sched_ext: Pass held rq to SCX_CALL_OP() for core_sched_before

[ Upstream commit 4155fb4 ]

scx_prio_less() runs from core-sched's pick_next_task() path with rq
locked but invokes ops.core_sched_before() with NULL locked_rq, leaving
scx_locked_rq_state NULL. If the BPF callback calls a kfunc that
re-acquires rq based on scx_locked_rq() - e.g. scx_bpf_cpuperf_set(cpu)
- it re-acquires the already-held rq.

Pass task_rq(a).

Fixes: 7b0888b ("sched_ext: Implement core-sched support")
Cc: stable@vger.kernel.org # v6.12+
Reported-by: Chris Mason <clm@meta.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
[ adapted call to use stable's single `sch`/`SCX_KF_REST` mask and `scx_rq_bypassing(task_rq(a))` signature ]
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Author: htejun

kernel/sched/ext.c

@@ -2522,7 +2522,7 @@ bool scx_prio_less(const struct task_struct *a, const struct task_struct *b,
 	if (SCX_HAS_OP(sch, core_sched_before) &&
 	    !scx_rq_bypassing(task_rq(a)))
 		return SCX_CALL_OP_2TASKS_RET(sch, SCX_KF_REST, core_sched_before,
-					      NULL,
+					      task_rq(a),
 					      (struct task_struct *)a,
 					      (struct task_struct *)b);
 	else