vkd3d-shader/hlsl: Flatten conditional branches containing stores.

For an if block

    if (cond)
    {
        <then_block>
    }
    else
    {
        <else_block>
    }

We flatten it by first replacing any store instruction `v[[k]] = x`
in the then_block with the following:

    1: load(v[[k]])
    2: cond ? x : @1
    3: v[[k]] = @2

Similarly, we replace any store instruction `v[[k]] = x` in the
else_block with the following:

    1: load(v[[k]])
    2: cond ? @1 : x
    3: v[[k]] = @2

Then we can concatenate <then_block> and <else_block> together and
get rid of the if block.

tests/hlsl/cf-cond-types.shader_test

@@ -113,7 +113,7 @@ float4 main() : sv_target
     return 0;
 }
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float1 f;
 float4 main() : sv_target
 {
@@ -124,8 +124,8 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float4 -2.0 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (1.0, 1.0, 1.0, 1.0)
+draw quad
+probe (0, 0) f32(1.0, 1.0, 1.0, 1.0)
 uniform 0 float4 -0.0 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.0, 0.0, 0.0, 0.0)
+draw quad
+probe (0, 0) f32(0.0, 0.0, 0.0, 0.0)

tests/hlsl/conditional.shader_test

@@ -1,4 +1,4 @@
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float4 u;
 
 float4 main() : sv_target
@@ -11,13 +11,13 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float4 0.0 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.9, 0.8, 0.7, 0.6)
+draw quad
+probe (0, 0) f32(0.9, 0.8, 0.7, 0.6)
 uniform 0 float4 0.1 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.1, 0.2, 0.3, 0.4)
+draw quad
+probe (0, 0) f32(0.1, 0.2, 0.3, 0.4)
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float4 u;
 
 float4 main() : sv_target
@@ -29,7 +29,7 @@ float4 main() : sv_target
         return float4(0.9, 0.8, 0.7, 0.6);
 }
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float4 u;
 
 float4 main() : sv_target
@@ -43,8 +43,8 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float4 0.0 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.9, 0.8, 0.7, 0.6)
+draw quad
+probe (0, 0) f32(0.9, 0.8, 0.7, 0.6)
 
 [pixel shader fail(sm<6)]
 float4 u;
@@ -370,7 +370,7 @@ todo(msl & sm>=6) draw quad
 probe (0, 0) rgba (1.0, 1.0, 0.0, 4.0)
 
 % Branches with resource stores can't be flattened.
-[pixel shader fail(sm<6) todo]
+[pixel shader fail(sm<6)]
 uniform float a;
 RWBuffer<float> u;
 

tests/hlsl/discard.shader_test

@@ -16,7 +16,7 @@ uniform 0 float4 9 8 7 6
 todo(sm<4 | glsl | msl & sm>=6) draw quad
 probe (0, 0) f32(1, 2, 3, 4)
 
-[pixel shader todo(sm<4)]
+[pixel shader todo]
 uniform float4 x;
 
 float4 main() : sv_target
@@ -29,10 +29,10 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float4 1 2 3 4
-todo(sm<4 | glsl | msl & sm>=6) draw quad
+todo(sm<6 | msl) draw quad
 probe (0, 0) f32(1, 2, 3, 4)
 uniform 0 float4 9 8 7 6
-todo(sm<4 | glsl | msl & sm>=6) draw quad
+todo(sm<6 | msl) draw quad
 probe (0, 0) f32(1, 2, 3, 4)
 
 [pixel shader todo(sm<4)]
@@ -56,7 +56,7 @@ uniform 0 float4 9 8 7 6
 todo(sm<4 | glsl | msl & sm>=6) draw quad
 probe (0, 0) f32(1, 2, 3, 4)
 
-[pixel shader todo(sm<4)]
+[pixel shader todo]
 uniform float4 x;
 
 float4 main() : sv_target
@@ -72,10 +72,10 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float4 1 2 3 4
-todo(sm<4 | glsl | msl & sm>=6) draw quad
+todo(sm<6 | msl) draw quad
 probe (0, 0) f32(1, 2, 3, 4)
 uniform 0 float4 9 8 7 6
-todo(sm<4 | glsl | msl & sm>=6) draw quad
+todo(sm<6 | msl) draw quad
 probe (0, 0) f32(1, 2, 3, 4)
 
 [require]

tests/hlsl/for.shader_test

@@ -5,7 +5,7 @@ void main(float4 pos : position, out float tex : texcoord, out float4 out_pos :
     out_pos = pos;
 }
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 float4 main(float tex : texcoord) : sv_target
 {
     int i;
@@ -23,10 +23,10 @@ float4 main(float tex : texcoord) : sv_target
 }
 
