* [PATCH] perf report: Fix regression when decoding intelPT traces @ 2017-12-14 0:57 Mathieu Poirier 2017-12-14 15:28 ` Jiri Olsa 0 siblings, 1 reply; 3+ messages in thread From: Mathieu Poirier @ 2017-12-14 0:57 UTC (permalink / raw) To: jolsa Cc: peterz, mingo, acme, alexander.shishkin, linux-kernel, mathieu.poirier Commit (93d10af26bb7 perf tools: Optimize sample parsing for ordered events) breaks intelPT trace decoding by invariably returning an error if the event type isn't a PERF_SAMPLE_TIME. With this patch the timestamp is initialised and processing is allowed to continue even if an error is returned by function perf_evlist__parse_sample_timestamp(). Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org> --- tools/perf/util/session.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 54e30f1bcbd7..20cdcf14232b 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -1512,7 +1512,7 @@ static s64 perf_session__process_event(struct perf_session *session, ret = perf_evlist__parse_sample_timestamp(evlist, event, ×tamp); if (ret) - return ret; + timestamp = -1ULL; ret = perf_session__queue_event(session, event, timestamp, file_offset); if (ret != -ETIME) -- 2.7.4 ^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] perf report: Fix regression when decoding intelPT traces 2017-12-14 0:57 [PATCH] perf report: Fix regression when decoding intelPT traces Mathieu Poirier @ 2017-12-14 15:28 ` Jiri Olsa 2017-12-14 16:03 ` Mathieu Poirier 0 siblings, 1 reply; 3+ messages in thread From: Jiri Olsa @ 2017-12-14 15:28 UTC (permalink / raw) To: Mathieu Poirier; +Cc: peterz, mingo, acme, alexander.shishkin, linux-kernel On Wed, Dec 13, 2017 at 05:57:54PM -0700, Mathieu Poirier wrote: > Commit (93d10af26bb7 perf tools: Optimize sample parsing for ordered > events) breaks intelPT trace decoding by invariably returning an error if > the event type isn't a PERF_SAMPLE_TIME. right, thanks for catchng that > > With this patch the timestamp is initialised and processing is allowed to > continue even if an error is returned by function > perf_evlist__parse_sample_timestamp(). > > Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org> > --- > tools/perf/util/session.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c > index 54e30f1bcbd7..20cdcf14232b 100644 > --- a/tools/perf/util/session.c > +++ b/tools/perf/util/session.c > @@ -1512,7 +1512,7 @@ static s64 perf_session__process_event(struct perf_session *session, > > ret = perf_evlist__parse_sample_timestamp(evlist, event, ×tamp); > if (ret) > - return ret; > + timestamp = -1ULL; we still have some -EFAULT error codes in there that we want to catch.. should it be more like in below? thanks, jirka --- diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 54e30f1bcbd7..9498aa0efe39 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -1508,10 +1508,10 @@ static s64 perf_session__process_event(struct perf_session *session, return perf_session__process_user_event(session, event, file_offset); if (tool->ordered_events) { - u64 timestamp; + u64 timestamp = -1ULL ret = perf_evlist__parse_sample_timestamp(evlist, event, ×tamp); - if (ret) + if (ret != -1) return ret; ret = perf_session__queue_event(session, event, timestamp, file_offset); ^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] perf report: Fix regression when decoding intelPT traces 2017-12-14 15:28 ` Jiri Olsa @ 2017-12-14 16:03 ` Mathieu Poirier 0 siblings, 0 replies; 3+ messages in thread From: Mathieu Poirier @ 2017-12-14 16:03 UTC (permalink / raw) To: Jiri Olsa Cc: Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Alexander Shishkin, linux-kernel On 14 December 2017 at 08:28, Jiri Olsa <jolsa@redhat.com> wrote: > On Wed, Dec 13, 2017 at 05:57:54PM -0700, Mathieu Poirier wrote: >> Commit (93d10af26bb7 perf tools: Optimize sample parsing for ordered >> events) breaks intelPT trace decoding by invariably returning an error if >> the event type isn't a PERF_SAMPLE_TIME. > > right, thanks for catchng that > >> >> With this patch the timestamp is initialised and processing is allowed to >> continue even if an error is returned by function >> perf_evlist__parse_sample_timestamp(). >> >> Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org> >> --- >> tools/perf/util/session.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c >> index 54e30f1bcbd7..20cdcf14232b 100644 >> --- a/tools/perf/util/session.c >> +++ b/tools/perf/util/session.c >> @@ -1512,7 +1512,7 @@ static s64 perf_session__process_event(struct perf_session *session, >> >> ret = perf_evlist__parse_sample_timestamp(evlist, event, ×tamp); >> if (ret) >> - return ret; >> + timestamp = -1ULL; > > we still have some -EFAULT error codes in there that we > want to catch.. should it be more like in below? > > thanks, > jirka > > > --- > diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c > index 54e30f1bcbd7..9498aa0efe39 100644 > --- a/tools/perf/util/session.c > +++ b/tools/perf/util/session.c > @@ -1508,10 +1508,10 @@ static s64 perf_session__process_event(struct perf_session *session, > return perf_session__process_user_event(session, event, file_offset); > > if (tool->ordered_events) { > - u64 timestamp; > + u64 timestamp = -1ULL > > ret = perf_evlist__parse_sample_timestamp(evlist, event, ×tamp); > - if (ret) > + if (ret != -1) > return ret; Yes, that's better - I'll send another revision. > > ret = perf_session__queue_event(session, event, timestamp, file_offset); ^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2017-12-14 16:03 UTC | newest] Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2017-12-14 0:57 [PATCH] perf report: Fix regression when decoding intelPT traces Mathieu Poirier 2017-12-14 15:28 ` Jiri Olsa 2017-12-14 16:03 ` Mathieu Poirier
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.