From: Sebastian Andrzej Siewior Date: Mon, 6 Sep 2021 11:40:48 +0200 Subject: [PATCH] sched: Make the idle timer expire always in hardirq context. The intel powerclamp driver will setup a per-CPU worker with RT priority. The worker will then invoke play_idle() in which it remains in the idle poll loop until it is stopped by the timer it started earlier. That timer needs to expire in hardirq context on PREEMPT_RT. Otherwise the timer will expire in ksoftirqd as a SOFT timer but that task won't be scheduled on the CPU because its priority is lower than the priority of the worker which is in the idle loop. Always expire the idle timer in hardirq context. Fixes:c1de45ca831ac ("sched/idle: Add support for tasks that inject idle") Reported-by: Thomas Gleixner Signed-off-by: Sebastian Andrzej Siewior Link: https://lkml.kernel.org/r/20210906113034.jgfxrjdvxnjqgtmc@linutronix.de --- kernel/sched/idle.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/kernel/sched/idle.c +++ b/kernel/sched/idle.c @@ -379,10 +379,10 @@ void play_idle_precise(u64 duration_ns, cpuidle_use_deepest_state(latency_ns); it.done = 0; - hrtimer_init_on_stack(&it.timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hrtimer_init_on_stack(&it.timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_HARD); it.timer.function = idle_inject_timer_fn; hrtimer_start(&it.timer, ns_to_ktime(duration_ns), - HRTIMER_MODE_REL_PINNED); + HRTIMER_MODE_REL_PINNED_HARD); while (!READ_ONCE(it.done)) do_idle();