RE: [PATCH 07/12] cpufreq: amd_pstate: map desired perf into pstate scope for powersave governor

[Yuan, Perry]

[AMD Official Use Only - General]

Hi Nathan:

> -----Original Message-----
> From: Fontenot, Nathan <Nathan.Fontenot@xxxxxxx>
> Sent: Friday, July 8, 2022 3:57 AM
> To: Yuan, Perry <Perry.Yuan@xxxxxxx>; rafael.j.wysocki@xxxxxxxxx;
> viresh.kumar@xxxxxxxxxx; Huang, Ray <Ray.Huang@xxxxxxx>; Rafael J.
> Wysocki <rafael@xxxxxxxxxx>; linux-pm@xxxxxxxxxxxxxxx; linux-
> kernel@xxxxxxxxxxxxxxx
> Cc: Sharma, Deepak <Deepak.Sharma@xxxxxxx>; Limonciello, Mario
> <Mario.Limonciello@xxxxxxx>; Fontenot, Nathan
> <Nathan.Fontenot@xxxxxxx>; Deucher, Alexander
> <Alexander.Deucher@xxxxxxx>; Su, Jinzhou (Joe) <Jinzhou.Su@xxxxxxx>;
> Huang, Shimmer <Shimmer.Huang@xxxxxxx>; Du, Xiaojian
> <Xiaojian.Du@xxxxxxx>; Meng, Li (Jassmine) <Li.Meng@xxxxxxx>
> Subject: Re: [PATCH 07/12] cpufreq: amd_pstate: map desired perf into
> pstate scope for powersave governor
> 
> On 7/7/22 12:00, Perry Yuan wrote:
> > The patch will fix the invalid desired perf value for powersave
> > governor. This issue is found when testing on one AMD EPYC system, the
> > actual des_perf is smaller than the min_perf value, that is invalid
> > value. because the min_perf is the lowest_perf system can support in
> > idle state.
> >
> > Signed-off-by: Perry Yuan <Perry.Yuan@xxxxxxx>
> > ---
> >  drivers/cpufreq/amd-pstate.c | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/drivers/cpufreq/amd-pstate.c
> > b/drivers/cpufreq/amd-pstate.c index 7c51f4125263..154eed849f38
> 100644
> > --- a/drivers/cpufreq/amd-pstate.c
> > +++ b/drivers/cpufreq/amd-pstate.c
> > @@ -317,6 +317,7 @@ static int amd_pstate_target(struct cpufreq_policy
> *policy,
> >  	des_perf = DIV_ROUND_CLOSEST(target_freq * cap_perf,
> >  				     cpudata->max_freq);
> >
> > +	des_perf = clamp_t(unsigned long, des_perf, min_perf, max_perf);
> >  	cpufreq_freq_transition_begin(policy, &freqs);
> >  	amd_pstate_update(cpudata, min_perf, des_perf,
> >  			  max_perf, false);
> 
> The clamping of the desired perf value should be moved to
> amd_pstate_update(). The only other caller of amd_pstate_update() is
> amd_pstate_adjust_perf() which already clamps the desired perf value
> before making the call.
> 
> -Nathan

Thanks for your suggestion.
Add this change into V2.

Perry.