linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] perf/core: Fix missing read event generation on task exit
@ 2025-07-20  0:04 thaumy.love
  2025-07-23  6:26 ` kernel test robot
  0 siblings, 1 reply; 2+ messages in thread
From: thaumy.love @ 2025-07-20  0:04 UTC (permalink / raw)
  To: Peter Zijlstra
  Cc: Ingo Molnar, Arnaldo Carvalho de Melo, Namhyung Kim,
	linux-perf-users, linux-kernel, Thaumy Cheng

From: Thaumy Cheng <thaumy.love@gmail.com>

For events with inherit_stat enabled, a "read" event will be generated
to collect per task event counts on task exit.

The call chain is as follows:

do_exit
  -> perf_event_exit_task
    -> perf_event_exit_task_context
      -> perf_event_exit_event
        -> perf_remove_from_context
          -> perf_child_detach
            -> sync_child_event
              -> perf_event_read_event

However, the child event context detaches the task too early in
perf_event_exit_task_context, which causes sync_child_event to never
generate the read event in this case, since child_event->ctx->task is
always set to TASK_TOMBSTONE. This patch intends to fix that.

This bug can be reproduced by running "perf record -s" and attaching to
any program that generates perf events in its child tasks. If we check
the result with "perf report -T", the last line of the report will leave
an empty table like "# PID  TID", which is expected to contain the
per-task event counts by design.

Signed-off-by: Thaumy Cheng <thaumy.love@gmail.com>
---
 kernel/events/core.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/kernel/events/core.c b/kernel/events/core.c
index 22fdf0c187cd..266b9eabb342 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -14057,6 +14057,17 @@ static void perf_event_exit_task_context(struct task_struct *task, bool exit)
 	 */
 	mutex_lock(&ctx->mutex);

+	/*
+	 * Report the task dead after unscheduling the events so that we
+	 * won't get any samples after PERF_RECORD_EXIT. We can however still
+	 * get a few PERF_RECORD_READ events.
+	 */
+	if (exit)
+		perf_event_task(task, ctx, 0);
+
+	list_for_each_entry_safe(child_event, next, &ctx->event_list, event_entry)
+		perf_event_exit_event(child_event, ctx, false);
+
 	/*
 	 * In a single ctx::lock section, de-schedule the events and detach the
 	 * context from the task such that we cannot ever get it scheduled back
@@ -14081,17 +14092,6 @@ static void perf_event_exit_task_context(struct task_struct *task, bool exit)
 	if (clone_ctx)
 		put_ctx(clone_ctx);

-	/*
-	 * Report the task dead after unscheduling the events so that we
-	 * won't get any samples after PERF_RECORD_EXIT. We can however still
-	 * get a few PERF_RECORD_READ events.
-	 */
-	if (exit)
-		perf_event_task(task, ctx, 0);
-
-	list_for_each_entry_safe(child_event, next, &ctx->event_list, event_entry)
-		perf_event_exit_event(child_event, ctx, false);
-
 	mutex_unlock(&ctx->mutex);

 	if (!exit) {
--
2.49.0


^ permalink raw reply related	[flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-07-23  6:26 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-07-20  0:04 [PATCH] perf/core: Fix missing read event generation on task exit thaumy.love
2025-07-23  6:26 ` kernel test robot

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).