perf_counter: Fix an ipi-deadlock

perf_pending_counter() is called from IRQ context and will call
perf_counter_disable(), however perf_counter_disable() uses
smp_call_function_single() which doesn't fancy being used with
IRQs disabled due to IPI deadlocks.

Fix this by making it use the local __perf_counter_disable()
call and teaching the counter_sched_out() code about pending
disables as well.

This should cover the case where a counter migrates before the
pending queue gets processed.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Corey J Ashford <cjashfor@us.ibm.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: stephane eranian <eranian@googlemail.com>
LKML-Reference: <20090813103655.244097721@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c
index b8c6b97..3f841be 100644 (file)
--- a/kernel/perf_counter.c
+++ b/kernel/perf_counter.c
@@ -307,6 +307,10 @@ counter_sched_out(struct perf_counter *counter,
                return;
 
        counter->state = PERF_COUNTER_STATE_INACTIVE;
+       if (counter->pending_disable) {
+               counter->pending_disable = 0;
+               counter->state = PERF_COUNTER_STATE_OFF;
+       }
        counter->tstamp_stopped = ctx->time;
        counter->pmu->disable(counter);
        counter->oncpu = -1;
@@ -2343,7 +2347,7 @@ static void perf_pending_counter(struct perf_pending_entry *entry)
 
        if (counter->pending_disable) {
                counter->pending_disable = 0;
-               perf_counter_disable(counter);
+               __perf_counter_disable(counter);
        }
 
        if (counter->pending_wakeup) {