X-Git-Url: http://plrg.eecs.uci.edu/git/?p=oota-llvm.git;a=blobdiff_plain;f=lib%2FTarget%2FAArch64%2FAArch64InstrInfo.cpp;h=f398117de953b2f2d9a55673cfbaa0ca7c11ecab;hp=14daab33ecfb0b95546adea136dfe628be58fc6a;hb=5666fc71f0e2ed2c0400d8bca079a1dd3f33fe53;hpb=b21ab43cfc3fa0dacf5c95f04e58b6d804b59a16 diff --git a/lib/Target/AArch64/AArch64InstrInfo.cpp b/lib/Target/AArch64/AArch64InstrInfo.cpp index 14daab33ecf..f398117de95 100644 --- a/lib/Target/AArch64/AArch64InstrInfo.cpp +++ b/lib/Target/AArch64/AArch64InstrInfo.cpp @@ -11,196 +11,94 @@ // //===----------------------------------------------------------------------===// -#include "AArch64.h" #include "AArch64InstrInfo.h" -#include "AArch64MachineFunctionInfo.h" -#include "AArch64TargetMachine.h" -#include "MCTargetDesc/AArch64MCTargetDesc.h" -#include "Utils/AArch64BaseInfo.h" -#include "llvm/CodeGen/MachineConstantPool.h" -#include "llvm/CodeGen/MachineDominators.h" +#include "AArch64Subtarget.h" +#include "MCTargetDesc/AArch64AddressingModes.h" #include "llvm/CodeGen/MachineFrameInfo.h" -#include "llvm/CodeGen/MachineFunctionPass.h" #include "llvm/CodeGen/MachineInstrBuilder.h" +#include "llvm/CodeGen/MachineMemOperand.h" #include "llvm/CodeGen/MachineRegisterInfo.h" -#include "llvm/IR/Function.h" +#include "llvm/CodeGen/PseudoSourceValue.h" +#include "llvm/MC/MCInst.h" #include "llvm/Support/ErrorHandling.h" #include "llvm/Support/TargetRegistry.h" -#include +using namespace llvm; -#define GET_INSTRINFO_CTOR +#define GET_INSTRINFO_CTOR_DTOR #include "AArch64GenInstrInfo.inc" -using namespace llvm; - AArch64InstrInfo::AArch64InstrInfo(const AArch64Subtarget &STI) - : AArch64GenInstrInfo(AArch64::ADJCALLSTACKDOWN, AArch64::ADJCALLSTACKUP), - Subtarget(STI) {} + : AArch64GenInstrInfo(AArch64::ADJCALLSTACKDOWN, AArch64::ADJCALLSTACKUP), + RI(STI.getTargetTriple()), Subtarget(STI) {} -void AArch64InstrInfo::copyPhysReg(MachineBasicBlock &MBB, - MachineBasicBlock::iterator I, DebugLoc DL, - unsigned DestReg, unsigned SrcReg, - bool KillSrc) const { - unsigned Opc = 0; - unsigned ZeroReg = 0; - if (DestReg == AArch64::XSP || SrcReg == AArch64::XSP) { - // E.g. ADD xDst, xsp, #0 (, lsl #0) - BuildMI(MBB, I, DL, get(AArch64::ADDxxi_lsl0_s), DestReg) - .addReg(SrcReg) - .addImm(0); - return; - } else if (DestReg == AArch64::WSP || SrcReg == AArch64::WSP) { - // E.g. ADD wDST, wsp, #0 (, lsl #0) - BuildMI(MBB, I, DL, get(AArch64::ADDwwi_lsl0_s), DestReg) - .addReg(SrcReg) - .addImm(0); - return; - } else if (DestReg == AArch64::NZCV) { - assert(AArch64::GPR64RegClass.contains(SrcReg)); - // E.g. MSR NZCV, xDST - BuildMI(MBB, I, DL, get(AArch64::MSRix)) - .addImm(A64SysReg::NZCV) - .addReg(SrcReg); - } else if (SrcReg == AArch64::NZCV) { - assert(AArch64::GPR64RegClass.contains(DestReg)); - // E.g. MRS xDST, NZCV - BuildMI(MBB, I, DL, get(AArch64::MRSxi), DestReg) - .addImm(A64SysReg::NZCV); - } else if (AArch64::GPR64RegClass.contains(DestReg)) { - if(AArch64::GPR64RegClass.contains(SrcReg)){ - Opc = AArch64::ORRxxx_lsl; - ZeroReg = AArch64::XZR; - } else{ - assert(AArch64::FPR64RegClass.contains(SrcReg)); - BuildMI(MBB, I, DL, get(AArch64::FMOVxd), DestReg) - .addReg(SrcReg); - return; - } - } else if (AArch64::GPR32RegClass.contains(DestReg)) { - if(AArch64::GPR32RegClass.contains(SrcReg)){ - Opc = AArch64::ORRwww_lsl; - ZeroReg = AArch64::WZR; - } else{ - assert(AArch64::FPR32RegClass.contains(SrcReg)); - BuildMI(MBB, I, DL, get(AArch64::FMOVws), DestReg) - .addReg(SrcReg); - return; - } - } else if (AArch64::FPR32RegClass.contains(DestReg)) { - if(AArch64::FPR32RegClass.contains(SrcReg)){ - BuildMI(MBB, I, DL, get(AArch64::FMOVss), DestReg) - .addReg(SrcReg); - return; - } - else { - assert(AArch64::GPR32RegClass.contains(SrcReg)); - BuildMI(MBB, I, DL, get(AArch64::FMOVsw), DestReg) - .addReg(SrcReg); - return; - } - } else if (AArch64::FPR64RegClass.contains(DestReg)) { - if(AArch64::FPR64RegClass.contains(SrcReg)){ - BuildMI(MBB, I, DL, get(AArch64::FMOVdd), DestReg) - .addReg(SrcReg); - return; - } - else { - assert(AArch64::GPR64RegClass.contains(SrcReg)); - BuildMI(MBB, I, DL, get(AArch64::FMOVdx), DestReg) - .addReg(SrcReg); - return; - } - } else if (AArch64::FPR128RegClass.contains(DestReg)) { - assert(AArch64::FPR128RegClass.contains(SrcReg)); - - // FIXME: there's no good way to do this, at least without NEON: - // + There's no single move instruction for q-registers - // + We can't create a spill slot and use normal STR/LDR because stack - // allocation has already happened - // + We can't go via X-registers with FMOV because register allocation has - // already happened. - // This may not be efficient, but at least it works. - BuildMI(MBB, I, DL, get(AArch64::LSFP128_PreInd_STR), AArch64::XSP) - .addReg(SrcReg) - .addReg(AArch64::XSP) - .addImm(0x1ff & -16); +/// GetInstSize - Return the number of bytes of code the specified +/// instruction may be. This returns the maximum number of bytes. +unsigned AArch64InstrInfo::GetInstSizeInBytes(const MachineInstr *MI) const { + const MachineBasicBlock &MBB = *MI->getParent(); + const MachineFunction *MF = MBB.getParent(); + const MCAsmInfo *MAI = MF->getTarget().getMCAsmInfo(); - BuildMI(MBB, I, DL, get(AArch64::LSFP128_PostInd_LDR), DestReg) - .addReg(AArch64::XSP, RegState::Define) - .addReg(AArch64::XSP) - .addImm(16); - return; - } else { - llvm_unreachable("Unknown register class in copyPhysReg"); - } + if (MI->getOpcode() == AArch64::INLINEASM) + return getInlineAsmLength(MI->getOperand(0).getSymbolName(), *MAI); - // E.g. ORR xDst, xzr, xSrc, lsl #0 - BuildMI(MBB, I, DL, get(Opc), DestReg) - .addReg(ZeroReg) - .addReg(SrcReg) - .addImm(0); -} + const MCInstrDesc &Desc = MI->getDesc(); + switch (Desc.getOpcode()) { + default: + // Anything not explicitly designated otherwise is a nomal 4-byte insn. + return 4; + case TargetOpcode::DBG_VALUE: + case TargetOpcode::EH_LABEL: + case TargetOpcode::IMPLICIT_DEF: + case TargetOpcode::KILL: + return 0; + } -/// Does the Opcode represent a conditional branch that we can remove and re-add -/// at the end of a basic block? -static bool isCondBranch(unsigned Opc) { - return Opc == AArch64::Bcc || Opc == AArch64::CBZw || Opc == AArch64::CBZx || - Opc == AArch64::CBNZw || Opc == AArch64::CBNZx || - Opc == AArch64::TBZwii || Opc == AArch64::TBZxii || - Opc == AArch64::TBNZwii || Opc == AArch64::TBNZxii; + llvm_unreachable("GetInstSizeInBytes()- Unable to determin insn size"); } -/// Takes apart a given conditional branch MachineInstr (see isCondBranch), -/// setting TBB to the destination basic block and populating the Cond vector -/// with data necessary to recreate the conditional branch at a later -/// date. First element will be the opcode, and subsequent ones define the -/// conditions being branched on in an instruction-specific manner. -static void classifyCondBranch(MachineInstr *I, MachineBasicBlock *&TBB, - SmallVectorImpl &Cond) { - switch(I->getOpcode()) { - case AArch64::Bcc: - case AArch64::CBZw: - case AArch64::CBZx: - case AArch64::CBNZw: - case AArch64::CBNZx: - // These instructions just have one predicate operand in position 0 (either - // a condition code or a register being compared). - Cond.push_back(MachineOperand::CreateImm(I->getOpcode())); - Cond.push_back(I->getOperand(0)); - TBB = I->getOperand(1).getMBB(); - return; - case AArch64::TBZwii: - case AArch64::TBZxii: - case AArch64::TBNZwii: - case AArch64::TBNZxii: - // These have two predicate operands: a register and a bit position. - Cond.push_back(MachineOperand::CreateImm(I->getOpcode())); - Cond.push_back(I->getOperand(0)); - Cond.push_back(I->getOperand(1)); - TBB = I->getOperand(2).getMBB(); - return; +static void parseCondBranch(MachineInstr *LastInst, MachineBasicBlock *&Target, + SmallVectorImpl &Cond) { + // Block ends with fall-through condbranch. + switch (LastInst->getOpcode()) { default: - llvm_unreachable("Unknown conditional branch to classify"); + llvm_unreachable("Unknown branch instruction?"); + case AArch64::Bcc: + Target = LastInst->getOperand(1).getMBB(); + Cond.push_back(LastInst->getOperand(0)); + break; + case AArch64::CBZW: + case AArch64::CBZX: + case AArch64::CBNZW: + case AArch64::CBNZX: + Target = LastInst->getOperand(1).getMBB(); + Cond.push_back(MachineOperand::CreateImm(-1)); + Cond.push_back(MachineOperand::CreateImm(LastInst->getOpcode())); + Cond.push_back(LastInst->getOperand(0)); + break; + case AArch64::TBZW: + case AArch64::TBZX: + case AArch64::TBNZW: + case AArch64::TBNZX: + Target = LastInst->getOperand(2).getMBB(); + Cond.push_back(MachineOperand::CreateImm(-1)); + Cond.push_back(MachineOperand::CreateImm(LastInst->getOpcode())); + Cond.push_back(LastInst->getOperand(0)); + Cond.push_back(LastInst->getOperand(1)); } } - -bool -AArch64InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,MachineBasicBlock *&TBB, - MachineBasicBlock *&FBB, - SmallVectorImpl &Cond, - bool AllowModify) const { +// Branch analysis. +bool AArch64InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB, + MachineBasicBlock *&TBB, + MachineBasicBlock *&FBB, + SmallVectorImpl &Cond, + bool AllowModify) const { // If the block has no terminators, it just falls into the block after it. - MachineBasicBlock::iterator I = MBB.end(); - if (I == MBB.begin()) + MachineBasicBlock::iterator I = MBB.getLastNonDebugInstr(); + if (I == MBB.end()) return false; - --I; - while (I->isDebugValue()) { - if (I == MBB.begin()) - return false; - --I; - } + if (!isUnpredicatedTerminator(I)) return false; @@ -210,15 +108,16 @@ AArch64InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,MachineBasicBlock *&TBB, // If there is only one terminator instruction, process it. unsigned LastOpc = LastInst->getOpcode(); if (I == MBB.begin() || !isUnpredicatedTerminator(--I)) { - if (LastOpc == AArch64::Bimm) { + if (isUncondBranchOpcode(LastOpc)) { TBB = LastInst->getOperand(0).getMBB(); return false; } - if (isCondBranch(LastOpc)) { - classifyCondBranch(LastInst, TBB, Cond); + if (isCondBranchOpcode(LastOpc)) { + // Block ends with fall-through condbranch. + parseCondBranch(LastInst, TBB, Cond); return false; } - return true; // Can't handle indirect branch. + return true; // Can't handle indirect branch. } // Get the instruction before it if it is a terminator. @@ -227,8 +126,8 @@ AArch64InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,MachineBasicBlock *&TBB, // If AllowModify is true and the block ends with two or more unconditional // branches, delete all but the first unconditional branch. - if (AllowModify && LastOpc == AArch64::Bimm) { - while (SecondLastOpc == AArch64::Bimm) { + if (AllowModify && isUncondBranchOpcode(LastOpc)) { + while (isUncondBranchOpcode(SecondLastOpc)) { LastInst->eraseFromParent(); LastInst = SecondLastInst; LastOpc = LastInst->getOpcode(); @@ -248,23 +147,15 @@ AArch64InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,MachineBasicBlock *&TBB, return true; // If the block ends with a B and a Bcc, handle it. - if (LastOpc == AArch64::Bimm) { - if (SecondLastOpc == AArch64::Bcc) { - TBB = SecondLastInst->getOperand(1).getMBB(); - Cond.push_back(MachineOperand::CreateImm(AArch64::Bcc)); - Cond.push_back(SecondLastInst->getOperand(0)); - FBB = LastInst->getOperand(0).getMBB(); - return false; - } else if (isCondBranch(SecondLastOpc)) { - classifyCondBranch(SecondLastInst, TBB, Cond); - FBB = LastInst->getOperand(0).getMBB(); - return false; - } + if (isCondBranchOpcode(SecondLastOpc) && isUncondBranchOpcode(LastOpc)) { + parseCondBranch(SecondLastInst, TBB, Cond); + FBB = LastInst->getOperand(0).getMBB(); + return false; } // If the block ends with two unconditional branches, handle it. The second // one is not executed, so remove it. - if (SecondLastOpc == AArch64::Bimm && LastOpc == AArch64::Bimm) { + if (isUncondBranchOpcode(SecondLastOpc) && isUncondBranchOpcode(LastOpc)) { TBB = SecondLastInst->getOperand(0).getMBB(); I = LastInst; if (AllowModify) @@ -272,84 +163,67 @@ AArch64InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,MachineBasicBlock *&TBB, return false; } + // ...likewise if it ends with an indirect branch followed by an unconditional + // branch. + if (isIndirectBranchOpcode(SecondLastOpc) && isUncondBranchOpcode(LastOpc)) { + I = LastInst; + if (AllowModify) + I->eraseFromParent(); + return true; + } + // Otherwise, can't handle this. return true; } bool AArch64InstrInfo::ReverseBranchCondition( - SmallVectorImpl &Cond) const { - switch (Cond[0].getImm()) { - case AArch64::Bcc: { - A64CC::CondCodes CC = static_cast(Cond[1].getImm()); - CC = A64InvertCondCode(CC); - Cond[1].setImm(CC); - return false; - } - case AArch64::CBZw: - Cond[0].setImm(AArch64::CBNZw); - return false; - case AArch64::CBZx: - Cond[0].setImm(AArch64::CBNZx); - return false; - case AArch64::CBNZw: - Cond[0].setImm(AArch64::CBZw); - return false; - case AArch64::CBNZx: - Cond[0].setImm(AArch64::CBZx); - return false; - case AArch64::TBZwii: - Cond[0].setImm(AArch64::TBNZwii); - return false; - case AArch64::TBZxii: - Cond[0].setImm(AArch64::TBNZxii); - return false; - case AArch64::TBNZwii: - Cond[0].setImm(AArch64::TBZwii); - return false; - case AArch64::TBNZxii: - Cond[0].setImm(AArch64::TBZxii); - return false; - default: - llvm_unreachable("Unknown branch type"); - } -} - - -unsigned -AArch64InstrInfo::InsertBranch(MachineBasicBlock &MBB, MachineBasicBlock *TBB, - MachineBasicBlock *FBB, - const SmallVectorImpl &Cond, - DebugLoc DL) const { - if (FBB == 0 && Cond.empty()) { - BuildMI(&MBB, DL, get(AArch64::Bimm)).addMBB(TBB); - return 1; - } else if (FBB == 0) { - MachineInstrBuilder MIB = BuildMI(&MBB, DL, get(Cond[0].getImm())); - for (int i = 1, e = Cond.size(); i != e; ++i) - MIB.addOperand(Cond[i]); - MIB.addMBB(TBB); - return 1; + SmallVectorImpl &Cond) const { + if (Cond[0].getImm() != -1) { + // Regular Bcc + AArch64CC::CondCode CC = (AArch64CC::CondCode)(int)Cond[0].getImm(); + Cond[0].setImm(AArch64CC::getInvertedCondCode(CC)); + } else { + // Folded compare-and-branch + switch (Cond[1].getImm()) { + default: + llvm_unreachable("Unknown conditional branch!"); + case AArch64::CBZW: + Cond[1].setImm(AArch64::CBNZW); + break; + case AArch64::CBNZW: + Cond[1].setImm(AArch64::CBZW); + break; + case AArch64::CBZX: + Cond[1].setImm(AArch64::CBNZX); + break; + case AArch64::CBNZX: + Cond[1].setImm(AArch64::CBZX); + break; + case AArch64::TBZW: + Cond[1].setImm(AArch64::TBNZW); + break; + case AArch64::TBNZW: + Cond[1].setImm(AArch64::TBZW); + break; + case AArch64::TBZX: + Cond[1].setImm(AArch64::TBNZX); + break; + case AArch64::TBNZX: + Cond[1].setImm(AArch64::TBZX); + break; + } } - MachineInstrBuilder MIB = BuildMI(&MBB, DL, get(Cond[0].getImm())); - for (int i = 1, e = Cond.size(); i != e; ++i) - MIB.addOperand(Cond[i]); - MIB.addMBB(TBB); - - BuildMI(&MBB, DL, get(AArch64::Bimm)).addMBB(FBB); - return 2; + return false; } unsigned AArch64InstrInfo::RemoveBranch(MachineBasicBlock &MBB) const { - MachineBasicBlock::iterator I = MBB.end(); - if (I == MBB.begin()) return 0; - --I; - while (I->isDebugValue()) { - if (I == MBB.begin()) - return 0; - --I; - } - if (I->getOpcode() != AArch64::Bimm && !isCondBranch(I->getOpcode())) + MachineBasicBlock::iterator I = MBB.getLastNonDebugInstr(); + if (I == MBB.end()) + return 0; + + if (!isUncondBranchOpcode(I->getOpcode()) && + !isCondBranchOpcode(I->getOpcode())) return 0; // Remove the branch. @@ -357,9 +231,10 @@ unsigned AArch64InstrInfo::RemoveBranch(MachineBasicBlock &MBB) const { I = MBB.end(); - if (I == MBB.begin()) return 1; + if (I == MBB.begin()) + return 1; --I; - if (!isCondBranch(I->getOpcode())) + if (!isCondBranchOpcode(I->getOpcode())) return 1; // Remove the branch. @@ -367,471 +242,2811 @@ unsigned AArch64InstrInfo::RemoveBranch(MachineBasicBlock &MBB) const { return 2; } -bool -AArch64InstrInfo::expandPostRAPseudo(MachineBasicBlock::iterator MBBI) const { - MachineInstr &MI = *MBBI; - MachineBasicBlock &MBB = *MI.getParent(); +void AArch64InstrInfo::instantiateCondBranch( + MachineBasicBlock &MBB, DebugLoc DL, MachineBasicBlock *TBB, + ArrayRef Cond) const { + if (Cond[0].getImm() != -1) { + // Regular Bcc + BuildMI(&MBB, DL, get(AArch64::Bcc)).addImm(Cond[0].getImm()).addMBB(TBB); + } else { + // Folded compare-and-branch + // Note that we use addOperand instead of addReg to keep the flags. + const MachineInstrBuilder MIB = + BuildMI(&MBB, DL, get(Cond[1].getImm())).addOperand(Cond[2]); + if (Cond.size() > 3) + MIB.addImm(Cond[3].getImm()); + MIB.addMBB(TBB); + } +} - unsigned Opcode = MI.getOpcode(); - switch (Opcode) { - case AArch64::TLSDESC_BLRx: { - MachineInstr *NewMI = - BuildMI(MBB, MBBI, MI.getDebugLoc(), get(AArch64::TLSDESCCALL)) - .addOperand(MI.getOperand(1)); - MI.setDesc(get(AArch64::BLRx)); - - llvm::finalizeBundle(MBB, NewMI, *++MBBI); - return true; - } +unsigned AArch64InstrInfo::InsertBranch( + MachineBasicBlock &MBB, MachineBasicBlock *TBB, MachineBasicBlock *FBB, + ArrayRef Cond, DebugLoc DL) const { + // Shouldn't be a fall through. + assert(TBB && "InsertBranch must not be told to insert a fallthrough"); + + if (!FBB) { + if (Cond.empty()) // Unconditional branch? + BuildMI(&MBB, DL, get(AArch64::B)).addMBB(TBB); + else + instantiateCondBranch(MBB, DL, TBB, Cond); + return 1; + } + + // Two-way conditional branch. + instantiateCondBranch(MBB, DL, TBB, Cond); + BuildMI(&MBB, DL, get(AArch64::B)).addMBB(FBB); + return 2; +} + +// Find the original register that VReg is copied from. +static unsigned removeCopies(const MachineRegisterInfo &MRI, unsigned VReg) { + while (TargetRegisterInfo::isVirtualRegister(VReg)) { + const MachineInstr *DefMI = MRI.getVRegDef(VReg); + if (!DefMI->isFullCopy()) + return VReg; + VReg = DefMI->getOperand(1).getReg(); + } + return VReg; +} + +// Determine if VReg is defined by an instruction that can be folded into a +// csel instruction. If so, return the folded opcode, and the replacement +// register. +static unsigned canFoldIntoCSel(const MachineRegisterInfo &MRI, unsigned VReg, + unsigned *NewVReg = nullptr) { + VReg = removeCopies(MRI, VReg); + if (!TargetRegisterInfo::isVirtualRegister(VReg)) + return 0; + + bool Is64Bit = AArch64::GPR64allRegClass.hasSubClassEq(MRI.getRegClass(VReg)); + const MachineInstr *DefMI = MRI.getVRegDef(VReg); + unsigned Opc = 0; + unsigned SrcOpNum = 0; + switch (DefMI->getOpcode()) { + case AArch64::ADDSXri: + case AArch64::ADDSWri: + // if NZCV is used, do not fold. + if (DefMI->findRegisterDefOperandIdx(AArch64::NZCV, true) == -1) + return 0; + // fall-through to ADDXri and ADDWri. + case AArch64::ADDXri: + case AArch64::ADDWri: + // add x, 1 -> csinc. + if (!DefMI->getOperand(2).isImm() || DefMI->getOperand(2).getImm() != 1 || + DefMI->getOperand(3).getImm() != 0) + return 0; + SrcOpNum = 1; + Opc = Is64Bit ? AArch64::CSINCXr : AArch64::CSINCWr; + break; + + case AArch64::ORNXrr: + case AArch64::ORNWrr: { + // not x -> csinv, represented as orn dst, xzr, src. + unsigned ZReg = removeCopies(MRI, DefMI->getOperand(1).getReg()); + if (ZReg != AArch64::XZR && ZReg != AArch64::WZR) + return 0; + SrcOpNum = 2; + Opc = Is64Bit ? AArch64::CSINVXr : AArch64::CSINVWr; + break; + } + + case AArch64::SUBSXrr: + case AArch64::SUBSWrr: + // if NZCV is used, do not fold. + if (DefMI->findRegisterDefOperandIdx(AArch64::NZCV, true) == -1) + return 0; + // fall-through to SUBXrr and SUBWrr. + case AArch64::SUBXrr: + case AArch64::SUBWrr: { + // neg x -> csneg, represented as sub dst, xzr, src. + unsigned ZReg = removeCopies(MRI, DefMI->getOperand(1).getReg()); + if (ZReg != AArch64::XZR && ZReg != AArch64::WZR) + return 0; + SrcOpNum = 2; + Opc = Is64Bit ? AArch64::CSNEGXr : AArch64::CSNEGWr; + break; + } default: - return false; + return 0; } + assert(Opc && SrcOpNum && "Missing parameters"); - return false; + if (NewVReg) + *NewVReg = DefMI->getOperand(SrcOpNum).getReg(); + return Opc; } -void -AArch64InstrInfo::storeRegToStackSlot(MachineBasicBlock &MBB, - MachineBasicBlock::iterator MBBI, - unsigned SrcReg, bool isKill, - int FrameIdx, - const TargetRegisterClass *RC, - const TargetRegisterInfo *TRI) const { - DebugLoc DL = MBB.findDebugLoc(MBBI); - MachineFunction &MF = *MBB.getParent(); - MachineFrameInfo &MFI = *MF.getFrameInfo(); - unsigned Align = MFI.getObjectAlignment(FrameIdx); - - MachineMemOperand *MMO - = MF.getMachineMemOperand(MachinePointerInfo::getFixedStack(FrameIdx), - MachineMemOperand::MOStore, - MFI.getObjectSize(FrameIdx), - Align); - - unsigned StoreOp = 0; - if (RC->hasType(MVT::i64) || RC->hasType(MVT::i32)) { - switch(RC->getSize()) { - case 4: StoreOp = AArch64::LS32_STR; break; - case 8: StoreOp = AArch64::LS64_STR; break; - default: - llvm_unreachable("Unknown size for regclass"); - } - } else { - assert((RC->hasType(MVT::f32) || RC->hasType(MVT::f64) || - RC->hasType(MVT::f128)) - && "Expected integer or floating type for store"); - switch (RC->getSize()) { - case 4: StoreOp = AArch64::LSFP32_STR; break; - case 8: StoreOp = AArch64::LSFP64_STR; break; - case 16: StoreOp = AArch64::LSFP128_STR; break; - default: - llvm_unreachable("Unknown size for regclass"); - } +bool AArch64InstrInfo::canInsertSelect( + const MachineBasicBlock &MBB, ArrayRef Cond, + unsigned TrueReg, unsigned FalseReg, int &CondCycles, int &TrueCycles, + int &FalseCycles) const { + // Check register classes. + const MachineRegisterInfo &MRI = MBB.getParent()->getRegInfo(); + const TargetRegisterClass *RC = + RI.getCommonSubClass(MRI.getRegClass(TrueReg), MRI.getRegClass(FalseReg)); + if (!RC) + return false; + + // Expanding cbz/tbz requires an extra cycle of latency on the condition. + unsigned ExtraCondLat = Cond.size() != 1; + + // GPRs are handled by csel. + // FIXME: Fold in x+1, -x, and ~x when applicable. + if (AArch64::GPR64allRegClass.hasSubClassEq(RC) || + AArch64::GPR32allRegClass.hasSubClassEq(RC)) { + // Single-cycle csel, csinc, csinv, and csneg. + CondCycles = 1 + ExtraCondLat; + TrueCycles = FalseCycles = 1; + if (canFoldIntoCSel(MRI, TrueReg)) + TrueCycles = 0; + else if (canFoldIntoCSel(MRI, FalseReg)) + FalseCycles = 0; + return true; } - MachineInstrBuilder NewMI = BuildMI(MBB, MBBI, DL, get(StoreOp)); - NewMI.addReg(SrcReg, getKillRegState(isKill)) - .addFrameIndex(FrameIdx) - .addImm(0) - .addMemOperand(MMO); + // Scalar floating point is handled by fcsel. + // FIXME: Form fabs, fmin, and fmax when applicable. + if (AArch64::FPR64RegClass.hasSubClassEq(RC) || + AArch64::FPR32RegClass.hasSubClassEq(RC)) { + CondCycles = 5 + ExtraCondLat; + TrueCycles = FalseCycles = 2; + return true; + } + // Can't do vectors. + return false; } -void -AArch64InstrInfo::loadRegFromStackSlot(MachineBasicBlock &MBB, - MachineBasicBlock::iterator MBBI, - unsigned DestReg, int FrameIdx, - const TargetRegisterClass *RC, - const TargetRegisterInfo *TRI) const { - DebugLoc DL = MBB.findDebugLoc(MBBI); - MachineFunction &MF = *MBB.getParent(); - MachineFrameInfo &MFI = *MF.getFrameInfo(); - unsigned Align = MFI.getObjectAlignment(FrameIdx); - - MachineMemOperand *MMO - = MF.getMachineMemOperand(MachinePointerInfo::getFixedStack(FrameIdx), - MachineMemOperand::MOLoad, - MFI.getObjectSize(FrameIdx), - Align); - - unsigned LoadOp = 0; - if (RC->hasType(MVT::i64) || RC->hasType(MVT::i32)) { - switch(RC->getSize()) { - case 4: LoadOp = AArch64::LS32_LDR; break; - case 8: LoadOp = AArch64::LS64_LDR; break; +void AArch64InstrInfo::insertSelect(MachineBasicBlock &MBB, + MachineBasicBlock::iterator I, DebugLoc DL, + unsigned DstReg, + ArrayRef Cond, + unsigned TrueReg, unsigned FalseReg) const { + MachineRegisterInfo &MRI = MBB.getParent()->getRegInfo(); + + // Parse the condition code, see parseCondBranch() above. + AArch64CC::CondCode CC; + switch (Cond.size()) { + default: + llvm_unreachable("Unknown condition opcode in Cond"); + case 1: // b.cc + CC = AArch64CC::CondCode(Cond[0].getImm()); + break; + case 3: { // cbz/cbnz + // We must insert a compare against 0. + bool Is64Bit; + switch (Cond[1].getImm()) { default: - llvm_unreachable("Unknown size for regclass"); + llvm_unreachable("Unknown branch opcode in Cond"); + case AArch64::CBZW: + Is64Bit = 0; + CC = AArch64CC::EQ; + break; + case AArch64::CBZX: + Is64Bit = 1; + CC = AArch64CC::EQ; + break; + case AArch64::CBNZW: + Is64Bit = 0; + CC = AArch64CC::NE; + break; + case AArch64::CBNZX: + Is64Bit = 1; + CC = AArch64CC::NE; + break; } - } else { - assert((RC->hasType(MVT::f32) || RC->hasType(MVT::f64) - || RC->hasType(MVT::f128)) - && "Expected integer or floating type for store"); - switch (RC->getSize()) { - case 4: LoadOp = AArch64::LSFP32_LDR; break; - case 8: LoadOp = AArch64::LSFP64_LDR; break; - case 16: LoadOp = AArch64::LSFP128_LDR; break; + unsigned SrcReg = Cond[2].getReg(); + if (Is64Bit) { + // cmp reg, #0 is actually subs xzr, reg, #0. + MRI.constrainRegClass(SrcReg, &AArch64::GPR64spRegClass); + BuildMI(MBB, I, DL, get(AArch64::SUBSXri), AArch64::XZR) + .addReg(SrcReg) + .addImm(0) + .addImm(0); + } else { + MRI.constrainRegClass(SrcReg, &AArch64::GPR32spRegClass); + BuildMI(MBB, I, DL, get(AArch64::SUBSWri), AArch64::WZR) + .addReg(SrcReg) + .addImm(0) + .addImm(0); + } + break; + } + case 4: { // tbz/tbnz + // We must insert a tst instruction. + switch (Cond[1].getImm()) { default: - llvm_unreachable("Unknown size for regclass"); + llvm_unreachable("Unknown branch opcode in Cond"); + case AArch64::TBZW: + case AArch64::TBZX: + CC = AArch64CC::EQ; + break; + case AArch64::TBNZW: + case AArch64::TBNZX: + CC = AArch64CC::NE; + break; } + // cmp reg, #foo is actually ands xzr, reg, #1<begin(), E = BB->end(); - I != E; ++I) { - for (unsigned i = 0, e = I->getNumOperands(); i != e; ++i) { - if (!I->getOperand(i).isFI()) continue; - - // When using ADDxxi_lsl0_s to get the address of a stack object, 0xfff - // is the largest offset guaranteed to fit in the immediate offset. - if (I->getOpcode() == AArch64::ADDxxi_lsl0_s) { - Limit = std::min(Limit, 0xfffu); - break; - } - - int AccessScale, MinOffset, MaxOffset; - getAddressConstraints(*I, AccessScale, MinOffset, MaxOffset); - Limit = std::min(Limit, static_cast(MaxOffset)); - - break; // At most one FI per instruction - } + unsigned Opc = 0; + const TargetRegisterClass *RC = nullptr; + bool TryFold = false; + if (MRI.constrainRegClass(DstReg, &AArch64::GPR64RegClass)) { + RC = &AArch64::GPR64RegClass; + Opc = AArch64::CSELXr; + TryFold = true; + } else if (MRI.constrainRegClass(DstReg, &AArch64::GPR32RegClass)) { + RC = &AArch64::GPR32RegClass; + Opc = AArch64::CSELWr; + TryFold = true; + } else if (MRI.constrainRegClass(DstReg, &AArch64::FPR64RegClass)) { + RC = &AArch64::FPR64RegClass; + Opc = AArch64::FCSELDrrr; + } else if (MRI.constrainRegClass(DstReg, &AArch64::FPR32RegClass)) { + RC = &AArch64::FPR32RegClass; + Opc = AArch64::FCSELSrrr; + } + assert(RC && "Unsupported regclass"); + + // Try folding simple instructions into the csel. + if (TryFold) { + unsigned NewVReg = 0; + unsigned FoldedOpc = canFoldIntoCSel(MRI, TrueReg, &NewVReg); + if (FoldedOpc) { + // The folded opcodes csinc, csinc and csneg apply the operation to + // FalseReg, so we need to invert