[Mesa-dev] [PATCH 6/9] i965/fs: Use the LRP instruction for ir_triop_lrp when possible.

[Ian Romanick]

On 02/26/2013 09:51 AM, Kenneth Graunke wrote:
> On 02/19/2013 05:03 PM, Matt Turner wrote:
>> diff --git a/src/mesa/drivers/dri/i965/brw_shader.cpp
>> b/src/mesa/drivers/dri/i965/brw_shader.cpp
>> index 9ab18cc..6965d72 100644
>> --- a/src/mesa/drivers/dri/i965/brw_shader.cpp
>> +++ b/src/mesa/drivers/dri/i965/brw_shader.cpp
>> @@ -156,7 +156,7 @@ brw_link_shader(struct gl_context *ctx, struct
>> gl_shader_program *shProg)
>>                SUB_TO_ADD_NEG |
>>                EXP_TO_EXP2 |
>>                LOG_TO_LOG2 |
>> -             LRP_TO_ARITH);
>> +             (stage == MESA_SHADER_FRAGMENT ? 0 : LRP_TO_ARITH));
>
> Doesn't this need to include a gen check as well?  Perhaps:
>
> const int lrp_to_arith = 0;
> if (intel->gen < 6 || stage != MESA_SHADER_FRAGMENT)
>      lrp_to_arith = LRP_TO_ARITH;
>
> then use lrp_to_arith here.

I like Ken's suggestion.  This is also a place where I would use ?:, but 
I won't be too pushy about it. ;)

	const int lrp_to_arith = (intel->gen < 6 || stage != 
MESA_SHADER_FRAGMENT) ? LRP_TO_ARITH : 0;

If you take Ken's snippet, you'll have to remove const.

>>         /* Pre-gen6 HW can only nest if-statements 16 deep.  Beyond this,
>>          * if-statements need to be flattened.
>>
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev