ext4: mark block group as corrupt on block bitmap error

When we notice a block-bitmap corruption (because of device failure or
something else), we should mark this group as corrupt and prevent
further block allocations/deallocations from it. Currently, we end up
generating one error message for every block in the bitmap. This
potentially could make the system unstable as noticed in some
bugs. With this patch, the error will be printed only the first time
and mark the entire block group as corrupted. This prevents future
access allocations/deallocations from it.

Also tested by corrupting the block
bitmap and forcefully introducing the mb_free_blocks error:
(1) create a largefile (2Gb)
$ dd if=/dev/zero of=largefile oflag=direct bs=10485760 count=200
(2) umount filesystem. use dumpe2fs to see which block-bitmaps
are in use by largefile and note their block numbers
(3) use dd to zero-out the used block bitmaps
$ dd if=/dev/zero of=/dev/hdc4 bs=4096 seek=14 count=8 oflag=direct
(4) mount the FS and delete the largefile.
(5) recreate the largefile. verify that the new largefile does not
get any blocks from the groups marked as bad.
Without the patch, we will see mb_free_blocks error for each bit in
each zero'ed out bitmap at (4). With the patch, we only see the error
once per blockgroup:
[  309.706803] EXT4-fs error (device sdb4): ext4_mb_generate_buddy:735: group 15: 32768 clusters in bitmap, 0 in gd. blk grp corrupted.
[  309.720824] EXT4-fs error (device sdb4): ext4_mb_generate_buddy:735: group 14: 32768 clusters in bitmap, 0 in gd. blk grp corrupted.
[  309.732858] EXT4-fs error (device sdb4) in ext4_free_blocks:4802: IO failure
[  309.748321] EXT4-fs error (device sdb4): ext4_mb_generate_buddy:735: group 13: 32768 clusters in bitmap, 0 in gd. blk grp corrupted.
[  309.760331] EXT4-fs error (device sdb4) in ext4_free_blocks:4802: IO failure
[  309.769695] EXT4-fs error (device sdb4): ext4_mb_generate_buddy:735: group 12: 32768 clusters in bitmap, 0 in gd. blk grp corrupted.
[  309.781721] EXT4-fs error (device sdb4) in ext4_free_blocks:4802: IO failure
[  309.798166] EXT4-fs error (device sdb4): ext4_mb_generate_buddy:735: group 11: 32768 clusters in bitmap, 0 in gd. blk grp corrupted.
[  309.810184] EXT4-fs error (device sdb4) in ext4_free_blocks:4802: IO failure
[  309.819532] EXT4-fs error (device sdb4): ext4_mb_generate_buddy:735: group 10: 32768 clusters in bitmap, 0 in gd. blk grp corrupted.

Google-Bug-Id: 7258357

[darrick.wong@oracle.com]
Further modifications (by Darrick) to make more obvious that this corruption
bit applies to blocks only.  Set the corruption flag if the block group bitmap
verification fails.

Original-author: Aditya Kali <adityakali@google.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

1 files changed, 3 insertions, 0 deletions

diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index a40b1f9..d9b66c4 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -356,6 +356,7 @@ void ext4_validate_block_bitmap(struct super_block *sb,
 			       struct buffer_head *bh)
 {
 	ext4_fsblk_t	blk;
+	struct ext4_group_info *grp = ext4_get_group_info(sb, block_group);
 
 	if (buffer_verified(bh))
 		return;
@@ -366,12 +367,14 @@ void ext4_validate_block_bitmap(struct super_block *sb,
 		ext4_unlock_group(sb, block_group);
 		ext4_error(sb, "bg %u: block %llu: invalid block bitmap",
 			   block_group, blk);
+		set_bit(EXT4_GROUP_INFO_BBITMAP_CORRUPT_BIT, &grp->bb_state);
 		return;
 	}
 	if (unlikely(!ext4_block_bitmap_csum_verify(sb, block_group,
 			desc, bh))) {
 		ext4_unlock_group(sb, block_group);
 		ext4_error(sb, "bg %u: bad block bitmap checksum", block_group);
+		set_bit(EXT4_GROUP_INFO_BBITMAP_CORRUPT_BIT, &grp->bb_state);
 		return;
 	}
 	set_buffer_verified(bh);