Re: kernel 5.2 read time tree block corruption

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




On 2019/10/15 下午11:03, José Luis wrote:
> Thanks for fast response Qu.
> 
> I booted into a pendrive live system for the test cause I'm using the
> involving fylesystem with kernel 4.19. This time when I mount
>> [manjaro@manjaro ~]$ sudo mount /dev/sdb2 /mnt
>> mount: /mnt: no se puede leer el superbloque en /dev/sdb2.
> and in the dmesg:
> [ +30,866472] BTRFS info (device sdb2): disk space caching is enabled
> [  +0,017443] BTRFS info (device sdb2): enabling ssd optimizations
> [  +0,000637] BTRFS critical (device sdb2): corrupt leaf: root=5
> block=32145457152 slot=99, invalid key objectid: has
> 18446744073709551605 expect 6 or [256, 18446744073709551360] or
> 18446744073709551604
> [  +0,000002] BTRFS error (device sdb2): block=32145457152 read time
> tree block corruption detected
> [  +0,000012] BTRFS warning (device sdb2): failed to read fs tree: -5
> [  +0,061995] BTRFS error (device sdb2): open_ctree failed
> 
> So I suppose you need dump output from the block 32145457152 so I pastebin that:
> sudo btrfs ins dump-tree -b 32145457152 /dev/sdb2
> output --> https://pastebin.com/ssB5HTn7

The output is way crazier than I thought...

I was only expecting some strange inode number, but what I got is
completely ridiculous.

From item 96, we are having completely impossible inodes.
From FREE_INO to DATA_RELOC, even EXTENT_CSUM.

All of these are impossible to exist in fs tree.
The most strange thing is, they are all last modified in 2019-2-15.

Anyway, the tree-checker is doing completely valid behavior for this
case. The data is really ridiculous.

Any history about the kernel used in that time?
I see something only possible in Windows, any clue?

> 
> Please provide the parameter to the grep redirection for: "btrfs ins
> dump-tree -t 5 /dev/sdb2 | grep -A 7"

My bad, the parameter is "(431"

It will output all info about inode 431, so we can make sure what's
going wrong.

Thanks,
Qu
> 
> El mar., 15 oct. 2019 a las 14:38, Qu Wenruo
> (<quwenruo.btrfs@xxxxxxx>) escribió:
>>
>>
>>
>> On 2019/10/15 下午8:24, Qu Wenruo wrote:
>>>
>>>
>>> On 2019/10/15 下午6:15, José Luis wrote:
>>>> Dear devs,
>>>>
>>>> I cannot use kernel >= 5.2, They cannot mount sdb2 nor sb3 both btrfs
>>>> filesystems. I can work as intended on 4.19 which is an LTS version,
>>>> previously using 5.1 but Manjaro removed it from their repositories.
>>>>
>>>> More info:
>>>> · dmesg:
>>>>> [oct15 13:47] BTRFS info (device sdb2): disk space caching is enabled
>>>>> [  +0,009974] BTRFS info (device sdb2): enabling ssd optimizations
>>>>> [  +0,000481] BTRFS critical (device sdb2): corrupt leaf: root=5 block=30622793728 slot=115, invalid key objectid: has 18446744073709551605 expect 6 or [256, 18446744073709551360] or 18446744073709551604
>>>
>>> In fs tree, you are hitting a free space cache inode?
>>> That doesn't sound good.
>>>
>>> Please provide the following dump:
>>>
>>> # btrfs ins dump-tree -b 30622793728 /dev/sdb2
>>>
>>> The output may contain filename, feel free to remove filenames.
>>>
>>>>> [  +0,000002] BTRFS error (device sdb2): block=30622793728 read time tree block corruption detected
>>>>> [  +0,000021] BTRFS warning (device sdb2): failed to read fs tree: -5
>>>>> [  +0,044643] BTRFS error (device sdb2): open_ctree failed
>>>>
>>>>
>>>>
>>>> · sudo mount  /dev/sdb2 /mnt/
>>>>> mount: /mnt: no se puede leer el superbloque en /dev/sdb2.
>>>>
>>>> (cannot read superblock on /dev...)
>>>>
>>>> · sudo btrfs rescue super-recover /dev/sdb2
>>>>> All supers are valid, no need to recover
>>>>
>>>>
>>>> · sudo btrfs check /dev/sdb2
>>>>> Opening filesystem to check...
>>>>> Checking filesystem on /dev/sdb2
>>>>> UUID: ff559c37-bc38-491c-9edc-fa6bb0874942
>>>>> [1/7] checking root items
>>>>> [2/7] checking extents
>>>>> [3/7] checking free space cache
>>>>> cache and super generation don't match, space cache will be invalidated
>>>>> [4/7] checking fs roots
>>>>> root 5 inode 431 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 755 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 2379 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 11721 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 12211 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 15368 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 35329 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 960427 errors 1040, bad file extent, some csum missing
>>>>> root 5 inode 18446744073709551605 errors 2001, no inode item, link count wrong
>>>>>         unresolved ref dir 256 index 0 namelen 12 name $RECYCLE.BIN filetype 2 errors 6, no dir index, no inode ref
>>>
>>> Check is reporting the same problem of the inode.
>>>
>>> We need to make sure what's going wrong on that leaf, based on the
>>> mentioned dump.
>>>
>>> For the csum missing error and bad file extent, it should be a big problem.
>>
>> s/should/should not/
>>
>>> if you want to make sure what's going wrong, please provide the
>>> following dump:
>>>
>>> # btrfs ins dump-tree -t 5 /dev/sdb2 | grep -A 7
>>>
>>> Also feel free the censor the filenames.
>>>
>>> Thanks,
>>> Qu
>>>
>>>>> root 388 inode 1245 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 1288 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 1292 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 1313 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 11870 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 68126 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 88051 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 88255 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 88455 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 88588 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 88784 errors 1040, bad file extent, some csum missing
>>>>> root 388 inode 88916 errors 1040, bad file extent, some csum missing
>>>>> ERROR: errors found in fs roots
>>>>> found 37167415296 bytes used, error(s) found
>>>>> total csum bytes: 33793568
>>>>> total tree bytes: 1676722176
>>>>> total fs tree bytes: 1540243456
>>>>> total extent tree bytes: 81510400
>>>>> btree space waste bytes: 306327457
>>>>> file data blocks allocated: 42200928256
>>>>>  referenced 52868354048
>>>>
>>>>
>>>>
>>>>
>>>> ---
>>>>
>>>> Regards,
>>>> José Luis.
>>>>
>>>
>>

Attachment: signature.asc
Description: OpenPGP digital signature


[Index of Archives]     [Linux Filesystem Development]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux