GitHub/LineageOS/G12/android_kernel_amlogic_linux-4.9.git
15 years agohighmem: Fix debug_kmap_atomic() to also handle KM_IRQ_PTE, KM_NMI, and KM_NMI_PTE
Soeren Sandmann [Wed, 28 Oct 2009 17:56:35 +0000 (18:56 +0100)]
highmem: Fix debug_kmap_atomic() to also handle KM_IRQ_PTE, KM_NMI, and KM_NMI_PTE

Previously calling debug_kmap_atomic() with these types would
cause spurious warnings.

(triggered by SysProf using perf events)

Signed-off-by: Soeren Sandmann Pedersen <sandmann@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: a.p.zijlstra@chello.nl
Cc: <stable@kernel.org> # .31.x
LKML-Reference: <ye8vdhz8krw.fsf@camel23.daimi.au.dk>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agohighmem: Fix race in debug_kmap_atomic() which could cause warn_count to underflow
Soeren Sandmann [Wed, 28 Oct 2009 17:55:36 +0000 (18:55 +0100)]
highmem: Fix race in debug_kmap_atomic() which could cause warn_count to underflow

debug_kmap_atomic() tries to prevent ever printing more than 10
warnings, but it does so by testing whether an unsigned integer
is equal to 0. However, if the warning is caused by a nested
IRQ, then this counter may underflow and the stream of warnings
will never end.

Fix that by using a signed integer instead.

Signed-off-by: Soeren Sandmann Pedersen <sandmann@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: a.p.zijlstra@chello.nl
Cc: <stable@kernel.org> # .31.x
LKML-Reference: <ye8zl7b8ktj.fsf@camel23.daimi.au.dk>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agorcu: Fix long-grace-period race between forcing and initialization
Paul E. McKenney [Wed, 28 Oct 2009 15:14:49 +0000 (08:14 -0700)]
rcu: Fix long-grace-period race between forcing and initialization

Very long RCU read-side critical sections (50 milliseconds or
so) can cause a race between force_quiescent_state() and
rcu_start_gp() as follows on kernel builds with multi-level
rcu_node hierarchies:

1. CPU 0 calls force_quiescent_state(), sees that there is a
grace period in progress, and acquires ->fsqlock.

2. CPU 1 detects the end of the grace period, and so
cpu_quiet_msk_finish() sets rsp->completed to rsp->gpnum.
This operation is carried out under the root rnp->lock,
but CPU 0 has not yet acquired that lock.  Note that
rsp->signaled is still RCU_SAVE_DYNTICK from the last
grace period.

3. CPU 1 calls rcu_start_gp(), but no one wants a new grace
period, so it drops the root rnp->lock and returns.

4. CPU 0 acquires the root rnp->lock and picks up rsp->completed
and rsp->signaled, then drops rnp->lock.  It then enters the
RCU_SAVE_DYNTICK leg of the switch statement.

5. CPU 2 invokes call_rcu(), and now needs a new grace period.
It calls rcu_start_gp(), which acquires the root rnp->lock, sets
rsp->signaled to RCU_GP_INIT (too bad that CPU 0 is already in
the RCU_SAVE_DYNTICK leg of the switch statement!)  and starts
initializing the rcu_node hierarchy.  If there are multiple
levels to the hierarchy, it will drop the root rnp->lock and
initialize the lower levels of the hierarchy.

6. CPU 0 notes that rsp->completed has not changed, which permits
        both CPU 2 and CPU 0 to try updating it concurrently.  If CPU 0's
update prevails, later calls to force_quiescent_state() can
count old quiescent states against the new grace period, which
can in turn result in premature ending of grace periods.

Not good.

This patch adds an RCU_GP_IDLE state for rsp->signaled that is
set initially at boot time and any time a grace period ends.
This prevents CPU 0 from getting into the workings of
force_quiescent_state() in step 4.  Additional locking and
checks prevent the concurrent update of rsp->signaled in step 6.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
LKML-Reference: <1256742889199-git-send-email->
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agouids: Prevent tear down race
Thomas Gleixner [Mon, 2 Nov 2009 12:01:56 +0000 (13:01 +0100)]
uids: Prevent tear down race

Ingo triggered the following warning:

WARNING: at lib/debugobjects.c:255 debug_print_object+0x42/0x50()
Hardware name: System Product Name
ODEBUG: init active object type: timer_list
Modules linked in:
Pid: 2619, comm: dmesg Tainted: G        W  2.6.32-rc5-tip+ #5298
Call Trace:
 [<81035443>] warn_slowpath_common+0x6a/0x81
 [<8120e483>] ? debug_print_object+0x42/0x50
 [<81035498>] warn_slowpath_fmt+0x29/0x2c
 [<8120e483>] debug_print_object+0x42/0x50
 [<8120ec2a>] __debug_object_init+0x279/0x2d7
 [<8120ecb3>] debug_object_init+0x13/0x18
 [<810409d2>] init_timer_key+0x17/0x6f
 [<81041526>] free_uid+0x50/0x6c
 [<8104ed2d>] put_cred_rcu+0x61/0x72
 [<81067fac>] rcu_do_batch+0x70/0x121

debugobjects warns about an enqueued timer being initialized. If
CONFIG_USER_SCHED=y the user management code uses delayed work to
remove the user from the hash table and tear down the sysfs objects.

free_uid is called from RCU and initializes/schedules delayed work if
the usage count of the user_struct is 0. The init/schedule happens
outside of the uidhash_lock protected region which allows a concurrent
caller of find_user() to reference the about to be destroyed
user_struct w/o preventing the work from being scheduled. If the next
free_uid call happens before the work timer expired then the active
timer is initialized and the work scheduled again.

The race was introduced in commit 5cb350ba (sched: group scheduling,
sysfs tunables) and made more prominent by commit 3959214f (sched:
delayed cleanup of user_struct)

Move the init/schedule_delayed_work inside of the uidhash_lock
protected region to prevent the race.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Cc: Paul E. McKenney <paulmck@us.ibm.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: stable@kernel.org
15 years agofutex: Fix spurious wakeup for requeue_pi really
Thomas Gleixner [Wed, 28 Oct 2009 19:26:48 +0000 (20:26 +0100)]
futex: Fix spurious wakeup for requeue_pi really

The requeue_pi path doesn't use unqueue_me() (and the racy lock_ptr ==
NULL test) nor does it use the wake_list of futex_wake() which where
the reason for commit 41890f2 (futex: Handle spurious wake up)

See debugging discussing on LKML Message-ID: <4AD4080C.20703@us.ibm.com>

The changes in this fix to the wait_requeue_pi path were considered to
be a likely unecessary, but harmless safety net. But it turns out that
due to the fact that for unknown $@#!*( reasons EWOULDBLOCK is defined
as EAGAIN we built an endless loop in the code path which returns
correctly EWOULDBLOCK.

Spurious wakeups in wait_requeue_pi code path are unlikely so we do
the easy solution and return EWOULDBLOCK^WEAGAIN to user space and let
it deal with the spurious wakeup.

Cc: Darren Hart <dvhltc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: John Stultz <johnstul@linux.vnet.ibm.com>
Cc: Dinakar Guniguntala <dino@in.ibm.com>
LKML-Reference: <4AE23C74.1090502@us.ibm.com>
Cc: stable@kernel.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
15 years agofutex: Move drop_futex_key_refs out of spinlock'ed region
Darren Hart [Thu, 15 Oct 2009 22:30:48 +0000 (15:30 -0700)]
futex: Move drop_futex_key_refs out of spinlock'ed region

When requeuing tasks from one futex to another, the reference held
by the requeued task to the original futex location needs to be
dropped eventually.

Dropping the reference may ultimately lead to a call to
"iput_final" and subsequently call into filesystem- specific code -
which may be non-atomic.

It is therefore safer to defer this drop operation until after the
futex_hash_bucket spinlock has been dropped.

