#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 <vector>
namespace llvm {
Wrapper, // Wrapper - A wrapper node for TargetConstantPool,
// TargetExternalSymbol, and TargetGlobalAddress.
- WrapperDYN, // WrapperDYN - A wrapper node for TargetGlobalAddress in
- // DYN mode.
WrapperPIC, // WrapperPIC - A wrapper node for TargetGlobalAddress in
// 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.
BR_JT, // Jumptable branch.
BR2_JT, // Jumptable branch (2 level - jumptable entry is a jump).
RET_FLAG, // Return with a flag operand.
+ INTRET_FLAG, // Interrupt return with an LR-offset and a flag operand.
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,
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.
DYN_ALLOC, // Dynamic allocation on the stack.
- MEMBARRIER, // Memory barrier (DMB)
MEMBARRIER_MCR, // Memory barrier (MCR)
PRELOAD, // Preload
VMOVIMM,
VMVNIMM,
+ // Vector move f32 immediate:
+ VMOVFPIMM,
+
// Vector duplicate:
VDUP,
VDUPLANE,
VMULLs, // ...signed
VMULLu, // ...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
// Floating-point max and min:
FMAX,
FMIN,
+ VMAXNM,
+ VMINNM,
// Bit-field insert
BFI,
VST4_UPD,
VST2LN_UPD,
VST3LN_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
+ VST4LN_UPD
};
}
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 EVT getSetCCResultType(LLVMContext &Context, EVT VT) const;
virtual MachineBasicBlock *
EmitInstrWithCustomInserter(MachineInstr *MI,
bool isDesirableToTransformToIntegerOp(unsigned Opc, EVT VT) const;
/// allowsUnalignedMemoryAccesses - Returns true if the target allows
- /// unaligned memory accesses. of the specified type.
- 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 NonScalarIntSafe,
+ bool IsMemset, bool ZeroMemset,
bool MemcpyStrSrc,
MachineFunction &MF) const;
+ using TargetLowering::isZExtFree;
+ virtual bool isZExtFree(SDValue Val, EVT VT2) const;
+
+ virtual bool allowTruncateForTailCall(Type *Ty1, Type *Ty2) 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, Type *Ty)const;
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;
+ MVT VT) const;
/// LowerAsmOperandForConstraint - Lower the specified operand into the Ops
/// vector. If it is invalid, don't add anything to Ops. If hasMemory is
/// getRegClassFor - Return the register class that should be used for the
/// specified value type.
- virtual TargetRegisterClass *getRegClassFor(EVT VT) const;
+ virtual const TargetRegisterClass *getRegClassFor(MVT VT) const;
/// getMaximalGlobalOffset - Returns the maximal possible offset which can
/// be used for loads / stores from the global.
virtual unsigned getMaximalGlobalOffset() const;
+ /// Returns true if a cast between SrcAS and DestAS is a noop.
+ virtual bool isNoopAddrSpaceCast(unsigned SrcAS, unsigned DestAS) const {
+ // Addrspacecasts are always noops.
+ return true;
+ }
+
/// 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,
+ void PassF64ArgInRegs(SDLoc dl, SelectionDAG &DAG,
SDValue Chain, SDValue &Arg,
RegsToPassVector &RegsToPass,
CCValAssign &VA, CCValAssign &NextVA,
SDValue &StackPtr,
- SmallVector<SDValue, 8> &MemOpChains,
+ SmallVectorImpl<SDValue> &MemOpChains,
ISD::ArgFlagsTy Flags) const;
SDValue GetF64FormalArgument(CCValAssign &VA, CCValAssign &NextVA,
SDValue &Root, SelectionDAG &DAG,
- DebugLoc dl) const;
+ SDLoc dl) const;
CCAssignFn *CCAssignFnForNode(CallingConv::ID CC, bool Return,
bool isVarArg) const;
SDValue LowerMemOpCallTo(SDValue Chain, SDValue StackPtr, SDValue Arg,
- DebugLoc dl, SelectionDAG &DAG,
+ SDLoc dl, SelectionDAG &DAG,
const CCValAssign &VA,
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 LowerConstantFP(SDValue Op, SelectionDAG &DAG,
+ const ARMSubtarget *ST) const;
SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG,
const ARMSubtarget *ST) const;
+ SDValue LowerFSINCOS(SDValue Op, SelectionDAG &DAG) const;
+ SDValue LowerDivRem(SDValue Op, SelectionDAG &DAG) const;
+
+ /// isFMAFasterThanFMulAndFAdd - Return true if an FMA operation is faster
+ /// than a pair of fmul and fadd instructions. fmuladd intrinsics will be
+ /// expanded to FMAs when this method returns true, otherwise fmuladd is
+ /// expanded to fmul + fadd.
