[PATCH] nvme: add nvme pci timeout testcase

Chaitanya Kulkarni chaitanyak at nvidia.com
Wed Jan 10 15:47:22 PST 2024


On 1/10/2024 1:24 AM, Hannes Reinecke wrote:
> On 1/10/24 09:12, Chaitanya Kulkarni wrote:
>>
>>>>>
>>>>> Should this be TEST_DEV instead ?
>>>>
>>>> why ?
>>>>
>>> My understanding of blktests is, add device which we want to test in
>>> config files under TEST_DEV (except null-blk and nvme-fabrics loopback
>>> devices, which are usually populated inside the tests).
>>> In this case, if someone do not want to disturb nvme0n1 device,
>>> this test doesn't allow it.
>>>
>>> Regards,
>>> Nitesh Shetty
>>>
>>
>> it is clearly stated in the documentation that blktests are destructive
>> to the entire system and including any devices you have, if your
>> system has sensitive data then _don't run these tests_ simple, when
>> you are running blktests you are bound to disturb system you can't
>> prevent that by using TEST_DEV.
>>
> I don't think this is the direction we want to go.
> NVMe for internal drives is becoming more and more prevalent (especially
> on laptops), making them unusable for running blktests.
> 
> We have been putting quite some effort into nvme-cli to ensure that
> blktest _can_ run concurrently with other NVMe drives, so really we
> should not hard-code any device names.
> 

Okay ...

> Or we discuss this at LSF/MM; Daniel or me will be happy to give an
> overview about concurrent NVMe-oF management applications on the same
> system (nvme-cli, nvme-stas, blktests all running at the same time).
> 

would be a great topic for the blktests session ...

> Cheers,
> 
> Hannes
> 

I'll make it TEST_DEV change ...

-ck




More information about the Linux-nvme mailing list