1 files changed, 62 insertions, 40 deletions

diff --git a/fs/super.c b/fs/super.c
index 2d762ce67f..076392396e 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -178,7 +178,7 @@ static void super_wake(struct super_block *sb, unsigned int flag)
  * One thing we have to be careful of with a per-sb shrinker is that we don't
  * drop the last active reference to the superblock from within the shrinker.
  * If that happens we could trigger unregistering the shrinker from within the
- * shrinker path and that leads to deadlock on the shrinker_rwsem. Hence we
+ * shrinker path and that leads to deadlock on the shrinker_mutex. Hence we
  * take a passive reference to the superblock to avoid this from occurring.
  */
 static unsigned long super_cache_scan(struct shrinker *shrink,
@@ -191,7 +191,7 @@ static unsigned long super_cache_scan(struct shrinker *shrink,
 	long	dentries;
 	long	inodes;
 
-	sb = container_of(shrink, struct super_block, s_shrink);
+	sb = shrink->private_data;
 
 	/*
 	 * Deadlock avoidance.  We may hold various FS locks, and we don't want
@@ -244,7 +244,7 @@ static unsigned long super_cache_count(struct shrinker *shrink,
 	struct super_block *sb;
 	long	total_objects = 0;
 
-	sb = container_of(shrink, struct super_block, s_shrink);
+	sb = shrink->private_data;
 
 	/*
 	 * We don't call super_trylock_shared() here as it is a scalability
@@ -306,7 +306,7 @@ static void destroy_unused_super(struct super_block *s)
 	security_sb_free(s);
 	put_user_ns(s->s_user_ns);
 	kfree(s->s_subtype);
-	free_prealloced_shrinker(&s->s_shrink);
+	shrinker_free(s->s_shrink);
 	/* no delays needed */
 	destroy_super_work(&s->destroy_work);
 }
@@ -383,16 +383,19 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags,
 	s->s_time_min = TIME64_MIN;
 	s->s_time_max = TIME64_MAX;
 
-	s->s_shrink.seeks = DEFAULT_SEEKS;
-	s->s_shrink.scan_objects = super_cache_scan;
-	s->s_shrink.count_objects = super_cache_count;
-	s->s_shrink.batch = 1024;
-	s->s_shrink.flags = SHRINKER_NUMA_AWARE | SHRINKER_MEMCG_AWARE;
-	if (prealloc_shrinker(&s->s_shrink, "sb-%s", type->name))
+	s->s_shrink = shrinker_alloc(SHRINKER_NUMA_AWARE | SHRINKER_MEMCG_AWARE,
+				     "sb-%s", type->name);
+	if (!s->s_shrink)
 		goto fail;
-	if (list_lru_init_memcg(&s->s_dentry_lru, &s->s_shrink))
+
+	s->s_shrink->scan_objects = super_cache_scan;
+	s->s_shrink->count_objects = super_cache_count;
+	s->s_shrink->batch = 1024;
+	s->s_shrink->private_data = s;
+
+	if (list_lru_init_memcg(&s->s_dentry_lru, s->s_shrink))
 		goto fail;
-	if (list_lru_init_memcg(&s->s_inode_lru, &s->s_shrink))
+	if (list_lru_init_memcg(&s->s_inode_lru, s->s_shrink))
 		goto fail;
 	return s;
 
@@ -477,7 +480,7 @@ void deactivate_locked_super(struct super_block *s)
 {
 	struct file_system_type *fs = s->s_type;
 	if (atomic_dec_and_test(&s->s_active)) {
-		unregister_shrinker(&s->s_shrink);
+		shrinker_free(s->s_shrink);
 		fs->kill_sb(s);
 
 		kill_super_notify(s);
@@ -818,7 +821,7 @@ retry:
 	hlist_add_head(&s->s_instances, &s->s_type->fs_supers);
 	spin_unlock(&sb_lock);
 	get_filesystem(s->s_type);
-	register_shrinker_prepared(&s->s_shrink);
+	shrinker_register(s->s_shrink);
 	return s;
 
 share_extant_sb:
@@ -901,7 +904,7 @@ retry:
 	hlist_add_head(&s->s_instances, &type->fs_supers);
 	spin_unlock(&sb_lock);
 	get_filesystem(type);
-	register_shrinker_prepared(&s->s_shrink);
+	shrinker_register(s->s_shrink);
 	return s;
 }
 EXPORT_SYMBOL(sget);
@@ -1419,32 +1422,48 @@ EXPORT_SYMBOL(sget_dev);
 
 #ifdef CONFIG_BLOCK
 /*
- * Lock a super block that the callers holds a reference to.
+ * Lock the superblock that is holder of the bdev. Returns the superblock
+ * pointer if we successfully locked the superblock and it is alive. Otherwise
+ * we return NULL and just unlock bdev->bd_holder_lock.
  *
- * The caller needs to ensure that the super_block isn't being freed while
- * calling this function, e.g. by holding a lock over the call to this function
- * and the place that clears the pointer to the superblock used by this function
- * before freeing the superblock.
+ * The function must be called with bdev->bd_holder_lock and releases it.
  */
-static bool super_lock_shared_active(struct super_block *sb)
+static struct super_block *bdev_super_lock_shared(struct block_device *bdev)
+	__releases(&bdev->bd_holder_lock)
 {
-	bool born = super_lock_shared(sb);
+	struct super_block *sb = bdev->bd_holder;
+	bool born;
+
+	lockdep_assert_held(&bdev->bd_holder_lock);
+	lockdep_assert_not_held(&sb->s_umount);
+	lockdep_assert_not_held(&bdev->bd_disk->open_mutex);
 
