firefly-linux-kernel-4.4.55.git
14 years agoInput: i8042 - add ALDI/MEDION netbook E1222 to qurik reset table
Christoph Fritz [Sun, 14 Mar 2010 06:26:23 +0000 (22:26 -0800)]
Input: i8042 - add ALDI/MEDION netbook E1222 to qurik reset table

commit 31968ecf584330b51a25b7bf881c2b632a02a3fb upstream.

ALDI/MEDION netbook E1222 needs to be in the reset quirk list for
its touchpad's proper function.

Reported-by: Michael Fischer <mifi@gmx.de>
Signed-off-by: Christoph Fritz <chf.fritz@googlemail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoInput: alps - add support for the touchpad on Toshiba Tecra A11-11L
Thomas Bächler [Wed, 10 Mar 2010 04:38:48 +0000 (20:38 -0800)]
Input: alps - add support for the touchpad on Toshiba Tecra A11-11L

commit eb8bff85c5bd5caef7c374ff32b86545029efb56 upstream.

Signed-off-by: Thomas Bächler <thomas@archlinux.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotimekeeping: Prevent oops when GENERIC_TIME=n
john stultz [Mon, 1 Mar 2010 20:34:43 +0000 (12:34 -0800)]
timekeeping: Prevent oops when GENERIC_TIME=n

commit ad6759fbf35d104dbf573cd6f4c6784ad6823f7e upstream.

Aaro Koskinen reported an issue in kernel.org bugzilla #15366, where
on non-GENERIC_TIME systems, accessing
/sys/devices/system/clocksource/clocksource0/current_clocksource
results in an oops.

It seems the timekeeper/clocksource rework missed initializing the
curr_clocksource value in the !GENERIC_TIME case.

Thanks to Aaro for reporting and diagnosing the issue as well as
testing the fix!

Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: John Stultz <johnstul@us.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
LKML-Reference: <1267475683.4216.61.camel@localhost.localdomain>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda: enable MSI for Gateway M-6866
Daniel T Chen [Mon, 15 Mar 2010 05:04:26 +0000 (01:04 -0400)]
ALSA: hda: enable MSI for Gateway M-6866

BugLink: https://bugs.launchpad.net/bugs/538918
The OR has verified that explicitly enabling MSI is necessary for audio
to be audible by default.

This patch is only applicable to 2.6.32.y; in 2.6.33, MSI is enabled by
default.

Reported-by: Sam Townsend <stownsend42@sbcglobal.net>
Tested-by: Sam Townsend <stownsend42@sbcglobal.net>
Acked-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Daniel T Chen <crimsun@ubuntu.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda - Fix input source elements of secondary ADCs on Realtek
Takashi Iwai [Mon, 8 Mar 2010 11:13:07 +0000 (12:13 +0100)]
ALSA: hda - Fix input source elements of secondary ADCs on Realtek

commit 5311114d4867113c00f78829d4ce14be458ec925 upstream.

Since alc_auto_create_input_ctls() doesn't set the elements for the
secondary ADCs, "Input Source" elemtns for these also get empty, resulting
in buggy outputs of alsactl like:
control.14 {
comment.access 'read write'
comment.type ENUMERATED
comment.count 1
iface MIXER
name 'Input Source'
index 1
value 0
}

This patch fixes alc_mux_enum_*() (and others) to fall back to the
first entry if the secondary input mux is empty.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotg3: Fix 5906 transmit hangs
Matt Carlson [Fri, 12 Mar 2010 23:12:13 +0000 (18:12 -0500)]
tg3: Fix 5906 transmit hangs

This is a resubmit backport of commit 92c6b8d16a36df3f28b2537bed2a56491fb08f11
to kernel version 2.6.32. The gentoo bug report can be found at
https://bugs.gentoo.org/show_bug.cgi?id=301091. Thanks to Matt Carlson for his
assistance and working me to fix a regression caused by the initial patch.  The
original description is as follows:

The 5906 has trouble with fragments that are less than 8 bytes in size.  This
patch works around the problem by pivoting the 5906's transmit routine to
tg3_start_xmit_dma_bug() and introducing a new SHORT_DMA_BUG flag that enables
code to detect and react to the problematic condition.

Signed-off-by: Mike Pagano <mpagano@gentoo.org>
Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotg3: Fix tg3_poll_controller() passing wrong pointer to tg3_interrupt()
Louis Rilling [Tue, 9 Mar 2010 06:14:41 +0000 (06:14 +0000)]
tg3: Fix tg3_poll_controller() passing wrong pointer to tg3_interrupt()

commit fe234f0e5cbb880792d2d1ac0743cf8c07e9dde3 upstream.

Commit 09943a1819a240ff4a72f924d0038818fcdd0a90
Author: Matt Carlson <mcarlson@broadcom.com>
Date:   Fri Aug 28 14:01:57 2009 +0000

tg3: Convert ISR parameter to tnapi

forgot to update tg3_poll_controller(), leading to intermittent crashes with
netpoll.

Fix this.

Signed-off-by: Louis Rilling <louis.rilling@kerlabs.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoMIPS: Cleanup forgotten label_module_alloc in tlbex.c
David Daney [Fri, 4 Dec 2009 01:43:54 +0000 (17:43 -0800)]
MIPS: Cleanup forgotten label_module_alloc in tlbex.c

commit abbdc3d88aa2d5c937b21044c336bcd056c1732f upstream.

commit c8af165342e83a4eb078c9607d29a7c399d30a53 (lmo) rsp.
e0cc87f59490d7d62a8ab2a76498dc8a2b64927a (kernel.org) left
label_module_alloc unused.  Remove it now.

Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Cc: linux-mips@linux-mips.org
Patchwork: http://patchwork.linux-mips.org/patch/752/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoARM: Fix decompressor's kernel size estimation for ROM=y
Russell King [Thu, 25 Feb 2010 23:56:38 +0000 (23:56 +0000)]
ARM: Fix decompressor's kernel size estimation for ROM=y

commit 98e12b5a6e05413420a7e3b3eca7fbfc2ff41b6d upstream.

Commit 2552fc2 changed the way the decompressor decides if it is safe
to decompress the kernel directly to its final location.  Unfortunately,
it took the top of the compressed data as being the stack pointer,
which it is for ROM=n cases.  However, for ROM=y, the stack pointer
is not relevant, and results in the wrong answer.

Fix this by explicitly storing the end of the biggybacked data in the
decompressor, and use that to calculate the compressed image size.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodecompress: fix new decompressor for PIC
Russell King [Wed, 10 Mar 2010 23:23:53 +0000 (15:23 -0800)]
decompress: fix new decompressor for PIC

commit 5ceaa2f39bfa73c4398cd01e78f1c3ebde3d3383 upstream.

The ARM kernel decompressor wants to be able to relocate r/w data
independently from the rest of the image, and we do this by ensuring that
r/w data has global visibility.  Define STATIC_RW_DATA to be empty to
achieve this.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Cc: Alain Knaff <alain@knaff.lu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodrivers/scsi/ses.c: eliminate double free
Julia Lawall [Wed, 10 Mar 2010 23:20:42 +0000 (15:20 -0800)]
drivers/scsi/ses.c: eliminate double free

commit 9b3a6549b2602ca30f58715a0071e29f9898cae9 upstream.

The few lines below the kfree of hdr_buf may go to the label err_free
which will also free hdr_buf.  The most straightforward solution seems to
be to just move the kfree of hdr_buf after these gotos.

A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)

// <smpl>
@r@
identifier E;
expression E1;
iterator I;
statement S;
@@

*kfree(E);
... when != E = E1
    when != I(E,...) S
    when != &E
*kfree(E);
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodrm/i915: fix gpio register detection logic for BIOS without VBT
Shaohua Li [Wed, 18 Nov 2009 07:15:02 +0000 (15:15 +0800)]
drm/i915: fix gpio register detection logic for BIOS without VBT

commit 29874f44fbcbc24b231b42c9956f8f9de9407231 upstream.

if no VBT is present, crt_ddc_bus will be left at 0, and cause us
to use that for the GPIO register offset. That's never a valid register
offset, so let the "undefined" value be 0 instead of -1.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
[anholt: clarified the commit message a bit]
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoLinux 2.6.32.10
Greg Kroah-Hartman [Mon, 15 Mar 2010 15:52:04 +0000 (08:52 -0700)]
Linux 2.6.32.10

14 years agox86, mm: Allow highmem user page tables to be disabled at boot time
Ian Campbell [Wed, 17 Feb 2010 10:38:10 +0000 (10:38 +0000)]
x86, mm: Allow highmem user page tables to be disabled at boot time

commit 14315592009c17035cac81f4954d5a1f4d71e489 upstream.

Distros generally (I looked at Debian, RHEL5 and SLES11) seem to
enable CONFIG_HIGHPTE for any x86 configuration which has highmem
enabled. This means that the overhead applies even to machines which
have a fairly modest amount of high memory and which therefore do not
really benefit from allocating PTEs in high memory but still pay the
price of the additional mapping operations.

Running kernbench on a 4G box I found that with CONFIG_HIGHPTE=y but
no actual highptes being allocated there was a reduction in system
time used from 59.737s to 55.9s.

