Btrfs: fix ->iterate_shared() by upgrading i_rwsem for delayed nodes

Commit fe742fd4f90f ("Revert "btrfs: switch to ->iterate_shared()"")
backed out the conversion to ->iterate_shared() for Btrfs because the
delayed inode handling in btrfs_real_readdir() is racy. However, we can
still do readdir in parallel if there are no delayed nodes.

This is a temporary fix which upgrades the shared inode lock to an
exclusive lock only when we have delayed items until we come up with a
more complete solution. While we're here, rename the
btrfs_{get,put}_delayed_items functions to make it very clear that
they're just for readdir.

Tested with xfstests and by doing a parallel kernel build:

	while make tinyconfig && make -j4 && git clean dqfx; do
		:
	done

along with a bunch of parallel finds in another shell:

	while true; do
		for ((i=0; i<4; i++)); do
			find . >/dev/null &
		done
		wait
	done

Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>

1 files changed, 22 insertions, 5 deletions

diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index 61561c2a3f96..d3aaabbfada0 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -1606,15 +1606,23 @@ int btrfs_inode_delayed_dir_index_count(struct inode *inode)
 	return 0;
 }
 
-void btrfs_get_delayed_items(struct inode *inode, struct list_head *ins_list,
-			     struct list_head *del_list)
+bool btrfs_readdir_get_delayed_items(struct inode *inode,
+				     struct list_head *ins_list,
+				     struct list_head *del_list)
 {
 	struct btrfs_delayed_node *delayed_node;
 	struct btrfs_delayed_item *item;
 
 	delayed_node = btrfs_get_delayed_node(inode);
 	if (!delayed_node)
-		return;
+		return false;
+
+	/*
+	 * We can only do one readdir with delayed items at a time because of
+	 * item->readdir_list.
+	 */
+	inode_unlock_shared(inode);
+	inode_lock(inode);
 
 	mutex_lock(&delayed_node->mutex);
 	item = __btrfs_first_delayed_insertion_item(delayed_node);
@@ -1641,10 +1649,13 @@ void btrfs_get_delayed_items(struct inode *inode, struct list_head *ins_list,
 	 * requeue or dequeue this delayed node.
 	 */
 	atomic_dec(&delayed_node->refs);
+
+	return true;
 }
 
-void btrfs_put_delayed_items(struct list_head *ins_list,
-			     struct list_head *del_list)
+void btrfs_readdir_put_delayed_items(struct inode *inode,
+				     struct list_head *ins_list,
+				     struct list_head *del_list)
 {
 	struct btrfs_delayed_item *curr, *next;
 
@@ -1659,6 +1670,12 @@ void btrfs_put_delayed_items(struct list_head *ins_list,
 		if (atomic_dec_and_test(&curr->refs))
 			kfree(curr);
 	}
+
+	/*
+	 * The VFS is going to do up_read(), so we need to downgrade back to a
+	 * read lock.
+	 */
+	downgrade_write(&inode->i_rwsem);
 }
 
 int btrfs_should_delete_dir_index(struct list_head *del_list,