Originally-From: Helge Bahmann <hcb@chaoticmind.net>
Signed-off-by: Darren Hart <dvhltc@us.ibm.com>
Cc: <stable@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Dinakar Guniguntala <dino@in.ibm.com>
Cc: John Stultz <johnstul@linux.vnet.ibm.com>
Cc: Sven-Thorsten Dietrich <sdietrich@novell.com>
Cc: John Kacur <jkacur@redhat.com>
LKML-Reference: <4AD7A298.5040802@us.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agorcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
Paul E. McKenney [Thu, 15 Oct 2009 16:26:14 +0000 (09:26 -0700)]
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang

If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:

o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.

o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.

o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.

o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.

At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.

The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure.  This will cause the next invocation of
force_quiescent_state() to end the grace period.

Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
LKML-Reference: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agorcu: Stopgap fix for synchronize_rcu_expedited() for TREE_PREEMPT_RCU
Paul E. McKenney [Wed, 14 Oct 2009 17:15:56 +0000 (10:15 -0700)]
rcu: Stopgap fix for synchronize_rcu_expedited() for TREE_PREEMPT_RCU

For the short term, map synchronize_rcu_expedited() to
synchronize_rcu() for TREE_PREEMPT_RCU and to
synchronize_sched_expedited() for TREE_RCU.

Longer term, there needs to be a real expedited grace period for
TREE_PREEMPT_RCU, but candidate patches to date are considerably
more complex and intrusive.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
Cc: npiggin@suse.de
Cc: jens.axboe@oracle.com
LKML-Reference: <12555405592331-git-send-email->
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agorcu: Prevent RCU IPI storms in presence of high call_rcu() load
Paul E. McKenney [Wed, 14 Oct 2009 17:15:55 +0000 (10:15 -0700)]
rcu: Prevent RCU IPI storms in presence of high call_rcu() load

As the number of callbacks on a given CPU rises, invoke
force_quiescent_state() only every blimit number of callbacks
(defaults to 10,000), and even then only if no other CPU has
invoked force_quiescent_state() in the meantime.

This should fix the performance regression reported by Nick.

Reported-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
Cc: jens.axboe@oracle.com
LKML-Reference: <12555405592133-git-send-email->
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agofutex: Check for NULL keys in match_futex
Darren Hart [Wed, 14 Oct 2009 17:12:39 +0000 (10:12 -0700)]
futex: Check for NULL keys in match_futex

If userspace tries to perform a requeue_pi on a non-requeue_pi waiter,
it will find the futex_q->requeue_pi_key to be NULL and OOPS.

Check for NULL in match_futex() instead of doing explicit NULL pointer
checks on all call sites.  While match_futex(NULL, NULL) returning
false is a little odd, it's still correct as we expect valid key
references.

Signed-off-by: Darren Hart <dvhltc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@elte.hu>
CC: Eric Dumazet <eric.dumazet@gmail.com>
CC: Dinakar Guniguntala <dino@in.ibm.com>
CC: John Stultz <johnstul@us.ibm.com>
Cc: stable@kernel.org
LKML-Reference: <4AD60687.10306@us.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
15 years agofutex: Handle spurious wake up
Thomas Gleixner [Tue, 13 Oct 2009 18:40:43 +0000 (20:40 +0200)]
futex: Handle spurious wake up

The futex code does not handle spurious wake up in futex_wait and
futex_wait_requeue_pi.

The code assumes that any wake up which was not caused by futex_wake /
requeue or by a timeout was caused by a signal wake up and returns one
of the syscall restart error codes.

In case of a spurious wake up the signal delivery code which deals
with the restart error codes is not invoked and we return that error
code to user space. That causes applications which actually check the
return codes to fail. Blaise reported that on preempt-rt a python test
program run into a exception trap. -rt exposed that due to a built in
spurious wake up accelerator :)

Solve this by checking signal_pending(current) in the wake up path and
handle the spurious wake up case w/o returning to user space.

Reported-by: Blaise Gassend <blaise@willowgarage.com>
Debugged-by: Darren Hart <dvhltc@us.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: stable@kernel.org
LKML-Reference: <new-submission>

15 years agoMerge branch 'urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/rric/oprofile...
Ingo Molnar [Mon, 12 Oct 2009 21:26:36 +0000 (23:26 +0200)]
Merge branch 'urgent' of git://git./linux/kernel/git/rric/oprofile into core/urgent

15 years agooprofile: warn on freeing event buffer too early
Robert Richter [Fri, 9 Oct 2009 01:17:44 +0000 (03:17 +0200)]
oprofile: warn on freeing event buffer too early

A race shouldn't happen since all workqueues or handlers are canceled
or flushed before the event buffer is freed. A warning is triggered
now if the buffer is freed too early.

Also, this patch adds some comments about event buffer protection,
reworks some code and adds code to clear buffer_pos during alloc and
free of the event buffer.

Cc: David Rientjes <rientjes@google.com>
Cc: Stephane Eranian <eranian@google.com>
Signed-off-by: Robert Richter <robert.richter@amd.com>
15 years agooprofile: fix race condition in event_buffer free
David Rientjes [Wed, 9 Sep 2009 13:02:33 +0000 (15:02 +0200)]
oprofile: fix race condition in event_buffer free

Looking at the 2.6.31-rc9 code, it appears there is a race condition
in the event_buffer cleanup code path (shutdown). This could lead to
kernel panic as some CPUs may be operating on the event buffer AFTER
it has been freed. The attached patch solves the problem and makes
sure CPUs check if the buffer is not NULL before they access it as
some may have been spinning on the mutex while the buffer was being
freed.

The race may happen if the buffer is freed during pending reads. But
it is not clear why there are races in add_event_entry() since all
workqueues or handlers are canceled or flushed before the event buffer
is freed.

Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Robert Richter <robert.richter@amd.com>
15 years agolockdep: Use cpu_clock() for lockstat
Peter Zijlstra [Fri, 9 Oct 2009 08:12:41 +0000 (10:12 +0200)]
lockdep: Use cpu_clock() for lockstat

Some tracepoint magic (TRACE_EVENT(lock_acquired)) relies on
the fact that lock hold times are positive and uses div64 on
that. That triggered a build warning on MIPS, and probably
causes bad output in certain circumstances as well.

Make it truly positive.

Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1254818502.21044.112.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzi...
Linus Torvalds [Thu, 8 Oct 2009 19:22:45 +0000 (12:22 -0700)]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  pata_atp867x: add Power Management support
  pata_atp867x: PIO support fixes
  pata_atp867x: clarifications in timings calculations and cable detection
  pata_atp867x: fix it to not claim MWDMA support
  libata: fix incorrect link online check during probe
  ahci: filter FPDMA non-zero offset enable for Aspire 3810T
  libata: make gtf_filter per-dev
  libata: implement more acpi filtering options
  libata: cosmetic updates
  ahci: display all AHCI 1.3 HBA capability flags (v2)
  pata_ali: trivial fix of a very frequent spelling mistake
  ahci: disable 64bit DMA by default on SB600s

15 years agoMerge branch 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 8 Oct 2009 19:16:35 +0000 (12:16 -0700)]
Merge branch 'core-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  futex: fix requeue_pi key imbalance
  futex: Fix typo in FUTEX_WAIT/WAKE_BITSET_PRIVATE definitions
  rcu: Place root rcu_node structure in separate lockdep class
  rcu: Make hot-unplugged CPU relinquish its own RCU callbacks
  rcu: Move rcu_barrier() to rcutree
  futex: Move exit_pi_state() call to release_mm()
  futex: Nullify robust lists after cleanup
  futex: Fix locking imbalance
  panic: Fix panic message visibility by calling bust_spinlocks(0) before dying
  rcu: Replace the rcu_barrier enum with pointer to call_rcu*() function
  rcu: Clean up code based on review feedback from Josh Triplett, part 4
  rcu: Clean up code based on review feedback from Josh Triplett, part 3
  rcu: Fix rcu_lock_map build failure on CONFIG_PROVE_LOCKING=y
  rcu: Clean up code to address Ingo's checkpatch feedback
  rcu: Clean up code based on review feedback from Josh Triplett, part 2
  rcu: Clean up code based on review feedback from Josh Triplett

