linux-kernel - Re: [Question] About unnecessary policy_has_boost_freq() calls in freq

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: <CAJZ5v0hshbbbTncpkZoS98jq+ChiARSZCNn5P8kaEduADbmHSw@mail.gmail.com>
Date: Tue, 23 Sep 2025 14:46:43 +0200
From: "Rafael J. Wysocki" <rafael@...nel.org>
To: Zihuan Zhang <zhangzihuan@...inos.cn>
Cc: "Rafael J. wysocki" <rafael@...nel.org>, Viresh Kumar <viresh.kumar@...aro.org>, 
	"zhenglifeng (A)" <zhenglifeng1@...wei.com>, Linux PM <linux-pm@...r.kernel.org>, 
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [Question] About unnecessary policy_has_boost_freq() calls in freq_table.c

Hi,

On Tue, Sep 23, 2025 at 10:53 AM Zihuan Zhang <zhangzihuan@...inos.cn> wrote:
>
> Hi all,
>
> While reviewing the cpufreq code, I noticed that in
> cpufreq_table_validate_and_sort() the function policy_has_boost_freq()
> is still being called, even though most cpufreq drivers do not make use
> of the CPUFREQ_BOOST_FREQ flag.
>
> code in cpufreq_table_validate_and_sort():
>
>          /* Driver's may have set this field already */
>          if (policy_has_boost_freq(policy))
>                  policy->boost_supported = true;
>
>
> For drivers like acpi-cpufreq that don’t rely on CPUFREQ_BOOST_FREQ
> flag, this extra check looks confusing and possibly unnecessary.
>
> Would it make sense to move this check into cpufreq_boost_set_sw()
> instead, so that the call is only made when boost is actually relevant?

First off, setting policy->boost_supported doesn't really belong to
cpufreq_table_validate_and_sort(), so the idea of splitting it off
that function sounds reasonable to me.

However, cpufreq_boost_set_sw() is used as a .set_boost() callback, so
it gets called every time the "boost enabled" setting is changed.  It
doesn't look like a good place for updating policy->boost_supported to
me.