+ MachineSDNode::mmo_iterator MemRefs0 = MF->allocateMemRefsArray(1);
+ MemRefs0[0] = cast<MemSDNode>(N)->getMemOperand();
+ cast<MachineSDNode>(ST)->setMemRefs(MemRefs0, MemRefs0 + 1);
+
+ return ST;
+}
+
+SDNode *NVPTXDAGToDAGISel::SelectLoadParam(SDNode *Node) {
+ SDValue Chain = Node->getOperand(0);
+ SDValue Offset = Node->getOperand(2);
+ SDValue Flag = Node->getOperand(3);
+ SDLoc DL(Node);
+ MemSDNode *Mem = cast<MemSDNode>(Node);
+
+ unsigned VecSize;
+ switch (Node->getOpcode()) {
+ default:
+ return nullptr;
+ case NVPTXISD::LoadParam:
+ VecSize = 1;
+ break;
+ case NVPTXISD::LoadParamV2:
+ VecSize = 2;
+ break;
+ case NVPTXISD::LoadParamV4:
+ VecSize = 4;
+ break;
+ }
+
+ EVT EltVT = Node->getValueType(0);
+ EVT MemVT = Mem->getMemoryVT();
+
+ unsigned Opc = 0;
+
+ switch (VecSize) {
+ default:
+ return nullptr;
+ case 1:
+ switch (MemVT.getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opc = NVPTX::LoadParamMemI8;
+ break;
+ case MVT::i8:
+ Opc = NVPTX::LoadParamMemI8;
+ break;
+ case MVT::i16:
+ Opc = NVPTX::LoadParamMemI16;
+ break;
+ case MVT::i32:
+ Opc = NVPTX::LoadParamMemI32;
+ break;
+ case MVT::i64:
+ Opc = NVPTX::LoadParamMemI64;
+ break;
+ case MVT::f32:
+ Opc = NVPTX::LoadParamMemF32;
+ break;
+ case MVT::f64:
+ Opc = NVPTX::LoadParamMemF64;
+ break;
+ }
+ break;
+ case 2:
+ switch (MemVT.getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opc = NVPTX::LoadParamMemV2I8;
+ break;
+ case MVT::i8:
+ Opc = NVPTX::LoadParamMemV2I8;
+ break;
+ case MVT::i16:
+ Opc = NVPTX::LoadParamMemV2I16;
+ break;
+ case MVT::i32:
+ Opc = NVPTX::LoadParamMemV2I32;
+ break;
+ case MVT::i64:
+ Opc = NVPTX::LoadParamMemV2I64;
+ break;
+ case MVT::f32:
+ Opc = NVPTX::LoadParamMemV2F32;
+ break;
+ case MVT::f64:
+ Opc = NVPTX::LoadParamMemV2F64;
+ break;
+ }
+ break;
+ case 4:
+ switch (MemVT.getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opc = NVPTX::LoadParamMemV4I8;
+ break;
+ case MVT::i8:
+ Opc = NVPTX::LoadParamMemV4I8;
+ break;
+ case MVT::i16:
+ Opc = NVPTX::LoadParamMemV4I16;
+ break;
+ case MVT::i32:
+ Opc = NVPTX::LoadParamMemV4I32;
+ break;
+ case MVT::f32:
+ Opc = NVPTX::LoadParamMemV4F32;
+ break;
+ }
+ break;
+ }
+
+ SDVTList VTs;
+ if (VecSize == 1) {
+ VTs = CurDAG->getVTList(EltVT, MVT::Other, MVT::Glue);
+ } else if (VecSize == 2) {
+ VTs = CurDAG->getVTList(EltVT, EltVT, MVT::Other, MVT::Glue);
+ } else {
+ EVT EVTs[] = { EltVT, EltVT, EltVT, EltVT, MVT::Other, MVT::Glue };
+ VTs = CurDAG->getVTList(EVTs);
+ }
+
+ unsigned OffsetVal = cast<ConstantSDNode>(Offset)->getZExtValue();
+
+ SmallVector<SDValue, 2> Ops;
+ Ops.push_back(CurDAG->getTargetConstant(OffsetVal, MVT::i32));
+ Ops.push_back(Chain);
+ Ops.push_back(Flag);
+
+ SDNode *Ret =
+ CurDAG->getMachineNode(Opc, DL, VTs, Ops);
+ return Ret;
+}
+
+SDNode *NVPTXDAGToDAGISel::SelectStoreRetval(SDNode *N) {
+ SDLoc DL(N);
+ SDValue Chain = N->getOperand(0);
+ SDValue Offset = N->getOperand(1);
+ unsigned OffsetVal = cast<ConstantSDNode>(Offset)->getZExtValue();
+ MemSDNode *Mem = cast<MemSDNode>(N);
+
+ // How many elements do we have?
+ unsigned NumElts = 1;
+ switch (N->getOpcode()) {
+ default:
+ return nullptr;
+ case NVPTXISD::StoreRetval:
+ NumElts = 1;
+ break;
+ case NVPTXISD::StoreRetvalV2:
+ NumElts = 2;
+ break;
+ case NVPTXISD::StoreRetvalV4:
+ NumElts = 4;
+ break;
+ }
+
+ // Build vector of operands
+ SmallVector<SDValue, 6> Ops;
+ for (unsigned i = 0; i < NumElts; ++i)
+ Ops.push_back(N->getOperand(i + 2));
+ Ops.push_back(CurDAG->getTargetConstant(OffsetVal, MVT::i32));
+ Ops.push_back(Chain);
+
+ // Determine target opcode
+ // If we have an i1, use an 8-bit store. The lowering code in
+ // NVPTXISelLowering will have already emitted an upcast.
