LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Vincent Guittot <vincent.guittot@linaro.org>
To: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Benjamin Segall <bsegall@google.com>,
	Mel Gorman <mgorman@techsingularity.net>,
	Peter Zijlstra <peterz@infradead.org>,
	Daniel Bristot de Oliveira <bristot@redhat.com>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	Josh Don <joshdon@google.com>, Juri Lelli <juri.lelli@redhat.com>,
	kvm@vger.kernel.org, linux-kernel <linux-kernel@vger.kernel.org>,
	linux-s390@vger.kernel.org,
	Rasmus Villemoes <linux@rasmusvillemoes.dk>,
	Mel Gorman <mgorman@suse.de>, Ingo Molnar <mingo@kernel.org>,
	Steven Rostedt <rostedt@goodmis.org>,
	Valentin Schneider <valentin.schneider@arm.com>
Subject: Re: [PATCH 1/1] sched/fair: improve yield_to vs fairness
Date: Tue, 10 Aug 2021 10:49:03 +0200	[thread overview]
Message-ID: <CAKfTPtCwP9cV0AQCuwUDAMPvXYdR9Bi0QMbbonCtVbFRpUv0XQ@mail.gmail.com> (raw)
In-Reply-To: <d9543747-c75f-28c2-6af3-8d9a134717a6@de.ibm.com>

On Wed, 28 Jul 2021 at 18:24, Christian Borntraeger
<borntraeger@de.ibm.com> wrote:
>
>
>
> On 27.07.21 20:57, Benjamin Segall wrote:
> > Christian Borntraeger <borntraeger@de.ibm.com> writes:
> >
> >> On 23.07.21 18:21, Mel Gorman wrote:
> >>> On Fri, Jul 23, 2021 at 02:36:21PM +0200, Christian Borntraeger wrote:
> >>>>> sched: Do not select highest priority task to run if it should be skipped
> >>>>>
> >>>>> <SNIP>
> >>>>>
> >>>>> index 44c452072a1b..ddc0212d520f 100644
> >>>>> --- a/kernel/sched/fair.c
> >>>>> +++ b/kernel/sched/fair.c
> >>>>> @@ -4522,7 +4522,8 @@ pick_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *curr)
> >>>>>                           se = second;
> >>>>>           }
> >>>>> - if (cfs_rq->next && wakeup_preempt_entity(cfs_rq->next, left) < 1) {
> >>>>> + if (cfs_rq->next &&
> >>>>> +     (cfs_rq->skip == left || wakeup_preempt_entity(cfs_rq->next, left) < 1)) {
> >>>>>                   /*
> >>>>>                    * Someone really wants this to run. If it's not unfair, run it.
> >>>>>                    */
> >>>>>
> >>>>
> >>>> I do see a reduction in ignored yields, but from a performance aspect for my
> >>>> testcases this patch does not provide a benefit, while the the simple
> >>>>    curr->vruntime += sysctl_sched_min_granularity;
> >>>> does.
> >>> I'm still not a fan because vruntime gets distorted. From the docs
> >>>      Small detail: on "ideal" hardware, at any time all tasks would have the
> >>> same
> >>>      p->se.vruntime value --- i.e., tasks would execute simultaneously and no task
> >>>      would ever get "out of balance" from the "ideal" share of CPU time
> >>> If yield_to impacts this "ideal share" then it could have other
> >>> consequences.
> >>> I think your patch may be performing better in your test case because every
> >>> "wrong" task selected that is not the yield_to target gets penalised and
> >>> so the yield_to target gets pushed up the list.
> >>>
> >>>> I still think that your approach is probably the cleaner one, any chance to improve this
> >>>> somehow?
> >>>>
> >>> Potentially. The patch was a bit off because while it noticed that skip