the condition. + CC = AArch64CC::getInvertedCondCode(CC); + TrueReg = FalseReg; + } else + FoldedOpc = canFoldIntoCSel(MRI, FalseReg, &NewVReg); + + // Fold the operation. Leave any dead instructions for DCE to clean up. + if (FoldedOpc) { + FalseReg = NewVReg; + Opc = FoldedOpc; + // The extends the live range of NewVReg. + MRI.clearKillFlags(NewVReg); } } - return Limit; + // Pull all virtual register into the appropriate class. + MRI.constrainRegClass(TrueReg, RC); + MRI.constrainRegClass(FalseReg, RC); + + // Insert the csel. + BuildMI(MBB, I, DL, get(Opc), DstReg).addReg(TrueReg).addReg(FalseReg).addImm( + CC); } -void AArch64InstrInfo::getAddressConstraints(const MachineInstr &MI, - int &AccessScale, int &MinOffset, - int &MaxOffset) const { - switch (MI.getOpcode()) { - default: llvm_unreachable("Unkown load/store kind"); - case TargetOpcode::DBG_VALUE: - AccessScale = 1; - MinOffset = INT_MIN; - MaxOffset = INT_MAX; - return; - case AArch64::LS8_LDR: case AArch64::LS8_STR: - case AArch64::LSFP8_LDR: case AArch64::LSFP8_STR: - case AArch64::LDRSBw: - case AArch64::LDRSBx: - AccessScale = 1; - MinOffset = 0; - MaxOffset = 0xfff; - return; - case AArch64::LS16_LDR: case AArch64::LS16_STR: - case AArch64::LSFP16_LDR: case AArch64::LSFP16_STR: - case AArch64::LDRSHw: - case AArch64::LDRSHx: - AccessScale = 2; - MinOffset = 0; - MaxOffset = 0xfff * AccessScale; - return; - case AArch64::LS32_LDR: case AArch64::LS32_STR: - case AArch64::LSFP32_LDR: case AArch64::LSFP32_STR: - case AArch64::LDRSWx: - case AArch64::LDPSWx: - AccessScale = 4; - MinOffset = 0; - MaxOffset = 0xfff * AccessScale; - return; - case AArch64::LS64_LDR: case AArch64::LS64_STR: - case AArch64::LSFP64_LDR: case AArch64::LSFP64_STR: - case AArch64::PRFM: - AccessScale = 8; - MinOffset = 0; - MaxOffset = 0xfff * AccessScale; - return; - case AArch64::LSFP128_LDR: case AArch64::LSFP128_STR: - AccessScale = 16; - MinOffset = 0; - MaxOffset = 0xfff * AccessScale; - return; - case AArch64::LSPair32_LDR: case AArch64::LSPair32_STR: - case AArch64::LSFPPair32_LDR: case AArch64::LSFPPair32_STR: - AccessScale = 4; - MinOffset = -0x40 * AccessScale; - MaxOffset = 0x3f * AccessScale; - return; - case AArch64::LSPair64_LDR: case AArch64::LSPair64_STR: - case AArch64::LSFPPair64_LDR: case AArch64::LSFPPair64_STR: - AccessScale = 8; - MinOffset = -0x40 * AccessScale; - MaxOffset = 0x3f * AccessScale; - return; - case AArch64::LSFPPair128_LDR: case AArch64::LSFPPair128_STR: - AccessScale = 16; - MinOffset = -0x40 * AccessScale; - MaxOffset = 0x3f * AccessScale; - return; - } + +/// Returns true if a MOVi32imm or MOVi64imm can be expanded to an ORRxx. +static bool canBeExpandedToORR(const MachineInstr *MI, unsigned BitSize) { + uint64_t Imm = MI->getOperand(1).getImm(); + uint64_t UImm = Imm << (64 - BitSize) >> (64 - BitSize); + uint64_t Encoding; + return AArch64_AM::processLogicalImmediate(UImm, BitSize, Encoding); } -unsigned AArch64InstrInfo::getInstSizeInBytes(const MachineInstr &MI) const { - const MCInstrDesc &MCID = MI.getDesc(); - const MachineBasicBlock &MBB = *MI.getParent(); - const MachineFunction &MF = *MBB.getParent(); - const MCAsmInfo &MAI = *MF.getTarget().getMCAsmInfo(); +// FIXME: this implementation should be micro-architecture dependent, so a +// micro-architecture target hook should be introduced here in future. +bool AArch64InstrInfo::isAsCheapAsAMove(const MachineInstr *MI) const { + if (!Subtarget.isCortexA57() && !Subtarget.isCortexA53()) + return MI->isAsCheapAsAMove(); - if (MCID.getSize()) - return MCID.getSize(); + switch (MI->getOpcode()) { + default: + return false; - if (MI.getOpcode() == AArch64::INLINEASM) - return getInlineAsmLength(MI.getOperand(0).getSymbolName(), MAI); + // add/sub on register without shift + case AArch64::ADDWri: + case AArch64::ADDXri: + case AArch64::SUBWri: + case AArch64::SUBXri: + return (MI->getOperand(3).getImm() == 0); + + // logical ops on immediate + case AArch64::ANDWri: + case AArch64::ANDXri: + case AArch64::EORWri: + case AArch64::EORXri: + case AArch64::ORRWri: + case AArch64::ORRXri: + return true; - if (MI.isLabel()) - return 0; + // logical ops on register without shift + case AArch64::ANDWrr: + case AArch64::ANDXrr: + case AArch64::BICWrr: + case AArch64::BICXrr: + case AArch64::EONWrr: + case AArch64::EONXrr: + case AArch64::EORWrr: + case AArch64::EORXrr: + case AArch64::ORNWrr: + case AArch64::ORNXrr: + case AArch64::ORRWrr: + case AArch64::ORRXrr: + return true; + // If MOVi32imm or MOVi64imm can be expanded into ORRWri or + // ORRXri, it is as cheap as MOV + case AArch64::MOVi32imm: + return canBeExpandedToORR(MI, 32); + case AArch64::MOVi64imm: + return canBeExpandedToORR(MI, 64); + } + + llvm_unreachable("Unknown opcode to check as cheap as a move!"); +} +bool AArch64InstrInfo::isCoalescableExtInstr(const MachineInstr &MI, + unsigned &SrcReg, unsigned &DstReg, + unsigned &SubIdx) const { switch (MI.getOpcode()) { - case TargetOpcode::BUNDLE: - return getInstBundleLength(MI); - case TargetOpcode::IMPLICIT_DEF: - case TargetOpcode::KILL: - case TargetOpcode::PROLOG_LABEL: - case TargetOpcode::EH_LABEL: - case TargetOpcode::DBG_VALUE: - return 0; - case AArch64::TLSDESCCALL: - return 0; default: - llvm_unreachable("Unknown instruction class"); + return false; + case AArch64::SBFMXri: // aka sxtw + case AArch64::UBFMXri: // aka uxtw + // Check for the 32 -> 64 bit extension case, these instructions can do + // much more. + if (MI.getOperand(2).getImm() != 0 || MI.getOperand(3).getImm() != 31) + return false; + // This is a signed or unsigned 32 -> 64 bit extension. + SrcReg = MI.getOperand(1).getReg(); + DstReg = MI.getOperand(0).getReg(); + SubIdx = AArch64::sub_32; + return true; } } -unsigned AArch64InstrInfo::getInstBundleLength(const MachineInstr &MI) const { - unsigned Size = 0; - MachineBasicBlock::const_instr_iterator I = MI; - MachineBasicBlock::const_instr_iterator E = MI.getParent()->instr_end(); - while (++I != E && I->isInsideBundle()) { - assert(!I->isBundle() && "No nested bundle!"); - Size += getInstSizeInBytes(*I); +bool +AArch64InstrInfo::areMemAccessesTriviallyDisjoint(MachineInstr *MIa, + MachineInstr *MIb, + AliasAnalysis *AA) const { + const TargetRegisterInfo *TRI = &getRegisterInfo(); + unsigned BaseRegA = 0, BaseRegB = 0; + int OffsetA = 0, OffsetB = 0; + int WidthA = 0, WidthB = 0; + + assert(MIa && MIa->mayLoadOrStore() && "MIa must be a load or store."); + assert(MIb && MIb->mayLoadOrStore() && "MIb must be a load or store."); + + if (MIa->hasUnmodeledSideEffects() || MIb->hasUnmodeledSideEffects() || + MIa->hasOrderedMemoryRef() || MIb->hasOrderedMemoryRef()) + return false; + + // Retrieve the base register, offset from the base register and width. Width + // is the size of memory that is being loaded/stored (e.g. 1, 2, 4, 8). If + // base registers are identical, and the offset of a lower memory access + + // the width doesn't overlap the offset of a higher memory access, + // then the memory accesses are different. + if (getMemOpBaseRegImmOfsWidth(MIa, BaseRegA, OffsetA, WidthA, TRI) && + getMemOpBaseRegImmOfsWidth(MIb, BaseRegB, OffsetB, WidthB, TRI)) { + if (BaseRegA == BaseRegB) { + int LowOffset = OffsetA < OffsetB ? OffsetA : OffsetB; + int HighOffset = OffsetA < OffsetB ? OffsetB : OffsetA; + int LowWidth = (LowOffset == OffsetA) ? WidthA : WidthB; + if (LowOffset + LowWidth <= HighOffset) + return true; + } } - return Size; + return false; } -bool llvm::rewriteA64FrameIndex(MachineInstr &MI, unsigned FrameRegIdx, - unsigned FrameReg, int &Offset, - const AArch64InstrInfo &TII) { - MachineBasicBlock &MBB = *MI.getParent(); - MachineFunction &MF = *MBB.getParent(); - MachineFrameInfo &MFI = *MF.getFrameInfo(); +/// analyzeCompare - For a comparison instruction, return the source registers +/// in SrcReg and SrcReg2, and the value it compares against in CmpValue. +/// Return true if the comparison instruction can be analyzed. +bool AArch64InstrInfo::analyzeCompare(const MachineInstr *MI, unsigned &SrcReg, + unsigned &SrcReg2, int &CmpMask, + int &CmpValue) const { + switch (MI->getOpcode()) { + default: + break; + case AArch64::SUBSWrr: + case AArch64::SUBSWrs: + case AArch64::SUBSWrx: + case AArch64::SUBSXrr: + case AArch64::SUBSXrs: + case AArch64::SUBSXrx: + case AArch64::ADDSWrr: + case AArch64::ADDSWrs: + case AArch64::ADDSWrx: + case AArch64::ADDSXrr: + case AArch64::ADDSXrs: + case AArch64::ADDSXrx: + // Replace SUBSWrr with SUBWrr if NZCV is not used. + SrcReg = MI->getOperand(1).getReg(); + SrcReg2 = MI->getOperand(2).getReg(); + CmpMask = ~0; + CmpValue = 0; + return true; + case AArch64::SUBSWri: + case AArch64::ADDSWri: + case AArch64::SUBSXri: + case AArch64::ADDSXri: + SrcReg = MI->getOperand(1).getReg(); + SrcReg2 = 0; + CmpMask = ~0; + // FIXME: In order to convert CmpValue to 0 or 1 + CmpValue = (MI->getOperand(2).getImm() != 0); + return true; + case AArch64::ANDSWri: + case AArch64::ANDSXri: + // ANDS does not use the same encoding scheme as the others xxxS + // instructions. + SrcReg = MI->getOperand(1).getReg(); + SrcReg2 = 0; + CmpMask = ~0; + // FIXME:The return val type of decodeLogicalImmediate is uint64_t, + // while the type of CmpValue is int. When converting uint64_t to int, + // the high 32 bits of uint64_t will be lost. + // In fact it causes a bug in spec2006-483.xalancbmk + // CmpValue is only used to compare with zero in OptimizeCompareInstr + CmpValue = (AArch64_AM::decodeLogicalImmediate( + MI->getOperand(2).getImm(), + MI->getOpcode() == AArch64::ANDSWri ? 