#ifndef ARMISELLOWERING_H
#define ARMISELLOWERING_H
+#include "ARM.h"
#include "ARMSubtarget.h"
-#include "llvm/Target/TargetLowering.h"
-#include "llvm/Target/TargetRegisterInfo.h"
+#include "llvm/CodeGen/CallingConvLower.h"
#include "llvm/CodeGen/FastISel.h"
#include "llvm/CodeGen/SelectionDAG.h"
-#include "llvm/CodeGen/CallingConvLower.h"
+#include "llvm/Target/TargetLowering.h"
+#include "llvm/Target/TargetRegisterInfo.h"
+#include "llvm/Target/TargetTransformImpl.h"
#include <vector>
namespace llvm {
// PIC mode.
WrapperJT, // WrapperJT - A wrapper node for TargetJumpTable
+ // Add pseudo op to model memcpy for struct byval.
+ COPY_STRUCT_BYVAL,
+
CALL, // Function call.
CALL_PRED, // Function call that's predicable.
CALL_NOLINK, // Function call with branch not branch-and-link.
PIC_ADD, // Add with a PC operand and a PIC label.
CMP, // ARM compare instructions.
+ CMN, // ARM CMN instructions.
CMPZ, // ARM compare that sets only Z flag.
CMPFP, // ARM VFP compare instruction, sets FPSCR.
CMPFPw0, // ARM VFP compare against zero instruction, sets FPSCR.
FMSTAT, // ARM fmstat instruction.
+
CMOV, // ARM conditional move instructions.
BCC_i64,
SRA_FLAG, // V,Flag = sra_flag X -> sra X, 1 + save carry out.
RRX, // V = RRX X, Flag -> srl X, 1 + shift in carry flag.
+ ADDC, // Add with carry
+ ADDE, // Add using carry
+ SUBC, // Sub with carry
+ SUBE, // Sub using carry
+
VMOVRRD, // double to two gprs.
VMOVDRR, // Two gprs to double.
EH_SJLJ_SETJMP, // SjLj exception handling setjmp.
EH_SJLJ_LONGJMP, // SjLj exception handling longjmp.
- EH_SJLJ_DISPATCHSETUP, // SjLj exception handling dispatch setup.
TC_RETURN, // Tail call return pseudo.
MEMBARRIER_MCR, // Memory barrier (MCR)
PRELOAD, // Preload
-
+
VCEQ, // Vector compare equal.
VCEQZ, // Vector compare equal to zero.
VCGE, // Vector compare greater than or equal.
VMOVIMM,
VMVNIMM,
+ // Vector move f32 immediate:
+ VMOVFPIMM,
+
// Vector duplicate:
VDUP,
VDUPLANE,
VMULLs, // ...signed
VMULLu, // ...unsigned
+ // Vector add long:
+ VADDLs, // ...signed
+ VADDLu, // ...unsigned
+
+ // Vector subtract long:
+ VSUBLs, // ...signed
+ VSUBLu, // ...unsigned
+
+ UMLAL, // 64bit Unsigned Accumulate Multiply
+ SMLAL, // 64bit Signed Accumulate Multiply
+
// Operands of the standard BUILD_VECTOR node are not legalized, which
// is fine if BUILD_VECTORs are always lowered to shuffles or other
// operations, but for ARM some BUILD_VECTORs are legal as-is and their
// Bit-field insert
BFI,
-
+
// Vector OR with immediate
VORRIMM,
// Vector AND with NOT of immediate
VBICIMM,
+ // Vector bitwise select
+ VBSL,
+
// Vector load N-element structure to all lanes:
VLD2DUP = ISD::FIRST_TARGET_MEMORY_OPCODE,
VLD3DUP,
VST4_UPD,
VST2LN_UPD,
VST3LN_UPD,
- VST4LN_UPD
+ VST4LN_UPD,
+
+ // 64-bit atomic ops (value split into two registers)
+ ATOMADD64_DAG,
+ ATOMSUB64_DAG,
+ ATOMOR64_DAG,
+ ATOMXOR64_DAG,
+ ATOMAND64_DAG,
+ ATOMNAND64_DAG,
+ ATOMSWAP64_DAG,
+ ATOMCMPXCHG64_DAG,
+ ATOMMIN64_DAG,
+ ATOMUMIN64_DAG,
+ ATOMMAX64_DAG,
+ ATOMUMAX64_DAG
};
}
/// Define some predicates that are used for node matching.
namespace ARM {
- /// getVFPf32Imm / getVFPf64Imm - If the given fp immediate can be
- /// materialized with a VMOV.f32 / VMOV.f64 (i.e. fconsts / fconstd)
- /// instruction, returns its 8-bit integer representation. Otherwise,
- /// returns -1.
