[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <57030672.8080405@mellanox.com>
Date: Mon, 4 Apr 2016 20:27:30 -0400
From: Chris Metcalf <cmetcalf@...lanox.com>
To: Rik van Riel <riel@...hat.com>,
Frederic Weisbecker <fweisbec@...il.com>,
Christoph Lameter <cl@...ux.com>,
Ingo Molnar <mingo@...nel.org>,
Luiz Capitulino <lcapitulino@...hat.com>,
Peter Zijlstra <peterz@...radead.org>,
Thomas Gleixner <tglx@...utronix.de>,
Viresh Kumar <viresh.kumar@...aro.org>,
<linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] nohz_full: Make sched_should_stop_tick() more
conservative
On 4/4/2016 3:36 PM, Rik van Riel wrote:
>>> On Fri, 2016-04-01 at 15:42 -0400, Chris Metcalf wrote:
>>>> On arm64, when calling enqueue_task_fair() from
>>>> migration_cpu_stop(),
>>>> we find the nr_running value updated by add_nr_running(), but the
>>>> cfs.nr_running value has not always yet been
>>>> updated. Accordingly,
>>>> the sched_can_stop_tick() false returns true when we are
>>>> migrating a
>>>> second task onto a core.
>>>> Correct this by using rq->nr_running instead of rq-
>>>>> cfs.nr_running.
>>>> This should always be more conservative, and reverts the test to
>>>> the
>>>> form it had before commit 76d92ac305f2 ("sched: Migrate sched to
>>>> use
>>>> new tick dependency mask model").
>>>>
>>>>
> [...]
>
> Patch gets my:
>
> Acked-by: Rik van Riel <riel@...hat.com>
Thanks! Whose tree should this go through: Frederic, PeterZ, Ingo?
Do any of you have any concerns with it?
--
Chris Metcalf, Mellanox Technologies
http://www.mellanox.com
Powered by blists - more mailing lists