Pull btrfs fix from Chris Mason: "Filipe nailed down a problem where tree log replay would do some work that orphan code wasn't expecting to be done yet, leading to BUG_ON" * 'for-linus-4.5' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: Btrfs: fix loading of orphan roots leading to BUG_ON
Unified · Split
| @@ -310,8 +310,16 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root) · 309 unmodified lines | ||
| 310 | 310 | set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state); |
| 311 | 311 | |
| 312 | 312 | err = btrfs_insert_fs_root(root->fs_info, root); |
| 313 | + /* | |
| 314 | + * The root might have been inserted already, as before we look | |
| 315 | + * for orphan roots, log replay might have happened, which | |
| 316 | + * triggers a transaction commit and qgroup accounting, which | |
| 317 | + * in turn reads and inserts fs roots while doing backref | |
| 318 | + * walking. | |
| 319 | + */ | |
| 320 | + if (err == -EEXIST) | |
| 321 | + err = 0; | |
| 313 | 322 | if (err) { |
| 314 | − BUG_ON(err == -EEXIST); | |
| 315 | 323 | btrfs_free_fs_root(root); |
| 316 | 324 | break; |
| 317 | 325 | } |