[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZnMvOht0usglJOzm@slm.duckdns.org>
Date: Wed, 19 Jun 2024 09:19:22 -1000
From: Tejun Heo <tj@...nel.org>
To: Christian Loehle <christian.loehle@....com>
Cc: rafael@...nel.org, viresh.kumar@...aro.org, linux-pm@...r.kernel.org,
void@...ifault.com, linux-kernel@...r.kernel.org,
kernel-team@...a.com, mingo@...hat.com, peterz@...radead.org,
David Vernet <dvernet@...a.com>,
"Rafael J . Wysocki" <rafael.j.wysocki@...el.com>
Subject: Re: [PATCH 2/2] sched_ext: Add cpuperf support
Hello,
On Wed, Jun 19, 2024 at 03:07:51PM +0100, Christian Loehle wrote:
> > static void sugov_get_util(struct sugov_cpu *sg_cpu, unsigned long boost)
> > {
> > - unsigned long min, max, util = cpu_util_cfs_boost(sg_cpu->cpu);
> > + unsigned long min, max;
> > + unsigned long util = cpu_util_cfs_boost(sg_cpu->cpu) +
> > + scx_cpuperf_target(sg_cpu->cpu);
>
> What does cpu_util_cfs_boost() contain if scx is active?
Hmm... when SCX is in partial mode, cpu_util_cfs_boost() reports the util
metric from cfs side as usual, but, yeah, when scx_switched_all(), this
could be stale. I'll update.
> NIT: reverse xmas
Ah, let me just make the initial assignment a separate statement. Reverse
xmas is awkward with broken lines.
Thanks.
--
tejun
Powered by blists - more mailing lists