//
//===----------------------------------------------------------------------===//
-#ifndef LLVM_TARGET_POWERPC_PPC32ISELLOWERING_H
-#define LLVM_TARGET_POWERPC_PPC32ISELLOWERING_H
+#ifndef LLVM_LIB_TARGET_POWERPC_PPCISELLOWERING_H
+#define LLVM_LIB_TARGET_POWERPC_PPCISELLOWERING_H
#include "PPC.h"
#include "PPCInstrInfo.h"
namespace llvm {
namespace PPCISD {
- enum NodeType {
+ enum NodeType : unsigned {
// Start the numbering where the builtin ops and target ops leave off.
FIRST_NUMBER = ISD::BUILTIN_OP_END,
///
VPERM,
+ /// The CMPB instruction (takes two operands of i32 or i64).
+ CMPB,
+
/// Hi/Lo - These represent the high and low 16-bit parts of a global
/// address respectively. These nodes have two operands, the first of
/// which must be a TargetGlobalAddress, and the second of which must be a
/// though these are usually folded into other nodes.
Hi, Lo,
- TOC_ENTRY,
-
/// The following two target-specific nodes are used for calls through
/// function pointers in the 64-bit SVR4 ABI.
- /// Like a regular LOAD but additionally taking/producing a flag.
- LOAD,
-
- /// Like LOAD (taking/producing a flag), but using r2 as hard-coded
- /// destination.
- LOAD_TOC,
-
/// OPRC, CHAIN = DYNALLOC(CHAIN, NEGSIZE, FRAME_INDEX)
/// This instruction is lowered in PPCRegisterInfo::eliminateFrameIndex to
/// compute an allocation on the stack.
/// code.
SRL, SRA, SHL,
+ /// The combination of sra[wd]i and addze used to implemented signed
+ /// integer division by a power of 2. The first operand is the dividend,
+ /// and the second is the constant shift amount (representing the
+ /// divisor).
+ SRA_ADDZE,
+
/// CALL - A direct function call.
/// CALL_NOP is a call with the special NOP which follows 64-bit
/// SVR4 calls.
/// BCTRL instruction.
BCTRL,
+ /// CHAIN,FLAG = BCTRL(CHAIN, ADDR, INFLAG) - The combination of a bctrl
+ /// instruction and the TOC reload required on SVR4 PPC64.
+ BCTRL_LOAD_TOC,
+
/// Return with a flag operand, matched by 'blr'
RET_FLAG,
/// resultant GPR. Bits corresponding to other CR regs are undefined.
MFOCRF,
+ /// Direct move from a VSX register to a GPR
+ MFVSR,
+
+ /// Direct move from a GPR to a VSX register (algebraic)
+ MTVSRA,
+
+ /// Direct move from a GPR to a VSX register (zero)
+ MTVSRZ,
+
// FIXME: Remove these once the ANDI glue bug is fixed:
/// i1 = ANDIo_1_[EQ|GT]_BIT(i32 or i64 x) - Represents the result of the
/// eq or gt bit of CR0 after executing andi. x, 1. This is used to
/// implement truncation of i32 or i64 to i1.
ANDIo_1_EQ_BIT, ANDIo_1_GT_BIT,
+ // READ_TIME_BASE - A read of the 64-bit time-base register on a 32-bit
+ // target (returns (Lo, Hi)). It takes a chain operand.
+ READ_TIME_BASE,
+
// EH_SJLJ_SETJMP - SjLj exception handling setjmp.
EH_SJLJ_SETJMP,
/// F8RC = MFFS - This moves the FPSCR (not modeled) into the register.
MFFS,
- /// LARX = This corresponds to PPC l{w|d}arx instrcution: load and
- /// reserve indexed. This is used to implement atomic operations.
- LARX,
-
- /// STCX = This corresponds to PPC stcx. instrcution: store conditional
- /// indexed. This is used to implement atomic operations.
- STCX,
-
/// TC_RETURN - A tail call return.