- int getVFPf32Imm(const APFloat &FPImm);
- int getVFPf64Imm(const APFloat &FPImm);
bool isBitFieldInvertedMask(unsigned v);
}
public:
explicit ARMTargetLowering(TargetMachine &TM);
- virtual unsigned getJumpTableEncoding(void) const;
+ virtual unsigned getJumpTableEncoding() const;
virtual SDValue LowerOperation(SDValue Op, SelectionDAG &DAG) const;
virtual const char *getTargetNodeName(unsigned Opcode) const;
+ virtual bool isSelectSupported(SelectSupportKind Kind) const {
+ // ARM does not support scalar condition selects on vectors.
+ return (Kind != ScalarCondVectorVal);
+ }
+
+ /// getSetCCResultType - Return the value type to use for ISD::SETCC.
+ virtual EVT getSetCCResultType(EVT VT) const;
+
virtual MachineBasicBlock *
EmitInstrWithCustomInserter(MachineInstr *MI,
MachineBasicBlock *MBB) const;
+ virtual void
+ AdjustInstrPostInstrSelection(MachineInstr *MI, SDNode *Node) const;
+
+ SDValue PerformCMOVCombine(SDNode *N, SelectionDAG &DAG) const;
virtual SDValue PerformDAGCombine(SDNode *N, DAGCombinerInfo &DCI) const;
bool isDesirableToTransformToIntegerOp(unsigned Opc, EVT VT) const;
/// allowsUnalignedMemoryAccesses - Returns true if the target allows
- /// unaligned memory accesses. of the specified type.
- /// FIXME: Add getOptimalMemOpType to implement memcpy with NEON?
- virtual bool allowsUnalignedMemoryAccesses(EVT VT) const;
+ /// unaligned memory accesses of the specified type. Returns whether it
+ /// is "fast" by reference in the second argument.
+ virtual bool allowsUnalignedMemoryAccesses(EVT VT, bool *Fast) const;
+
+ virtual EVT getOptimalMemOpType(uint64_t Size,
+ unsigned DstAlign, unsigned SrcAlign,
+ bool IsMemset, bool ZeroMemset,
+ bool MemcpyStrSrc,
+ MachineFunction &MF) const;
+
+ using TargetLowering::isZExtFree;
+ virtual bool isZExtFree(SDValue Val, EVT VT2) const;
/// isLegalAddressingMode - Return true if the addressing mode represented
/// by AM is legal for this target, for a load/store of the specified type.
- virtual bool isLegalAddressingMode(const AddrMode &AM, const Type *Ty)const;
+ virtual bool isLegalAddressingMode(const AddrMode &AM, Type *Ty)const;
bool isLegalT2ScaledAddressingMode(const AddrMode &AM, EVT VT) const;
/// isLegalICmpImmediate - Return true if the specified immediate is legal
/// the immediate into a register.
virtual bool isLegalICmpImmediate(int64_t Imm) const;
+ /// isLegalAddImmediate - Return true if the specified immediate is legal
+ /// add immediate, that is the target has add instructions which can
+ /// add a register and the immediate without having to materialize
+ /// the immediate into a register.
+ virtual bool isLegalAddImmediate(int64_t Imm) const;
+
/// getPreIndexedAddressParts - returns true by value, base pointer and
/// offset pointer and addressing mode by reference if the node's address
/// can be legally represented as pre-indexed load / store address.