> >>> was not being obeyed, the fix was clumsy and isolated. The current flow is
> >>> 1. pick se == left as the candidate
> >>> 2. try pick a different se if the "ideal" candidate is a skip candidate
> >>> 3. Ignore the se update if next or last are set
> >>> Step 3 looks off because it ignores skip if next or last buddies are set
> >>> and I don't think that was intended. Can you try this?
> >>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> >>> index 44c452072a1b..d56f7772a607 100644
> >>> --- a/kernel/sched/fair.c
> >>> +++ b/kernel/sched/fair.c
> >>> @@ -4522,12 +4522,12 @@ pick_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *curr)
> >>>                     se = second;
> >>>     }
> >>>    -        if (cfs_rq->next && wakeup_preempt_entity(cfs_rq->next, left) < 1) {
> >>> +   if (cfs_rq->next && wakeup_preempt_entity(cfs_rq->next, se) < 1) {
> >>>             /*
> >>>              * Someone really wants this to run. If it's not unfair, run it.
> >>>              */
> >>>             se = cfs_rq->next;
> >>> -   } else if (cfs_rq->last && wakeup_preempt_entity(cfs_rq->last, left) < 1) {
> >>> +   } else if (cfs_rq->last && wakeup_preempt_entity(cfs_rq->last, se) < 1) {
> >>>             /*
> >>>              * Prefer last buddy, try to return the CPU to a preempted task.
> >>>              */
> >>>
> >>
> >> This one alone does not seem to make a difference. Neither in ignored yield, nor
> >> in performance.
> >>
> >> Your first patch does really help in terms of ignored yields when
> >> all threads are pinned to one host CPU. After that we do have no ignored yield
> >> it seems. But it does not affect the performance of my testcase.
> >> I did some more experiments and I removed the wakeup_preempt_entity checks in
> >> pick_next_entity - assuming that this will result in source always being stopped
> >> and target always being picked. But still, no performance difference.
> >> As soon as I play with vruntime I do see a difference (but only without the cpu cgroup
> >> controller). I will try to better understand the scheduler logic and do some more
> >> testing. If you have anything that I should test, let me know.
> >>
> >> Christian
> >
> > If both yielder and target are in the same cpu cgroup or the cpu cgroup
> > is disabled (ie, if cfs_rq_of(p->se) matches), you could try
> >
> > if (p->se.vruntime > rq->curr->se.vruntime)
> >       swap(p->se.vruntime, rq->curr->se.vruntime)
>
> I tried that and it does not show the performance benefit. I then played with my
> patch (uses different values to add) and the benefit seems to be depending on the
> size that is being added, maybe when swapping it was just not large enough.
>
> I have to say that this is all a bit unclear what and why performance improves.
> It just seems that the cpu cgroup controller has a fair share of the performance
> problems.
>
> I also asked the performance people to run some measurements and the numbers of
> some transactional workload under KVM was
> baseline: 11813
> with much smaller sched_latency_ns and sched_migration_cost_ns: 16419

