kernel/lib/kernel_lock.c, branch linux-2.6.34.y

bkl: Fixup core_lock fallout

2009-12-14T22:55:33Z

kernel_lock.c emits a warning because a raw spinlock function is used with a spinlock. Convert BKL to raw_spinlock. Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra Acked-by: Ingo Molnar

locking: Further name space cleanups

2009-12-14T22:55:33Z

The name space hierarchy for the internal lock functions is now a bit backwards. raw_spin* functions map to _spin* which use __spin*, while we would like to have _raw_spin* and __raw_spin*. _raw_spin* is already used by lock debugging, so rename those funtions to do_raw_spin* to free up the _raw_spin* name space. No functional change. Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra Acked-by: Ingo Molnar

[BKL] add 'might_sleep()' to the outermost lock taker

2009-12-12T22:46:33Z

As shown by the previous patch (6698e3472: "tty: Fix BKL taken under a spinlock bug introduced in the BKL split") the BKL removal is prone to some subtle issues, where removing the BKL in one place may in fact make a previously nested BKL call the new outer call, and then prone to nasty deadlocks with other spinlocks. In general, we should never take the BKL while we're holding a spinlock, so let's just add a "might_sleep()" to it (even though the BKL doesn't technically sleep - at least not yet), and we'll get nice warnings the next time this kind of problem happens during BKL removal. Acked-and-Tested-by: Thomas Gleixner Signed-off-by: Linus Torvalds

tracing: Pushdown the bkl tracepoints calls

2009-09-28T16:00:48Z

Currently we are calling the bkl tracepoint callbacks just before the bkl lock/unlock operations, ie the tracepoint call is not inside a lock_kernel() function but inside a lock_kernel() macro. Hence the bkl trace event header must be included from smp_lock.h. This raises some nasty circular header dependencies: linux/smp_lock.h -> trace/events/bkl.h -> trace/define_trace.h -> trace/ftrace.h -> linux/ftrace_event.h -> linux/hardirq.h -> linux/smp_lock.h This results in incomplete event declarations, spurious event definitions and other kind of funny behaviours. This is hardly fixable without ugly workarounds. So instead, we push the file name, line number and function name as lock_kernel() parameters, so that we only deal with the trace event header from lib/kernel_lock.c This adds two parameters to lock_kernel() and unlock_kernel() but it should be fine wrt to performances because this pair dos not seem to be called in fast paths. Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt Cc: Ingo Molnar Cc: Li Zefan

tracing/bkl: Add bkl ftrace events

2009-09-24T13:16:31Z

Add two events lock_kernel and unlock_kernel() to trace the bkl uses. This opens the door for userspace tools to perform statistics about the callsites that use it, dependencies with other locks (by pairing the trace with lock events), use with recursivity and so on... The {__reacquire,release}_kernel_lock() events are not traced because these are called from schedule, thus the sched events are sufficient to trace them. Example of a trace: hald-addon-stor-4152 [000] 165.875501: unlock_kernel: depth: 0, fs/block_dev.c:1358 __blkdev_put() hald-addon-stor-4152 [000] 167.832974: lock_kernel: depth: 0, fs/block_dev.c:1167 __blkdev_get() How to get the callsites that acquire it recursively: cd /debug/tracing/events/bkl echo "lock_depth > 0" > filter firefox-4951 [001] 206.276967: unlock_kernel: depth: 1, fs/reiserfs/super.c:575 reiserfs_dirty_inode() You can also filter by file and/or line. v2: Use of FILTER_PTR_STRING attribute for files and lines fields to make them traceable. Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt Cc: Li Zefan

sched: TIF_NEED_RESCHED -> need_reshed() cleanup

2009-03-06T11:48:55Z

Impact: cleanup Use test_tsk_need_resched(), set_tsk_need_resched(), need_resched() instead of using TIF_NEED_RESCHED. Signed-off-by: Lai Jiangshan Cc: Peter Zijlstra LKML-Reference: <49B10BA4.9070209@cn.fujitsu.com> Signed-off-by: Ingo Molnar

BKL: revert back to the old spinlock implementation

2008-05-11T03:58:02Z

The generic semaphore rewrite had a huge performance regression on AIM7 (and potentially other BKL-heavy benchmarks) because the generic semaphores had been rewritten to be simple to understand and fair. The latter, in particular, turns a semaphore-based BKL implementation into a mess of scheduling. The attempt to fix the performance regression failed miserably (see the previous commit 00b41ec2611dc98f87f30753ee00a53db648d662 'Revert "semaphore: fix"'), and so for now the simple and sane approach is to instead just go back to the old spinlock-based BKL implementation that never had any issues like this. This patch also has the advantage of being reported to fix the regression completely according to Yanmin Zhang, unlike the semaphore hack which still left a couple percentage point regression. As a spinlock, the BKL obviously has the potential to be a latency issue, but it's not really any different from any other spinlock in that respect. We do want to get rid of the BKL asap, but that has been the plan for several years. These days, the biggest users are in the tty layer (open/release in particular) and Alan holds out some hope: "tty release is probably a few months away from getting cured - I'm afraid it will almost certainly be the very last user of the BKL in tty to get fixed as it depends on everything else being sanely locked." so while we're not there yet, we do have a plan of action. Tested-by: Yanmin Zhang Cc: Ingo Molnar Cc: Andi Kleen Cc: Matthew Wilcox Cc: Alexander Viro Cc: Andrew Morton Signed-off-by: Linus Torvalds

Convert asm/semaphore.h users to linux/semaphore.h

2008-04-19T02:22:54Z

Signed-off-by: Matthew Wilcox

Add semaphore.h to kernel_lock.c

2008-04-17T14:42:27Z

kernel_lock.c uses DECLARE_MUTEX, up() and down() without explicitly including asm/semaphore.h. This is fragile and leaves it vulnerable to breakage during header reorganisations. Signed-off-by: Matthew Wilcox

sched: remove the !PREEMPT_BKL code

2008-01-25T20:08:33Z

remove the !PREEMPT_BKL code. this removes 160 lines of legacy code. Signed-off-by: Ingo Molnar