linux-kernel - Re: [PATCH] selftests: cpufreq: Check cpuinfo_cur

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] [day] [month] [year] [list]

Message-ID: <3568375.ceg5RryLQe@aspire.rjw.lan>
Date:   Wed, 19 Jul 2017 14:47:30 +0200
From:   "Rafael J. Wysocki" <rjw@...ysocki.net>
To:     Viresh Kumar <viresh.kumar@...aro.org>
Cc:     Leonard Crestez <leonard.crestez@....com>,
        Shuah Khan <shuah@...nel.org>, linux-kselftest@...r.kernel.org,
        Octavian Purdila <octavian.purdila@....com>,
        linux-pm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] selftests: cpufreq: Check cpuinfo_cur_freq set as expected

On Wednesday, July 19, 2017 12:24:06 PM Viresh Kumar wrote:
> On 18-07-17, 22:34, Leonard Crestez wrote:
> > The semantics of scaling_cur_freq and cpuinfo_cur_freq are not very
> > clear to me.
> 
> cpuinfo_cur_freq reads the frequency right from hardware all the time
> and so can be slow. It can only be read by root if I remember
> correctly.
> 
> Whereas scaling_cur_freq tries to read the cached frequency. But it
> has changed a bit with the below mentioned patch.
> 
> > In my particular case I need to check cpuinfo_cur_freq because this is
> > what ends up returning the rate of the arm clk. Otherwise
> > scaling_cur_freq just returns policy->cur
> 
> Yeah, we may actually need to use cpuinfo_cur_freq as that is what
> ends up giving the real freq.
> 
> > unless the driver has a
> > setpolicy function (I don't understand that condition).
> 
> That's because the core doesn't know the cached freq for setpolicy
> drivers and so we need to call the ->get() callback. But for non
> setpolicy drivers, core already has the cached value.

Please remember that cpuinfo_cur_freq may not be present.

Thanks,
Rafael