All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] perf: fix comm for processes with named threads
@ 2011-12-09 16:00 David Ahern
  2011-12-09 17:08 ` David Ahern
  0 siblings, 1 reply; 2+ messages in thread
From: David Ahern @ 2011-12-09 16:00 UTC (permalink / raw)
  To: acme, linux-kernel; +Cc: mingo, peterz, fweisbec, David Ahern

perf does not properly handle monitoring of processes with named threads.
For example:

$ ps -C myapp -L
  PID   LWP TTY          TIME CMD
25118 25118 ?        00:00:00 myapp
25118 25119 ?        00:00:00 myapp:worker

perf record -e cs -c 1 -fo /tmp/perf.data -p 25118 -- sleep 10
perf report --stdio -i /tmp/perf.data
   100.00%  myapp:worker  [kernel.kallsyms]  [k] perf_event_task_sched_out

The process name is set to the name of the last thread it finds for the
process.

The Problem:
perf-top and perf-record both create a thread_map of threads to be
monitored. That map is used in perf_event__synthesize_thread_map which
loops over the entries in thread_map and calls __event__synthesize_thread
to generate COMM and MMAP events.

__event__synthesize_thread calls perf_event__synthesize_comm which opens
/proc/pid/status, reads the name of the task and its thread group id.
That's all fine. The problem is that it then reads /proc/pid/task and
generates COMM events for each task it finds - but using the name found
in /proc/pid/status where pid is the thread of interest. This means the
name of the last thread processed sets the name for all threads in the
process - which is not good for multithreaded processes with named
threads.

The Fix:
perf_event__synthesize_comm has an input argument (full) that decides
whether to process task entries for each pid it is passed. It currently
never set to 0 (perf_event__synthesize_comm has a single caller and it
always passes the value 1). Let's fix that.

Add the full input argument to __event__synthesize_thread which passes
it to perf_event__synthesize_comm. For thread/process monitoring set full
to 0 which means COMM and MMAP events are only generated for the pid
passed to it. For system wide monitoring set full to 1 so that COMM events
are generated for all threads in a process.

Signed-off-by: David Ahern <dsahern@gmail.com>
---
 tools/perf/util/event.c |   35 ++++++++++++++++++++++++++++++-----
 1 files changed, 30 insertions(+), 5 deletions(-)

diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
index 437f8ca..6085727 100644
--- a/tools/perf/util/event.c
+++ b/tools/perf/util/event.c
@@ -260,10 +260,10 @@ int perf_event__synthesize_modules(perf_event__handler_t process,
 
 static int __event__synthesize_thread(union perf_event *comm_event,
 				      union perf_event *mmap_event,
-				      pid_t pid, perf_event__handler_t process,
+				      pid_t pid, int full, perf_event__handler_t process,
 				      struct perf_session *session)
 {
-	pid_t tgid = perf_event__synthesize_comm(comm_event, pid, 1, process,
+	pid_t tgid = perf_event__synthesize_comm(comm_event, pid, full, process,
 					    session);
 	if (tgid == -1)
 		return -1;
@@ -276,7 +276,7 @@ int perf_event__synthesize_thread_map(struct thread_map *threads,
 				      struct perf_session *session)
 {
 	union perf_event *comm_event, *mmap_event;
-	int err = -1, thread;
+	int err = -1, thread, j;
 
 	comm_event = malloc(sizeof(comm_event->comm) + session->id_hdr_size);
 	if (comm_event == NULL)
@@ -289,11 +289,36 @@ int perf_event__synthesize_thread_map(struct thread_map *threads,
 	err = 0;
 	for (thread = 0; thread < threads->nr; ++thread) {
 		if (__event__synthesize_thread(comm_event, mmap_event,
-					       threads->map[thread],
+					       threads->map[thread], 0,
 					       process, session)) {
 			err = -1;
 			break;
 		}
+
+		/* 
+		 * comm.pid is set to thread group id by
+		 * perf_event__synthesize_comm
+		 */
+		if ((int) comm_event->comm.pid != threads->map[thread]) {
+			bool need_leader = true;
+
+			/* is thread group leader in thread_map? */
+			for (j = 0; j < threads->nr; ++j) {
+				if ((int) comm_event->comm.pid == threads->map[j]) {
+					need_leader = false;
+					break;
+				}
+			}
+
+			/* if not, generate events for it */
+			if (need_leader &&
+			    __event__synthesize_thread(comm_event, mmap_event,
+					       comm_event->comm.pid, 0,
+					       process, session)) {
+				err = -1;
+				break;
+			}
+		}
 	}
 	free(mmap_event);
 out_free_comm:
@@ -329,7 +354,7 @@ int perf_event__synthesize_threads(perf_event__handler_t process,
 		if (*end) /* only interested in proper numerical dirents */
 			continue;
 
-		__event__synthesize_thread(comm_event, mmap_event, pid,
+		__event__synthesize_thread(comm_event, mmap_event, pid, 1,
 					   process, session);
 	}
 
-- 
1.7.7.3


^ permalink raw reply related	[flat|nested] 2+ messages in thread
end of thread, other threads:[~2011-12-09 17:08 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-12-09 16:00 [PATCH] perf: fix comm for processes with named threads David Ahern
2011-12-09 17:08 ` David Ahern

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.