[SPIRV] Implement translation for llvm.modf.* intrinsics

llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp

@@ -296,6 +296,8 @@ class SPIRVInstructionSelector : public InstructionSelector {
   bool selectImageWriteIntrinsic(MachineInstr &I) const;
   bool selectResourceGetPointer(Register &ResVReg, const SPIRVType *ResType,
                                 MachineInstr &I) const;
+  bool selectModf(Register ResVReg, const SPIRVType *ResType,
+                  MachineInstr &I) const;
 
   // Utilities
   std::pair<Register, bool>
@@ -3207,6 +3209,9 @@ bool SPIRVInstructionSelector::selectIntrinsic(Register ResVReg,
   case Intrinsic::spv_discard: {
     return selectDiscard(ResVReg, ResType, I);
   }
+  case Intrinsic::modf: {
+    return selectModf(ResVReg, ResType, I);
+  }
   default: {
     std::string DiagMsg;
     raw_string_ostream OS(DiagMsg);
@@ -3990,6 +3995,104 @@ bool SPIRVInstructionSelector::selectLog10(Register ResVReg,
                        .constrainAllUses(TII, TRI, RBI);
 }
 
+bool SPIRVInstructionSelector::selectModf(Register ResVReg,
+                                          const SPIRVType *ResType,
+                                          MachineInstr &I) const {
+  // llvm.modf has a single arg --the number to be decomposed-- and returns a
+  // struct { restype, restype }, while OpenCLLIB::modf has two args --the
+  // number to be decomposed and a pointer--, returns the fractional part and
+  // the integral part is stored in the pointer argument. Therefore, we can't
+  // use directly the OpenCLLIB::modf intrinsic. However, we can do some
+  // scaffolding to make it work. The idea is to create an alloca instruction
+  // to get a ptr, pass this ptr to OpenCL::modf, and then load the value
+  // from this ptr to place it in the struct. llvm.modf returns the fractional
+  // part as the first element of the result, and the integral part as the
+  // second element of the result.
+
+  // At this point, the return type is not a struct anymore, but rather two
+  // independent elements of SPIRVResType. We can get each independent element
+  // from I.getDefs() or I.getOperands().
+  if (STI.canUseExtInstSet(SPIRV::InstructionSet::OpenCL_std)) {
+    uint32_t Opcode = CL::modf;
+    MachineIRBuilder MIRBuilder(I);
+    // Get pointer type for alloca variable.
+    const SPIRVType *PtrType = GR.getOrCreateSPIRVPointerType(
+        ResType, MIRBuilder, SPIRV::StorageClass::Function);
+    // Create new register for the pointer type of alloca variable.
+    Register PtrTyReg =
+        MIRBuilder.getMRI()->createVirtualRegister(&SPIRV::iIDRegClass);
+    MIRBuilder.getMRI()->setType(
+        PtrTyReg,
+        LLT::pointer(storageClassToAddressSpace(SPIRV::StorageClass::Function),
+                     GR.getPointerSize()));
+    // Assign SPIR-V type of the pointer type of the alloca variable to the
+    // new register.
+    GR.assignSPIRVTypeToVReg(PtrType, PtrTyReg, MIRBuilder.getMF());
+    MachineBasicBlock &EntryBB = I.getMF()->front();
+    // At this point it's difficult to find the right position to insert the
+    // variable, because most instructions are still MachineInstruction and
+    // don't have SPIRV opcodes yet. OpFunction and OpFunctionParameter are
+    // already translated, so we will aim to insert the variable just after the
+    // last OpFunctionParameter, if any, or just after OpFunction otherwise.
+    auto VarPos = EntryBB.begin();
+    while (VarPos != EntryBB.end() &&
+           VarPos->getOpcode() != SPIRV::OpFunction) {
+      ++VarPos;
+    }
+    // Advance VarPos to the next instruction after OpFunction, it will either
+    // be an OpFunctionParameter, so that we can start the next loop, or the
+    // position to insert the OpVariable instruction.
+    ++VarPos;
+    while (VarPos != EntryBB.end() &&
+           VarPos->getOpcode() == SPIRV::OpFunctionParameter) {
+      ++VarPos;
+    }
+    // VarPos is now pointing at after the last OpFunctionParameter, if any,
+    // or after OpFunction, if no parameters.
+    // Create a new MachineInstruction for alloca variable in the
+    // entry block.
+    auto AllocaMIB =
+        BuildMI(EntryBB, VarPos, I.getDebugLoc(), TII.get(SPIRV::OpVariable))
+            .addDef(PtrTyReg)
+            .addUse(GR.getSPIRVTypeID(PtrType))
+            .addImm(static_cast<uint32_t>(SPIRV::StorageClass::Function));
+    Register Variable = AllocaMIB->getOperand(0).getReg();
+    // Modf must have 4 operands, the first two are the 2 parts of the result,
+    // the third is the operand, and the last one is the floating point value.
+    assert(I.getNumOperands() == 4 &&
+           "Expected 4 operands for modf instruction");
+    MachineBasicBlock &BB = *I.getParent();
+    // Create the OpenCLLIB::modf instruction.
+    auto MIB =
+        BuildMI(BB, I, I.getDebugLoc(), TII.get(SPIRV::OpExtInst))
+            .addDef(ResVReg)
+            .addUse(GR.getSPIRVTypeID(ResType))
+            .addImm(static_cast<uint32_t>(SPIRV::InstructionSet::OpenCL_std))
+            .addImm(Opcode)
+            .setMIFlags(I.getFlags())
+            .add(I.getOperand(3)) // Floating point value.
+            .addUse(Variable);    // Pointer to integral part.
+    // Assign the integral part stored in the ptr to the second element of the
+    // result.
+    Register IntegralPartReg = I.getOperand(1).getReg();
+    if (IntegralPartReg.isValid()) {
+      // Load the value from the pointer to integral part.
+      auto LoadMIB = BuildMI(BB, I, I.getDebugLoc(), TII.get(SPIRV::OpLoad))
+                         .addDef(IntegralPartReg)
+                         .addUse(GR.getSPIRVTypeID(ResType))
+                         .addUse(Variable);
+      return LoadMIB.constrainAllUses(TII, TRI, RBI);
+    }
+
+    return MIB.constrainAllUses(TII, TRI, RBI);
+  } else if (STI.canUseExtInstSet(SPIRV::InstructionSet::GLSL_std_450)) {
+    assert(false && "GLSL::Modf is deprecated.");
+    // FIXME: GL::Modf is deprecated, use Modfstruct instead.
+    return false;
+  }
+  return false;
+}
+
 // Generate the instructions to load 3-element vector builtin input
 // IDs/Indices.
 // Like: GlobalInvocationId, LocalInvocationId, etc....