+ unsigned Opcode = 0;
+ switch (NumElts) {
+ default:
+ return nullptr;
+ case 1:
+ switch (Mem->getMemoryVT().getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opcode = NVPTX::StoreRetvalI8;
+ break;
+ case MVT::i8:
+ Opcode = NVPTX::StoreRetvalI8;
+ break;
+ case MVT::i16:
+ Opcode = NVPTX::StoreRetvalI16;
+ break;
+ case MVT::i32:
+ Opcode = NVPTX::StoreRetvalI32;
+ break;
+ case MVT::i64:
+ Opcode = NVPTX::StoreRetvalI64;
+ break;
+ case MVT::f32:
+ Opcode = NVPTX::StoreRetvalF32;
+ break;
+ case MVT::f64:
+ Opcode = NVPTX::StoreRetvalF64;
+ break;
+ }
+ break;
+ case 2:
+ switch (Mem->getMemoryVT().getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opcode = NVPTX::StoreRetvalV2I8;
+ break;
+ case MVT::i8:
+ Opcode = NVPTX::StoreRetvalV2I8;
+ break;
+ case MVT::i16:
+ Opcode = NVPTX::StoreRetvalV2I16;
+ break;
+ case MVT::i32:
+ Opcode = NVPTX::StoreRetvalV2I32;
+ break;
+ case MVT::i64:
+ Opcode = NVPTX::StoreRetvalV2I64;
+ break;
+ case MVT::f32:
+ Opcode = NVPTX::StoreRetvalV2F32;
+ break;
+ case MVT::f64:
+ Opcode = NVPTX::StoreRetvalV2F64;
+ break;
+ }
+ break;
+ case 4:
+ switch (Mem->getMemoryVT().getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opcode = NVPTX::StoreRetvalV4I8;
+ break;
+ case MVT::i8:
+ Opcode = NVPTX::StoreRetvalV4I8;
+ break;
+ case MVT::i16:
+ Opcode = NVPTX::StoreRetvalV4I16;
+ break;
+ case MVT::i32:
+ Opcode = NVPTX::StoreRetvalV4I32;
+ break;
+ case MVT::f32:
+ Opcode = NVPTX::StoreRetvalV4F32;
+ break;
+ }
+ break;
+ }
+
+ SDNode *Ret =
+ CurDAG->getMachineNode(Opcode, DL, MVT::Other, Ops);
+ MachineSDNode::mmo_iterator MemRefs0 = MF->allocateMemRefsArray(1);
+ MemRefs0[0] = cast<MemSDNode>(N)->getMemOperand();
+ cast<MachineSDNode>(Ret)->setMemRefs(MemRefs0, MemRefs0 + 1);
+
+ return Ret;
+}
+
+SDNode *NVPTXDAGToDAGISel::SelectStoreParam(SDNode *N) {
+ SDLoc DL(N);
+ SDValue Chain = N->getOperand(0);
+ SDValue Param = N->getOperand(1);
+ unsigned ParamVal = cast<ConstantSDNode>(Param)->getZExtValue();
+ SDValue Offset = N->getOperand(2);
+ unsigned OffsetVal = cast<ConstantSDNode>(Offset)->getZExtValue();
+ MemSDNode *Mem = cast<MemSDNode>(N);
+ SDValue Flag = N->getOperand(N->getNumOperands() - 1);
+
+ // How many elements do we have?
+ unsigned NumElts = 1;
+ switch (N->getOpcode()) {
+ default:
+ return nullptr;
+ case NVPTXISD::StoreParamU32:
+ case NVPTXISD::StoreParamS32:
+ case NVPTXISD::StoreParam:
+ NumElts = 1;
+ break;
+ case NVPTXISD::StoreParamV2:
+ NumElts = 2;
+ break;
+ case NVPTXISD::StoreParamV4:
+ NumElts = 4;
+ break;
+ }
+
+ // Build vector of operands
+ SmallVector<SDValue, 8> Ops;
+ for (unsigned i = 0; i < NumElts; ++i)
+ Ops.push_back(N->getOperand(i + 3));
+ Ops.push_back(CurDAG->getTargetConstant(ParamVal, MVT::i32));
+ Ops.push_back(CurDAG->getTargetConstant(OffsetVal, MVT::i32));
+ Ops.push_back(Chain);
+ Ops.push_back(Flag);
+
+ // Determine target opcode
+ // If we have an i1, use an 8-bit store. The lowering code in
+ // NVPTXISelLowering will have already emitted an upcast.
+ unsigned Opcode = 0;
+ switch (N->getOpcode()) {
+ default:
+ switch (NumElts) {
+ default:
+ return nullptr;
+ case 1:
+ switch (Mem->getMemoryVT().getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opcode = NVPTX::StoreParamI8;
+ break;
+ case MVT::i8:
+ Opcode = NVPTX::StoreParamI8;
+ break;
+ case MVT::i16:
+ Opcode = NVPTX::StoreParamI16;
+ break;
+ case MVT::i32:
+ Opcode = NVPTX::StoreParamI32;
+ break;
+ case MVT::i64:
+ Opcode = NVPTX::StoreParamI64;
+ break;
+ case MVT::f32:
+ Opcode = NVPTX::StoreParamF32;
+ break;
+ case MVT::f64:
+ Opcode = NVPTX::StoreParamF64;
+ break;
+ }
+ break;
+ case 2:
+ switch (Mem->getMemoryVT().getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opcode = NVPTX::StoreParamV2I8;
+ break;
+ case MVT::i8:
+ Opcode = NVPTX::StoreParamV2I8;
+ break;
+ case MVT::i16:
+ Opcode = NVPTX::StoreParamV2I16;
+ break;
+ case MVT::i32:
+ Opcode = NVPTX::StoreParamV2I32;
+ break;
+ case MVT::i64:
+ Opcode = NVPTX::StoreParamV2I64;
+ break;
+ case MVT::f32:
+ Opcode = NVPTX::StoreParamV2F32;
+ break;
+ case MVT::f64:
+ Opcode = NVPTX::StoreParamV2F64;
+ break;
+ }
+ break;
+ case 4:
+ switch (Mem->getMemoryVT().getSimpleVT().SimpleTy) {
+ default:
+ return nullptr;
+ case MVT::i1:
+ Opcode = NVPTX::StoreParamV4I8;
+ break;
+ case MVT::i8:
+ Opcode = NVPTX::StoreParamV4I8;
+ break;
+ case MVT::i16:
+ Opcode = NVPTX::StoreParamV4I16;
+ break;
+ case MVT::i32:
+ Opcode = NVPTX::StoreParamV4I32;
+ break;
+ case MVT::f32:
+ Opcode = NVPTX::StoreParamV4F32;
+ break;
+ }
+ break;
+ }
+ break;
+ // Special case: if we have a sign-extend/zero-extend node, insert the
+ // conversion instruction first, and use that as the value operand to
+ // the selected StoreParam node.
