[Mesa-dev] [PATCH] mesa: update some old-style (K&R?) function pointer calls

[Ian Romanick]

On 11/20/2015 12:46 PM, Emil Velikov wrote:
> On 20 November 2015 at 20:40, Brian Paul <brianp at vmware.com> wrote:
>> ---
>>  src/mesa/main/blend.c     | 4 ++--
>>  src/mesa/main/buffers.c   | 2 +-
>>  src/mesa/main/fog.c       | 2 +-
>>  src/mesa/main/getstring.c | 2 +-
>>  src/mesa/main/points.c    | 2 +-
>>  src/mesa/main/texenv.c    | 2 +-
>>  6 files changed, 7 insertions(+), 7 deletions(-)
>>
>> diff --git a/src/mesa/main/blend.c b/src/mesa/main/blend.c
>> index f07552b..2ae22e9 100644
>> --- a/src/mesa/main/blend.c
>> +++ b/src/mesa/main/blend.c
>> @@ -404,7 +404,7 @@ _mesa_BlendEquation( GLenum mode )
>>     ctx->Color._BlendEquationPerBuffer = GL_FALSE;
>>
>>     if (ctx->Driver.BlendEquationSeparate)
>> -      (*ctx->Driver.BlendEquationSeparate)( ctx, mode, mode );
>> +      ctx->Driver.BlendEquationSeparate(ctx, mode, mode);
>>  }
>>
>>
>> @@ -582,7 +582,7 @@ _mesa_BlendColor( GLclampf red, GLclampf green, GLclampf blue, GLclampf alpha )
>>     ctx->Color.BlendColor[3] = CLAMP(tmp[3], 0.0F, 1.0F);
>>
>>     if (ctx->Driver.BlendColor)
>> -      (*ctx->Driver.BlendColor)(ctx, ctx->Color.BlendColor);
>> +      ctx->Driver.BlendColor(ctx, ctx->Color.BlendColor);
>>  }
>>
>>
>> diff --git a/src/mesa/main/buffers.c b/src/mesa/main/buffers.c
>> index 93588a2..83e238a 100644
>> --- a/src/mesa/main/buffers.c
>> +++ b/src/mesa/main/buffers.c
>> @@ -731,7 +731,7 @@ _mesa_read_buffer(struct gl_context *ctx, struct gl_framebuffer *fb,
>>     /* Call the device driver function only if fb is the bound read buffer */
>>     if (fb == ctx->ReadBuffer) {
>>        if (ctx->Driver.ReadBuffer)
>> -         (*ctx->Driver.ReadBuffer)(ctx, buffer);
>> +         ctx->Driver.ReadBuffer(ctx, buffer);
>>     }
>>  }
>>
>> diff --git a/src/mesa/main/fog.c b/src/mesa/main/fog.c
>> index 45f343d..1ad939c 100644
>> --- a/src/mesa/main/fog.c
>> +++ b/src/mesa/main/fog.c
>> @@ -190,7 +190,7 @@ _mesa_Fogfv( GLenum pname, const GLfloat *params )
>>     }
>>
>>     if (ctx->Driver.Fogfv) {
>> -      (*ctx->Driver.Fogfv)( ctx, pname, params );
>> +      ctx->Driver.Fogfv( ctx, pname, params );
>>     }
>>
>>     return;
>> diff --git a/src/mesa/main/getstring.c b/src/mesa/main/getstring.c
>> index 06ba17c..87c5a3a 100644
>> --- a/src/mesa/main/getstring.c
>> +++ b/src/mesa/main/getstring.c
>> @@ -121,7 +121,7 @@ _mesa_GetString( GLenum name )
>>     assert(ctx->Driver.GetString);
>>     {
>>        /* Give the driver the chance to handle this query */
>> -      const GLubyte *str = (*ctx->Driver.GetString)(ctx, name);
>> +      const GLubyte *str = ctx->Driver.GetString(ctx, name);
>>        if (str)
>>           return str;
>>     }
>> diff --git a/src/mesa/main/points.c b/src/mesa/main/points.c
>> index 863e3c1..c2f2b63 100644
>> --- a/src/mesa/main/points.c
>> +++ b/src/mesa/main/points.c
>> @@ -209,7 +209,7 @@ _mesa_PointParameterfv( GLenum pname, const GLfloat *params)
>>     }
>>
>>     if (ctx->Driver.PointParameterfv)
>> -      (*ctx->Driver.PointParameterfv)(ctx, pname, params);
>> +      ctx->Driver.PointParameterfv(ctx, pname, params);
>>  }
>>
>>
>> diff --git a/src/mesa/main/texenv.c b/src/mesa/main/texenv.c
>> index 0919221..93c6806 100644
>> --- a/src/mesa/main/texenv.c
>> +++ b/src/mesa/main/texenv.c
>> @@ -495,7 +495,7 @@ _mesa_TexEnvfv( GLenum target, GLenum pname, const GLfloat *param )
>>
>>     /* Tell device driver about the new texture environment */
>>     if (ctx->Driver.TexEnv) {
>> -      (*ctx->Driver.TexEnv)( ctx, target, pname, param );
>> +      ctx->Driver.TexEnv(ctx, target, pname, param);
>>     }
>>  }
>>
> Fwiw I'm all for nuking these. Any non-historical reasons why these exist ?

We used to do it like that because it was more explicit that you will
calling a function pointer.  In this sort of case it's pretty obvious.
But if you see

    (*foo)(param1, param2);

versus

    foo(param1, param2);

the former is obviously a function pointer whereas it's not obvious
about the latter.

I think it's fine to make this change (or other
function-pointer-in-a-struct) changes, but I'd /mildly/ prefer the
explicit way for other cases... though I suspect there's a lot of
deviation from it in the codebase already.

> The is a ton in src/glx which makes the code even less pleasant (is
> that even possible) to read... in case you're bored :-P

That sounds like a job for sed.

> Reviewed-by: Emil Velikov <emil.l.velikov at gmail.com>
> 
> -Emil
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev
>