Re: [PATCH v1 1/2] perf test attr: Update no event/metric expectations

[Arnaldo Carvalho de Melo]

Em Thu, May 18, 2023 at 01:45:06PM -0400, Liang, Kan escreveu:
> Hi Arnaldo,
> 
> On 2023-05-18 8:55 a.m., Liang, Kan wrote:
> > On 2023-05-18 8:20 a.m., Arnaldo Carvalho de Melo wrote:
> >> Em Thu, May 18, 2023 at 08:50:14AM +0100, James Clark escreveu:
> >>>
> >>> On 17/05/2023 23:57, Ian Rogers wrote:
> >>>> Previously hard coded events/metrics were used, update for the use of
> >>>> the TopdownL1 json metric group.
> >>>>
> >>>> Fixes: 94b1a603fca7 ("perf stat: Add TopdownL1 metric as a default if present")
> >>>> Signed-off-by: Ian Rogers <irogers@xxxxxxxxxx>
> >>>> ---
> >>>>  tools/perf/tests/attr/base-stat            |   2 +-
> >>>>  tools/perf/tests/attr/test-stat-default    |  80 ++++++++-----
> >>>>  tools/perf/tests/attr/test-stat-detailed-1 |  95 +++++++++------
> >>>>  tools/perf/tests/attr/test-stat-detailed-2 | 119 +++++++++++--------
> >>>>  tools/perf/tests/attr/test-stat-detailed-3 | 127 ++++++++++++---------
> >>>>  5 files changed, 249 insertions(+), 174 deletions(-)
> >>>>
> >>> Reviewed-by: James Clark <james.clark@xxxxxxx>
> >> Thanks, applied both patches to perf-tools.
> >>
> > Could you please hold patch 1 for a while?
> 
> Sorry for the noise.
> 
> Ian has addressed my concerns. The patch looks good to me now.
> It also fixes a "Failed" on my Cascade Lake machine.
> 
> Tested-by: Kan Liang <kan.liang@xxxxxxxxxxxxxxx>

Thanks, added it,

- Arnaldo
 
> But for long term, we probably need a better solution to avoid keeping
> adding such non-architectural events in the test case.