Merge tag 'locking-core-2022-03-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull locking updates from Ingo Molnar:
 "Changes in this cycle were:

  Bitops & cpumask:
   - Always inline various generic helpers, to improve code generation,
     but also for instrumentation, found by noinstr validation.

   - Add a x86-specific cpumask_clear_cpu() helper to improve code
     generation

  Atomics:
   - Fix atomic64_{read_acquire,set_release} fallbacks

  Lockdep:
   - Fix /proc/lockdep output loop iteration for classes

   - Fix /proc/lockdep potential access to invalid memory

   - Add Mark Rutland as reviewer for atomic primitives

   - Minor cleanups

  Jump labels:
   - Clean up the code a bit

  Misc:
   - Add __sched annotations to percpu rwsem primitives

   - Enable RT_MUTEXES on PREEMPT_RT by default

   - Stray v8086_mode() inlining fix, result of noinstr objtool
     validation"

* tag 'locking-core-2022-03-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  jump_label: Refactor #ifdef of struct static_key
  jump_label: Avoid unneeded casts in STATIC_KEY_INIT_{TRUE,FALSE}
  locking/lockdep: Iterate lock_classes directly when reading lockdep files
  x86/ptrace: Always inline v8086_mode() for instrumentation
  cpumask: Add a x86-specific cpumask_clear_cpu() helper
  locking: Enable RT_MUTEXES by default on PREEMPT_RT.
  locking/local_lock: Make the empty local_lock_*() function a macro.
  atomics: Fix atomic64_{read_acquire,set_release} fallbacks
  locking: Add missing __sched attributes
  cpumask: Always inline helpers which use bit manipulation functions
  asm-generic/bitops: Always inline all bit manipulation helpers
  locking/lockdep: Avoid potential access of invalid memory in lock_class
  lockdep: Use memset_startat() helper in reinit_class()
  MAINTAINERS: add myself as reviewer for atomics

kernel/locking/lockdep.c

@@ -183,11 +183,9 @@ static DECLARE_BITMAP(list_entries_in_use, MAX_LOCKDEP_ENTRIES);
 static struct hlist_head lock_keys_hash[KEYHASH_SIZE];
 unsigned long nr_lock_classes;
 unsigned long nr_zapped_classes;
-#ifndef CONFIG_DEBUG_LOCKDEP
-static
-#endif
+unsigned long max_lock_class_idx;
 struct lock_class lock_classes[MAX_LOCKDEP_KEYS];
-static DECLARE_BITMAP(lock_classes_in_use, MAX_LOCKDEP_KEYS);
+DECLARE_BITMAP(lock_classes_in_use, MAX_LOCKDEP_KEYS);
 
 static inline struct lock_class *hlock_class(struct held_lock *hlock)
 {
@@ -338,7 +336,7 @@ static inline void lock_release_holdtime(struct held_lock *hlock)
  * elements. These elements are linked together by the lock_entry member in
  * struct lock_class.
  */
-LIST_HEAD(all_lock_classes);
+static LIST_HEAD(all_lock_classes);
 static LIST_HEAD(free_lock_classes);
 
 /**
@@ -1252,6 +1250,7 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force)
 	struct lockdep_subclass_key *key;
 	struct hlist_head *hash_head;
 	struct lock_class *class;
+	int idx;
 
 	DEBUG_LOCKS_WARN_ON(!irqs_disabled());
 
@@ -1317,6 +1316,9 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force)
 	 * of classes.
 	 */
 	list_move_tail(&class->lock_entry, &all_lock_classes);
+	idx = class - lock_classes;
+	if (idx > max_lock_class_idx)
+		max_lock_class_idx = idx;
 
 	if (verbose(class)) {
 		graph_unlock();
@@ -6000,6 +6002,8 @@ static void zap_class(struct pending_free *pf, struct lock_class *class)
 		WRITE_ONCE(class->name, NULL);
 		nr_lock_classes--;
 		__clear_bit(class - lock_classes, lock_classes_in_use);
+		if (class - lock_classes == max_lock_class_idx)
+			max_lock_class_idx--;
 	} else {
 		WARN_ONCE(true, "%s() failed for class %s\n", __func__,
 			  class->name);
@@ -6011,13 +6015,10 @@ static void zap_class(struct pending_free *pf, struct lock_class *class)
 
 static void reinit_class(struct lock_class *class)
 {
-	void *const p = class;
-	const unsigned int offset = offsetof(struct lock_class, key);
-
 	WARN_ON_ONCE(!class->lock_entry.next);
 	WARN_ON_ONCE(!list_empty(&class->locks_after));
 	WARN_ON_ONCE(!list_empty(&class->locks_before));
-	memset(p + offset, 0, sizeof(*class) - offset);
+	memset_startat(class, 0, key);
 	WARN_ON_ONCE(!class->lock_entry.next);
 	WARN_ON_ONCE(!list_empty(&class->locks_after));
 	WARN_ON_ONCE(!list_empty(&class->locks_before));
@@ -6290,7 +6291,13 @@ void lockdep_reset_lock(struct lockdep_map *lock)
 		lockdep_reset_lock_reg(lock);
 }
 