15 years agoMerge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 8 Oct 2009 19:07:24 +0000 (12:07 -0700)]
Merge branch 'sched-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  sched: Set correct normal_prio and prio values in sched_fork()

15 years agoMerge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Thu, 8 Oct 2009 19:06:36 +0000 (12:06 -0700)]
Merge branch 'x86-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86, pci: Correct spelling in a comment
  x86: Simplify bound checks in the MTRR code
  x86: EDAC: carve out AMD MCE decoding logic
  initcalls: Add early_initcall() for modules
  x86: EDAC: MCE: Fix MCE decoding callback logic

15 years agoMerge branch 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 8 Oct 2009 19:06:09 +0000 (12:06 -0700)]
Merge branch 'tracing-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  tracing: user local buffer variable for trace branch tracer
  tracing: fix warning on kernel/trace/trace_branch.c andtrace_hw_branches.c
  ftrace: check for failure for all conversions
  tracing: correct module boundaries for ftrace_release
  tracing: fix transposed numbers of lock_depth and preempt_count
  trace: Fix missing assignment in trace_ctxwake_*
  tracing: Use free_percpu instead of kfree
  tracing: Check total refcount before releasing bufs in profile_enable failure

15 years agoMerge branch 'sparc-perf-events-fixes-for-linus' of git://git.kernel.org/pub/scm...
Linus Torvalds [Thu, 8 Oct 2009 19:05:50 +0000 (12:05 -0700)]
Merge branch 'sparc-perf-events-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'sparc-perf-events-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  mm, perf_event: Make vmalloc_user() align base kernel virtual address to SHMLBA
  perf_event: Provide vmalloc() based mmap() backing

15 years agoMerge branch 'perf-fixes-for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 8 Oct 2009 19:05:00 +0000 (12:05 -0700)]
Merge branch 'perf-fixes-for-linus-2' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'perf-fixes-for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  perf_events: Make ABI definitions available to userspace
  perf tools: elf_sym__is_function() should accept "zero" sized functions
  tracing/syscalls: Use long for syscall ret format and field definitions
  perf trace: Update eval_flag() flags array to match interrupt.h
  perf trace: Remove unused code in builtin-trace.c
  perf: Propagate term signal to child

15 years agoMerge branch 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 8 Oct 2009 19:04:04 +0000 (12:04 -0700)]
Merge branch 'timers-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86, timers: Check for pending timers after (device) interrupts
  NOHZ: update idle state also when NOHZ is inactive

15 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Thu, 8 Oct 2009 19:03:21 +0000 (12:03 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: ice1724: increase SPDIF and independent stereo buffer sizes
  ALSA: opl3: circular locking in the snd_opl3_note_on() and snd_opl3_note_off()
  ALSA: ICE1712/24 - Change the Multi Track Peak control (level meters) from MIXER to PCM type
  ALSA: hda - Fix yet another auto-mic bug in ALC268
  ASoC: WM8350 capture PGA mutes are inverted
  ASoC: Remove absent SYNC and TDM DAI format options from i.MX SSI
  sound: via82xx: move DXS volume controls to PCM interface
  ALSA: hda - Don't pick up invalid HP pins in alc_subsystem_id()
  ALSA: hda - Add a workaround for ASUS A7K
  ALSA: hda - Fix invalid initializations for ALC861 auto mode
  ASoC: wm8940: Fix check on error code form snd_soc_codec_set_cache_io
  ASoC: Fix SND_SOC_DAPM_LINE handling

15 years agoMerge branch 'drm-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied...
Linus Torvalds [Thu, 8 Oct 2009 19:02:06 +0000 (12:02 -0700)]
Merge branch 'drm-linus' of git://git./linux/kernel/git/airlied/drm-2.6

* 'drm-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6: (24 commits)
  drm/radeon/kms: fix vline register for second head.
  drm/r600: avoid assigning vb twice in blit code
  drm/radeon: use list_for_each_entry instead of list_for_each
  drm/radeon/kms: Fix AGP support for R600/RV770 family (v2)
  drm/radeon/kms: Fallback to non AGP when acceleration fails to initialize (v2)
  drm/radeon/kms: Fix RS600/RV515/R520/RS690 IRQ
  drm/radeon: Fix setting of bits
  drm/ttm: fix refcounting in ttm global code.
  drm/fb: add more correct 8/16/24/32 bpp fb support.
  drm/fb: add setcmap and fix 8-bit support.
  drm/radeon/kms: respect single crtc cards, only create one crtc. (v2)
  drm: Delete the DRM_DEBUG_KMS in drm_mode_cursor_ioctl
  drm/radeon/kms: add support for "Surround View"
  drm/radeon/kms: Fix irq handling on AVIVO hw
  drm/radeon/kms: R600/RV770 remove dead code and print message for wrong BIOS
  drm/radeon/kms: Fix R600/RV770 disable acceleration path
  drm/radeon/kms: Fix R600/RV770 startup path & reset
  drm/radeon/kms: Fix R600 write back buffer
  drm/radeon/kms: Remove old init path as no hw use it anymore
  drm/radeon/kms: Convert RS600 to new init path
  ...

15 years agoMerge branch 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 8 Oct 2009 19:01:01 +0000 (12:01 -0700)]
Merge branch 'omap-fixes-for-linus' of git://git./linux/kernel/git/tmlind/linux-omap-2.6

* 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6:
  omapfb: Blizzard: constify register address tables
  omapfb: Blizzard: fix pointer to be const
  omapfb: Condition mutex acquisition
  omap: iovmm: Add missing mutex_unlock
  omap: iovmm: Fix incorrect spelling
  omap: SRAM: flush the right address after memcpy in omap_sram_push
  omap: Lock DPLL5 at boot
  omap: Fix incorrect 730 vs 850 detection
  OMAP3: PM: introduce a new powerdomain walk helper
  OMAP3: PM: Enable GPIO module-level wakeups
  OMAP3: PM: USBHOST: clear wakeup events on both hosts
  OMAP3: PM: PRCM interrupt: only handle selected PRCM interrupts
  OMAP3: PM: PRCM interrupt: check MPUGRPSEL register
  OMAP3: PM: Prevent hang in prcm_interrupt_handler

15 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/bp/bp
Linus Torvalds [Thu, 8 Oct 2009 19:00:39 +0000 (12:00 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/bp/bp

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/bp/bp:
  amd64_edac: beef up DRAM error injection
  amd64_edac: fix DRAM base and limit extraction
  amd64_edac: fix chip select handling
  amd64_edac: simple fix to allow reporting of CECC errors
  amd64_edac: fix K8 intlv_sel check
  amd64_edac: fix interleave enable tests
  amd64_edac: fix DRAM base and limit address extraction
  amd64_edac: fix driver instance lookup table allocation

15 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Thu, 8 Oct 2009 18:59:30 +0000 (11:59 -0700)]
Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (40 commits)
  ethoc: limit the number of buffers to 128
  ethoc: use system memory as buffer
  ethoc: align received packet to make IP header at word boundary
  ethoc: fix buffer address mapping
  ethoc: fix typo to compute number of tx descriptors
  au1000_eth: Duplicate test of RX_OVERLEN bit in update_rx_stats()
  netxen: Fix Unlikely(x) > y
  pasemi_mac: ethtool get settings fix
  add maintainer for network drop monitor kernel service
  tg3: Fix phylib locking strategy
  rndis_host: support ETHTOOL_GPERMADDR
  ipv4: arp_notify address list bug
  gigaset: add kerneldoc comments
  gigaset: correct debugging output selection
  gigaset: improve error recovery
  gigaset: fix device ERROR response handling
  gigaset: announce if built with debugging
  gigaset: handle isoc frame errors more gracefully
  gigaset: linearize skb
  gigaset: fix reject/hangup handling
  ...

15 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/ide-2.6
Linus Torvalds [Thu, 8 Oct 2009 18:59:06 +0000 (11:59 -0700)]
Merge git://git./linux/kernel/git/davem/ide-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/ide-2.6:
  Revert "Revert "ide: try to use PIO Mode 0 during probe if possible""
  sis5513: fix PIO setup for ATAPI devices

15 years agox86, timers: Check for pending timers after (device) interrupts
Arjan van de Ven [Thu, 8 Oct 2009 13:40:41 +0000 (06:40 -0700)]
x86, timers: Check for pending timers after (device) interrupts

Now that range timers and deferred timers are common, I found a
problem with these using the "perf timechart" tool. Frans Pop also
reported high scheduler latencies via LatencyTop, when using
iwlagn.

