Re: [Intel-gfx] [PATCH] drm/i915/selftests: Fix uninitialized variable

[Jani Nikula <jani.nikula@linux.intel.com>]

On Wed, 04 Mar 2020, Jani Nikula <jani.nikula@linux.intel.com> wrote:
> On Tue, 03 Mar 2020, Matt Roper <matthew.d.roper@intel.com> wrote:
>> On Tue, Mar 03, 2020 at 05:25:21PM +0200, Jani Nikula wrote:
>>> On Tue, 03 Mar 2020, Aditya Swarup <aditya.swarup@intel.com> wrote:
>>> > -	struct lrc_timestamp data;
>>> > +	struct lrc_timestamp data = { 0 };
>>> 
>>> {} is preferred over {0}.
>>
>> Is there a reference for this (e.g., in the kernel coding style)?  I
>> thought this came up a couple years ago and the consensus was the other
>> way, although I could be misremembering.  Unless it's changed in a
>> recent standard, I think {} is only legal in C++, so using it in C code
>> is a gcc-ism?
>
> Both are widely used in the kernel. I think we've mostly converged to {}
> in i915. Yes, it's a gcc-ism in C code, but the kernel is gcc, not
> standard C.
>
> I can't find a reference right now, but ISTR there are some warnings
> issued in some cases with the {0} initializer, depending on the struct
> and perhaps on the compiler.

Here's one [1].

BR,
Jani.


[1] http://patchwork.freedesktop.org/patch/msgid/20200304183654.GA9011@paulmck-ThinkPad-P72


-- 
Jani Nikula, Intel Open Source Graphics Center
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx