Re: [PATCH v2 1/2] cpufreq/schedutil: Remove fast_switch_possible flag if driver doesn't set fast_switch

[Rafael J. Wysocki]

On Wed, May 10, 2023 at 7:43 AM Wyes Karny <wyes.karny@xxxxxxx> wrote:
>
> Hi Rafael,
>
> Thanks for reviewing the patch.
>
> On 09 May 20:39, Rafael J. Wysocki wrote:
> ------------------------------------------>8--------------------------------------
> > > > diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c
> > > > index 2548ec92faa2..007893514c87 100644
> > > > --- a/drivers/cpufreq/intel_pstate.c
> > > > +++ b/drivers/cpufreq/intel_pstate.c
> > > > @@ -2698,8 +2698,6 @@ static int __intel_pstate_cpu_init(struct cpufreq_policy *policy)
> > > >
> > > >         intel_pstate_init_acpi_perf_limits(policy);
> > > >
> > > > -       policy->fast_switch_possible = true;
> > > > -
> > > >         return 0;
> > > >  }
> > > >
> > > > @@ -2955,6 +2953,7 @@ static int intel_cpufreq_cpu_init(struct cpufreq_policy *policy)
> > > >         if (ret)
> > > >                 return ret;
> > > >
> > > > +       policy->fast_switch_possible = true;
> >
> > I'm not sure what this is about.  Is it a cleanup of intel_pstate?
>
> This patch intends to remove fast_switch_possible flag dependency from
> drivers which only use adjust_perf as frequency/pref update callback. As
> intel_pstate and amd_pstate driver has only adjust_perf and not
> fast_switch, therefore I'm removing that flag from these drivers. But
> intel_cpufreq has fast_switch therefore, only adding that flag for
> intel_cpufreq driver.

But is it really better to change it?  It works correctly as-is AFAICS.

In any case, the intel_pstate change should be a separate patch,
because it is not directly related to the other changes in the
$subject one IMV.