lkml.org 
[lkml]   [2011]   [Apr]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [patch 09/15] sched: add exports tracking cfs bandwidth control statistics
FromPeter Zijlstra <>
DateTue, 05 Apr 2011 15:28:11 +0200
On Tue, 2011-03-22 at 20:03 -0700, Paul Turner wrote:
> @@ -1431,10 +1432,15 @@ static void unthrottle_cfs_rq(struct cfs
> struct rq *rq = rq_of(cfs_rq);
> struct sched_entity *se;
> struct tg_unthrottle_down_data udd;
> + struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg);
>
> se = cfs_rq->tg->se[cpu_of(rq_of(cfs_rq))];
>
> update_rq_clock(rq);
> + raw_spin_lock(&cfs_b->lock);
> + cfs_b->throttled_time += (rq->clock - cfs_rq->throttled_timestamp);
> + raw_spin_unlock(&cfs_b->lock);
> + cfs_rq->throttled_timestamp = 0;
>
> /* don't include throttled window for load statistics */
> udd.cpu = rq->cpu;

> @@ -1523,6 +1530,11 @@ static int do_sched_cfs_period_timer(str
> raw_spin_lock(&cfs_b->lock);
> cfs_b->runtime = runtime;
> cfs_b->runtime_assigned = runtime_assigned;
> +
> + /* update throttled stats */
> + cfs_b->nr_periods++;

Aren't you under-counting the periods, shouldn't that be += overrun?

> + if (throttled)
> + cfs_b->nr_throttled++;

idem?

> raw_spin_unlock(&cfs_b->lock);
>
> return idle;



\
 
 \ /
  Last update: 2011-04-05 15:31    [from the cache]
©2003-2010