workqueue: Changes for v6.12

Nothing major:
 
 - workqueue.panic_on_stall boot param added.
 
 - alloc_workqueue_lockdep_map() added (used by DRM).
 
 - Other cleanusp and doc updates.
 -----BEGIN PGP SIGNATURE-----
 
 iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCZuN3gQ4cdGpAa2VybmVs
 Lm9yZwAKCRCxYfJx3gVYGR1hAP0XObdExeNsVWe1JUUUX061+H+aA6aVffb9+J/t
 b32u3QEAsn+oNWzuvzlGlSQKQMpPk+dT2na0Q0yZNxkNEzUiEQQ=
 =TeDS
 -----END PGP SIGNATURE-----

Merge tag 'wq-for-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq

Pull workqueue updates from Tejun Heo:
 "Nothing major:

   - workqueue.panic_on_stall boot param added

   - alloc_workqueue_lockdep_map() added (used by DRM)

   - Other cleanusp and doc updates"

* tag 'wq-for-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
  kernel/workqueue.c: fix DEFINE_PER_CPU_SHARED_ALIGNED expansion
  workqueue: Fix another htmldocs build warning
  workqueue: fix null-ptr-deref on __alloc_workqueue() error
  workqueue: Don't call va_start / va_end twice
  workqueue: Fix htmldocs build warning
  workqueue: Add interface for user-defined workqueue lockdep map
  workqueue: Change workqueue lockdep map to pointer
  workqueue: Split alloc_workqueue into internal function and lockdep init
  Documentation: kernel-parameters: add workqueue.panic_on_stall
  workqueue: add cmdline parameter workqueue.panic_on_stall

Documentation/admin-guide/kernel-parameters.txt

@@ -7386,6 +7386,13 @@
 			it can be updated at runtime by writing to the
 			corresponding sysfs file.
 
+	workqueue.panic_on_stall=<uint>
+			Panic when workqueue stall is detected by
+			CONFIG_WQ_WATCHDOG. It sets the number times of the
+			stall to trigger panic.
+
+			The default is 0, which disables the panic on stall.
+
 	workqueue.cpu_intensive_thresh_us=
 			Per-cpu work items which run for longer than this
 			threshold are automatically considered CPU intensive

include/linux/workqueue.h

@@ -507,6 +507,47 @@ void workqueue_softirq_dead(unsigned int cpu);
 __printf(1, 4) struct workqueue_struct *
 alloc_workqueue(const char *fmt, unsigned int flags, int max_active, ...);
 
+#ifdef CONFIG_LOCKDEP
+/**
+ * alloc_workqueue_lockdep_map - allocate a workqueue with user-defined lockdep_map
+ * @fmt: printf format for the name of the workqueue
+ * @flags: WQ_* flags
+ * @max_active: max in-flight work items, 0 for default
+ * @lockdep_map: user-defined lockdep_map
+ * @...: args for @fmt
+ *
+ * Same as alloc_workqueue but with the a user-define lockdep_map. Useful for
+ * workqueues created with the same purpose and to avoid leaking a lockdep_map
+ * on each workqueue creation.
+ *
+ * RETURNS:
+ * Pointer to the allocated workqueue on success, %NULL on failure.
+ */
+__printf(1, 5) struct workqueue_struct *
+alloc_workqueue_lockdep_map(const char *fmt, unsigned int flags, int max_active,
+			    struct lockdep_map *lockdep_map, ...);
+
+/**
+ * alloc_ordered_workqueue_lockdep_map - allocate an ordered workqueue with
+ * user-defined lockdep_map
+ *
+ * @fmt: printf format for the name of the workqueue
+ * @flags: WQ_* flags (only WQ_FREEZABLE and WQ_MEM_RECLAIM are meaningful)
+ * @lockdep_map: user-defined lockdep_map
+ * @args: args for @fmt
+ *
+ * Same as alloc_ordered_workqueue but with the a user-define lockdep_map.
+ * Useful for workqueues created with the same purpose and to avoid leaking a
+ * lockdep_map on each workqueue creation.
+ *
+ * RETURNS:
+ * Pointer to the allocated workqueue on success, %NULL on failure.
+ */
+#define alloc_ordered_workqueue_lockdep_map(fmt, flags, lockdep_map, args...)	\
+	alloc_workqueue_lockdep_map(fmt, WQ_UNBOUND | __WQ_ORDERED | (flags),	\
+				    1, lockdep_map, ##args)
+#endif
+
 /**
  * alloc_ordered_workqueue - allocate an ordered workqueue
  * @fmt: printf format for the name of the workqueue

kernel/workqueue.c

@@ -364,7 +364,8 @@ struct workqueue_struct {
 #ifdef CONFIG_LOCKDEP
 	char			*lock_name;
 	struct lock_class_key	key;
-	struct lockdep_map	lockdep_map;
+	struct lockdep_map	__lockdep_map;
+	struct lockdep_map	*lockdep_map;
 #endif
 	char			name[WQ_NAME_LEN]; /* I: workqueue name */
 
