linux ext filesystem errors

DISCUSSION Markus Moeller 2 participants 3 messages Oct 20, 2013

lore

Share & tools

lore

Markus Moeller

author

2013-10-20 12:55 UTC #0 lore raw

Hi,

I have an SD disk which was running fine with an ext4 filesystem. Lately
I got the following errors:

[  133.736890] EXT4-fs error (device mmcblk0p2): ext4_ext_check_inode:462: inode #131171: comm tput: bad header/extent: invalid eh_entries - magic f30a, entries 65, max 4(4), depth 0(0)
[  133.775135] EXT4-fs error (device mmcblk0p2): ext4_ext_check_inode:462: inode #131171: comm tput: bad header/extent: invalid eh_entries - magic f30a, entries 65, max 4(4), depth 0(0)
[  133.874077] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm start-stop-daem: bad extra_isize (32924 != 256)
[  134.298519] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm startpar: bad extra_isize (32924 != 256)
[  134.317035] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm startpar: bad extra_isize (32924 != 256)
[  134.815006] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm init: bad extra_isize (32924 != 256)
[  134.845380] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm init: bad extra_isize (32924 != 256)
[  134.866181] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm init: bad extra_isize (32924 != 256)
[  134.894042] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm sulogin: bad extra_isize (32924 != 256)
[  134.912231] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46: comm sulogin: bad extra_isize (32924 != 256)

which I tried to fix with fsck but it gets killed.

linux:~ # e2fsck -y /dev/sdb2
e2fsck 1.42.6 (21-Sep-2012)
Backing up journal inode block information.

/dev/sdb2 contains a file system with errors, check forced. Resize inode not valid. Recreate? yes

Pass 1: Checking inodes, blocks, and sizes
Inode 131179 has an invalid extent

       (logical block 0, invalid physical block 1342766638, len 1)

Clear? yes

Inode 131179, i_blocks is 8, should be 0. Fix? yes

Inodes that were part of a corrupted orphan linked list found. Fix? yes

Inode 131180 was part of the orphaned inode list. FIXED. Inode 131180 has a extra size (2076) which is invalid Fix? yes

Error while reading over extent tree in inode 131180: Corrupt extent header Clear inode? yes

Inode 131180, i_blocks is 8, should be 0. Fix? yes

Inode 131185 was part of the orphaned inode list. FIXED. Inode 131185, i_size is 144185556820644144, should be 614400. Fix? yes

Inode 131186 was part of the orphaned inode list. FIXED. Inode 131186 has a extra size (540) which is invalid Fix? yes

Inode 131186 has an invalid extent

       (logical block 0, invalid physical block 155898722910210, len 2)

Clear? yes

Inode 131186, i_blocks is 16, should be 0. Fix? yes

Error while reading over extent tree in inode 131187: Corrupt extent header Clear inode? yes

Inode 131187, i_blocks is 9, should be 0. Fix? yes

Inode 131188 has an invalid extent

       (logical block 0, invalid physical block 4400194552889, len 1)

Clear? yes

Inode 131188, i_blocks is 24, should be 0. Fix? yes

Inode 131193, i_size is 18014398509484072, should be 2199023259648. Fix? yes

Killed

Does this mean it is a HW error ?

Thank you
Markus

Markus Moeller

author

2013-10-20 14:47 UTC #1 lore raw

I did created a new filesystem and it looks fine, which I think excludes
a disk HW problems. So how can I fix the ext4 filesystem ?

Thank you
Markus

"Markus Moeller" wrote in message news:[email protected]...

Hi,

I have an SD disk which was running fine with an ext4 filesystem. Lately
I got the following errors:

[  133.736890] EXT4-fs error (device mmcblk0p2): ext4_ext_check_inode:462:
inode #131171: comm tput: bad header/extent: invalid eh_entries - magic
f30a, entries 65, max 4(4), depth 0(0)
[  133.775135] EXT4-fs error (device mmcblk0p2): ext4_ext_check_inode:462:
inode #131171: comm tput: bad header/extent: invalid eh_entries - magic
f30a, entries 65, max 4(4), depth 0(0)
[  133.874077] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:
comm start-stop-daem: bad extra_isize (32924 != 256)
[  134.298519] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:
comm startpar: bad extra_isize (32924 != 256)
[  134.317035] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:
comm startpar: bad extra_isize (32924 != 256)
[  134.815006] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:
comm init: bad extra_isize (32924 != 256)
[  134.845380] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:
comm init: bad extra_isize (32924 != 256)
[  134.866181] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:
comm init: bad extra_isize (32924 != 256)
[  134.894042] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:
comm sulogin: bad extra_isize (32924 != 256)
[  134.912231] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #46:

comm sulogin: bad extra_isize (32924 != 256)

which I tried to fix with fsck but it gets killed.

linux:~ # e2fsck -y /dev/sdb2
e2fsck 1.42.6 (21-Sep-2012)
Backing up journal inode block information.

/dev/sdb2 contains a file system with errors, check forced. Resize inode not valid. Recreate? yes

Pass 1: Checking inodes, blocks, and sizes
Inode 131179 has an invalid extent

       (logical block 0, invalid physical block 1342766638, len 1)

Clear? yes

Inode 131179, i_blocks is 8, should be 0. Fix? yes

Inodes that were part of a corrupted orphan linked list found. Fix? yes

Inode 131180 was part of the orphaned inode list. FIXED. Inode 131180 has a extra size (2076) which is invalid Fix? yes

Error while reading over extent tree in inode 131180: Corrupt extent header Clear inode? yes

Inode 131180, i_blocks is 8, should be 0. Fix? yes

Inode 131185 was part of the orphaned inode list. FIXED. Inode 131185, i_size is 144185556820644144, should be 614400. Fix? yes

Inode 131186 was part of the orphaned inode list. FIXED. Inode 131186 has a extra size (540) which is invalid Fix? yes

Inode 131186 has an invalid extent

       (logical block 0, invalid physical block 155898722910210, len 2)

Clear? yes

Inode 131186, i_blocks is 16, should be 0. Fix? yes

Error while reading over extent tree in inode 131187: Corrupt extent header Clear inode? yes

Inode 131187, i_blocks is 9, should be 0. Fix? yes

Inode 131188 has an invalid extent

       (logical block 0, invalid physical block 4400194552889, len 1)

Clear? yes

Inode 131188, i_blocks is 24, should be 0. Fix? yes

Inode 131193, i_size is 18014398509484072, should be 2199023259648. Fix? yes

Killed

Does this mean it is a HW error ?

Thank you
Markus

--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Theodore Ts'o

2013-10-20 15:24 UTC #2 lore raw

In reply to Markus Moeller

It's not clear how or why e2fsck is getting killed, so it's hard to say. It could be bug in e2fsck; or it could be any number of things. If you're willing to compile your own version of e2fsprogs, it would be great to download the latest e2fsprogs git tree, use the maint branch, and then build the latest version of e2fsck, which by default will come with debugging enabled, and see if you can replicate the failure using the latest version of e2fsck. (1.42.6 is a bit old, so it might not be applicable with the latest version; I also don't know how your distro might have compiled e2fsprogs).

If you can still get it to fail, it would be good to run it under gdb so we can see the stack trace when it got killed. Also, if you can grab a raw e2image dump while the file system is unmounted, that could be useful for us to try to debug what is going on.

Regards,

- Ted