Improve texture loads and stores on CUDA (#8644)

- fix handling layer and mip level
- add support for 1D layered textures
- reduce code by using macros
- assert when trying to emit unsupported intrinsics

There is a new set of unit tests in slang-rhi for exhaustive testing of
shader loads/stores on textures. These fixes allow to enable most of
these tests. Formatted loads/stores on surfaces are not supported in PTX
ISA, so this would require codegen for the conversion which in theory
should be possible but not as part of the CUDA prelude.

1 files changed, 19 insertions, 22 deletions

diff --git a/source/slang/hlsl.meta.slang b/source/slang/hlsl.meta.slang
index ff2762b23..26c2a43cb 100644
--- a/source/slang/hlsl.meta.slang
+++ b/source/slang/hlsl.meta.slang
@@ -3654,32 +3654,29 @@ extension _Texture<T,Shape,isArray,0,sampleCount,0,isShadow,isCombined,format>
             __intrinsic_asm ".Load";
 		case cuda:
             if (isArray != 0)
+            {
+                switch(Shape.flavor)
                 {
-     				static_assert(Shape.flavor == $(SLANG_TEXTURE_2D) || Shape.flavor == $(SLANG_TEXTURE_3D),
-                              "Integer coordinates are supported for texture reads only for 2D and 3D textures and 2D array textures.");
-
-                    if (Shape.flavor == $(SLANG_TEXTURE_2D))
-                    {
-						__intrinsic_asm "tex2DArrayfetch_int<$T0>($0, ($1).x, ($1).y, ($1).z)";
-				    }
-                    else
-					{
-                        __intrinsic_asm "<invalid intrinsic>";
-                    }
+                case $(SLANG_TEXTURE_1D):
+                    __intrinsic_asm "tex1DArrayfetch_int<$T0>($0, ($1).x, ($1).y, ($1).z)";
+                case $(SLANG_TEXTURE_2D):
+                    __intrinsic_asm "tex2DArrayfetch_int<$T0>($0, ($1).x, ($1).y, ($1).z, ($1).w)";
                 }
-                else
+            }
+            else
+            {
+                switch(Shape.flavor)
                 {
-                    switch(Shape.flavor)
-                    {
-                    case $(SLANG_TEXTURE_2D):
-                        __intrinsic_asm "tex2Dfetch_int<$T0>($0, ($1).x, ($1).y)";
-                    case $(SLANG_TEXTURE_3D):
-                        __intrinsic_asm "tex3Dfetch_int<$T0>($0, ($1).x, ($1).y, ($1).z)";
-					case $(SLANG_TEXTURE_CUBE):
-                    default:
-                        __intrinsic_asm "<invalid intrinsic>";
-                    }
+                case $(SLANG_TEXTURE_1D):
+                    __intrinsic_asm "tex1Dfetch_int<$T0>($0, ($1).x, ($1).y)";
+                case $(SLANG_TEXTURE_2D):
+                    __intrinsic_asm "tex2Dfetch_int<$T0>($0, ($1).x, ($1).y, ($1).z)";
+                case $(SLANG_TEXTURE_3D):
+                    __intrinsic_asm "tex3Dfetch_int<$T0>($0, ($1).x, ($1).y, ($1).z, ($1).w)";
                 }
+            }
+            static_assert(false, "Unsupported 'Load' of 'texture' for 'cuda' target");
+            __intrinsic_asm "<invalid intrinsic>";
         case metal:
             switch (Shape.flavor)
             {