-/* Unregister a dynamically allocated key. */
+/*
+ * Unregister a dynamically allocated key.
+ *
+ * Unlike lockdep_register_key(), a search is always done to find a matching
+ * key irrespective of debug_locks to avoid potential invalid access to freed
+ * memory in lock_class entry.
+ */
 void lockdep_unregister_key(struct lock_class_key *key)
 {
 	struct hlist_head *hash_head = keyhashentry(key);
@@ -6305,10 +6312,8 @@ void lockdep_unregister_key(struct lock_class_key *key)
 		return;
 
 	raw_local_irq_save(flags);
-	if (!graph_lock())
-		goto out_irq;
+	lockdep_lock();
 
-	pf = get_pending_free();
 	hlist_for_each_entry_rcu(k, hash_head, hash_entry) {
 		if (k == key) {
 			hlist_del_rcu(&k->hash_entry);
@@ -6316,11 +6321,13 @@ void lockdep_unregister_key(struct lock_class_key *key)
 			break;
 		}
 	}
-	WARN_ON_ONCE(!found);
-	__lockdep_free_key_range(pf, key, 1);
-	call_rcu_zapped(pf);
-	graph_unlock();
-out_irq:
+	WARN_ON_ONCE(!found && debug_locks);
+	if (found) {
+		pf = get_pending_free();
+		__lockdep_free_key_range(pf, key, 1);
+		call_rcu_zapped(pf);
+	}
+	lockdep_unlock();
 	raw_local_irq_restore(flags);
 
 	/* Wait until is_dynamic_key() has finished accessing k->hash_entry. */

kernel/locking/lockdep_internals.h

@@ -121,7 +121,6 @@ static const unsigned long LOCKF_USED_IN_IRQ_READ =
 
 #define MAX_LOCKDEP_CHAIN_HLOCKS (MAX_LOCKDEP_CHAINS*5)
 
-extern struct list_head all_lock_classes;
 extern struct lock_chain lock_chains[];
 
 #define LOCK_USAGE_CHARS (2*XXX_LOCK_USAGE_STATES + 1)
@@ -151,6 +150,10 @@ extern unsigned int nr_large_chain_blocks;
 
 extern unsigned int max_lockdep_depth;
 extern unsigned int max_bfs_queue_depth;
+extern unsigned long max_lock_class_idx;
+
+extern struct lock_class lock_classes[MAX_LOCKDEP_KEYS];
+extern unsigned long lock_classes_in_use[];
 
 #ifdef CONFIG_PROVE_LOCKING
 extern unsigned long lockdep_count_forward_deps(struct lock_class *);
@@ -205,7 +208,6 @@ struct lockdep_stats {
 };
 
 DECLARE_PER_CPU(struct lockdep_stats, lockdep_stats);
-extern struct lock_class lock_classes[MAX_LOCKDEP_KEYS];
 
 #define __debug_atomic_inc(ptr)					\
 	this_cpu_inc(lockdep_stats.ptr);

kernel/locking/lockdep_proc.c

@@ -24,14 +24,33 @@
 
 #include "lockdep_internals.h"
 
+/*
+ * Since iteration of lock_classes is done without holding the lockdep lock,
+ * it is not safe to iterate all_lock_classes list directly as the iteration
+ * may branch off to free_lock_classes or the zapped list. Iteration is done
+ * directly on the lock_classes array by checking the lock_classes_in_use
+ * bitmap and max_lock_class_idx.
+ */
+#define iterate_lock_classes(idx, class)				\
+	for (idx = 0, class = lock_classes; idx <= max_lock_class_idx;	\
+	     idx++, class++)
+
 static void *l_next(struct seq_file *m, void *v, loff_t *pos)
 {
-	return seq_list_next(v, &all_lock_classes, pos);
+	struct lock_class *class = v;
+
+	++class;
+	*pos = class - lock_classes;
+	return (*pos > max_lock_class_idx) ? NULL : class;
 }
 
 static void *l_start(struct seq_file *m, loff_t *pos)
 {
-	return seq_list_start_head(&all_lock_classes, *pos);
+	unsigned long idx = *pos;
+
+	if (idx > max_lock_class_idx)
+		return NULL;
+	return lock_classes + idx;
 }
 
 static void l_stop(struct seq_file *m, void *v)
@@ -57,14 +76,16 @@ static void print_name(struct seq_file *m, struct lock_class *class)
 
 static int l_show(struct seq_file *m, void *v)
 {
-	struct lock_class *class = list_entry(v, struct lock_class, lock_entry);
+	struct lock_class *class = v;
 	struct lock_list *entry;
 	char usage[LOCK_USAGE_CHARS];
+	int idx = class - lock_classes;
 
