On Tue, 2021-02-02 at 20:11 +0100, Rafael J. Wysocki wrote:

On Tue, Feb 2, 2021 at 7:45 PM Rafael J. Wysocki [off-list ref] wrote:

quoted

On Tue, Jan 26, 2021 at 5:19 PM Giovanni Gherdovich [off-list ref] wrote:

quoted

cpufreq core asks the driver what's the f_max. What's the answer?

intel_pstate says: 1C

Yes, unless turbo is disabled, in which case it is P0.

BTW, and that actually is quite important, the max_freq reported by
intel_pstate doesn't matter for schedutil after the new ->adjust_perf
callback has been added, because that doesn't even use the frequency.

So, as a long-term remedy, it may just be better to implement
->adjust_perf in acpi_cpufreq().

Thanks for pointing this out.

I agree that in the long term adding ->adjust_perf to acpi_cpufreq is
the best solution.

Yet for this submission, considering it's late in the 5.11 cycle,
the patch I propose is a reasonable candidate: the footprint is small and
it's gone through a fair amount of testing.


Thanks,
Giovanni