+ case NVPTXISD::StoreParamU32: {
+ Opcode = NVPTX::StoreParamI32;
+ SDValue CvtNone = CurDAG->getTargetConstant(NVPTX::PTXCvtMode::NONE,
+ MVT::i32);
+ SDNode *Cvt = CurDAG->getMachineNode(NVPTX::CVT_u32_u16, DL,
+ MVT::i32, Ops[0], CvtNone);
+ Ops[0] = SDValue(Cvt, 0);
+ break;
+ }
+ case NVPTXISD::StoreParamS32: {
+ Opcode = NVPTX::StoreParamI32;
+ SDValue CvtNone = CurDAG->getTargetConstant(NVPTX::PTXCvtMode::NONE,
+ MVT::i32);
+ SDNode *Cvt = CurDAG->getMachineNode(NVPTX::CVT_s32_s16, DL,
+ MVT::i32, Ops[0], CvtNone);
+ Ops[0] = SDValue(Cvt, 0);
+ break;
+ }
+ }
+
+ SDVTList RetVTs = CurDAG->getVTList(MVT::Other, MVT::Glue);
+ SDNode *Ret =
+ CurDAG->getMachineNode(Opcode, DL, RetVTs, Ops);
+ MachineSDNode::mmo_iterator MemRefs0 = MF->allocateMemRefsArray(1);
+ MemRefs0[0] = cast<MemSDNode>(N)->getMemOperand();
+ cast<MachineSDNode>(Ret)->setMemRefs(MemRefs0, MemRefs0 + 1);
+
+ return Ret;
+}
+
+SDNode *NVPTXDAGToDAGISel::SelectTextureIntrinsic(SDNode *N) {
+ SDValue Chain = N->getOperand(0);
+ SDNode *Ret = nullptr;
+ unsigned Opc = 0;
+ SmallVector<SDValue, 8> Ops;
+
+ switch (N->getOpcode()) {
+ default: return nullptr;
+ case NVPTXISD::Tex1DFloatS32:
+ Opc = NVPTX::TEX_1D_F32_S32;
+ break;
+ case NVPTXISD::Tex1DFloatFloat:
+ Opc = NVPTX::TEX_1D_F32_F32;
+ break;
+ case NVPTXISD::Tex1DFloatFloatLevel:
+ Opc = NVPTX::TEX_1D_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex1DFloatFloatGrad:
+ Opc = NVPTX::TEX_1D_F32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex1DS32S32:
+ Opc = NVPTX::TEX_1D_S32_S32;
+ break;
+ case NVPTXISD::Tex1DS32Float:
+ Opc = NVPTX::TEX_1D_S32_F32;
+ break;
+ case NVPTXISD::Tex1DS32FloatLevel:
+ Opc = NVPTX::TEX_1D_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex1DS32FloatGrad:
+ Opc = NVPTX::TEX_1D_S32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex1DU32S32:
+ Opc = NVPTX::TEX_1D_U32_S32;
+ break;
+ case NVPTXISD::Tex1DU32Float:
+ Opc = NVPTX::TEX_1D_U32_F32;
+ break;
+ case NVPTXISD::Tex1DU32FloatLevel:
+ Opc = NVPTX::TEX_1D_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex1DU32FloatGrad:
+ Opc = NVPTX::TEX_1D_U32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex1DArrayFloatS32:
+ Opc = NVPTX::TEX_1D_ARRAY_F32_S32;
+ break;
+ case NVPTXISD::Tex1DArrayFloatFloat:
+ Opc = NVPTX::TEX_1D_ARRAY_F32_F32;
+ break;
+ case NVPTXISD::Tex1DArrayFloatFloatLevel:
+ Opc = NVPTX::TEX_1D_ARRAY_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex1DArrayFloatFloatGrad:
+ Opc = NVPTX::TEX_1D_ARRAY_F32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex1DArrayS32S32:
+ Opc = NVPTX::TEX_1D_ARRAY_S32_S32;
+ break;
+ case NVPTXISD::Tex1DArrayS32Float:
+ Opc = NVPTX::TEX_1D_ARRAY_S32_F32;
+ break;
+ case NVPTXISD::Tex1DArrayS32FloatLevel:
+ Opc = NVPTX::TEX_1D_ARRAY_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex1DArrayS32FloatGrad:
+ Opc = NVPTX::TEX_1D_ARRAY_S32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex1DArrayU32S32:
+ Opc = NVPTX::TEX_1D_ARRAY_U32_S32;
+ break;
+ case NVPTXISD::Tex1DArrayU32Float:
+ Opc = NVPTX::TEX_1D_ARRAY_U32_F32;
+ break;
+ case NVPTXISD::Tex1DArrayU32FloatLevel:
+ Opc = NVPTX::TEX_1D_ARRAY_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex1DArrayU32FloatGrad:
+ Opc = NVPTX::TEX_1D_ARRAY_U32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex2DFloatS32:
+ Opc = NVPTX::TEX_2D_F32_S32;
+ break;
+ case NVPTXISD::Tex2DFloatFloat:
+ Opc = NVPTX::TEX_2D_F32_F32;
+ break;
+ case NVPTXISD::Tex2DFloatFloatLevel:
+ Opc = NVPTX::TEX_2D_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex2DFloatFloatGrad:
+ Opc = NVPTX::TEX_2D_F32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex2DS32S32:
+ Opc = NVPTX::TEX_2D_S32_S32;
+ break;
+ case NVPTXISD::Tex2DS32Float:
+ Opc = NVPTX::TEX_2D_S32_F32;
+ break;
+ case NVPTXISD::Tex2DS32FloatLevel:
+ Opc = NVPTX::TEX_2D_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex2DS32FloatGrad:
+ Opc = NVPTX::TEX_2D_S32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex2DU32S32:
+ Opc = NVPTX::TEX_2D_U32_S32;
+ break;
+ case NVPTXISD::Tex2DU32Float:
+ Opc = NVPTX::TEX_2D_U32_F32;
+ break;
+ case NVPTXISD::Tex2DU32FloatLevel:
+ Opc = NVPTX::TEX_2D_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex2DU32FloatGrad:
+ Opc = NVPTX::TEX_2D_U32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex2DArrayFloatS32:
+ Opc = NVPTX::TEX_2D_ARRAY_F32_S32;
+ break;
+ case NVPTXISD::Tex2DArrayFloatFloat:
+ Opc = NVPTX::TEX_2D_ARRAY_F32_F32;
+ break;
+ case NVPTXISD::Tex2DArrayFloatFloatLevel:
+ Opc = NVPTX::TEX_2D_ARRAY_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex2DArrayFloatFloatGrad:
+ Opc = NVPTX::TEX_2D_ARRAY_F32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex2DArrayS32S32:
+ Opc = NVPTX::TEX_2D_ARRAY_S32_S32;
+ break;
+ case NVPTXISD::Tex2DArrayS32Float:
+ Opc = NVPTX::TEX_2D_ARRAY_S32_F32;
+ break;
+ case NVPTXISD::Tex2DArrayS32FloatLevel:
+ Opc = NVPTX::TEX_2D_ARRAY_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex2DArrayS32FloatGrad:
+ Opc = NVPTX::TEX_2D_ARRAY_S32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex2DArrayU32S32:
+ Opc = NVPTX::TEX_2D_ARRAY_U32_S32;
+ break;
+ case NVPTXISD::Tex2DArrayU32Float:
+ Opc = NVPTX::TEX_2D_ARRAY_U32_F32;
+ break;
+ case NVPTXISD::Tex2DArrayU32FloatLevel:
+ Opc = NVPTX::TEX_2D_ARRAY_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex2DArrayU32FloatGrad:
+ Opc = NVPTX::TEX_2D_ARRAY_U32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex3DFloatS32:
+ Opc = NVPTX::TEX_3D_F32_S32;
+ break;
+ case NVPTXISD::Tex3DFloatFloat:
+ Opc = NVPTX::TEX_3D_F32_F32;
+ break;
+ case NVPTXISD::Tex3DFloatFloatLevel:
+ Opc = NVPTX::TEX_3D_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex3DFloatFloatGrad:
+ Opc = NVPTX::TEX_3D_F32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex3DS32S32:
+ Opc = NVPTX::TEX_3D_S32_S32;
+ break;
+ case NVPTXISD::Tex3DS32Float:
+ Opc = NVPTX::TEX_3D_S32_F32;
+ break;
+ case NVPTXISD::Tex3DS32FloatLevel:
+ Opc = NVPTX::TEX_3D_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex3DS32FloatGrad:
+ Opc = NVPTX::TEX_3D_S32_F32_GRAD;
+ break;
+ case NVPTXISD::Tex3DU32S32:
+ Opc = NVPTX::TEX_3D_U32_S32;
+ break;
+ case NVPTXISD::Tex3DU32Float:
+ Opc = NVPTX::TEX_3D_U32_F32;
+ break;
+ case NVPTXISD::Tex3DU32FloatLevel:
+ Opc = NVPTX::TEX_3D_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tex3DU32FloatGrad:
+ Opc = NVPTX::TEX_3D_U32_F32_GRAD;
+ break;
+ case NVPTXISD::TexCubeFloatFloat:
+ Opc = NVPTX::TEX_CUBE_F32_F32;
+ break;
+ case NVPTXISD::TexCubeFloatFloatLevel:
+ Opc = NVPTX::TEX_CUBE_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexCubeS32Float:
+ Opc = NVPTX::TEX_CUBE_S32_F32;
