Re: [RFC][PATCH] sched/core: Tweak wait_task_inactive() to force dequeue sched_delayed tasks

[K Prateek Nayak <kprateek.nayak@amd.com>]

Hello John,

On 4/22/2025 10:13 AM, John Stultz wrote:
> It was reported that in 6.12, smpboot_create_threads() was
> taking much longer then in 6.6.
> 
> I narrowed down the call path to:
>   smpboot_create_threads()
>   -> kthread_create_on_cpu()
>      -> kthread_bind()
>         -> __kthread_bind_mask()
>            ->wait_task_inactive()
> 
> Where in wait_task_inactive() we were regularly hitting the
> queued case, which sets a 1 tick timeout, which when called
> multiple times in a row, accumulates quickly into a long
> delay.
> 
> I noticed disabling the DELAY_DEQUEUE sched feature recovered
> the performance, and it seems the newly create tasks are usually
> sched_delayed and left on the runqueue.
> 
> So in wait_task_inactive() when we see the task
> task_on_rq_queued(p) and p->se.sched_delayed, manually dequeue
> the sched_delayed task with DEQUEUE_DELAYED, so we don't have to
> constantly wait a tick.
> 
> This seems to work, but I've only lightly tested it, so I'd love
> close review and feedback in case I've missed something in
> wait_task_inactive(), or if there is a simpler alternative
> approach.
> 
> Cc: Ingo Molnar <mingo@redhat.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Juri Lelli <juri.lelli@redhat.com>
> Cc: Vincent Guittot <vincent.guittot@linaro.org>
> Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
> Cc: Steven Rostedt <rostedt@goodmis.org>
> Cc: Ben Segall <bsegall@google.com>
> Cc: Mel Gorman <mgorman@suse.de>
> Cc: Valentin Schneider <vschneid@redhat.com>
> Cc: K Prateek Nayak <kprateek.nayak@amd.com>
> Cc: kernel-team@android.com
> Signed-off-by: John Stultz <jstultz@google.com>
> ---
>   kernel/sched/core.c | 8 ++++++++
>   1 file changed, 8 insertions(+)
> 
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index c81cf642dba05..43f0931a3cd8a 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -2287,6 +2287,14 @@ unsigned long wait_task_inactive(struct task_struct *p, unsigned int match_state
>   		running = task_on_cpu(rq, p);
>   		queued = task_on_rq_queued(p);
>   		ncsw = 0;
> +		/*
> +		 * If task is sched_delayed, force dequeue it, to avoid always
> +		 * hitting the tick timeout in the queued case
> +		 */
> +		if (!running && queued && p->se.sched_delayed) {

I think an "update_rq_clock(rq)" is required here before dequeue but other
than that ...

> +			dequeue_task(rq, p, DEQUEUE_SLEEP | DEQUEUE_DELAYED);

I think this makes sense since the comment above the hrtimeout considers
a "queued" task to be preempted and still runnable but a delayed task
doesn't fit that description.

You can perhaps move dequeuing of delayed task to just after the
task_rq_lock() bit since once the rq lock and the pi_lock are acquired,
task cannot be requeued and simply checking "p->se.sched_delayed"
should be enough. Something like:

     rq = task_rq_lock(p, &rf);
     trace_sched_wait_task(p);

     if (p->se.sched_delayed) {
             update_rq_clock(rq);
             dequeue_task(rq, p, DEQUEUE_SLEEP | DEQUEUE_DELAYED);
     }

     queued = task_on_rq_queued(p);
     ...

Similar pattern exists for __sched_setscheduler() and
rt_mutex_setprio().

Other than that, feel free to include:

Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>

> +			queued = task_on_rq_queued(p);
> +		}
>   		if ((match = __task_state_match(p, match_state))) {
>   			/*
>   			 * When matching on p->saved_state, consider this task

-- 
Thanks and Regards,
Prateek