Re: [PATCH v2 3/3] sched_ext: Guarantee rq lock on scx_bpf_cpu_rq()

[Andrea Righi <arighi@nvidia.com>]

On Mon, Aug 04, 2025 at 12:27:43PM +0100, Christian Loehle wrote:
> Most fields in scx_bpf_cpu_rq() assume that its rq_lock is held.
> Furthermore they become meaningless without rq lock, too.
> Only return scx_bpf_cpu_rq() if we hold rq lock of that rq.
> 
> All upstream scx schedulers can be converted into the new
> scx_bpf_remote_curr() instead.
> 
> Signed-off-by: Christian Loehle <christian.loehle@arm.com>
> ---
>  kernel/sched/ext.c | 10 +++++++++-
>  1 file changed, 9 insertions(+), 1 deletion(-)
> 
> diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c
> index 1d9d9cbed0aa..0b05ddc1f100 100644
> --- a/kernel/sched/ext.c
> +++ b/kernel/sched/ext.c
> @@ -7420,10 +7420,18 @@ __bpf_kfunc s32 scx_bpf_task_cpu(const struct task_struct *p)
>   */
>  __bpf_kfunc struct rq *scx_bpf_cpu_rq(s32 cpu)
>  {
> +	struct rq *rq;
> +
>  	if (!kf_cpu_valid(cpu, NULL))
>  		return NULL;
>  
> -	return cpu_rq(cpu);
> +	rq = cpu_rq(cpu);
> +	if (rq != scx_locked_rq_state) {

I think you want to check rq != scx_locked_rq(), since scx_locked_rq_state
is a per-CPU variable.

We may also want to add a preempt_disable/enable() for consistency. How
about something like this?

	preempt_disable();
	rq = cpu_rq(cpu);
	if (rq != scx_locked_rq()) {
		scx_kf_error("Accessing CPU%d rq from CPU%d without holding its lock",
			     cpu, smp_processor_id());
		rq = NULL;
	}
	preempt_enable();

Thanks,
-Andrea

> +		scx_kf_error("Accessing not locked rq %d", cpu);
> +		return NULL;
> +	}
> +
> +	return rq;
>  }
>  
>  struct task_struct *bpf_task_acquire(struct task_struct *p);
> -- 
> 2.34.1
>