 [test]
-todo(sm<4) draw quad
-probe (  0, 0, 159, 480) rgba (10.0, 35.0, 0.0, 0.0)
-probe (161, 0, 479, 480) rgba (10.0, 38.0, 0.0, 0.0)
-probe (481, 0, 640, 480) rgba ( 5.0, 10.0, 0.0, 0.0)
+draw quad
+probe (  0, 0, 159, 480) f32(10.0, 35.0, 0.0, 0.0)
+probe (161, 0, 479, 480) f32(10.0, 38.0, 0.0, 0.0)
+probe (481, 0, 640, 480) f32( 5.0, 10.0, 0.0, 0.0)
 
 [require]
 shader model >= 4.0

tests/hlsl/function-return.shader_test

@@ -32,7 +32,7 @@ float4 main() : sv_target
 draw quad
 probe (0, 0) rgba (0.2, 0.1, 0.8, 0.5);
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 float func(out float o)
@@ -79,19 +79,19 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 0.1
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.2, 0.6, 0.3) 1
+draw quad
+probe (0, 0) f32(0.3, 0.2, 0.6, 0.3) 1
 uniform 0 float 0.4
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.6, 0.5, 0.6, 0.3) 1
+draw quad
+probe (0, 0) f32(0.6, 0.5, 0.6, 0.3) 1
 uniform 0 float 0.6
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.6, 0.5, 0.4, 0.5) 1
+draw quad
+probe (0, 0) f32(0.6, 0.5, 0.4, 0.5) 1
 uniform 0 float 0.8
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.8, 0.7, 0.4, 0.5) 1
+draw quad
+probe (0, 0) f32(0.8, 0.7, 0.4, 0.5) 1
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 float func(out float o)
@@ -134,14 +134,14 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 0.1
-todo(sm<4 | msl & sm>=6) draw quad
-probe (0, 0) rgba (0.2, 0.1, 0.2, 0.1) 1
+todo(msl & sm>=6) draw quad
+probe (0, 0) f32(0.2, 0.1, 0.2, 0.1) 1
 uniform 0 float 0.5
-todo(sm<4 | msl & sm>=6) draw quad
-probe (0, 0) rgba (0.5, 0.4, 1.0, 0.9) 1
+todo(msl & sm>=6) draw quad
+probe (0, 0) f32(0.5, 0.4, 1.0, 0.9) 1
 uniform 0 float 0.9
-todo(sm<4 | msl & sm>=6) draw quad
-probe (0, 0) rgba (1.0, 0.9, 1.0, 0.6) 1
+todo(msl & sm>=6) draw quad
+probe (0, 0) f32(1.0, 0.9, 1.0, 0.6) 1
 
 [pixel shader]
 float func(out float o)
@@ -184,7 +184,7 @@ float4 main() : sv_target
 draw quad
 probe (0, 0) rgba (0.4, 0.3, 0.3, 0.9) 1
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 float func(out float o)
@@ -235,26 +235,26 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.2, 0.3, 0.3) 1
+draw quad
+probe (0, 0) f32(0.3, 0.2, 0.3, 0.3) 1
 
 uniform 0 float 0.1
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.3, 0.3, 0.3) 1
+draw quad
+probe (0, 0) f32(0.3, 0.3, 0.3, 0.3) 1
 
 uniform 0 float 0.3
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.5, 0.3, 0.3) 1
+draw quad
+probe (0, 0) f32(0.3, 0.5, 0.3, 0.3) 1
 
 uniform 0 float 0.7
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.9, 0.7, 0.6) 1
+draw quad
+probe (0, 0) f32(0.3, 0.9, 0.7, 0.6) 1
 
 uniform 0 float 0.9
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.4, 0.1, 0.7, 0.6) 1
+draw quad
+probe (0, 0) f32(0.4, 0.1, 0.7, 0.6) 1
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 
 uniform float4 f[3];
 
@@ -291,24 +291,24 @@ float4 main() : sv_target
 uniform 0 float4 0.3 0.0 0.0 0.0
 uniform 4 float4 0.0 0.0 0.0 0.0
 uniform 8 float4 0.1 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.2, 0.6, 0.6) 1
+draw quad
+probe (0, 0) f32(0.3, 0.2, 0.6, 0.6) 1
 
 uniform 4 float4 0.35 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.3, 0.6, 0.6) 1
+draw quad
+probe (0, 0) f32(0.3, 0.3, 0.6, 0.6) 1
 
