sched_ext: Fix exit_cpu accuracy for lockup paths

handle_lockup() uses raw_smp_processor_id() for exit_cpu, which is wrong
for two paths:

- scx_hardlockup_irq_workfn() has the hung CPU in a local variable but
  irq_work may run elsewhere. Pass the local cpu explicitly.
- scx_rcu_cpu_stall() records the detector CPU rather than the stalled
  one. Pass -1 for now. The next patch fixes this properly.

Signed-off-by: Cheng-Yang Chou <yphbchou0911@gmail.com>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>

2 files changed, 9 insertions, 8 deletions

diff --git a/kernel/sched/ext/ext.c b/kernel/sched/ext/ext.c
index f08ef32b9bd6..7044a591e4c5 100644
--- a/kernel/sched/ext/ext.c
+++ b/kernel/sched/ext/ext.c
@@ -5092,6 +5092,7 @@ bool scx_allow_ttwu_queue(const struct task_struct *p)
 
 /**
  * handle_lockup - sched_ext common lockup handler
+ * @exit_cpu: CPU to record in exit_info. Pass the stalled/hung CPU, not current.
  * @fmt: format string
  *
  * Called on system stall or lockup condition and initiates abort of sched_ext
@@ -5101,7 +5102,7 @@ bool scx_allow_ttwu_queue(const struct task_struct *p)
  * resolve the lockup. %false if sched_ext is not enabled or abort was already
  * initiated by someone else.
  */
-static __printf(1, 2) bool handle_lockup(const char *fmt, ...)
+static __printf(2, 3) bool handle_lockup(int exit_cpu, const char *fmt, ...)
 {
 	struct scx_sched *sch;
 	va_list args;
@@ -5117,7 +5118,7 @@ static __printf(1, 2) bool handle_lockup(const char *fmt, ...)
 	case SCX_ENABLING:
 	case SCX_ENABLED:
 		va_start(args, fmt);
-		ret = scx_verror(sch, fmt, args);
+		ret = scx_vexit(sch, SCX_EXIT_ERROR, 0, exit_cpu, fmt, args);
 		va_end(args);
 		return ret;
 	default:
@@ -5139,7 +5140,7 @@ static __printf(1, 2) bool handle_lockup(const char *fmt, ...)
  */
 bool scx_rcu_cpu_stall(void)
 {
-	return handle_lockup("RCU CPU stall detected!");
+	return handle_lockup(-1, "RCU CPU stall detected!");
 }
 
 /**
@@ -5154,11 +5155,13 @@ bool scx_rcu_cpu_stall(void)
  */
 void scx_softlockup(u32 dur_s)
 {
-	if (!handle_lockup("soft lockup - CPU %d stuck for %us", smp_processor_id(), dur_s))
+	int cpu = smp_processor_id();
+
+	if (!handle_lockup(cpu, "soft lockup - CPU %d stuck for %us", cpu, dur_s))
 		return;
 
 	printk_deferred(KERN_ERR "sched_ext: Soft lockup - CPU %d stuck for %us, disabling BPF scheduler\n",
-			smp_processor_id(), dur_s);
+			cpu, dur_s);
 }
 
 /*
@@ -5173,7 +5176,7 @@ static void scx_hardlockup_irq_workfn(struct irq_work *work)
 {
 	int cpu = atomic_xchg(&scx_hardlockup_cpu, -1);
 
-	if (cpu >= 0 && handle_lockup("hard lockup - CPU %d", cpu))
+	if (cpu >= 0 && handle_lockup(cpu, "hard lockup - CPU %d", cpu))
 		printk_deferred(KERN_ERR "sched_ext: Hard lockup - CPU %d, disabling BPF scheduler\n",
 				cpu);
 }
diff --git a/kernel/sched/ext/internal.h b/kernel/sched/ext/internal.h
index f4ba67799b0f..65eceefcf5e2 100644
--- a/kernel/sched/ext/internal.h
+++ b/kernel/sched/ext/internal.h
@@ -1543,8 +1543,6 @@ __printf(5, 6) bool __scx_exit(struct scx_sched *sch, enum scx_exit_kind kind,
 	__scx_exit(sch, kind, exit_code, raw_smp_processor_id(), fmt, ##args)
 #define scx_error(sch, fmt, args...)						\
 	scx_exit((sch), SCX_EXIT_ERROR, 0, fmt, ##args)
-#define scx_verror(sch, fmt, args)						\
-	scx_vexit((sch), SCX_EXIT_ERROR, 0, raw_smp_processor_id(), fmt, args)
 
 /*
  * Return the rq currently locked from an scx callback, or NULL if no rq is