Em Tue, Aug 30, 2022 at 09:48:40AM -0700, Ian Rogers escreveu:
> Previous behavior is to segfault if there is no CPU PMU table and a
> metric is sought. To reproduce compile with NO_JEVENTS=1 then request
> a metric, for example, "perf stat -M IPC true".
>
> Fixes: 00facc760903 ("perf jevents: Switch build to use jevents.py")
> Signed-off-by: Ian Rogers <irogers@google.com>
Before:
$ make -k NO_JEVENTS=1 BUILD_BPF_SKEL=1 O=/tmp/build/perf-urgent -C tools/perf install-bin
$ perf stat -M IPC true
Segmentation fault (core dumped)
$
After:
$ perf stat -M IPC true
Usage: perf stat [<options>] [<command>]
-M, --metrics <metric/metric group list>
monitor specified metrics or metric groups (separated by ,)
$
Better, but perhaps we could provide some better message as to why the
metric provided isn't acceptable?
Anyway, applied to perf/urgent.
Thanks,
- Arnaldo
> ---
> tools/perf/util/metricgroup.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/tools/perf/util/metricgroup.c b/tools/perf/util/metricgroup.c
> index ad5cacdecd81..18aae040d61d 100644
> --- a/tools/perf/util/metricgroup.c
> +++ b/tools/perf/util/metricgroup.c
> @@ -1655,6 +1655,9 @@ int metricgroup__parse_groups(const struct option *opt,
> struct evlist *perf_evlist = *(struct evlist **)opt->value;
> const struct pmu_events_table *table = pmu_events_table__find();
>
> + if (!table)
> + return -EINVAL;
> +
> return parse_groups(perf_evlist, str, metric_no_group,
> metric_no_merge, NULL, metric_events, table);
> }
> --
> 2.37.2.672.g94769d06f0-goog
--
- Arnaldo