 uniform 8 float4 0.5 0.0 0.0 0.0
-todo(sm<4) draw quad
+draw quad
 % Apparently fixed by Mesa commit 9785fa460c41b9498c24a82b98069655a91224c5,
 % which only adds optimization; so it might just be hiding the bug rather than
 % solving it; see also https://gitlab.freedesktop.org/mesa/mesa/-/issues/13092
 bug(mesa<25.1 & llvm>=16) probe (0, 0) f32(0.3, 0.5, 0.6, 0.6) 1
 
 uniform 0 float4 1.0 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.5, 0.6, 0.6) 1
+draw quad
+probe (0, 0) f32(0.3, 0.5, 0.6, 0.6) 1
 
 uniform 4 float4 2.0 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.4, 0.1, 0.6, 0.6) 1
+draw quad
+probe (0, 0) f32(0.4, 0.1, 0.6, 0.6) 1

tests/hlsl/hard-copy-prop.shader_test

@@ -1,4 +1,4 @@
-[pixel shader todo(sm<4)]
+[pixel shader]
 float cond;
 
 float4 main() : sv_target
@@ -17,14 +17,13 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (2.0, 2.0, 2.0, 2.0)
+draw quad
+probe (0, 0) f32(2.0, 2.0, 2.0, 2.0)
 uniform 0 float 1.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (-2.0, -2.0, -2.0, -2.0)
+draw quad
+probe (0, 0) f32(-2.0, -2.0, -2.0, -2.0)
 
-
-[pixel shader todo(sm<4)]
+[pixel shader]
 float cond;
 
 float4 main() : sv_target
@@ -43,14 +42,13 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (2.0, 2.0, 2.0, 2.0)
+draw quad
+probe (0, 0) f32(2.0, 2.0, 2.0, 2.0)
 uniform 0 float 1.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (20.0, 20.0, 20.0, 20.0)
+draw quad
+probe (0, 0) f32(20.0, 20.0, 20.0, 20.0)
 
-
-[pixel shader todo(sm<4)]
+[pixel shader]
 float cond;
 
 float4 main() : sv_target
@@ -69,11 +67,11 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (1.0, 4.0, 0.0, 0.0)
+draw quad
+probe (0, 0) f32(1.0, 4.0, 0.0, 0.0)
 uniform 0 float 1.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (1.0, 40.0, 0.0, 0.0)
+draw quad
+probe (0, 0) f32(1.0, 40.0, 0.0, 0.0)
 
 [pixel shader]
 float x[6];
@@ -206,7 +204,7 @@ float4 main() : sv_target
     return 0;
 }
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 float a;
 
 float4 main() : sv_target
@@ -221,8 +219,8 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 11.0
-todo(sm<4) draw quad
-probe (0, 0) rgba(1, 2, 3, 5)
+draw quad
+probe (0, 0) f32(1, 2, 3, 5)
 
 [require]
 shader model < 5.1

tests/hlsl/return.shader_test

@@ -25,7 +25,7 @@ void main(out float4 ret : sv_target)
 draw quad
 probe (0, 0) rgba (0.1, 0.2, 0.3, 0.4)
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 float4 main() : sv_target
@@ -37,13 +37,13 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 0.2
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.1, 0.2, 0.3, 0.4)
+draw quad
+probe (0, 0) f32(0.1, 0.2, 0.3, 0.4)
 uniform 0 float 0.8
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.5, 0.6, 0.7, 0.8)
+draw quad
+probe (0, 0) f32(0.5, 0.6, 0.7, 0.8)
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 void main(out float4 ret : sv_target)
@@ -63,13 +63,13 @@ void main(out float4 ret : sv_target)
 
 [test]
 uniform 0 float 0.2
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.3, 0.4, 0.5, 0.6)
+draw quad
+probe (0, 0) f32(0.3, 0.4, 0.5, 0.6)
 uniform 0 float 0.8
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.1, 0.2, 0.3, 0.4)
+draw quad
+probe (0, 0) f32(0.1, 0.2, 0.3, 0.4)
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 void main(out float4 ret : sv_target)
@@ -89,16 +89,16 @@ void main(out float4 ret : sv_target)
 
 [test]
 uniform 0 float 0.1
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.1, 0.2, 0.3, 0.4) 1
+draw quad
+probe (0, 0) f32(0.1, 0.2, 0.3, 0.4) 1
 uniform 0 float 0.5
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.2, 0.3, 0.4, 0.5) 1
+draw quad
+probe (0, 0) f32(0.2, 0.3, 0.4, 0.5) 1
 uniform 0 float 0.9
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.5, 0.6, 0.7, 0.8) 1
+draw quad
+probe (0, 0) f32(0.5, 0.6, 0.7, 0.8) 1
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 void main(out float4 ret : sv_target)
@@ -115,14 +115,14 @@ void main(out float4 ret : sv_target)
 
