Re: [PATCH v4 00/18] All-stop on top of non-stop

[Luis Machado <lgustavo@codesourcery.com>]

On 08/12/2015 05:50 PM, Joel Brobecker wrote:
>> My idea of a testcase comment is at the beginning of the testcase file,
>> explaining what the test does and why it does it. I'd mention the amd64
>> example as well, since it is part of why the test was created in the first
>> place.
>>
>> That should give others enough background to pursue an investigation about
>> why this potentially fails for them.
>>
>> My 2 cents anyway.
>
> There is this perception that the testcase was created because
> of the issue on amd64, but that's not true. The testcase was
> created, albeit in AdaCore's infrastructure only, to test that
> "next" in that context works as expected. Only later on did it
> allow us to find another bug which actually has nothing to do
> with the initial reason for creating the testcase. I hope I'm not
> looking like I'm splitting hair, but I feel like there is a bit
> of a misunderstanding somewhere, probably because the testcase
> appears as new to the  GDB community and was combined with an
> amd64-specific fix.
>

What caused confusion was the fact that the test was sent together with 
a fix, which seemed to imply it was a fix-specific test as opposed to a 
testcase-less fix and a generic new testcase in the same patch.

> That being said, I propose the attached patch. I confess I'm not
> super convinced about the comment on amd64, as I think it might
> become one day irrelevant. But I don't mind it that much; if
> it is helpful to others...

The description looks good. Feel free to drop the amd64 reference if you 
think it is not worth it. I don't have a strong opinion on it.

It just felt like the testcase needed a bit more information on why it 
was created in the first place.