Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1533334

Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group

From Morten Rasmussen <morten.rasmussen@arm.com>
Newsgroups linux.kernel
Subject Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group
Date 2016-11-30 15:30 +0100
Message-ID <sJe3v-7fx-15@gated-at.bofh.it> (permalink)
References <sHqLw-2kA-3@gated-at.bofh.it> <sHqLw-2kA-1@gated-at.bofh.it> <sJcuK-6aW-9@gated-at.bofh.it> <sJdAt-6My-9@gated-at.bofh.it> <sJdAt-6My-11@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


On Wed, Nov 30, 2016 at 02:54:00PM +0100, Vincent Guittot wrote:
> On 30 November 2016 at 14:49, Vincent Guittot
> <vincent.guittot@linaro.org> wrote:
> > On 30 November 2016 at 13:49, Morten Rasmussen <morten.rasmussen@arm.com> wrote:
> >> On Fri, Nov 25, 2016 at 04:34:33PM +0100, Vincent Guittot wrote:
> >>> find_idlest_group() only compares the runnable_load_avg when looking for
> >>> the least loaded group. But on fork intensive use case like hackbench
> >
> > [snip]
> >
> >>> +                             min_avg_load = avg_load;
> >>> +                             idlest = group;
> >>> +                     } else if ((runnable_load < (min_runnable_load + imbalance)) &&
> >>> +                                     (100*min_avg_load > imbalance_scale*avg_load)) {
> >>> +                             /*
> >>> +                              * The runnable loads are close so we take
> >>> +                              * into account blocked load through avg_load
> >>> +                              *  which is blocked + runnable load
> >>> +                              */
> >>> +                             min_avg_load = avg_load;
> >>>                               idlest = group;
> >>>                       }
> >>>
> >>> @@ -5470,13 +5495,16 @@ find_idlest_group(struct sched_domain *sd, struct task_struct *p,
> >>>               goto no_spare;
> >>>
> >>>       if (this_spare > task_util(p) / 2 &&
> >>> -         imbalance*this_spare > 100*most_spare)
> >>> +         imbalance_scale*this_spare > 100*most_spare)
> >>>               return NULL;
> >>>       else if (most_spare > task_util(p) / 2)
> >>>               return most_spare_sg;
> >>>
> >>>  no_spare:
> >>> -     if (!idlest || 100*this_load < imbalance*min_load)
> >>> +     if (!idlest ||
> >>> +         (min_runnable_load > (this_runnable_load + imbalance)) ||
> >>> +         ((this_runnable_load < (min_runnable_load + imbalance)) &&
> >>> +                     (100*min_avg_load > imbalance_scale*this_avg_load)))
> >>
> >> I don't get why you have imbalance_scale applied to this_avg_load and
> >> not min_avg_load. IIUC, you end up preferring non-local groups?
> >
> > In fact, I have keep the same condition that is used when looping the group.
> > You're right that we should prefer local rq if avg_load are close and
> > test the condition
> > (100*this_avg_load > imbalance_scale*min_avg_load) instead
> 
> Of course the correct condition is
>  (100*this_avg_load < imbalance_scale*min_avg_load)

Agreed, I should have read the entire thread before replying :-)

Back to linux.kernel | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

[PATCH 2/2 v2] sched: use load_avg for selecting idlest group Vincent Guittot <vincent.guittot@linaro.org> - 2016-11-25 16:40 +0100
  Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Morten Rasmussen <morten.rasmussen@arm.com> - 2016-11-30 13:50 +0100
    Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Vincent Guittot <vincent.guittot@linaro.org> - 2016-11-30 15:00 +0100
      Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Vincent Guittot <vincent.guittot@linaro.org> - 2016-11-30 15:00 +0100
        Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Morten Rasmussen <morten.rasmussen@arm.com> - 2016-11-30 15:30 +0100
          Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Vincent Guittot <vincent.guittot@linaro.org> - 2016-12-02 16:30 +0100
            Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Matt Fleming <matt@codeblueprint.co.uk> - 2016-12-02 23:30 +0100
      Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Morten Rasmussen <morten.rasmussen@arm.com> - 2016-11-30 15:30 +0100
  Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Brendan Gregg <brendan.d.gregg@gmail.com> - 2016-12-03 04:40 +0100
    Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Matt Fleming <matt@codeblueprint.co.uk> - 2016-12-03 23:00 +0100
      Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Vincent Guittot <vincent.guittot@linaro.org> - 2016-12-05 10:30 +0100
        Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Matt Fleming <matt@codeblueprint.co.uk> - 2016-12-05 14:40 +0100
          Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Matt Fleming <matt@codeblueprint.co.uk> - 2016-12-08 15:20 +0100
            Re: [PATCH 2/2 v2] sched: use load_avg for selecting idlest group Vincent Guittot <vincent.guittot@linaro.org> - 2016-12-08 15:40 +0100

csiph-web