linux-kernel - Re: [PATCH 0/4] Task based throttle follow ups

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <xhsmhzfaqwlse.mognet@vschneid-thinkpadt14sgen2i.remote.csb>
Date: Fri, 19 Sep 2025 16:37:37 +0200
From: Valentin Schneider <vschneid@...hat.com>
To: Aaron Lu <ziqianlu@...edance.com>, Ben Segall <bsegall@...gle.com>, K
 Prateek Nayak <kprateek.nayak@....com>, Peter Zijlstra
 <peterz@...radead.org>, Chengming Zhou <chengming.zhou@...ux.dev>, Josh
 Don <joshdon@...gle.com>, Ingo Molnar <mingo@...hat.com>, Vincent Guittot
 <vincent.guittot@...aro.org>, Xi Wang <xii@...gle.com>
Cc: linux-kernel@...r.kernel.org, Juri Lelli <juri.lelli@...hat.com>,
 Dietmar Eggemann <dietmar.eggemann@....com>, Steven Rostedt
 <rostedt@...dmis.org>, Mel Gorman <mgorman@...e.de>, Chuyi Zhou
 <zhouchuyi@...edance.com>, Jan Kiszka <jan.kiszka@...mens.com>, Florian
 Bezdeka <florian.bezdeka@...mens.com>, Songtang Liu
 <liusongtang@...edance.com>, Chen Yu <yu.c.chen@...el.com>, Matteo
 Martelli <matteo.martelli@...ethink.co.uk>, Michal Koutný
 <mkoutny@...e.com>, Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Subject: Re: [PATCH 0/4] Task based throttle follow ups

On 10/09/25 17:50, Aaron Lu wrote:
> Peter noticed the inconsistency in load propagation for throttled cfs_rq
> and Ben pointed out several other places regarding throttled cfs_rq that
> could be no longer needed after task based throttle model.
>
> To ease discussing and reviewing, I've come up with this follow up
> series which implements the individual changes.
>
> Patch1 deals with load propagation. According to Peter and Prateek's
> discussion, previously, load propagation for throttled cfs_rq happened
> on unthrottle time but now with per-task throttle, it's no longer the
> case so load propagation should happen immediately or we could lose this
> propagated part.
>
> Patch2 made update_cfs_group() to continue function for cfs_rqs in
> throttled hierarchy so that cfs_rq's entity can get an up2date weight. I
> think this is mostly useful when a cfs_rq in throttled hierarchy still
> has tasks running and on tick/enqueue/dequeue, update_cfs_group() can
> update this cfs_rq's entity weight.
>
> Patch3 removed special treatment of tasks in throttled hierarchy,
> including: dequeue_entities(), check_preempt_wakeup_fair() and
> yield_task_to_fair().
>
> Patch4 inhibited load balancing to a throttled cfs_rq to make hackbench
> happy.
>
> I think patch1 is needed for correctness, patch2-4 is open for
> discussion as there are pros/cons doing things either way. Comments are
> welcome, thanks.
>
> BTW, I also noticed there is the task_is_throttled sched class callback
> and in fair, it is task_is_throttled_fair(). IIUC, it is used by core
> scheduling to find a matching cookie task to run on the sibling SMT CPU.
> For this reason, it doesn't seem very useful if we find it a task that
> is to be throttled so I kept the current implementation; but I guess
> this is also two folded if that to be throttled task is holding some
> kernel resources. Anyway, I didn't write a patch to change it in this
> series, but feel free to let me know if it should be changed.
>

I saw these already made it to tip/sched/core, but FWIW that all LGTM.

Reviewed-by: Valentin Schneider <vschneid@...hat.com>