[HLSL][SPIRV][DXIL] Implement dot4add_u8packed intrinsic

clang/include/clang/Basic/Builtins.td

@@ -4798,6 +4798,12 @@ def HLSLDot4AddI8Packed : LangBuiltin<"HLSL_LANG"> {
   let Prototype = "int(unsigned int, unsigned int, int)";
 }
 
+def HLSLDot4AddU8Packed : LangBuiltin<"HLSL_LANG"> {
+  let Spellings = ["__builtin_hlsl_dot4add_u8packed"];
+  let Attributes = [NoThrow, Const];
+  let Prototype = "unsigned int(unsigned int, unsigned int, unsigned int)";
+}
+
 def HLSLFirstBitHigh : LangBuiltin<"HLSL_LANG"> {
   let Spellings = ["__builtin_hlsl_elementwise_firstbithigh"];
   let Attributes = [NoThrow, Const];

clang/lib/CodeGen/CGBuiltin.cpp

@@ -18881,6 +18881,16 @@ Value *CodeGenFunction::EmitHLSLBuiltinExpr(unsigned BuiltinID,
         /*ReturnType=*/C->getType(), ID, ArrayRef<Value *>{A, B, C}, nullptr,
         "hlsl.dot4add.i8packed");
   }
+  case Builtin::BI__builtin_hlsl_dot4add_u8packed: {
+    Value *A = EmitScalarExpr(E->getArg(0));
+    Value *B = EmitScalarExpr(E->getArg(1));
+    Value *C = EmitScalarExpr(E->getArg(2));
+
+    Intrinsic::ID ID = CGM.getHLSLRuntime().getDot4AddU8PackedIntrinsic();
+    return Builder.CreateIntrinsic(
+        /*ReturnType=*/C->getType(), ID, ArrayRef<Value *>{A, B, C}, nullptr,
+        "hlsl.dot4add.u8packed");
+  }
   case Builtin::BI__builtin_hlsl_elementwise_firstbithigh: {
 
     Value *X = EmitScalarExpr(E->getArg(0));

clang/lib/CodeGen/CGHLSLRuntime.h

@@ -90,6 +90,7 @@ class CGHLSLRuntime {
   GENERATE_HLSL_INTRINSIC_FUNCTION(SDot, sdot)
   GENERATE_HLSL_INTRINSIC_FUNCTION(UDot, udot)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Dot4AddI8Packed, dot4add_i8packed)
+  GENERATE_HLSL_INTRINSIC_FUNCTION(Dot4AddU8Packed, dot4add_u8packed)
   GENERATE_HLSL_INTRINSIC_FUNCTION(WaveIsFirstLane, wave_is_first_lane)
   GENERATE_HLSL_INTRINSIC_FUNCTION(WaveReadLaneAt, wave_readlane)
   GENERATE_HLSL_INTRINSIC_FUNCTION(FirstBitUHigh, firstbituhigh)

clang/lib/Headers/hlsl/hlsl_intrinsics.h

@@ -942,7 +942,13 @@ uint64_t dot(uint64_t4, uint64_t4);
 
 _HLSL_AVAILABILITY(shadermodel, 6.4)
 _HLSL_BUILTIN_ALIAS(__builtin_hlsl_dot4add_i8packed)
-int dot4add_i8packed(unsigned int, unsigned int, int);
+int dot4add_i8packed(uint, uint, int);
+
+/// \fn uint dot4add_u8packed(uint A, uint B, uint C)
+
+_HLSL_AVAILABILITY(shadermodel, 6.4)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_dot4add_u8packed)
+uint dot4add_u8packed(uint, uint, uint);
 
 //===----------------------------------------------------------------------===//
 // exp builtins

clang/test/CodeGenHLSL/builtins/dot4add_u8packed.hlsl

@@ -0,0 +1,18 @@
+
+// RUN: %clang_cc1 -finclude-default-header -triple \
+// RUN:   dxil-pc-shadermodel6.3-compute %s -emit-llvm -disable-llvm-passes -o - | \
+// RUN:   FileCheck %s -DTARGET=dx
+// RUN: %clang_cc1 -finclude-default-header -triple \
+// RUN:   spirv-pc-vulkan-compute %s -emit-llvm -disable-llvm-passes -o - | \
+// RUN:   FileCheck %s -DTARGET=spv
+
+// Test basic lowering to runtime function call.
+
+// CHECK-LABEL: define {{.*}}test
+uint test(uint a, uint b, uint c) {
+  // CHECK:  %[[RET:.*]] = call [[TY:i32]] @llvm.[[TARGET]].dot4add.u8packed([[TY]] %[[#]], [[TY]] %[[#]], [[TY]] %[[#]])
+  // CHECK:  ret [[TY]] %[[RET]]
+  return dot4add_u8packed(a, b, c);
+}
+
+// CHECK: declare [[TY]] @llvm.[[TARGET]].dot4add.u8packed([[TY]], [[TY]], [[TY]])