With CONFIG_HIGHPTE=y and highmem PTEs being allocated:
  Average Optimal load -j 4 Run (std deviation):
  Elapsed Time 175.396 (0.238914)
  User Time 515.983 (5.85019)
  System Time 59.737 (1.26727)
  Percent CPU 263.8 (71.6796)
  Context Switches 39989.7 (4672.64)
  Sleeps 42617.7 (246.307)

With CONFIG_HIGHPTE=y but with no highmem PTEs being allocated:
  Average Optimal load -j 4 Run (std deviation):
  Elapsed Time 174.278 (0.831968)
  User Time 515.659 (6.07012)
  System Time 55.9 (1.07799)
  Percent CPU 263.8 (71.266)
  Context Switches 39929.6 (4485.13)
  Sleeps 42583.7 (373.039)

This patch allows the user to control the allocation of PTEs in
highmem from the command line ("userpte=nohigh") but retains the
status-quo as the default.

It is possible that some simple heuristic could be developed which
allows auto-tuning of this option however I don't have a sufficiently
large machine available to me to perform any particularly meaningful
experiments. We could probably handwave up an argument for a threshold
at 16G of total RAM.

Assuming 768M of lowmem we have 196608 potential lowmem PTE
pages. Each page can map 2M of RAM in a PAE-enabled configuration,
meaning a maximum of 384G of RAM could potentially be mapped using
lowmem PTEs.

Even allowing generous factor of 10 to account for other required
lowmem allocations, generous slop to account for page sharing (which
reduces the total amount of RAM mappable by a given number of PT
pages) and other innacuracies in the estimations it would seem that
even a 32G machine would not have a particularly pressing need for
highmem PTEs. I think 32G could be considered to be at the upper bound
of what might be sensible on a 32 bit machine (although I think in
practice 64G is still supported).

It's seems questionable if HIGHPTE is even a win for any amount of RAM
you would sensibly run a 32 bit kernel on rather than going 64 bit.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
LKML-Reference: <1266403090-20162-1-git-send-email-ian.campbell@citrix.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosched: Don't use possibly stale sched_class
Thomas Gleixner [Wed, 17 Feb 2010 08:05:48 +0000 (09:05 +0100)]
sched: Don't use possibly stale sched_class

commit 83ab0aa0d5623d823444db82c3b3c34d7ec364ae upstream.

setscheduler() saves task->sched_class outside of the rq->lock held
region for a check after the setscheduler changes have become
effective. That might result in checking a stale value.

rtmutex_setprio() has the same problem, though it is protected by
p->pi_lock against setscheduler(), but for correctness sake (and to
avoid bad examples) it needs to be fixed as well.

Retrieve task->sched_class inside of the rq->lock held region.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosched: Fix SMT scheduler regression in find_busiest_queue()
Suresh Siddha [Sat, 13 Feb 2010 01:14:22 +0000 (17:14 -0800)]
sched: Fix SMT scheduler regression in find_busiest_queue()

commit 9000f05c6d1607f79c0deacf42b09693be673f4c upstream.

Fix a SMT scheduler performance regression that is leading to a scenario
where SMT threads in one core are completely idle while both the SMT threads
in another core (on the same socket) are busy.

This is caused by this commit (with the problematic code highlighted)

   commit bdb94aa5dbd8b55e75f5a50b61312fe589e2c2d1
   Author: Peter Zijlstra <a.p.zijlstra@chello.nl>
   Date:   Tue Sep 1 10:34:38 2009 +0200

   sched: Try to deal with low capacity

   @@ -4203,15 +4223,18 @@ find_busiest_queue()
   ...
for_each_cpu(i, sched_group_cpus(group)) {
   + unsigned long power = power_of(i);

   ...

   - wl = weighted_cpuload(i);
   + wl = weighted_cpuload(i) * SCHED_LOAD_SCALE;
   + wl /= power;

   - if (rq->nr_running == 1 && wl > imbalance)
   + if (capacity && rq->nr_running == 1 && wl > imbalance)
continue;

On a SMT system, power of the HT logical cpu will be 589 and
the scheduler load imbalance (for scenarios like the one mentioned above)
can be approximately 1024 (SCHED_LOAD_SCALE). The above change of scaling
the weighted load with the power will result in "wl > imbalance" and
ultimately resulting in find_busiest_queue() return NULL, causing
load_balance() to think that the load is well balanced. But infact
one of the tasks can be moved to the idle core for optimal performance.

We don't need to use the weighted load (wl) scaled by the cpu power to
compare with  imabalance. In that condition, we already know there is only a
single task "rq->nr_running == 1" and the comparison between imbalance,
wl is to make sure that we select the correct priority thread which matches
imbalance. So we really need to compare the imabalnce with the original
weighted load of the cpu and not the scaled load.

But in other conditions where we want the most hammered(busiest) cpu, we can
use scaled load to ensure that we consider the cpu power in addition to the
actual load on that cpu, so that we can move the load away from the
guy that is getting most hammered with respect to the actual capacity,
as compared with the rest of the cpu's in that busiest group.

Fix it.

Reported-by: Ma Ling <ling.ma@intel.com>
Initial-Analysis-by: Zhang, Yanmin <yanmin_zhang@linux.intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1266023662.2808.118.camel@sbs-t61.sc.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosched: Fix sched_mv_power_savings for !SMT
Vaidyanathan Srinivasan [Mon, 8 Feb 2010 10:05:55 +0000 (15:35 +0530)]
sched: Fix sched_mv_power_savings for !SMT

commit 28f5318167adf23b16c844b9c2253f355cb21796 upstream.

Fix for sched_mc_powersavigs for pre-Nehalem platforms.
Child sched domain should clear SD_PREFER_SIBLING if parent will have
SD_POWERSAVINGS_BALANCE because they are contradicting.

Sets the flags correctly based on sched_mc_power_savings.

Signed-off-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100208100555.GD2931@dirshya.in.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoKVM: x86 emulator: Check CPL level during privilege instruction emulation
Gleb Natapov [Wed, 10 Feb 2010 12:21:35 +0000 (14:21 +0200)]
KVM: x86 emulator: Check CPL level during privilege instruction emulation

commit e92805ac1228626c59c865f2f4e9059b9fb8c97b upstream.

Add CPL checking in case emulator is tricked into emulating
privilege instruction from userspace.

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoKVM: x86 emulator: Add group9 instruction decoding
Gleb Natapov [Wed, 10 Feb 2010 12:21:30 +0000 (14:21 +0200)]
KVM: x86 emulator: Add group9 instruction decoding

commit 60a29d4ea4e7b6b95d9391ebc8625b0426f3a363 upstream.

Use groups mechanism to decode 0F C7 instructions.

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoKVM: x86 emulator: Forbid modifying CS segment register by mov instruction
Gleb Natapov [Thu, 18 Feb 2010 10:14:59 +0000 (12:14 +0200)]
KVM: x86 emulator: Forbid modifying CS segment register by mov instruction

commit 8b9f44140bc4afd2698413cd9960c3912168ee91 upstream.

Inject #UD if guest attempts to do so. This is in accordance to Intel
SDM.

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoKVM: x86 emulator: Add group8 instruction decoding
Gleb Natapov [Wed, 10 Feb 2010 12:21:29 +0000 (14:21 +0200)]
KVM: x86 emulator: Add group8 instruction decoding

commit 2db2c2eb6226e30f8059b82512a1364db98da8e3 upstream.

Use groups mechanism to decode 0F BA instructions.

Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodm: free dm_io before bio_endio not after
Mikulas Patocka [Sat, 6 Mar 2010 02:32:29 +0000 (02:32 +0000)]
dm: free dm_io before bio_endio not after

commit a97f925a32aad2a37971d7bfb657006acf04e42d upstream.

Free the dm_io structure before calling bio_endio() instead of after it,
to ensure that the io_pool containing it is not referenced after it is
freed.

This partially fixes a problem described here
  https://www.redhat.com/archives/dm-devel/2010-February/msg00109.html

thread 1:
bio_endio(bio, io_error);
/* scheduling happens */
thread 2:
close the device
remove the device
thread 1:
free_io(md, io);

Thread 2, when removing the device, sees non-empty md->io_pool (because the
io hasn't been freed by thread 1 yet) and may crash with BUG in mempool_free.
Thread 1 may also crash, when freeing into a nonexisting mempool.

To fix this we must make sure that bio_endio() is the last call and
the md structure is not accessed afterwards.

There is another bio_endio in process_barrier, but it is called from the thread
and the thread is destroyed prior to freeing the mempools, so this call is
not affected by the bug.

A similar bug exists with module unloads - the module may be unloaded
immediately after bio_endio - but that is more difficult to fix.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoNFS: Fix an allocation-under-spinlock bug
Trond Myklebust [Tue, 2 Mar 2010 18:06:22 +0000 (13:06 -0500)]
NFS: Fix an allocation-under-spinlock bug

commit ebed9203b68a4f333ce5d17e874b26c3afcfeff1 upstream.

sunrpc_cache_update() will always call detail->update() from inside the
detail->hash_lock, so it cannot allocate memory.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agortc-coh901331: fix braces in resume code
James Hogan [Fri, 5 Mar 2010 21:44:31 +0000 (13:44 -0800)]
rtc-coh901331: fix braces in resume code

commit 5a98c04d78c896d52baef20ffc11f6d1ba6eb786 upstream.

The else part of the if statement is indented but does not have braces
around it. It clearly should since it uses clk_enable and clk_disable
which are supposed to balance.

Signed-off-by: James Hogan <james@albanarts.com>
Acked-by: Linus Walleij <linus.walleij@stericsson.com>
Acked-by: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agos3cmci: s3cmci_card_present: Use no_detect to decide whether there is a card detect pin
Lars-Peter Clausen [Fri, 5 Mar 2010 21:43:37 +0000 (13:43 -0800)]
s3cmci: s3cmci_card_present: Use no_detect to decide whether there is a card detect pin

commit dc2ed552804f3a2ae41c0ffe4bc09879ec8f7396 upstream.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Cc: Ben Dooks <ben-linux@fluff.org>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoSUNRPC: Handle EINVAL error returns from the TCP connect operation
Trond Myklebust [Tue, 2 Mar 2010 18:06:21 +0000 (13:06 -0500)]
SUNRPC: Handle EINVAL error returns from the TCP connect operation

commit 9fcfe0c83c3b04a759cde6b8c5f961237f17808b upstream.

This can, for instance, happen if the user specifies a link local IPv6
address.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosunrpc: remove unnecessary svc_xprt_put
Neil Brown [Fri, 26 Feb 2010 22:33:40 +0000 (09:33 +1100)]
sunrpc: remove unnecessary svc_xprt_put

commit ab1b18f70a007ea6caeb007d269abb75b131a410 upstream.

The 'struct svc_deferred_req's on the xpt_deferred queue do not
own a reference to the owning xprt.  This is seen in svc_revisit
which is where things are added to this queue.  dr->xprt is set to
NULL and the reference to the xprt it put.

So when this list is cleaned up in svc_delete_xprt, we mustn't
put the reference.

Also, replace the 'for' with a 'while' which is arguably
simpler and more likely to compile efficiently.

Cc: Tom Tucker <tom@opengridcomputing.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodrm/ttm: handle OOM in ttm_tt_swapout
Maarten Maathuis [Sat, 20 Feb 2010 02:22:21 +0000 (03:22 +0100)]
drm/ttm: handle OOM in ttm_tt_swapout

commit 290e55056ec3d25c72088628245d8cae037b30db upstream.

- Without this change I get a general protection fault.
- Also use PTR_ERR where applicable.

Signed-off-by: Maarten Maathuis <madman2003@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Acked-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodrm/i915: Use a dmi quirk to skip a broken SDVO TV output.
Zhao Yakui [Mon, 8 Feb 2010 13:35:12 +0000 (21:35 +0800)]
drm/i915: Use a dmi quirk to skip a broken SDVO TV output.

commit 6070a4a928f8c92b9fae7d6717ebbb05f425d6b2 upstream.

This IBM system has a multi-function SDVO card that reports both VGA
and TV, but the system has no TV connector.  The TV connector always
reported as connected, which would lead to poor modesetting choices.

https://bugs.freedesktop.org/show_bug.cgi?id=25787

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Tested-by: Vance <liangghv@sg.ibm.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: unusual_devs: Add support for multiple Option 3G sticks
Jan Dumon [Tue, 5 Jan 2010 14:53:26 +0000 (15:53 +0100)]
USB: unusual_devs: Add support for multiple Option 3G sticks

commit 46216e4fbe8c62059b5440dec0b236f386248a41 upstream.

Enable the SD-Card interface on multiple Option 3G sticks.
The unusual_devs.h entry is necessary because the device descriptor is
vendor-specific. That prevents usb-storage from binding to it as an interface
driver.

Signed-off-by: Jan Dumon <j.dumon@option.com>
Signed-off-by: Phil Dibowitz <phil@ipom.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: cp210x: Add 81E8 (Zephyr Bioharness)
Alan Cox [Mon, 8 Feb 2010 10:10:44 +0000 (10:10 +0000)]
USB: cp210x: Add 81E8 (Zephyr Bioharness)

commit bd07c551aae5d2200c7b195142e5ba63f26424da upstream.

As reported in
http://bugzilla.kernel.org/show_bug.cgi?id=10980

Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: serial: ftdi: add CONTEC vendor and product id
Daniel Sangorrin [Mon, 22 Feb 2010 02:03:11 +0000 (11:03 +0900)]
USB: serial: ftdi: add CONTEC vendor and product id

commit 46b72d78cb022714c89a9ebc00b9581b550cfca7 upstream.

This is a patch to ftdi_sio_ids.h and ftdi_sio.c that adds
identifiers for CONTEC USB serial converter. I tested it
with the device COM-1(USB)H

Signed-off-by: Daniel Sangorrin <daniel.sangorrin@gmail.com>
Cc: Andreas Mohr <andi@lisas.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: add new ftdi_sio device ids
Mitchell Solomon [Fri, 12 Feb 2010 18:23:18 +0000 (13:23 -0500)]
USB: add new ftdi_sio device ids

commit 9714080d20f2ec4b671a06ce69367d91fa9e227e upstream.

PID patch for my products

Signed-off-by: Mitchell Solomon <mitchjs@rush2112.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: ftdi_sio: add device IDs (several ELV, one Mindstorms NXT)
Andreas Mohr [Sun, 17 Jan 2010 10:45:38 +0000 (11:45 +0100)]
USB: ftdi_sio: add device IDs (several ELV, one Mindstorms NXT)

commit 65e1ec6751b3eefee6d94161185e78736366126f upstream.

- add FTDI device IDs for several ELV devices and NXTCam of Lego Mindstorms NXT
- add hopefully helpful new_id comment
- remove less helpful "Due to many user requests for multiple ELV devices we enable
  them by default." comment (we simply add _all_ known devices - an
  enduser shouldn't have to fiddle with obscure module parameters...).
- add myself to DRIVER_AUTHOR

The missing NXTCam ID has been found at
http://www.unixboard.de/vb3/showthread.php?t=44155
, ELV devices taken from ELV Windows .inf file.

Signed-off-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: ftdi_sio: new device id for papouch AD4USB
Radek Liboska [Wed, 27 Jan 2010 14:38:34 +0000 (15:38 +0100)]
USB: ftdi_sio: new device id for papouch AD4USB

commit a7787e508acb4378d62f4584bae3dd1cd0ba3eac upstream.

added new device pid (PAPOUCH_AD4USB_PID) to ftdi_sio.h and ftdi_sio.c

AD4USB measuring converter is a 4-input A/D converter which enables the
user to measure to four current inputs ranging from 0(4) to 20 mA or
voltage between 0 and 10 V. The measured values are then transferred to
a superior system in digital form. The AD4USB communicates via USB.
Powered is also via USB.  datasheet in english is here:
http://www.papouch.com/shop/scripts/pdf/ad4usb_en.pdf

Signed-off-by: Radek Liboska <liboska@uochb.cas.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: ftdi_sio: sort PID/VID entries in new ftdi_sio_ids.h header
Andreas Mohr [Thu, 17 Dec 2009 10:56:09 +0000 (11:56 +0100)]
USB: ftdi_sio: sort PID/VID entries in new ftdi_sio_ids.h header

commit 4e092d110fe931db3878865db185be1c9df3e182 upstream.

This is a (almost) sort-only patch to sort FTDI device
product ID definitions in new ftdi_sio_ids.h header.

Advantage is that new device ID submissions will now have a specific (sorted)
position - less future merge conflicts.

Compile-tested, based on _current_ mainline git.
Minor checkpatch.pl warnings were eliminated whereever it made sense,
very minor text changes.

Signed-off-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: ftdi_sio: isolate all device IDs to new ftdi_sio_ids.h header
Andreas Mohr [Wed, 16 Dec 2009 20:45:10 +0000 (21:45 +0100)]
USB: ftdi_sio: isolate all device IDs to new ftdi_sio_ids.h header

commit 31844d55800e1b93fe75c4d6188a4a44db2e1bbe upstream.

This is a strictly move-only patch to relocate all FTDI device
product ID definitions to their own ftdi_sio_ids.h header
(following the usual *_ids.h kernel tree convention, too),
thus correcting the slightly too messy appearance
(crucial driver defines were stuck somewhere in the decaying middle swamp
of the huge existing header).

Compile-tested, based on latest mainline git.

Signed-off-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: Move hcd free_dev call into usb_disconnect to fix oops
Herbert Xu [Sun, 10 Jan 2010 09:15:03 +0000 (20:15 +1100)]
USB: Move hcd free_dev call into usb_disconnect to fix oops

commit f7410ced7f931bb1ad79d1336412cf7b7a33cb14 upstream.

USB: Move hcd free_dev call into usb_disconnect

I found a way to oops the kernel:

1. Open a USB device through devio.
2. Remove the hcd module in the host kernel.
3. Close the devio file descriptor.

The problem is that closing the file descriptor does usb_release_dev
as it is the last reference.  usb_release_dev then tries to invoke
the hcd free_dev function (or rather dereferencing the hcd driver
struct).  This causes an oops as the hcd driver has already been
unloaded so the struct is gone.

This patch tries to fix this by bringing the free_dev call earlier
and into usb_disconnect.  I have verified that repeating the
above steps no longer crashes with this patch applied.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: remove debugging message for uevent constructions
Alan Stern [Mon, 8 Feb 2010 14:45:12 +0000 (09:45 -0500)]
USB: remove debugging message for uevent constructions

commit cceffe9348f93188d7811bda95924d4bd3040d0f upstream.

This patch (as1332) removes an unneeded and annoying debugging message
announcing all USB uevent constructions.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: fix crash in uhci_scan_schedule
Pete Zaitcev [Fri, 8 Jan 2010 22:39:22 +0000 (15:39 -0700)]
USB: fix crash in uhci_scan_schedule

commit d23356da714595b888686d22cd19061323c09190 upstream.

