When we fail to write out inode when adding new xattr block, we backed out some changes (like quota modifications) but left others (like block allocation or inode xattr block modification). Just leave the inode intact in this case because we can hardly reach a consistent state and defer fixing to e2fsck. Also remove outdated comment and special data handling since sync_inode_metadata() does not write out any data. Signed-off-by: Jan Kara --- fs/ext2/xattr.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) The inconsistency was notice by unrelated Sashiko review. I plan to merge this through my tree. diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c index e55d16abf422..5f49ec4afc36 100644 --- a/fs/ext2/xattr.c +++ b/fs/ext2/xattr.c @@ -779,16 +779,12 @@ ext2_xattr_set2(struct inode *inode, struct buffer_head *old_bh, inode_set_ctime_current(inode); if (IS_SYNC(inode)) { error = sync_inode_metadata(inode, 1); - /* In case sync failed due to ENOSPC the inode was actually - * written (only some dirty data were not) so we just proceed - * as if nothing happened and cleanup the unused block */ - if (error && error != -ENOSPC) { - if (new_bh && new_bh != old_bh) { - dquot_free_block_nodirty(inode, 1); - mark_inode_dirty(inode); - } + /* + * Inode writeout failed. Backing everything out is complex so + * let's just leave it for e2fsck to cleanup the mess. + */ + if (error) goto cleanup; - } } else mark_inode_dirty(inode); -- 2.51.0