Control: tag -1 confirmed patch pending On Sun, 2018-07-15 at 18:03 +0200, debian@heinopeters.de wrote: > Package: linux-image-4.9.0-7-amd64 > Version: 4.9.110-1 > > Mounting ext4 filesystem with meta_bg set fails with "structure needs cleaning". > Mounting the same filesystem with the kernel from linux-image-4.9.0-6-amd64 > works as expected. [...] I can reproduce this and found that it is fixed by a commit in Ted Ts'o's ext4 development repository. I'm attaching a backported version which applies to stretch. As a workaround, you can mount a filesystems affected by this bug as read-only and then remount it read-write. In this case the consistency checks do not prevent mounting. (This is probably a bug in itself!) Ben. -- Ben Hutchings Once a job is fouled up, anything done to improve it makes it worse.
From: Theodore Ts'o <tytso@mit.edu>
Date: Sun, 8 Jul 2018 19:35:02 -0400
Subject: ext4: fix false negatives *and* false positives in
ext4_check_descriptors()
Origin: https://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4.git/commit?id=44de022c4382541cebdd6de4465d1f4f465ff1dd
Bug-Debian: https://bugs.debian.org/903838
Ext4_check_descriptors() was getting called before s_gdb_count was
initialized. So for file systems w/o the meta_bg feature, allocation
bitmaps could overlap the block group descriptors and ext4 wouldn't
notice.
For file systems with the meta_bg feature enabled, there was a
fencepost error which would cause the ext4_check_descriptors() to
incorrectly believe that the block allocation bitmap overlaps with the
block group descriptor blocks, and it would reject the mount.
Fix both of these problems.
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
[bwh: Backported to 4.9: adjust context]
---
fs/ext4/super.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -2231,7 +2231,7 @@ static int ext4_check_descriptors(struct
struct ext4_sb_info *sbi = EXT4_SB(sb);
ext4_fsblk_t first_block = le32_to_cpu(sbi->s_es->s_first_data_block);
ext4_fsblk_t last_block;
- ext4_fsblk_t last_bg_block = sb_block + ext4_bg_num_gdb(sb, 0) + 1;
+ ext4_fsblk_t last_bg_block = sb_block + ext4_bg_num_gdb(sb, 0);
ext4_fsblk_t block_bitmap;
ext4_fsblk_t inode_bitmap;
ext4_fsblk_t inode_table;
@@ -3912,13 +3912,13 @@ static int ext4_fill_super(struct super_
goto failed_mount2;
}
}
+ sbi->s_gdb_count = db_count;
if (!ext4_check_descriptors(sb, logical_sb_block, &first_not_zeroed)) {
ext4_msg(sb, KERN_ERR, "group descriptors corrupted!");
ret = -EFSCORRUPTED;
goto failed_mount2;
}
- sbi->s_gdb_count = db_count;
get_random_bytes(&sbi->s_next_generation, sizeof(u32));
spin_lock_init(&sbi->s_next_gen_lock);
Attachment:
signature.asc
Description: This is a digitally signed message part