Re: [Mesa-dev] [PATCH] i965: compute DDX in a subspan based only on top row

2013-10-02 Thread Chia-I Wu
On Wed, Oct 2, 2013 at 2:14 AM, Chris Forbes wrote: > With those fixes: > > Reviewed-by: Chris Forbes Thanks, I will push it shortly. With this change landed, the slowness of sample_d is no longer the bottleneck. Instead, the lack of native SIMD16 sample_d becomes the problem. I have posted my

Re: [Mesa-dev] [PATCH] i965: compute DDX in a subspan based only on top row

2013-10-01 Thread Chris Forbes
With those fixes: Reviewed-by: Chris Forbes On Wed, Oct 2, 2013 at 6:38 AM, Ian Romanick wrote: > On 09/30/2013 10:54 PM, Chia-I Wu wrote: >> From: Chia-I Wu > > I agree with both of Ken's comments. With those fixed, this patch is > > Reviewed-by: Ian Romanick > >> Consider only the top-left

Re: [Mesa-dev] [PATCH] i965: compute DDX in a subspan based only on top row

2013-10-01 Thread Ian Romanick
On 09/30/2013 10:54 PM, Chia-I Wu wrote: > From: Chia-I Wu I agree with both of Ken's comments. With those fixed, this patch is Reviewed-by: Ian Romanick > Consider only the top-left and top-right pixels to approximate DDX in a 2x2 > subspan, unless the application requests a more accurate ap

Re: [Mesa-dev] [PATCH] i965: compute DDX in a subspan based only on top row

2013-09-30 Thread Kenneth Graunke
On 09/30/2013 10:54 PM, Chia-I Wu wrote: > From: Chia-I Wu > > Consider only the top-left and top-right pixels to approximate DDX in a 2x2 > subspan, unless the application requests a more accurate approximation via > GL_FRAGMENT_SHADER_DERIVATIVE_HINT or this optimization is disabled from the >

[Mesa-dev] [PATCH] i965: compute DDX in a subspan based only on top row

2013-09-30 Thread Chia-I Wu
From: Chia-I Wu Consider only the top-left and top-right pixels to approximate DDX in a 2x2 subspan, unless the application requests a more accurate approximation via GL_FRAGMENT_SHADER_DERIVATIVE_HINT or this optimization is disabled from the new driconf option disable_derivative_optimization.