/// operand #0 chain
/// operand #1 callee (register or absolute)
PPC32_GOT,
/// GPRC = address of _GLOBAL_OFFSET_TABLE_. Used by general dynamic and
- /// local dynamic TLS on PPC32.
+ /// local dynamic TLS on PPC32.
PPC32_PICGOT,
/// G8RC = ADDIS_GOT_TPREL_HA %X2, Symbol - Used by the initial-exec
/// register to sym\@got\@tlsgd\@ha.
ADDIS_TLSGD_HA,
- /// G8RC = ADDI_TLSGD_L G8RReg, Symbol - For the general-dynamic TLS
+ /// %X3 = ADDI_TLSGD_L G8RReg, Symbol - For the general-dynamic TLS
/// model, produces an ADDI8 instruction that adds G8RReg to
- /// sym\@got\@tlsgd\@l.
+ /// sym\@got\@tlsgd\@l and stores the result in X3. Hidden by
+ /// ADDIS_TLSGD_L_ADDR until after register assignment.
ADDI_TLSGD_L,
- /// G8RC = GET_TLS_ADDR %X3, Symbol - For the general-dynamic TLS
- /// model, produces a call to __tls_get_addr(sym\@tlsgd).
+ /// %X3 = GET_TLS_ADDR %X3, Symbol - For the general-dynamic TLS
+ /// model, produces a call to __tls_get_addr(sym\@tlsgd). Hidden by
+ /// ADDIS_TLSGD_L_ADDR until after register assignment.
GET_TLS_ADDR,
+ /// G8RC = ADDI_TLSGD_L_ADDR G8RReg, Symbol, Symbol - Op that
+ /// combines ADDI_TLSGD_L and GET_TLS_ADDR until expansion following
+ /// register assignment.
+ ADDI_TLSGD_L_ADDR,
+
/// G8RC = ADDIS_TLSLD_HA %X2, Symbol - For the local-dynamic TLS
/// model, produces an ADDIS8 instruction that adds the GOT base
/// register to sym\@got\@tlsld\@ha.
ADDIS_TLSLD_HA,
- /// G8RC = ADDI_TLSLD_L G8RReg, Symbol - For the local-dynamic TLS
+ /// %X3 = ADDI_TLSLD_L G8RReg, Symbol - For the local-dynamic TLS
/// model, produces an ADDI8 instruction that adds G8RReg to
- /// sym\@got\@tlsld\@l.
+ /// sym\@got\@tlsld\@l and stores the result in X3. Hidden by
+ /// ADDIS_TLSLD_L_ADDR until after register assignment.
ADDI_TLSLD_L,
- /// G8RC = GET_TLSLD_ADDR %X3, Symbol - For the local-dynamic TLS
- /// model, produces a call to __tls_get_addr(sym\@tlsld).
+ /// %X3 = GET_TLSLD_ADDR %X3, Symbol - For the local-dynamic TLS
+ /// model, produces a call to __tls_get_addr(sym\@tlsld). Hidden by
+ /// ADDIS_TLSLD_L_ADDR until after register assignment.
GET_TLSLD_ADDR,
- /// G8RC = ADDIS_DTPREL_HA %X3, Symbol, Chain - For the
- /// local-dynamic TLS model, produces an ADDIS8 instruction
- /// that adds X3 to sym\@dtprel\@ha. The Chain operand is needed
- /// to tie this in place following a copy to %X3 from the result
- /// of a GET_TLSLD_ADDR.
+ /// G8RC = ADDI_TLSLD_L_ADDR G8RReg, Symbol, Symbol - Op that
+ /// combines ADDI_TLSLD_L and GET_TLSLD_ADDR until expansion
+ /// following register assignment.
+ ADDI_TLSLD_L_ADDR,
+
+ /// G8RC = ADDIS_DTPREL_HA %X3, Symbol - For the local-dynamic TLS
+ /// model, produces an ADDIS8 instruction that adds X3 to
+ /// sym\@dtprel\@ha.
