mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "tip-bot2 for Peter Zijlstra" <tip-bot2@linutronix.de>
To: linux-tip-commits@vger.kernel.org
Cc: "Peter Zijlstra (Intel)" <peterz@infradead.org>,
	x86@kernel.org, linux-kernel@vger.kernel.org
Subject: [tip: sched/urgent] sched/fair: Fix update_cfs_group() vs DELAY_DEQUEUE
Date: Mon, 13 Jan 2025 12:58:52 -0000	[thread overview]
Message-ID: <173677313261.399.16609836040504211717.tip-bot2@tip-bot2> (raw)
In-Reply-To: <20250110115720.GA17405@noisy.programming.kicks-ass.net>

The following commit has been merged into the sched/urgent branch of tip:

Commit-ID:     66951e4860d3c688bfa550ea4a19635b57e00eca
Gitweb:        https://git.kernel.org/tip/66951e4860d3c688bfa550ea4a19635b57e00eca
Author:        Peter Zijlstra <peterz@infradead.org>
AuthorDate:    Mon, 13 Jan 2025 13:50:11 +01:00
Committer:     Peter Zijlstra <peterz@infradead.org>
CommitterDate: Mon, 13 Jan 2025 13:50:56 +01:00

sched/fair: Fix update_cfs_group() vs DELAY_DEQUEUE

Normally dequeue_entities() will continue to dequeue an empty group entity;
except DELAY_DEQUEUE changes things -- it retains empty entities such that they
might continue to compete and burn off some lag.

However, doing this results in update_cfs_group() re-computing the cgroup
weight 'slice' for an empty group, which it (rightly) figures isn't much at
all. This in turn means that the delayed entity is not competing at the
expected weight. Worse, the very low weight causes its lag to be inflated,
which combined with avg_vruntime() using scale_load_down(), leads to artifacts.

As such, don't adjust the weight for empty group entities and let them compete
at their original weight.

Fixes: 152e11f6df29 ("sched/fair: Implement delayed dequeue")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250110115720.GA17405@noisy.programming.kicks-ass.net
---
 kernel/sched/fair.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index eeed8e3..2695843 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -3956,7 +3956,11 @@ static void update_cfs_group(struct sched_entity *se)
 	struct cfs_rq *gcfs_rq = group_cfs_rq(se);
 	long shares;
 
-	if (!gcfs_rq)
+	/*
+	 * When a group becomes empty, preserve its weight. This matters for
+	 * DELAY_DEQUEUE.
+	 */
+	if (!gcfs_rq || !gcfs_rq->load.weight)
 		return;
 
 	if (throttled_hierarchy(gcfs_rq))

  parent reply	other threads:[~2025-01-13 12:58 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-12-29 22:51 [REGRESSION] Re: [PATCH 00/24] Complete EEVDF Doug Smythies
2025-01-06 11:57 ` Peter Zijlstra
2025-01-06 15:01   ` Doug Smythies
2025-01-06 16:59     ` Peter Zijlstra
2025-01-06 17:04       ` Peter Zijlstra
2025-01-06 17:14         ` Peter Zijlstra
2025-01-07  1:24           ` Doug Smythies
2025-01-07 10:49             ` Peter Zijlstra
2025-01-06 22:28         ` Doug Smythies
2025-01-07 11:26           ` Peter Zijlstra
2025-01-07 15:04             ` Doug Smythies
2025-01-07 16:25               ` Doug Smythies
2025-01-07 19:23             ` Peter Zijlstra
2025-01-08  5:15               ` Doug Smythies
2025-01-08 13:12                 ` Peter Zijlstra
2025-01-08 15:48                   ` Doug Smythies
2025-01-09 10:59                     ` Peter Zijlstra
2025-01-09 12:18                       ` [tip: sched/urgent] sched/fair: Fix EEVDF entity placement bug causing scheduling lag tip-bot2 for Peter Zijlstra
2025-04-17  9:56                         ` Alexander Egorenkov
2025-04-22  5:40                           ` ll"RE: " Doug Smythies
2025-04-24  7:56                             ` Alexander Egorenkov
2025-04-26 15:09                               ` Doug Smythies
2025-01-10  5:09                       ` [REGRESSION] Re: [PATCH 00/24] Complete EEVDF Doug Smythies
2025-01-10 11:57                         ` Peter Zijlstra
2025-01-12 23:14                           ` Doug Smythies
2025-01-13 11:03                             ` Peter Zijlstra
2025-01-14 10:58                               ` Peter Zijlstra
2025-01-14 15:15                                 ` Doug Smythies
2025-01-15  2:08                                   ` Len Brown
2025-01-15 16:47                                     ` Doug Smythies
2025-01-19  0:09                                 ` Doug Smythies
2025-01-20  3:55                                   ` Doug Smythies
2025-01-21 11:06                                     ` Peter Zijlstra
2025-01-21  8:49                                   ` Peter Zijlstra
2025-01-21 11:21                                     ` Peter Zijlstra
2025-01-21 15:58                                       ` Doug Smythies
2025-01-24  4:34                                         ` Doug Smythies
2025-01-24 11:04                                           ` Peter Zijlstra
2025-01-13 11:05                             ` Peter Zijlstra
2025-01-13 16:01                               ` Doug Smythies
2025-01-13 12:58                           ` tip-bot2 for Peter Zijlstra [this message]
2025-01-12 19:59                         ` Doug Smythies

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=173677313261.399.16609836040504211717.tip-bot2@tip-bot2 \
    --to=tip-bot2@linutronix.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-tip-commits@vger.kernel.org \
    --cc=peterz@infradead.org \
    --cc=x86@kernel.org \
    /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

all inboxes | Powered by JetHome®