Re: [PATCH] perf/rapl: restart perf rapl counter after resume

[Liang, Kan]

On 6/20/2019 8:50 AM, Peter Zijlstra wrote:
> On Mon, Jun 17, 2019 at 09:41:37PM +0800, Zhang Rui wrote:
>
> > After S3 suspend/resume, "perf stat -I 1000 -e power/energy-pkg/ -a"
> > reports an insane value for the very first sampling period after resume
> > as shown below.
> >
> >  ÂÂÂÂ19.278989977ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ2.16 Joules power/energy-pkg/
> >  ÂÂÂÂ20.279373569ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ1.96 Joules power/energy-pkg/
> >  ÂÂÂÂ21.279765481ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ2.09 Joules power/energy-pkg/
> >  ÂÂÂÂ22.280305420ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ2.10 Joules power/energy-pkg/
> >  ÂÂÂÂ25.504782277ÂÂÂ4,294,966,686.01 Joules power/energy-pkg/
> >  ÂÂÂÂ26.505114993ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ3.58 Joules power/energy-pkg/
> >  ÂÂÂÂ27.505471758ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ1.66 Joules power/energy-pkg/
> >
> > Fix this by resetting the counter right after resume.
>
> Cute...
>
>
> > +#ifdef CONFIG_PM
> > +
> > +static int perf_rapl_suspend(void)
> > +{
> > +	int i;
> > +
> > +	get_online_cpus();
> > +	for (i = 0; i < rapl_pmus->maxpkg; i++)
> > +		rapl_pmu_update_all(rapl_pmus->pmus[i]);
> > +	put_online_cpus();
> > +	return 0;
> > +}
> > +
> > +static void perf_rapl_resume(void)
> > +{
> > +	int i;
> > +
> > +	get_online_cpus();
> > +	for (i = 0; i < rapl_pmus->maxpkg; i++)
> > +		rapl_pmu_restart_all(rapl_pmus->pmus[i]);
> > +	put_online_cpus();
> > +}
>
> What's the reason for that get/put_online_cpus() here ?

It looks like syscore_* functions are executed with one CPU on-line.
If so, they may not be the right place for the rapl callback.

Rapl is per socket. The driver manipulates the registers on the first 
CPU of each socket. I think we need to update/restart the counters on 
all sockets. That's the reason I add get/put_online_cpus() in the 
original patch.

Besides, I think we also need to call rapl_pmu_restart/update_all() on 
the target cpu.


Thanks,
Kan