Commit 927b54fc authored by Ben Segall's avatar Ben Segall Committed by Ingo Molnar

sched: Fix hrtimer_cancel()/rq->lock deadlock

__start_cfs_bandwidth calls hrtimer_cancel while holding rq->lock,
waiting for the hrtimer to finish. However, if sched_cfs_period_timer
runs for another loop iteration, the hrtimer can attempt to take
rq->lock, resulting in deadlock.

Fix this by ensuring that cfs_b->timer_active is cleared only if the
_latest_ call to do_sched_cfs_period_timer is returning as idle. Then
__start_cfs_bandwidth can just call hrtimer_try_to_cancel and wait for
that to succeed or timer_active == 1.
Signed-off-by: default avatarBen Segall <bsegall@google.com>
Signed-off-by: default avatarPeter Zijlstra <peterz@infradead.org>
Cc: pjt@google.com
Link: http://lkml.kernel.org/r/20131016181622.22647.16643.stgit@sword-of-the-dawn.mtv.corp.google.comSigned-off-by: default avatarIngo Molnar <mingo@kernel.org>
parent db06e78c
...@@ -3225,6 +3225,13 @@ static int do_sched_cfs_period_timer(struct cfs_bandwidth *cfs_b, int overrun) ...@@ -3225,6 +3225,13 @@ static int do_sched_cfs_period_timer(struct cfs_bandwidth *cfs_b, int overrun)
if (idle) if (idle)
goto out_unlock; goto out_unlock;
/*
* if we have relooped after returning idle once, we need to update our
* status as actually running, so that other cpus doing
* __start_cfs_bandwidth will stop trying to cancel us.
*/
cfs_b->timer_active = 1;
__refill_cfs_bandwidth_runtime(cfs_b); __refill_cfs_bandwidth_runtime(cfs_b);
if (!throttled) { if (!throttled) {
...@@ -3493,11 +3500,11 @@ void __start_cfs_bandwidth(struct cfs_bandwidth *cfs_b) ...@@ -3493,11 +3500,11 @@ void __start_cfs_bandwidth(struct cfs_bandwidth *cfs_b)
* (timer_active==0 becomes visible before the hrtimer call-back * (timer_active==0 becomes visible before the hrtimer call-back
* terminates). In either case we ensure that it's re-programmed * terminates). In either case we ensure that it's re-programmed
*/ */
while (unlikely(hrtimer_active(&cfs_b->period_timer))) { while (unlikely(hrtimer_active(&cfs_b->period_timer)) &&
hrtimer_try_to_cancel(&cfs_b->period_timer) < 0) {
/* bounce the lock to allow do_sched_cfs_period_timer to run */
raw_spin_unlock(&cfs_b->lock); raw_spin_unlock(&cfs_b->lock);
/* ensure cfs_b->lock is available while we wait */ cpu_relax();
hrtimer_cancel(&cfs_b->period_timer);
raw_spin_lock(&cfs_b->lock); raw_spin_lock(&cfs_b->lock);
/* if someone else restarted the timer then we're done */ /* if someone else restarted the timer then we're done */
if (cfs_b->timer_active) if (cfs_b->timer_active)
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment