linux-kernel - Re: [PATCH v2 0/3] Clarify abstract scale usage for power values in Energy Model, EAS and IPA

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20201016121844.GA2420691@google.com>
Date:   Fri, 16 Oct 2020 13:18:44 +0100
From:   Quentin Perret <qperret@...gle.com>
To:     Daniel Lezcano <daniel.lezcano@...aro.org>
Cc:     "Rafael J. Wysocki" <rafael@...nel.org>,
        Lukasz Luba <lukasz.luba@....com>,
        "Rafael J. Wysocki" <rjw@...ysocki.net>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Linux PM <linux-pm@...r.kernel.org>,
        "open list:DOCUMENTATION" <linux-doc@...r.kernel.org>,
        "devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
        Rob Herring <robh+dt@...nel.org>,
        Amit Kucheria <amitk@...nel.org>,
        Jonathan Corbet <corbet@....net>,
        Dietmar Eggemann <Dietmar.Eggemann@....com>,
        Doug Anderson <dianders@...omium.org>,
        Matthias Kaehlcke <mka@...omium.org>,
        "Nayak, Rajendra" <rnayak@...eaurora.org>
Subject: Re: [PATCH v2 0/3] Clarify abstract scale usage for power values in
 Energy Model, EAS and IPA

On Friday 16 Oct 2020 at 13:48:33 (+0200), Daniel Lezcano wrote:
> If the SCMI is returning abstract numbers, the thermal IPA governor will
> use these numbers as a reference to mitigate the temperature at the
> specified sustainable power which is expressed in mW in the DT. So it
> does not work and we can not detect such conflict.
> 
> That is why I'm advocating to keep mW for the energy model and make the
> SCMI and DT power numbers incompatible.

I think it's fair to say SCMI-provided number should only be compared to
other SCMI-provided numbers, so +1 on that. But what I don't understand
is why specifying the EM in mW helps with that? Can we not let the
providers specify the unit? And then it's up to the clients to decide
what they want to do. The scheduler wouldn't care, and IPA would have to
check things are comparable, but all in all that should work out fine
without a strong requirement on the actual unit.

Also, I thought SCMI had a notion of sustained performance too, why
can't we use that for IPA? Lukasz?

Thanks.
Quentin