mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: Vincent Guittot <vincent.guittot@linaro.org>
Cc: mingo@redhat.com, juri.lelli@redhat.com,
	dietmar.eggemann@arm.com, rostedt@goodmis.org,
	bsegall@google.com, mgorman@suse.de, vschneid@redhat.com,
	kprateek.nayak@amd.com, linux-kernel@vger.kernel.org,
	qyousef@layalina.io
Subject: Re: [PATCH 6/6 v3] sched/eevdf: Speedup short slice task scheduling
Date: Thu, 25 Jun 2026 10:33:55 +0200	[thread overview]
Message-ID: <20260625083355.GQ42921@noisy.programming.kicks-ass.net> (raw)
In-Reply-To: <20260624151229.1710703-7-vincent.guittot@linaro.org>

On Wed, Jun 24, 2026 at 05:12:29PM +0200, Vincent Guittot wrote:
> When a task with a shorter slice is enqueued, we protect the running
> task which has a longer slice until it becomes ineligible instead of a
> full slice in order to speedup the switch to other tasks until the task
> with the shortest slice is scheduled. This helps to the task to not wait
> too many full slices before running.
> 
> Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
> Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
> ---
>  kernel/sched/fair.c | 52 +++++++++++++++++++++++++++++++++++++++++++--
>  1 file changed, 50 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index f972987618e7..7c541f27a1ed 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -813,6 +813,48 @@ u64 avg_vruntime(struct cfs_rq *cfs_rq)
>  	return cfs_rq->zero_vruntime;
>  }
>  
> +/*
> + * Compute the vruntime until which the entity remains eligible when it runs
> + * or is about to run on the CPU. We use this value to set vprot to the min
> + * value until which other entities would not be picked anyway.
> + *     \Sum (v_i - v0)*w_i
> + * V = ------------------- + v0
> + *          \Sum w_i
> + *
> + * We want V' for (v_se - v0) == 0. Previous entity has already been enqueued
> + * in the rb tree and next is already dequeued so
> + *
> + *      cfs_rq->sum_w_vruntime
> + * V' = ------------------------- + v0
> + *      cfs_rq->sum_weight + w_se
> +
> + */
> +static u64 eligible_vruntime(struct cfs_rq *cfs_rq, struct sched_entity *se)
> +{
> +	struct sched_entity *curr = cfs_rq->curr;
> +	long weight = cfs_rq->sum_weight;
> +	s64 delta = 0;

'curr' goes unused in this function, did you want:

	if (curr && !curr->on_rq)
		curr = NULL;

> +
> +	if (weight) {
> +		s64 runtime = cfs_rq->sum_w_vruntime;


		if (curr) {
			unsigned long w = avg_vruntime_weight(cfs_rq, curr->load.weight);

			runtime += entity_key(cfs_rq, curr) * w;
			weight += w;
		}

?

> +
> +		weight += avg_vruntime_weight(cfs_rq, se->load.weight);
> +
> +		/* sign flips effective floor / ceiling */
> +		if (runtime < 0)
> +			runtime -= (weight - 1);
> +
> +		delta = div64_long(runtime, weight);
> +	} else {
> +		/*
> +		 * When there is but one element, it is the average.
> +		 */
> +		delta = 0;
> +	}
> +
> +	return cfs_rq->zero_vruntime + delta + 1;
> +}



  parent reply	other threads:[~2026-06-25  8:34 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-24 15:12 [PATCH 0/6 v3] sched/eevdf: Improve scheduling latency of short slice task Vincent Guittot
2026-06-24 15:12 ` [PATCH 1/6 v3] sched/fair: Set next buddy for preempt short Vincent Guittot
2026-06-25  6:24   ` K Prateek Nayak
2026-06-25 12:40     ` Vincent Guittot
2026-06-25 12:43       ` Peter Zijlstra
2026-06-26  6:54       ` Peter Zijlstra
2026-06-26  7:02         ` Vincent Guittot
2026-06-26  8:08           ` Peter Zijlstra
2026-06-26  8:55             ` Vincent Guittot
2026-06-30  9:03   ` [tip: sched/core] " tip-bot2 for Vincent Guittot
2026-06-24 15:12 ` [PATCH 2/6 v3] sched/eevdf: Take into account current's lag when updating slice protection Vincent Guittot
2026-06-30  9:03   ` [tip: sched/core] " tip-bot2 for Vincent Guittot
2026-06-24 15:12 ` [PATCH 3/6 v3] sched/eevdf: Update slice protection even when resched is already set Vincent Guittot
2026-06-26  7:21   ` Peter Zijlstra
2026-06-26  7:46     ` Peter Zijlstra
2026-06-30  9:03       ` [tip: sched/core] sched/core: Fix inter-class wakeup_preempt() tip-bot2 for Peter Zijlstra
2026-06-30  9:03   ` [tip: sched/core] sched/eevdf: Update slice protection even when resched is already set tip-bot2 for Vincent Guittot
2026-06-24 15:12 ` [PATCH 4/6 v3] sched/eevdf: Cancel slice protection if short slice task is eligible Vincent Guittot
2026-06-25  6:00   ` K Prateek Nayak
2026-06-25 12:40     ` Vincent Guittot
2026-06-26  5:51       ` Shubhang Kaushik
2026-06-26  7:39         ` Vincent Guittot
2026-06-30  9:03   ` [tip: sched/core] " tip-bot2 for Vincent Guittot
2026-06-24 15:12 ` [PATCH 5/6 v3] sched/eevdf: Always update slice protection Vincent Guittot
2026-06-30  9:03   ` [tip: sched/core] " tip-bot2 for Vincent Guittot
2026-06-24 15:12 ` [PATCH 6/6 v3] sched/eevdf: Speedup short slice task scheduling Vincent Guittot
2026-06-25  7:37   ` K Prateek Nayak
2026-06-25  8:37     ` Peter Zijlstra
2026-06-25 10:09       ` Peter Zijlstra
2026-06-25 12:57         ` Vincent Guittot
2026-06-25 12:59           ` Vincent Guittot
2026-06-25 22:28             ` Peter Zijlstra
2026-06-26  3:57               ` K Prateek Nayak
2026-06-26  6:41                 ` Peter Zijlstra
2026-06-26  7:05                 ` Vincent Guittot
2026-06-26  7:18                   ` Peter Zijlstra
2026-06-26  7:54                     ` Peter Zijlstra
2026-06-26  8:55                       ` Vincent Guittot
2026-06-26 13:55                         ` Vincent Guittot
2026-06-29 14:25                           ` Vincent Guittot
2026-06-25 14:55       ` Vincent Guittot
2026-06-25 12:51     ` Vincent Guittot
2026-06-25  8:33   ` Peter Zijlstra [this message]
2026-06-30  9:03   ` [tip: sched/core] " tip-bot2 for Vincent Guittot

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=20260625083355.GQ42921@noisy.programming.kicks-ass.net \
    --to=peterz@infradead.org \
    --cc=bsegall@google.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=juri.lelli@redhat.com \
    --cc=kprateek.nayak@amd.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mgorman@suse.de \
    --cc=mingo@redhat.com \
    --cc=qyousef@layalina.io \
    --cc=rostedt@goodmis.org \
    --cc=vincent.guittot@linaro.org \
    --cc=vschneid@redhat.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

Powered by JetHome