[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <7fdfff24-80ed-acbf-810f-b641570141fd@arm.com>
Date: Mon, 21 Aug 2023 18:41:00 +0200
From: Dietmar Eggemann <dietmar.eggemann@....com>
To: Qais Yousef <qyousef@...alina.io>,
"Rafael J. Wysocki" <rafael@...nel.org>,
Viresh Kumar <viresh.kumar@...aro.org>,
Ingo Molnar <mingo@...nel.org>,
Peter Zijlstra <peterz@...radead.org>
Cc: linux-kernel@...r.kernel.org, linux-pm@...r.kernel.org,
Vincent Guittot <vincent.guittot@...aro.org>,
Lukasz Luba <lukasz.luba@....com>
Subject: Re: [PATCH RFC 4/4] sched: cpufreq: Apply DVFS headroom to CFS only
On 20/08/2023 23:06, Qais Yousef wrote:
> RT and Deadline have exact performance requirement when running. RT runs
> at max or a specific OPP defined by uclamp_min. Deadline's OPP is
> defined by its bandwidth. Both of which are known ahead of time and
> don't require a headroom to grow into.
>
> IRQs on the other hand have no specific performance requirement and
> cruises along at whatever the current OPP happens to be when they occur.
>
> Now they all have PELT pressure signals that does impact frequency
> selection and task placement. The question is do they need DVFS
> headroom?
>
> I think the answer is no because when CFS is not running at all, these
> pressure signal has no real impact on performance for RT, DL or IRQ.
>
> If CFS util is not zero, we already add their pressure as an
> *additional* headroom to account for the lost/stolen time. So I argue
> that the pressure are headroom themselves and shouldn't need an
> additional DVFS headroom applied on top.
>
> In summary final outcome should be:
>
> CFS + DVFS headroom + (RT, DT, IRQ) pressure headroom
I assume here you want to align the difference that EAS deals with
`util_cfs` vs `capacity` whereas power deals with `util` vs
`capacity_orig`? You want that power should only apply the 1.25 to util_cfs?
> Signed-off-by: Qais Yousef (Google) <qyousef@...alina.io>
> ---
> kernel/sched/core.c | 12 +++++-------
> 1 file changed, 5 insertions(+), 7 deletions(-)
>
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index 441d433c83cd..602e369753a3 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -7438,10 +7438,11 @@ unsigned long effective_cpu_util(int cpu, unsigned long util_cfs,
> * When there are no CFS RUNNABLE tasks, clamps are released and
> * frequency will be gracefully reduced with the utilization decay.
> */
> - util = util_cfs + cpu_util_rt(rq);
> if (type == FREQUENCY_UTIL) {
> - util = apply_dvfs_headroom(util);
> + util = apply_dvfs_headroom(util_cfs) + cpu_util_rt(rq);
> util = uclamp_rq_util_with(rq, util, p);
> + } else {
> + util = util_cfs + cpu_util_rt(rq);
> }
>
> dl_util = cpu_util_dl(rq);
> @@ -7473,12 +7474,9 @@ unsigned long effective_cpu_util(int cpu, unsigned long util_cfs,
> * max - irq
> * U' = irq + --------- * U
> * max
> - *
> - * We only need to apply dvfs headroom to irq part since the util part
> - * already had it applied.
> */
> util = scale_irq_capacity(util, irq, max);
> - util += type == FREQUENCY_UTIL ? apply_dvfs_headroom(irq) : irq;
> + util += irq;
>
> /*
> * Bandwidth required by DEADLINE must always be granted while, for
> @@ -7491,7 +7489,7 @@ unsigned long effective_cpu_util(int cpu, unsigned long util_cfs,
> * an interface. So, we only do the latter for now.
> */
> if (type == FREQUENCY_UTIL)
> - util += apply_dvfs_headroom(cpu_bw_dl(rq));
> + util += cpu_bw_dl(rq);
>
> return min(max, util);
> }
Powered by blists - more mailing lists