Re: [LSF/MM/BPF ATTEND][LSF/MM/BPF TOPIC] : blktests: status, expansion plan for the storage stack test framework

[Nilay Shroff <nilay@linux.ibm.com>]

On 4/21/26 11:49 AM, Shin'ichiro Kawasaki wrote:
> On Feb 16, 2026 / 00:08, Nilay Shroff wrote:
>>
>>
>> On 2/13/26 4:53 PM, Shinichiro Kawasaki wrote:
>>> On Feb 12, 2026 / 08:52, Daniel Wagner wrote:
>>>> On Wed, Feb 11, 2026 at 08:35:30PM +0000, Chaitanya Kulkarni wrote:
>>>>>     For the storage track at LSFMMBPF2026, I propose a session dedicated to
>>>>>     blktests to discuss expansion plan and CI integration progress.
>>>>
>>>> Thanks for proposing this topic.
>>>
>>> Chaitanya, my thank also goes to you.
>>>
>> Yes thanks for proposing this!
>>
>>>> Just a few random topics which come to mind we could discuss:
>>>>
>>>> - blktests has gain a bit of traction and some folks run on regular
>>>>    basis these tests. Can we gather feedback from them, what is working
>>>>    good, what is not? Are there feature wishes?
>>>
>>> Good topic, I also would like to hear about it.
>>>
>> One improvement I’d like to highlight is related to how blktests are executed
>> today. So far, we’ve been running blktests serially, but if it's possible to
>> run tests in parallel to improve test turnaround time and make large-scale or
>> CI-based testing more efficient? For instance, adding parallel_safe Tags: Marking tests
>> that don't modify global kernel state so they can be safely offloaded to parallel
>> workers. Marking parallel_safe tags would allow the runner to distinguish:
>>
>> Safe Tests: Tests that only perform I/O on a specific, non-shared device or
>> check static kernel parameters.
>>
>> Unsafe Tests: Tests that reload kernel modules, modify global /sys or /proc entries,
>> or require exclusive access to specific hardware addresses.
>>
>> Yes adding parallel execution support shall require framework/design changes.
> 
> Hi Nilay, thanks for the idea. I understand that shorter test time will make CI
> cycles faster and improve the development efficiency.
> 
> Said that, the safe/unsafe testing idea may not be enough. I think majority of
> test case does kernel module set up using null_blk, scsi_debug, or nvme target
> drivers. Then I foresee the majority of the test cases will be "unsafe", and
> cannot be run in parallel.
> 
> Also, parallel runs on single system will affect dmesg or kmemleak checking.
> We cannot tell which run caused a dmesg message or a memory leak.
> 
> For the runtime reduction by parallel runs, I guess blktests run on VMs might be
> the good approach as Haris pointed out. Anyway, this topic will need more
> discussion.
> 
> [...]

Alright, see if we may discuss this during LSFMM.

> 
>>>   4. Long standing failures make test result reports dirty
>>>      - I feel lockdep WARNs are tend to be left unfixed rather long period.
>>>        How can we gather effort to fix them?
>>
>> I agree regarding lockdep; recently we did see quite a few lockdep splats.
>> That said, I believe the number has dropped significantly and only a small
>> set remains. From what I can tell, most of the outstanding lockdep issues
>> are related to fs-reclaim paths recursing into the block layer while the
>> queue is frozen. We should be able to resolve most of these soon, or at
>> least before the conference. If anything is still outstanding after that,
>> we can discuss it during the conference and work toward addressing it as
>> quickly as possible.
> 
> Taking this chance, I'd like to express my appreciation for the effort to
> resolve the lockdep issues. It is great that a number of lockdeps are already
> fixed. Said that, two lockdep issues are still observed with v7.0 kernel at
> nvme/005 and nbd/002 [1]. I would like to gather attentions to the failures.
> 
> [1] https://lore.kernel.org/linux-block/ynmi72x5wt5ooljjafebhcarit3pvu6axkslqenikb2p5txe57@ldytqa2t4i2x/
> 
I think nvme/005 and nbd/002 failures shall be addressed with this
patch: https://lore.kernel.org/all/20260413171628.6204-1-kch@nvidia.com/

It's currently applied to nvme-7.1 and not there yet to mainline kernel.

Thanks,
--Nilay