perf annotate: Fix segmentation fault
authorIngo Molnar <mingo@elte.hu>
Tue, 18 Aug 2009 11:52:28 +0000 (13:52 +0200)
committerIngo Molnar <mingo@elte.hu>
Tue, 18 Aug 2009 12:00:52 +0000 (14:00 +0200)
Linus reported this perf annotate segfault:

        [torvalds@nehalem git]$ perf annotate unmap_vmas
        Segmentation fault

        #0  map__clone (self=<value optimized out>) at builtin-annotate.c:236
        #1  thread__fork (self=<value optimized out>) at builtin-annotate.c:372

The bug here was that builtin-annotate.c was a copy of
builtin-report.c and a threading related fix to builtin-report.c
didnt get propagated to builtin-annotate.c ...

Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
tools/perf/builtin-annotate.c

index 1dba568..343e7b1 100644 (file)
@@ -980,6 +980,13 @@ process_fork_event(event_t *event, unsigned long offset, unsigned long head)
                (void *)(long)(event->header.size),
                event->fork.pid, event->fork.ppid);
 
+       /*
+        * A thread clone will have the same PID for both
+        * parent and child.
+        */
+       if (thread == parent)
+               return 0;
+
        if (!thread || !parent || thread__fork(thread, parent)) {
                dprintf("problem processing PERF_EVENT_FORK, skipping event.\n");
                return -1;