fs,block: get holder during claim

Now that we open block devices as files we need to deal with the
realities that closing is a deferred operation. An operation on the
block device such as e.g., freeze, thaw, or removal that runs
concurrently with umount, tries to acquire a stable reference on the
holder. The holder might already be gone though. Make that reliable by
grabbing a passive reference to the holder during bdev_open() and
releasing it during bdev_release().

Fixes: f3a608827d1f ("bdev: open block device as files") # mainline only
Reported-by: Christoph Hellwig <hch@infradead.org>
Link: https://lore.kernel.org/r/ZfEQQ9jZZVes0WCZ@infradead.org
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Christoph Hellwig <hch@infradead.org>
Tested-by: Yi Zhang <yi.zhang@redhat.com>
Reported-by: https://lore.kernel.org/r/CAHj4cs8tbDwKRwfS1=DmooP73ysM__xAb2PQc6XsAmWR+VuYmg@mail.gmail.com
Link: https://lore.kernel.org/r/20240315-freibad-annehmbar-ca68c375af91@brauner
Signed-off-by: Christian Brauner <brauner@kernel.org>

1 files changed, 18 insertions, 0 deletions

diff --git a/fs/super.c b/fs/super.c
index ee05ab6b37e7..71d9779c42b1 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -1515,11 +1515,29 @@ static int fs_bdev_thaw(struct block_device *bdev)
 	return error;
 }
 
+static void fs_bdev_super_get(void *data)
+{
+	struct super_block *sb = data;
+
+	spin_lock(&sb_lock);
+	sb->s_count++;
+	spin_unlock(&sb_lock);
+}
+
+static void fs_bdev_super_put(void *data)
+{
+	struct super_block *sb = data;
+
+	put_super(sb);
+}
+
 const struct blk_holder_ops fs_holder_ops = {
 	.mark_dead		= fs_bdev_mark_dead,
 	.sync			= fs_bdev_sync,
 	.freeze			= fs_bdev_freeze,
 	.thaw			= fs_bdev_thaw,
+	.get_holder		= fs_bdev_super_get,
+	.put_holder		= fs_bdev_super_put,
 };
 EXPORT_SYMBOL_GPL(fs_holder_ops);