compiler: Add lowering support for 64-bit saturate operations to software

Fixes 7 Khronos GL CTS tests:
KHR-GL45.gpu_shader_fp64.builtin.smoothstep_dvec{double, 2, 3, 4}
KHR-GL45.gpu_shader_fp64.builtin.smoothstep_against_scalar_dvec{2, 3, 4}

Suggested-by: Jason Ekstrand <jason@jlekstrand.net>
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>

diff --git a/src/compiler/glsl/float64.glsl b/src/compiler/glsl/float64.glsl
index 415dde3..babfad2 100644 (file)
--- a/src/compiler/glsl/float64.glsl
+++ b/src/compiler/glsl/float64.glsl
@@ -218,6 +218,18 @@ __fge64(uint64_t a, uint64_t b)
    return !__flt64_nonnan(a, b);
 }
 
+uint64_t
+__fsat64(uint64_t __a)
+{
+   if (__flt64(__a, 0ul))
+      return 0ul;
+
+   if (__fge64(__a, 0x3FF0000000000000ul /* 1.0 */))
+      return 0x3FF0000000000000ul;
+
+   return __a;
+}
+
 /* Adds the 64-bit value formed by concatenating `a0' and `a1' to the 64-bit
  * value formed by concatenating `b0' and `b1'.  Addition is modulo 2^64, so
  * any carry out is lost.  The result is broken into two 32-bit pieces which

diff --git a/src/compiler/nir/nir_lower_double_ops.c b/src/compiler/nir/nir_lower_double_ops.c
index 18fe08c..04ec2a8 100644 (file)
--- a/src/compiler/nir/nir_lower_double_ops.c
+++ b/src/compiler/nir/nir_lower_double_ops.c
@@ -550,6 +550,9 @@ lower_doubles_instr_to_soft(nir_builder *b, nir_alu_instr *instr,
    case nir_op_ffma:
       name = "__ffma64";
       break;
+   case nir_op_fsat:
+      name = "__fsat64";
+      break;
    default:
       return false;
    }