SelectionDAG &DAG) const;
virtual void computeMaskedBitsForTargetNode(const SDValue Op,
- const APInt &Mask,
APInt &KnownZero,
APInt &KnownOne,
const SelectionDAG &DAG,
std::pair<unsigned, const TargetRegisterClass*>
getRegForInlineAsmConstraint(const std::string &Constraint,
EVT VT) const;
- std::vector<unsigned>
- getRegClassForInlineAsmConstraint(const std::string &Constraint,
- EVT VT) const;
/// LowerAsmOperandForConstraint - Lower the specified operand into the Ops
/// vector. If it is invalid, don't add anything to Ops. If hasMemory is
/// true it means one of the asm constraint of the inline asm instruction
/// being processed is 'm'.
virtual void LowerAsmOperandForConstraint(SDValue Op,
- char ConstraintLetter,
+ std::string &Constraint,
std::vector<SDValue> &Ops,
SelectionDAG &DAG) const;
/// getRegClassFor - Return the register class that should be used for the
/// specified value type.
- virtual TargetRegisterClass *getRegClassFor(EVT VT) const;
-
- /// getFunctionAlignment - Return the Log2 alignment of this function.
- virtual unsigned getFunctionAlignment(const Function *F) const;
+ virtual const TargetRegisterClass *getRegClassFor(MVT VT) const;
/// getMaximalGlobalOffset - Returns the maximal possible offset which can
/// be used for loads / stores from the global.
/// createFastISel - This method returns a target specific FastISel object,
/// or null if the target does not support "fast" ISel.
- virtual FastISel *createFastISel(FunctionLoweringInfo &funcInfo) const;
+ virtual FastISel *createFastISel(FunctionLoweringInfo &funcInfo,
+ const TargetLibraryInfo *libInfo) const;
Sched::Preference getSchedulingPreference(SDNode *N) const;
unsigned Intrinsic) const;
protected:
std::pair<const TargetRegisterClass*, uint8_t>
- findRepresentativeClass(EVT VT) const;
+ findRepresentativeClass(MVT VT) const;
private:
/// Subtarget - Keep a pointer to the ARMSubtarget around so that we can
///
unsigned ARMPCLabelIndex;
- void addTypeForNEON(EVT VT, EVT PromotedLdStVT, EVT PromotedBitwiseVT);
- void addDRTypeForNEON(EVT VT);
- void addQRTypeForNEON(EVT VT);
+ void addTypeForNEON(MVT VT, MVT PromotedLdStVT, MVT PromotedBitwiseVT);
+ void addDRTypeForNEON(MVT VT);
+ void addQRTypeForNEON(MVT VT);
typedef SmallVector<std::pair<unsigned, SDValue>, 8> RegsToPassVector;
void PassF64ArgInRegs(DebugLoc dl, SelectionDAG &DAG,
ISD::ArgFlagsTy Flags) const;
SDValue LowerEH_SJLJ_SETJMP(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerEH_SJLJ_LONGJMP(SDValue Op, SelectionDAG &DAG) const;
- SDValue LowerEH_SJLJ_DISPATCHSETUP(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerINTRINSIC_WO_CHAIN(SDValue Op, SelectionDAG &DAG,
const ARMSubtarget *Subtarget) const;
SDValue LowerBlockAddress(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerToTLSGeneralDynamicModel(GlobalAddressSDNode *GA,
SelectionDAG &DAG) const;
SDValue LowerToTLSExecModels(GlobalAddressSDNode *GA,
- SelectionDAG &DAG) const;
+ SelectionDAG &DAG,
+ TLSModel::Model model) const;
SDValue LowerGLOBAL_OFFSET_TABLE(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerBR_JT(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerSELECT(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerShiftRightParts(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerShiftLeftParts(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerFLT_ROUNDS_(SDValue Op, SelectionDAG &DAG) const;
- SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG,
+ SDValue LowerConstantFP(SDValue Op, SelectionDAG &DAG,
+ const ARMSubtarget *ST) const;
+ SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG,
const ARMSubtarget *ST) const;
SDValue ReconstructShuffle(SDValue Op, SelectionDAG &DAG) const;
DebugLoc dl, SelectionDAG &DAG,
SmallVectorImpl<SDValue> &InVals) const;
+ void VarArgStyleRegisters(CCState &CCInfo, SelectionDAG &DAG,
+ DebugLoc dl, SDValue &Chain,
+ const Value *OrigArg,
+ unsigned OffsetFromOrigArg,
+ unsigned ArgOffset,
+ bool ForceMutable = false)
+ const;
+
+ void computeRegArea(CCState &CCInfo, MachineFunction &MF,
+ unsigned &VARegSize, unsigned &VARegSaveSize) const;
+
virtual SDValue
- LowerCall(SDValue Chain, SDValue Callee,
- CallingConv::ID CallConv, bool isVarArg,
- bool &isTailCall,
- const SmallVectorImpl<ISD::OutputArg> &Outs,
- const SmallVectorImpl<SDValue> &OutVals,
- const SmallVectorImpl<ISD::InputArg> &Ins,
- DebugLoc dl, SelectionDAG &DAG,
+ LowerCall(TargetLowering::CallLoweringInfo &CLI,
SmallVectorImpl<SDValue> &InVals) const;
/// HandleByVal - Target-specific cleanup for ByVal support.