+	/* Make sure sb doesn't go away from under us */
+	spin_lock(&sb_lock);
+	sb->s_count++;
+	spin_unlock(&sb_lock);
+	mutex_unlock(&bdev->bd_holder_lock);
+
+	born = super_lock_shared(sb);
 	if (!born || !sb->s_root || !(sb->s_flags & SB_ACTIVE)) {
 		super_unlock_shared(sb);
-		return false;
+		put_super(sb);
+		return NULL;
 	}
-	return true;
+	/*
+	 * The superblock is active and we hold s_umount, we can drop our
+	 * temporary reference now.
+	 */
+	put_super(sb);
+	return sb;
 }
 
 static void fs_bdev_mark_dead(struct block_device *bdev, bool surprise)
 {
-	struct super_block *sb = bdev->bd_holder;
-
-	/* bd_holder_lock ensures that the sb isn't freed */
-	lockdep_assert_held(&bdev->bd_holder_lock);
+	struct super_block *sb;
 
-	if (!super_lock_shared_active(sb))
+	sb = bdev_super_lock_shared(bdev);
+	if (!sb)
 		return;
 
 	if (!surprise)
@@ -1459,11 +1478,10 @@ static void fs_bdev_mark_dead(struct block_device *bdev, bool surprise)
 
 static void fs_bdev_sync(struct block_device *bdev)
 {
-	struct super_block *sb = bdev->bd_holder;
-
-	lockdep_assert_held(&bdev->bd_holder_lock);
+	struct super_block *sb;
 
-	if (!super_lock_shared_active(sb))
+	sb = bdev_super_lock_shared(bdev);
+	if (!sb)
 		return;
 	sync_filesystem(sb);
 	super_unlock_shared(sb);
@@ -1479,14 +1497,16 @@ int setup_bdev_super(struct super_block *sb, int sb_flags,
 		struct fs_context *fc)
 {
 	blk_mode_t mode = sb_open_mode(sb_flags);
+	struct bdev_handle *bdev_handle;
 	struct block_device *bdev;
 
-	bdev = blkdev_get_by_dev(sb->s_dev, mode, sb, &fs_holder_ops);
-	if (IS_ERR(bdev)) {
+	bdev_handle = bdev_open_by_dev(sb->s_dev, mode, sb, &fs_holder_ops);
+	if (IS_ERR(bdev_handle)) {
 		if (fc)
 			errorf(fc, "%s: Can't open blockdev", fc->source);
-		return PTR_ERR(bdev);
+		return PTR_ERR(bdev_handle);
 	}
+	bdev = bdev_handle->bdev;
 
 	/*
 	 * This really should be in blkdev_get_by_dev, but right now can't due
@@ -1494,7 +1514,7 @@ int setup_bdev_super(struct super_block *sb, int sb_flags,
 	 * writable from userspace even for a read-only block device.
 	 */
 	if ((mode & BLK_OPEN_WRITE) && bdev_read_only(bdev)) {
-		blkdev_put(bdev, sb);
+		bdev_release(bdev_handle);
 		return -EACCES;
 	}
 
@@ -1510,10 +1530,11 @@ int setup_bdev_super(struct super_block *sb, int sb_flags,
 		mutex_unlock(&bdev->bd_fsfreeze_mutex);
 		if (fc)
 			warnf(fc, "%pg: Can't mount, blockdev is frozen", bdev);
-		blkdev_put(bdev, sb);
+		bdev_release(bdev_handle);
 		return -EBUSY;
 	}
 	spin_lock(&sb_lock);
+	sb->s_bdev_handle = bdev_handle;
 	sb->s_bdev = bdev;
 	sb->s_bdi = bdi_get(bdev->bd_disk->bdi);
 	if (bdev_stable_writes(bdev))
@@ -1522,7 +1543,7 @@ int setup_bdev_super(struct super_block *sb, int sb_flags,
 	mutex_unlock(&bdev->bd_fsfreeze_mutex);
 
 	snprintf(sb->s_id, sizeof(sb->s_id), "%pg", bdev);
-	shrinker_debugfs_rename(&sb->s_shrink, "sb-%s:%s", sb->s_type->name,
+	shrinker_debugfs_rename(sb->s_shrink, "sb-%s:%s", sb->s_type->name,
 				sb->s_id);
 	sb_set_blocksize(sb, block_size(bdev));
 	return 0;
@@ -1646,7 +1667,7 @@ void kill_block_super(struct super_block *sb)
 	generic_shutdown_super(sb);
 	if (bdev) {
 		sync_blockdev(bdev);
-		blkdev_put(bdev, sb);
+		bdev_release(sb->s_bdev_handle);
 	}
 }
 
@@ -2139,3 +2160,4 @@ int sb_init_dio_done_wq(struct super_block *sb)
 		destroy_workqueue(wq);
 	return 0;
 }
+EXPORT_SYMBOL_GPL(sb_init_dio_done_wq);