When hardware is removed on a Stratus, the system may crash like this:

ACPI: PCI interrupt for device 0000:7c:00.1 disabled
Trying to free nonexistent resource <00000000a8000000-00000000afffffff>
Trying to free nonexistent resource <00000000a4800000-00000000a480ffff>
uhci_hcd 0000:7e:1d.0: remove, state 1
usb usb2: USB disconnect, address 1
usb 2-1: USB disconnect, address 2
Unable to handle kernel paging request at 0000000000100100 RIP:
 [<ffffffff88021950>] :uhci_hcd:uhci_scan_schedule+0xa2/0x89c

 #4 [ffff81011de17e50] uhci_scan_schedule at ffffffff88021918
 #5 [ffff81011de17ed0] uhci_irq at ffffffff88023cb8
 #6 [ffff81011de17f10] usb_hcd_irq at ffffffff801f1c1f
 #7 [ffff81011de17f20] handle_IRQ_event at ffffffff8001123b
 #8 [ffff81011de17f50] __do_IRQ at ffffffff800ba749

This occurs because an interrupt scans uhci->skelqh, which is
being freed. We do the right thing: disable the interrupts in the
device, and do not do any processing if the interrupt is shared
with other source, but it's possible that another CPU gets
delayed somewhere (e.g. loops) until we started freeing.

The agreed-upon solution is to wait for interrupts to play out
before proceeding. No other bareers are neceesary.

A backport of this patch was tested on a 2.6.18 based kernel.
Testing of 2.6.32-based kernels is under way, but it takes us
forever (months) to turn this around. So I think it's a good
patch and we should keep it.

Tracked in RH bz#516851

Signed-Off-By: Pete Zaitcev <zaitcev@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: fix the idProduct value for USB-3.0 root hubs
Alan Stern [Thu, 25 Feb 2010 18:19:37 +0000 (13:19 -0500)]
USB: fix the idProduct value for USB-3.0 root hubs

commit cd780694920fbf869b23c8afb0bd083e7b0448c7 upstream.

This patch (as1346) changes the idProduct value for USB-3.0 root hubs
from 0x0002 (which we already use for USB-2.0 root hubs) to 0x0003.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: xhci: Fix finding extended capabilities registers
Edward Shao [Wed, 10 Feb 2010 19:37:30 +0000 (03:37 +0800)]
USB: xhci: Fix finding extended capabilities registers

commit 05197921ff3dad52d99fd1647974c57d9c28d40e upstream.

According "5.3.6 Capability Parameters (HCCPARAMS)" of xHCI rev0.96 spec,
value of xECP register indicates a relative offset, in 32-bit words,
from Base to the beginning of the first extended capability.
The wrong calculation will cause BIOS handoff fail (not handoff from BIOS)
in some platform with BIOS USB legacy sup support.

Signed-off-by: Edward Shao <laface.tw@gmail.com>
Cc: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agox86: Fix SCI on IOAPIC != 0
Yinghai Lu [Wed, 10 Feb 2010 09:20:05 +0000 (01:20 -0800)]
x86: Fix SCI on IOAPIC != 0

commit 18dce6ba5c8c6bd0f3ab4efa4cbdd698dab5c40a upstream.

Thomas Renninger <trenn@suse.de> reported on IBM x3330

booting a latest kernel on this machine results in:

PCI: PCI BIOS revision 2.10 entry at 0xfd61c, last bus=1
PCI: Using configuration type 1 for base access bio: create slab <bio-0> at 0
ACPI: SCI (IRQ30) allocation failed
ACPI Exception: AE_NOT_ACQUIRED, Unable to install System Control Interrupt handler (20090903/evevent-161)
ACPI: Unable to start the ACPI Interpreter

Later all kind of devices fail...

and bisect it down to this commit:
commit b9c61b70075c87a8612624736faf4a2de5b1ed30

    x86/pci: update pirq_enable_irq() to setup io apic routing

it turns out we need to set irq routing for the sci on ioapic1 early.

-v2: make it work without sparseirq too.
-v3: fix checkpatch.pl warning, and cc to stable

Reported-by: Thomas Renninger <trenn@suse.de>
Bisected-by: Thomas Renninger <trenn@suse.de>
Tested-by: Thomas Renninger <trenn@suse.de>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
LKML-Reference: <1265793639-15071-2-git-send-email-yinghai@kernel.org>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agox86: Avoid race condition in pci_enable_msix()
Brandon Phiilps [Wed, 10 Feb 2010 09:20:06 +0000 (01:20 -0800)]
x86: Avoid race condition in pci_enable_msix()

commit ced5b697a76d325e7a7ac7d382dbbb632c765093 upstream.

Keep chip_data in create_irq_nr and destroy_irq.

When two drivers are setting up MSI-X at the same time via
pci_enable_msix() there is a race.  See this dmesg excerpt:

[   85.170610] ixgbe 0000:02:00.1: irq 97 for MSI/MSI-X
[   85.170611]   alloc irq_desc for 99 on node -1
[   85.170613] igb 0000:08:00.1: irq 98 for MSI/MSI-X
[   85.170614]   alloc kstat_irqs on node -1
[   85.170616] alloc irq_2_iommu on node -1
[   85.170617]   alloc irq_desc for 100 on node -1
[   85.170619]   alloc kstat_irqs on node -1
[   85.170621] alloc irq_2_iommu on node -1
[   85.170625] ixgbe 0000:02:00.1: irq 99 for MSI/MSI-X
[   85.170626]   alloc irq_desc for 101 on node -1
[   85.170628] igb 0000:08:00.1: irq 100 for MSI/MSI-X
[   85.170630]   alloc kstat_irqs on node -1
[   85.170631] alloc irq_2_iommu on node -1
[   85.170635]   alloc irq_desc for 102 on node -1
[   85.170636]   alloc kstat_irqs on node -1
[   85.170639] alloc irq_2_iommu on node -1
[   85.170646] BUG: unable to handle kernel NULL pointer dereference
at 0000000000000088

As you can see igb and ixgbe are both alternating on create_irq_nr()
via pci_enable_msix() in their probe function.

ixgbe: While looping through irq_desc_ptrs[] via create_irq_nr() ixgbe
choses irq_desc_ptrs[102] and exits the loop, drops vector_lock and
calls dynamic_irq_init. Then it sets irq_desc_ptrs[102]->chip_data =
NULL via dynamic_irq_init().

igb: Grabs the vector_lock now and starts looping over irq_desc_ptrs[]
via create_irq_nr(). It gets to irq_desc_ptrs[102] and does this:

cfg_new = irq_desc_ptrs[102]->chip_data;
if (cfg_new->vector != 0)
continue;

This hits the NULL deref.

Another possible race exists via pci_disable_msix() in a driver or in
the number of error paths that call free_msi_irqs():

destroy_irq()
dynamic_irq_cleanup() which sets desc->chip_data = NULL
...race window...
desc->chip_data = cfg;

Remove the save and restore code for cfg in create_irq_nr() and
destroy_irq() and take the desc->lock when checking the irq_cfg.

Reported-and-analyzed-by: Brandon Philips <bphilips@suse.de>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
LKML-Reference: <1265793639-15071-3-git-send-email-yinghai@kernel.org>
Signed-off-by: Brandon Phililps <bphilips@suse.de>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agox86, xen: Disable highmem PTE allocation even when CONFIG_HIGHPTE=y
Ian Campbell [Fri, 26 Feb 2010 17:16:00 +0000 (17:16 +0000)]
x86, xen: Disable highmem PTE allocation even when CONFIG_HIGHPTE=y

commit 817a824b75b1475f1b067c8cee318c7b4d66fcde upstream.

There's a path in the pagefault code where the kernel deliberately
breaks its own locking rules by kmapping a high pte page without
holding the pagetable lock (in at least page_check_address). This
breaks Xen's ability to track the pinned/unpinned state of the
page. There does not appear to be a viable workaround for this
behaviour so simply disable HIGHPTE for all Xen guests.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
LKML-Reference: <1267204562-11844-1-git-send-email-ian.campbell@citrix.com>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Pasi Kärkkäinen <pasik@iki.fi>
Cc: <xen-devel@lists.xensource.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agox86: Add iMac9,1 to pci_reboot_dmi_table
Justin P. Mattock [Tue, 16 Feb 2010 23:17:29 +0000 (15:17 -0800)]
x86: Add iMac9,1 to pci_reboot_dmi_table

commit 0a832320f1bae6a4169bf683e201378f2437cfc1 upstream.

On the iMac9,1 /sbin/reboot results in a black mangled screen. Adding
this DMI entry gets the machine to reboot cleanly as it should.

Signed-off-by: Justin P. Mattock <justinmattock@gmail.com>
LKML-Reference: <1266362249-3337-1-git-send-email-justinmattock@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agox86, ia32_aout: do not kill argument mapping
Jiri Slaby [Wed, 10 Feb 2010 19:55:16 +0000 (20:55 +0100)]
x86, ia32_aout: do not kill argument mapping

commit 318f6b228ba88a394ef560efc1bfe028ad5ae6b6 upstream.

Do not set current->mm->mmap to NULL in 32-bit emulation on 64-bit
load_aout_binary after flush_old_exec as it would destroy already
set brpm mapping with arguments.

Introduced by b6a2fea39318e43fee84fa7b0b90d68bed92d2ba
mm: variable length argument support
where the argument mapping in bprm was added.

