Re: [BUG report] fstests/btrfs/080 fails

[Glass Su <glass.su@suse.com>]

> On Feb 14, 2025, at 13:26, Qu Wenruo <quwenruo.btrfs@gmx.com> wrote:
> 
> 
> 
> 在 2025/2/14 15:41, Glass Su 写道:
>> 
>> Hi
>> 
>> Recently I found btrfs/080 fails like:
>> 
>> btrfs/080 124s ... [failed, exit status 1]- output mismatch (see /root/xfstests-dev/results//btrfs/080.out.bad)
>>     --- tests/btrfs/080.out     2024-08-29 09:10:14.933333334 +0800
>>     +++ /root/xfstests-dev/results//btrfs/080.out.bad   2025-02-14 12:53:24.667572260 +0800
>>     @@ -1,2 +1,3 @@
>>      QA output created by 080
>>     -Silence is golden
>>     +Unexpected digest for file /mnt/scratch/12_52_59_984815662_snap/foobar_39
>>     +(see /root/xfstests-dev/results//btrfs/080.full for details)
>>     ...
>>     (Run 'diff -u /root/xfstests-dev/tests/btrfs/080.out /root/xfstests-dev/results//btrfs/080.out.bad'  to see the entire diff)
>> Ran: btrfs/080
>> Failures: btrfs/080
>> Failed 1 of 1 tests
>> 
>> It can be reproduced once in about 20 times on v6.13, misc-next(HEAD: 1c08f86eeadab89e8f6ad8559df54633afb7a3ba)
>> in my VM with 32 cores.
>> 
>> Configs and log are attached.
> 
> I checked your kernel config, it looks like it has a config that is
> known to cause problems:
> 
> - CONFIG_PT_RECLAIM=y
> 
> I'm unable to reproduce the bug locally, with 64 runs.

The test stresses CPU I guess reproduce possibility is CPU cores sensitive.

> But that's with CONFIG_PT_RECLAIM=n, as I use that config to workaround
> the bug.
> 
> Mind to test with either that config disabled, or apply this hotfix and
> retry?
> 

> https://lore.kernel.org/linux-mm/20250211072625.89188-1-zhengqi.arch@bytedance.com/
> 
Unfortunately it still fails after patching this.

— 
Su
> Thanks,
> Qu
>> 
>> 
>> —
>> Su