rtmutex: add BUG_ON if a task attempts to block on two locks
in [Kernel]
|
Prev: [PATCH net-next] drivers/net/mlx4: Use %pV, pr_<level>, printk_once
Next: parser: fix and simplify support of asm goto
From: Darren Hart on 10 Jul 2010 13:40 On 07/09/2010 05:30 PM, Steven Rostedt wrote: > On Fri, 2010-07-09 at 15:32 -0700, Darren Hart wrote: >> rtmutex proxy locking complicates the logic a bit and opens up >> the possibility for a task to wake and attempt to take another >> sleeping lock without knowing it has been enqueued on another >> lock already. Add a BUG_ON to catch this scenario early. >> >> Signed-off-by: Darren Hart<dvhltc(a)us.ibm.com> >> Cc: Thomas Gleixner<tglx(a)linutronix.de> >> Cc: Peter Zijlstra<peterz(a)infradead.org> >> Cc: Ingo Molnar<mingo(a)elte.hu> >> Cc: Eric Dumazet<eric.dumazet(a)gmail.com> >> Cc: John Kacur<jkacur(a)redhat.com> >> Cc: Steven Rostedt<rostedt(a)goodmis.org> >> Cc: Mike Galbraith<efault(a)gmx.de> >> --- >> kernel/rtmutex.c | 3 +++ >> 1 files changed, 3 insertions(+), 0 deletions(-) >> >> diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c >> index baac7d9..22f9d18 100644 >> --- a/kernel/rtmutex.c >> +++ b/kernel/rtmutex.c >> @@ -459,6 +459,9 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, >> top_waiter = rt_mutex_top_waiter(lock); >> plist_add(&waiter->list_entry,&lock->wait_list); >> >> + /* Tasks can only block on one lock at a time. */ >> + BUG_ON(task->pi_blocked_on != NULL); > > WARN_ON may be better. Since it may not cause a system crash or other > huge bug if it is not true. > No objection. From 31c7b6c5657bcc897ddc79d7f9bb1942eb4c854a Mon Sep 17 00:00:00 2001 From: Darren Hart <dvhltc(a)us.ibm.com> Date: Fri, 9 Jul 2010 17:46:05 -0400 Subject: [PATCH 2/4] rtmutex: add WARN_ON if a task attempts to block on two locks rtmutex proxy locking complicates the logic a bit and opens up the possibility for a task to wake and attempt to take another sleeping lock without knowing it has been enqueued on another lock already. Add a WARN_ON to catch this scenario early. V2: use a WARN_ON instead of a BUG_ON per Steven's request. Signed-off-by: Darren Hart <dvhltc(a)us.ibm.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Ingo Molnar <mingo(a)elte.hu> Cc: Eric Dumazet <eric.dumazet(a)gmail.com> Cc: John Kacur <jkacur(a)redhat.com> Cc: Steven Rostedt <rostedt(a)goodmis.org> Cc: Mike Galbraith <efault(a)gmx.de> --- kernel/rtmutex.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c index baac7d9..5262d0f 100644 --- a/kernel/rtmutex.c +++ b/kernel/rtmutex.c @@ -459,6 +459,9 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, top_waiter = rt_mutex_top_waiter(lock); plist_add(&waiter->list_entry, &lock->wait_list); + /* Tasks can only block on one lock at a time. */ + WARN_ON(task->pi_blocked_on != NULL); + task->pi_blocked_on = waiter; raw_spin_unlock(&task->pi_lock); -- 1.7.0.4 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |