work.

The list-based DMA engine has the following behaviour:

* When the DMA engine is in the init state, you can write the first
  descriptor address to the QCU TxDP register and it will work.

* Then when it hits the end of the list (ie, it either hits a NULL
  link pointer, OR it hits a descriptor with VEOL set) the QCU
  stops, and the TxDP points to the last descriptor that was transmitted.

* Then when you want to transmit a new frame, you can then either:
  + write the head of the new list into TxDP, or
  + you write the head of the new list into the link pointer of the
    last completed descriptor (ie, where TxDP points), then kick
    TxE to restart transmission on that QCU>

* The hardware then will re-read the descriptor to pick up the link
  pointer and then jump to that.

Now, the quirks:

* If you write a TxDP when there's been no previous TxDP (ie, it's 0),
  it works.

* If you write a TxDP in any other instance, the TxDP write may actually
  fail.  Thus, when you start transmission, it will re-read the last
  transmitted descriptor to get the link pointer, NOT just start a new
  transmission.

So the correct thing to do here is:

* ALWAYS use the holding descriptor (ie, the last transmitted descriptor
  that we've kept safe) and use the link pointer in _THAT_ to transmit
  the next frame.

* NEVER write to the TxDP after you've done the initial write.

* .. also, don't do this whilst you're also resetting the NIC.

With this in mind, the following patch does basically the above.

* Since this encapsulates Sam's issues with the QCU behaviour w/ TDMA,
  kill the TDMA special case and replace it with the above.

* Add a new TXQ flag - PUTRUNNING - which indicates that we've started
  DMA.

* Clear that flag when DMA has been shutdown.

* Ensure that we're not restarting DMA with PUTRUNNING enabled.

* Fix the link pointer logic during TXQ drain - we should always ensure
  the link pointer does point to something if there's a list of frames.
  Having it be NULL as an indication that DMA has finished or during
  a reset causes trouble.

Now, given all of this, i want to nuke axq_link from orbit.  There's now HAL
methods to get and set the link pointer of a descriptor, so what we
should do instead is to update the right link pointer.

* If there's a holding descriptor and an empty TXQ list, set the
  link pointer of said holding descriptor to the new frame.

* If there's a non-empty TXQ list, set the link pointer of the
  last descriptor in the list to the new frame.

* Nuke axq_link from orbit.

Note:

* The AR9380 doesn't need this.  FIFO TX writes are atomic.  As long as
  we don't append to a list of frames that we've already passed to the
  hardware, all of the above doesn't apply.  The holding descriptor stuff
  is still needed to ensure the hardware can re-read a completed
  descriptor to move onto the next one, but we restart DMA by pushing in
  a new FIFO entry into the TX QCU.  That doesn't require any real
  gymnastics.

Tested:

* AR5210, AR5211, AR5212, AR5416, AR9380 - STA mode.

11 years agoHave the ipropd-master listen on an IPv6 socket in addition to an IPv4
Bjoern A. Zeeb [Sat, 18 May 2013 18:01:21 +0000 (18:01 +0000)]
Have the ipropd-master listen on an IPv6 socket in addition to an IPv4
socket to allow propagation of changes to a Heimdal Kerberos database
from the KDC master to the slave(s) work on IPv6 as well.

Update the stats logging to also handle IPv6 addresses.

Reported by: peter (found on FreeBSD cluster)
X-to-be-tested-by: peter
MFC after: 3 weeks

11 years agoUse my @freebsd.org username, rather than ${WORK} username.
Adrian Chadd [Sat, 18 May 2013 14:05:32 +0000 (14:05 +0000)]
Use my @freebsd.org username, rather than ${WORK} username.

Pointy hat provided by: kib, thankyou!

11 years agoRe-add some code to exclude transmitting if we're in reset.
Adrian Chadd [Sat, 18 May 2013 13:58:07 +0000 (13:58 +0000)]
Re-add some code to exclude transmitting if we're in reset.

This fixes some "transmitting during reset" bugs that crept in after
I messed around with this part of the transmit path.

11 years agoUpdate maintainers with ath/net80211.
Adrian Chadd [Sat, 18 May 2013 13:56:26 +0000 (13:56 +0000)]
Update maintainers with ath/net80211.

11 years agoFix wildcard srcs for phony targets
Simon J. Gerraty [Sat, 18 May 2013 13:24:53 +0000 (13:24 +0000)]
Fix wildcard srcs for phony targets

11 years agoMatch what is in netbsd.
Simon J. Gerraty [Sat, 18 May 2013 13:15:19 +0000 (13:15 +0000)]
Match what is in netbsd.

11 years agoVar_Delete: expand name to delete if needed.
Simon J. Gerraty [Sat, 18 May 2013 13:07:01 +0000 (13:07 +0000)]
Var_Delete: expand name to delete if needed.

11 years agoCorrect the spelling of "okay".
Tim Kientzle [Sat, 18 May 2013 12:54:37 +0000 (12:54 +0000)]
Correct the spelling of "okay".
Add pinmux setting for the Reset GPIO pin for MMC1.

11 years agoLabel the mmc child after the parent.
Tim Kientzle [Sat, 18 May 2013 12:53:20 +0000 (12:53 +0000)]
Label the mmc child after the parent.

11 years agoUse separate function to update mbuf checksum flags instead of
Alexander V. Chernikov [Sat, 18 May 2013 08:14:21 +0000 (08:14 +0000)]
Use separate function to update mbuf checksum flags instead of
duplicating the same code in different places.

MFC after: 2 weeks

11 years agoDon't clear stall at first time use of USB MIDI endpoints.
Hans Petter Selasky [Sat, 18 May 2013 07:16:20 +0000 (07:16 +0000)]
Don't clear stall at first time use of USB MIDI endpoints.
Most likely some non-USB compliant devices will choke on it
sooner or later. Clear stall is strictly speaking not needed.
If the first MIDI command sent or transmitted is lost, this
is not a big problem for us.

MFC after: 1 week

11 years agoFix rte leak introduced in r248070.
Alexander V. Chernikov [Sat, 18 May 2013 07:10:22 +0000 (07:10 +0000)]
Fix rte leak introduced in r248070.

MFC after: 2 weeks

11 years agoFix issue with "Logitech Webcam C525":
Hans Petter Selasky [Sat, 18 May 2013 07:03:06 +0000 (07:03 +0000)]
Fix issue with "Logitech Webcam C525":

Set a valid alternate interface setting
when enumerating USB audio devices else
the device mentioned will not work like
expected.

PR: usb/178722
MFC after: 1 week

11 years agoFix ipfw(8) sets of ipv6 addresses handling.
Alexander V. Chernikov [Sat, 18 May 2013 04:49:00 +0000 (04:49 +0000)]
Fix ipfw(8) sets of ipv6 addresses handling.
Conditionally use stack buffer instead of calling strdup().

PR: bin/104921
MFC after: 2 weeks

11 years agoImport ACPICA 20130517.
Jung-uk Kim [Fri, 17 May 2013 23:13:40 +0000 (23:13 +0000)]
Import ACPICA 20130517.

11 years agoSet errno to ETIMEDOUT if an SCTP association times out during
Michael Tuexen [Fri, 17 May 2013 22:26:05 +0000 (22:26 +0000)]
Set errno to ETIMEDOUT if an SCTP association times out during
setup.

MFC after: 1 week

11 years agoPer the comment, we cannot rely on bsd.own.mk
Simon J. Gerraty [Fri, 17 May 2013 22:18:27 +0000 (22:18 +0000)]
Per the comment, we cannot rely on bsd.own.mk
we could be on an old system that knows noting of MK_BMAKE
or on an almost up to date one that is defaulting it to "no"
neither of which will work.

11 years agoDon't send an ABORT chunk with verification 0.
Michael Tuexen [Fri, 17 May 2013 21:45:52 +0000 (21:45 +0000)]
Don't send an ABORT chunk with verification 0.

