On Tue, Aug 19, 2025 at 09:49:20AM +0200, Nam Cao wrote:

On Fri, Aug 15, 2025 at 03:40:16PM +0200, Peter Zijlstra wrote:

quoted

On Wed, Aug 06, 2025 at 10:01:20AM +0200, Nam Cao wrote:

quoted

+/*
+ * The two trace points below may not work as expected for fair tasks due
+ * to delayed dequeue. See:
+ * https://lore.kernel.org/lkml/179674c6-f82a-4718-ace2-67b5e672fdee@amd.com/ (local)
+ */

quoted

+DECLARE_TRACE(dequeue_task,
+	TP_PROTO(int cpu, struct task_struct *task),
+	TP_ARGS(cpu, task));
+

quoted

@@ -2119,7 +2121,11 @@ inline bool dequeue_task(struct rq *rq, struct task_struct *p, int flags)
 	 * and mark the task ->sched_delayed.
 	 */
 	uclamp_rq_dec(rq, p);
-	return p->sched_class->dequeue_task(rq, p, flags);
+	if (p->sched_class->dequeue_task(rq, p, flags)) {
+		trace_dequeue_task_tp(rq->cpu, p);
+		return true;
+	}
+	return false;
 }

Hurmpff.. that's not very nice.

How about something like:

dequeue_task():
	...
	ret = p->sched_class->dequeue_task(rq, p, flags);
	if (trace_dequeue_task_p_enabled() && !(flags & DEQUEUE_SLEEP))
		__trace_dequeue_task_tp(rq->cpu, p);
	return ret;


__block_task():
	trace_dequeue_task_tp(rq->cpu, p);
	...


Specifically, only DEQUEUE_SLEEP is allowed to fail, and DEQUEUE_SLEEP
will eventually cause __block_task() to be called, either directly, or
delayed.

Thanks for the suggestion, this makes sense.

From my understanding, it makes the tracepoints work correctly for fair
tasks too, so I will get rid of the comment.

Just so indeed :-)