Re: [PATCH v2 3/8] s390/vfio-ap: Fix use of wrong lock in mdev probe function

[Matthew Rosato]

>> Sashiko reports that this patch introduces a regression, which seems
>> valid.
>>
>> AFAICT you resolve this regression with patch 7 of this series.
>>
>> For the sake of bisect, can you look at whether squashing these 2
>> patches together would work so that we don't have an interim regression?
> 
> The reason for the two different patches is because the shah on the
> Fixes: tag
> differ.
> 

That's not a good enough reason to break bisect and introduce a regression.