linux-kernel - Re: [PATCH] sched/fair: Use rq->lock when checking cfs

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

Message-ID: <CAFpoUr1AO_qStNOYrFWGnFfc=uSFrXSYD8A5cQ8h0t2pioQzDA@mail.gmail.com>
Date:   Wed, 13 Oct 2021 19:52:34 +0100
From:   Odin Ugedal <odin@...d.al>
To:     Tao Zhou <tao.zhou@...ux.dev>
Cc:     Odin Ugedal <odin@...d.al>,
        Michal Koutný <mkoutny@...e.com>,
        open list <linux-kernel@...r.kernel.org>,
        Ingo Molnar <mingo@...hat.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Juri Lelli <juri.lelli@...hat.com>,
        Vincent Guittot <vincent.guittot@...aro.org>,
        Dietmar Eggemann <dietmar.eggemann@....com>,
        Steven Rostedt <rostedt@...dmis.org>,
        Ben Segall <bsegall@...gle.com>, Mel Gorman <mgorman@...e.de>,
        Daniel Bristot de Oliveira <bristot@...hat.com>
Subject: Re: [PATCH] sched/fair: Use rq->lock when checking cfs_rq list presence

Hi,

tir. 12. okt. 2021 kl. 19:31 skrev Tao Zhou <tao.zhou@...ux.dev>:
>
> Er.. this is considered specific to this fix I think. Normal unthrottle(not
> race with delete, avg maybe used in after) also need the normal avg decay.

Yeah, it was more meant as a way to express my idea, should probably have
said that more explicitly. It would essentially be a revert of a7b359fc6a37
("sched/fair: Correctly insert cfs_rq's to list on unthrottle"),
while "temporary" fixing the problem it fixed, by removing the load from the tg.

But yeah, it would still need to decay the load properly to _actually_ be fully
correct and work as it does today. But in most cases nr_running>0 on unthrottle,
but yeah, it is definitely not always the case.

Thanks
Odin