 [test]
 uniform 0 float 0.1
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.1, 0.2, 0.3, 0.4) 1
+draw quad
+probe (0, 0) f32(0.1, 0.2, 0.3, 0.4) 1
 uniform 0 float 0.5
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.5, 0.6, 0.7, 0.8) 1
+draw quad
+probe (0, 0) f32(0.5, 0.6, 0.7, 0.8) 1
 uniform 0 float 0.9
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.4, 0.5, 0.6, 0.7) 1
+draw quad
+probe (0, 0) f32(0.4, 0.5, 0.6, 0.7) 1
 
 [pixel shader]
 void main(out float4 ret : sv_target)
@@ -141,7 +141,7 @@ void main(out float4 ret : sv_target)
 draw quad
 probe (0, 0) rgba (0.2, 0.4, 0.6, 0.8)
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 void main(out float4 ret : sv_target)
@@ -160,26 +160,26 @@ void main(out float4 ret : sv_target)
 
 [test]
 uniform 0 float 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.1, 0.1, 0.1, 0.1) 1
+draw quad
+probe (0, 0) f32(0.1, 0.1, 0.1, 0.1) 1
 
 uniform 0 float 0.1
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.2, 0.2, 0.2, 0.2) 1
+draw quad
+probe (0, 0) f32(0.2, 0.2, 0.2, 0.2) 1
 
 uniform 0 float 0.3
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.4, 0.4, 0.4, 0.4) 1
+draw quad
+probe (0, 0) f32(0.4, 0.4, 0.4, 0.4) 1
 
 uniform 0 float 0.7
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.8, 0.8, 0.8, 0.8) 1
+draw quad
+probe (0, 0) f32(0.8, 0.8, 0.8, 0.8) 1
 
 uniform 0 float 0.9
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.9, 0.9, 0.9, 0.9) 1
+draw quad
+probe (0, 0) f32(0.9, 0.9, 0.9, 0.9) 1
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 uniform float f;
 
 void main(out float4 ret : sv_target)
@@ -204,13 +204,13 @@ void main(out float4 ret : sv_target)
 
 [test]
 uniform 0 float 0.2
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.2, 0.2, 0.2, 0.2)
+draw quad
+probe (0, 0) f32(0.2, 0.2, 0.2, 0.2)
 uniform 0 float 0.8
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.5, 0.5, 0.5, 0.5)
+draw quad
+probe (0, 0) f32(0.5, 0.5, 0.5, 0.5)
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 
 uniform float4 f[3];
 
@@ -236,23 +236,23 @@ void main(out float4 ret : sv_target)
 uniform 0 float4 0.3 0.0 0.0 0.0
 uniform 4 float4 0.0 0.0 0.0 0.0
 uniform 8 float4 0.1 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.1, 0.1, 0.1, 0.1) 1
+draw quad
+probe (0, 0) f32(0.1, 0.1, 0.1, 0.1) 1
 
 uniform 4 float4 0.35 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.2, 0.2, 0.2, 0.2) 1
+draw quad
+probe (0, 0) f32(0.2, 0.2, 0.2, 0.2) 1
 
 uniform 8 float4 0.5 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.4, 0.4, 0.4, 0.4) 1
+draw quad
+probe (0, 0) f32(0.4, 0.4, 0.4, 0.4) 1
 
 uniform 0 float4 1.0 0.0 0.0 0.0
-todo(sm<4) draw quad
-probe (0, 0) rgba (0.4, 0.4, 0.4, 0.4) 1
+draw quad
+probe (0, 0) f32(0.4, 0.4, 0.4, 0.4) 1
 
 uniform 4 float4 2.0 0.0 0.0 0.0
-todo(sm<4) draw quad
+draw quad
 % Apparently fixed by Mesa commit 9785fa460c41b9498c24a82b98069655a91224c5,
 % which only adds optimization; so it might just be hiding the bug rather than
 % solving it; see also https://gitlab.freedesktop.org/mesa/mesa/-/issues/13092

tests/hlsl/unroll.shader_test

@@ -1,6 +1,6 @@
 % Unrolling edge cases.
 
-[pixel shader todo(sm<4)]
+[pixel shader]
 float a : register(c0);
 
 float4 main() : sv_target
@@ -19,8 +19,8 @@ float4 main() : sv_target
 
 [test]
 uniform 0 float 1
-todo(sm<4) draw quad
-probe (0,0) rgba(10.0, 1.0, 2.0, 3.0)
+draw quad
+probe (0,0) f32(10.0, 1.0, 2.0, 3.0)
 
 [pixel shader]
 float4 main() : sv_target