@@ -476,16 +477,13 @@ static bool wq_debug_force_rr_cpu = false;
 module_param_named(debug_force_rr_cpu, wq_debug_force_rr_cpu, bool, 0644);
 
 /* to raise softirq for the BH worker pools on other CPUs */
-static DEFINE_PER_CPU_SHARED_ALIGNED(struct irq_work [NR_STD_WORKER_POOLS],
-				     bh_pool_irq_works);
+static DEFINE_PER_CPU_SHARED_ALIGNED(struct irq_work [NR_STD_WORKER_POOLS], bh_pool_irq_works);
 
 /* the BH worker pools */
-static DEFINE_PER_CPU_SHARED_ALIGNED(struct worker_pool [NR_STD_WORKER_POOLS],
-				     bh_worker_pools);
+static DEFINE_PER_CPU_SHARED_ALIGNED(struct worker_pool [NR_STD_WORKER_POOLS], bh_worker_pools);
 
 /* the per-cpu worker pools */
-static DEFINE_PER_CPU_SHARED_ALIGNED(struct worker_pool [NR_STD_WORKER_POOLS],
-				     cpu_worker_pools);
+static DEFINE_PER_CPU_SHARED_ALIGNED(struct worker_pool [NR_STD_WORKER_POOLS], cpu_worker_pools);
 
 static DEFINE_IDR(worker_pool_idr);	/* PR: idr of all pools */
 
@@ -3203,7 +3201,7 @@ __acquires(&pool->lock)
 	lockdep_start_depth = lockdep_depth(current);
 	/* see drain_dead_softirq_workfn() */
 	if (!bh_draining)
-		lock_map_acquire(&pwq->wq->lockdep_map);
+		lock_map_acquire(pwq->wq->lockdep_map);
 	lock_map_acquire(&lockdep_map);
 	/*
 	 * Strictly speaking we should mark the invariant state without holding
@@ -3237,7 +3235,7 @@ __acquires(&pool->lock)
 	pwq->stats[PWQ_STAT_COMPLETED]++;
 	lock_map_release(&lockdep_map);
 	if (!bh_draining)
-		lock_map_release(&pwq->wq->lockdep_map);
+		lock_map_release(pwq->wq->lockdep_map);
 
 	if (unlikely((worker->task && in_atomic()) ||
 		     lockdep_depth(current) != lockdep_start_depth ||
@@ -3873,11 +3871,14 @@ static bool flush_workqueue_prep_pwqs(struct workqueue_struct *wq,
 static void touch_wq_lockdep_map(struct workqueue_struct *wq)
 {
 #ifdef CONFIG_LOCKDEP
+	if (unlikely(!wq->lockdep_map))
+		return;
+
 	if (wq->flags & WQ_BH)
 		local_bh_disable();
 
-	lock_map_acquire(&wq->lockdep_map);
-	lock_map_release(&wq->lockdep_map);
+	lock_map_acquire(wq->lockdep_map);
+	lock_map_release(wq->lockdep_map);
 
 	if (wq->flags & WQ_BH)
 		local_bh_enable();
@@ -3911,7 +3912,7 @@ void __flush_workqueue(struct workqueue_struct *wq)
 	struct wq_flusher this_flusher = {
 		.list = LIST_HEAD_INIT(this_flusher.list),
 		.flush_color = -1,
-		.done = COMPLETION_INITIALIZER_ONSTACK_MAP(this_flusher.done, wq->lockdep_map),
+		.done = COMPLETION_INITIALIZER_ONSTACK_MAP(this_flusher.done, (*wq->lockdep_map)),
 	};
 	int next_color;
 
@@ -4776,16 +4777,23 @@ static void wq_init_lockdep(struct workqueue_struct *wq)
 		lock_name = wq->name;
 
 	wq->lock_name = lock_name;
-	lockdep_init_map(&wq->lockdep_map, lock_name, &wq->key, 0);
+	wq->lockdep_map = &wq->__lockdep_map;
+	lockdep_init_map(wq->lockdep_map, lock_name, &wq->key, 0);
 }
 
 static void wq_unregister_lockdep(struct workqueue_struct *wq)
 {
+	if (wq->lockdep_map != &wq->__lockdep_map)
+		return;
+
 	lockdep_unregister_key(&wq->key);
 }
 
 static void wq_free_lockdep(struct workqueue_struct *wq)
 {
+	if (wq->lockdep_map != &wq->__lockdep_map)
+		return;
+
 	if (wq->lock_name != wq->name)
 		kfree(wq->lock_name);
 }
@@ -5619,12 +5627,10 @@ static void wq_adjust_max_active(struct workqueue_struct *wq)
 	} while (activated);
 }
 
