Re: [PATCH 3/8] sched/fair: Tweak misfit-related capacity checks

[Vincent Guittot]

On Fri, 5 Feb 2021 at 18:00, Valentin Schneider
<valentin.schneider@xxxxxxx> wrote:
>
> On 05/02/21 15:31, Vincent Guittot wrote:
> > On Thu, 28 Jan 2021 at 19:32, Valentin Schneider
> > <valentin.schneider@xxxxxxx> wrote:
> >>
> >> During load-balance, groups classified as group_misfit_task are filtered
> >> out if they do not pass
> >>
> >>   group_smaller_max_cpu_capacity(<candidate group>, <local group>);
> >
> > group_smaller_max_cpu_capacity and  group_smaller_max_cpu_capacity are
> > removed in the next patch. Merge this and the next and directly remove
> > them
> >
>
> OK.
>
> >> @@ -8253,7 +8260,7 @@ check_cpu_capacity(struct rq *rq, struct sched_domain *sd)
> >>  static inline int check_misfit_status(struct rq *rq, struct sched_domain *sd)
> >>  {
> >>         return rq->misfit_task_load &&
> >> -               (rq->cpu_capacity_orig < rq->rd->max_cpu_capacity ||
> >> +               (capacity_greater(rq->rd->max_cpu_capacity, rq->cpu_capacity_orig) ||
> >
> > Why do you add a margin here whereas there was no margin before ?
> >
>
> Comparing capacities without any sort of filter can lead to ping-ponging
> tasks around (capacity values very easily fluctuate by +/- 1, if not more).

max_cpu_capacity reflects the max of the cpu_capacity_orig values
don't aim to change and can be considered as static values.
It would be better to fix this rounding problem (if any) in
topology_get_cpu_scale instead of computing a margin every time it's
used

> I'm guilty of doing two things at once here: replace existing users, and
> convert callsites that should be existing users. I can split the conversion
> in a separate patch.