linux-kernel - Re: [RFC PATCH 8/9] sched/rt: Remove support for cgroups-v1

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: <aEfvmM-VpKru33Ql@jlelli-thinkpadt14gen4.remote.csb>
Date: Tue, 10 Jun 2025 10:40:56 +0200
From: Juri Lelli <juri.lelli@...hat.com>
To: Yuri Andriaccio <yurand2000@...il.com>
Cc: Ingo Molnar <mingo@...hat.com>, Peter Zijlstra <peterz@...radead.org>,
	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>,
	Valentin Schneider <vschneid@...hat.com>,
	linux-kernel@...r.kernel.org,
	Luca Abeni <luca.abeni@...tannapisa.it>,
	Yuri Andriaccio <yuri.andriaccio@...tannapisa.it>
Subject: Re: [RFC PATCH 8/9] sched/rt: Remove support for cgroups-v1

Hello,

On 05/06/25 09:14, Yuri Andriaccio wrote:
> Disable control files for cgroups-v1, and allow only cgroups-v2. This should
> simplify maintaining the code, also because cgroups-v1 is deprecated.

I wonder if the change above and what below need to be split in separate
patches. They look orthogonal and we would have a single separate patch
disabling v1 support which is probably desirable.

> Set the default rt-cgroups runtime to zero, otherwise a cgroup-v1 kernel will
> not be able to start SCHED_DEADLINE tasks.
> 
> Allow zeroing the runtime of the root control group. This runtime only affects
> the available bandwidth of the rt-cgroup hierarchy but not the SCHED_FIFO /
> SCHED_RR tasks on the global runqueue.
> 
> Notes:
> Disabling the root control group bandwidth should not cause any side effect, as
> SCHED_FIFO / SCHED_RR tasks do not depend on it since the introduction of
> fair_servers.

Also, what summarized above needs to be well documented in the kernel
docs, as I am pretty sure it will surprise people (actually surprised
even myself while playing with the set :).

Thanks,
Juri