[SPIR-V] Add support for OpFMod intrinsic (#193172)

Add the `spv.fmod` intrinsic and lower it directly to `SPIRV::OpFMod`
covering scalar and vector cases

llvm/lib/Target/SPIRV/SPIRVBuiltins.cpp

diff --git a/llvm/lib/Target/SPIRV/SPIRVBuiltins.cpp b/llvm/lib/Target/SPIRV/SPIRVBuiltins.cpp
index 4b663d7..70644c4 100644
--- a/llvm/lib/Target/SPIRV/SPIRVBuiltins.cpp
+++ b/llvm/lib/Target/SPIRV/SPIRVBuiltins.cpp
@@ -2139,6 +2139,21 @@
   return true;
 }
 
+static bool generateArithmeticInst(const SPIRV::IncomingCall *Call,
+                                   MachineIRBuilder &MIRBuilder,
+                                   SPIRVGlobalRegistry *GR) {
+  const SPIRV::DemangledBuiltin *Builtin = Call->Builtin;
+  unsigned Opcode =
+      SPIRV::lookupNativeBuiltin(Builtin->Name, Builtin->Set)->Opcode;
+
+  auto MIB = MIRBuilder.buildInstr(Opcode)
+                 .addDef(Call->ReturnRegister)
+                 .addUse(GR->getSPIRVTypeID(Call->ReturnType));
+  for (Register Arg : Call->Arguments)
+    MIB.addUse(Arg);
+  return true;
+}
+
 static bool generateGetQueryInst(const SPIRV::IncomingCall *Call,
                                  MachineIRBuilder &MIRBuilder,
                                  SPIRVGlobalRegistry *GR) {
@@ -3346,6 +3361,7 @@
   case SPIRV::AsyncCopy:
   case SPIRV::LoadStore:
   case SPIRV::CoopMatr:
+  case SPIRV::Arithmetic:
     if (const auto *R =
             SPIRV::lookupNativeBuiltin(Call->Builtin->Name, Call->Builtin->Set))
       return std::make_tuple(Call->Builtin->Group, R->Opcode, 0);
@@ -3459,6 +3475,8 @@
     return generateICarryBorrowInst(Call.get(), MIRBuilder, GR);
   case SPIRV::MulExtended:
     return generateMulExtendedInst(Call.get(), MIRBuilder, GR);
+  case SPIRV::Arithmetic:
+    return generateArithmeticInst(Call.get(), MIRBuilder, GR);
   case SPIRV::GetQuery:
     return generateGetQueryInst(Call.get(), MIRBuilder, GR);
   case SPIRV::ImageSizeQuery:

llvm/lib/Target/SPIRV/SPIRVBuiltins.td

diff --git a/llvm/lib/Target/SPIRV/SPIRVBuiltins.td b/llvm/lib/Target/SPIRV/SPIRVBuiltins.td
index ad19288..806d283 100644
--- a/llvm/lib/Target/SPIRV/SPIRVBuiltins.td
+++ b/llvm/lib/Target/SPIRV/SPIRVBuiltins.td
@@ -72,6 +72,7 @@
 def Block2DLoadStore : BuiltinGroup;
 def Pipe : BuiltinGroup;
 def PredicatedLoadStore : BuiltinGroup;
+def Arithmetic : BuiltinGroup;
 def ArbitraryPrecisionFixedPoint : BuiltinGroup;
 def BlockingPipes : BuiltinGroup;
 def ImageChannelDataTypes : BuiltinGroup;
@@ -697,6 +698,9 @@
 defm : DemangledNativeBuiltin<"__spirv_SMulExtended", OpenCL_std, MulExtended, 2, 3, OpSMulExtended>;
 defm : DemangledNativeBuiltin<"__spirv_SMulExtended", GLSL_std_450, MulExtended, 2, 3, OpSMulExtended>;
 
+// Arithmetic builtin records:
+defm : DemangledNativeBuiltin<"__spirv_FMod", OpenCL_std, Arithmetic, 2, 2, OpFMod>;
+
 // cl_intel_split_work_group_barrier
 defm : DemangledNativeBuiltin<"intel_work_group_barrier_arrive", OpenCL_std, Barrier, 1, 2, OpControlBarrierArriveINTEL>;
 defm : DemangledNativeBuiltin<"__spirv_ControlBarrierArriveINTEL", OpenCL_std, Barrier, 3, 3, OpControlBarrierArriveINTEL>;

