[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180925081448.GA4287@localhost.localdomain>
Date: Tue, 25 Sep 2018 10:14:48 +0200
From: Juri Lelli <juri.lelli@...hat.com>
To: peterz@...radead.org, mingo@...hat.com, rostedt@...dmis.org
Cc: linux-kernel@...r.kernel.org, luca.abeni@...tannapisa.it,
claudio@...dence.eu.com, tommaso.cucinotta@...tannapisa.it,
bristot@...hat.com, mathieu.poirier@...aro.org, lizefan@...wei.com,
cgroups@...r.kernel.org
Subject: Re: [PATCH v5 0/5] sched/deadline: fix cpusets bandwidth accounting
Hi,
On 03/09/18 16:27, Juri Lelli wrote:
> Hi,
>
> v5 of a series of patches, originally authored by Mathieu, with the intent
> of fixing a long standing issue of SCHED_DEADLINE bandwidth accounting.
> As originally reported by Steve [1], when hotplug and/or (certain)
> cpuset reconfiguration operations take place, DEADLINE bandwidth
> accounting information is lost since root domains are destroyed and
> recreated.
>
> Mathieu's approach is based on restoring bandwidth accounting info on
> the newly created root domains by iterating through the (DEADLINE) tasks
> belonging to the configured cpuset(s).
>
> Main problem of v4 was caused by the trylocking of cpuset_mutex. As
> noticed by Steve [2], if multiple tasks are created at they same time
> only the first gets to grab the mutex, the other get -EBUSY and need to
> retry. Not really nice. So, in v5 I'm proposing to use callback_lock
> instead of cpuset_mutex, which AFAIU should be enough to grant read-only
> safe access to cpusets.
>
> 01/05 has been dropped because it wasn't really adding much and was
> only causing false positives.
>
> 05/05 is still too much DEADLINE specific I guess, but let's first agree
> on foundations patches.
>
> Set also available at
>
> https://github.com/jlelli/linux.git fixes/deadline/root-domain-accounting-v5
Gentle ping about this.
Best,
- Juri
Powered by blists - more mailing lists