# \[PATCH 1/2\] Ext2: return ENOMEM rather than EIO if sb\_getblk fails

- From: Wang shilong \<hidden\>
- Date: 2011-01-13 10:46:25 
- List: linux-ext4
- Page: https://ratatoskr.run/linux-ext4/2011/01/50641
- Thread: https://ratatoskr.run/linux-fsdevel/2011/01/50641/t.md

```text
From: Wang Shilong <redacted>

As the only reason that sb_getblk fails is that allocation fails.
It will be better to use ENOMEM rather than EIO.

Signed-off-by: Wang Shilong <redacted>
---
 fs/ext2/super.c | 2 +-
 fs/ext2/xattr.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/ext2/super.c b/fs/ext2/super.c
index 663fdcd..3308a55 100644
--- a/fs/ext2/super.c
+++ b/fs/ext2/super.c
@@ -1457,7 +1457,7 @@ static ssize_t ext2_quota_write(struct super_block *sb, int type,
 		else
 			bh = sb_getblk(sb, tmp_bh.b_blocknr);
 		if (unlikely(!bh)) {
-			err = -EIO;
+			err = -ENOMEM;
 			goto out;
 		}
 		lock_buffer(bh);
diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c
index 06209ec..2d7557d 100644
--- a/fs/ext2/xattr.c
+++ b/fs/ext2/xattr.c
@@ -665,7 +665,7 @@ ext2_xattr_set2(struct inode *inode, struct buffer_head *old_bh,
 			if (unlikely(!new_bh)) {
 				ext2_free_blocks(inode, block, 1);
 				mark_inode_dirty(inode);
-				error = -EIO;
+				error = -ENOMEM;
 				goto cleanup;
 			}
 			lock_buffer(new_bh);
-- 
1.7.11.7

```
