[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <706278eb-1906-79f3-7a9f-6ab5080ecb63@arm.com>
Date: Fri, 8 Nov 2019 11:53:05 +0100
From: Dietmar Eggemann <dietmar.eggemann@....com>
To: Vincent Guittot <vincent.guittot@...aro.org>,
Thara Gopinath <thara.gopinath@...aro.org>
Cc: mingo@...hat.com, peterz@...radead.org, ionela.voinescu@....com,
rui.zhang@...el.com, edubezval@...il.com, qperret@...gle.com,
linux-kernel@...r.kernel.org, amit.kachhap@...il.com,
javi.merino@...nel.org, daniel.lezcano@...aro.org
Subject: Re: [Patch v5 6/6] sched/fair: Enable tuning of decay period
On 07/11/2019 11:49, Vincent Guittot wrote:
> Le Tuesday 05 Nov 2019 à 13:49:46 (-0500), Thara Gopinath a écrit :
[...]
>> /**
>> @@ -10444,8 +10465,8 @@ void update_thermal_pressure(int cpu, unsigned long capped_capacity)
>> static void trigger_thermal_pressure_average(struct rq *rq)
>> {
>> #ifdef CONFIG_SMP
>> - update_thermal_load_avg(rq_clock_task(rq), rq,
>> - per_cpu(thermal_pressure, cpu_of(rq)));
>> + update_thermal_load_avg(rq_clock_task(rq) >> sched_thermal_decay_shift,
>> + rq, per_cpu(thermal_pressure, cpu_of(rq)));
>
> Would be better to create
>
> +static inline u64 rq_clock_thermal(struct rq *rq)
> +{
> + lockdep_assert_held(&rq->lock);
> + assert_clock_updated(rq);
IMHO, the asserts can be skipped here since they're already done in
rq_clock_task().
> + return rq_clock_task(rq) >> sched_thermal_decay_shift;
> +}
Powered by blists - more mailing lists