ADDIS_DTPREL_HA,
/// G8RC = ADDI_DTPREL_L G8RReg, Symbol - For the local-dynamic TLS
/// operand identifies the operating system entry point.
SC,
+ /// VSRC, CHAIN = XXSWAPD CHAIN, VSRC - Occurs only for little
+ /// endian. Maps to an xxswapd instruction that corrects an lxvd2x
+ /// or stxvd2x instruction. The chain is necessary because the
+ /// sequence replaces a load and needs to provide the same number
+ /// of outputs.
+ XXSWAPD,
+
+ /// QVFPERM = This corresponds to the QPX qvfperm instruction.
+ QVFPERM,
+
+ /// QVGPCI = This corresponds to the QPX qvgpci instruction.
+ QVGPCI,
+
+ /// QVALIGNI = This corresponds to the QPX qvaligni instruction.
+ QVALIGNI,
+
+ /// QVESPLATI = This corresponds to the QPX qvesplati instruction.
+ QVESPLATI,
+
+ /// QBFLT = Access the underlying QPX floating-point boolean
+ /// representation.
+ QBFLT,
+
/// CHAIN = STBRX CHAIN, GPRC, Ptr, Type - This is a
/// byte-swapping store instruction. It byte-swaps the low "Type" bits of
/// the GPRC input, then stores it through Ptr. Type can be either i16 or
/// destination 64-bit register.
LFIWZX,
- /// G8RC = ADDIS_TOC_HA %X2, Symbol - For medium and large code model,
- /// produces an ADDIS8 instruction that adds the TOC base register to
- /// sym\@toc\@ha.
- ADDIS_TOC_HA,
+ /// VSRC, CHAIN = LXVD2X_LE CHAIN, Ptr - Occurs only for little endian.
+ /// Maps directly to an lxvd2x instruction that will be followed by
+ /// an xxswapd.
+ LXVD2X,
- /// G8RC = LD_TOC_L Symbol, G8RReg - For medium and large code model,
- /// produces a LD instruction with base register G8RReg and offset
- /// sym\@toc\@l. Preceded by an ADDIS_TOC_HA to form a full 32-bit offset.
- LD_TOC_L,
+ /// CHAIN = STXVD2X CHAIN, VSRC, Ptr - Occurs only for little endian.
+ /// Maps directly to an stxvd2x instruction that will be preceded by
+ /// an xxswapd.
+ STXVD2X,
- /// G8RC = ADDI_TOC_L G8RReg, Symbol - For medium code model, produces
- /// an ADDI8 instruction that adds G8RReg to sym\@toc\@l.
- /// Preceded by an ADDIS_TOC_HA to form a full 32-bit offset.
- ADDI_TOC_L
+ /// QBRC, CHAIN = QVLFSb CHAIN, Ptr
+ /// The 4xf32 load used for v4i1 constants.
+ QVLFSb,
+
+ /// GPRC = TOC_ENTRY GA, TOC
+ /// Loads the entry for GA from the TOC, where the TOC base is given by
+ /// the last operand.
+ TOC_ENTRY
};
}
/// VSPLTB/VSPLTH/VSPLTW.
bool isSplatShuffleMask(ShuffleVectorSDNode *N, unsigned EltSize);
- /// isAllNegativeZeroVector - Returns true if all elements of build_vector
- /// are -0.0.
- bool isAllNegativeZeroVector(SDNode *N);
-
/// getVSPLTImmediate - Return the appropriate VSPLT* immediate to splat the
/// specified isSplatShuffleMask VECTOR_SHUFFLE mask.
unsigned getVSPLTImmediate(SDNode *N, unsigned EltSize, SelectionDAG &DAG);
/// size, return the constant being splatted. The ByteSize field indicates
/// the number of bytes of each element [124] -> [bhw].
SDValue get_VSPLTI_elt(SDNode *N, unsigned ByteSize, SelectionDAG &DAG);
+
+ /// If this is a qvaligni shuffle mask, return the shift
+ /// amount, otherwise return -1.
