linux-kernel - RE: [PATCH v2] cpufreq: intel_pstate: Rework iowait boosting to be less aggressive

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 for Android: free password hash cracker in your pocket

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

Message-ID: <000c01d4c6f6$914a7f20$b3df7d60$@net>
Date:   Sun, 17 Feb 2019 11:25:37 -0800
From:   "Doug Smythies" <dsmythies@...us.net>
To:     "'Rafael J. Wysocki'" <rjw@...ysocki.net>,
        "'Linux PM'" <linux-pm@...r.kernel.org>
Cc:     "'Srinivas Pandruvada'" <srinivas.pandruvada@...ux.intel.com>,
        "'LKML'" <linux-kernel@...r.kernel.org>,
        "Doug Smythies" <dsmythies@...us.net>
Subject: RE: [PATCH v2] cpufreq: intel_pstate: Rework iowait boosting to be less aggressive

On 2019.02.07 03:51 Rafael J. Wysocki wrote:

> From: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
> 
> The current iowait boosting mechanism in intel_pstate_update_util()
> is quite aggressive, as it goes to the maximum P-state right away,
> and may cause excessive amounts of energy to be used, which is not
> desirable and arguably isn't necessary too.
>
> Follow commit a5a0809bc58e ("cpufreq: schedutil: Make iowait boost
> more energy efficient") that reworked the analogous iowait boost
> mechanism in the schedutil governor and make the iowait boosting
> in intel_pstate_update_util() work along the same lines.
>
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
> ---
>
> -> v2:
>   * Follow the Doug's suggestion and drop the immediate jump to
>     max P-state if boost is max.  The code is simpler this way and
>     the perf impact should not be noticeable on average.

Hi Rafael,

Something has broken on my incoming e-mail sorting stuff, and I
missed this one (and some others).

This V2 is not actually what I was proposing. I was O.K. with
the immediate jump, but I didn't want the set_pstate step
by-passed if it was already at max because that would also
by-pass the trace sample, if it was enabled.

Anyway, this V2 seems O.K. to me. I tested it compared to V1
and, as you mentioned, wasn't able to detect any energy consumption
or performance differences.

... Doug