It turns out that on x86, these two 'opportunistic' timers only get
checked when another "real" timer happens. These opportunistic
timers have the objective to save power by hitchhiking on other
wakeups, as to avoid CPU wakeups by themselves as much as possible.

The change in this patch runs this check not only at timer
interrupts, but at all (device) interrupts. The effect is that:

 1) the deferred timers/range timers get delayed less

 2) the range timers cause less wakeups by themselves because
    the percentage of hitchhiking on existing wakeup events goes up.

I've verified the working of the patch using "perf timechart", the
original exposed bug is gone with this patch. Frans also reported
success - the latencies are now down in the expected ~10 msec
range.

Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Tested-by: Frans Pop <elendil@planet.nl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mike Galbraith <efault@gmx.de>
LKML-Reference: <20091008064041.67219b13@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agomm, perf_event: Make vmalloc_user() align base kernel virtual address to SHMLBA
David Miller [Mon, 21 Sep 2009 19:22:34 +0000 (12:22 -0700)]
mm, perf_event: Make vmalloc_user() align base kernel virtual address to SHMLBA

When a vmalloc'd area is mmap'd into userspace, some kind of
co-ordination is necessary for this to work on platforms with cpu
D-caches which can have aliases.

Otherwise kernel side writes won't be seen properly in userspace
and vice versa.

If the kernel side mapping and the user side one have the same
alignment, modulo SHMLBA, this can work as long as VM_SHARED is
shared of VMA and for all current users this is true.  VM_SHARED
will force SHMLBA alignment of the user side mmap on platforms with
D-cache aliasing matters.

The bulk of this patch is just making it so that a specific
alignment can be passed down into __get_vm_area_node().  All
existing callers pass in '1' which preserves existing behavior.
vmalloc_user() gives SHMLBA for the alignment.

As a side effect this should get the video media drivers and other
vmalloc_user() users into more working shape on such systems.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Jens Axboe <jens.axboe@oracle.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
LKML-Reference: <200909211922.n8LJMYjw029425@imap1.linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/kyle/parisc-2.6
Linus Torvalds [Thu, 8 Oct 2009 14:40:19 +0000 (07:40 -0700)]
Merge branch 'fixes' of git://git./linux/kernel/git/kyle/parisc-2.6

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/kyle/parisc-2.6:
  agp: parisc-agp.c - use correct page_mask function
  parisc: Fix linker script breakage.
  parisc: convert to asm-generic/hardirq.h
  parisc: Make THREAD_SIZE available to assembly files and linker scripts.
  parisc: correct use of SHF_ALLOC
  parisc: rename parisc's vmalloc_start to parisc_vmalloc_start
  parisc: add me to Maintainers
  parisc: includecheck fix: signal.c
  parisc: HAVE_ARCH_TRACEHOOK
  parisc: add skeleton syscall.h
  parisc: stop using task->ptrace for {single,block}step flags
  parisc: split syscall_trace into two halves
  parisc: add missing TI_TASK macro in syscall.S
  parisc: tracehook_signal_handler
  parisc: tracehook_report_syscall

15 years agolis3lv02d_spi: module unload didn't remove sysfs entry
Samu Onkalo [Wed, 7 Oct 2009 23:32:35 +0000 (16:32 -0700)]
lis3lv02d_spi: module unload didn't remove sysfs entry

In module unload, lis3lv02d core driver sysfs clean up was not called.

Signed-off-by: Samu Onkalo <samu.p.onkalo@nokia.com>
Acked-by: Daniel Mack <daniel@caiaq.de>
Cc: Éric Piel <eric.piel@tremplin-utc.net>
Cc: "Trisal, Kalhan" <kalhan.trisal@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agommc: sdio: don't require CISTPL_VERS_1 to contain 4 strings
David Vrabel [Wed, 7 Oct 2009 23:32:33 +0000 (16:32 -0700)]
mmc: sdio: don't require CISTPL_VERS_1 to contain 4 strings

The PC Card 8.0 specification (vol.  4, section 3.2.10) says the
TPLLV1_INFO field of the CISTPL_VERS_1 tuple must contain 4 strings.  Some
cards don't have all 4 so just parse as many as we can.

[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: David Vrabel <david.vrabel@csr.com>
Tested-by: Jonathan Cameron <jic23@cam.ac.uk>
Tested-by: Bing Zhao <bzhao@marvell.com>
Cc: Roel Kluin <roel.kluin@gmail.com>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopage-types: add hwpoison/unpoison feature
Wu Fengguang [Wed, 7 Oct 2009 23:32:32 +0000 (16:32 -0700)]
page-types: add hwpoison/unpoison feature

For hwpoison stress testing.  The debugfs mount point is assumed to be
/debug/.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopage-types: introduce kpageflags_flags()
Wu Fengguang [Wed, 7 Oct 2009 23:32:31 +0000 (16:32 -0700)]
page-types: introduce kpageflags_flags()

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopage-types: make voffset local variables
Wu Fengguang [Wed, 7 Oct 2009 23:32:30 +0000 (16:32 -0700)]
page-types: make voffset local variables

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopage-types: make standalone pagemap/kpageflags read routines
Wu Fengguang [Wed, 7 Oct 2009 23:32:30 +0000 (16:32 -0700)]
page-types: make standalone pagemap/kpageflags read routines

Refactor the code to be more modular and easier to reuse.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopage-types: introduce checked_open()
Wu Fengguang [Wed, 7 Oct 2009 23:32:29 +0000 (16:32 -0700)]
page-types: introduce checked_open()

This helps merge duplicate code (now and future) and outstand the main
logic.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopage-types: add GPL note
Wu Fengguang [Wed, 7 Oct 2009 23:32:28 +0000 (16:32 -0700)]
page-types: add GPL note

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopagemap: document KPF_KSM and show it in page-types
Wu Fengguang [Wed, 7 Oct 2009 23:32:28 +0000 (16:32 -0700)]
pagemap: document KPF_KSM and show it in page-types

It indicates to the system admin that processes mapping such pages may be
eating less physical memory than the reported numbers by legacy tools.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Cc: Izik Eidus <ieidus@redhat.com>
Acked-by: Chris Wright <chrisw@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agopagemap: export KPF_HWPOISON
Wu Fengguang [Wed, 7 Oct 2009 23:32:27 +0000 (16:32 -0700)]
pagemap: export KPF_HWPOISON

This flag indicates a hardware detected memory corruption on the page.
Any future access of the page data may bring down the machine.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agocgroups: update documentation of cgroups tasks and procs files
Paul Menage [Wed, 7 Oct 2009 23:32:26 +0000 (16:32 -0700)]
cgroups: update documentation of cgroups tasks and procs files

Update documentation of cgroups tasks and procs files

Document the cgroup.procs file.

Clarify the semantics of the cgroup.procs and tasks files.  Although the
current cgroup.procs interface returns a sorted and uniqified list of
pids, potential future performance enhancements could result in those
properties being removed - explicitly document this aspect of the API.

There are no existing users of cgroup.procs, so compatibility isn't an
issue.  There are users of the "tasks" file, but none that would appear to
break in the event of the sorted property being broken.  The standard
"libcpuset" explicitly sorts the results of reading from the tasks file,
and "libcg" and other users don't appear to care about ordering.

Signed-off-by: Paul Menage <menage@google.com>
Reviewed-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agovideo: includecheck fix: da8xx-fb.c
Jaswinder Singh Rajput [Wed, 7 Oct 2009 23:32:25 +0000 (16:32 -0700)]
video: includecheck fix: da8xx-fb.c

fix the following 'make includecheck' warning:

  drivers/video/da8xx-fb.c: linux/device.h is included more than once.

Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agovideo: includecheck fix: msm, mddi.c
Jaswinder Singh Rajput [Wed, 7 Oct 2009 23:32:24 +0000 (16:32 -0700)]
video: includecheck fix: msm, mddi.c

fix the following 'make includecheck' warning:

  drivers/video/msm/mddi.c: linux/delay.h is included more than once.

Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agofs: includecheck fix: proc, kcore.c
Jaswinder Singh Rajput [Wed, 7 Oct 2009 23:32:24 +0000 (16:32 -0700)]
fs: includecheck fix: proc, kcore.c

fix the following 'make includecheck' warning:

  fs/proc/kcore.c: linux/mm.h is included more than once.

Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agomm: includecheck fix: vmalloc.c
Jaswinder Singh Rajput [Wed, 7 Oct 2009 23:32:23 +0000 (16:32 -0700)]
mm: includecheck fix: vmalloc.c

fix the following 'make includecheck' warning:

  mm/vmalloc.c: linux/highmem.h is included more than once.

Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoksm: more on default values
Hugh Dickins [Wed, 7 Oct 2009 23:32:22 +0000 (16:32 -0700)]
ksm: more on default values

Adjust the max_kernel_pages default to a quarter of totalram_pages,
instead of nr_free_buffer_pages() / 4: the KSM pages themselves come from
highmem, and even on a 16GB PAE machine, 4GB of KSM pages would only be
pinning 32MB of lowmem with their rmap_items, so no need for the more
obscure calculation (nor for its own special init function).

There is no way for the user to switch KSM on if CONFIG_SYSFS is not
enabled, so in that case default run to KSM_RUN_MERGE.

Update KSM Documentation and Kconfig to reflect the new defaults.

Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Cc: Izik Eidus <ieidus@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
15 years agoMerge branch 'fix/misc' into for-linus
Takashi Iwai [Thu, 8 Oct 2009 11:00:02 +0000 (13:00 +0200)]
Merge branch 'fix/misc' into for-linus

15 years agoMerge branch 'fix/hda' into for-linus
Takashi Iwai [Thu, 8 Oct 2009 10:59:58 +0000 (12:59 +0200)]
Merge branch 'fix/hda' into for-linus

15 years agoALSA: ice1724: increase SPDIF and independent stereo buffer sizes
Robert Hancock [Thu, 8 Oct 2009 02:19:21 +0000 (20:19 -0600)]
ALSA: ice1724: increase SPDIF and independent stereo buffer sizes

Increase the default and maximum PCM buffer prellocation size for ice1724's
SPDIF and independent stereo pair outputs to 256K, which is the hardware's
maximum supported size. This allows a reduction in interrupt rate and
potentially power usage when an application is not latency-critical.

Signed-off-by: Robert Hancock <hancockrwd@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
15 years agoALSA: opl3: circular locking in the snd_opl3_note_on() and snd_opl3_note_off()
Krzysztof Helt [Wed, 7 Oct 2009 20:51:34 +0000 (22:51 +0200)]
ALSA: opl3: circular locking in the snd_opl3_note_on() and snd_opl3_note_off()

Fix following circular locking in the opl3 driver.

=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.32-rc3 #87
-------------------------------------------------------
swapper/0 is trying to acquire lock:
 (&opl3->voice_lock){..-...}, at: [<cca748fe>] snd_opl3_note_off+0x1e/0xe0 [snd_opl3_synth]

but task is already holding lock:
 (&opl3->sys_timer_lock){..-...}, at: [<cca75169>] snd_opl3_timer_func+0x19/0xc0 [snd_opl3_synth]

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #1 (&opl3->sys_timer_lock){..-...}:
       [<c02461d5>] validate_chain+0xa25/0x1040
       [<c0246aca>] __lock_acquire+0x2da/0xab0
       [<c024731a>] lock_acquire+0x7a/0xa0
       [<c044c300>] _spin_lock_irqsave+0x40/0x60
       [<cca75046>] snd_opl3_note_on+0x686/0x790 [snd_opl3_synth]
       [<cca68912>] snd_midi_process_event+0x322/0x590 [snd_seq_midi_emul]
       [<cca74245>] snd_opl3_synth_event_input+0x15/0x20 [snd_opl3_synth]
       [<cca4dcc0>] snd_seq_deliver_single_event+0x100/0x200 [snd_seq]
       [<cca4de07>] snd_seq_deliver_event+0x47/0x1f0 [snd_seq]
       [<cca4e50b>] snd_seq_dispatch_event+0x3b/0x140 [snd_seq]
       [<cca5008c>] snd_seq_check_queue+0x10c/0x120 [snd_seq]
       [<cca5037b>] snd_seq_enqueue_event+0x6b/0xe0 [snd_seq]
       [<cca4e0fd>] snd_seq_client_enqueue_event+0xdd/0x100 [snd_seq]
       [<cca4eb7a>] snd_seq_write+0xea/0x190 [snd_seq]
       [<c02827b6>] vfs_write+0x96/0x160
       [<c0282c9d>] sys_write+0x3d/0x70
       [<c0202c45>] syscall_call+0x7/0xb

-> #0 (&opl3->voice_lock){..-...}:
       [<c02467e6>] validate_chain+0x1036/0x1040
       [<c0246aca>] __lock_acquire+0x2da/0xab0
       [<c024731a>] lock_acquire+0x7a/0xa0
       [<c044c300>] _spin_lock_irqsave+0x40/0x60
       [<cca748fe>] snd_opl3_note_off+0x1e/0xe0 [snd_opl3_synth]
       [<cca751f0>] snd_opl3_timer_func+0xa0/0xc0 [snd_opl3_synth]
       [<c022ac46>] run_timer_softirq+0x166/0x1e0
       [<c02269e8>] __do_softirq+0x78/0x110
       [<c0226ac6>] do_softirq+0x46/0x50
       [<c0226e26>] irq_exit+0x36/0x40
       [<c0204bd2>] do_IRQ+0x42/0xb0
       [<c020328e>] common_interrupt+0x2e/0x40
       [<c021092f>] apm_cpu_idle+0x10f/0x290
       [<c0201b11>] cpu_idle+0x21/0x40
       [<c04443cd>] rest_init+0x4d/0x60
       [<c055c835>] start_kernel+0x235/0x280
       [<c055c066>] i386_start_kernel+0x66/0x70

other info that might help us debug this:

2 locks held by swapper/0:
 #0:  (&opl3->tlist){+.-...}, at: [<c022abd0>] run_timer_softirq+0xf0/0x1e0
 #1:  (&opl3->sys_timer_lock){..-...}, at: [<cca75169>] snd_opl3_timer_func+0x19/0xc0 [snd_opl3_synth]

stack backtrace:
Pid: 0, comm: swapper Not tainted 2.6.32-rc3 #87
Call Trace:
 [<c0245188>] print_circular_bug+0xc8/0xd0
 [<c02467e6>] validate_chain+0x1036/0x1040
 [<c0247f14>] ? check_usage_forwards+0x54/0xd0
 [<c0246aca>] __lock_acquire+0x2da/0xab0
 [<c024731a>] lock_acquire+0x7a/0xa0
 [<cca748fe>] ? snd_opl3_note_off+0x1e/0xe0 [snd_opl3_synth]
 [<c044c300>] _spin_lock_irqsave+0x40/0x60
 [<cca748fe>] ? snd_opl3_note_off+0x1e/0xe0 [snd_opl3_synth]
 [<cca748fe>] snd_opl3_note_off+0x1e/0xe0 [snd_opl3_synth]
 [<c044c307>] ? _spin_lock_irqsave+0x47/0x60
 [<cca751f0>] snd_opl3_timer_func+0xa0/0xc0 [snd_opl3_synth]
 [<c022ac46>] run_timer_softirq+0x166/0x1e0
 [<c022abd0>] ? run_timer_softirq+0xf0/0x1e0
 [<cca75150>] ? snd_opl3_timer_func+0x0/0xc0 [snd_opl3_synth]
 [<c02269e8>] __do_softirq+0x78/0x110
 [<c044c0fd>] ? _spin_unlock+0x1d/0x20
 [<c025915f>] ? handle_level_irq+0xaf/0xe0
 [<c0226ac6>] do_softirq+0x46/0x50
 [<c0226e26>] irq_exit+0x36/0x40
 [<c0204bd2>] do_IRQ+0x42/0xb0
 [<c024463c>] ? trace_hardirqs_on_caller+0x12c/0x180
 [<c020328e>] common_interrupt+0x2e/0x40
 [<c0208d88>] ? default_idle+0x38/0x50
 [<c021092f>] apm_cpu_idle+0x10f/0x290
 [<c0201b11>] cpu_idle+0x21/0x40
 [<c04443cd>] rest_init+0x4d/0x60
 [<c055c835>] start_kernel+0x235/0x280
 [<c055c210>] ? unknown_bootoption+0x0/0x210
 [<c055c066>] i386_start_kernel+0x66/0x70

