[Mesa-dev] [PATCH 11/19] anv: Use GNU C empty brace initializer

[Ian Romanick]

On 08/29/2017 10:19 AM, Emil Velikov wrote:
> On 29 August 2017 at 18:10, Matt Turner <mattst88 at gmail.com> wrote:
>> On Tue, Aug 29, 2017 at 3:35 AM, Emil Velikov <emil.l.velikov at gmail.com> wrote:
>>> On 29 August 2017 at 11:11, Eric Engestrom <eric.engestrom at imgtec.com> wrote:
>>>> On Monday, 2017-08-28 14:57:13 -0700, Matt Turner wrote:
>>>>> Avoids Clang's warning about the current code:
>>>>>
>>>>>    warning: suggest braces around initialization of subobject
>>>>
>>>> I'm not sure about this patch [1], but everything else in this series is:
>>> IIRC "{}" is the least likely way to avoid the warnings, across
>>> GCC/Clang versions.
>>> It's not part of the C spec, but since it works - not sure how much to bother.
>>
>> I'm surprised to hear you say that. I've tested with gcc-4.9.4, 6.4.0,
>> 7.1.0; and clang-4.0, and none of them warn about {}
>>
> Eric corrected me on IRC, but I forgot to mention here - "s/least/most/"
> 
>> I think when we went through this before (in NIR) we couldn't use {}
>> because it needs to compile with MSVC, and MSVC doesn't allow that in
>> C.
>>
> Precisely. The "most" portable solution that I know of is memset().

Except memset generates different code.  That will clear the padding
area between fields, and x = {} (and variations there of) do not.

>>> Would be great to have the issue reported (and fixed) in Clang though.
>>> AFAICT both {0} and {0,} are valid, if memory serves me right.
>>
>> I don't really understand the nuances of {0}, {0,}, {{0}}, etc, and I
>> get the sense that that's the case for nearly everyone else as well.
>>
> AFAICT {0} and {0,} should just work everywhere, modulo bugs.

And I believe most of these generate warnings if you enable other
warning options (see also this patch series
https://patchwork.freedesktop.org/patch/68141/).

> [1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53119
> [2] https://bugs.llvm.org/show_bug.cgi?id=21689
> 
>> The most prominent bug I see in all of this is that {} isn't part of
>> standard C (it is part of standard C++!)
> Agreed. Not sure how easy it will be to convince the committee.
> 
> -Emil
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/mesa-dev