MFC after: 1 week

11 years agoFor ARM, MIPS, and PowerPC, default to 32-byte alignment, but allow it
Warner Losh [Fri, 17 May 2013 20:53:15 +0000 (20:53 +0000)]
For ARM, MIPS, and PowerPC, default to 32-byte alignment, but allow it
to be as small as 8.

11 years agoUse readlocking now that assertions on vm_page_lookup() are relaxed.
Attilio Rao [Fri, 17 May 2013 20:03:55 +0000 (20:03 +0000)]
Use readlocking now that assertions on vm_page_lookup() are relaxed.

Sponsored by: EMC / Isilon storage division
Reviewed by: alc
Tested by: flo, pho

11 years agoWe really need to get the bsd.own.mk from this tree so the
Simon J. Gerraty [Fri, 17 May 2013 19:37:16 +0000 (19:37 +0000)]
We really need to get the bsd.own.mk from this tree so the
correct options are set.
Also defined NO_PWD_OVERRIDE to match behavior of fmake.

11 years agoo Retrive the part number (CP2103 etc) from the hardware on attach.
Gavin Atkinson [Fri, 17 May 2013 19:13:31 +0000 (19:13 +0000)]
o  Retrive the part number (CP2103 etc) from the hardware on attach.
o  The CP2101 and CP2102 do not support GPIO pin use at all, enforce this.
o  Support reading the GPIO status on the second port of the CP2105.  More
   work is needed before the CP2105 GPIO pins can be used as outputs.

Hardware donated by: Silicon Labs
MFC after: 3 weeks

11 years agoRelax the object locking in vm_fault_prefault(). A read lock suffices.
Alan Cox [Fri, 17 May 2013 19:02:36 +0000 (19:02 +0000)]
Relax the object locking in vm_fault_prefault().  A read lock suffices.

Reviewed by: attilio
Sponsored by: EMC / Isilon Storage Division

11 years agoRelax the object locking assertion in pmap_enter_locked().
Alan Cox [Fri, 17 May 2013 18:59:00 +0000 (18:59 +0000)]
Relax the object locking assertion in pmap_enter_locked().

Reviewed by: attilio
Sponsored by: EMC / Isilon Storage Division

11 years ago- Update the release notes to DocBook 4.5 XML to build with the
Gabor Kovesdan [Fri, 17 May 2013 18:53:29 +0000 (18:53 +0000)]
- Update the release notes to DocBook 4.5 XML to build with the
  updated documentation tree

11 years agoRelax the object locking assertion in vm_page_lookup(). Now that a radix
Alan Cox [Fri, 17 May 2013 18:49:43 +0000 (18:49 +0000)]
Relax the object locking assertion in vm_page_lookup().  Now that a radix
tree is used to maintain the object's collection of resident pages,
vm_page_lookup() no longer needs an exclusive lock.

Reviewed by:    attilio
Sponsored by:   EMC / Isilon Storage Division

11 years agoRevert r250692. We'll use 1 kernel config file for the BeagleBone models.
Rui Paulo [Fri, 17 May 2013 15:20:45 +0000 (15:20 +0000)]
Revert r250692. We'll use 1 kernel config file for the BeagleBone models.

11 years agoDecode new HT 3.00 and 3.10 capabilities.
Konstantin Belousov [Fri, 17 May 2013 14:05:31 +0000 (14:05 +0000)]
Decode new HT 3.00 and 3.10 capabilities.

Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com>
MFC after: 1 week

11 years agoAdd new capability types encodings from HyperTransport I/O Link
Konstantin Belousov [Fri, 17 May 2013 14:04:31 +0000 (14:04 +0000)]
Add new capability types encodings from HyperTransport I/O Link
Specification revisions 3.00 and 3.10.

Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com>
MFC after: 1 week

11 years agoUpgrade to OpenSSH 6.2p2. Mostly a no-op since I had already patched
Dag-Erling Smørgrav [Fri, 17 May 2013 09:12:33 +0000 (09:12 +0000)]
Upgrade to OpenSSH 6.2p2.  Mostly a no-op since I had already patched
the issues that affected us.