Have you also tried to increase sched_wakeup_granularity which is used
to decide whether we can preempt curr ?

> with cpu controller disabled: 15962
> with cpu controller disabled + my patch: 16782

Your patch modifies the vruntime of the task but cgroup sched_entity
stays unchanged. Scheduler starts to compare the vruntime of the
sched_entity of the group before reaching your task. That's probably
why your patch doesn't help with cgroup and will be difficult to
extend to cgroup because the yield of the task should not impact the
other entities in the group.

>
> I will be travelling the next 2 weeks, so I can continue with more debugging
> after that.
>
> Thanks for all the ideas and help so far.
>
> Christian
>
> > as well as the existing buddy flags, as an entirely fair vruntime boost
> > to the target.
> >
> > For when they aren't direct siblings, you /could/ use find_matching_se,
> > but it's much less clear that's desirable, since it would yield vruntime
> > for the entire hierarchy to the target's hierarchy.
> >

  reply	other threads:[~2021-08-10  8:49 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-12 10:14 [PATCH v2 0/9] sched: Clean up SCHED_DEBUG Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 1/9] sched/numa: Allow runtime enabling/disabling of NUMA balance without SCHED_DEBUG Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 2/9] sched: Remove sched_schedstats sysctl out from under SCHED_DEBUG Peter Zijlstra
2021-04-16 15:53   ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 3/9] sched: Dont make LATENCYTOP select SCHED_DEBUG Peter Zijlstra
2021-04-16 15:53   ` [tip: sched/core] sched: Don't " tip-bot2 for Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 4/9] sched: Move SCHED_DEBUG sysctl to debugfs Peter Zijlstra
2021-04-15 16:29   ` [PATCH] sched/debug: Rename the sched_debug parameter to sched_debug_verbose Peter Zijlstra
2021-04-19 19:26     ` Josh Don
2021-04-16 15:53   ` [tip: sched/core] sched: Move SCHED_DEBUG sysctl to debugfs tip-bot2 for Peter Zijlstra
2021-04-27 14:59   ` Christian Borntraeger
2021-04-27 15:09     ` Steven Rostedt
2021-04-27 15:17       ` Christian Borntraeger
2021-04-28  8:47       ` Peter Zijlstra
2021-04-28  8:46     ` Peter Zijlstra
2021-04-28  8:54       ` Christian Borntraeger
2021-04-28  8:58         ` Christian Borntraeger
2021-04-28  9:25         ` Peter Zijlstra
2021-04-28  9:31           ` Christian Borntraeger
2021-04-28  9:42       ` Christian Borntraeger
2021-04-28 12:38         ` Peter Zijlstra
2021-04-28 14:49           ` Christian Borntraeger
2021-07-07 12:34           ` [PATCH 0/1] Improve yield (was: sched: Move SCHED_DEBUG sysctl to debugfs) Christian Borntraeger
2021-07-07 12:34             ` [PATCH 1/1] sched/fair: improve yield_to vs fairness Christian Borntraeger
2021-07-07 18:07               ` kernel test robot
2021-07-23  9:35               ` Mel Gorman
2021-07-23 12:36                 ` Christian Borntraeger
2021-07-23 16:21                   ` Mel Gorman
2021-07-26 18:41                     ` Christian Borntraeger
2021-07-26 19:32                       ` Mel Gorman
2021-07-27  6:59                         ` Christian Borntraeger
2021-07-27 18:57                       ` Benjamin Segall
2021-07-28 16:23                         ` Christian Borntraeger
2021-08-10  8:49                           ` Vincent Guittot [this message]
2021-07-27 13:29                     ` Peter Zijlstra
2021-07-27 13:33                 ` Peter Zijlstra
2021-07-27 14:31                   ` Mel Gorman
2021-04-12 10:14 ` [PATCH v2 5/9] sched,preempt: Move preempt_dynamic to debug.c Peter Zijlstra
2021-04-16 15:53   ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 6/9] debugfs: Implement debugfs_create_str() Peter Zijlstra
2021-04-16 15:53   ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 7/9] sched,debug: Convert sysctl sched_domains to debugfs Peter Zijlstra
2021-04-13 14:55   ` Valentin Schneider
2021-04-15  9:06     ` Peter Zijlstra
2021-04-15 12:16       ` Dietmar Eggemann
2021-04-15 12:34       ` Valentin Schneider
2021-04-15 13:02         ` Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 8/9] sched: Move /proc/sched_debug " Peter Zijlstra
2021-04-16 15:53   ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2021-04-12 10:14 ` [PATCH v2 9/9] sched,fair: Alternative sched_slice() Peter Zijlstra
2021-04-12 10:26   ` Peter Zijlstra
2021-04-16 15:53   ` [tip: sched/core] " tip-bot2 for Peter Zijlstra

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAKfTPtCwP9cV0AQCuwUDAMPvXYdR9Bi0QMbbonCtVbFRpUv0XQ@mail.gmail.com \
    --to=vincent.guittot@linaro.org \
    --cc=borntraeger@de.ibm.com \
    --cc=bristot@redhat.com \
    --cc=bsegall@google.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=joshdon@google.com \
    --cc=juri.lelli@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=linux@rasmusvillemoes.dk \
    --cc=mgorman@suse.de \
    --cc=mgorman@techsingularity.net \
    --cc=mingo@kernel.org \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=valentin.schneider@arm.com \
    --subject='Re: [PATCH 1/1] sched/fair: improve yield_to vs fairness' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).