+ int isQVALIGNIShuffleMask(SDNode *N);
}
- class PPCSubtarget;
class PPCTargetLowering : public TargetLowering {
const PPCSubtarget &Subtarget;
public:
- explicit PPCTargetLowering(PPCTargetMachine &TM);
+ explicit PPCTargetLowering(const PPCTargetMachine &TM,
+ const PPCSubtarget &STI);
/// getTargetNodeName() - This method returns the name of a target specific
/// DAG node.
MVT getScalarShiftAmountTy(EVT LHSTy) const override { return MVT::i32; }
+ bool isCheapToSpeculateCttz() const override {
+ return true;
+ }
+
+ bool isCheapToSpeculateCtlz() const override {
+ return true;
+ }
+
/// getSetCCResultType - Return the ISD::SETCC ValueType
EVT getSetCCResultType(LLVMContext &Context, EVT VT) const override;
+ /// Return true if target always beneficiates from combining into FMA for a
+ /// given value type. This must typically return false on targets where FMA
+ /// takes more cycles to execute than FADD.
+ bool enableAggressiveFMAFusion(EVT VT) const override;
+
/// 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.
void ReplaceNodeResults(SDNode *N, SmallVectorImpl<SDValue>&Results,
SelectionDAG &DAG) const override;
+ SDValue expandVSXLoadForLE(SDNode *N, DAGCombinerInfo &DCI) const;
+ SDValue expandVSXStoreForLE(SDNode *N, DAGCombinerInfo &DCI) const;
+
SDValue PerformDAGCombine(SDNode *N, DAGCombinerInfo &DCI) const override;
+ SDValue BuildSDIVPow2(SDNode *N, const APInt &Divisor, SelectionDAG &DAG,
+ std::vector<SDNode *> *Created) const override;
+
unsigned getRegisterByName(const char* RegName, EVT VT) const override;
void computeKnownBitsForTargetNode(const SDValue Op,
const SelectionDAG &DAG,
unsigned Depth = 0) const override;
+ unsigned getPrefLoopAlignment(MachineLoop *ML) const override;
+
+ Instruction* emitLeadingFence(IRBuilder<> &Builder, AtomicOrdering Ord,
+ bool IsStore, bool IsLoad) const override;
+ Instruction* emitTrailingFence(IRBuilder<> &Builder, AtomicOrdering Ord,
+ bool IsStore, bool IsLoad) const override;
+
MachineBasicBlock *
EmitInstrWithCustomInserter(MachineInstr *MI,
MachineBasicBlock *MBB) const override;
MachineBasicBlock *EmitAtomicBinary(MachineInstr *MI,
- MachineBasicBlock *MBB, bool is64Bit,
+ MachineBasicBlock *MBB,
+ unsigned AtomicSize,
unsigned BinOpcode) const;
MachineBasicBlock *EmitPartwordAtomicBinary(MachineInstr *MI,
MachineBasicBlock *MBB,
ConstraintWeight getSingleConstraintMatchWeight(
AsmOperandInfo &info, const char *constraint) const override;
- std::pair<unsigned, const TargetRegisterClass*>
- getRegForInlineAsmConstraint(const std::string &Constraint,
- MVT VT) const override;
+ std::pair<unsigned, const TargetRegisterClass *>
+ getRegForInlineAsmConstraint(const TargetRegisterInfo *TRI,
+ const std::string &Constraint,
+ MVT VT) const override;
/// getByValTypeAlignment - Return the desired alignment for ByVal aggregate
/// function arguments in the caller parameter area. This is the actual
std::vector<SDValue> &Ops,
SelectionDAG &DAG) const override;
+ unsigned getInlineAsmMemConstraint(
+ const std::string &ConstraintCode) const override {
+ if (ConstraintCode == "es")
+ return InlineAsm::Constraint_es;
+ else if (ConstraintCode == "o")
+ return InlineAsm::Constraint_o;
+ else if (ConstraintCode == "Q")
+ return InlineAsm::Constraint_Q;
+ else if (ConstraintCode == "Z")
+ return InlineAsm::Constraint_Z;
+ else if (ConstraintCode == "Zy")
+ return InlineAsm::Constraint_Zy;
+ return TargetLowering::getInlineAsmMemConstraint(ConstraintCode);
+ }
+
/// isLegalAddressingMode - Return true if the addressing mode represented
/// by AM is legal for this target, for a load/store of the specified type.
