[Piglit] [PATCH 1/1] cl: Fix tolerance for division in scalar arithmetic test

[Tom Stellard]

On Mon, Jan 26, 2015 at 03:20:34PM -0500, Jan Vesely wrote:
> Signed-off-by: Jan Vesely <jan.vesely at rutgers.edu>

Reviewed-by: Tom Stellard <thomas.stellard at amd.com>

> ---
> RFC: I only tagged numerical results,
>  it did not make sense to me to do it for NaNs and Infs
>

I think this is fine.

> R600: Fail -> Pass
> 
>  tests/cl/program/execute/scalar-arithmetic-float.cl | 14 +++++++-------
>  1 file changed, 7 insertions(+), 7 deletions(-)
> 
> diff --git a/tests/cl/program/execute/scalar-arithmetic-float.cl b/tests/cl/program/execute/scalar-arithmetic-float.cl
> index 33a44d4..84c3970 100644
> --- a/tests/cl/program/execute/scalar-arithmetic-float.cl
> +++ b/tests/cl/program/execute/scalar-arithmetic-float.cl
> @@ -178,42 +178,42 @@ name: pos div pos
>  kernel_name: div
>  arg_in:  1 float 8.5
>  arg_in:  2 float 4.25
> -arg_out: 0 buffer float[1] 2
> +arg_out: 0 buffer float[1] 2 tolerance 2
>  
>  [test]
>  name: pos div neg
>  kernel_name: div
>  arg_in:  1 float 11.25
>  arg_in:  2 float -3
> -arg_out: 0 buffer float[1] -3.75
> +arg_out: 0 buffer float[1] -3.75 tolerance 2
>  
>  [test]
>  name: neg div pos
>  kernel_name: div
>  arg_in:  1 float -21
>  arg_in:  2 float 5.25
> -arg_out: 0 buffer float[1] -4
> +arg_out: 0 buffer float[1] -4 tolerance 2
>  
>  [test]
>  name: neg div neg
>  kernel_name: div
>  arg_in:  1 float -21.25
>  arg_in:  2 float -5
> -arg_out: 0 buffer float[1] 4.25
> +arg_out: 0 buffer float[1] 4.25 tolerance 2
>  
>  [test]
>  name: 0 div pos
>  kernel_name: div
>  arg_in:  1 float 0
>  arg_in:  2 float 3.7
> -arg_out: 0 buffer float[1] 0
> +arg_out: 0 buffer float[1] 0 tolerance 2
>  
>  [test]
>  name: 0 div neg
>  kernel_name: div
>  arg_in:  1 float 0
>  arg_in:  2 float -3.7
> -arg_out: 0 buffer float[1] -0
> +arg_out: 0 buffer float[1] -0 tolerance 2
>  
>  [test]
>  name: num div 0
> @@ -234,7 +234,7 @@ name: 0 div inf
>  kernel_name: div
>  arg_in:  1 float 0
>  arg_in:  2 float inf
> -arg_out: 0 buffer float[1] 0
> +arg_out: 0 buffer float[1] 0 tolerance 2
>  
>  [test]
>  name: inf div 0
> -- 
> 2.1.0
> 
> _______________________________________________
> Piglit mailing list
> Piglit at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/piglit