[ hpa: this is a regression from 2.6.22... time to kill a.out? ]

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
LKML-Reference: <1265831716-7668-1-git-send-email-jslaby@suse.cz>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ollie Wild <aaw@google.com>
Cc: x86@kernel.org
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoocfs2: Only bug out in direct io write for reflinked extent.
Tao Ma [Fri, 26 Feb 2010 02:54:52 +0000 (10:54 +0800)]
ocfs2: Only bug out in direct io write for reflinked extent.

commit cbaee472f274ea9a98aabe47025f6e5551acadcb upstream.

In ocfs2_direct_IO_get_blocks, we only need to bug out
in case of we are going to write a recounted extent rec.

What a silly bug introduced by me!

Signed-off-by: Tao Ma <tao.ma@oracle.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agothinkpad-acpi: fix bluetooth/wwan resume
Henrique de Moraes Holschuh [Fri, 26 Feb 2010 01:22:07 +0000 (22:22 -0300)]
thinkpad-acpi: fix bluetooth/wwan resume

commit 08fedfc903c78e380b0baa7b57c52d367794d0a5 upstream.

Studying the DSDTs of various thinkpads, it looks like bit 3 of the
argument to SBDC and SWAN is not "set radio to last state on resume".
Rather, it seems to be "if this bit is set, enable radio on resume,
otherwise disable it on resume".

So, the proper way to prepare the radios for S3 suspend is: disable
radio and clear bit 3 on the SBDC/SWAN call to to resume with radio
disabled, and enable radio and set bit 3 on the SBDC/SWAN call to
resume with the radio enabled.

Also, for persistent devices, the rfkill core does not restore state,
so we really need to get the firmware to do the right thing.

We don't sync the radio state on suspend, instead we trust the BIOS to
not do anything weird if we never touched the radio state since boot.
Time will tell if that's a wise way of doing things...

Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agothinkpad-acpi: make driver events work in NVRAM poll mode
Henrique de Moraes Holschuh [Fri, 26 Feb 2010 00:29:00 +0000 (21:29 -0300)]
thinkpad-acpi: make driver events work in NVRAM poll mode

commit 7f0cf712a74fcc3ad21f0bde95bd32c2f2cc3888 upstream.

Thadeu Lima de Souza Cascardo reports this:

Brightness notification does not work until the user writes to
hotkey_mask attribute.  That's because the polling thread will only run
if hotkey_user_mask is set and someone is reading the input device or
if hotkey_driver_mask is set.  In this second case, this condition is
not tested after the mask is changed, because the brightness and
volume drivers are started after the hotkey drivers.

Fix tpacpi_hotkey_driver_mask_set() to call hotkey_poll_setup(), so
that the poller kthread will be started when needed.

Reported-by: Thadeu Lima de Souza Cascardo <cascardo@holoscopio.com>
Tested-by: Thadeu Lima de Souza Cascardo <cascardo@holoscopio.com>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agothinkpad-acpi: document HKEY event 3006
Henrique de Moraes Holschuh [Fri, 26 Feb 2010 00:28:56 +0000 (21:28 -0300)]
thinkpad-acpi: document HKEY event 3006

commit bf8b29c8f7f8269e99eca8b19048ed5b34b51810 upstream.

Event 0x3006 is used to help power management of the ODD in the
UltraBay.  The EC generates this event when the ODD eject button is
pressed (even if the bay is powered down).

Normally, Linux doesn't need this as we keep the SATA link powered
up (which wastes power).  The EC powers up the bay by itself when the
ODD eject button is pressed, and the SATA PHY reports the hotplug.

However, we could also power that SATA link down (and for that matter,
also power down the Ultrabay) if the ODD is left idle for a while with
no disk inside, and use event 0x3006 to know when we need that SATA link
powered back up.

For now, just stop asking for more information when event 0x3006 is
seen, there is no point in pestering users about it anymore.

Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agothinkpad-acpi: R52 brightness_mode has been confirmed
Henrique de Moraes Holschuh [Fri, 26 Feb 2010 00:28:56 +0000 (21:28 -0300)]
thinkpad-acpi: R52 brightness_mode has been confirmed

commit 7d1894d8d1c411d2dad95abfe0f65bacf68c4afa upstream.

We can stop pestering users for confirmation of the brightness_mode
default for firmware TP-76.

While at it, add a few missing comments in that quirk table.

Reported-by: Whoopie <whoopie79@gmx.net>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agothinkpad-acpi: fix poll thread auto-start
Henrique de Moraes Holschuh [Fri, 26 Feb 2010 00:28:58 +0000 (21:28 -0300)]
thinkpad-acpi: fix poll thread auto-start

commit b589ea4c44170d3f7a845684e2d1b3b9571663af upstream.

The driver was not starting the NVRAM polling thread if the input
device was bound immediately after registration.

This fixes:
http://bugzilla.kernel.org/show_bug.cgi?id=15118

Reported-by: Florian Zumbiehl <florz@florz.de>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosunxvr500: Additional PCI id for sunxvr500 driver
Ben Hutchings [Fri, 26 Feb 2010 12:37:09 +0000 (04:37 -0800)]
sunxvr500: Additional PCI id for sunxvr500 driver

commit 275143e9b237dd7e0b6d01660fd9b8acd9922fa7 upstream.

Intergraph bought 3D Labs and some XVR-500 chips have Intergraph's
vendor id.

Reported-by: Jurij Smakov <jurij@wooyd.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agonetfilter: xt_recent: fix false match
Tim Gardner [Tue, 23 Feb 2010 13:59:12 +0000 (14:59 +0100)]
netfilter: xt_recent: fix false match

commit 8ccb92ad41cb311e52ad1b1fe77992c7f47a3b63 upstream.

A rule with a zero hit_count will always match.

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agonetfilter: xt_recent: fix buffer overflow
Tim Gardner [Tue, 23 Feb 2010 13:55:21 +0000 (14:55 +0100)]
netfilter: xt_recent: fix buffer overflow

commit 2c08522e5d2f0af2d6f05be558946dcbf8173683 upstream.

e->index overflows e->stamps[] every ip_pkt_list_tot packets.

Consider the case when ip_pkt_list_tot==1; the first packet received is stored
in e->stamps[0] and e->index is initialized to 1. The next received packet
timestamp is then stored at e->stamps[1] in recent_entry_update(),
a buffer overflow because the maximum e->stamps[] index is 0.

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agob43/b43legacy: Wake queues in wireless_core_start
Larry Finger [Wed, 3 Feb 2010 19:33:44 +0000 (13:33 -0600)]
b43/b43legacy: Wake queues in wireless_core_start

commit 0866b03c7d7dee8a34ffa527ecda426c0f405518 upstream.

If b43 or b43legacy are deauthenticated or disconnected, there is a
possibility that a reconnection is tried with the queues stopped in
mac80211. To prevent this, start the queues before setting
STAT_INITIALIZED.

In b43, a similar change has been in place (twice) in the
wireless_core_init() routine. Remove the duplicate and add similar
code to b43legacy.

Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoath5k: use correct packet type when transmitting
Bob Copeland [Tue, 9 Feb 2010 18:06:54 +0000 (13:06 -0500)]
ath5k: use correct packet type when transmitting

commit 2ac2927a953a01c83df255118922cce1523d1a18 upstream.

The hardware needs to know what type of frames are being
sent in order to fill in various fields, for example the
timestamp in probe responses (before this patch, it was
always 0).  Set it correctly when initializing the TX
descriptor.

Signed-off-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoath9k: disable RIFS search for AR91xx based chips
Felix Fietkau [Wed, 24 Feb 2010 03:43:05 +0000 (04:43 +0100)]
ath9k: disable RIFS search for AR91xx based chips

commit 7bfbae10dc10a5c94a780d117a57e875d77e8e5a upstream.

While ath9k does not support RIFS yet, the ability to receive RIFS
frames is currently enabled for most chipsets in the initvals.
This is causing baseband related issues on AR9160 and AR9130 based
chipsets, which can lock up under certain conditions.

This patch fixes these issues by overriding the initvals, effectively
disabling RIFS for all affected chipsets.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Acked-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoath9k: fix rate control fallback rate selection
Felix Fietkau [Fri, 19 Feb 2010 00:46:36 +0000 (01:46 +0100)]
ath9k: fix rate control fallback rate selection

commit 5c0ba62fd4b2dce08055a89600f1d834f9f0fe9e upstream.

When selecting the tx fallback rate, rc.c used a separate variable
'nrix' for storing the next rate index, however it did not use that as
reference for further rate index lowering. Because of that, it ended up
reusing the same rate for multiple multi-rate retry stages, thus
decreasing delivery probability under changing link conditions.

This patch removes the separate (unnecessary) variable and fixes
fallback the way it was intended to work.
This should result in increased throughput and better link stability.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoath9k: fix beacon timer restart after a card reset
Felix Fietkau [Fri, 19 Feb 2010 17:21:42 +0000 (18:21 +0100)]
ath9k: fix beacon timer restart after a card reset

commit d8728ee919282c7b01b65cd479ec1e2a9c5d3ba8 upstream.

In AP mode, ath_beacon_config_ap only restarts the timer if a TSF
restart is requested. Apparently this was added, because this function
unconditionally sets the flag for TSF reset.

The problem with this is, that ath9k_hw_reset() clobbers the timer
registers (specified in the initvals), thus effectively disabling the
SWBA interrupt whenever a card reset without TSF reset is issued
(happens in a few places in the code).