bool isLegalAddressingMode(const AddrMode &AM, Type *Ty) const override;
bool isTruncateFree(Type *Ty1, Type *Ty2) const override;
bool isTruncateFree(EVT VT1, EVT VT2) const override;
+ bool isZExtFree(SDValue Val, EVT VT2) const override;
+
+ bool isFPExtFree(EVT VT) const override;
+
/// \brief Returns true if it is beneficial to convert a load of a constant
/// to just the constant itself.
bool shouldConvertConstantLoadToIntImm(const APInt &Imm,
bool isOffsetFoldingLegal(const GlobalAddressSDNode *GA) const override;
+ bool getTgtMemIntrinsic(IntrinsicInfo &Info,
+ const CallInst &I,
+ unsigned Intrinsic) const override;
+
/// getOptimalMemOpType - Returns the target specific optimal type for load
/// and store operations as a result of memset, memcpy, and memmove
/// lowering. If DstAlign is zero that means it's safe to destination
/// expanded to fmul + fadd.
bool isFMAFasterThanFMulAndFAdd(EVT VT) const override;
+ const MCPhysReg *getScratchRegisters(CallingConv::ID CC) const override;
+
// Should we expand the build vector with shuffles?
bool
shouldExpandBuildVectorWithShuffles(EVT VT,
}
private:
+
+ struct ReuseLoadInfo {
+ SDValue Ptr;
+ SDValue Chain;
+ SDValue ResChain;
+ MachinePointerInfo MPI;
+ bool IsInvariant;
+ unsigned Alignment;
+ AAMDNodes AAInfo;
+ const MDNode *Ranges;
+
+ ReuseLoadInfo() : IsInvariant(false), Alignment(0), Ranges(nullptr) {}
+ };
+
+ bool canReuseLoadAddress(SDValue Op, EVT MemVT, ReuseLoadInfo &RLI,
+ SelectionDAG &DAG,
+ ISD::LoadExtType ET = ISD::NON_EXTLOAD) const;
+ void spliceIntoChain(SDValue ResChain, SDValue NewResChain,
+ SelectionDAG &DAG) const;
+
+ void LowerFP_TO_INTForReuse(SDValue Op, ReuseLoadInfo &RLI,
+ SelectionDAG &DAG, SDLoc dl) const;
+ SDValue LowerFP_TO_INTDirectMove(SDValue Op, SelectionDAG &DAG,
+ SDLoc dl) const;
+ SDValue LowerINT_TO_FPDirectMove(SDValue Op, SelectionDAG &DAG,
+ SDLoc dl) const;
+
SDValue getFramePointerFrameIndex(SelectionDAG & DAG) const;
SDValue getReturnAddrFrameIndex(SelectionDAG & DAG) const;
SDValue LowerSRA_PARTS(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerVECTOR_SHUFFLE(SDValue Op, SelectionDAG &DAG) const;
+ SDValue LowerEXTRACT_VECTOR_ELT(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerINTRINSIC_WO_CHAIN(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerSCALAR_TO_VECTOR(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerSIGN_EXTEND_INREG(SDValue Op, SelectionDAG &DAG) const;
SDValue LowerMUL(SDValue Op, SelectionDAG &DAG) const;
+ SDValue LowerVectorLoad(SDValue Op, SelectionDAG &DAG) const;
+ SDValue LowerVectorStore(SDValue Op, SelectionDAG &DAG) const;
+
SDValue LowerCallResult(SDValue Chain, SDValue InFlag,
CallingConv::ID CallConv, bool isVarArg,
const SmallVectorImpl<ISD::InputArg> &Ins,
