Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 29 Aug 2022 16:34:39 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Quentin Perret <qperret@google.com>
Cc:     Vincent Guittot <vincent.guittot@linaro.org>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Ingo Molnar <mingo@kernel.org>,
        Morten Rasmussen <morten.rasmussen@arm.com>,
        Vincent Donnefort <vdonnefort@google.com>,
        Patrick Bellasi <patrick.bellasi@matbug.net>,
        Abhijeet Dharmapurikar <adharmap@quicinc.com>,
        Jian-Min <Jian-Min.Liu@mediatek.com>,
        Qais Yousef <qais.yousef@arm.com>, linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCH 1/1] sched/pelt: Introduce PELT multiplier
Message-ID: <YwzOf5WV5G8dQBGb@hirez.programming.kicks-ass.net>
References: <20220829055450.1703092-1-dietmar.eggemann@arm.com>
 <20220829055450.1703092-2-dietmar.eggemann@arm.com>
 <Ywxz7cuR+/+1CFMR@hirez.programming.kicks-ass.net>
 <YwyOzgbbUbB+JmSH@hirez.programming.kicks-ass.net>
 <CAKfTPtAnC4escfufhjJkqsYyhDPzpmQH4xYT4nj=EjCP7gzifg@mail.gmail.com>
 <YwzL1eJUIReAEv0l@google.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <YwzL1eJUIReAEv0l@google.com>
Precedence: bulk

On Mon, Aug 29, 2022 at 02:23:17PM +0000, Quentin Perret wrote:

> I'll let folks in CC comment about their use-case in more details, but
> there's definitely been an interest in tuning this thing at run-time

An interest and it making sense are two very distinct things that bear
no relation to one another in any way.

> FWIW. Typically a larger half-life will be fine with predictable
> workloads with little inputs from users (e.g. fullscreen video playback)
> while a lower one can be preferred in highly interactive cases (games,

As per always; consider the combined workload.

> ...). The transient state is fun to reason about, but it really
> shouldn't be too common AFAIK.

Once you give away control there is no taking it back, and userspace
*will* do stupid things and expect unicorns.

> With that said I'd quite like to see numbers to back that claim.
> Measuring power while running a video (for instance) with various HL
> values should help. And similarly it shouldn't be too hard to get
> performance numbers.

I'm thinking this all needs far more than mere numbers to justify.