perf tools: Fix Intel PT timestamp handling
authorAdrian Hunter <adrian.hunter@intel.com>
Thu, 20 Aug 2015 08:51:32 +0000 (11:51 +0300)
committerArnaldo Carvalho de Melo <acme@redhat.com>
Fri, 21 Aug 2015 13:29:23 +0000 (10:29 -0300)
Events that don't sample the timestamp have a timestamp value of -1.

Intel PT processing wasn't taking that into account.

This is particularly noticeable with Intel BTS because timestamps are
not requested by default.

Then, if the conversion of -1 to TSC results in a small number, the
processing is unaffected.

However if the conversion results in a big number, then the data is
processed prematurely before relevant sideband data like mmap events,
which in turn results in samples with unknown dsos.

Commiter note:

Since BTS wasn't upstream, I split the patch to fold the BTS part with
the patch introducing it, to avoid having this bug in the commit
history. PT was already upstream, so this patch contains that part.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1440060692-5585-1-git-send-email-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
tools/perf/util/intel-pt.c

index 2a4a412..a5acd2f 100644 (file)
@@ -1450,7 +1450,7 @@ static int intel_pt_process_event(struct perf_session *session,
                return -EINVAL;
        }
 
-       if (sample->time)
+       if (sample->time && sample->time != (u64)-1)
                timestamp = perf_time_to_tsc(sample->time, &pt->tc);
        else
                timestamp = 0;