[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAM9d7cgJEdkTRoQr7vP6MktDxemL3dJXjBuO-YtXNtFxjVMx6A@mail.gmail.com>
Date: Fri, 15 Jan 2021 16:01:34 +0900
From: Namhyung Kim <namhyung@...nel.org>
To: Jiri Olsa <jolsa@...hat.com>
Cc: Arnaldo Carvalho de Melo <acme@...nel.org>,
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>,
Stephane Eranian <eranian@...gle.com>,
Andi Kleen <ak@...ux.intel.com>,
Jin Yao <yao.jin@...ux.intel.com>,
Ian Rogers <irogers@...gle.com>
Subject: Re: [PATCH v2 1/2] perf stat: Introduce struct runtime_stat_data
On Thu, Jan 14, 2021 at 10:22 PM Jiri Olsa <jolsa@...hat.com> wrote:
>
> On Thu, Jan 14, 2021 at 12:25:39PM +0900, Namhyung Kim wrote:
> > Hi Jiri,
> >
> > On Wed, Jan 13, 2021 at 8:19 PM Jiri Olsa <jolsa@...hat.com> wrote:
> > >
> > > On Tue, Jan 12, 2021 at 03:14:30PM +0900, Namhyung Kim wrote:
> > > > To pass more info to the saved_value in the runtime_stat, add a new
> > > > struct runtime_stat_data. Currently it only has 'ctx' field but later
> > > > patch will add more.
> > > >
> > > > Suggested-by: Andi Kleen <ak@...ux.intel.com>
> > > > Signed-off-by: Namhyung Kim <namhyung@...nel.org>
> > > > ---
> > > > tools/perf/util/stat-shadow.c | 346 +++++++++++++++++-----------------
> > > > 1 file changed, 173 insertions(+), 173 deletions(-)
> > > >
> > > > diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
> > > > index 901265127e36..a1565b6e38f2 100644
> > > > --- a/tools/perf/util/stat-shadow.c
> > > > +++ b/tools/perf/util/stat-shadow.c
> > > > @@ -114,6 +114,10 @@ static struct saved_value *saved_value_lookup(struct evsel *evsel,
> > > >
> > > > rblist = &st->value_list;
> > > >
> > > > + /* don't use context info for clock events */
> > > > + if (type == STAT_NSECS)
> > > > + dm.ctx = 0;
> > > > +
> > >
> > > I think this should go to separate patch and be explained,
> > > the change is advertised as adding struct for arguments
> >
> > Actually it was already there and I found it during this work.
> > Basically it passes ctx for lookup but it uses 0 for time.
> > Please see below..
>
> ah nice, did not see that.. could be mentioned in the changelog ;-)
OK, will send v3.
Thanks,
Namhyung
Powered by blists - more mailing lists