Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:3 as permitted sender) client-ip=2620:137:e000::3:3;
MIME-Version: 1.0
References: <20230916040915.1075620-1-irogers@google.com> <CAL715WJxmMbXkZSE3p_+ODGxabgrXREsBo9aFu9G9=qkYZeH9A@mail.gmail.com>
 <CAP-5=fUjNiDv=KQ7t8jqfOfOt5i8HGvt8Vv1hn2-hLxX_Kqucg@mail.gmail.com>
In-Reply-To: <CAP-5=fUjNiDv=KQ7t8jqfOfOt5i8HGvt8Vv1hn2-hLxX_Kqucg@mail.gmail.com>
From:   Mingwei Zhang <mizhang@google.com>
Date:   Wed, 20 Sep 2023 22:04:33 -0700
Message-ID: <CAL715W+GQuCJm-1SEsNN2qnHghNL1SrzwH9Km5K8UxubEFfYVw@mail.gmail.com>
Subject: Re: [PATCH v1] perf evlist: Avoid frequency mode for the dummy event
To:     Ian Rogers <irogers@google.com>
Cc:     Peter Zijlstra <peterz@infradead.org>,
        Ingo Molnar <mingo@redhat.com>,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        Mark Rutland <mark.rutland@arm.com>,
        Alexander Shishkin <alexander.shishkin@linux.intel.com>,
        Jiri Olsa <jolsa@kernel.org>,
        Namhyung Kim <namhyung@kernel.org>,
        Adrian Hunter <adrian.hunter@intel.com>,
        Kan Liang <kan.liang@linux.intel.com>,
        linux-perf-users@vger.kernel.org, linux-kernel@vger.kernel.org,
        Yang Jihong <yangjihong1@huawei.com>,
        Stephane Eranian <eranian@google.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, Sep 18, 2023 at 3:43=E2=80=AFPM Ian Rogers <irogers@google.com> wro=
te:
>
> On Sat, Sep 16, 2023 at 5:46=E2=80=AFPM Mingwei Zhang <mizhang@google.com=
> wrote:
> >
> > On Fri, Sep 15, 2023 at 9:10=E2=80=AFPM Ian Rogers <irogers@google.com>=
 wrote:
> > >
> > > Dummy events are created with an attribute where the period and freq
> > > are zero. evsel__config will then see the uninitialized values and
> > > initialize them in evsel__default_freq_period. As fequency mode is
> > > used by default the dummy event would be set to use frequency
> > > mode. However, this has no effect on the dummy event but does cause
> > > unnecessary timers/interrupts. Avoid this overhead by setting the
> > > period to 1 for dummy events.
> > >
> > > evlist__add_aux_dummy calls evlist__add_dummy then sets freq=3D0 and
> > > period=3D1. This isn't necessary after this change and so the setting=
 is
> > > removed.
> > >
> > > From Stephane:
> > >
> > > The dummy event is not counting anything. It is used to collect mmap
> > > records and avoid a race condition during the synthesize mmap phase o=
f
> > > perf record. As such, it should not cause any overhead during active
> > > profiling. Yet, it did. Because of a bug the dummy event was
> > > programmed as a sampling event in frequency mode. Events in that mode
> > > incur more kernel overheads because on timer tick, the kernel has to
> > > look at the number of samples for each event and potentially adjust
> > > the sampling period to achieve the desired frequency. The dummy event
> > > was therefore adding a frequency event to task and ctx contexts we ma=
y
> > > otherwise not have any, e.g., perf record -a -e
> > > cpu/event=3D0x3c,period=3D10000000/. On each timer tick the
> > > perf_adjust_freq_unthr_context() is invoked and if ctx->nr_freq is
> > > non-zero, then the kernel will loop over ALL the events of the contex=
t
> > > looking for frequency mode ones. In doing, so it locks the context,
> > > and enable/disable the PMU of each hw event. If all the events of the
> > > context are in period mode, the kernel will have to traverse the list=
 for
> > > nothing incurring overhead. The overhead is multiplied by a very larg=
e
> > > factor when this happens in a guest kernel. There is no need for the
> > > dummy event to be in frequency mode, it does not count anything and
> > > therefore should not cause extra overhead for no reason.
> > >
> > > Fixes: 5bae0250237f ("perf evlist: Introduce perf_evlist__new_dummy c=
onstructor")
> > > Reported-by: Stephane Eranian <eranian@google.com>
> > > Signed-off-by: Ian Rogers <irogers@google.com>
> > > ---
> > >  tools/perf/util/evlist.c | 5 +++--
> > >  1 file changed, 3 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
> > > index 25c3ebe2c2f5..e36da58522ef 100644
> > > --- a/tools/perf/util/evlist.c
> > > +++ b/tools/perf/util/evlist.c
> > > @@ -251,6 +251,9 @@ static struct evsel *evlist__dummy_event(struct e=
vlist *evlist)
> > >                 .type   =3D PERF_TYPE_SOFTWARE,
> > >                 .config =3D PERF_COUNT_SW_DUMMY,
> > >                 .size   =3D sizeof(attr), /* to capture ABI version *=
/
> > > +               /* Avoid frequency mode for dummy events to avoid ass=
ociated timers. */
> > > +               .freq =3D 0,
> > > +               .sample_period =3D 1,
> > >         };
> > >
> > >         return evsel__new_idx(&attr, evlist->core.nr_entries);
> > > @@ -277,8 +280,6 @@ struct evsel *evlist__add_aux_dummy(struct evlist=
 *evlist, bool system_wide)
> > >         evsel->core.attr.exclude_kernel =3D 1;
> > >         evsel->core.attr.exclude_guest =3D 1;
> > >         evsel->core.attr.exclude_hv =3D 1;
> > > -       evsel->core.attr.freq =3D 0;
> > > -       evsel->core.attr.sample_period =3D 1;
> > >         evsel->core.system_wide =3D system_wide;
> > >         evsel->no_aux_samples =3D true;
> > >         evsel->name =3D strdup("dummy:u");
> > > --
> > > 2.42.0.459.ge4e396fd5e-goog
> > >
> >
> > Thank you very much for the change. I have one quick question about
> > the PMU unthrottling logic. When I am looking into the function
> > perf_adjust_freq_unthr_context(), I see the loop with PMU stop and
> > start in each iteration. Is there a good way to avoid this PMU reset
> > operation while quickly figuring out the event in frequency mode?
>
> Agreed. I think before the pmu_disable could be avoided for this conditio=
n:
> ```
> if (event->hw.interrupts !=3D MAX_INTERRUPTS &&
>     (!event->attr.freq || !event->attr.sample_freq))
>         continue;
> ```
> Fixing up the event stop/start looks harder.
>

Right, I think putting the check early before pmu_disable() is already
a great optimization. The only concern I initially had was whether
event->hw.interrupts can be accessed before we disable the pmu. But
after checking this field in other locations, I don't see any problem
at all.

Thanks.
-Mingwei