11 years agoVendor import of OpenSSH 6.2p2.
Dag-Erling Smørgrav [Fri, 17 May 2013 08:55:26 +0000 (08:55 +0000)]
Vendor import of OpenSSH 6.2p2.

11 years agoAdd a -N option that prints the jail name rather than its number.
Dag-Erling Smørgrav [Fri, 17 May 2013 08:48:16 +0000 (08:48 +0000)]
Add a -N option that prints the jail name rather than its number.

MFC after: 3 weeks

11 years agoAdd some more debugging printf()s to complain if the ath_buf tx queue
Adrian Chadd [Fri, 17 May 2013 05:16:30 +0000 (05:16 +0000)]
Add some more debugging printf()s to complain if the ath_buf tx queue
doesn't match the actual hardware queue this frame is queued to.

I'm trying to ensure that the holding buffers are actually being queued
to the same TX queue as the holding buffer that they end up on.
I'm pretty sure this is all correct so if this complains, it'll be due
to some kind of subtle broken-ness that needs fixing.

This is only done for legacy hardware, not EDMA hardware.

Tested:

* AR5416 STA mode, very lightly

11 years agoComplete revision 243104.
Greg Lehey [Fri, 17 May 2013 03:52:20 +0000 (03:52 +0000)]
Complete revision 243104.

11 years agoDrop any connection to newsyslog. I haven't worked on it for quite
Garance A Drosehn [Fri, 17 May 2013 03:14:55 +0000 (03:14 +0000)]
Drop any connection to newsyslog.  I haven't worked on it for quite
some time.  Note that I do want to keep the pre-commit review for
usr.sbin/lpr.  I am actively working on some updates for that.

11 years agoguard against MK_SHARED_TOOLCHAIN not being defined.
Xin LI [Thu, 16 May 2013 22:35:33 +0000 (22:35 +0000)]
guard against MK_SHARED_TOOLCHAIN not being defined.

11 years agoSimilar to r250719: guard against MK_BMAKE not being defined.
Xin LI [Thu, 16 May 2013 22:25:17 +0000 (22:25 +0000)]
Similar to r250719: guard against MK_BMAKE not being defined.

11 years agoFix cut-n-paste error.
Julian Elischer [Thu, 16 May 2013 21:27:47 +0000 (21:27 +0000)]
Fix cut-n-paste error.

11 years agoReverse the sense of the test wrt bmake, and guard against
Simon J. Gerraty [Thu, 16 May 2013 21:26:46 +0000 (21:26 +0000)]
Reverse the sense of the test wrt bmake, and guard against
MK_BMAKE not being defined.

11 years agoConstify string pointers.
Xin LI [Thu, 16 May 2013 21:04:56 +0000 (21:04 +0000)]
Constify string pointers.

Verified with: sha256(1)
MFC after: 2 weeks

11 years agoMinor mdoc improvements.
Joel Dahl [Thu, 16 May 2013 20:55:20 +0000 (20:55 +0000)]
Minor mdoc improvements.

11 years agoWhen running the -m option to generate a newfs(8) command suitable for
Kirk McKusick [Thu, 16 May 2013 20:07:08 +0000 (20:07 +0000)]
When running the -m option to generate a newfs(8) command suitable for
recreating the filesystem, check for and output the -i, -k, and -l
options if appropriate.

Note the remaining deficiencies of the -m option in the dumpfs(8)
manual page. Specifically that newfs(8) options -E, -R, -S, and -T
options are not handled and that -p is not useful so is omitted.
Also document that newfs(8) options -n and -r are neither checked
for nor output but should be. The -r flag is needed if the filesystem
uses gjournal(8).

PR:           bin/163992
Reported by:  Dieter <freebsd@sopwith.solgatos.com>
Submitted by: Andy Kosela <akosela@andykosela.com>
MFC after:    1 week