Bug#685726: linux-image-3.2.0-3-amd64: return error when trying to format image file (mkfs -t ext4 file.img)

[hrayr@infotech.am]

> On Fri, 2012-08-24 at 03:30 +0400, Hrayr Grigoryan wrote:
>> Package: src:linux
>> Version: 3.2.23-1
>> Severity: critical
>> Tags: lfs
>> Justification: causes serious data loss
>>
>> Dear Maintainer,
>>
>>    When I'm trying to format file image with the command "mkfs -t ext4
>> file.img", it returns the following errors
>> [  142.328065] EXT4-fs error (device sdb1): ext4_ext_map_blocks:3769:
>> inode #12: comm mkfs.ext4: bad extent address lblock: 1022, depth: 2
>> pblock 0
>> [  142.328387] EXT4-fs error (device sdb1): ext4_ext_map_blocks:3769:
>> inode #12: comm mkfs.ext4: bad extent address lblock: 1023, depth: 2
>> pblock 0
>> [  142.328699] EXT4-fs error (device sdb1): ext4_ext_map_blocks:3769:
>> inode #12: comm mkfs.ext4: bad extent address lblock: 9254, depth: 2
>> pblock 0
>> [  142.329018] EXT4-fs error (device sdb1): ext4_ext_map_blocks:3769:
>> inode #12: comm mkfs.ext4: bad extent address lblock: 1057, depth: 2
>> pblock 0
>
> So you have an ext4 filesystem in file.img, on top of an ext4 filesystem
> on /dev/sdb1?
yes, all are ext4, no other file systems I have on that machine.


>
>>    The problem visible only for big images f.x. 160Gb and more.
>>    I did the tests on different hardwares, but result the same.
> [...]
>
> How is test.img created, before you run mkfs?  Is it a sparse file or
> does it have all data blocks allocated?
No it is not sparse file, test.img was created with the following command
dd if=/dev/zero of=test.img bs=1M count=160k


> Have you tested this on any other kernel versions (earlier or later)?
Yes, I have tested with the earlier versions, all earlier versions of
3.x.x kernels have this bug.

Hrayr.



>
> Ben.
>
> --
> Ben Hutchings
> Experience is what causes a person to make new mistakes instead of old
> ones.
>