This patch fixes ath_beacon_config_ap to only issue the TSF reset flag
when necessary, but reinitialize the timer unconditionally. Tests show,
that this is enough to keep the SWBA interrupt going after a call to
ath_reset()

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoscm: Only support SCM_RIGHTS on unix domain sockets.
Eric W. Biederman [Sun, 28 Feb 2010 01:20:36 +0000 (01:20 +0000)]
scm: Only support SCM_RIGHTS on unix domain sockets.

commit 76dadd76c265a0cdb5a76aa4eef03fcc9639b388 upstream.

We use scm_send and scm_recv on both unix domain and
netlink sockets, but only unix domain sockets support
everything required for file descriptor passing,
so error if someone attempts to pass file descriptors
over netlink sockets.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agopata_hpt3x2n: always stretch UltraDMA timing
Sergei Shtylyov [Mon, 7 Dec 2009 19:25:52 +0000 (23:25 +0400)]
pata_hpt3x2n: always stretch UltraDMA timing

commit 60661933995bc7a09686c901439e17c2a4ea7d5d upstream.

The UltraDMA Tss timing must be stretched with ATA clock of 66 MHz, but the
driver only does this when PCI clock is 66 MHz, whereas it always programs
DPLL clock (which is used as the ATA clock) to 66 MHz.

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agohwmon: (tmp421) Restore missing inputs
Jean Delvare [Fri, 5 Mar 2010 21:17:25 +0000 (22:17 +0100)]
hwmon: (tmp421) Restore missing inputs

commit 8d59582a867470a3e0c3eced4a01625ae8dc546b upstream.

An off-by-one error caused some inputs to not be created by the driver
when they should. TMP421 gets only one input instead of two, TMP422
gets two instead of three, etc. Fix the bug by listing explicitly the
number of inputs each device has.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Tested-by: Andre Prendel <andre.prendel@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agohwmon: (tmp421) Fix temperature conversions
Jean Delvare [Fri, 5 Mar 2010 21:17:25 +0000 (22:17 +0100)]
hwmon: (tmp421) Fix temperature conversions

commit a44908d742a577fb5ccb9a8c082326d4cea234c2 upstream.

The low bits of temperature registers are status bits, they must be
masked out before converting the register values to temperatures.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Tested-by: Andre Prendel <andre.prendel@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agogpiolib: Actually set output state in wm831x_gpio_direction_output()
Mark Brown [Wed, 17 Feb 2010 18:04:35 +0000 (18:04 +0000)]
gpiolib: Actually set output state in wm831x_gpio_direction_output()

commit 3383d23d86791503559cb87837491af37469d9e5 upstream.

wm831x_gpio_direction_output() ignored the state passed into it.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agop54pci: handle dma mapping errors
Christian Lamparter [Sun, 17 Jan 2010 22:17:29 +0000 (23:17 +0100)]
p54pci: handle dma mapping errors

commit 288c8ce8047695fd8872dd5db3ef21a9679c402f upstream.

This patch adds error-paths to handle pci_dma_mapping errors.

Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agop54usb: Add the USB ID for Belkin (Accton) FD7050E ver 1010ec
Jean-François Moine [Wed, 17 Feb 2010 16:59:31 +0000 (10:59 -0600)]
p54usb: Add the USB ID for Belkin (Accton) FD7050E ver 1010ec

commit 5b9a919a97ac8bdda8020c9b366491b5b91b196e upstream.

Yet another USB ID.

Signed-off-by: Jean-François Moine <moinejf@free.fr>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: fix I2C API usage in ohci-pnx4008.
Luotao Fu [Fri, 19 Feb 2010 14:42:00 +0000 (15:42 +0100)]
USB: fix I2C API usage in ohci-pnx4008.

commit 8740cc7d0c532e098cc428251c08befd14f087d8 upstream.

i2c_board_info doesn't contain a member called name. i2c_register_client
call does not exist.

Signed-off-by: Luotao Fu <l.fu@pengutronix.de>
Acked-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: SIS USB2VGA DRIVER: support KAIREN's USB VGA adaptor USB20SVGA-MB-PLUS
Tanaka Akira [Wed, 20 Jan 2010 17:31:09 +0000 (02:31 +0900)]
USB: SIS USB2VGA DRIVER: support KAIREN's USB VGA adaptor USB20SVGA-MB-PLUS

commit bbcb8bbad52b8795912e8f02c2b319092b96078e upstream.

This patch adds the USB product ID of KAIREN's USB VGA Adaptor,
USB20SVGA-MB-PLUS, to sisusbvga work with it.

Signed-off-by: Tanaka Akira <akr@fsij.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: serial: sierra driver indat_callback fix
Elina Pasheva [Mon, 15 Feb 2010 22:50:14 +0000 (14:50 -0800)]
USB: serial: sierra driver indat_callback fix

commit b87c6e86dac1bb5222279cc8ff7e09529e1c4ed9 upstream.

A crash has been reported with sierra driver on disconnect with
Ubuntu/Lucid distribution based on kernel-2.6.32.
The cause of the crash was determined as "NULL tty pointer was being
referenced" and the NULL pointer was passed by sierra_indat_callback().

This patch modifies sierra_indat_callback() function to check for NULL
tty structure pointer. This modification prevents a crash from happening
when the device is disconnected.

This patch fixes the bug reported in Launchpad:
  https://bugs.launchpad.net/ubuntu/+source/linux/+bug/511157

Signed-off-by: Elina Pasheva <epasheva@sierrawireless.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoserial: imx: fix NULL dereference Oops when pdata == NULL
Baruch Siach [Tue, 22 Dec 2009 00:26:46 +0000 (16:26 -0800)]
serial: imx: fix NULL dereference Oops when pdata == NULL

commit bbcd18d1b37413d25eaf4580682b1b8e4a09ff5e upstream.

The platform code doesn't have to provide platform data to get sensible
default behaviour from the imx serial driver.

This patch does not handle NULL dereference in the IrDA case, which still
requires a valid platform data pointer (in imx_startup()/imx_shutdown()),
since I don't know whether there is a sensible default behaviour, or
should the operation just fail cleanly.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Cc: Baruch Siach <baruch@tkos.co.il>
Cc: Alan Cox <alan@linux.intel.com>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Oskar Schirmer <os@emlix.com>
Cc: Fabian Godehardt <fg@emlix.com>
Cc: Daniel Glöckner <dg@emlix.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotty: Fix the ldisc hangup race
Alan Cox [Mon, 8 Feb 2010 10:09:26 +0000 (10:09 +0000)]
tty: Fix the ldisc hangup race

commit 638b9648ab51c9c549ff5735d3de519ef6199df3 upstream.

This was noticed by Matthias Urlichs and he proposed a fix. This patch
does the fixing a different way to avoid introducing several new race
conditions into the code.

The problem case is TTY_DRIVER_RESET_TERMIOS = 0. In that case while we
abort the ldisc change, the hangup processing has not cleaned up and restarted
the ldisc either.

We can't restart the ldisc stuff in the set_ldisc as we don't know what
the hangup did and may touch stuff we shouldn't as we are no longer
supposed to influence the tty at that point in case it has been re-opened
before we get rescheduled.

Instead do it the simple way. Always re-init the ldisc on the hangup, but
use TTY_DRIVER_RESET_TERMIOS to indicate that we should force N_TTY.

Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoDriver-Core: devtmpfs - reset inode permissions before unlinking
Kay Sievers [Wed, 13 Jan 2010 13:16:36 +0000 (14:16 +0100)]
Driver-Core: devtmpfs - reset inode permissions before unlinking

commit 5e31d76f2817bd50258a092a7c5b15b3006fd61c upstream.

Before unlinking the inode, reset the current permissions of possible
references like hardlinks, so granted permissions can not be retained
across the device lifetime by creating hardlinks, in the unusual case
that there is a user-writable directory on the same filesystem.

Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodriver-core: fix race condition in get_device_parent()
Tejun Heo [Fri, 5 Feb 2010 08:57:02 +0000 (17:57 +0900)]
driver-core: fix race condition in get_device_parent()

commit 77d3d7c1d561f49f755d7390f0764dff90765974 upstream.

sysfs is creating several devices in cuse class concurrently and with
CONFIG_SYSFS_DEPRECATED turned off, it triggers the following oops.