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
15 years agoALSA: ICE1712/24 - Change the Multi Track Peak control (level meters) from MIXER...
Pavel Hofman [Tue, 6 Oct 2009 14:04:11 +0000 (16:04 +0200)]
ALSA: ICE1712/24 - Change the Multi Track Peak control (level meters) from MIXER to PCM type

* PLEASE NOTE - this change requires the corresponding update of
  envy24control for ice1712 - kind of an ABI change.
* The "Multi Track Peak" control is read-only level meters indicator.
* The control is VERY confusing to most users since it is currently displayed
  in regular mixers. E.g. alsamixer ignores its read-only status
  and allows changing the levels with keys which makes no sense.

Signed-off-by: Pavel Hofman <pavel.hofman@ivitera.com>
Acked-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
15 years agoMerge branch 'drm-next' of ../drm-next into drm-linus
Dave Airlie [Thu, 8 Oct 2009 04:03:05 +0000 (14:03 +1000)]
Merge branch 'drm-next' of ../drm-next into drm-linus

conflict in radeon since new init path merged with vga arb code.

Conflicts:
drivers/gpu/drm/radeon/radeon.h
drivers/gpu/drm/radeon/radeon_asic.h
drivers/gpu/drm/radeon/radeon_device.c

15 years agotracing: user local buffer variable for trace branch tracer
Steven Rostedt [Thu, 8 Oct 2009 01:53:41 +0000 (21:53 -0400)]
tracing: user local buffer variable for trace branch tracer

Just using the tr->buffer for the API to trace_buffer_lock_reserve
is not good enough. This is because the tr->buffer may change, and we
do not want to commit with a different buffer that we reserved from.

This patch uses a local variable to hold the buffer that was used to
reserve and commit with.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
15 years agotracing: fix warning on kernel/trace/trace_branch.c andtrace_hw_branches.c
Zhenwen Xu [Thu, 8 Oct 2009 01:21:46 +0000 (09:21 +0800)]
tracing: fix warning on kernel/trace/trace_branch.c andtrace_hw_branches.c

fix warnings that caused the API change of trace_buffer_lock_reserve()
change files: kernel/trace/trace_hw_branch.c
              kernel/trace/trace_branch.c

Signed-off-by: Zhenwen Xu <helight.xu@gmail.com>
LKML-Reference: <20091008012146.GA4170@helight>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
15 years agodrm/radeon/kms: fix vline register for second head.
Dave Airlie [Thu, 8 Oct 2009 01:32:49 +0000 (11:32 +1000)]
drm/radeon/kms: fix vline register for second head.

Both r100/r600 had this wrong, use the macro to extract the register
to relocate.

Signed-off-by: Dave Airlie <airlied@redhat.com>
15 years agodrm/r600: avoid assigning vb twice in blit code
Robert Noland [Mon, 5 Oct 2009 15:56:44 +0000 (11:56 -0400)]
drm/r600: avoid assigning vb twice in blit code

There is no need to assign vb before you know that space is available.

[agd5f: adapted for kernel tree.]

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
15 years agodrm/radeon: use list_for_each_entry instead of list_for_each
Dave Airlie [Wed, 7 Oct 2009 23:28:19 +0000 (09:28 +1000)]
drm/radeon: use list_for_each_entry instead of list_for_each

This is just a cleanup of the list macro usage.

Signed-off-by: Dave Airlie <airlied@redhat.com>
15 years agodrm/radeon/kms: Fix AGP support for R600/RV770 family (v2)
Jerome Glisse [Tue, 6 Oct 2009 17:04:30 +0000 (19:04 +0200)]
drm/radeon/kms: Fix AGP support for R600/RV770 family (v2)

For AGP to work unmapped access must cover VRAM & AGP as
AGP is treated like VRAM by the GPU (ie physical address).
This patch properly setup the virtual memory system aperture
to cover AGP if AGP is enabled. It seems that there is memory
corruption after resume when using AGP (RV770 seems unaffected
thought). Version 2 just fix merge issue with updated AGP
fallback patch.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
15 years agodrm/radeon/kms: Fallback to non AGP when acceleration fails to initialize (v2)
Jerome Glisse [Tue, 6 Oct 2009 17:04:29 +0000 (19:04 +0200)]
drm/radeon/kms: Fallback to non AGP when acceleration fails to initialize (v2)

When GPU acceleration is not working with AGP try to fallback to non
AGP GART (either PCI or PCIE GART). This should make KMS failure on
AGP less painfull. We still need to find out what is wrong when AGP
fails but at least user have a lot of more chances to get a working
configuration with acceleration. This patch also cleanup R600/RV770
fallback path so they use same code as others asics. Version 2
factorize agp disabling logic to avoid code duplication and bugs.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
15 years agodrm/radeon/kms: Fix RS600/RV515/R520/RS690 IRQ
Jerome Glisse [Wed, 7 Oct 2009 09:08:22 +0000 (11:08 +0200)]
drm/radeon/kms: Fix RS600/RV515/R520/RS690 IRQ

Bad generated header file leaded to use wrong register
to check IRQ status and acknowledge them. Fix the header
and use proper registers.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
15 years agoftrace: check for failure for all conversions
Steven Rostedt [Wed, 7 Oct 2009 20:57:56 +0000 (16:57 -0400)]
ftrace: check for failure for all conversions

Due to legacy code from back when the dynamic tracer used a daemon,
only core kernel code was checking for failures. This is no longer
the case. We must check for failures any time we perform text modifications.

Cc: stable@kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
15 years agotracing: correct module boundaries for ftrace_release
jolsa@redhat.com [Wed, 7 Oct 2009 17:00:35 +0000 (19:00 +0200)]
tracing: correct module boundaries for ftrace_release

When the module is about the unload we release its call records.
The ftrace_release function was given wrong values representing
the module core boundaries, thus not releasing its call records.

Plus making ftrace_release function module specific.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
LKML-Reference: <1254934835-363-3-git-send-email-jolsa@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
15 years agofutex: fix requeue_pi key imbalance
Darren Hart [Wed, 7 Oct 2009 18:46:54 +0000 (11:46 -0700)]
futex: fix requeue_pi key imbalance

If futex_wait_requeue_pi() wakes prior to requeue, we drop the
reference to the source futex_key twice, once in
handle_early_requeue_pi_wakeup() and once on our way out.

Remove the drop from the handle_early_requeue_pi_wakeup() and keep
the get/drops together in futex_wait_requeue_pi().

Reported-by: Helge Bahmann <hcb@chaoticmind.net>
Signed-off-by: Darren Hart <dvhltc@us.ibm.com>
Cc: Helge Bahmann <hcb@chaoticmind.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Dinakar Guniguntala <dino@in.ibm.com>
Cc: John Stultz <johnstul@us.ibm.com>
Cc: stable-2.6.31 <stable@kernel.org>
LKML-Reference: <4ACCE21E.5030805@us.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
15 years agotracing: fix transposed numbers of lock_depth and preempt_count
Steven Rostedt [Sun, 27 Sep 2009 11:02:07 +0000 (07:02 -0400)]
tracing: fix transposed numbers of lock_depth and preempt_count

The lock_depth and preempt_count numbers in the latency format is
transposed.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
15 years agoamd64_edac: beef up DRAM error injection
Borislav Petkov [Thu, 24 Sep 2009 09:05:30 +0000 (11:05 +0200)]
amd64_edac: beef up DRAM error injection

When injecting DRAM ECC errors (F3xBC_x8), EccVector[15:0] is a bitmask
of which bits should be error injected when written to and holds the
payload of 16-bit DRAM word when read, respectively.

