[HLSL] NonUniformResourceIndex implementation (llvm#160430)

Adds HLSL function `NonUniformResourceIndex` to `hlsl_intrinsics.h.` The function calls a builtin `__builtin_hlsl_resource_nonuniformindex` which gets translated to LLVM intrinsic `llvm.{dx|spv}.resource_nonuniformindex`.

Closes llvm#157923

Author: hekota

clang/include/clang/Basic/Builtins.td

@@ -4945,6 +4945,12 @@ def HLSLResourceHandleFromImplicitBinding : LangBuiltin<"HLSL_LANG"> {
   let Prototype = "void(...)";
 }
 
+def HLSLResourceNonUniformIndex : LangBuiltin<"HLSL_LANG"> {
+  let Spellings = ["__builtin_hlsl_resource_nonuniformindex"];
+  let Attributes = [NoThrow];
+  let Prototype = "uint32_t(uint32_t)";
+}
+
 def HLSLAll : LangBuiltin<"HLSL_LANG"> {
   let Spellings = ["__builtin_hlsl_all"];
   let Attributes = [NoThrow, Const];

clang/lib/CodeGen/CGHLSLBuiltins.cpp

@@ -352,6 +352,13 @@ Value *CodeGenFunction::EmitHLSLBuiltinExpr(unsigned BuiltinID,
     SmallVector<Value *> Args{OrderID, SpaceOp, RangeOp, IndexOp, Name};
     return Builder.CreateIntrinsic(HandleTy, IntrinsicID, Args);
   }
+  case Builtin::BI__builtin_hlsl_resource_nonuniformindex: {
+    Value *IndexOp = EmitScalarExpr(E->getArg(0));
+    llvm::Type *RetTy = ConvertType(E->getType());
+    return Builder.CreateIntrinsic(
+        RetTy, CGM.getHLSLRuntime().getNonUniformResourceIndexIntrinsic(),
+        ArrayRef<Value *>{IndexOp});
+  }
   case Builtin::BI__builtin_hlsl_all: {
     Value *Op0 = EmitScalarExpr(E->getArg(0));
     return Builder.CreateIntrinsic(

clang/lib/CodeGen/CGHLSLRuntime.h

@@ -129,6 +129,8 @@ class CGHLSLRuntime {
                                    resource_handlefrombinding)
   GENERATE_HLSL_INTRINSIC_FUNCTION(CreateHandleFromImplicitBinding,
                                    resource_handlefromimplicitbinding)
+  GENERATE_HLSL_INTRINSIC_FUNCTION(NonUniformResourceIndex,
+                                   resource_nonuniformindex)
   GENERATE_HLSL_INTRINSIC_FUNCTION(BufferUpdateCounter, resource_updatecounter)
   GENERATE_HLSL_INTRINSIC_FUNCTION(GroupMemoryBarrierWithGroupSync,
                                    group_memory_barrier_with_group_sync)

clang/lib/Headers/hlsl/hlsl_intrinsics.h

@@ -422,6 +422,30 @@ constexpr int4 D3DCOLORtoUBYTE4(float4 V) {
   return __detail::d3d_color_to_ubyte4_impl(V);
 }
 
+//===----------------------------------------------------------------------===//
+// NonUniformResourceIndex builtin
+//===----------------------------------------------------------------------===//
+
+/// \fn uint NonUniformResourceIndex(uint I)
+/// \brief A compiler hint to indicate that a resource index varies across
+/// threads within a wave (i.e., it is non-uniform).
+/// \param I [in] Resource array index
+///
+/// The return value is the \Index parameter.
+///
+/// When indexing into an array of shader resources (e.g., textures, buffers),
+/// some GPU hardware and drivers require the compiler to know whether the index
+/// is uniform (same for all threads) or non-uniform (varies per thread).
+///
+/// Using NonUniformResourceIndex explicitly marks an index as non-uniform,
+/// disabling certain assumptions or optimizations that could lead to incorrect
+/// behavior when dynamically accessing resource arrays with non-uniform
+/// indices.
+
+constexpr uint32_t NonUniformResourceIndex(uint32_t Index) {
+  return __builtin_hlsl_resource_nonuniformindex(Index);
+}
+
 //===----------------------------------------------------------------------===//
 // reflect builtin
 //===----------------------------------------------------------------------===//