 BUG: unable to handle kernel NULL pointer dereference at 0000000000000038
 IP: [<ffffffff81158b0a>] sysfs_addrm_start+0x4a/0xf0
 PGD 75bb067 PUD 75be067 PMD 0
 Oops: 0000 [#1] PREEMPT SMP
 last sysfs file: /sys/devices/system/cpu/cpu7/topology/core_siblings
 CPU 1
 Modules linked in: cuse fuse
 Pid: 4737, comm: osspd Not tainted 2.6.31-work #77
 RIP: 0010:[<ffffffff81158b0a>]  [<ffffffff81158b0a>] sysfs_addrm_start+0x4a/0xf0
 RSP: 0018:ffff88000042f8f8  EFLAGS: 00010296
 RAX: ffff88000042ffd8 RBX: 0000000000000000 RCX: 0000000000000000
 RDX: 0000000000000000 RSI: ffff880007eef660 RDI: 0000000000000001
 RBP: ffff88000042f918 R08: 0000000000000000 R09: 0000000000000000
 R10: 0000000000000001 R11: ffffffff81158b0a R12: ffff88000042f928
 R13: 00000000fffffff4 R14: 0000000000000000 R15: ffff88000042f9a0
 FS:  00007fe93905a950(0000) GS:ffff880008600000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
 CR2: 0000000000000038 CR3: 00000000077c9000 CR4: 00000000000006e0
 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
 Process osspd (pid: 4737, threadinfo ffff88000042e000, task ffff880007eef040)
 Stack:
  ffff880005da10e8 0000000011cc8d6e ffff88000042f928 ffff880003d28a28
 <0> ffff88000042f988 ffffffff811592d7 0000000000000000 0000000000000000
 <0> 0000000000000000 0000000000000000 ffff88000042f958 0000000011cc8d6e
 Call Trace:
  [<ffffffff811592d7>] create_dir+0x67/0xe0
  [<ffffffff811593a8>] sysfs_create_dir+0x58/0xb0
  [<ffffffff8128ca7c>] ? kobject_add_internal+0xcc/0x220
  [<ffffffff812942e1>] ? vsnprintf+0x3c1/0xb90
  [<ffffffff8128cab7>] kobject_add_internal+0x107/0x220
  [<ffffffff8128cd37>] kobject_add_varg+0x47/0x80
  [<ffffffff8128ce53>] kobject_add+0x53/0x90
  [<ffffffff81357d84>] device_add+0xd4/0x690
  [<ffffffff81356c2b>] ? dev_set_name+0x4b/0x70
  [<ffffffffa001a884>] cuse_process_init_reply+0x2b4/0x420 [cuse]
  ...

The problem is that kobject_add_internal() first adds a kobject to the
kset and then try to create sysfs directory for it.  If the creation
fails, it remove the kobject from the kset.  get_device_parent()
accesses class_dirs kset while only holding class_dirs.list_lock to
see whether the cuse class dir exists.  But when it exists, it may not
have finished initialization yet or may fail and get removed soon.  In
the above case, the former happened so the second one ends up trying
to create subdirectory under NULL sysfs_dirent.

Fix it by grabbing a mutex in get_device_parent().

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Colin Guthrie <cguthrie@mandriva.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoASoC: fix ak4104 register array access
Daniel Mack [Fri, 26 Feb 2010 06:36:54 +0000 (14:36 +0800)]
ASoC: fix ak4104 register array access

commit e555317c083fda01f516d2153589e82514e20e70 upstream.

Don't touch the variable 'reg' to construct the value for the actual SPI
transport. This variable is again used to access the driver's register
cache, and so random memory is overwritten.
Compute the value in-place instead.

Signed-off-by: Daniel Mack <daniel@caiaq.de>
Acked-by: Liam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda-intel: Add position_fix quirk for ASUS M2V-MX SE.
Paul Menzel [Mon, 8 Feb 2010 19:42:46 +0000 (20:42 +0100)]
ALSA: hda-intel: Add position_fix quirk for ASUS M2V-MX SE.

commit 0708cc582f0fe2578eaab722841caf2b4f8cfe37 upstream.

With PulseAudio and an application accessing an input device like `gnome-volume-manager` both have high CPU load as reported in [1].

Loading `snd-hda-intel` with `position_fix=1` fixes this issue. Therefore add a quirk for ASUS M2V-MX SE.

The only downside is, when now exiting for example MPlayer when it is playing an audio file a high pitched sound is outputted by the speaker.

$ lspci -vvnn | grep -A10 Audio
20:01.0 Audio device [0403]: VIA Technologies, Inc. VT1708/A [Azalia HDAC] (VIA High Definition Audio Controller) [1106:3288] (rev 10)
Subsystem: ASUSTeK Computer Inc. Device [1043:8290]
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 0, Cache Line Size: 64 bytes
Interrupt: pin A routed to IRQ 17
Region 0: Memory at fbffc000 (64-bit, non-prefetchable) [size=16K]
Capabilities: <access denied>
Kernel driver in use: HDA Intel

[1] http://sourceforge.net/mailarchive/forum.php?thread_name=1265550675.4642.24.camel%40mattotaupa&forum_name=alsa-user

Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda - Add position_fix quirk for HP dv3
Takashi Iwai [Tue, 1 Dec 2009 13:17:37 +0000 (14:17 +0100)]
ALSA: hda - Add position_fix quirk for HP dv3

commit 2f703e7a2ea5f6d5ea14a7b2cd0d31be07839ac6 upstream.

HP dv3 requires position_fix=1.

Reference: Novell bnc#555935
https://bugzilla.novell.com/show_bug.cgi?id=555935

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda - Add a position_fix quirk for MSI Wind U115
Takashi Iwai [Mon, 30 Nov 2009 10:58:30 +0000 (11:58 +0100)]
ALSA: hda - Add a position_fix quirk for MSI Wind U115

commit 45d4ebf1a6255f2234a041685789cbecac3453f1 upstream.

MSI Wind U115 seems to require position_fix=1 explicitly.
Otherwise it screws up PulseAudio.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda: Use LPIB for a Biostar Microtech board
Daniel T Chen [Fri, 5 Mar 2010 14:04:49 +0000 (09:04 -0500)]
ALSA: hda: Use LPIB for a Biostar Microtech board

commit 0321b69569eadbc13242922925a4316754c5f744 upstream.

BugLink: https://launchpad.net/bugs/523953
The OR has verified that position_fix=1 is necessary to work around
errors on his machine.

Reported-by: MMarking
Signed-off-by: Daniel T Chen <crimsun@ubuntu.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda: Use LPIB for Dell Latitude 131L
Daniel T Chen [Wed, 3 Mar 2010 23:24:26 +0000 (18:24 -0500)]
ALSA: hda: Use LPIB for Dell Latitude 131L

commit 9919c7619c52d01e89103bca405cc3d4a2b1ac31 upstream.

BugLink: https://launchpad.net/bugs/530346
The OR has verified that position_fix=1 is necessary to work around
errors on his machine.

Reported-by: Tom Louwrier
Signed-off-by: Daniel T Chen <crimsun@ubuntu.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: USB MIDI support for Access Music VirusTI
Sebastien Alaiwan [Tue, 16 Feb 2010 07:55:08 +0000 (08:55 +0100)]
ALSA: USB MIDI support for Access Music VirusTI

commit d39e82db73eb876c60d00f00219d767b3be30307 upstream.

Here's a patch that adds MIDI support through USB for one of the Access
Music synths, the VirusTI.

The synth uses standard USBMIDI protocol on its USB interface 3, although
it does signal "vendor specific" class. A magic string has to be sent on
interface 3 to enable the sending of MIDI from the synth (this string was
found by sniffing usb communication of the Windows driver). This is all
my patch does, and it works on my computer.

Please note that the synth can also do standard usb audio I/O on its
interfaces 2&3, which already works with the current snd-usb-audio driver,
except for the audio input from the synth. I'm going to work on it when I
have some time.

Signed-off-by: Sebastien Alaiwan <sebastien.alaiwan@gmail.com>
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: pcm core - fix fifo_size channels interval check
Jaroslav Kysela [Tue, 16 Feb 2010 10:55:43 +0000 (11:55 +0100)]
ALSA: pcm core - fix fifo_size channels interval check

commit 3be522a9514f58e0596db34898a514df206cadc5 upstream.

Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: via82xx: add quirk for D1289 motherboard
Clemens Ladisch [Mon, 22 Feb 2010 09:33:13 +0000 (10:33 +0100)]
ALSA: via82xx: add quirk for D1289 motherboard

commit bf30a4309d4294d3eca248ea8a20c1c3570f5e74 upstream.

Add a headphones-only quirk for the Fujitsu Siemens D1289.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Reported-and-tested-by: Marc Haber <mh+alsa201002@zugschlus.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoALSA: hda: Use 3stack quirk for Toshiba Satellite L40-10Q
Daniel T Chen [Sat, 20 Feb 2010 16:16:30 +0000 (11:16 -0500)]
ALSA: hda: Use 3stack quirk for Toshiba Satellite L40-10Q

commit ba579eb7b30791751f556ee01905636cda50c864 upstream.

BugLink: https://bugs.launchpad.net/bugs/524948
The OR has verified that the existing model=laptop-eapd quirk does not
function correctly but instead needs model=3stack.  Make this change
so that manual corrections to module-init-tools file(s) are not
required.

Reported-by: Lasse Havelund <lasse@havelund.org>
Signed-off-by: Daniel T Chen <crimsun@ubuntu.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agooprofile/x86: fix msr access to reserved counters
Robert Richter [Fri, 26 Feb 2010 12:45:24 +0000 (13:45 +0100)]
oprofile/x86: fix msr access to reserved counters

commit cfc9c0b450176a077205ef39092f0dc1a04e020a upstream.

During switching virtual counters there is access to perfctr msrs. If
the counter is not available this fails due to an invalid
address. This patch fixes this.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agooprofile/x86: use kzalloc() instead of kmalloc()
Robert Richter [Thu, 25 Feb 2010 19:20:25 +0000 (20:20 +0100)]
oprofile/x86: use kzalloc() instead of kmalloc()

commit c17c8fbf349482e89b57d1b800e83e9f4cf40c47 upstream.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agooprofile/x86: remove node check in AMD IBS initialization
Robert Richter [Thu, 28 Jan 2010 15:50:45 +0000 (16:50 +0100)]
oprofile/x86: remove node check in AMD IBS initialization

commit 89baaaa98a10cad5cc8516c7208b02d9fc711890 upstream.

