Re: EXT4 regression caused 4eec7

[Eric Sandeen <sandeen@redhat.com>]

On 5/14/13 2:11 AM, Dmitry Monakhov wrote:
> On Mon, 13 May 2013 12:09:22 -0500, Eric Sandeen <sandeen@redhat.com> wrote:
>> On 5/13/13 12:01 PM, Jan Kara wrote:
>>> On Mon 13-05-13 11:34:12, Eric Sandeen wrote:
>>>> On 5/12/13 4:01 AM, Dmitry Monakhov wrote:
>>>>> In fact '4eec70' are vexing because I have reviewed and tested this patch before
>>>>> it was marked as Review-by, but missed the bug. This is because xfstests
>>>>> was executed manually logs was full of warnings but tainted flag was not
>>>>> checked at the end. 
>>>>
>>>> Can you elaborate on this?  What was logged, and is it something we could
>>>> try to pick up post-test in xfstests?
>>>   Generally I think it might be useful if xfstests would fail / warn if
>>> kernel became tainted during the test (e.g. due to WARN_ON or oops, or
>>> something like that). It should be even relatively easy to implement
>>> (just compare /proc/sys/kernel/tainted before and after each test).
>>>
>>> 								Honza
>>>
>>
>> Ah, right.  That should be easy, I'll see if I can cook that up.
> Also we can use abrt's kernel-oops handler to collect messages.

I sent a pretty simple patch to just check the sysctl to the xfs list
yesterday.

-Eric