32 : 64) != 0); + return true; + } - MFI.getObjectOffset(FrameRegIdx); - llvm_unreachable("Unimplemented rewriteFrameIndex"); + return false; } -void llvm::emitRegUpdate(MachineBasicBlock &MBB, - MachineBasicBlock::iterator MBBI, - DebugLoc dl, const TargetInstrInfo &TII, - unsigned DstReg, unsigned SrcReg, unsigned ScratchReg, - int64_t NumBytes, MachineInstr::MIFlag MIFlags) { - if (NumBytes == 0 && DstReg == SrcReg) - return; - else if (abs64(NumBytes) & ~0xffffff) { - // Generically, we have to materialize the offset into a temporary register - // and subtract it. There are a couple of ways this could be done, for now - // we'll use a movz/movk or movn/movk sequence. - uint64_t Bits = static_cast(abs64(NumBytes)); - BuildMI(MBB, MBBI, dl, TII.get(AArch64::MOVZxii), ScratchReg) - .addImm(0xffff & Bits).addImm(0) - .setMIFlags(MIFlags); - - Bits >>= 16; - if (Bits & 0xffff) { - BuildMI(MBB, MBBI, dl, TII.get(AArch64::MOVKxii), ScratchReg) - .addReg(ScratchReg) - .addImm(0xffff & Bits).addImm(1) - .setMIFlags(MIFlags); - } - - Bits >>= 16; - if (Bits & 0xffff) { - BuildMI(MBB, MBBI, dl, TII.get(AArch64::MOVKxii), ScratchReg) - .addReg(ScratchReg) - .addImm(0xffff & Bits).addImm(2) - .setMIFlags(MIFlags); - } - - Bits >>= 16; - if (Bits & 0xffff) { - BuildMI(MBB, MBBI, dl, TII.get(AArch64::MOVKxii), ScratchReg) - .addReg(ScratchReg) - .addImm(0xffff & Bits).addImm(3) - .setMIFlags(MIFlags); - } - - // ADD DST, SRC, xTMP (, lsl #0) - unsigned AddOp = NumBytes > 0 ? AArch64::ADDxxx_uxtx : AArch64::SUBxxx_uxtx; - BuildMI(MBB, MBBI, dl, TII.get(AddOp), DstReg) - .addReg(SrcReg, RegState::Kill) - .addReg(ScratchReg, RegState::Kill) - .addImm(0) - .setMIFlag(MIFlags); - return; +static bool UpdateOperandRegClass(MachineInstr *Instr) { + MachineBasicBlock *MBB = Instr->getParent(); + assert(MBB && "Can't get MachineBasicBlock here"); + MachineFunction *MF = MBB->getParent(); + assert(MF && "Can't get MachineFunction here"); + const TargetInstrInfo *TII = MF->getSubtarget().getInstrInfo(); + const TargetRegisterInfo *TRI = MF->getSubtarget().getRegisterInfo(); + MachineRegisterInfo *MRI = &MF->getRegInfo(); + + for (unsigned OpIdx = 0, EndIdx = Instr->getNumOperands(); OpIdx < EndIdx; + ++OpIdx) { + MachineOperand &MO = Instr->getOperand(OpIdx); + const TargetRegisterClass *OpRegCstraints = + Instr->getRegClassConstraint(OpIdx, TII, TRI); + + // If there's no constraint, there's nothing to do. + if (!OpRegCstraints) + continue; + // If the operand is a frame index, there's nothing to do here. + // A frame index operand will resolve correctly during PEI. + if (MO.isFI()) + continue; + + assert(MO.isReg() && + "Operand has register constraints without being a register!"); + + unsigned Reg = MO.getReg(); + if (TargetRegisterInfo::isPhysicalRegister(Reg)) { + if (!OpRegCstraints->contains(Reg)) + return false; + } else if (!OpRegCstraints->hasSubClassEq(MRI->getRegClass(Reg)) && + !MRI->constrainRegClass(Reg, OpRegCstraints)) + return false; } - // Now we know that the adjustment can be done in at most two add/sub - // (immediate) instructions, which is always more efficient than a - // literal-pool load, or even a hypothetical movz/movk/add sequence + return true; +} - // Decide whether we're doing addition or subtraction - unsigned LowOp, HighOp; - if (NumBytes >= 0) { - LowOp = AArch64::ADDxxi_lsl0_s; - HighOp = AArch64::ADDxxi_lsl12_s; - } else { - LowOp = AArch64::SUBxxi_lsl0_s; - HighOp = AArch64::SUBxxi_lsl12_s; - NumBytes = abs64(NumBytes); +/// \brief Return the opcode that does not set flags when possible - otherwise +/// return the original opcode. The caller is responsible to do the actual +/// substitution and legality checking. +static unsigned convertFlagSettingOpcode(const MachineInstr *MI) { + // Don't convert all compare instructions, because for some the zero register + // encoding becomes the sp register. + bool MIDefinesZeroReg = false; + if (MI->definesRegister(AArch64::WZR) || MI->definesRegister(AArch64::XZR)) + MIDefinesZeroReg = true; + + switch (MI->getOpcode()) { + default: + return MI->getOpcode(); + case AArch64::ADDSWrr: + return AArch64::ADDWrr; + case AArch64::ADDSWri: + return MIDefinesZeroReg ? AArch64::ADDSWri : AArch64::ADDWri; + case AArch64::ADDSWrs: + return MIDefinesZeroReg ? AArch64::ADDSWrs : AArch64::ADDWrs; + case AArch64::ADDSWrx: + return AArch64::ADDWrx; + case AArch64::ADDSXrr: + return AArch64::ADDXrr; + case AArch64::ADDSXri: + return MIDefinesZeroReg ? AArch64::ADDSXri : AArch64::ADDXri; + case AArch64::ADDSXrs: + return MIDefinesZeroReg ? AArch64::ADDSXrs : AArch64::ADDXrs; + case AArch64::ADDSXrx: + return AArch64::ADDXrx; + case AArch64::SUBSWrr: + return AArch64::SUBWrr; + case AArch64::SUBSWri: + return MIDefinesZeroReg ? AArch64::SUBSWri : AArch64::SUBWri; + case AArch64::SUBSWrs: + return MIDefinesZeroReg ? AArch64::SUBSWrs : AArch64::SUBWrs; + case AArch64::SUBSWrx: + return AArch64::SUBWrx; + case AArch64::SUBSXrr: + return AArch64::SUBXrr; + case AArch64::SUBSXri: + return MIDefinesZeroReg ? AArch64::SUBSXri : AArch64::SUBXri; + case AArch64::SUBSXrs: + return MIDefinesZeroReg ? AArch64::SUBSXrs : AArch64::SUBXrs; + case AArch64::SUBSXrx: + return AArch64::SUBXrx; } +} - // If we're here, at the very least a move needs to be produced, which just - // happens to be materializable by an ADD. - if ((NumBytes & 0xfff) || NumBytes == 0) { - BuildMI(MBB, MBBI, dl, TII.get(LowOp), DstReg) - .addReg(SrcReg, RegState::Kill) - .addImm(NumBytes & 0xfff) - .setMIFlag(MIFlags); +/// True when condition code could be modified on the instruction +/// trace starting at from and ending at to. +static bool modifiesConditionCode(MachineInstr *From, MachineInstr *To, + const bool CheckOnlyCCWrites, + const TargetRegisterInfo *TRI) { + // We iterate backward starting \p To until we hit \p From + MachineBasicBlock::iterator I = To, E = From, B = To->getParent()->begin(); - // Next update should use the register we've just defined. - SrcReg = DstReg; - } + // Early exit if To is at the beginning of the BB. + if (I == B) + return true; + + // Check whether the definition of SrcReg is in the same basic block as + // Compare. If not, assume the condition code gets modified on some path. + if (To->getParent() != From->getParent()) + return true; - if (NumBytes & 0xfff000) { - BuildMI(MBB, MBBI, dl, TII.get(HighOp), DstReg) - .addReg(SrcReg, RegState::Kill) - .addImm(NumBytes >> 12) - .setMIFlag(MIFlags); + // Check that NZCV isn't set on the trace. + for (--I; I != E; --I) { + const MachineInstr &Instr = *I; + + if (Instr.modifiesRegister(AArch64::NZCV, TRI) || + (!CheckOnlyCCWrites && Instr.readsRegister(AArch64::NZCV, TRI))) + // This instruction modifies or uses NZCV after the one we want to + // change. + return true; + if (I == B) + // We currently don't allow the instruction trace to cross basic + // block boundaries + return true; } + return false; } +/// optimizeCompareInstr - Convert the instruction supplying the argument to the +/// comparison into one that sets the zero bit in the flags register. +bool AArch64InstrInfo::optimizeCompareInstr( + MachineInstr *CmpInstr, unsigned SrcReg, unsigned SrcReg2, int CmpMask, + int CmpValue, const MachineRegisterInfo *MRI) const { + + // Replace SUBSWrr with SUBWrr if NZCV is not used. + int Cmp_NZCV = CmpInstr->findRegisterDefOperandIdx(AArch64::NZCV, true); + if (Cmp_NZCV != -1) { + if (CmpInstr->definesRegister(AArch64::WZR) || + CmpInstr->definesRegister(AArch64::XZR)) { + CmpInstr->eraseFromParent(); + return true; + } + unsigned Opc = CmpInstr->getOpcode(); + unsigned NewOpc = convertFlagSettingOpcode(CmpInstr); + if (NewOpc == Opc) + return false; + const MCInstrDesc &MCID = get(NewOpc); + CmpInstr->setDesc(MCID); + CmpInstr->RemoveOperand(Cmp_NZCV); + bool succeeded = UpdateOperandRegClass(CmpInstr); + (void)succeeded; + assert(succeeded && "Some operands reg class are incompatible!"); + return true; + } -void llvm::emitSPUpdate(MachineBasicBlock &MBB, MachineBasicBlock::iterator MI, - DebugLoc dl, const TargetInstrInfo &TII, - unsigned ScratchReg, int64_t NumBytes, - MachineInstr::MIFlag MIFlags) { - emitRegUpdate(MBB, MI, dl, TII, AArch64::XSP, AArch64::XSP, AArch64::X16, - NumBytes, MIFlags); -} + // Continue only if we have a "ri" where immediate is zero. + // FIXME:CmpValue has already been converted to 0 or 1 in analyzeCompare + // function. + assert((CmpValue == 0 || CmpValue == 1) && "CmpValue must be 0 or 1!"); + if (CmpValue != 0 || SrcReg2 != 0) + return