1 ext4: don't perform block validity checks on the journal inode
3 Since the journal inode is already checked when we added it to the
4 block validity's system zone, if we check it again, we'll just trigger
7 This was causing failures like this:
9 [ 53.897001] EXT4-fs error (device sda): ext4_find_extent:909: inode
10 #8: comm jbd2/sda-8: pblk 121667583 bad header/extent: invalid extent entries - magic f30a, entries 8, max 340(340), depth 0(0)
11 [ 53.931430] jbd2_journal_bmap: journal block not found at offset 49 on sda-8
12 [ 53.938480] Aborting journal on device sda-8.
14 ... but only if the system was under enough memory pressure that
15 logical->physical mapping for the journal inode gets pushed out of the
16 extent cache. (This is why it wasn't noticed earlier.)
18 Fixes: 345c0dbf3a30 ("ext4: protect journal inode's blocks using block_validity")
19 Reported-by: Dan Rue <dan.rue@linaro.org>
20 Signed-off-by: Theodore Ts'o <tytso@mit.edu>
21 Tested-by: Naresh Kamboju <naresh.kamboju@linaro.org>
23 fs/ext4/extents.c | 12 ++++++++----
24 1 file changed, 8 insertions(+), 4 deletions(-)
26 diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
27 index f2c62e2a0c98..d40ed940001e 100644
28 --- a/fs/ext4/extents.c
29 +++ b/fs/ext4/extents.c
30 @@ -518,10 +518,14 @@ __read_extent_tree_block(const char *function, unsigned int line,
32 if (buffer_verified(bh) && !(flags & EXT4_EX_FORCE_CACHE))
34 - err = __ext4_ext_check(function, line, inode,
35 - ext_block_hdr(bh), depth, pblk);
38 + if (!ext4_has_feature_journal(inode->i_sb) ||
40 + le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum))) {
41 + err = __ext4_ext_check(function, line, inode,
42 + ext_block_hdr(bh), depth, pblk);
46 set_buffer_verified(bh);
48 * If this is a leaf block, cache all of its entries