ext4: fix corruption when online resizing a 1K bigalloc fs
authorBaokun Li <libaokun1@huawei.com>
Thu, 17 Nov 2022 04:03:41 +0000 (12:03 +0800)
committerTheodore Ts'o <tytso@mit.edu>
Fri, 9 Dec 2022 02:49:24 +0000 (21:49 -0500)
commit0aeaa2559d6d53358fca3e3fce73807367adca74
tree65343ef5e10e63effc81fa432841fa0bc351d114
parent8f49ec603ae3e213bfab2799182724e3abac55a1
ext4: fix corruption when online resizing a 1K bigalloc fs

When a backup superblock is updated in update_backups(), the primary
superblock's offset in the group (that is, sbi->s_sbh->b_blocknr) is used
as the backup superblock's offset in its group. However, when the block
size is 1K and bigalloc is enabled, the two offsets are not equal. This
causes the backup group descriptors to be overwritten by the superblock
in update_backups(). Moreover, if meta_bg is enabled, the file system will
be corrupted because this feature uses backup group descriptors.

To solve this issue, we use a more accurate ext4_group_first_block_no() as
the offset of the backup superblock in its group.

Fixes: d77147ff443b ("ext4: add support for online resizing with bigalloc")
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20221117040341.1380702-4-libaokun1@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
fs/ext4/resize.c