Standard AMD systems have the same number of nodes as there are
northbridge devices. However, there may kernel configurations
(especially for 32 bit) or system setups exist, where the node number
is different or it can not be detected properly. Thus the check is not
reliable and may fail though IBS setup was fine. For this reason it is
better to remove the check.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agooprofile: remove tracing build dependency
Robert Richter [Wed, 10 Feb 2010 09:03:34 +0000 (10:03 +0100)]
oprofile: remove tracing build dependency

commit 18b4a4d59e97e7ff13ee84b5bec79f3fc70a9f0a upstream.

The commit

 1155de4 ring-buffer: Make it generally available

already made ring-buffer available without the TRACING option
enabled. This patch removes the TRACING dependency from oprofile.

Fixes also oprofile configuration on ia64.

The patch also applies to the 2.6.32-stable kernel.

Reported-by: Tony Jones <tonyj@suse.de>
Signed-off-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agooprofile/x86: fix perfctr nmi reservation for mulitplexing
Robert Richter [Thu, 25 Feb 2010 18:16:46 +0000 (19:16 +0100)]
oprofile/x86: fix perfctr nmi reservation for mulitplexing

commit 68dc819ce829f7e7977a56524e710473bdb55115 upstream.

Multiple virtual counters share one physical counter. The reservation
of virtual counters fails due to duplicate allocation of the same
counter. The counters are already reserved. Thus, virtual counter
reservation may removed at all. This also makes the code easier.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agomacintosh/hwmon/ams: Fix device removal sequence
Jean Delvare [Sun, 31 Jan 2010 04:03:23 +0000 (04:03 +0000)]
macintosh/hwmon/ams: Fix device removal sequence

commit 98ceb75c7c14eada76b0aa9f03a635a735cee3cb upstream.

Some code that is in ams_exit() (the module exit code) should instead
be called when the device (not module) is removed. It probably doesn't
make much of a difference in the PMU case, but in the I2C case it does
matter.

I make no guarantee that my fix isn't racy, I'm not familiar enough
with the ams driver code to tell for sure.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Tested-by: Christian Kujau <lists@nerdbynature.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Stelian Pop <stelian@popies.net>
Cc: Michael Hanselmann <linux-kernel@hansmi.ch>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agomacintosh/therm_adt746x: Fix sysfs attributes lifetime
Jean Delvare [Sun, 31 Jan 2010 04:00:30 +0000 (04:00 +0000)]
macintosh/therm_adt746x: Fix sysfs attributes lifetime

commit 33a470f6d5e1879c26f16f6b34dc09f82d44f6e9 upstream.

Looking at drivers/macintosh/therm_adt746x.c, the sysfs files are
created in thermostat_init() and removed in thermostat_exit(), which
are the driver's init and exit functions. These files are backed-up by
a per-device structure, so it looks like the wrong thing to do: the
sysfs files have a lifetime longer than the data structure that is
backing it up.

I think that sysfs files creation should be moved to the end of
probe_thermostat() and sysfs files removal should be moved to the
beginning of remove_thermostat().

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Tested-by: Christian Kujau <lists@nerdbynature.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Colin Leroy <colin@colino.net>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoPM / Hibernate: Fix preallocating of memory
Rafael J. Wysocki [Thu, 25 Feb 2010 21:32:37 +0000 (22:32 +0100)]
PM / Hibernate: Fix preallocating of memory

commit a9c9b4429df437982d2fbfab1f4b46b01329e9ed upstream.

The hibernate memory preallocation code allocates memory to push some
user space data out of physical RAM, so that the hibernation image is
not too large.  It allocates more memory than necessary for creating
the image, so it has to release some pages to make room for
allocations made while suspending devices and disabling nonboot CPUs,
or the system will hang due to the lack of free pages to allocate
from.  Unfortunately, the function used for freeing these pages,
free_unnecessary_pages(), contains a bug that prevents it from doing
the job on all systems without highmem.

Fix this problem, which is a regression from the 2.6.30 kernel, by
using the right condition for the termination of the loop in
free_unnecessary_pages().

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Reported-and-tested-by: Alan Jenkins <sourcejedi.lkml@googlemail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoV4L/DVB (13148): uvcvideo: Handle V4L2_CTRL_TYPE_BUTTON control type in VIDIOC_QUERYCTRL
Laurent Pinchart [Thu, 6 Aug 2009 09:05:40 +0000 (06:05 -0300)]
V4L/DVB (13148): uvcvideo: Handle V4L2_CTRL_TYPE_BUTTON control type in VIDIOC_QUERYCTRL

commit f4eabafeb3ea41801260fba624cbf2da971d19f8 upstream.

Return minimum, maximum and step set to 0 without querying the hardware.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Cc: maximilian attems <max@stro.at>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodvb-core: Fix DoS bug in ULE decapsulation code that can be triggered by an invalid...
Ang Way Chuang [Thu, 25 Feb 2010 01:45:03 +0000 (09:45 +0800)]
dvb-core: Fix DoS bug in ULE decapsulation code that can be triggered by an invalid Payload Pointer

commit 29e1fa3565a7951cc415c634eb2b78dbdbee151d upstream.

ULE (Unidirectional Lightweight Encapsulation RFC 4326) decapsulation
has a bug that causes endless loop when Payload Pointer of MPEG2-TS
frame is 182 or 183.  Anyone who sends malicious MPEG2-TS frame will
cause the receiver of ULE SNDU to go into endless loop.

This patch was generated and tested against linux-2.6.32.9 and should
apply cleanly to linux-2.6.33 as well because there was only one typo
fix to dvb_net.c since v2.6.32.

This bug was brought to you by modern day Santa Claus who decided to
shower the satellite dish at Keio University with heavy snow causing
huge burst of errors.  We, receiver end, received Santa Claus's gift in
the form of kernel bug.

Care has been taken not to introduce more bug by fixing this bug, but
please scrutinize the code for I always produces buggy code.

Signed-off-by: Ang Way Chuang <wcang79@gmail.com>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoStaging: mimio: remove the mimio driver
Greg Kroah-Hartman [Fri, 12 Feb 2010 01:15:14 +0000 (17:15 -0800)]
Staging: mimio: remove the mimio driver

commit e37bcc0de040127281de13a84a608570355c20eb upstream.

It turns out that Mimio has a userspace solution for this product using
libusb, and the in-kernel driver is just getting in the way now and
causing problems.  So they have asked that the in-kernel driver be
removed.  As the staging driver wasn't quite working anyway, and Mimio
supports their libusb solution for all distros, I am removing the
in-kernel driver.

The libusb solution can be downloaded from:
http://www.mimio.com/downloads/mimio_studio_software/linux.asp

Cc: <mwilder@cs.nmsu.edu>
Cc: Phil Hannent <phil@hannent.co.uk>
Cc: Marc Rousseau <Marc.Rousseau@mimio.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoStaging: hv: match on DMI values to know if we should run.
Greg Kroah-Hartman [Fri, 26 Feb 2010 00:43:15 +0000 (16:43 -0800)]
Staging: hv: match on DMI values to know if we should run.

commit c22090facd354749cfe99a46e903449c7ac07788 upstream.

The HV core mucks around with specific irqs and other low-level stuff
and takes forever to determine that it really shouldn't be running on a
machine.  So instead, trigger off of the DMI system information and
error out much sooner.  This also allows the module loading tools to
recognize that this code should be loaded on this type of system.

Cc: Haiyang Zhang <haiyangz@microsoft.com>
Cc: Hank Janssen <hjanssen@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoStaging: hv: add a pci device table
Greg Kroah-Hartman [Fri, 26 Feb 2010 00:42:10 +0000 (16:42 -0800)]
Staging: hv: add a pci device table

commit 9a775dbd4e8c87b7d35549183145321c7205404e upstream.

This allows the HV core to be properly found and autoloaded
by the system tools.

It uses the Microsoft virtual VGA device to trigger this.

Cc: Haiyang Zhang <haiyangz@microsoft.com>
Cc: Hank Janssen <hjanssen@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoSCSI: qla1280: Drop host_lock while requesting firmware
Ben Hutchings [Sat, 23 Jan 2010 18:40:29 +0000 (18:40 +0000)]
SCSI: qla1280: Drop host_lock while requesting firmware

commit 2cec802980727f1daa46d8c31b411e083d49d7a2 upstream.

request_firmware() may sleep and it appears to be safe to release the
spinlock here.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoclocksource: Fix up a registration/IRQ race in the sh drivers.
Paul Mundt [Thu, 25 Feb 2010 07:37:46 +0000 (16:37 +0900)]
clocksource: Fix up a registration/IRQ race in the sh drivers.

commit da64c2a8dee66ca03f4f3e15d84be7bedf73db3d upstream.

All of the SH clocksource drivers follow the scheme that the IRQ is setup
prior to registering the clockevent. The interrupt handler in the
clockevent cases looks to the event handler function pointer being filled
in by the registration code, permitting us to get in to situations where
asserted IRQs step in to the handler before registration has had a chance
to complete and hitting a NULL pointer deref.

In practice this is not an issue for most platforms, but some of them
with fairly special loaders (or that are chain-loading from another
kernel) may enter in to this situation. This fixes up the oops reported
by Rafael on hp6xx.

Reported-and-tested-by: Rafael Ignacio Zurita <rafaelignacio.zurita@gmail.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>