Re: perf test fail :: "perf stat --bpf-counters --for-each-cgroup test"

[Michael Petlan <mpetlan@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 1884 bytes --]

On Fri, 19 Jul 2024, Namhyung Kim wrote:
> Hello,
> 
> On Fri, Jul 19, 2024 at 2:50 AM Michael Petlan <mpetlan@redhat.com> wrote:
> >
> > Hello Namhyung,
> >
> > we were investigating some test failures of the testcase mentioned
> > in $subj. We have narrowed it down to:
> >
> >     # perf stat -C 0,1 --for-each-cgroup system.slice,user.slice -e cycles -- taskset -c 1 perf test -w thloop
> >
> >     Performance counter stats for 'CPU(s) 0,1':
> >          <not counted>      cycles                           system.slice
> >          3,020,401,084      cycles                           user.slice
> >
> >          1.009787097 seconds time elapsed
> >
> > As seen, the system.slice is not counted properly in our case. It
> > happens even without bpf-counters being involved.
> >
> > There were rumours that it might be caused due to too small system
> > load, but it apparently happens even when the load was replaced by
> > "thloop" workload from perf-test's workload library. However, even
> > so, if the load was insufficient, we'd see a value – 0 instead of
> > "not counted". The "<not counted>" result is printed if the counter
> > wasn't properly enabled and running.
> >
> > Have you encountered this problem? What could cause it?
> 
> Sometimes, it depends on the system activity.
> I assumed there would be some system operation running
> on the given CPUs at any time.  If not, it can fail..

Thanks for response. However, in such case, I'd expect "0" value instead
of "<not counted>". As far as I understand, "<not supported>" occurs when
an event cannot be even opened, "<not counted>" appears when the event
can be opened, but for example there aren't enough free hardware counters,
so it does not run any counting actually. But for cases when there is not
enough load, we get simply low numerical values, or 0, don't we...

Michael
> 
> Thanks,
> Namhyung
> 
>