[Piglit] [PATCH] arb_shader_image_load_store: fix MSVC build

Francisco Jerez currojerez at riseup.net
Tue Feb 3 13:28:10 PST 2015


Jose Fonseca <jfonseca at vmware.com> writes:

> On 03/02/15 20:08, Francisco Jerez wrote:
>> Francisco Jerez <currojerez at riseup.net> writes:
>>
>>> Francisco Jerez <currojerez at riseup.net> writes:
>>>
>>>> Brian Paul <brianp at vmware.com> writes:
>>>>
>>>>> MSVC doesn't seem to like function calls inside struct initializers.
>>>>
>>>> That's awful.  But fine:
>>>> Reviewed-by: Francisco Jerez <currojerez at riseup.net>
>>>>
>>> Actually it's surprising that you got it to build with so few changes.
>>> I'd expect you to find *many* more function calls inside struct
>>> initializers.  Does it actually work with this change or did you forget
>>> to include some files?
>>>
>>
>> Brian, if that's the case don't bother to fix all the function calls in
>> struct initalizers, just let me know, I'd like to try a less intrusive
>> solution.
>
> If this was C++ we could use C++ constructors.  In fact, the code could 
> be substantially simplified.
>
> I think we should make more use of C++ in piglit.  MSVC C++ support is 
> far more polished than it's C support.  But above all, C++ features like 
> classes and templates could save a lot of hand typing.
>
> Jose

Yes, you're probably right.  Actually except for the implicit void
pointer conversions it might be possible to get these tests to compile
as C++ source with just minor changes, MSVC would likely be more
forgiving with that.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 212 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/piglit/attachments/20150203/60b6a100/attachment.sig>


More information about the Piglit mailing list