示例#1
0
bool AMDGPULegalizerInfo::legalizeAddrSpaceCast(
  MachineInstr &MI, MachineRegisterInfo &MRI,
  MachineIRBuilder &MIRBuilder) const {
  MachineFunction &MF = MIRBuilder.getMF();

  MIRBuilder.setInstr(MI);

  unsigned Dst = MI.getOperand(0).getReg();
  unsigned Src = MI.getOperand(1).getReg();

  LLT DstTy = MRI.getType(Dst);
  LLT SrcTy = MRI.getType(Src);
  unsigned DestAS = DstTy.getAddressSpace();
  unsigned SrcAS = SrcTy.getAddressSpace();

  // TODO: Avoid reloading from the queue ptr for each cast, or at least each
  // vector element.
  assert(!DstTy.isVector());

  const AMDGPUTargetMachine &TM
    = static_cast<const AMDGPUTargetMachine &>(MF.getTarget());

  const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();
  if (ST.getTargetLowering()->isNoopAddrSpaceCast(SrcAS, DestAS)) {
    MI.setDesc(MIRBuilder.getTII().get(TargetOpcode::G_BITCAST));
    return true;
  }

  if (SrcAS == AMDGPUAS::FLAT_ADDRESS) {
    assert(DestAS == AMDGPUAS::LOCAL_ADDRESS ||
           DestAS == AMDGPUAS::PRIVATE_ADDRESS);
    unsigned NullVal = TM.getNullPointerValue(DestAS);

    auto SegmentNull = MIRBuilder.buildConstant(DstTy, NullVal);
    auto FlatNull = MIRBuilder.buildConstant(SrcTy, 0);

    unsigned PtrLo32 = MRI.createGenericVirtualRegister(DstTy);

    // Extract low 32-bits of the pointer.
    MIRBuilder.buildExtract(PtrLo32, Src, 0);

    unsigned CmpRes = MRI.createGenericVirtualRegister(LLT::scalar(1));
    MIRBuilder.buildICmp(CmpInst::ICMP_NE, CmpRes, Src, FlatNull.getReg(0));
    MIRBuilder.buildSelect(Dst, CmpRes, PtrLo32, SegmentNull.getReg(0));

    MI.eraseFromParent();
    return true;
  }

  assert(SrcAS == AMDGPUAS::LOCAL_ADDRESS ||
         SrcAS == AMDGPUAS::PRIVATE_ADDRESS);

  auto SegmentNull =
      MIRBuilder.buildConstant(SrcTy, TM.getNullPointerValue(SrcAS));
  auto FlatNull =
      MIRBuilder.buildConstant(DstTy, TM.getNullPointerValue(DestAS));

  unsigned ApertureReg = getSegmentAperture(DestAS, MRI, MIRBuilder);

  unsigned CmpRes = MRI.createGenericVirtualRegister(LLT::scalar(1));
  MIRBuilder.buildICmp(CmpInst::ICMP_NE, CmpRes, Src, SegmentNull.getReg(0));

  unsigned BuildPtr = MRI.createGenericVirtualRegister(DstTy);

  // Coerce the type of the low half of the result so we can use merge_values.
  unsigned SrcAsInt = MRI.createGenericVirtualRegister(LLT::scalar(32));
  MIRBuilder.buildInstr(TargetOpcode::G_PTRTOINT)
    .addDef(SrcAsInt)
    .addUse(Src);

  // TODO: Should we allow mismatched types but matching sizes in merges to
  // avoid the ptrtoint?
  MIRBuilder.buildMerge(BuildPtr, {SrcAsInt, ApertureReg});
  MIRBuilder.buildSelect(Dst, CmpRes, BuildPtr, FlatNull.getReg(0));

