linux-kernel - Re: [PATCH 0/4] CPUFreq statistics retrieved by drivers

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]

Date:   Tue, 4 Aug 2020 10:27:13 -0700
From:   Florian Fainelli <f.fainelli@...il.com>
To:     Lukasz Luba <lukasz.luba@....com>, linux-kernel@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org, linux-pm@...r.kernel.org
Cc:     sudeep.holla@....com, cristian.marussi@....com,
        viresh.kumar@...aro.org, rjw@...ysocki.net
Subject: Re: [PATCH 0/4] CPUFreq statistics retrieved by drivers



On 7/29/2020 8:12 AM, Lukasz Luba wrote:
> Hi all,
> 
> The existing CPUFreq framework does not tracks the statistics when the
> 'fast switch' is used or when firmware changes the frequency independently
> due to e.g. thermal reasons. However, the firmware might track the frequency
> changes and expose this to the kernel.

Or the firmware might have changed the CPU frequency in response to a
request from the secure world for instance.

> 
> This patch set aims to introduce CPUfreq statistics gathered by firmware
> and retrieved by CPUFreq driver. It would require a new API functions
> in the CPUFreq, which allows to poke drivers to get these stats.

>From a debugging perspective, it would be helpful if the firmware
maintained statistics were exposed as a super-set of the Linux cpufreq
statistics and aggregated into them such that you could view the normal
world vs. secure world residency of a given frequency point. This would
help because a lot of times, Linux requests freq X, but the secure world
requires freq Y (with X >= Y) and people do not really understand why
the resulting power usage is higher for instance.

What are your thoughts on this?
-- 
Florian