llvm/test/CodeGen/SPIRV/llvm-intrinsics/fp-intrinsics.ll

@@ -1,4 +1,5 @@
 ; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s
+; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64-unknown-unknown %s -o - -filetype=obj | spirv-val %}
 
 ; CHECK: %[[#extinst_id:]] = OpExtInstImport "OpenCL.std"
 
@@ -337,3 +338,61 @@ entry:
 }
 
 declare float @llvm.fma.f32(float, float, float)
+
+; CHECK: OpFunction
+; CHECK: %[[#d:]] = OpFunctionParameter %[[#]]
+; CHECK: %[[#fracPtr:]] = OpFunctionParameter %[[#]]
+; CHECK: %[[#integralPtr:]] = OpFunctionParameter %[[#]]
+; CHECK: %[[#varPtr:]] = OpVariable %[[#]] Function
+; CHECK: %[[#frac:]] = OpExtInst %[[#var2]] %[[#extinst_id]] modf %[[#d]] %[[#varPtr]]
+; CHECK: %[[#integral:]] = OpLoad %[[#var2]] %[[#varPtr]]
+; CHECK: OpStore %[[#fracPtr]] %[[#frac]]
+; CHECK: OpStore %[[#integralPtr]] %[[#integral]]
+; CHECK: OpFunctionEnd
+define void @TestModf(double %d, ptr addrspace(1) %frac, ptr addrspace(1) %integral) {
+entry:
+  %4 = tail call { double, double } @llvm.modf.f64(double %d)
+  %5 = extractvalue { double, double } %4, 0
+  %6 = extractvalue { double, double } %4, 1
+  store double %5, ptr addrspace(1) %frac, align 8
+  store double %6, ptr addrspace(1) %integral, align 8
+  ret void
+}
+
+define dso_local void @TestModf2(double noundef %d, ptr noundef %frac, ptr noundef %integral) {
+entry:
+  %d.addr = alloca double, align 4
+  %frac.addr = alloca ptr, align 8
+  %integral.addr = alloca ptr, align 8
+  store double %d, ptr %d.addr, align 4
+  store ptr %frac, ptr %frac.addr, align 8
+  store ptr %integral, ptr %integral.addr, align 8
+  %0 = load ptr, ptr %frac.addr, align 8
+  %tobool = icmp ne ptr %0, null
+  br i1 %tobool, label %lor.lhs.false, label %if.then
+
+lor.lhs.false:
+  %1 = load ptr, ptr %integral.addr, align 8
+  %tobool1 = icmp ne ptr %1, null
+  br i1 %tobool1, label %if.end, label %if.then
+
+if.then:
+  br label %return
+
+if.end:
+  %2 = load ptr, ptr %frac.addr, align 8
+  %3 = load double, ptr %2, align 4
+  %4 = load ptr, ptr %integral.addr, align 8
+  %5 = load double, ptr %4, align 4
+  %6 = tail call { double, double } @llvm.modf.f64(double %d)
+  %7 = extractvalue { double, double } %6, 0
+  %8 = extractvalue { double, double } %6, 1
+  store double %7, ptr %frac, align 4
+  store double %7, ptr %integral, align 4
+  br label %return
+
+return:
+  ret void
+}
+
+declare { double, double } @llvm.modf.f64(double)