clang/test/SemaHLSL/BuiltIns/dot4add_u8packed-errors.hlsl

@@ -0,0 +1,28 @@
+// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -emit-llvm-only -disable-llvm-passes -verify
+
+int test_too_few_arg0() {
+  return __builtin_hlsl_dot4add_u8packed();
+  // expected-error@-1 {{too few arguments to function call, expected 3, have 0}}
+}
+
+int test_too_few_arg1(int p0) {
+  return __builtin_hlsl_dot4add_u8packed(p0);
+  // expected-error@-1 {{too few arguments to function call, expected 3, have 1}}
+}
+
+int test_too_few_arg2(uint p0) {
+  return __builtin_hlsl_dot4add_u8packed(p0, p0);
+  // expected-error@-1 {{too few arguments to function call, expected 3, have 2}}
+}
+
+int test_too_many_arg(uint p0) {
+  return __builtin_hlsl_dot4add_u8packed(p0, p0, p0, p0);
+  // expected-error@-1 {{too many arguments to function call, expected 3, have 4}}
+}
+
+struct S { float f; };
+
+int test_expr_struct_type_check(S p0, uint p1) {
+  return __builtin_hlsl_dot4add_u8packed(p1, p1, p0);
+  // expected-error@-1 {{no viable conversion from 'S' to 'unsigned int'}}
+}

llvm/include/llvm/IR/IntrinsicsDirectX.td

@@ -69,7 +69,8 @@ def int_dx_udot :
     DefaultAttrsIntrinsic<[LLVMVectorElementType<0>],
     [llvm_anyint_ty, LLVMScalarOrSameVectorWidth<0, LLVMVectorElementType<0>>],
     [IntrNoMem, Commutative] >;
-  def int_dx_dot4add_i8packed : DefaultAttrsIntrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty], [IntrNoMem]>;
+def int_dx_dot4add_i8packed : DefaultAttrsIntrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty], [IntrNoMem]>;
+def int_dx_dot4add_u8packed : DefaultAttrsIntrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty], [IntrNoMem]>;
 
 def int_dx_frac  : DefaultAttrsIntrinsic<[llvm_anyfloat_ty], [LLVMMatchType<0>], [IntrNoMem]>;
 def int_dx_degrees : DefaultAttrsIntrinsic<[LLVMMatchType<0>], [llvm_anyfloat_ty], [IntrNoMem]>;

llvm/include/llvm/IR/IntrinsicsSPIRV.td

@@ -84,6 +84,7 @@ let TargetPrefix = "spv" in {
     [llvm_anyint_ty, LLVMScalarOrSameVectorWidth<0, LLVMVectorElementType<0>>],
     [IntrNoMem, Commutative] >;
   def int_spv_dot4add_i8packed : DefaultAttrsIntrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty], [IntrNoMem]>;
+  def int_spv_dot4add_u8packed : DefaultAttrsIntrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty], [IntrNoMem]>;
   def int_spv_wave_is_first_lane : DefaultAttrsIntrinsic<[llvm_i1_ty], [], [IntrConvergent]>;
   def int_spv_wave_readlane : DefaultAttrsIntrinsic<[llvm_any_ty], [LLVMMatchType<0>, llvm_i32_ty], [IntrConvergent, IntrNoMem]>;
   def int_spv_sign : DefaultAttrsIntrinsic<[LLVMScalarOrSameVectorWidth<0, llvm_i32_ty>], [llvm_any_ty], [IntrNoMem]>;

llvm/lib/Target/DirectX/DXIL.td

@@ -822,6 +822,16 @@ def Dot4AddI8Packed : DXILOp<163, dot4AddPacked> {
   let stages = [Stages<DXIL1_0, [all_stages]>];
 }
 
