sched: Convert BUG_ON()s in try_to_wake_up_local() to WARN_ON_ONCE()s
Tejun Heo [Mon, 18 Mar 2013 19:22:34 +0000 (12:22 -0700)]
commit 383efcd00053ec40023010ce5034bd702e7ab373 upstream.

try_to_wake_up_local() should only be invoked to wake up another
task in the same runqueue and BUG_ON()s are used to enforce the
rule. Missing try_to_wake_up_local() can stall workqueue
execution but such stalls are likely to be finite either by
another work item being queued or the one blocked getting
unblocked.  There's no reason to trigger BUG while holding rq
lock crashing the whole system.

Convert BUG_ON()s in try_to_wake_up_local() to WARN_ON_ONCE()s.

Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20130318192234.GD3042@htj.dyndns.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

kernel/sched/core.c

index e1718bc..4b6c546 100644 (file)
@@ -1653,8 +1653,10 @@ static void try_to_wake_up_local(struct task_struct *p)
 {
        struct rq *rq = task_rq(p);
 
-       BUG_ON(rq != this_rq());
-       BUG_ON(p == current);
+       if (WARN_ON_ONCE(rq != this_rq()) ||
+           WARN_ON_ONCE(p == current))
+               return;
+
        lockdep_assert_held(&rq->lock);
 
        if (!raw_spin_trylock(&p->pi_lock)) {