sched_ext: Disable preemption between scx_claim_exit() and kicking helper work

scx_claim_exit() atomically sets exit_kind, which prevents scx_error() from triggering further error handling. After claiming exit, the caller must kick the helper kthread work which initiates bypass mode and teardown.

If the calling task gets preempted between claiming exit and kicking the helper work, and the BPF scheduler fails to schedule it back (since error handling is now disabled), the helper work is never queued, bypass mode never activates, tasks stop being dispatched, and the system wedges.

Disable preemption across scx_claim_exit() and the subsequent work kicking in all callers - scx_disable() and scx_vexit(). Add lockdep_assert_preemption_disabled() to scx_claim_exit() to enforce the requirement.

Fixes a69040ed57f5 ("sched_ext: Simplify breather mechanism with scx_aborting flag")

Cc [email protected] # v6.19+

Signed-off-by Tejun Heo <[email protected]>

---
 kernel/sched/ext.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

kernel/sched/ext.c +12 -0

index c18e81e8ef51..9280381f8923 100644
@@ -4423,10 +4423,19 @@ static void scx_disable_workfn(struct kthread_work *work)
 	scx_bypass(false);
 }
 
+/*
+ * Claim the exit on @sch. The caller must ensure that the helper kthread work
+ * is kicked before the current task can be preempted. Once exit_kind is
+ * claimed, scx_error() can no longer trigger, so if the current task gets
+ * preempted and the BPF scheduler fails to schedule it back, the helper work
+ * will never be kicked and the whole system can wedge.
+ */
 static bool scx_claim_exit(struct scx_sched *sch, enum scx_exit_kind kind)
 {
 	int none = SCX_EXIT_NONE;
 
+	lockdep_assert_preemption_disabled();
+
 	if (!atomic_try_cmpxchg(&sch->exit_kind, &none, kind))
 		return false;
 

@@ -4449,6 +4458,7 @@ static void scx_disable(enum scx_exit_kind kind)
 	rcu_read_lock();
 	sch = rcu_dereference(scx_root);
 	if (sch) {
+		guard(preempt)();
 		scx_claim_exit(sch, kind);
 		kthread_queue_work(sch->helper, &sch->disable_work);
 	}

@@ -4771,6 +4781,8 @@ static bool scx_vexit(struct scx_sched *sch,
 {
 	struct scx_exit_info *ei = sch->exit_info;
 
+	guard(preempt)();
+
 	if (!scx_claim_exit(sch, kind))
 		return false;
 

Signature · 1 line

2.53.0

I think the same race window already existed even before this commit, we were just doing atomic_try_cmpxchg() directly, instead of using the scx_claim_exit() helper.

So, probably the right target should be f0e1a0643a59b ("sched_ext:
Implement BPF extensible scheduler class").