clang/test/CodeGenHLSL/resources/NonUniformResourceIndex.hlsl

@@ -0,0 +1,38 @@
+// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.3-compute -emit-llvm -disable-llvm-passes -o - %s \
+// RUN:   | llvm-cxxfilt | FileCheck %s --check-prefixes=CHECK,DXIL
+// RUN: %clang_cc1 -finclude-default-header -triple spirv-pc-vulkan1.3-compute -emit-llvm -disable-llvm-passes -o - %s \
+// RUN:   | llvm-cxxfilt | FileCheck %s --check-prefixes=CHECK,SPV
+
+RWBuffer<float> A[10];
+
+[numthreads(4,1,1)]
+void main(uint GI : SV_GroupID) {
+  // CHECK: %[[GI:.*]] = load i32, ptr %GI.addr
+  // CHECK: %[[NURI_1:.*]] = call {{.*}} i32 @hlsl::NonUniformResourceIndex(unsigned int)(i32 noundef %[[GI]])
+  // CHECK: call void @hlsl::RWBuffer<float>::__createFromImplicitBinding(unsigned int, unsigned int, int, unsigned int, char const*)
+  // CHECK-SAME: (ptr {{.*}}, i32 noundef 0, i32 noundef 0, i32 noundef 10, i32 noundef %[[NURI_1]], ptr noundef @A.str)
+  float a = A[NonUniformResourceIndex(GI)][0];
+
+  // CHECK: %[[GI:.*]] = load i32, ptr %GI.addr
+  // CHECK: %[[ADD:.*]] = add i32 %[[GI]], 1
+  // CHECK: %[[NURI_2:.*]] = call {{.*}} i32 @hlsl::NonUniformResourceIndex(unsigned int)(i32 noundef %[[ADD]])
+  // CHECK: %[[MOD:.*]] = urem i32 %[[NURI_2]], 10
+  // CHECK: call void @hlsl::RWBuffer<float>::__createFromImplicitBinding(unsigned int, unsigned int, int, unsigned int, char const*)
+  // CHECK-SAME: (ptr {{.*}}, i32 noundef 0, i32 noundef 0, i32 noundef 10, i32 noundef %[[MOD]], ptr noundef @A.str)
+  float b = A[NonUniformResourceIndex(GI + 1) % 10][0];
+
+  // CHECK: %[[GI:.*]] = load i32, ptr %GI.addr
+  // CHECK: %[[NURI_3:.*]] = call {{.*}} i32 @hlsl::NonUniformResourceIndex(unsigned int)(i32 noundef %[[GI]])
+  // CHECK: %[[MUL:.*]] = mul i32 3, %[[NURI_3]]
+  // CHECK: %[[ADD2:.*]] = add i32 10, %[[MUL]]
+  // CHECK: call void @hlsl::RWBuffer<float>::__createFromImplicitBinding(unsigned int, unsigned int, int, unsigned int, char const*)
+  // CHECK-SAME: (ptr {{.*}}, i32 noundef 0, i32 noundef 0, i32 noundef 10, i32 noundef %[[ADD2]], ptr noundef @A.str)
+  float c = A[10 + 3 * NonUniformResourceIndex(GI)][0];
+  A[0][0] = a + b + c;
+}
+
+// CHECK: define {{.*}} i32 @hlsl::NonUniformResourceIndex(unsigned int)(i32 noundef %Index)
+// CHECK: %[[INDEX1:.*]] = load i32, ptr %Index.addr, align 4
+// DXIL: %[[INDEX2:.*]] = call i32 @llvm.dx.resource.nonuniformindex(i32 %[[INDEX1]])
+// SPV: %[[INDEX2:.*]] = call i32 @llvm.spv.resource.nonuniformindex(i32 %[[INDEX1]])
+// CHECK: ret i32 %[[INDEX2]]