+ break;
+ case NVPTXISD::TexCubeS32FloatLevel:
+ Opc = NVPTX::TEX_CUBE_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexCubeU32Float:
+ Opc = NVPTX::TEX_CUBE_U32_F32;
+ break;
+ case NVPTXISD::TexCubeU32FloatLevel:
+ Opc = NVPTX::TEX_CUBE_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexCubeArrayFloatFloat:
+ Opc = NVPTX::TEX_CUBE_ARRAY_F32_F32;
+ break;
+ case NVPTXISD::TexCubeArrayFloatFloatLevel:
+ Opc = NVPTX::TEX_CUBE_ARRAY_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexCubeArrayS32Float:
+ Opc = NVPTX::TEX_CUBE_ARRAY_S32_F32;
+ break;
+ case NVPTXISD::TexCubeArrayS32FloatLevel:
+ Opc = NVPTX::TEX_CUBE_ARRAY_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexCubeArrayU32Float:
+ Opc = NVPTX::TEX_CUBE_ARRAY_U32_F32;
+ break;
+ case NVPTXISD::TexCubeArrayU32FloatLevel:
+ Opc = NVPTX::TEX_CUBE_ARRAY_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tld4R2DFloatFloat:
+ Opc = NVPTX::TLD4_R_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4G2DFloatFloat:
+ Opc = NVPTX::TLD4_G_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4B2DFloatFloat:
+ Opc = NVPTX::TLD4_B_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4A2DFloatFloat:
+ Opc = NVPTX::TLD4_A_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4R2DS64Float:
+ Opc = NVPTX::TLD4_R_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4G2DS64Float:
+ Opc = NVPTX::TLD4_G_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4B2DS64Float:
+ Opc = NVPTX::TLD4_B_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4A2DS64Float:
+ Opc = NVPTX::TLD4_A_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4R2DU64Float:
+ Opc = NVPTX::TLD4_R_2D_U32_F32;
+ break;
+ case NVPTXISD::Tld4G2DU64Float:
+ Opc = NVPTX::TLD4_G_2D_U32_F32;
+ break;
+ case NVPTXISD::Tld4B2DU64Float:
+ Opc = NVPTX::TLD4_B_2D_U32_F32;
+ break;
+ case NVPTXISD::Tld4A2DU64Float:
+ Opc = NVPTX::TLD4_A_2D_U32_F32;
+ break;
+ case NVPTXISD::TexUnified1DFloatS32:
+ Opc = NVPTX::TEX_UNIFIED_1D_F32_S32;
+ break;
+ case NVPTXISD::TexUnified1DFloatFloat:
+ Opc = NVPTX::TEX_UNIFIED_1D_F32_F32;
+ break;
+ case NVPTXISD::TexUnified1DFloatFloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_1D_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified1DFloatFloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_1D_F32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified1DS32S32:
+ Opc = NVPTX::TEX_UNIFIED_1D_S32_S32;
+ break;
+ case NVPTXISD::TexUnified1DS32Float:
+ Opc = NVPTX::TEX_UNIFIED_1D_S32_F32;
+ break;
+ case NVPTXISD::TexUnified1DS32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_1D_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified1DS32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_1D_S32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified1DU32S32:
+ Opc = NVPTX::TEX_UNIFIED_1D_U32_S32;
+ break;
+ case NVPTXISD::TexUnified1DU32Float:
+ Opc = NVPTX::TEX_UNIFIED_1D_U32_F32;
+ break;
+ case NVPTXISD::TexUnified1DU32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_1D_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified1DU32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_1D_U32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified1DArrayFloatS32:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_F32_S32;
+ break;
+ case NVPTXISD::TexUnified1DArrayFloatFloat:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_F32_F32;
+ break;
+ case NVPTXISD::TexUnified1DArrayFloatFloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified1DArrayFloatFloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_F32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified1DArrayS32S32:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_S32_S32;
+ break;
+ case NVPTXISD::TexUnified1DArrayS32Float:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_S32_F32;
+ break;
+ case NVPTXISD::TexUnified1DArrayS32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified1DArrayS32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_S32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified1DArrayU32S32:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_U32_S32;
+ break;
+ case NVPTXISD::TexUnified1DArrayU32Float:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_U32_F32;
+ break;
+ case NVPTXISD::TexUnified1DArrayU32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified1DArrayU32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_1D_ARRAY_U32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified2DFloatS32:
+ Opc = NVPTX::TEX_UNIFIED_2D_F32_S32;
+ break;
+ case NVPTXISD::TexUnified2DFloatFloat:
+ Opc = NVPTX::TEX_UNIFIED_2D_F32_F32;
+ break;
+ case NVPTXISD::TexUnified2DFloatFloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_2D_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified2DFloatFloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_2D_F32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified2DS32S32:
+ Opc = NVPTX::TEX_UNIFIED_2D_S32_S32;
+ break;
+ case NVPTXISD::TexUnified2DS32Float:
+ Opc = NVPTX::TEX_UNIFIED_2D_S32_F32;
+ break;
+ case NVPTXISD::TexUnified2DS32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_2D_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified2DS32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_2D_S32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified2DU32S32:
+ Opc = NVPTX::TEX_UNIFIED_2D_U32_S32;
+ break;
+ case NVPTXISD::TexUnified2DU32Float:
+ Opc = NVPTX::TEX_UNIFIED_2D_U32_F32;
+ break;
+ case NVPTXISD::TexUnified2DU32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_2D_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified2DU32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_2D_U32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified2DArrayFloatS32:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_F32_S32;
+ break;
+ case NVPTXISD::TexUnified2DArrayFloatFloat:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_F32_F32;
+ break;
+ case NVPTXISD::TexUnified2DArrayFloatFloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified2DArrayFloatFloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_F32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified2DArrayS32S32:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_S32_S32;
+ break;
+ case NVPTXISD::TexUnified2DArrayS32Float:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_S32_F32;
+ break;
+ case NVPTXISD::TexUnified2DArrayS32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified2DArrayS32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_S32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified2DArrayU32S32:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_U32_S32;