Add /sysfs members to show the DRAM ECC section/word/vector.

Fail wrong injection values entered over /sysfs instead of truncating
them.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoamd64_edac: fix DRAM base and limit extraction
Borislav Petkov [Tue, 22 Sep 2009 14:48:37 +0000 (16:48 +0200)]
amd64_edac: fix DRAM base and limit extraction

On Fam10h and above, F1x[1, 0][7C:40] are DRAM Base/Limit registers
which specify the destination node of a DRAM address. Those address
boundaries are being extracted into ->dram_base[] and ->dram_limit[].
Correct the extraction masks to match the respective address bits.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoamd64_edac: fix chip select handling
Borislav Petkov [Mon, 21 Sep 2009 12:35:51 +0000 (14:35 +0200)]
amd64_edac: fix chip select handling

Different processor families support a different number of chip selects.
Handle this in a family-dependent way with the proper values assigned at
init time (see amd64_set_dct_base_and_mask).

Remove _DCSM_COUNT defines since they're used at one place and originate
from public documentation.

CC: Keith Mannthey <kmannth@us.ibm.com>
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoamd64_edac: simple fix to allow reporting of CECC errors
Keith Mannthey [Fri, 18 Sep 2009 12:35:23 +0000 (14:35 +0200)]
amd64_edac: simple fix to allow reporting of CECC errors

This allows the errors to be further decoded and mapped to csrows.
Tested with ECC debug dimms and an Rev F cpu based system.

Signed-off-by: Keith Mannthey <kmannth@us.ibm.com>
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoamd64_edac: fix K8 intlv_sel check
Borislav Petkov [Fri, 18 Sep 2009 10:39:19 +0000 (12:39 +0200)]
amd64_edac: fix K8 intlv_sel check

The check when DRAM interleaving is enabled should be done against the
pvt->dram_IntlvSel field and not against the ->dram_limit.

Simplify first loop and fixup printk formatting while at it.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoamd64_edac: fix interleave enable tests
Borislav Petkov [Fri, 18 Sep 2009 10:27:27 +0000 (12:27 +0200)]
amd64_edac: fix interleave enable tests

The pvt->dram_IntlvEn saves the 3 "Interleave Enable" bits already
right-shifted by 8 so the check in find_mc_by_sys_addr() by shifting the
values to the left 8 bits is wrong.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoamd64_edac: fix DRAM base and limit address extraction
Borislav Petkov [Fri, 18 Sep 2009 10:12:46 +0000 (12:12 +0200)]
amd64_edac: fix DRAM base and limit address extraction

K8 DRAM base and limit addresses from F1x40 +8*i and F1x44 + 8*i, where
i in (0..7) are both bits 39-24 and therefore the shifting should be
done by 24 and not by 8.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoamd64_edac: fix driver instance lookup table allocation
Borislav Petkov [Mon, 21 Sep 2009 11:23:34 +0000 (13:23 +0200)]
amd64_edac: fix driver instance lookup table allocation

Allocate memory statically for 8-node machines max for simplicity
instead of relying on MAX_NUMNODES which is 0 on !CONFIG_NUMA builds.

Spotted by Jan Beulich.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
15 years agoALSA: hda - Fix yet another auto-mic bug in ALC268
Takashi Iwai [Wed, 7 Oct 2009 13:12:27 +0000 (15:12 +0200)]
ALSA: hda - Fix yet another auto-mic bug in ALC268

Since patch_alc268() doesn't call set_capture_mixer() (due to its h/w
design different from other siblings), it needs to call fixup_automic_adc()
explicitly to set up the auto-mic routing.  Otherwise the indices for
int/ext mics aren't set properly.

Reference: Novell bnc#544899
http://bugzilla.novell.com/show_bug.cgi?id=544899

Signed-off-by: Takashi Iwai <tiwai@suse.de>
15 years agoRevert "Revert "ide: try to use PIO Mode 0 during probe if possible""
Bartlomiej Zolnierkiewicz [Tue, 6 Oct 2009 12:27:45 +0000 (12:27 +0000)]
Revert "Revert "ide: try to use PIO Mode 0 during probe if possible""

This reverts commit 24df31acaff8465d797f0006437b45ad0f2a5cb1.

The root cause of reported system hangs was (now fixed) sis5513 bug
and not "ide: try to use PIO Mode 0 during probe if possible" change
(commit 6029336426a2b43e4bc6f4a84be8789a047d139e) so the revert was
incorrect (it simply replaced one regression with the other one).

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agosis5513: fix PIO setup for ATAPI devices
Bartlomiej Zolnierkiewicz [Tue, 6 Oct 2009 14:46:05 +0000 (14:46 +0000)]
sis5513: fix PIO setup for ATAPI devices

Clear prefetch setting before potentially (re-)enabling it in
config_drive_art_rwp() so the transition of the device type on
the port from ATA to ATAPI (i.e. during warm-plug operation)
is handled correctly.

This is a really old bug (it probably goes back to very early
days of the driver) but it was only affecting warm-plug operation
until the recent "ide: try to use PIO Mode 0 during probe if
possible" change (commit 6029336426a2b43e4bc6f4a84be8789a047d139e).

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Tested-by: David Fries <david@fries.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoNOHZ: update idle state also when NOHZ is inactive
Eero Nurkkala [Wed, 7 Oct 2009 08:54:26 +0000 (11:54 +0300)]
NOHZ: update idle state also when NOHZ is inactive

Commit f2e21c9610991e95621a81407cdbab881226419b had unfortunate side
effects with cpufreq governors on some systems.

If the system did not switch into NOHZ mode ts->inidle is not set when
tick_nohz_stop_sched_tick() is called from the idle routine. Therefor
all subsequent calls from irq_exit() to tick_nohz_stop_sched_tick()
fail to call tick_nohz_start_idle(). This results in bogus idle
accounting information which is passed to cpufreq governors.

Set the inidle flag unconditionally of the NOHZ active state to keep
the idle time accounting correct in any case.

[ tglx: Added comment and tweaked the changelog ]

Reported-by: Steven Noonan <steven@uplinklabs.net>
Signed-off-by: Eero Nurkkala <ext-eero.nurkkala@nokia.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Cc: Greg KH <greg@kroah.com>
Cc: Steven Noonan <steven@uplinklabs.net>
Cc: stable@kernel.org
LKML-Reference: <1254907901.30157.93.camel@eenurkka-desktop>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
15 years agoethoc: limit the number of buffers to 128
Thomas Chou [Tue, 6 Oct 2009 03:25:25 +0000 (03:25 +0000)]
ethoc: limit the number of buffers to 128

Only 128 buffer descriptors are supported in the core. Limit the
number in case we have more memory.

Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoethoc: use system memory as buffer
Thomas Chou [Sun, 4 Oct 2009 23:33:20 +0000 (23:33 +0000)]
ethoc: use system memory as buffer

This patch enabled the ethoc to allocate system memory as buffer
when there is no dedicated buffer memory.

Some hardware designs may not have dedicated buffer memory such as
on chip or off chip SRAM. In this case, only one memory resource is
supplied in the platform data instead of two. Then a DMA buffer can
be allocated from system memory and used for the transfer.

Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoethoc: align received packet to make IP header at word boundary
Thomas Chou [Sun, 4 Oct 2009 23:33:19 +0000 (23:33 +0000)]
ethoc: align received packet to make IP header at word boundary

The packet buffer is allocated at 4 bytes boundary, but the IP header
length and version bits is located at byte 14. These bit fields access
as 32 bits word and caused exception on processors that do not support
unaligned access.

The patch adds 2 bytes offset to make the bit fields word aligned.

Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoethoc: fix buffer address mapping
Thomas Chou [Sun, 4 Oct 2009 23:33:18 +0000 (23:33 +0000)]
ethoc: fix buffer address mapping

The pointer address in buffer descriptors is physical address. The
pointer that processor used to access packet is virtual address.