  MI.eraseFromParent();
  return true;
}
示例#2
0
void LegalizerInfo::computeTables() {
  assert(TablesInitialized == false);

  for (unsigned OpcodeIdx = 0; OpcodeIdx <= LastOp - FirstOp; ++OpcodeIdx) {
    const unsigned Opcode = FirstOp + OpcodeIdx;
    for (unsigned TypeIdx = 0; TypeIdx != SpecifiedActions[OpcodeIdx].size();
         ++TypeIdx) {
      // 0. Collect information specified through the setAction API, i.e.
      // for specific bit sizes.
      // For scalar types:
      SizeAndActionsVec ScalarSpecifiedActions;
      // For pointer types:
      std::map<uint16_t, SizeAndActionsVec> AddressSpace2SpecifiedActions;
      // For vector types:
      std::map<uint16_t, SizeAndActionsVec> ElemSize2SpecifiedActions;
      for (auto LLT2Action : SpecifiedActions[OpcodeIdx][TypeIdx]) {
        const LLT Type = LLT2Action.first;
        const LegalizeAction Action = LLT2Action.second;

        auto SizeAction = std::make_pair(Type.getSizeInBits(), Action);
        if (Type.isPointer())
          AddressSpace2SpecifiedActions[Type.getAddressSpace()].push_back(
              SizeAction);
        else if (Type.isVector())
          ElemSize2SpecifiedActions[Type.getElementType().getSizeInBits()]
              .push_back(SizeAction);
        else
          ScalarSpecifiedActions.push_back(SizeAction);
      }

      // 1. Handle scalar types
      {
        // Decide how to handle bit sizes for which no explicit specification
        // was given.
        SizeChangeStrategy S = &unsupportedForDifferentSizes;
        if (TypeIdx < ScalarSizeChangeStrategies[OpcodeIdx].size() &&
            ScalarSizeChangeStrategies[OpcodeIdx][TypeIdx] != nullptr)
          S = ScalarSizeChangeStrategies[OpcodeIdx][TypeIdx];
        std::sort(ScalarSpecifiedActions.begin(), ScalarSpecifiedActions.end());
        checkPartialSizeAndActionsVector(ScalarSpecifiedActions);
        setScalarAction(Opcode, TypeIdx, S(ScalarSpecifiedActions));
      }

      // 2. Handle pointer types
      for (auto PointerSpecifiedActions : AddressSpace2SpecifiedActions) {
        std::sort(PointerSpecifiedActions.second.begin(),
                  PointerSpecifiedActions.second.end());
        checkPartialSizeAndActionsVector(PointerSpecifiedActions.second);
        // For pointer types, we assume that there isn't a meaningfull way
        // to change the number of bits used in the pointer.
        setPointerAction(
            Opcode, TypeIdx, PointerSpecifiedActions.first,
            unsupportedForDifferentSizes(PointerSpecifiedActions.second));
      }

      // 3. Handle vector types
      SizeAndActionsVec ElementSizesSeen;
      for (auto VectorSpecifiedActions : ElemSize2SpecifiedActions) {
        std::sort(VectorSpecifiedActions.second.begin(),
                  VectorSpecifiedActions.second.end());
        const uint16_t ElementSize = VectorSpecifiedActions.first;
        ElementSizesSeen.push_back({ElementSize, Legal});
        checkPartialSizeAndActionsVector(VectorSpecifiedActions.second);
        // For vector types, we assume that the best way to adapt the number
        // of elements is to the next larger number of elements type for which
        // the vector type is legal, unless there is no such type. In that case,
        // legalize towards a vector type with a smaller number of elements.
        SizeAndActionsVec NumElementsActions;
        for (SizeAndAction BitsizeAndAction : VectorSpecifiedActions.second) {
          assert(BitsizeAndAction.first % ElementSize == 0);
          const uint16_t NumElements = BitsizeAndAction.first / ElementSize;
          NumElementsActions.push_back({NumElements, BitsizeAndAction.second});
        }
        setVectorNumElementAction(
            Opcode, TypeIdx, ElementSize,
            moreToWiderTypesAndLessToWidest(NumElementsActions));
      }
      std::sort(ElementSizesSeen.begin(), ElementSizesSeen.end());
      SizeChangeStrategy VectorElementSizeChangeStrategy =
          &unsupportedForDifferentSizes;
      if (TypeIdx < VectorElementSizeChangeStrategies[OpcodeIdx].size() &&
          VectorElementSizeChangeStrategies[OpcodeIdx][TypeIdx] != nullptr)
        VectorElementSizeChangeStrategy =
            VectorElementSizeChangeStrategies[OpcodeIdx][TypeIdx];
      setScalarInVectorAction(
          Opcode, TypeIdx, VectorElementSizeChangeStrategy(ElementSizesSeen));
    }
  }

  TablesInitialized = true;
}