[PATCH 12/12] hrtimer: Apply READ_ONCE() to lockless base->running loads
COOLING5d
From: "Paul E. McKenney" <paulmck@kernel.org>
Date: 2026-09-19 00:14:33
Also in:
io-uring, linux-fsdevel, lkml
Subsystem:
high-resolution timers, timer wheel, clockevents, the rest · Maintainers:
Anna-Maria Behnsen, Frederic Weisbecker, Thomas Gleixner, Linus Torvalds
Updates to base->running are protected by the hrtimer base lock, but some loads are lockless. Therefore, prevent compiler mischief by applying READ_ONCE() to the lockless loads. KCSAN located this issue. Signed-off-by: Paul E. McKenney <paulmck@kernel.org> Cc: Anna-Maria Behnsen <anna-maria@linutronix.de> Cc: Frederic Weisbecker <frederic@kernel.org> Cc: Thomas Gleixner <tglx@kernel.org> --- kernel/time/hrtimer.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c
index 91ac2b517637..b3b4776c17a1 100644
--- a/kernel/time/hrtimer.c
+++ b/kernel/time/hrtimer.c@@ -1997,7 +1997,7 @@ bool hrtimer_active(const struct hrtimer *timer) base = READ_ONCE(timer->base); seq = raw_read_seqcount_begin(&base->seq); - if (timer->is_queued || base->running == timer) + if (timer->is_queued || READ_ONCE(base->running) == timer) return true; } while (read_seqcount_retry(&base->seq, seq) || base != READ_ONCE(timer->base));
@@ -2034,7 +2034,7 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, struct hrtimer_cloc lockdep_assert_held(&cpu_base->lock); debug_hrtimer_deactivate(timer); - base->running = timer; + WRITE_ONCE(base->running, timer); /* * Separate the ->running assignment from the ->is_queued assignment.
@@ -2093,7 +2093,7 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, struct hrtimer_cloc raw_write_seqcount_barrier(&base->seq); WARN_ON_ONCE(base->running != timer); - base->running = NULL; + WRITE_ONCE(base->running, NULL); } static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now,
--
2.40.1