[V3] Btrfs: kill some BUG_ONs() in the find_parent_nodes()

Commit Message

Wang Shilong March 29, 2013, 11:03 p.m. UTC

From: Wang Shilong <[e-mail]>

The reason that BUG_ON() happens in these places is just
because of ENOMEM.

We try ro return ENOMEM rather than trigger BUG_ON(), the
caller will abort the transaction thus avoiding the kernel panic.

Signed-off-by: Wang Shilong <[e-mail]>
Reviewed-by: Miao Xie <[e-mail]>
Reviewed-by: Jan Schmidt <[e-mail]>
---
Changelog:
	V2->V3: correct wrong check about ulist_add()
	and revert the check about ret=0
---
 fs/btrfs/backref.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

Patch

diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
index bd605c8..3dd0693 100644
--- a/fs/btrfs/backref.c
+++ b/fs/btrfs/backref.c
@@ -900,7 +900,8 @@  again:
 		if (ref->count && ref->root_id && ref->parent == 0) {
 			/* no parent == root of tree */
 			ret = ulist_add(roots, ref->root_id, 0, GFP_NOFS);
-			BUG_ON(ret < 0);
+			if (ret < 0)
+				goto out;
 		}
 		if (ref->count && ref->parent) {
 			struct extent_inode_elem *eie = NULL;
@@ -920,6 +921,8 @@  again:
 			ret = ulist_add_merge(refs, ref->parent,
 					      (uintptr_t)ref->inode_list,
 					      (u64 *)&eie, GFP_NOFS);
+			if (ret < 0)
+				goto out;
 			if (!ret && extent_item_pos) {
 				/*
 				 * we've recorded that parent, so we must extend
@@ -930,7 +933,6 @@  again:
 					eie = eie->next;
 				eie->next = ref->inode_list;
 			}
-			BUG_ON(ret < 0);
 		}
 		kfree(ref);
 	}