[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20181023100057.GJ3109@worktop.c.hoisthospitality.com>
Date: Tue, 23 Oct 2018 12:00:57 +0200
From: Peter Zijlstra <peterz@...radead.org>
To: Vincent Guittot <vincent.guittot@...aro.org>
Cc: mingo@...nel.org, linux-kernel@...r.kernel.org, rjw@...ysocki.net,
dietmar.eggemann@....com, Morten.Rasmussen@....com,
patrick.bellasi@....com, pjt@...gle.com, bsegall@...gle.com,
thara.gopinath@...aro.org
Subject: Re: [PATCH v4 2/2] sched/fair: update scale invariance of PELT
On Fri, Oct 19, 2018 at 06:17:51PM +0200, Vincent Guittot wrote:
> In order to achieve this time scaling, a new clock_pelt is created per rq.
> diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
> index 3990818..d987f50 100644
> --- a/kernel/sched/sched.h
> +++ b/kernel/sched/sched.h
> @@ -848,6 +848,8 @@ struct rq {
> unsigned int clock_update_flags;
> u64 clock;
> u64 clock_task;
> + u64 clock_pelt;
> + unsigned long lost_idle_time;
Very clever that. Seems to work out nicely. We should maybe look at
ensuring all these clock fields are indeed on the same cacheline.
Powered by blists - more mailing lists