Though the higher bits of pointer address used by the MAC may be
truncated to zero in special case, it is not always true in larger
designs.

Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoethoc: fix typo to compute number of tx descriptors
Thomas Chou [Sun, 4 Oct 2009 23:33:17 +0000 (23:33 +0000)]
ethoc: fix typo to compute number of tx descriptors

It should be max() instead of min(). Use 1/4 of available
descriptors for tx, and there should be at least 2 tx
descriptors.

Signed-off-by: Thomas Chou <thomas@wytron.com.tw>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoau1000_eth: Duplicate test of RX_OVERLEN bit in update_rx_stats()
roel kluin [Tue, 6 Oct 2009 09:54:18 +0000 (09:54 +0000)]
au1000_eth: Duplicate test of RX_OVERLEN bit in update_rx_stats()

in update_rx_stats() the RX_OVERLEN bit is set twice, replace it by RX_RUNT.
in au1000_rx() the RX_MISSED_FRAME bit was tested a few lines earlier already

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Acked-by: Manuel Lauss <manuel.lauss@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agonetxen: Fix Unlikely(x) > y
Roel Kluin [Tue, 6 Oct 2009 19:34:39 +0000 (19:34 +0000)]
netxen: Fix Unlikely(x) > y

The closing parenthesis was not on the right location.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Acked-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agopasemi_mac: ethtool get settings fix
Valentine Barshak [Mon, 5 Oct 2009 03:27:56 +0000 (03:27 +0000)]
pasemi_mac: ethtool get settings fix

Not all pasemi mac interfaces can have a phy attached.
For example, XAUI has no phy and phydev is NULL for it.
In this case ethtool get settings causes kernel crash.
Fix it by returning -EOPNOTSUPP if there's no PHY attached.

Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com>
Acked-by: Olof Johansson <olof@lixom.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agoadd maintainer for network drop monitor kernel service
Neil Horman [Mon, 5 Oct 2009 03:56:55 +0000 (03:56 +0000)]
add maintainer for network drop monitor kernel service

I was getting ribbed about this earlier, so I figured I'd make it
official.  Add myself as the maintainer of the drop monitor bits, so
people don't just gripe at Dave when it breaks (I'm sure it will never
break, but just in case :) ).

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agotg3: Fix phylib locking strategy
Matt Carlson [Mon, 5 Oct 2009 17:55:29 +0000 (17:55 +0000)]
tg3: Fix phylib locking strategy

Felix Radensky noted that chip resets were generating stack trace dumps.
This is because the driver is attempting to acquire the mdio bus mutex
while holding the tp->lock spinlock.  The fix is to change the code such
that every phy access takes the tp->lock spinlock instead.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agorndis_host: support ETHTOOL_GPERMADDR
John W. Linville [Tue, 6 Oct 2009 11:05:42 +0000 (11:05 +0000)]
rndis_host: support ETHTOOL_GPERMADDR

Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agofutex: Fix typo in FUTEX_WAIT/WAKE_BITSET_PRIVATE definitions
Anton Blanchard [Wed, 7 Oct 2009 00:13:58 +0000 (11:13 +1100)]
futex: Fix typo in FUTEX_WAIT/WAKE_BITSET_PRIVATE definitions

Looks like a typo, FUTEX_WAKE_BITS should be FUTEX_WAIT_BITSET.

Signed-off-by: Anton Blanchard <anton@samba.org>
LKML-Reference: <20091007001358.GE16073@kryten>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
15 years agoipv4: arp_notify address list bug
Stephen Hemminger [Wed, 7 Oct 2009 10:18:17 +0000 (03:18 -0700)]
ipv4: arp_notify address list bug

This fixes a bug with arp_notify.

If arp_notify is enabled, kernel will crash if address is changed
and no IP address is assigned.
  http://bugzilla.kernel.org/show_bug.cgi?id=14330

Reported-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agorcu: Place root rcu_node structure in separate lockdep class
Paul E. McKenney [Wed, 7 Oct 2009 04:48:18 +0000 (21:48 -0700)]
rcu: Place root rcu_node structure in separate lockdep class

Before this patch, all of the rcu_node structures were in the same lockdep
class, so that lockdep would complain when rcu_preempt_offline_tasks()
acquired the root rcu_node structure's lock while holding one of the leaf
rcu_nodes' locks.

This patch changes rcu_init_one() to use a separate
spin_lock_init() for the root rcu_node structure's lock than is
used for that of all of the rest of the rcu_node structures, which
puts the root rcu_node structure's lock in its own lockdep class.

Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: akpm@linux-foundation.org
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
LKML-Reference: <12548908983277-git-send-email->
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agorcu: Make hot-unplugged CPU relinquish its own RCU callbacks
Paul E. McKenney [Wed, 7 Oct 2009 04:48:17 +0000 (21:48 -0700)]
rcu: Make hot-unplugged CPU relinquish its own RCU callbacks

The current interaction between RCU and CPU hotplug requires that
RCU block in CPU notifiers waiting for callbacks to drain.

This can be greatly simplified by having each CPU relinquish its
own callbacks, and for both _rcu_barrier() and CPU_DEAD notifiers
to adopt all callbacks that were previously relinquished.

This change also eliminates the possibility of certain types of
hangs due to the previous practice of waiting for callbacks to be
invoked from within CPU notifiers.  If you don't every wait, you
cannot hang.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: akpm@linux-foundation.org
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
LKML-Reference: <1254890898456-git-send-email->
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agorcu: Move rcu_barrier() to rcutree
Paul E. McKenney [Wed, 7 Oct 2009 04:48:16 +0000 (21:48 -0700)]
rcu: Move rcu_barrier() to rcutree

Move the existing rcu_barrier() implementation to rcutree.c,
consistent with the fact that the rcu_barrier() implementation is
tied quite tightly to the RCU implementation.

This opens the way to simplify and fix rcutree.c's rcu_barrier()
implementation in a later patch.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: akpm@linux-foundation.org
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
LKML-Reference: <12548908982563-git-send-email->
Signed-off-by: Ingo Molnar <mingo@elte.hu>
15 years agogigaset: add kerneldoc comments
Tilman Schmidt [Tue, 6 Oct 2009 12:19:01 +0000 (12:19 +0000)]
gigaset: add kerneldoc comments

Add kerneldoc comments to some functions in the Gigaset driver.

Impact: documentation
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agogigaset: correct debugging output selection
Tilman Schmidt [Tue, 6 Oct 2009 12:18:56 +0000 (12:18 +0000)]
gigaset: correct debugging output selection

Dump payload data consistently only when DEBUG_STREAM_DUMP debug bit
is set.

Impact: debugging aid
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agogigaset: improve error recovery
Tilman Schmidt [Tue, 6 Oct 2009 12:18:51 +0000 (12:18 +0000)]
gigaset: improve error recovery

When the Gigaset base stops responding, try resetting the USB
connection to recover.

Impact: error handling improvement
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agogigaset: fix device ERROR response handling
Tilman Schmidt [Tue, 6 Oct 2009 12:18:46 +0000 (12:18 +0000)]
gigaset: fix device ERROR response handling

Clear out pending command that got rejected with 'ERROR' response.
This fixes the bug where unloading the driver module would hang
with the message: "gigaset: not searching scheduled commands: busy"
after a device communication error.

Impact: error handling bugfix
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agogigaset: announce if built with debugging
Tilman Schmidt [Tue, 6 Oct 2009 12:18:41 +0000 (12:18 +0000)]
gigaset: announce if built with debugging

Mention in the driver load announcement whether the driver was built
with debugging messages enabled, to facilitate support.

Impact: informational message
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>
15 years agogigaset: handle isoc frame errors more gracefully
Tilman Schmidt [Tue, 6 Oct 2009 12:18:36 +0000 (12:18 +0000)]
gigaset: handle isoc frame errors more gracefully

Don't drop the remainder of an URB if an isochronous frame has an error.

Impact: error handling improvement
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Signed-off-by: David S. Miller <davem@davemloft.net>