-	if (v == &all_lock_classes) {
+	if (v == lock_classes)
 		seq_printf(m, "all lock classes:\n");
+
+	if (!test_bit(idx, lock_classes_in_use))
 		return 0;
-	}
 
 	seq_printf(m, "%p", class->key);
 #ifdef CONFIG_DEBUG_LOCKDEP
@@ -220,8 +241,11 @@ static int lockdep_stats_show(struct seq_file *m, void *v)
 
 #ifdef CONFIG_PROVE_LOCKING
 	struct lock_class *class;
+	unsigned long idx;
 
-	list_for_each_entry(class, &all_lock_classes, lock_entry) {
+	iterate_lock_classes(idx, class) {
+		if (!test_bit(idx, lock_classes_in_use))
+			continue;
 
 		if (class->usage_mask == 0)
 			nr_unused++;
@@ -254,6 +278,7 @@ static int lockdep_stats_show(struct seq_file *m, void *v)
 
 		sum_forward_deps += lockdep_count_forward_deps(class);
 	}
+
 #ifdef CONFIG_DEBUG_LOCKDEP
 	DEBUG_LOCKS_WARN_ON(debug_atomic_read(nr_unused_locks) != nr_unused);
 #endif
@@ -345,6 +370,8 @@ static int lockdep_stats_show(struct seq_file *m, void *v)
 	seq_printf(m, " max bfs queue depth:           %11u\n",
 			max_bfs_queue_depth);
 #endif
+	seq_printf(m, " max lock class index:          %11lu\n",
+			max_lock_class_idx);
 	lockdep_stats_debug_show(m);
 	seq_printf(m, " debug_locks:                   %11u\n",
 			debug_locks);
@@ -622,12 +649,16 @@ static int lock_stat_open(struct inode *inode, struct file *file)
 	if (!res) {
 		struct lock_stat_data *iter = data->stats;
 		struct seq_file *m = file->private_data;
+		unsigned long idx;
 
-		list_for_each_entry(class, &all_lock_classes, lock_entry) {
+		iterate_lock_classes(idx, class) {
+			if (!test_bit(idx, lock_classes_in_use))
+				continue;
 			iter->class = class;
 			iter->stats = lock_stats(class);
 			iter++;
 		}
+
 		data->iter_end = iter;
 
 		sort(data->stats, data->iter_end - data->stats,
@@ -645,6 +676,7 @@ static ssize_t lock_stat_write(struct file *file, const char __user *buf,
 			       size_t count, loff_t *ppos)
 {
 	struct lock_class *class;
+	unsigned long idx;
 	char c;
 
 	if (count) {
@@ -654,8 +686,11 @@ static ssize_t lock_stat_write(struct file *file, const char __user *buf,
 		if (c != '0')
 			return count;
 
-		list_for_each_entry(class, &all_lock_classes, lock_entry)
+		iterate_lock_classes(idx, class) {
+			if (!test_bit(idx, lock_classes_in_use))
+				continue;
 			clear_lock_stats(class);
+		}
 	}
 	return count;
 }

kernel/locking/percpu-rwsem.c

@@ -7,6 +7,7 @@
 #include <linux/rcupdate.h>
 #include <linux/sched.h>
 #include <linux/sched/task.h>
+#include <linux/sched/debug.h>
 #include <linux/errno.h>
 
 int __percpu_init_rwsem(struct percpu_rw_semaphore *sem,
@@ -162,7 +163,7 @@ static void percpu_rwsem_wait(struct percpu_rw_semaphore *sem, bool reader)
 	__set_current_state(TASK_RUNNING);
 }
 
-bool __percpu_down_read(struct percpu_rw_semaphore *sem, bool try)
+bool __sched __percpu_down_read(struct percpu_rw_semaphore *sem, bool try)
 {
 	if (__percpu_down_read_trylock(sem))
 		return true;
@@ -211,7 +212,7 @@ static bool readers_active_check(struct percpu_rw_semaphore *sem)
 	return true;
 }
 
-void percpu_down_write(struct percpu_rw_semaphore *sem)
+void __sched percpu_down_write(struct percpu_rw_semaphore *sem)
 {
 	might_sleep();
 	rwsem_acquire(&sem->dep_map, 0, 0, _RET_IP_);

kernel/locking/rwsem.c

@@ -1048,7 +1048,7 @@ out_nolock:
 /*
  * Wait until we successfully acquire the write lock
  */
-static struct rw_semaphore *
+static struct rw_semaphore __sched *
 rwsem_down_write_slowpath(struct rw_semaphore *sem, int state)
 {
 	long count;