Messages in this thread | | | Date | Mon, 01 Jun 2015 18:15:47 -0400 | From | Rik van Riel <> | Subject | Re: [PATCH RESEND] sched: prefer an idle cpu vs an idle sibling for BALANCE_WAKE |
| |
On 06/01/2015 03:38 PM, Josef Bacik wrote:
> Ok I got this patch to give me the same performance as all our other > crap, just need to apply this incremental > > > diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c > index b71eb2b..e11cfec 100644 > --- a/kernel/sched/fair.c > +++ b/kernel/sched/fair.c > @@ -4761,13 +4761,10 @@ select_task_rq_fair(struct task_struct *p, int > prev_cpu, int sd_flag, int wake_f > > if (tmp->flags & sd_flag) > sd = tmp; > - else if (!want_affine || (want_affine && affine_sd)) > - break; > } > > if (affine_sd && cpu != prev_cpu && wake_affine(affine_sd, p, sync)) { > prev_cpu = cpu; > - sd = NULL; /* WAKE_AFFINE trumps BALANCE_WAKE */
Given Peter's worries about wake_affine and affine_sd, should the above be sd = affine_sd, in case select_idle_sibling cannot find an idle sibling?
That way we can attempt to at least find an idle cpu inside the affine_sd.
Of course, there may be subtleties here I am overlooking...
> } > > if (sd_flag & SD_BALANCE_WAKE) {
-- All rights reversed
| |