+ break;
+ case NVPTXISD::TexUnified2DArrayU32Float:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_U32_F32;
+ break;
+ case NVPTXISD::TexUnified2DArrayU32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified2DArrayU32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_2D_ARRAY_U32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified3DFloatS32:
+ Opc = NVPTX::TEX_UNIFIED_3D_F32_S32;
+ break;
+ case NVPTXISD::TexUnified3DFloatFloat:
+ Opc = NVPTX::TEX_UNIFIED_3D_F32_F32;
+ break;
+ case NVPTXISD::TexUnified3DFloatFloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_3D_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified3DFloatFloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_3D_F32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified3DS32S32:
+ Opc = NVPTX::TEX_UNIFIED_3D_S32_S32;
+ break;
+ case NVPTXISD::TexUnified3DS32Float:
+ Opc = NVPTX::TEX_UNIFIED_3D_S32_F32;
+ break;
+ case NVPTXISD::TexUnified3DS32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_3D_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified3DS32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_3D_S32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnified3DU32S32:
+ Opc = NVPTX::TEX_UNIFIED_3D_U32_S32;
+ break;
+ case NVPTXISD::TexUnified3DU32Float:
+ Opc = NVPTX::TEX_UNIFIED_3D_U32_F32;
+ break;
+ case NVPTXISD::TexUnified3DU32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_3D_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnified3DU32FloatGrad:
+ Opc = NVPTX::TEX_UNIFIED_3D_U32_F32_GRAD;
+ break;
+ case NVPTXISD::TexUnifiedCubeFloatFloat:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_F32_F32;
+ break;
+ case NVPTXISD::TexUnifiedCubeFloatFloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnifiedCubeS32Float:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_S32_F32;
+ break;
+ case NVPTXISD::TexUnifiedCubeS32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnifiedCubeU32Float:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_U32_F32;
+ break;
+ case NVPTXISD::TexUnifiedCubeU32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnifiedCubeArrayFloatFloat:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_ARRAY_F32_F32;
+ break;
+ case NVPTXISD::TexUnifiedCubeArrayFloatFloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_ARRAY_F32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnifiedCubeArrayS32Float:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_ARRAY_S32_F32;
+ break;
+ case NVPTXISD::TexUnifiedCubeArrayS32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_ARRAY_S32_F32_LEVEL;
+ break;
+ case NVPTXISD::TexUnifiedCubeArrayU32Float:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_ARRAY_U32_F32;
+ break;
+ case NVPTXISD::TexUnifiedCubeArrayU32FloatLevel:
+ Opc = NVPTX::TEX_UNIFIED_CUBE_ARRAY_U32_F32_LEVEL;
+ break;
+ case NVPTXISD::Tld4UnifiedR2DFloatFloat:
+ Opc = NVPTX::TLD4_UNIFIED_R_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedG2DFloatFloat:
+ Opc = NVPTX::TLD4_UNIFIED_G_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedB2DFloatFloat:
+ Opc = NVPTX::TLD4_UNIFIED_B_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedA2DFloatFloat:
+ Opc = NVPTX::TLD4_UNIFIED_A_2D_F32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedR2DS64Float:
+ Opc = NVPTX::TLD4_UNIFIED_R_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedG2DS64Float:
+ Opc = NVPTX::TLD4_UNIFIED_G_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedB2DS64Float:
+ Opc = NVPTX::TLD4_UNIFIED_B_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedA2DS64Float:
+ Opc = NVPTX::TLD4_UNIFIED_A_2D_S32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedR2DU64Float:
+ Opc = NVPTX::TLD4_UNIFIED_R_2D_U32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedG2DU64Float:
+ Opc = NVPTX::TLD4_UNIFIED_G_2D_U32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedB2DU64Float:
+ Opc = NVPTX::TLD4_UNIFIED_B_2D_U32_F32;
+ break;
+ case NVPTXISD::Tld4UnifiedA2DU64Float:
+ Opc = NVPTX::TLD4_UNIFIED_A_2D_U32_F32;
+ break;
+ }
+
+ // Copy over operands
+ for (unsigned i = 1; i < N->getNumOperands(); ++i) {
+ Ops.push_back(N->getOperand(i));
+ }
+
+ Ops.push_back(Chain);
+ Ret = CurDAG->getMachineNode(Opc, SDLoc(N), N->getVTList(), Ops);
+ return Ret;
+}
+
+SDNode *NVPTXDAGToDAGISel::SelectSurfaceIntrinsic(SDNode *N) {
+ SDValue Chain = N->getOperand(0);
+ SDValue TexHandle = N->getOperand(1);
+ SDNode *Ret = nullptr;
+ unsigned Opc = 0;
+ SmallVector<SDValue, 8> Ops;
+ switch (N->getOpcode()) {
+ default: return nullptr;
+ case NVPTXISD::Suld1DI8Clamp:
+ Opc = NVPTX::SULD_1D_I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI16Clamp:
+ Opc = NVPTX::SULD_1D_I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI32Clamp:
+ Opc = NVPTX::SULD_1D_I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI64Clamp:
+ Opc = NVPTX::SULD_1D_I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I8Clamp:
+ Opc = NVPTX::SULD_1D_V2I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I16Clamp:
+ Opc = NVPTX::SULD_1D_V2I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I32Clamp:
+ Opc = NVPTX::SULD_1D_V2I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I64Clamp:
+ Opc = NVPTX::SULD_1D_V2I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I8Clamp:
+ Opc = NVPTX::SULD_1D_V4I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I16Clamp:
+ Opc = NVPTX::SULD_1D_V4I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I32Clamp:
+ Opc = NVPTX::SULD_1D_V4I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI8Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI16Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI32Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI64Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I8Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I16Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I32Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I64Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I8Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I16Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I32Clamp:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI8Clamp:
+ Opc = NVPTX::SULD_2D_I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI16Clamp:
+ Opc = NVPTX::SULD_2D_I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI32Clamp:
+ Opc = NVPTX::SULD_2D_I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI64Clamp:
+ Opc = NVPTX::SULD_2D_I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I8Clamp:
+ Opc = NVPTX::SULD_2D_V2I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I16Clamp:
+ Opc = NVPTX::SULD_2D_V2I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I32Clamp:
+ Opc = NVPTX::SULD_2D_V2I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I64Clamp:
+ Opc = NVPTX::SULD_2D_V2I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I8Clamp:
+ Opc = NVPTX::SULD_2D_V4I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I16Clamp:
+ Opc = NVPTX::SULD_2D_V4I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I32Clamp:
+ Opc = NVPTX::SULD_2D_V4I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI8Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI16Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI32Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI64Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I8Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I16Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I32Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I64Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I8Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I16Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I32Clamp:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI8Clamp:
+ Opc = NVPTX::SULD_3D_I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI16Clamp:
+ Opc = NVPTX::SULD_3D_I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI32Clamp:
+ Opc = NVPTX::SULD_3D_I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI64Clamp:
+ Opc = NVPTX::SULD_3D_I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I8Clamp:
+ Opc = NVPTX::SULD_3D_V2I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I16Clamp:
+ Opc = NVPTX::SULD_3D_V2I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I32Clamp:
+ Opc = NVPTX::SULD_3D_V2I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I64Clamp:
+ Opc = NVPTX::SULD_3D_V2I64_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I8Clamp:
+ Opc = NVPTX::SULD_3D_V4I8_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I16Clamp:
+ Opc = NVPTX::SULD_3D_V4I16_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I32Clamp:
+ Opc = NVPTX::SULD_3D_V4I32_CLAMP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI8Trap:
+ Opc = NVPTX::SULD_1D_I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI16Trap:
+ Opc = NVPTX::SULD_1D_I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI32Trap:
+ Opc = NVPTX::SULD_1D_I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI64Trap:
+ Opc = NVPTX::SULD_1D_I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I8Trap:
+ Opc = NVPTX::SULD_1D_V2I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I16Trap:
+ Opc = NVPTX::SULD_1D_V2I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I32Trap:
+ Opc = NVPTX::SULD_1D_V2I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I64Trap:
+ Opc = NVPTX::SULD_1D_V2I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I8Trap:
+ Opc = NVPTX::SULD_1D_V4I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I16Trap:
+ Opc = NVPTX::SULD_1D_V4I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I32Trap:
+ Opc = NVPTX::SULD_1D_V4I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI8Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI16Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI32Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI64Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I8Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I16Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I32Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I64Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I8Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I16Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I32Trap:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI8Trap:
+ Opc = NVPTX::SULD_2D_I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI16Trap:
+ Opc = NVPTX::SULD_2D_I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI32Trap:
+ Opc = NVPTX::SULD_2D_I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI64Trap:
+ Opc = NVPTX::SULD_2D_I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I8Trap:
+ Opc = NVPTX::SULD_2D_V2I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I16Trap:
+ Opc = NVPTX::SULD_2D_V2I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I32Trap:
+ Opc = NVPTX::SULD_2D_V2I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I64Trap:
+ Opc = NVPTX::SULD_2D_V2I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I8Trap:
+ Opc = NVPTX::SULD_2D_V4I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I16Trap:
+ Opc = NVPTX::SULD_2D_V4I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I32Trap:
+ Opc = NVPTX::SULD_2D_V4I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI8Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI16Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI32Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI64Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I8Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I16Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I32Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I64Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I8Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I16Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I32Trap:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI8Trap:
+ Opc = NVPTX::SULD_3D_I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI16Trap:
+ Opc = NVPTX::SULD_3D_I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI32Trap:
+ Opc = NVPTX::SULD_3D_I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI64Trap:
+ Opc = NVPTX::SULD_3D_I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I8Trap:
+ Opc = NVPTX::SULD_3D_V2I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I16Trap:
+ Opc = NVPTX::SULD_3D_V2I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I32Trap:
+ Opc = NVPTX::SULD_3D_V2I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I64Trap:
+ Opc = NVPTX::SULD_3D_V2I64_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I8Trap:
+ Opc = NVPTX::SULD_3D_V4I8_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I16Trap:
+ Opc = NVPTX::SULD_3D_V4I16_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I32Trap:
+ Opc = NVPTX::SULD_3D_V4I32_TRAP;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI8Zero:
+ Opc = NVPTX::SULD_1D_I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI16Zero:
+ Opc = NVPTX::SULD_1D_I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI32Zero:
+ Opc = NVPTX::SULD_1D_I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DI64Zero:
+ Opc = NVPTX::SULD_1D_I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I8Zero:
+ Opc = NVPTX::SULD_1D_V2I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I16Zero:
+ Opc = NVPTX::SULD_1D_V2I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I32Zero:
+ Opc = NVPTX::SULD_1D_V2I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV2I64Zero:
+ Opc = NVPTX::SULD_1D_V2I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I8Zero:
+ Opc = NVPTX::SULD_1D_V4I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I16Zero:
+ Opc = NVPTX::SULD_1D_V4I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DV4I32Zero:
+ Opc = NVPTX::SULD_1D_V4I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI8Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI16Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI32Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayI64Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I8Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I16Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I32Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV2I64Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_V2I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I8Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I16Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld1DArrayV4I32Zero:
+ Opc = NVPTX::SULD_1D_ARRAY_V4I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI8Zero:
+ Opc = NVPTX::SULD_2D_I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI16Zero:
+ Opc = NVPTX::SULD_2D_I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI32Zero:
+ Opc = NVPTX::SULD_2D_I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DI64Zero:
+ Opc = NVPTX::SULD_2D_I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I8Zero:
+ Opc = NVPTX::SULD_2D_V2I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I16Zero:
+ Opc = NVPTX::SULD_2D_V2I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I32Zero:
+ Opc = NVPTX::SULD_2D_V2I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV2I64Zero:
+ Opc = NVPTX::SULD_2D_V2I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I8Zero:
+ Opc = NVPTX::SULD_2D_V4I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I16Zero:
+ Opc = NVPTX::SULD_2D_V4I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DV4I32Zero:
+ Opc = NVPTX::SULD_2D_V4I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI8Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI16Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI32Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayI64Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I8Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I16Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I32Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV2I64Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_V2I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I8Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I16Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld2DArrayV4I32Zero:
+ Opc = NVPTX::SULD_2D_ARRAY_V4I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI8Zero:
+ Opc = NVPTX::SULD_3D_I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI16Zero:
+ Opc = NVPTX::SULD_3D_I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI32Zero:
+ Opc = NVPTX::SULD_3D_I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DI64Zero:
+ Opc = NVPTX::SULD_3D_I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I8Zero:
+ Opc = NVPTX::SULD_3D_V2I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I16Zero:
+ Opc = NVPTX::SULD_3D_V2I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I32Zero:
+ Opc = NVPTX::SULD_3D_V2I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV2I64Zero:
+ Opc = NVPTX::SULD_3D_V2I64_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I8Zero:
+ Opc = NVPTX::SULD_3D_V4I8_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I16Zero:
+ Opc = NVPTX::SULD_3D_V4I16_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ case NVPTXISD::Suld3DV4I32Zero:
+ Opc = NVPTX::SULD_3D_V4I32_ZERO;
+ Ops.push_back(TexHandle);
+ Ops.push_back(N->getOperand(2));
+ Ops.push_back(N->getOperand(3));
+ Ops.push_back(N->getOperand(4));
+ Ops.push_back(Chain);
+ break;
+ }
+ Ret = CurDAG->getMachineNode(Opc, SDLoc(N), N->getVTList(), Ops);
+ return Ret;
+}
+
+
+/// SelectBFE - Look for instruction sequences that can be made more efficient
+/// by using the 'bfe' (bit-field extract) PTX instruction
+SDNode *NVPTXDAGToDAGISel::SelectBFE(SDNode *N) {
+ SDValue LHS = N->getOperand(0);
+ SDValue RHS = N->getOperand(1);
+ SDValue Len;
+ SDValue Start;
+ SDValue Val;
+ bool IsSigned = false;
+
+ if (N->getOpcode() == ISD::AND) {
+ // Canonicalize the operands
+ // We want 'and %val, %mask'
+ if (isa<ConstantSDNode>(LHS) && !isa<ConstantSDNode>(RHS)) {
+ std::swap(LHS, RHS);
+ }
+
+ ConstantSDNode *Mask = dyn_cast<ConstantSDNode>(RHS);
+ if (!Mask) {
+ // We need a constant mask on the RHS of the AND
+ return NULL;
+ }
+
+ // Extract the mask bits
+ uint64_t MaskVal = Mask->getZExtValue();
+ if (!isMask_64(MaskVal)) {
+ // We *could* handle shifted masks here, but doing so would require an
+ // 'and' operation to fix up the low-order bits so we would trade
+ // shr+and for bfe+and, which has the same throughput
+ return NULL;
+ }
+
+ // How many bits are in our mask?
+ uint64_t NumBits = countTrailingOnes(MaskVal);
+ Len = CurDAG->getTargetConstant(NumBits, MVT::i32);
+
+ if (LHS.getOpcode() == ISD::SRL || LHS.getOpcode() == ISD::SRA) {
+ // We have a 'srl/and' pair, extract the effective start bit and length
+ Val = LHS.getNode()->getOperand(0);
+ Start = LHS.getNode()->getOperand(1);
+ ConstantSDNode *StartConst = dyn_cast<ConstantSDNode>(Start);
+ if (StartConst) {
+ uint64_t StartVal = StartConst->getZExtValue();
+ // How many "good" bits do we have left? "good" is defined here as bits
+ // that exist in the original value, not shifted in.
+ uint64_t GoodBits = Start.getValueType().getSizeInBits() - StartVal;
+ if (NumBits > GoodBits) {
+ // Do not handle the case where bits have been shifted in. In theory
+ // we could handle this, but the cost is likely higher than just
+ // emitting the srl/and pair.
+ return NULL;
+ }
+ Start = CurDAG->getTargetConstant(StartVal, MVT::i32);
+ } else {
+ // Do not handle the case where the shift amount (can be zero if no srl
+ // was found) is not constant. We could handle this case, but it would
+ // require run-time logic that would be more expensive than just
+ // emitting the srl/and pair.
+ return NULL;
+ }
+ } else {
+ // Do not handle the case where the LHS of the and is not a shift. While
+ // it would be trivial to handle this case, it would just transform
+ // 'and' -> 'bfe', but 'and' has higher-throughput.
