Re: [PATCH] btrfs: rescue: allow ibadroots to skip bad extent tree when reading block group items

[Qu Wenruo <wqu@suse.com>]

On 2021/7/15 上午10:17, Anand Jain wrote:
> On 15/07/2021 09:24, Qu Wenruo wrote:
>> When extent tree gets corrupted, normally it's not extent tree root, but
>> one toasted tree leaf/node.
>>
>> In that case, rescue=ibadroots mount option won't help as it can only
>> handle the extent tree root corruption.
>>
>> This patch will enhance the behavior by:
>>
>> - Allow fill_dummy_bgs() to ignore -EEXIST error
>>
>>    This means we may have some block group items read from disk, but
>>    then hit some error halfway.
>>
>> - Fallback to fill_dummy_bgs() if any error gets hit in
>>    btrfs_read_block_groups()
>>
>>    Of course, this still needs rescue=ibadroots mount option.
>>
>> With that, rescue=ibadroots can handle extent tree corruption more
>> gracefully and allow a better recover chance.
>>
>> Reported-by: Zhenyu Wu <wuzy001@gmail.com>
>> Link: https://www.spinics.net/lists/linux-btrfs/msg114424.html
>> Signed-off-by: Qu Wenruo <wqu@suse.com>
>> ---
>>   fs/btrfs/block-group.c | 20 +++++++++++++++++---
>>   1 file changed, 17 insertions(+), 3 deletions(-)
>>
>> diff --git a/fs/btrfs/block-group.c b/fs/btrfs/block-group.c
>> index 5bd76a45037e..33086b882fe8 100644
>> --- a/fs/btrfs/block-group.c
>> +++ b/fs/btrfs/block-group.c
>> @@ -2105,11 +2105,16 @@ static int fill_dummy_bgs(struct btrfs_fs_info 
>> *fs_info)
>>           bg->used = em->len;
>>           bg->flags = map->type;
>>           ret = btrfs_add_block_group_cache(fs_info, bg);
>> -        if (ret) {
>> +        /*
>> +         * We may have some block groups filled already, thus ignore
>> +         * the -EEXIST error.
>> +         */
>> +        if (ret && ret != -EEXIST) {
>>               btrfs_remove_free_space_cache(bg);
>>               btrfs_put_block_group(bg);
>>               break;
>>           }
>> +        ret = 0;
>>           btrfs_update_space_info(fs_info, bg->flags, em->len, em->len,
>>                       0, 0, &space_info);
>>           bg->space_info = space_info;
> 
> 
>> @@ -2139,8 +2144,10 @@ int btrfs_read_block_groups(struct 
>> btrfs_fs_info *info)
>>       key.offset = 0;
>>       key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
>>       path = btrfs_alloc_path();
>> -    if (!path)
>> -        return -ENOMEM;
>> +    if (!path) {
>> +        ret = -ENOMEM;
>> +        goto error;
>> +    }
> 
>    Return -ENOMEM; is correct as alloc failure is not part of the 
> corruption?

Oh, right. In that case we shouldn't try to fill dummy eb, especially it 
will also fail with -ENOMEM anyway.

Thanks,
Qu
> 
> Thanks, Anand
> 
>>       cache_gen = btrfs_super_cache_generation(info->super_copy);
>>       if (btrfs_test_opt(info, SPACE_CACHE) &&
>> @@ -2212,6 +2219,13 @@ int btrfs_read_block_groups(struct 
>> btrfs_fs_info *info)
>>       ret = check_chunk_block_group_mappings(info);
>>   error:
>>       btrfs_free_path(path);
>> +    /*
>> +     * Either we have no extent_root (already implies 
>> IGNOREBADROOTS), or
>> +     * we hit error and have IGNOREBADROOTS, then we can try to use 
>> dummy
>> +     * bgs.
>> +     */
>> +    if (!info->extent_root || (ret && btrfs_test_opt(info, 
>> IGNOREBADROOTS)))
>> +        ret = fill_dummy_bgs(info);
>>       return ret;
>>   }
>>
>