Re: [bug report] fstests generic/774 hang

[John Garry <john.g.garry@oracle.com>]

On 05/11/2025 02:19, Shinichiro Kawasaki wrote:
> On Nov 04, 2025 / 16:33, Darrick J. Wong wrote:
>> [add jogarry/ojaswin since this is a new atomic writes test]
>>
>> On Thu, Oct 30, 2025 at 08:45:05AM +0000, Shinichiro Kawasaki wrote:
>>> I observe the fstests test case generic/774 hangs, when I run it for xfs on 8GiB
>>> TCMU fileio devices. It was observed with v6.17 and v6.18-rcX kernel versions.
>>> FYI, here I attach the kernel message log that was taken with v6.18-rc3 kernel
>>> [1]. The hang is recreated in stable manner by repeating the test case a few
>>> times in my environment.
>>>
>>> Actions for fix will be appreciated. If I can do any help, please let me know.
>>
>> I wonder, does your disk support atomic writes or are we just using the
>> software fallback in xfs?
> 
> I don't think the disk supports atomic writes. It is just a regular TCMU device,
> and its atomic write related sysfs attributes have value 0:
> 
>    $ grep -rne . /sys/block/sdh/queue/ | grep atomic
>    /sys/block/sdh/queue/atomic_write_unit_max_bytes:1:0
>    /sys/block/sdh/queue/atomic_write_boundary_bytes:1:0
>    /sys/block/sdh/queue/atomic_write_max_bytes:1:0
>    /sys/block/sdh/queue/atomic_write_unit_min_bytes:1:0
> 
> FYI, I attach the all sysfs queue attribute values of the device [2].

Yes, this would only be using software-based atomic writes.

Shinichiro, do the other atomic writes tests run ok, like 775, 767? You 
can check group "atomicwrites" to know which tests they are.

774 is the fio test.

Some things to try:
- use a physical disk for the TEST_DEV
- Don't set LOAD_FACTOR (if you were setting it). If not, bodge 774 to 
reduce $threads to a low value, say, 2
- trying turning on XFS_DEBUG config

BTW, Darrick has posted some xfs atomics fixes @ 
https://lore.kernel.org/linux-xfs/20251105001200.GV196370@frogsfrogsfrogs/T/#t. 
I doubt that they will help this, but worth trying.

I will try to recreate.

Thanks,
John