Re: [PATCH] perf jevents: Silence warning for ArchStd files

[Ian Rogers]

On Fri, May 7, 2021 at 2:00 AM John Garry <john.garry@xxxxxxxxxx> wrote:
>
> On 06/05/2021 23:56, Ian Rogers wrote:
> > json files in the level 1 directory are used for ArchStd events (see
> > preprocess_arch_std_files), as such they shouldn't be warned about.
> >
> > Signed-off-by: Ian Rogers <irogers@xxxxxxxxxx>
> > ---
> >   tools/perf/pmu-events/jevents.c | 6 ++++--
> >   1 file changed, 4 insertions(+), 2 deletions(-)
> >
> > diff --git a/tools/perf/pmu-events/jevents.c b/tools/perf/pmu-events/jevents.c
> > index ed4f0bd72e5a..7422b0ea8790 100644
> > --- a/tools/perf/pmu-events/jevents.c
> > +++ b/tools/perf/pmu-events/jevents.c
> > @@ -1123,8 +1123,10 @@ static int process_one_file(const char *fpath, const struct stat *sb,
> >                       mapfile = strdup(fpath);
> >                       return 0;
> >               }
> > -
> > -             pr_info("%s: Ignoring file %s\n", prog, fpath);
> > +             if (is_json_file(bname))
> > +                     pr_debug("%s: ArchStd json is preprocessed %s\n", prog, fpath)
> We could get more elaborate and add the same first debug print in
> process_one_file() to preprocess_arch_std_file() to give the allusion
> that they are preprocessed, and change the logic not print that for arch
> std files (in process_one_file()). But not sure it's worth it.
>
> Or else we could also just omit any print here for archstd files here.

I thought about just dropping the print in the json case but then a
comment would be nice, the pr_debug is a comment and is somewhat
intention revealing. If you think it is overkill then it is ok to
change.

Thanks,
Ian

> But I suppose what you are doing is ok...
>
> Thanks,
> John
>
> > +             else
> > +                     pr_info("%s: Ignoring file %s\n", prog, fpath);
> >               return 0;
> >       }
> >
> >
>