On 2023/2/17 1:31, Jan Kara wrote:
On Tue 14-02-23 10:29:04, Ye Bin wrote:
From: Ye Bin <yebin10@xxxxxxxxxx>
Now, 'es->s_state' maybe covered by recover journal. And journal errno
maybe not recorded in journal sb as IO error. ext4_update_super() only
update error information when 'sbi->s_add_error_count' large than zero.
Then 'EXT4_ERROR_FS' flag maybe lost.
To solve above issue commit error information after recover journal.
Signed-off-by: Ye Bin <yebin10@xxxxxxxxxx>
---
fs/ext4/super.c | 12 ++++++++++++
1 file changed, 12 insertions(+)
Hum, I'm not sure I follow here. If journal replay has overwritten the
superblock (and thus the stored error info), then I'd expect
es->s_error_count got overwritten (possibly to 0) as well. And this is
actually relatively realistic scenario with errors=remount-ro behavior when
the first fs error happens.
What I intended in my original suggestion was to save es->s_error_count,
es->s_state & EXT4_ERROR_FS, es->s_first_error_*, es->s_last_error_* before
doing journal replay in ext4_load_journal() and then after journal replay
merge this info back to the superblock - if EXT4_ERROR_FS was set, set it
now as well, take max of old and new s_error_count, set s_first_error_* if
it is now unset, set s_last_error_* if stored timestamp is newer than
current timestamp.
Or am I overengineering it now? :)
Honza
This is exactly how the code is designed now！
The code has now saved all the above information except EXT4_ERROR_FS by
the following two pieces of logic, as follows：
---------------- In struct ext4_super_block ----------------
1412 #define EXT4_S_ERR_START offsetof(struct ext4_super_block,
s_error_count)
1413 __le32 s_error_count; /* number of fs errors */
1414 __le32 s_first_error_time; /* first time an error
happened */
1415 __le32 s_first_error_ino; /* inode involved in first
error */
1416 __le64 s_first_error_block; /* block involved of first
error */
1417 __u8 s_first_error_func[32] __nonstring; /* function
where the error happened */
1418 __le32 s_first_error_line; /* line number where error
happened */
1419 __le32 s_last_error_time; /* most recent time of an
error */
1420 __le32 s_last_error_ino; /* inode involved in last
error */
1421 __le32 s_last_error_line; /* line number where error
happened */
1422 __le64 s_last_error_block; /* block involved of last
error */
1423 __u8 s_last_error_func[32] __nonstring; /* function
where the error happened */
1424 #define EXT4_S_ERR_END offsetof(struct ext4_super_block, s_mount_opts)
-----------------------------------------------------------
---------------- In ext4_load_journal() ----------------
5929 char *save = kmalloc(EXT4_S_ERR_LEN, GFP_KERNEL);
5930 if (save)
5931 memcpy(save, ((char *) es) +
5932 EXT4_S_ERR_START, EXT4_S_ERR_LEN);
5933 err = jbd2_journal_load(journal);
5934 if (save)
5935 memcpy(((char *) es) + EXT4_S_ERR_START,
5936 save, EXT4_S_ERR_LEN);
5937 kfree(save);
--------------------------------------------------------
As you said, we should also save EXT4_ERROR_FS to es->s_state.
But we are not saving this now, so I think we just need to add
`es->s_state |= cpu_to_le16(EXT4_SB(sb)->s_mount_state & EXT4_ERROR_FS);`
to save the possible EXT4_ERROR_FS flag after copying the error area
data to es.
--
With Best Regards,
Baokun Li
.
