Btrfs: check duplicate backrefs for both data and metadata

lookup_inline_extent_backref only checks for duplicate backref for data
extents. It assumes backrefs for tree block never conflict.

This patch makes lookup_inline_extent_backref check for duplicate backrefs
for both data and tree block, so that we can detect potential bug earlier.
This is a safety check, strictly speaking it is not required.

Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 33a65f2..edc7d20 100644 (file)
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -1056,8 +1056,7 @@ int lookup_inline_extent_backref(struct btrfs_trans_handle *trans,
        want = extent_ref_type(parent, owner);
        if (insert) {
                extra_size = btrfs_extent_inline_ref_size(want);
-               if (owner >= BTRFS_FIRST_FREE_OBJECTID)
-                       path->keep_locks = 1;
+               path->keep_locks = 1;
        } else
                extra_size = -1;
        ret = btrfs_search_slot(trans, root, &key, path, extra_size, 1);
@@ -1087,12 +1086,6 @@ int lookup_inline_extent_backref(struct btrfs_trans_handle *trans,
 #endif
        BUG_ON(item_size < sizeof(*ei));
 
-       if (owner < BTRFS_FIRST_FREE_OBJECTID && insert &&
-           item_size + extra_size >= BTRFS_MAX_EXTENT_ITEM_SIZE(root)) {
-               err = -EAGAIN;
-               goto out;
-       }
-
        ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
        flags = btrfs_extent_flags(leaf, ei);
 
@@ -1165,15 +1158,15 @@ int lookup_inline_extent_backref(struct btrfs_trans_handle *trans,
                 * For simplicity, we just do not add new inline back
                 * ref if there is any kind of item for this block
                 */
-               if (owner >= BTRFS_FIRST_FREE_OBJECTID &&
-                   find_next_key(path, &key) == 0 && key.objectid == bytenr) {
+               if (find_next_key(path, &key) == 0 && key.objectid == bytenr &&
+                   key.type < BTRFS_BLOCK_GROUP_ITEM_KEY) {
                        err = -EAGAIN;
                        goto out;
                }
        }
        *ref_ret = (struct btrfs_extent_inline_ref *)ptr;
 out:
-       if (insert && owner >= BTRFS_FIRST_FREE_OBJECTID) {
+       if (insert) {
                path->keep_locks = 0;
                btrfs_unlock_up_safe(path, 1);
        }