[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20231114193840.4041-3-frederic@kernel.org>
Date: Tue, 14 Nov 2023 14:38:40 -0500
From: Frederic Weisbecker <frederic@...nel.org>
To: Peter Zijlstra <peterz@...radead.org>
Cc: LKML <linux-kernel@...r.kernel.org>,
Frederic Weisbecker <frederic@...nel.org>,
"Rafael J . Wysocki" <rafael@...nel.org>,
Daniel Lezcano <daniel.lezcano@...aro.org>,
Thomas Gleixner <tglx@...utronix.de>,
Anna-Maria Behnsen <anna-maria@...utronix.de>,
Ingo Molnar <mingo@...nel.org>
Subject: [PATCH 2/2] sched/timers: Explain why idle task schedules out on remote timer enqueue
Trying to avoid that didn't bring much value after testing, add comment
about this.
Acked-by: Rafael J. Wysocki <rafael@...nel.org>
Signed-off-by: Frederic Weisbecker <frederic@...nel.org>
---
kernel/sched/core.c | 22 ++++++++++++++++++++++
1 file changed, 22 insertions(+)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index a708d225c28e..50abc7eddb82 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1131,6 +1131,28 @@ static void wake_up_idle_cpu(int cpu)
if (cpu == smp_processor_id())
return;
+ /*
+ * Set TIF_NEED_RESCHED and send an IPI if in the non-polling
+ * part of the idle loop. This forces an exit from the idle loop
+ * and a round trip to schedule(). Now this could be optimized
+ * because a simple new idle loop iteration is enough to
+ * re-evaluate the next tick. Provided some re-ordering of tick
+ * nohz functions that would need to follow TIF_NR_POLLING
+ * clearing:
+ *
+ * - On most archs, a simple fetch_or on ti::flags with a
+ * "0" value would be enough to know if an IPI needs to be sent.
+ *
+ * - x86 needs to perform a last need_resched() check between
+ * monitor and mwait which doesn't take timers into account.
+ * There a dedicated TIF_TIMER flag would be required to
+ * fetch_or here and be checked along with TIF_NEED_RESCHED
+ * before mwait().
+ *
+ * However, remote timer enqueue is not such a frequent event
+ * and testing of the above solutions didn't appear to report
+ * much benefits.
+ */
if (set_nr_and_not_polling(rq->idle))
smp_send_reschedule(cpu);
else
--
2.42.1
Powered by blists - more mailing lists