-__printf(1, 4)
-struct workqueue_struct *alloc_workqueue(const char *fmt,
-					 unsigned int flags,
-					 int max_active, ...)
+static struct workqueue_struct *__alloc_workqueue(const char *fmt,
+						  unsigned int flags,
+						  int max_active, va_list args)
 {
-	va_list args;
 	struct workqueue_struct *wq;
 	size_t wq_size;
 	int name_len;
@@ -5656,9 +5662,7 @@ struct workqueue_struct *alloc_workqueue(const char *fmt,
 			goto err_free_wq;
 	}
 
-	va_start(args, max_active);
 	name_len = vsnprintf(wq->name, sizeof(wq->name), fmt, args);
-	va_end(args);
 
 	if (name_len >= WQ_NAME_LEN)
 		pr_warn_once("workqueue: name exceeds WQ_NAME_LEN. Truncating to: %s\n",
@@ -5688,12 +5692,11 @@ struct workqueue_struct *alloc_workqueue(const char *fmt,
 	INIT_LIST_HEAD(&wq->flusher_overflow);
 	INIT_LIST_HEAD(&wq->maydays);
 
-	wq_init_lockdep(wq);
 	INIT_LIST_HEAD(&wq->list);
 
 	if (flags & WQ_UNBOUND) {
 		if (alloc_node_nr_active(wq->node_nr_active) < 0)
-			goto err_unreg_lockdep;
+			goto err_free_wq;
 	}
 
 	/*
@@ -5732,9 +5735,6 @@ err_unlock_free_node_nr_active:
 		kthread_flush_worker(pwq_release_worker);
 		free_node_nr_active(wq->node_nr_active);
 	}
-err_unreg_lockdep:
-	wq_unregister_lockdep(wq);
-	wq_free_lockdep(wq);
 err_free_wq:
 	free_workqueue_attrs(wq->unbound_attrs);
 	kfree(wq);
@@ -5745,8 +5745,49 @@ err_destroy:
 	destroy_workqueue(wq);
 	return NULL;
 }
+
+__printf(1, 4)
+struct workqueue_struct *alloc_workqueue(const char *fmt,
+					 unsigned int flags,
+					 int max_active, ...)
+{
+	struct workqueue_struct *wq;
+	va_list args;
+
+	va_start(args, max_active);
+	wq = __alloc_workqueue(fmt, flags, max_active, args);
+	va_end(args);
+	if (!wq)
+		return NULL;
+
+	wq_init_lockdep(wq);
+
+	return wq;
+}
 EXPORT_SYMBOL_GPL(alloc_workqueue);
 
+#ifdef CONFIG_LOCKDEP
+__printf(1, 5)
+struct workqueue_struct *
+alloc_workqueue_lockdep_map(const char *fmt, unsigned int flags,
+			    int max_active, struct lockdep_map *lockdep_map, ...)
+{
+	struct workqueue_struct *wq;
+	va_list args;
+
+	va_start(args, lockdep_map);
+	wq = __alloc_workqueue(fmt, flags, max_active, args);
+	va_end(args);
+	if (!wq)
+		return NULL;
+
+	wq->lockdep_map = lockdep_map;
+
+	return wq;
+}
+EXPORT_SYMBOL_GPL(alloc_workqueue_lockdep_map);
+#endif
+
 static bool pwq_busy(struct pool_workqueue *pwq)
 {
 	int i;
@@ -7406,6 +7447,9 @@ static struct timer_list wq_watchdog_timer;
 static unsigned long wq_watchdog_touched = INITIAL_JIFFIES;
 static DEFINE_PER_CPU(unsigned long, wq_watchdog_touched_cpu) = INITIAL_JIFFIES;
 
+static unsigned int wq_panic_on_stall;
+module_param_named(panic_on_stall, wq_panic_on_stall, uint, 0644);
+
 /*
  * Show workers that might prevent the processing of pending work items.
  * The only candidates are CPU-bound workers in the running state.
@@ -7457,6 +7501,16 @@ static void show_cpu_pools_hogs(void)
 	rcu_read_unlock();
 }
 
+static void panic_on_wq_watchdog(void)
+{
+	static unsigned int wq_stall;
+
+	if (wq_panic_on_stall) {
+		wq_stall++;
+		BUG_ON(wq_stall >= wq_panic_on_stall);
+	}
+}
+
 static void wq_watchdog_reset_touched(void)
 {
 	int cpu;
@@ -7529,6 +7583,9 @@ static void wq_watchdog_timer_fn(struct timer_list *unused)
 	if (cpu_pool_stall)
 		show_cpu_pools_hogs();
 
+	if (lockup_detected)
+		panic_on_wq_watchdog();
+
 	wq_watchdog_reset_touched();
 	mod_timer(&wq_watchdog_timer, jiffies + thresh);
 }