btrfs: unify the error handling of btrfs_read_buffer()

There is one oddball error handling of btrfs_read_buffer():

	ret = btrfs_read_buffer(tmp, gen, parent_level - 1, &first_key);
	if (!ret) {
		*eb_ret = tmp;
		return 0;
	}
	free_extent_buffer(tmp);
	btrfs_release_path(p);
	return -EIO;

While all other call sites check the error first.  Unify the behavior.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>

fs/btrfs/ctree.c

@@ -1438,14 +1438,14 @@ read_block_for_search(struct btrfs_root *root, struct btrfs_path *p,
 
 		/* now we're allowed to do a blocking uptodate check */
 		ret = btrfs_read_buffer(tmp, gen, parent_level - 1, &first_key);
-		if (!ret) {
-			*eb_ret = tmp;
-			return 0;
-		}
+		if (ret) {
 			free_extent_buffer(tmp);
 			btrfs_release_path(p);
 			return -EIO;
 		}
+		*eb_ret = tmp;
+		return 0;
+	}
 
 	/*
 	 * reduce lock contention at high levels