Re: [PATCH] fix scheduler regression from "sched/fair: Rework load_balance()"

[Vincent Guittot]

On Fri, 30 Oct 2020 at 03:11, Rik van Riel <riel@xxxxxxxxxxx> wrote:
>
> On Mon, 2020-10-26 at 17:52 +0100, Vincent Guittot wrote:
> > On Mon, 26 Oct 2020 at 17:48, Chris Mason <clm@xxxxxx> wrote:
> > > On 26 Oct 2020, at 12:20, Vincent Guittot wrote:
> > >
> > > > what you are suggesting is something like:
> > > > diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> > > > index 4978964e75e5..3b6fbf33abc2 100644
> > > > --- a/kernel/sched/fair.c
> > > > +++ b/kernel/sched/fair.c
> > > > @@ -9156,7 +9156,8 @@ static inline void
> > > > calculate_imbalance(struct
> > > > lb_env *env, struct sd_lb_stats *s
> > > >          * emptying busiest.
> > > >          */
> > > >         if (local->group_type == group_has_spare) {
> > > > -               if (busiest->group_type > group_fully_busy) {
> > > > +               if ((busiest->group_type > group_fully_busy) &&
> > > > +                   !(env->sd->flags & SD_SHARE_PKG_RESOURCES)) {
> > > >                         /*
> > > >                          * If busiest is overloaded, try to fill
> > > > spare
> > > >                          * capacity. This might end up creating
> > > > spare
> > > > capacity
> > > >
> > > > which also fixes the problem for me and alignes LB with wakeup
> > > > path
> > > > regarding the migration
> > > > in the LLC
> > >
> > > Vincent’s patch on top of 5.10-rc1 looks pretty great:
> > >
> > > Latency percentiles (usec) runtime 90 (s) (3320 total samples)
> > >          50.0th: 161 (1687 samples)
> > >          75.0th: 200 (817 samples)
> > >          90.0th: 228 (488 samples)
> > >          95.0th: 254 (164 samples)
> > >          *99.0th: 314 (131 samples)
> > >          99.5th: 330 (17 samples)
> > >          99.9th: 356 (13 samples)
> > >          min=29, max=358
> > >
> > > Next we test in prod, which probably won’t have answers until
> > > tomorrow.  Thanks again Vincent!
> >
> > Great !
> >
> > I'm going to run more tests on my setup as well to make sure that it
> > doesn't generate unexpected side effects on other kinds of use cases.
>
> We have tested the patch with several pretty demanding
> workloads for the past several days, and it seems to
> do the trick!
>
> With all the current scheduler code from the Linus tree,
> plus this patch on top, performance is as good as it ever
> was before with one workload, and slightly better with
> the other.

Thanks for the test results.
I still have a few tests to run on my systems but current results look
good for me too.


>
> --
> All Rights Reversed.