- virtual void HandleByVal(CCState *) const;
+ virtual void HandleByVal(CCState *, unsigned &, unsigned) const;
/// IsEligibleForTailCallOptimization - Check whether the call is eligible
/// for tail call optimization. Targets which want to do tail call
const SmallVectorImpl<SDValue> &OutVals,
const SmallVectorImpl<ISD::InputArg> &Ins,
SelectionDAG& DAG) const;
+
+ virtual bool CanLowerReturn(CallingConv::ID CallConv,
+ MachineFunction &MF, bool isVarArg,
+ const SmallVectorImpl<ISD::OutputArg> &Outs,
+ LLVMContext &Context) const;
+
virtual SDValue
LowerReturn(SDValue Chain,
CallingConv::ID CallConv, bool isVarArg,
const SmallVectorImpl<SDValue> &OutVals,
DebugLoc dl, SelectionDAG &DAG) const;
- virtual bool isUsedByReturnOnly(SDNode *N) const;
+ virtual bool isUsedByReturnOnly(SDNode *N, SDValue &Chain) const;
+
+ virtual bool mayBeEmittedAsTailCall(CallInst *CI) const;
SDValue getARMCmp(SDValue LHS, SDValue RHS, ISD::CondCode CC,
SDValue &ARMcc, SelectionDAG &DAG, DebugLoc dl) const;
MachineBasicBlock *BB,
unsigned Size,
unsigned BinOpcode) const;
-
+ MachineBasicBlock *EmitAtomicBinary64(MachineInstr *MI,
+ MachineBasicBlock *BB,
+ unsigned Op1,
+ unsigned Op2,
+ bool NeedsCarry = false,
+ bool IsCmpxchg = false,
+ bool IsMinMax = false,
+ ARMCC::CondCodes CC = ARMCC::AL) const;
+ MachineBasicBlock * EmitAtomicBinaryMinMax(MachineInstr *MI,
+ MachineBasicBlock *BB,
+ unsigned Size,
+ bool signExtend,
+ ARMCC::CondCodes Cond) const;
+
+ void SetupEntryBlockForSjLj(MachineInstr *MI,
+ MachineBasicBlock *MBB,
+ MachineBasicBlock *DispatchBB, int FI) const;
+
+ MachineBasicBlock *EmitSjLjDispatchBlock(MachineInstr *MI,
+ MachineBasicBlock *MBB) const;
+
+ bool RemapAddSubWithFlags(MachineInstr *MI, MachineBasicBlock *BB) const;
+
+ MachineBasicBlock *EmitStructByval(MachineInstr *MI,
+ MachineBasicBlock *MBB) const;
};
-
+
enum NEONModImmType {
VMOVModImm,
VMVNModImm,
OtherModImm
};
-
-
+
+
namespace ARM {
- FastISel *createFastISel(FunctionLoweringInfo &funcInfo);
+ FastISel *createFastISel(FunctionLoweringInfo &funcInfo,
+ const TargetLibraryInfo *libInfo);
}
+
+ class ARMScalarTargetTransformImpl : public ScalarTargetTransformImpl {
+ const ARMSubtarget *Subtarget;
+ public:
+ explicit ARMScalarTargetTransformImpl(const TargetLowering *TL) :
+ ScalarTargetTransformImpl(TL),
+ Subtarget(&TL->getTargetMachine().getSubtarget<ARMSubtarget>()) {};
+
+ virtual unsigned getIntImmCost(const APInt &Imm, Type *Ty) const;
+ };
}
#endif // ARMISELLOWERING_H
projects / oota-llvm.git / blobdiff