Re: [PATCH sched_ext/for-6.15] sched_ext: Implement SCX_OPS_ALLOW_QUEUED_WAKEUP

[Tejun Heo <tj@kernel.org>]

On Wed, Feb 12, 2025 at 01:08:31PM -1000, Tejun Heo wrote:
> A task wakeup can be either processed on the waker's CPU or bounced to the
> wakee's previous CPU using an IPI (ttwu_queue). Bouncing to the wakee's CPU
> avoids the waker's CPU locking and accessing the wakee's rq which can be
> expensive across cache and node boundaries.
> 
> When ttwu_queue path is taken, select_task_rq() and thus ops.select_cpu()
> are skipped. As this confused some BPF schedulers, there wasn't a good way
> for a BPF scheduler to tell whether idle CPU selection has been skipped,
> ops.enqueue() couldn't insert tasks into foreign local DSQs, and the
> performance difference on machines with simple toplogies were minimal,
> sched_ext disabled ttwu_queue.
> 
> However, this optimization makes noticeable difference on more complex
> topologies and a BPF scheduler now has an easy way tell whether
> ops.select_cpu() was skipped since 9b671793c7d9 ("sched_ext, scx_qmap: Add
> and use SCX_ENQ_CPU_SELECTED") and can insert tasks into foreign local DSQs
> since 5b26f7b920f7 ("sched_ext: Allow SCX_DSQ_LOCAL_ON for direct
> dispatches").
> 
> Implement SCX_OPS_ALLOW_QUEUED_WAKEUP which allows BPF schedulers to choose
> to enable ttwu_queue optimization.
> 
> Signed-off-by: Tejun Heo <tj@kernel.org>
> Reported-by: Neel Natu <neelnatu@google.com>
> Reported-by: Barret Rhoden <brho@google.com>
> Cc: Peter Zijlstra (Intel) <peterz@infradead.org>

Applying to sched_ext/for-6.15. Peter, the core side change is trivial but
please holler for any concerns.

Thanks.

-- 
tejun