Re: [PATCH 2/2] thermal: cpufreq_cooling: Reuse effective_cpu_util()

[Rafael J. Wysocki]

On Thu, Oct 22, 2020 at 1:58 PM Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>
> On Thu, Oct 22, 2020 at 01:30:01PM +0200, Rafael J. Wysocki wrote:
>
> > Many people use intel_pstate in the active mode with HWP enabled too.
>
> We now have HWP-passive supported, afaict. So we should discourage that.

Which is kind of hard, because plain HWP does better in some cases,
especially performance-focused.

I am still not sure why this is the case given how the passive mode
with HWP enabled is implemented, but that's what Srinivas sees in his
tests.

> That is; I'll care less and less about people not using schedutil as
> time goes on.
>
> > Arguably, that doesn't need to compute the effective utilization, so I
> > guess it is not relevant for the discussion here, but it is not
> > negligible in general.
>
> Why not? cpufreq-cooling should still be able to throttle the system by
> setting HWP.Highest_Performance no?

Well, in theory, but it is not used on x86 AFAICS.

> In which case it still needs an energy estimate.