Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <DB4481A8-FD4E-4879-9CD2-275ABAFC09CF@fb.com> <CAKfTPtBiOFXwV9SkZ=YBw16xoS6LSrKVR4sFX6r2hZPZ9_5-+A@mail.gmail.com>
 <0014CA62-A632-495A-92B0-4B14C8CA193C@fb.com> <20201026142455.GA13495@vingu-book>
 <465597a2250d69346cff73dd07817794d3e80244.camel@surriel.com>
 <CAKfTPtCVzass7GM5oj3o3y0ru4HQViWZc2+D-RpFoLvg=__FrA@mail.gmail.com>
 <334f491d2887a6ed7c5347d5125412849feb8a0a.camel@surriel.com>
 <CAKfTPtAfKn0jzOpPNR4NUb0zLs02iLQq2_UCDSCEwhTB2LDAig@mail.gmail.com>
 <20201026120445.6a5dbbbe@imladris.surriel.com> <20201026162029.GA11367@vingu-book>
 <B23038A4-945F-48E9-8D38-EABE8204F208@fb.com>
In-Reply-To: <B23038A4-945F-48E9-8D38-EABE8204F208@fb.com>
From:   Vincent Guittot <vincent.guittot@linaro.org>
Date:   Mon, 26 Oct 2020 17:52:41 +0100
Message-ID: <CAKfTPtCROc_+rPYm8FGHDNJ-U6h1iZ0nm2+xy+tZ+L1q09h28w@mail.gmail.com>
Subject: Re: [PATCH] fix scheduler regression from "sched/fair: Rework load_balance()"
To:     Chris Mason <clm@fb.com>
Cc:     Rik van Riel <riel@surriel.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Johannes Weiner <hannes@cmpxchg.org>,
        linux-kernel <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, 26 Oct 2020 at 17:48, Chris Mason <clm@fb.com> wrote:
>
> On 26 Oct 2020, at 12:20, Vincent Guittot wrote:
>
> > Le lundi 26 oct. 2020 =C3=A0 12:04:45 (-0400), Rik van Riel a =C3=A9cri=
t :
> >> On Mon, 26 Oct 2020 16:42:14 +0100
> >> Vincent Guittot <vincent.guittot@linaro.org> wrote:
> >>> On Mon, 26 Oct 2020 at 16:04, Rik van Riel <riel@surriel.com> wrote:
> >>
> >>>> Could utilization estimates be off, either lagging or
> >>>> simply having a wrong estimate for a task, resulting
> >>>> in no task getting pulled sometimes, while doing a
> >>>> migrate_task imbalance always moves over something?
> >>>
> >>> task and cpu utilization are not always up to fully synced and may
> >>> lag
> >>> a bit which explains that sometimes LB can fail to migrate for a
> >>> small
> >>> diff
> >>
> >> OK, running with this little snippet below, I see latencies
> >> improve back to near where they used to be:
> >>
> >> Latency percentiles (usec) runtime 150 (s)
> >>         50.0th: 13
> >>         75.0th: 31
> >>         90.0th: 69
> >>         95.0th: 90
> >>         *99.0th: 761
> >>         99.5th: 2268
> >>         99.9th: 9104
> >>         min=3D1, max=3D16158
> >>
> >> I suspect the right/cleaner approach might be to use
> >> migrate_task more in !CPU_NOT_IDLE cases?
> >>
> >> Running a task to an idle CPU immediately, instead of refusing
> >> to have the load balancer move it, improves latencies for fairly
> >> obvious reasons.
> >>
> >> I am not entirely clear on why the load balancer should need to
> >> be any more conservative about moving tasks than the wakeup
> >> path is in eg. select_idle_sibling.
> >
> >
> > 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 =3D=3D 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=E2=80=99s 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=3D29, max=3D358
>
> Next we test in prod, which probably won=E2=80=99t 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.

>
> -chris