+def Dot4AddU8Packed : DXILOp<164, dot4AddPacked> {
+  let Doc = "unsigned dot product of 4 x i8 vectors packed into i32, with "
+            "accumulate to i32";
+  let LLVMIntrinsic = int_dx_dot4add_u8packed;
+  let arguments = [Int32Ty, Int32Ty, Int32Ty];
+  let result = Int32Ty;
+  let attributes = [Attributes<DXIL1_0, [ReadNone]>];
+  let stages = [Stages<DXIL1_0, [all_stages]>];
+}
+
 def AnnotateHandle : DXILOp<216, annotateHandle> {
   let Doc = "annotate handle with resource properties";
   let arguments = [HandleTy, ResPropsTy];

llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp

@@ -2687,6 +2687,11 @@ bool SPIRVInstructionSelector::selectIntrinsic(Register ResVReg,
         STI.isAtLeastSPIRVVer(VersionTuple(1, 6)))
       return selectDot4AddPacked<true>(ResVReg, ResType, I);
     return selectDot4AddPackedExpansion<true>(ResVReg, ResType, I);
+  case Intrinsic::spv_dot4add_u8packed:
+    if (STI.canUseExtension(SPIRV::Extension::SPV_KHR_integer_dot_product) ||
+        STI.isAtLeastSPIRVVer(VersionTuple(1, 6)))
+      return selectDot4AddPacked<false>(ResVReg, ResType, I);
+    return selectDot4AddPackedExpansion<false>(ResVReg, ResType, I);
   case Intrinsic::spv_all:
     return selectAll(ResVReg, ResType, I);
   case Intrinsic::spv_any:

llvm/test/CodeGen/DirectX/dot4add_u8packed.ll

@@ -0,0 +1,10 @@
+; RUN: opt -S -dxil-op-lower -mtriple=dxil-pc-shadermodel6.3-compute %s | FileCheck %s
+
+define void @main(i32 %a, i32 %b, i32 %c) {
+entry:
+; CHECK: call i32 @dx.op.dot4AddPacked(i32 164, i32 %a, i32 %b, i32 %c)
+  %0 = call i32 @llvm.dx.dot4add.u8packed(i32 %a, i32 %b, i32 %c)
+  ret void
+}
+
+declare i32 @llvm.dx.dot4add.u8packed(i32, i32, i32)

