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:   Mon, 6 Aug 2018 10:30:39 +0200
From:   "Rafael J. Wysocki" <rafael@...nel.org>
To:     Gabriele Mazzotta <gabriele.mzt@...il.com>
Cc:     Srinivas Pandruvada <srinivas.pandruvada@...ux.intel.com>,
        "Rafael J. Wysocki" <rjw@...ysocki.net>,
        Len Brown <lenb@...nel.org>,
        Viresh Kumar <viresh.kumar@...aro.org>,
        Linux PM <linux-pm@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] Revert "cpufreq: intel_pstate: Fix ->set_policy()
 interface for no_turbo"

On Sat, Aug 4, 2018 at 7:31 PM, Gabriele Mazzotta
<gabriele.mzt@...il.com> wrote:
> On 04/08/2018 17:29, Gabriele Mazzotta wrote:
>> This change does not take into account that some BIOSes change
>> MSR_IA32_MISC_ENABLE_TURBO_DISABLE depending on the power source.
>> If the turbo is disabled when the system boots, policy.max_freq
>> is set to pstate.max_pstate. However, if the BIOS later enables
>> the turbo, the CPU will never be able to run at pstate.turbo_pstate.
>>
>> Since now intel_pstate_set_policy() does its calculations using
>> pstate.max_freq and pstate.turbo_freq, we can always calculate
>> cpuinfo.max_freq using pstate.turbo_pstate, thus allowing system
>> with varying MSR_IA32_MISC_ENABLE_TURBO_DISABLE to run at full
>> speed when the turbo is enabled.

Well, the problem with this approach is that always using
pstate.turbo_pstate as the max causes the governor to overestimate the
target frequency when the turbo range is not available (the target
depends on the width of the entire available P-state range including
turbo, so if the turbo range is not available, the number take into
that computation is too large).

Are we expected to get notified when the BIOS updates
MSR_IA32_MISC_ENABLE_TURBO_DISABLE?

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