[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAM9d7ciMfPqY+eJsFG36V82wOiNR-2g1Siytc=vRjaZsCdfDWQ@mail.gmail.com>
Date: Mon, 3 May 2021 14:27:36 -0700
From: Namhyung Kim <namhyung@...nel.org>
To: Arnaldo Carvalho de Melo <acme@...nel.org>
Cc: Ingo Molnar <mingo@...nel.org>,
Peter Zijlstra <peterz@...radead.org>,
Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
LKML <linux-kernel@...r.kernel.org>,
Andi Kleen <ak@...ux.intel.com>,
Ian Rogers <irogers@...gle.com>,
Jin Yao <yao.jin@...ux.intel.com>, Jiri Olsa <jolsa@...hat.com>
Subject: Re: [PATCH] perf stat: Use aggregated counts directly
Hi Arnaldo,
I'm not sure it's in your tree, can you please take a look and merge it?
Thanks,
Namhyung
On Sun, Apr 25, 2021 at 11:00 AM Namhyung Kim <namhyung@...nel.org> wrote:
>
> On Mon, Apr 26, 2021 at 12:50 AM Jiri Olsa <jolsa@...hat.com> wrote:
> >
> > On Thu, Apr 22, 2021 at 07:38:33PM -0700, Namhyung Kim wrote:
> > > The ps->res_stats is for repeated runs, so the interval code should
> > > not touch it. Actually the aggregated counts are available in the
> > > counter->counts->aggr, so we can (and should) use it directly IMHO.
> > >
> > > No functional change intended.
> >
> > it looks ok, but it should fix the noise output then, right?
>
> Yeah, but only if -r and -I options are used at the same time.
>
> >
> > Acked-by: Jiri Olsa <jolsa@...hat.com>
>
> Thanks,
> Namhyung
>
> >
> > >
> > > Signed-off-by: Namhyung Kim <namhyung@...nel.org>
> > > ---
> > > tools/perf/util/stat-display.c | 8 ++++----
> > > tools/perf/util/stat.c | 12 ------------
> > > 2 files changed, 4 insertions(+), 16 deletions(-)
> > >
> > > diff --git a/tools/perf/util/stat-display.c b/tools/perf/util/stat-display.c
> > > index d3137bc17065..a38fa6527586 100644
> > > --- a/tools/perf/util/stat-display.c
> > > +++ b/tools/perf/util/stat-display.c
> > > @@ -807,11 +807,11 @@ static void counter_aggr_cb(struct perf_stat_config *config __maybe_unused,
> > > bool first __maybe_unused)
> > > {
> > > struct caggr_data *cd = data;
> > > - struct perf_stat_evsel *ps = counter->stats;
> > > + struct perf_counts_values *aggr = &counter->counts->aggr;
> > >
> > > - cd->avg += avg_stats(&ps->res_stats[0]);
> > > - cd->avg_enabled += avg_stats(&ps->res_stats[1]);
> > > - cd->avg_running += avg_stats(&ps->res_stats[2]);
> > > + cd->avg += aggr->val;
> > > + cd->avg_enabled += aggr->ena;
> > > + cd->avg_running += aggr->run;
> > > }
> > >
> > > /*
> > > diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c
> > > index 2db46b9bebd0..d3ec2624e036 100644
> > > --- a/tools/perf/util/stat.c
> > > +++ b/tools/perf/util/stat.c
> > > @@ -437,18 +437,6 @@ int perf_stat_process_counter(struct perf_stat_config *config,
> > >
> > > aggr->val = aggr->ena = aggr->run = 0;
> > >
> > > - /*
> > > - * We calculate counter's data every interval,
> > > - * and the display code shows ps->res_stats
> > > - * avg value. We need to zero the stats for
> > > - * interval mode, otherwise overall avg running
> > > - * averages will be shown for each interval.
> > > - */
> > > - if (config->interval || config->summary) {
> > > - for (i = 0; i < 3; i++)
> > > - init_stats(&ps->res_stats[i]);
> > > - }
> > > -
> > > if (counter->per_pkg)
> > > evsel__zero_per_pkg(counter);
> > >
> > > --
> > > 2.31.1.498.g6c1eba8ee3d-goog
> > >
> >
Powered by blists - more mailing lists