+ return NULL;
+ }
+ } else if (N->getOpcode() == ISD::SRL || N->getOpcode() == ISD::SRA) {
+ if (LHS->getOpcode() == ISD::AND) {
+ ConstantSDNode *ShiftCnst = dyn_cast<ConstantSDNode>(RHS);
+ if (!ShiftCnst) {
+ // Shift amount must be constant
+ return NULL;
+ }
+
+ uint64_t ShiftAmt = ShiftCnst->getZExtValue();
+
+ SDValue AndLHS = LHS->getOperand(0);
+ SDValue AndRHS = LHS->getOperand(1);
+
+ // Canonicalize the AND to have the mask on the RHS
+ if (isa<ConstantSDNode>(AndLHS)) {
+ std::swap(AndLHS, AndRHS);
+ }
+
+ ConstantSDNode *MaskCnst = dyn_cast<ConstantSDNode>(AndRHS);
+ if (!MaskCnst) {
+ // Mask must be constant
+ return NULL;
+ }
+
+ uint64_t MaskVal = MaskCnst->getZExtValue();
+ uint64_t NumZeros;
+ uint64_t NumBits;
+ if (isMask_64(MaskVal)) {
+ NumZeros = 0;
+ // The number of bits in the result bitfield will be the number of
+ // trailing ones (the AND) minus the number of bits we shift off
+ NumBits = countTrailingOnes(MaskVal) - ShiftAmt;
+ } else if (isShiftedMask_64(MaskVal)) {
+ NumZeros = countTrailingZeros(MaskVal);
+ unsigned NumOnes = countTrailingOnes(MaskVal >> NumZeros);
+ // The number of bits in the result bitfield will be the number of
+ // trailing zeros plus the number of set bits in the mask minus the
+ // number of bits we shift off
+ NumBits = NumZeros + NumOnes - ShiftAmt;
+ } else {
+ // This is not a mask we can handle
+ return NULL;
+ }
+
+ if (ShiftAmt < NumZeros) {
+ // Handling this case would require extra logic that would make this
+ // transformation non-profitable
+ return NULL;
+ }
+
+ Val = AndLHS;
+ Start = CurDAG->getTargetConstant(ShiftAmt, MVT::i32);
+ Len = CurDAG->getTargetConstant(NumBits, MVT::i32);
+ } else if (LHS->getOpcode() == ISD::SHL) {
+ // Here, we have a pattern like:
+ //
+ // (sra (shl val, NN), MM)
+ // or
+ // (srl (shl val, NN), MM)
+ //
+ // If MM >= NN, we can efficiently optimize this with bfe
+ Val = LHS->getOperand(0);
+
+ SDValue ShlRHS = LHS->getOperand(1);
+ ConstantSDNode *ShlCnst = dyn_cast<ConstantSDNode>(ShlRHS);
+ if (!ShlCnst) {
+ // Shift amount must be constant
+ return NULL;
+ }
+ uint64_t InnerShiftAmt = ShlCnst->getZExtValue();
+
+ SDValue ShrRHS = RHS;
+ ConstantSDNode *ShrCnst = dyn_cast<ConstantSDNode>(ShrRHS);
+ if (!ShrCnst) {
+ // Shift amount must be constant
+ return NULL;
+ }
+ uint64_t OuterShiftAmt = ShrCnst->getZExtValue();
+
+ // To avoid extra codegen and be profitable, we need Outer >= Inner
+ if (OuterShiftAmt < InnerShiftAmt) {
+ return NULL;
+ }
+
+ // If the outer shift is more than the type size, we have no bitfield to
+ // extract (since we also check that the inner shift is <= the outer shift
+ // then this also implies that the inner shift is < the type size)
+ if (OuterShiftAmt >= Val.getValueType().getSizeInBits()) {
+ return NULL;
+ }
+
+ Start =
+ CurDAG->getTargetConstant(OuterShiftAmt - InnerShiftAmt, MVT::i32);
+ Len =
+ CurDAG->getTargetConstant(Val.getValueType().getSizeInBits() -
+ OuterShiftAmt, MVT::i32);
+
+ if (N->getOpcode() == ISD::SRA) {
+ // If we have a arithmetic right shift, we need to use the signed bfe
+ // variant
+ IsSigned = true;
+ }
+ } else {
+ // No can do...
+ return NULL;
+ }
+ } else {
+ // No can do...
+ return NULL;
+ }
+
+
+ unsigned Opc;
+ // For the BFE operations we form here from "and" and "srl", always use the
+ // unsigned variants.
+ if (Val.getValueType() == MVT::i32) {
+ if (IsSigned) {
+ Opc = NVPTX::BFE_S32rii;
+ } else {
+ Opc = NVPTX::BFE_U32rii;
+ }
+ } else if (Val.getValueType() == MVT::i64) {
+ if (IsSigned) {
+ Opc = NVPTX::BFE_S64rii;
+ } else {
+ Opc = NVPTX::BFE_U64rii;
+ }
+ } else {
+ // We cannot handle this type
+ return NULL;
+ }
+
+ SDValue Ops[] = {
+ Val, Start, Len
+ };
+
+ SDNode *Ret =
+ CurDAG->getMachineNode(Opc, SDLoc(N), N->getVTList(), Ops);
+
+ return Ret;
+}
+
+// SelectDirectAddr - Match a direct address for DAG.
+// A direct address could be a globaladdress or externalsymbol.
+bool NVPTXDAGToDAGISel::SelectDirectAddr(SDValue N, SDValue &Address) {
+ // Return true if TGA or ES.
+ if (N.getOpcode() == ISD::TargetGlobalAddress ||
+ N.getOpcode() == ISD::TargetExternalSymbol) {
+ Address = N;
+ return true;
+ }
+ if (N.getOpcode() == NVPTXISD::Wrapper) {
+ Address = N.getOperand(0);
+ return true;
+ }
+ if (N.getOpcode() == ISD::INTRINSIC_WO_CHAIN) {
+ unsigned IID = cast<ConstantSDNode>(N.getOperand(0))->getZExtValue();
+ if (IID == Intrinsic::nvvm_ptr_gen_to_param)
+ if (N.getOperand(1).getOpcode() == NVPTXISD::MoveParam)
+ return (SelectDirectAddr(N.getOperand(1).getOperand(0), Address));
+ }
+ return false;
+}
+
+// symbol+offset
+bool NVPTXDAGToDAGISel::SelectADDRsi_imp(
+ SDNode *OpNode, SDValue Addr, SDValue &Base, SDValue &Offset, MVT mvt) {
+ if (Addr.getOpcode() == ISD::ADD) {
+ if (ConstantSDNode *CN = dyn_cast<ConstantSDNode>(Addr.getOperand(1))) {
+ SDValue base = Addr.getOperand(0);
+ if (SelectDirectAddr(base, Base)) {
+ Offset = CurDAG->getTargetConstant(CN->getZExtValue(), mvt);
+ return true;
+ }
+ }
+ }
+ return false;
+}
+
+// symbol+offset
+bool NVPTXDAGToDAGISel::SelectADDRsi(SDNode *OpNode, SDValue Addr,
+ SDValue &Base, SDValue &Offset) {
+ return SelectADDRsi_imp(OpNode, Addr, Base, Offset, MVT::i32);
+}
+
+// symbol+offset
+bool NVPTXDAGToDAGISel::SelectADDRsi64(SDNode *OpNode, SDValue Addr,
+ SDValue &Base, SDValue &Offset) {
+ return SelectADDRsi_imp(OpNode, Addr, Base, Offset, MVT::i64);
+}
+
+// register+offset
+bool NVPTXDAGToDAGISel::SelectADDRri_imp(
+ SDNode *OpNode, SDValue Addr, SDValue &Base, SDValue &Offset, MVT mvt) {
+ if (FrameIndexSDNode *FIN = dyn_cast<FrameIndexSDNode>(Addr)) {
+ Base = CurDAG->getTargetFrameIndex(FIN->getIndex(), mvt);
+ Offset = CurDAG->getTargetConstant(0, mvt);
+ return true;
+ }
+ if (Addr.getOpcode() == ISD::TargetExternalSymbol ||
+ Addr.getOpcode() == ISD::TargetGlobalAddress)
+ return false; // direct calls.