Re: [PATCH] perf stat: Fix shadow stats for clock events

[Arnaldo Carvalho de Melo]

Em Tue, Nov 27, 2018 at 01:50:44PM +0530, Ravi Bangoria escreveu:
> 
> 
> On 11/16/18 7:05 PM, Jiri Olsa wrote:
> > On Fri, Nov 16, 2018 at 09:58:43AM +0530, Ravi Bangoria wrote:
> >> Commit 0aa802a79469 ("perf stat: Get rid of extra clock display
> >> function") introduced scale and unit for clock events. Thus,
> >> perf_stat__update_shadow_stats() now saves scaled values of
> >> clock events in msecs, instead of original nsecs. But while
> >> calculating values of shadow stats we still consider clock
> >> event values in nsecs. This results in a wrong shadow stat
> >> values. Ex,
> >>
> >>   # ./perf stat -e task-clock,cycles ls
> >>     <SNIP>
> >>               2.60 msec task-clock:u    #    0.877 CPUs utilized
> >>          2,430,564      cycles:u        # 1215282.000 GHz
> >>
> >> Fix this by saving original nsec values for clock events in
> >> perf_stat__update_shadow_stats(). After patch:
> >>
> >>   # ./perf stat -e task-clock,cycles ls
> >>     <SNIP>
> >>               3.14 msec task-clock:u    #    0.839 CPUs utilized
> >>          3,094,528      cycles:u        #    0.985 GHz
> >>
> >> Reported-by: Anton Blanchard <anton@xxxxxxxxx>
> >> Suggested-by: Jiri Olsa <jolsa@xxxxxxxxxx>
> >> Fixes: 0aa802a79469 ("perf stat: Get rid of extra clock display function")
> >> Signed-off-by: Ravi Bangoria <ravi.bangoria@xxxxxxxxxxxxx>
> > 
> > Reviewed-by: Jiri Olsa <jolsa@xxxxxxxxxx>
> 
> Hi Arnaldo, Please pull this.

I did it yesterday, its just that I installed a new notebook and the
MTA was down :-\ All should be fixed by now and the patch should go
Ingo's way soon.

- Arnaldo