linux-kernel - Re: [PATCH] sched/fair: Prevent cpu_busy_time from exceeding actual_cpu

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: <CAKfTPtAQvQv0WfRD5siqJKLBbY-kKu8jcrNvr_bBf0Bc4n-tqQ@mail.gmail.com>
Date: Fri, 21 Jun 2024 15:00:53 +0200
From: Vincent Guittot <vincent.guittot@...aro.org>
To: Dietmar Eggemann <dietmar.eggemann@....com>
Cc: Xuewen Yan <xuewen.yan94@...il.com>, Qais Yousef <qyousef@...alina.io>, 
	Xuewen Yan <xuewen.yan@...soc.com>, mingo@...hat.com, peterz@...radead.org, 
	juri.lelli@...hat.com, rostedt@...dmis.org, bsegall@...gle.com, 
	mgorman@...e.de, bristot@...hat.com, vschneid@...hat.com, 
	vincent.donnefort@....com, ke.wang@...soc.com, linux-kernel@...r.kernel.org, 
	christian.loehle@....com
Subject: Re: [PATCH] sched/fair: Prevent cpu_busy_time from exceeding actual_cpu_capacity

On Fri, 21 Jun 2024 at 12:41, Dietmar Eggemann <dietmar.eggemann@....com> wrote:
>
> On 19/06/2024 04:46, Xuewen Yan wrote:
> > On Tue, Jun 18, 2024 at 10:58 PM Qais Yousef <qyousef@...alina.io> wrote:
> >>
> >> On 06/17/24 12:03, Qais Yousef wrote:
>
> [...]
>
> > Sorry for the late reply...
> > In our own tree, we removed the check for rd overutil in feec(), so
> > the above case often occurs.
>
> How to you schedule hackbench on this thing then? Via EAS or do you just
> exclude this kind of workload?

Don't know the details for Xuewen but that's also what I'm doing as
part of the rework I'm doing on EAS. As I said in at OSPM, I'm also
calling feec() every time even when overutilized. feec() return -1
when it can't find a suitable cpu and we then fallback to default
performance path

>
> > And now it seems that on the mainline, uclamp_max is the only way to
> > override this.
>
> [...]
>