llvm/test/CodeGen/SPIRV/instructions/scalar-floating-point-arithmetic.ll

diff --git a/llvm/test/CodeGen/SPIRV/instructions/scalar-floating-point-arithmetic.ll b/llvm/test/CodeGen/SPIRV/instructions/scalar-floating-point-arithmetic.ll
index b04db94..bdad95a 100644
--- a/llvm/test/CodeGen/SPIRV/instructions/scalar-floating-point-arithmetic.ll
+++ b/llvm/test/CodeGen/SPIRV/instructions/scalar-floating-point-arithmetic.ll
@@ -8,7 +8,7 @@
 ; CHECK-DAG: OpName [[SCALAR_FDIV:%.+]] "scalar_fdiv"
 ; CHECK-DAG: OpName [[SCALAR_FREM:%.+]] "scalar_frem"
 ; CHECK-DAG: OpName [[SCALAR_FMA:%.+]] "scalar_fma"
-;; FIXME: add test for OpFMod
+; CHECK-DAG: OpName [[SCALAR_FMOD:%.+]] "scalar_fmod"
 
 ; CHECK-NOT: DAG-FENCE
 
@@ -109,6 +109,22 @@
 ; CHECK:      OpReturnValue [[C]]
 ; CHECK-NEXT: OpFunctionEnd
 
+;; Test fmod on scalar:
+define spir_func float @scalar_fmod(float %a, float %b) {
+    %c = call spir_func float @_Z12__spirv_FModff(float %a, float %b)
+    ret float %c
+}
+
+declare spir_func float @_Z12__spirv_FModff(float, float)
+
+; CHECK:      [[SCALAR_FMOD]] = OpFunction [[SCALAR]] None [[SCALAR_FN]]
+; CHECK-NEXT: [[A:%.+]] = OpFunctionParameter [[SCALAR]]
+; CHECK-NEXT: [[B:%.+]] = OpFunctionParameter [[SCALAR]]
+; CHECK:      OpLabel
+; CHECK:      [[C:%.+]] = OpFMod [[SCALAR]] [[A]] [[B]]
+; CHECK:      OpReturnValue [[C]]
+; CHECK-NEXT: OpFunctionEnd
+
 declare float @llvm.fma.f32(float, float, float)
 
 ;; Test fma on scalar:

llvm/test/CodeGen/SPIRV/instructions/vector-floating-point-arithmetic.ll

diff --git a/llvm/test/CodeGen/SPIRV/instructions/vector-floating-point-arithmetic.ll b/llvm/test/CodeGen/SPIRV/instructions/vector-floating-point-arithmetic.ll
index 0b0e505..e79b3a6 100644
--- a/llvm/test/CodeGen/SPIRV/instructions/vector-floating-point-arithmetic.ll
+++ b/llvm/test/CodeGen/SPIRV/instructions/vector-floating-point-arithmetic.ll
@@ -7,7 +7,7 @@
 ; CHECK-DAG: OpName [[VECTOR_FMUL:%.+]] "vector_fmul"
 ; CHECK-DAG: OpName [[VECTOR_FDIV:%.+]] "vector_fdiv"
 ; CHECK-DAG: OpName [[VECTOR_FREM:%.+]] "vector_frem"
-;; TODO: add test for OpFMod
+; CHECK-DAG: OpName [[VECTOR_FMOD:%.+]] "vector_fmod"
 
 ; CHECK-NOT: DAG-FENCE
 
@@ -106,3 +106,20 @@
 ; CHECK:      [[C:%.+]] = OpFRem [[VECTOR]] [[A]] [[B]]
 ; CHECK:      OpReturnValue [[C]]
 ; CHECK-NEXT: OpFunctionEnd
+
+
+;; Test fmod on vector:
+define spir_func <2 x half> @vector_fmod(<2 x half> %a, <2 x half> %b) {
+    %c = call spir_func <2 x half> @_Z12__spirv_FModDv2_DhS_(<2 x half> %a, <2 x half> %b)
+    ret <2 x half> %c
+}
+
+declare spir_func <2 x half> @_Z12__spirv_FModDv2_DhS_(<2 x half>, <2 x half>)
+
+; CHECK:      [[VECTOR_FMOD]] = OpFunction [[VECTOR]] None [[VECTOR_FN]]
+; CHECK-NEXT: [[A:%.+]] = OpFunctionParameter [[VECTOR]]
+; CHECK-NEXT: [[B:%.+]] = OpFunctionParameter [[VECTOR]]
+; CHECK:      OpLabel
+; CHECK:      [[C:%.+]] = OpFMod [[VECTOR]] [[A]] [[B]]
+; CHECK:      OpReturnValue [[C]]
+; CHECK-NEXT: OpFunctionEnd