[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <YexjN1tV27bBQBkJ@kernel.org>
Date: Sat, 22 Jan 2022 17:04:07 -0300
From: Arnaldo Carvalho de Melo <acme@...nel.org>
To: German Gomez <german.gomez@....com>
Cc: Wei Li <liwei391@...wei.com>, Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Jiri Olsa <jolsa@...hat.com>,
Namhyung Kim <namhyung@...nel.org>,
Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...hat.com>,
linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org,
guohanjun@...wei.com, jinyao5@...wei.com
Subject: Re: [PATCH] perf script: Fix printing 'phys_addr' failure issue
Em Fri, Jan 21, 2022 at 11:24:57AM +0000, German Gomez escreveu:
> Hi Wei,
>
> This looks good to me. As a followup we could also fix the 'weight'
> field, which was introduced recently.
>
> On 21/01/2022 06:59, Wei Li wrote:
> > From: Yao Jin <jinyao5@...wei.com>
> >
> > Perf script was failed to print the phys_addr for SPE profiling.
> > One 'dummy' event is added by SPE profiling but it doesn't have PHYS_ADDR
> > attribute set, perf script then exits with error.
> >
> > Now referring to 'addr', use evsel__do_check_stype() to check the type.
> >
> > Before:
> >
> > # perf record -e arm_spe_0/branch_filter=0,ts_enable=1,pa_enable=1,load_filter=1,jitter=0,\
> > store_filter=0,min_latency=0,event_filter=2/ -p 4064384 -- sleep 3
> > # perf script -F pid,tid,addr,phys_addr
> > Samples for 'dummy:u' event do not have PHYS_ADDR attribute set. Cannot print 'phys_addr' field.
> >
> > After:
> >
> > # perf record -e arm_spe_0/branch_filter=0,ts_enable=1,pa_enable=1,load_filter=1,jitter=0,\
> > store_filter=0,min_latency=0,event_filter=2/ -p 4064384 -- sleep 3
> > # perf script -F pid,tid,addr,phys_addr
> > 4064384/4064384 ffff802f921be0d0 2f921be0d0
> > 4064384/4064384 ffff802f921be0d0 2f921be0d0
> >
> > Signed-off-by: Yao Jin <jinyao5@...wei.com>
> > Signed-off-by: Wei Li <liwei391@...wei.com>
>
> Reviewed-by: German Gomez <german.gomez@....com>
>
> Also it looks like the SPE samples are also missing the PHYS_ADDR flag.
> I think we'll need to fix the consistency of the flags.
Thanks, applied.
- Arnaldo
> Many thanks,
> German
>
> > ---
> > tools/perf/builtin-script.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
> > index ecd4f99a6c14..abae8184e171 100644
> > --- a/tools/perf/builtin-script.c
> > +++ b/tools/perf/builtin-script.c
> > @@ -515,7 +515,7 @@ static int evsel__check_attr(struct evsel *evsel, struct perf_session *session)
> > return -EINVAL;
> >
> > if (PRINT_FIELD(PHYS_ADDR) &&
> > - evsel__check_stype(evsel, PERF_SAMPLE_PHYS_ADDR, "PHYS_ADDR", PERF_OUTPUT_PHYS_ADDR))
> > + evsel__do_check_stype(evsel, PERF_SAMPLE_PHYS_ADDR, "PHYS_ADDR", PERF_OUTPUT_PHYS_ADDR, allow_user_set))
> > return -EINVAL;
> >
> > if (PRINT_FIELD(DATA_PAGE_SIZE) &&
--
- Arnaldo
Powered by blists - more mailing lists