SDLoc dl, SelectionDAG &DAG,
SmallVectorImpl<SDValue> &InVals) const;
SDValue FinishCall(CallingConv::ID CallConv, SDLoc dl, bool isTailCall,
- bool isVarArg,
+ bool isVarArg, bool IsPatchPoint,
SelectionDAG &DAG,
SmallVector<std::pair<unsigned, SDValue>, 8>
&RegsToPass,
- SDValue InFlag, SDValue Chain,
+ SDValue InFlag, SDValue Chain, SDValue CallSeqStart,
SDValue &Callee,
int SPDiff, unsigned NumBytes,
const SmallVectorImpl<ISD::InputArg> &Ins,
- SmallVectorImpl<SDValue> &InVals) const;
+ SmallVectorImpl<SDValue> &InVals,
+ ImmutableCallSite *CS) const;
SDValue
LowerFormalArguments(SDValue Chain,
SDValue
LowerCall_Darwin(SDValue Chain, SDValue Callee,
CallingConv::ID CallConv,
- bool isVarArg, bool isTailCall,
+ bool isVarArg, bool isTailCall, bool IsPatchPoint,
const SmallVectorImpl<ISD::OutputArg> &Outs,
const SmallVectorImpl<SDValue> &OutVals,
const SmallVectorImpl<ISD::InputArg> &Ins,
SDLoc dl, SelectionDAG &DAG,
- SmallVectorImpl<SDValue> &InVals) const;
+ SmallVectorImpl<SDValue> &InVals,
+ ImmutableCallSite *CS) const;
SDValue
LowerCall_64SVR4(SDValue Chain, SDValue Callee,
CallingConv::ID CallConv,
- bool isVarArg, bool isTailCall,
+ bool isVarArg, bool isTailCall, bool IsPatchPoint,
const SmallVectorImpl<ISD::OutputArg> &Outs,
const SmallVectorImpl<SDValue> &OutVals,
const SmallVectorImpl<ISD::InputArg> &Ins,
SDLoc dl, SelectionDAG &DAG,
- SmallVectorImpl<SDValue> &InVals) const;
+ SmallVectorImpl<SDValue> &InVals,
+ ImmutableCallSite *CS) const;
SDValue
LowerCall_32SVR4(SDValue Chain, SDValue Callee, CallingConv::ID CallConv,
- bool isVarArg, bool isTailCall,
+ bool isVarArg, bool isTailCall, bool IsPatchPoint,
const SmallVectorImpl<ISD::OutputArg> &Outs,
const SmallVectorImpl<SDValue> &OutVals,
const SmallVectorImpl<ISD::InputArg> &Ins,
SDLoc dl, SelectionDAG &DAG,
- SmallVectorImpl<SDValue> &InVals) const;
+ SmallVectorImpl<SDValue> &InVals,
+ ImmutableCallSite *CS) const;
SDValue lowerEH_SJLJ_SETJMP(SDValue Op, SelectionDAG &DAG) const;
SDValue lowerEH_SJLJ_LONGJMP(SDValue Op, SelectionDAG &DAG) const;
SDValue DAGCombineExtBoolTrunc(SDNode *N, DAGCombinerInfo &DCI) const;
SDValue DAGCombineTruncBoolExt(SDNode *N, DAGCombinerInfo &DCI) const;
- SDValue DAGCombineFastRecip(SDValue Op, DAGCombinerInfo &DCI) const;
- SDValue DAGCombineFastRecipFSQRT(SDValue Op, DAGCombinerInfo &DCI) const;
+ SDValue combineFPToIntToFP(SDNode *N, DAGCombinerInfo &DCI) const;
+
+ SDValue getRsqrtEstimate(SDValue Operand, DAGCombinerInfo &DCI,
+ unsigned &RefinementSteps,
+ bool &UseOneConstNR) const override;
+ SDValue getRecipEstimate(SDValue Operand, DAGCombinerInfo &DCI,
+ unsigned &RefinementSteps) const override;
+ bool combineRepeatedFPDivisors(unsigned NumUsers) const override;
CCAssignFn *useFastISelCCs(unsigned Flag) const;
};