Re: [PATCH v2 1/1] x86,sched: On AMD EPYC set freq_max = max_boost in schedutil invariant formula

[Giovanni Gherdovich]

On Fri, 2021-01-29 at 16:23 +0100, Giovanni Gherdovich wrote:
> On Tue, 2021-01-26 at 11:05 +0100, Peter Zijlstra wrote:
> > On Tue, Jan 26, 2021 at 09:31:40AM +0000, Mel Gorman wrote:
> > 
> > > So, should this patch be merged for 5.11 as a stopgap, fix up
> > > schedutil/cpufreq and then test both AMD and Intel chips reporting the
> > > correct max non-turbo and max-turbo frequencies? That would give time to
> > > give some testing in linux-next before merging to reduce the chance
> > > something else falls out.
> > 
> > Yeah, we should probably do this now. Rafael, you want this or should I
> > take it?
> 
> Hello Rafael,
> 
> did you have a chance to check this patch?
> 
> It fixes a performance regression from 5.11-rc1, I hoped it could be included
> before v5.11 is released.

Hello Rafael,

you haven't replied to this patch, which was written aiming at v5.11.

Do you see any problem with it?
Frequency-invariant schedutil needs the driver to advertise a high max_freq to
work properly; the patch addresses this for AMD EPYC.


Thanks,
Giovanni