linux-kernel - Re: [RFC][PATCH 08/10] sched/fair: Implement delayed dequeue

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: <747627a1414f1f33d0c237f555494149d6937800.camel@gmx.de>
Date: Tue, 16 Apr 2024 07:36:28 +0200
From: Mike Galbraith <efault@....de>
To: K Prateek Nayak <kprateek.nayak@....com>, Peter Zijlstra
 <peterz@...radead.org>, mingo@...hat.com, juri.lelli@...hat.com, 
 vincent.guittot@...aro.org, dietmar.eggemann@....com, rostedt@...dmis.org, 
 bsegall@...gle.com, mgorman@...e.de, bristot@...hat.com,
 vschneid@...hat.com,  linux-kernel@...r.kernel.org
Cc: wuyun.abel@...edance.com, tglx@...utronix.de, Chen Yu
 <yu.c.chen@...el.com>,  Oliver Sang <oliver.sang@...el.com>
Subject: Re: [RFC][PATCH 08/10] sched/fair: Implement delayed dequeue

On Tue, 2024-04-16 at 08:48 +0530, K Prateek Nayak wrote:
>
> Wondering if you are running into the issue where pick_eevdf() returns
> NULL despite there being runnable CFS task on the runqueue. Can you try
> running with the following patch from Chenyu -
> https://lore.kernel.org/lkml/20240226082349.302363-1-yu.c.chen@intel.com/

I hadn't tried it in master branch, but had in my 6.1-eevdf tree which
contains the previous delay dequeue version, because I found that it
too would explode, which surprised me having run it quite a bit woe
free. Patch helped 6.1-eevdf not at all.  I just tried it in master to
be sure, and while the death throes were not identical, this one began
with SCHED_WARN_ON() in put_prev_entity(), box still went belly up.

BTW/FWIW, turning on AUTOGROUP seemingly rendered box immune (unless
booted with noautogroup 'course), and that immunity even held when I
launched all players in the same cgroup (hmm).

	-Mike