+ ///
+ /// ARM supports both fused and unfused multiply-add operations; we already
+ /// lower a pair of fmul and fadd to the latter so it's not clear that there
+ /// would be a gain or that the gain would be worthwhile enough to risk
+ /// correctness bugs.
+ virtual bool isFMAFasterThanFMulAndFAdd(EVT VT) const { return false; }
SDValue ReconstructShuffle(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerCallResult(SDValue Chain, SDValue InFlag,
CallingConv::ID CallConv, bool isVarArg,
const SmallVectorImpl<ISD::InputArg> &Ins,
- DebugLoc dl, SelectionDAG &DAG,
- SmallVectorImpl<SDValue> &InVals) const;
+ SDLoc dl, SelectionDAG &DAG,
+ SmallVectorImpl<SDValue> &InVals,
+ bool isThisReturn, SDValue ThisVal) const;
virtual SDValue
LowerFormalArguments(SDValue Chain,
CallingConv::ID CallConv, bool isVarArg,
const SmallVectorImpl<ISD::InputArg> &Ins,
- DebugLoc dl, SelectionDAG &DAG,
+ SDLoc dl, SelectionDAG &DAG,
SmallVectorImpl<SDValue> &InVals) const;
+ int StoreByValRegs(CCState &CCInfo, SelectionDAG &DAG,
+ SDLoc dl, SDValue &Chain,
+ const Value *OrigArg,
+ unsigned InRegsParamRecordIdx,
+ unsigned OffsetFromOrigArg,
+ unsigned ArgOffset,
+ unsigned ArgSize,
+ bool ForceMutable) const;
+
void VarArgStyleRegisters(CCState &CCInfo, SelectionDAG &DAG,
- DebugLoc dl, SDValue &Chain, unsigned ArgOffset)
- const;
+ SDLoc dl, SDValue &Chain,
+ unsigned ArgOffset,
+ bool ForceMutable = false) const;
void computeRegArea(CCState &CCInfo, MachineFunction &MF,
- unsigned &VARegSize, unsigned &VARegSaveSize) const;
+ unsigned InRegsParamRecordIdx,
+ unsigned ArgSize,
+ unsigned &ArgRegsSize,
+ unsigned &ArgRegsSaveSize) 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 *, unsigned &) 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<ISD::OutputArg> &Outs,
const SmallVectorImpl<SDValue> &OutVals,
- DebugLoc dl, SelectionDAG &DAG) const;
+ SDLoc 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;
+ SDValue &ARMcc, SelectionDAG &DAG, SDLoc dl) const;
SDValue getVFPCmp(SDValue LHS, SDValue RHS,
- SelectionDAG &DAG, DebugLoc dl) const;
+ SelectionDAG &DAG, SDLoc dl) const;
SDValue duplicateCmp(SDValue Cmp, SelectionDAG &DAG) const;
SDValue OptimizeVFPBrcond(SDValue Op, SelectionDAG &DAG) const;
unsigned Op1,
unsigned Op2,
bool NeedsCarry = false,
- bool IsCmpxchg = false) const;
+ 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 EmitBasePointerRecalculation(MachineInstr *MI, MachineBasicBlock *MBB,
- MachineBasicBlock *DispatchBB) const;
+ MachineBasicBlock *EmitAtomicLoad64(MachineInstr *MI,
+ MachineBasicBlock *BB) const;
void SetupEntryBlockForSjLj(MachineInstr *MI,
MachineBasicBlock *MBB,
MachineBasicBlock *MBB) const;
bool RemapAddSubWithFlags(MachineInstr *MI, MachineBasicBlock *BB) const;
+
+ MachineBasicBlock *EmitStructByval(MachineInstr *MI,
+ MachineBasicBlock *MBB) const;
};
enum NEONModImmType {
namespace ARM {
- FastISel *createFastISel(FunctionLoweringInfo &funcInfo);
+ FastISel *createFastISel(FunctionLoweringInfo &funcInfo,
+ const TargetLibraryInfo *libInfo);
}
}