llvm/test/CodeGen/SPIRV/hlsl-intrinsics/dot4add_u8packed.ll

@@ -0,0 +1,65 @@
+; RUN: llc -O0 -mtriple=spirv1.5-unknown-unknown %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-EXP
+; RUN: llc -O0 -mtriple=spirv1.6-unknown-unknown %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-DOT
+; RUN: llc -O0 -mtriple=spirv-unknown-unknown -spirv-ext=+SPV_KHR_integer_dot_product %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-DOT,CHECK-EXT
+; RUN: %if spirv-tools %{ llc -verify-machineinstrs -O0 -mtriple=spirv1.5-unknown-unknown %s -o - -filetype=obj | spirv-val %}
+; RUN: %if spirv-tools %{ llc -verify-machineinstrs -O0 -mtriple=spirv1.6-unknown-unknown %s -o - -filetype=obj | spirv-val %}
+; RUN: %if spirv-tools %{ llc -verify-machineinstrs -O0 -mtriple=spirv-unknown-unknown -spirv-ext=+SPV_KHR_integer_dot_product %s -o - -filetype=obj | spirv-val %}
+
+; CHECK-DOT: OpCapability DotProduct
+; CHECK-DOT: OpCapability DotProductInput4x8BitPacked
+; CHECK-EXT: OpExtension "SPV_KHR_integer_dot_product"
+
+; CHECK: %[[#int_32:]] = OpTypeInt 32 0
+; CHECK-EXP-DAG: %[[#int_8:]] = OpTypeInt 8 0
+; CHECK-EXP-DAG: %[[#zero:]] = OpConstantNull %[[#int_8]]
+; CHECK-EXP-DAG: %[[#eight:]] = OpConstant %[[#int_8]] 8
+; CHECK-EXP-DAG: %[[#sixteen:]] = OpConstant %[[#int_8]] 16
+; CHECK-EXP-DAG: %[[#twentyfour:]] = OpConstant %[[#int_8]] 24
+
+; CHECK-LABEL: Begin function test_dot
+define noundef i32 @test_dot(i32 noundef %a, i32 noundef %b, i32 noundef %c) {
+entry:
+; CHECK: %[[#A:]] = OpFunctionParameter %[[#int_32]]
+; CHECK: %[[#B:]] = OpFunctionParameter %[[#int_32]]
+; CHECK: %[[#C:]] = OpFunctionParameter %[[#int_32]]
+
+; Test that we use the dot product op when capabilities allow
+
+; CHECK-DOT: %[[#DOT:]] = OpUDot %[[#int_32]] %[[#A]] %[[#B]]
+; CHECK-DOT: %[[#RES:]] = OpIAdd %[[#int_32]] %[[#DOT]] %[[#C]]
+
+; Test expansion is used when spirv dot product capabilities aren't available:
+
+; First element of the packed vector
+; CHECK-EXP: %[[#A0:]] = OpBitFieldUExtract %[[#int_32]] %[[#A]] %[[#zero]] %[[#eight]]
+; CHECK-EXP: %[[#B0:]] = OpBitFieldUExtract %[[#int_32]] %[[#B]] %[[#zero]] %[[#eight]]
+; CHECK-EXP: %[[#MUL0:]] = OpIMul %[[#int_32]] %[[#A0]] %[[#B0]]
+; CHECK-EXP: %[[#MASK0:]] = OpBitFieldUExtract %[[#int_32]] %[[#MUL0]] %[[#zero]] %[[#eight]]
+; CHECK-EXP: %[[#ACC0:]] = OpIAdd %[[#int_32]] %[[#C]] %[[#MASK0]]
+
+; Second element of the packed vector
+; CHECK-EXP: %[[#A1:]] = OpBitFieldUExtract %[[#int_32]] %[[#A]] %[[#eight]] %[[#eight]]
+; CHECK-EXP: %[[#B1:]] = OpBitFieldUExtract %[[#int_32]] %[[#B]] %[[#eight]] %[[#eight]]
+; CHECK-EXP: %[[#MUL1:]] = OpIMul %[[#int_32]] %[[#A1]] %[[#B1]]
+; CHECK-EXP: %[[#MASK1:]] = OpBitFieldUExtract %[[#int_32]] %[[#MUL1]] %[[#zero]] %[[#eight]]
+; CHECK-EXP: %[[#ACC1:]] = OpIAdd %[[#int_32]] %[[#ACC0]] %[[#MASK1]]
+
+; Third element of the packed vector
+; CHECK-EXP: %[[#A2:]] = OpBitFieldUExtract %[[#int_32]] %[[#A]] %[[#sixteen]] %[[#eight]]
+; CHECK-EXP: %[[#B2:]] = OpBitFieldUExtract %[[#int_32]] %[[#B]] %[[#sixteen]] %[[#eight]]
+; CHECK-EXP: %[[#MUL2:]] = OpIMul %[[#int_32]] %[[#A2]] %[[#B2]]
+; CHECK-EXP: %[[#MASK2:]] = OpBitFieldUExtract %[[#int_32]] %[[#MUL2]] %[[#zero]] %[[#eight]]
+; CHECK-EXP: %[[#ACC2:]] = OpIAdd %[[#int_32]] %[[#ACC1]] %[[#MASK2]]
+
+; Fourth element of the packed vector
+; CHECK-EXP: %[[#A3:]] = OpBitFieldUExtract %[[#int_32]] %[[#A]] %[[#twentyfour]] %[[#eight]]
+; CHECK-EXP: %[[#B3:]] = OpBitFieldUExtract %[[#int_32]] %[[#B]] %[[#twentyfour]] %[[#eight]]
+; CHECK-EXP: %[[#MUL3:]] = OpIMul %[[#int_32]] %[[#A3]] %[[#B3]]
+; CHECK-EXP: %[[#MASK3:]] = OpBitFieldUExtract %[[#int_32]] %[[#MUL3]] %[[#zero]] %[[#eight]]
+
+; CHECK-EXP: %[[#RES:]] = OpIAdd %[[#int_32]] %[[#ACC2]] %[[#MASK3]]
+; CHECK: OpReturnValue %[[#RES]]
+  %spv.dot = call i32 @llvm.spv.dot4add.u8packed(i32 %a, i32 %b, i32 %c)
+
+  ret i32 %spv.dot
+}