[LLVMdev] teaching FileCheck to handle variations in order

[Krzysztof Parzyszek]

On 9/13/2012 4:49 PM, Chandler Carruth wrote:
>
> I mean, it's inconvenient, mostly for the person changing the scheduler,
> but provided there are helpful comments for them, I suspect they're not
> going to have a hard time updating the test cases.

Why would they have to do that in the first place?  Like I said, tests 
should not be checking things that they were not specifically meant to 
test.  Hard time or not, I don't think anyone would be excited about 
having to go through a list of potential false positives.

-K


-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, 
hosted by The Linux Foundation