[Mesa-dev] [PATCH 8/8] nir/spirv/glsl450: Implement IEEE-compliant handling of atan2(±∞, ±∞).

Ian Romanick idr at freedesktop.org
Wed Jan 25 02:03:08 UTC 2017


This appears to do the same thing as the GLSL change.  This patch is

Reviewed-by: Ian Romanick <ian.d.romanick at intel.com>


On 01/24/2017 03:26 PM, Francisco Jerez wrote:
> ---
>  src/compiler/spirv/vtn_glsl450.c | 22 +++++++++++++++++++++-
>  1 file changed, 21 insertions(+), 1 deletion(-)
> 
> diff --git a/src/compiler/spirv/vtn_glsl450.c b/src/compiler/spirv/vtn_glsl450.c
> index 508f218..7af2dad 100644
> --- a/src/compiler/spirv/vtn_glsl450.c
> +++ b/src/compiler/spirv/vtn_glsl450.c
> @@ -325,12 +325,32 @@ build_atan2(nir_builder *b, nir_ssa_def *y, nir_ssa_def *x)
>     nir_ssa_def *rcp_scaled_t = nir_frcp(b, nir_fmul(b, t, scale));
>     nir_ssa_def *s_over_t = nir_fmul(b, nir_fmul(b, s, scale), rcp_scaled_t);
>  
> +   /* For |x| = |y| assume tan = 1 even if infinite (i.e. pretend momentarily
> +    * that ∞/∞ = 1) in order to comply with the rather artificial rules
> +    * inherited from IEEE 754-2008, namely:
> +    *
> +    *  "atan2(±∞, −∞) is ±3π/4
> +    *   atan2(±∞, +∞) is ±π/4"
> +    *
> +    * Note that this is inconsistent with the rules for the neighborhood of
> +    * zero that are based on iterated limits:
> +    *
> +    *  "atan2(±0, −0) is ±π
> +    *   atan2(±0, +0) is ±0"
> +    *
> +    * but GLSL specifically allows implementations to deviate from IEEE rules
> +    * at (0,0), so we take that license (i.e. pretend that 0/0 = 1 here as
> +    * well).
> +    */
> +   nir_ssa_def *tan = nir_bcsel(b, nir_feq(b, nir_fabs(b, x), nir_fabs(b, y)),
> +                                one, nir_fabs(b, s_over_t));
> +
>     /* Calculate the arctangent and fix up the result if we had flipped the
>      * coordinate system.
>      */
>     nir_ssa_def *arc = nir_fadd(b, nir_fmul(b, nir_b2f(b, flip),
>                                             nir_imm_float(b, M_PI_2f)),
> -                               build_atan(b, nir_fabs(b, s_over_t)));
> +                               build_atan(b, tan));
>  
>     /* Rather convoluted calculation of the sign of the result.  When x < 0 we
>      * cannot use fsign because we need to be able to distinguish between
> 



More information about the mesa-dev mailing list