false; + + // CmpInstr is a Compare instruction if destination register is not used. + if (!MRI->use_nodbg_empty(CmpInstr->getOperand(0).getReg())) + return false; + // Get the unique definition of SrcReg. + MachineInstr *MI = MRI->getUniqueVRegDef(SrcReg); + if (!MI) + return false; -namespace { - struct LDTLSCleanup : public MachineFunctionPass { - static char ID; - LDTLSCleanup() : MachineFunctionPass(ID) {} + bool CheckOnlyCCWrites = false; + const TargetRegisterInfo *TRI = &getRegisterInfo(); + if (modifiesConditionCode(MI, CmpInstr, CheckOnlyCCWrites, TRI)) + return false; - virtual bool runOnMachineFunction(MachineFunction &MF) { - AArch64MachineFunctionInfo* MFI - = MF.getInfo(); - if (MFI->getNumLocalDynamicTLSAccesses() < 2) { - // No point folding accesses if there isn't at least two. - return false; - } + unsigned NewOpc = MI->getOpcode(); + switch (MI->getOpcode()) { + default: + return false; + case AArch64::ADDSWrr: + case AArch64::ADDSWri: + case AArch64::ADDSXrr: + case AArch64::ADDSXri: + case AArch64::SUBSWrr: + case AArch64::SUBSWri: + case AArch64::SUBSXrr: + case AArch64::SUBSXri: + break; + case AArch64::ADDWrr: NewOpc = AArch64::ADDSWrr; break; + case AArch64::ADDWri: NewOpc = AArch64::ADDSWri; break; + case AArch64::ADDXrr: NewOpc = AArch64::ADDSXrr; break; + case AArch64::ADDXri: NewOpc = AArch64::ADDSXri; break; + case AArch64::ADCWr: NewOpc = AArch64::ADCSWr; break; + case AArch64::ADCXr: NewOpc = AArch64::ADCSXr; break; + case AArch64::SUBWrr: NewOpc = AArch64::SUBSWrr; break; + case AArch64::SUBWri: NewOpc = AArch64::SUBSWri; break; + case AArch64::SUBXrr: NewOpc = AArch64::SUBSXrr; break; + case AArch64::SUBXri: NewOpc = AArch64::SUBSXri; break; + case AArch64::SBCWr: NewOpc = AArch64::SBCSWr; break; + case AArch64::SBCXr: NewOpc = AArch64::SBCSXr; break; + case AArch64::ANDWri: NewOpc = AArch64::ANDSWri; break; + case AArch64::ANDXri: NewOpc = AArch64::ANDSXri; break; + } - MachineDominatorTree *DT = &getAnalysis(); - return VisitNode(DT->getRootNode(), 0); - } - - // Visit the dominator subtree rooted at Node in pre-order. - // If TLSBaseAddrReg is non-null, then use that to replace any - // TLS_base_addr instructions. Otherwise, create the register - // when the first such instruction is seen, and then use it - // as we encounter more instructions. - bool VisitNode(MachineDomTreeNode *Node, unsigned TLSBaseAddrReg) { - MachineBasicBlock *BB = Node->getBlock(); - bool Changed = false; - - // Traverse the current block. - for (MachineBasicBlock::iterator I = BB->begin(), E = BB->end(); I != E; - ++I) { - switch (I->getOpcode()) { - case AArch64::TLSDESC_BLRx: - // Make sure it's a local dynamic access. - if (!I->getOperand(1).isSymbol() || - strcmp(I->getOperand(1).getSymbolName(), "_TLS_MODULE_BASE_")) - break; - - if (TLSBaseAddrReg) - I = ReplaceTLSBaseAddrCall(I, TLSBaseAddrReg); - else - I = SetRegister(I, &TLSBaseAddrReg); - Changed = true; - break; - default: - break; - } + // Scan forward for the use of NZCV. + // When checking against MI: if it's a conditional code requires + // checking of V bit, then this is not safe to do. + // It is safe to remove CmpInstr if NZCV is redefined or killed. + // If we are done with the basic block, we need to check whether NZCV is + // live-out. + bool IsSafe = false; + for (MachineBasicBlock::iterator I = CmpInstr, + E = CmpInstr->getParent()->end(); + !IsSafe && ++I != E;) { + const MachineInstr &Instr = *I; + for (unsigned IO = 0, EO = Instr.getNumOperands(); !IsSafe && IO != EO; + ++IO) { + const MachineOperand &MO = Instr.getOperand(IO); + if (MO.isRegMask() && MO.clobbersPhysReg(AArch64::NZCV)) { + IsSafe = true; + break; + } + if (!MO.isReg() || MO.getReg() != AArch64::NZCV) + continue; + if (MO.isDef()) { + IsSafe = true; + break; } - // Visit the children of this block in the dominator tree. - for (MachineDomTreeNode::iterator I = Node->begin(), E = Node->end(); - I != E; ++I) { - Changed |= VisitNode(*I, TLSBaseAddrReg); + // Decode the condition code. + unsigned Opc = Instr.getOpcode(); + AArch64CC::CondCode CC; + switch (Opc) { + default: + return false; + case AArch64::Bcc: + CC = (AArch64CC::CondCode)Instr.getOperand(IO - 2).getImm(); + break; + case AArch64::CSINVWr: + case AArch64::CSINVXr: + case AArch64::CSINCWr: + case AArch64::CSINCXr: + case AArch64::CSELWr: + case AArch64::CSELXr: + case AArch64::CSNEGWr: + case AArch64::CSNEGXr: + case AArch64::FCSELSrrr: + case AArch64::FCSELDrrr: + CC = (AArch64CC::CondCode)Instr.getOperand(IO - 1).getImm(); + break; } - return Changed; + // It is not safe to remove Compare instruction if Overflow(V) is used. + switch (CC) { + default: + // NZCV can be used multiple times, we should continue. + break; + case AArch64CC::VS: + case AArch64CC::VC: + case AArch64CC::GE: + case AArch64CC::LT: + case AArch64CC::GT: + case AArch64CC::LE: + return false; + } } + } - // Replace the TLS_base_addr instruction I with a copy from - // TLSBaseAddrReg, returning the new instruction. - MachineInstr *ReplaceTLSBaseAddrCall(MachineInstr *I, - unsigned TLSBaseAddrReg) { - MachineFunction *MF = I->getParent()->getParent(); - const AArch64TargetMachine *TM = - static_cast(&MF->getTarget()); - const AArch64InstrInfo *TII = TM->getInstrInfo(); - - // Insert a Copy from TLSBaseAddrReg to x0, which is where the rest of the - // code sequence assumes the address will be. - MachineInstr *Copy = BuildMI(*I->getParent(), I, I->getDebugLoc(), - TII->get(TargetOpcode::COPY), - AArch64::X0) - .addReg(TLSBaseAddrReg); + // If NZCV is not killed nor re-defined, we should check whether it is + // live-out. If it is live-out, do not optimize. + if (!IsSafe) { + MachineBasicBlock *ParentBlock = CmpInstr->getParent(); + for (auto *MBB : ParentBlock->successors()) + if (MBB->isLiveIn(AArch64::NZCV)) + return false; + } - // Erase the TLS_base_addr instruction. - I->eraseFromParent(); + // Update the instruction to set NZCV. + MI->setDesc(get(NewOpc)); + CmpInstr->eraseFromParent(); + bool succeeded = UpdateOperandRegClass(MI); + (void)succeeded; + assert(succeeded && "Some operands reg class are incompatible!"); + MI->addRegisterDefined(AArch64::NZCV, TRI); + return true; +} - return Copy; - } +bool +AArch64InstrInfo::expandPostRAPseudo(MachineBasicBlock::iterator MI) const { + if (MI->getOpcode() != TargetOpcode::LOAD_STACK_GUARD) + return false; - // Create a virtal register in *TLSBaseAddrReg, and populate it by - // inserting a copy instruction after I. Returns the new instruction. - MachineInstr *SetRegister(MachineInstr *I, unsigned *TLSBaseAddrReg) { - MachineFunction *MF = I->getParent()->getParent(); - const AArch64TargetMachine *TM = - static_cast(&MF->getTarget()); - const AArch64InstrInfo *TII = TM->getInstrInfo(); + MachineBasicBlock &MBB = *MI->getParent(); + DebugLoc DL = MI->getDebugLoc(); + unsigned Reg = MI->getOperand(0).getReg(); + const GlobalValue *GV = + cast((*MI->memoperands_begin())->getValue()); + const TargetMachine &TM = MBB.getParent()->getTarget(); + unsigned char OpFlags = Subtarget.ClassifyGlobalReference(GV, TM); + const unsigned char MO_NC = AArch64II::MO_NC; + + if ((OpFlags & AArch64II::MO_GOT) != 0) { + BuildMI(MBB, MI, DL, get(AArch64::LOADgot), Reg) + .addGlobalAddress(GV, 0, AArch64II::MO_GOT); + BuildMI(MBB, MI, DL, get(AArch64::LDRXui), Reg) + .addReg(Reg, RegState::Kill).addImm(0) + .addMemOperand(*MI->memoperands_begin()); + } else if (TM.getCodeModel() == CodeModel::Large) { + BuildMI(MBB, MI, DL, get(AArch64::MOVZXi), Reg) + .addGlobalAddress(GV, 0, AArch64II::MO_G3).addImm(48); + BuildMI(MBB, MI, DL, get(AArch64::MOVKXi), Reg) + .addReg(Reg, RegState::Kill) + .addGlobalAddress(GV, 0, AArch64II::MO_G2 | MO_NC).addImm(32); + BuildMI(MBB, MI, DL, get(AArch64::MOVKXi), Reg) + .addReg(Reg, RegState::Kill) + .addGlobalAddress(GV, 0, AArch64II::MO_G1 | MO_NC).addImm(16); + BuildMI(MBB, MI, DL, get(AArch64::MOVKXi), Reg) + .addReg(Reg, RegState::Kill) + .addGlobalAddress(GV, 0, AArch64II::MO_G0 | MO_NC).addImm(0); + BuildMI(MBB, MI, DL, get(AArch64::LDRXui), Reg) + .addReg(Reg, RegState::Kill).addImm(0) + .addMemOperand(*MI->memoperands_begin()); + } else { + BuildMI(MBB, MI, DL, get(AArch64::ADRP), Reg) + .addGlobalAddress(GV, 0, OpFlags | AArch64II::MO_PAGE); + unsigned char LoFlags = OpFlags | AArch64II::MO_PAGEOFF | MO_NC; + BuildMI(MBB, MI, DL, get(AArch64::LDRXui), Reg) + .addReg(Reg, RegState::Kill) + .addGlobalAddress(GV, 0, LoFlags) + .addMemOperand(*MI->memoperands_begin()); + } - // Create a virtual register for the TLS base address. - MachineRegisterInfo &RegInfo = MF->getRegInfo(); - *TLSBaseAddrReg = RegInfo.createVirtualRegister(&AArch64::GPR64RegClass); + MBB.erase(MI); - // Insert a copy from X0 to TLSBaseAddrReg for later. - MachineInstr *Next = I->getNextNode(); - MachineInstr *Copy = BuildMI(*I->getParent(), Next, I->getDebugLoc(), - TII->get(TargetOpcode::COPY), - *TLSBaseAddrReg) - .addReg(AArch64::X0); + return true; +} - return Copy; +/// Return true if this is this instruction has a non-zero immediate +bool AArch64InstrInfo::hasShiftedReg(const MachineInstr *MI) const { + switch (MI->getOpcode()) { + default: + break; + case AArch64::ADDSWrs: + case AArch64::ADDSXrs: + case AArch64::ADDWrs: + case AArch64::ADDXrs: + case AArch64::ANDSWrs: + case AArch64::ANDSXrs: + case AArch64::ANDWrs: + case AArch64::ANDXrs: + case AArch64::BICSWrs: + case AArch64::BICSXrs: + case AArch64::BICWrs: + case AArch64::BICXrs: + case AArch64::CRC32Brr: + case AArch64::CRC32CBrr: + case AArch64::CRC32CHrr: + case AArch64::CRC32CWrr: + case AArch64::CRC32CXrr: + case AArch64::CRC32Hrr: + case AArch64::CRC32Wrr: + case AArch64::CRC32Xrr: + case AArch64::EONWrs: + case AArch64::EONXrs: + case AArch64::EORWrs: + case AArch64::EORXrs: + case AArch64::ORNWrs: + case AArch64::ORNXrs: + case AArch64::ORRWrs: + case AArch64::ORRXrs: + case AArch64::SUBSWrs: + case AArch64::SUBSXrs: + case AArch64::SUBWrs: + case AArch64::SUBXrs: + if (MI->getOperand(3).isImm()) { + unsigned val = MI->getOperand(3).getImm(); + return (val != 0); } + break; + } + return false; +} - virtual const char *getPassName() const { - return "Local Dynamic TLS Access Clean-up"; +/// Return true if this is this instruction has a non-zero immediate +bool AArch64InstrInfo::hasExtendedReg(const MachineInstr *MI) const { + switch (MI->getOpcode()) { + default: + break; + case AArch64::ADDSWrx: + case AArch64::ADDSXrx: + case AArch64::ADDSXrx64: + case AArch64::ADDWrx: + case AArch64::ADDXrx: + case AArch64::ADDXrx64: + case AArch64::SUBSWrx: + case AArch64::SUBSXrx: + case AArch64::SUBSXrx64: + case AArch64::SUBWrx: + case AArch64::SUBXrx: + case AArch64::SUBXrx64: + if (MI->getOperand(3).isImm()) { + unsigned val = MI->getOperand(3).getImm(); + return (val != 0); } + break; + } + + return false; +} - virtual void getAnalysisUsage(AnalysisUsage &AU) const { - AU.setPreservesCFG(); - AU.addRequired(); - MachineFunctionPass::getAnalysisUsage(AU); +// Return true if this instruction simply sets its single destination register +// to zero. This is equivalent to a register rename of the zero-register. +bool AArch64InstrInfo::isGPRZero(const MachineInstr *MI) const { + switch (MI->getOpcode()) { + default: + break; + case AArch64::MOVZWi: + case AArch64::MOVZXi: // movz Rd, #0 (LSL #0) + if (MI->getOperand(1).isImm() && MI->getOperand(1).getImm() == 0) { + assert(MI->getDesc().getNumOperands() == 3 && + MI->getOperand(2).getImm() == 0 && "invalid MOVZi operands"); + return true; } - }; + break; + case AArch64::ANDWri: // and Rd, Rzr, #imm + return MI->getOperand(1).getReg() == AArch64::WZR; + case AArch64::ANDXri: + return MI->getOperand(1).getReg() == AArch64::XZR; + case TargetOpcode::COPY: + return MI->getOperand(1).getReg() == AArch64::WZR; + } + return false; } -char LDTLSCleanup::ID = 0; -FunctionPass* -llvm::createAArch64CleanupLocalDynamicTLSPass() { return new LDTLSCleanup(); } +// Return true if this instruction simply renames a general register without +// modifying bits. +bool AArch64InstrInfo::isGPRCopy(const MachineInstr *MI) const { + switch (MI->getOpcode()) { + default: + break; + case TargetOpcode::COPY: { + // GPR32 copies will by lowered to ORRXrs + unsigned DstReg = MI->getOperand(0).getReg(); + return (AArch64::GPR32RegClass.contains(DstReg) || + AArch64::GPR64RegClass.contains(DstReg)); + } + case AArch64::ORRXrs: // orr Xd, Xzr, Xm (LSL #0) + if (MI->getOperand(1).getReg() == AArch64::XZR) { + assert(MI->getDesc().getNumOperands() == 4 && + MI->getOperand(3).getImm() == 0 && "invalid ORRrs operands"); + return true; + } + break; + case AArch64::ADDXri: // add Xd, Xn, #0 (LSL #0) + if (MI->getOperand(2).getImm() == 0) { + assert(MI->getDesc().getNumOperands() == 4 && + MI->getOperand(3).getImm() == 0 && "invalid ADDXri operands"); + return true; + } + break; + } + return false; +} + +// Return true if this instruction simply renames a general register without +// modifying bits. +bool AArch64InstrInfo::isFPRCopy(const MachineInstr *MI) const { + switch (MI->getOpcode()) { + default: + break; + case TargetOpcode::COPY: { + // FPR64 copies will by lowered to ORR.16b + unsigned DstReg = MI->getOperand(0).getReg(); + return (AArch64::FPR64RegClass.contains(DstReg) || + AArch64::FPR128RegClass.contains(DstReg)); + } + case AArch64::ORRv16i8: + if (MI->getOperand(1).getReg() == MI->getOperand(2).getReg()) { + assert(MI->getDesc().getNumOperands() == 3 && MI->getOperand(0).isReg() && + "invalid ORRv16i8 operands"); + return true; + } + break; + } + return false; +} + +unsigned AArch64InstrInfo::isLoadFromStackSlot(const MachineInstr *MI, + int &FrameIndex) const { + switch (MI->getOpcode()) { + default: + break; + case AArch64::LDRWui: + case AArch64::LDRXui: + case AArch64::LDRBui: + case AArch64::LDRHui: + case AArch64::LDRSui: + case AArch64::LDRDui: + case AArch64::LDRQui: + if (MI->getOperand(0).getSubReg() == 0 && MI->getOperand(1).isFI() && + MI->getOperand(2).isImm() && MI->getOperand(2).getImm() == 0) { + FrameIndex = MI->getOperand(1).getIndex(); + return MI->getOperand(0).getReg(); + } + break; + } + + return 0; +} + +unsigned AArch64InstrInfo::isStoreToStackSlot(const MachineInstr *MI, + int &FrameIndex) const { + switch (MI->getOpcode()) { + default: + break; + case AArch64::STRWui: + case AArch64::STRXui: + case AArch64::STRBui: + case AArch64::STRHui: + case AArch64::STRSui: + case AArch64::STRDui: + case AArch64::STRQui: + if (MI->getOperand(0).getSubReg() == 0 && MI->getOperand(1).isFI() && + MI->getOperand(2).isImm() && MI->getOperand(2).getImm() == 0) { + FrameIndex = MI->getOperand(1).getIndex(); + return MI->getOperand(0).getReg(); + } + break; + } + return 0; +} + +/// Return true if this is load/store scales or extends its register offset. +/// This refers to scaling a dynamic index as opposed to scaled immediates. +/// MI should be a memory op that allows scaled addressing. +bool AArch64InstrInfo::isScaledAddr(const MachineInstr *MI) const { + switch (MI->getOpcode()) { + default: + break; + case AArch64::LDRBBroW: + case AArch64::LDRBroW: + case AArch64::LDRDroW: + case AArch64::LDRHHroW: + case AArch64::LDRHroW: + case AArch64::LDRQroW: + case AArch64::LDRSBWroW: + case AArch64::LDRSBXroW: + case AArch64::LDRSHWroW: + case AArch64::LDRSHXroW: + case AArch64::LDRSWroW: + case AArch64::LDRSroW: + case AArch64::LDRWroW: + case AArch64::LDRXroW: + case AArch64::STRBBroW: + case AArch64::STRBroW: + case AArch64::STRDroW: + case AArch64::STRHHroW: + case AArch64::STRHroW: + case AArch64::STRQroW: + case AArch64::STRSroW: + case AArch64::STRWroW: + case AArch64::STRXroW: + case AArch64::LDRBBroX: + case AArch64::LDRBroX: + case AArch64::LDRDroX: + case AArch64::LDRHHroX: + case AArch64::LDRHroX: + case AArch64::LDRQroX: + case AArch64::LDRSBWroX: + case AArch64::LDRSBXroX: + case AArch64::LDRSHWroX: + case AArch64::LDRSHXroX: + case AArch64::LDRSWroX: + case AArch64::LDRSroX: + case AArch64::LDRWroX: + case AArch64::LDRXroX: + case AArch64::STRBBroX: + case AArch64::STRBroX: + case AArch64::STRDroX: + case AArch64::STRHHroX: + case AArch64::STRHroX: + case AArch64::STRQroX: + case AArch64::STRSroX: + case AArch64::STRWroX: + case AArch64::STRXroX: + + unsigned Val = MI->getOperand(3).getImm(); + AArch64_AM::ShiftExtendType ExtType = AArch64_AM::getMemExtendType(Val); + return (ExtType != AArch64_AM::UXTX) || AArch64_AM::getMemDoShift(Val); + } + return false; +} + +/// Check all MachineMemOperands for a hint to suppress pairing. +bool AArch64InstrInfo::isLdStPairSuppressed(const MachineInstr *MI) const { + assert(MOSuppressPair < (1 << MachineMemOperand::MOTargetNumBits) && + "Too many target MO flags"); + for (auto *MM : MI->memoperands()) { + if (MM->getFlags() & + (MOSuppressPair << MachineMemOperand::MOTargetStartBit)) { + return true; + } + } + return false; +} + +/// Set a flag on the first MachineMemOperand to suppress pairing. +void AArch64InstrInfo::suppressLdStPair(MachineInstr *MI) const { + if (MI->memoperands_empty()) + return; + + assert(MOSuppressPair < (1 << MachineMemOperand::MOTargetNumBits) && + "Too many target MO flags"); + (*MI->memoperands_begin()) + ->setFlags(MOSuppressPair << MachineMemOperand::MOTargetStartBit); +} + +bool +AArch64InstrInfo::getMemOpBaseRegImmOfs(MachineInstr *LdSt, unsigned &BaseReg, + unsigned &Offset, + const TargetRegisterInfo *TRI) const { + switch (LdSt->getOpcode()) { + default: + return false; + case AArch64::STRSui: + case AArch64::STRDui: + case AArch64::STRQui: + case AArch64::STRXui: + case AArch64::STRWui: + case AArch64::LDRSui: + case AArch64::LDRDui: + case AArch64::LDRQui: + case AArch64::LDRXui: + case AArch64::LDRWui: + if (!LdSt->getOperand(1).isReg() || !LdSt->getOperand(2).isImm()) + return false; + BaseReg = LdSt->getOperand(1).getReg(); + MachineFunction &MF = *LdSt->getParent()->getParent(); + unsigned Width = getRegClass(LdSt->getDesc(), 0, TRI, MF)->getSize(); + Offset = LdSt->getOperand(2).getImm() * Width; + return true; + }; +} + +bool AArch64InstrInfo::getMemOpBaseRegImmOfsWidth( + MachineInstr *LdSt, unsigned &BaseReg, int &Offset, int &Width, + const TargetRegisterInfo *TRI) const { + // Handle only loads/stores with base register followed by immediate offset. + if (LdSt->getNumOperands() != 3) + return false; + if (!LdSt->getOperand(1).isReg() || !LdSt->getOperand(2).isImm()) + return false; + + // Offset is calculated as the immediate operand multiplied by the scaling factor. + // Unscaled instructions have scaling factor set to 1. + int Scale = 0; + switch (LdSt->getOpcode()) { + default: + return false; + case AArch64::LDURQi: + case AArch64::STURQi: + Width = 16; + Scale = 1; + break; + case AArch64::LDURXi: + case AArch64::LDURDi: + case AArch64::STURXi: + case AArch64::STURDi: + Width = 8; + Scale = 1; + break; + case AArch64::LDURWi: + case AArch64::LDURSi: + case AArch64::LDURSWi: + case AArch64::STURWi: + case AArch64::STURSi: + Width = 4; + Scale = 1; + break; + case AArch64::LDURHi: + case AArch64::LDURHHi: + case AArch64::LDURSHXi: + case AArch64::LDURSHWi: + case AArch64::STURHi: + case AArch64::STURHHi: + Width = 2; + Scale = 1; + break; + case AArch64::LDURBi: + case AArch64::LDURBBi: + case AArch64::LDURSBXi: + case AArch64::LDURSBWi: + case AArch64::STURBi: + case AArch64::STURBBi: + Width = 1; + Scale = 1; + break; + case AArch64::LDRQui: + case AArch64::STRQui: + Scale = Width = 16; + break; + case AArch64::LDRXui: + case AArch64::LDRDui: + case AArch64::STRXui: + case AArch64::STRDui: + Scale = Width = 8; + break; + case AArch64::LDRWui: + case AArch64::LDRSui: + case AArch64::STRWui: + case AArch64::STRSui: + Scale = Width = 4; + break; + case AArch64::LDRHui: + case AArch64::LDRHHui: + case AArch64::STRHui: + case AArch64::STRHHui: + Scale = Width = 2; + break; + case AArch64::LDRBui: + case AArch64::LDRBBui: + case AArch64::STRBui: + case AArch64::STRBBui: + Scale = Width = 1; + break; + }; + + BaseReg = LdSt->getOperand(1).getReg(); + Offset = LdSt->getOperand(2).getImm() * Scale; + return true; +} + +/// Detect opportunities for ldp/stp formation. +/// +/// Only called for LdSt for which getMemOpBaseRegImmOfs returns true. +bool AArch64InstrInfo::shouldClusterLoads(MachineInstr *FirstLdSt, + MachineInstr *SecondLdSt, + unsigned NumLoads) const { + // Only cluster up to a single pair. + if (NumLoads > 1) + return false; + if (FirstLdSt->getOpcode() != SecondLdSt->getOpcode()) + return false; + // getMemOpBaseRegImmOfs guarantees that oper 2 isImm. + unsigned Ofs1 = FirstLdSt->getOperand(2).getImm(); + // Allow 6 bits of positive range. + if (Ofs1 > 64) + return false; + // The caller should already have ordered First/SecondLdSt by offset. + unsigned Ofs2 = SecondLdSt->getOperand(2).getImm(); + return Ofs1 + 1 == Ofs2; +} + +bool AArch64InstrInfo::shouldScheduleAdjacent(MachineInstr *First, + MachineInstr *Second) const { + if (Subtarget.isCyclone()) { + // Cyclone can fuse CMN, CMP, TST followed by Bcc. + unsigned SecondOpcode = Second->getOpcode(); + if (SecondOpcode == AArch64::Bcc) { + switch (First->getOpcode()) { + default: + return false; + case AArch64::SUBSWri: + case AArch64::ADDSWri: + case AArch64::ANDSWri: + case AArch64::SUBSXri: + case AArch64::ADDSXri: + case AArch64::ANDSXri: + return true; + } + } + // Cyclone B0 also supports ALU operations followed by CBZ/CBNZ. + if (SecondOpcode == AArch64::CBNZW || SecondOpcode == AArch64::CBNZX || + SecondOpcode == AArch64::CBZW || SecondOpcode == AArch64::CBZX) { + switch (First->getOpcode()) { + default: + return false; + case AArch64::ADDWri: + case AArch64::ADDXri: + case AArch64::ANDWri: + case AArch64::ANDXri: + case AArch64::EORWri: + case AArch64::EORXri: + case AArch64::ORRWri: + case AArch64::ORRXri: + case AArch64::SUBWri: + case AArch64::SUBXri: + return true; + } + } + } + return false; +} + +MachineInstr *AArch64InstrInfo::emitFrameIndexDebugValue( + MachineFunction &MF, int FrameIx, uint64_t Offset, const MDNode *Var, + const MDNode *Expr, DebugLoc DL) const { + MachineInstrBuilder MIB = BuildMI(MF, DL, get(AArch64::DBG_VALUE)) + .addFrameIndex(FrameIx) + .addImm(0) + .addImm(Offset) + .addMetadata(Var) + .addMetadata(Expr); + return &*MIB; +} + +static const MachineInstrBuilder &AddSubReg(const MachineInstrBuilder &MIB, + unsigned Reg, unsigned SubIdx, + unsigned State, + const TargetRegisterInfo *TRI) { + if (!SubIdx) + return MIB.addReg(Reg, State); + + if (TargetRegisterInfo::isPhysicalRegister(Reg)) + return MIB.addReg(TRI->getSubReg(Reg, SubIdx), State); + return MIB.addReg(Reg, State, SubIdx); +} + +static bool forwardCopyWillClobberTuple(unsigned DestReg, unsigned SrcReg, + unsigned NumRegs) { + // We really want the positive remainder mod 32 here, that happens to be + // easily obtainable with a mask. + return ((DestReg - SrcReg) & 0x1f) < NumRegs; +} + +void AArch64InstrInfo::copyPhysRegTuple( + MachineBasicBlock &MBB, MachineBasicBlock::iterator I, DebugLoc DL, + unsigned DestReg, unsigned SrcReg, bool KillSrc, unsigned Opcode, + llvm::ArrayRef Indices) const { + assert(Subtarget.hasNEON() && + "Unexpected register copy without NEON"); + const TargetRegisterInfo *TRI = &getRegisterInfo(); + uint16_t DestEncoding = TRI->getEncodingValue(DestReg); + uint16_t SrcEncoding = TRI->getEncodingValue(SrcReg); + unsigned NumRegs = Indices.size(); + + int SubReg = 0, End = NumRegs, Incr = 1; + if (forwardCopyWillClobberTuple(DestEncoding, SrcEncoding, NumRegs)) { + SubReg = NumRegs - 1; + End = -1; + Incr = -1; + } + + for (; SubReg != End; SubReg += Incr) { + const MachineInstrBuilder MIB = BuildMI(MBB, I, DL, get(Opcode)); + AddSubReg(MIB, DestReg, Indices[SubReg], RegState::Define, TRI); + AddSubReg(MIB, SrcReg, Indices[SubReg], 0, TRI); + AddSubReg(MIB, SrcReg, Indices[SubReg], getKillRegState(KillSrc), TRI); + } +} + +void AArch64InstrInfo::copyPhysReg(MachineBasicBlock &MBB, + MachineBasicBlock::iterator I, DebugLoc DL, + unsigned DestReg, unsigned SrcReg, + bool KillSrc) const { + if (AArch64::GPR32spRegClass.contains(DestReg) && + (AArch64::GPR32spRegClass.contains(SrcReg) || SrcReg == AArch64::WZR)) { + const TargetRegisterInfo *TRI = &getRegisterInfo(); + + if (DestReg == AArch64::WSP || SrcReg == AArch64::WSP) { + // If either operand is WSP, expand to ADD #0. + if (Subtarget.hasZeroCycleRegMove()) { + // Cyclone recognizes "ADD Xd, Xn, #0" as a zero-cycle register move. + unsigned DestRegX = TRI->getMatchingSuperReg(DestReg, AArch64::sub_32, + &AArch64::GPR64spRegClass); + unsigned SrcRegX = TRI->getMatchingSuperReg(SrcReg, AArch64::sub_32, + &AArch64::GPR64spRegClass); + // This instruction is reading and writing X registers. This may upset + // the register scavenger and machine verifier, so we need to indicate + // that we are reading an undefined value from SrcRegX, but a proper + // value from SrcReg. + BuildMI(MBB, I, DL, get(AArch64::ADDXri), DestRegX) + .addReg(SrcRegX, RegState::Undef) + .addImm(0) + .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, 0)) + .addReg(SrcReg, RegState::Implicit | getKillRegState(KillSrc)); + } else { + BuildMI(MBB, I, DL, get(AArch64::ADDWri), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)) + .addImm(0) + .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, 0)); + } + } else if (SrcReg == AArch64::WZR && Subtarget.hasZeroCycleZeroing()) { + BuildMI(MBB, I, DL, get(AArch64::MOVZWi), DestReg).addImm(0).addImm( + AArch64_AM::getShifterImm(AArch64_AM::LSL, 0)); + } else { + if (Subtarget.hasZeroCycleRegMove()) { + // Cyclone recognizes "ORR Xd, XZR, Xm" as a zero-cycle register move. + unsigned DestRegX = TRI->getMatchingSuperReg(DestReg, AArch64::sub_32, + &AArch64::GPR64spRegClass); + unsigned SrcRegX = TRI->getMatchingSuperReg(SrcReg, AArch64::sub_32, + &AArch64::GPR64spRegClass); + // This instruction is reading and writing X registers. This may upset + // the register scavenger and machine verifier, so we need to indicate + // that we are reading an undefined value from SrcRegX, but a proper + // value from SrcReg. + BuildMI(MBB, I, DL, get(AArch64::ORRXrr), DestRegX) + .addReg(AArch64::XZR) + .addReg(SrcRegX, RegState::Undef) + .addReg(SrcReg, RegState::Implicit | getKillRegState(KillSrc)); + } else { + // Otherwise, expand to ORR WZR. + BuildMI(MBB, I, DL, get(AArch64::ORRWrr), DestReg) + .addReg(AArch64::WZR) + .addReg(SrcReg, getKillRegState(KillSrc)); + } + } + return; + } + + if (AArch64::GPR64spRegClass.contains(DestReg) && + (AArch64::GPR64spRegClass.contains(SrcReg) || SrcReg == AArch64::XZR)) { + if (DestReg == AArch64::SP || SrcReg == AArch64::SP) { + // If either operand is SP, expand to ADD #0. + BuildMI(MBB, I, DL, get(AArch64::ADDXri), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)) + .addImm(0) + .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, 0)); + } else if (SrcReg == AArch64::XZR && Subtarget.hasZeroCycleZeroing()) { + BuildMI(MBB, I, DL, get(AArch64::MOVZXi), DestReg).addImm(0).addImm( + AArch64_AM::getShifterImm(AArch64_AM::LSL, 0)); + } else { + // Otherwise, expand to ORR XZR. + BuildMI(MBB, I, DL, get(AArch64::ORRXrr), DestReg) + .addReg(AArch64::XZR) + .addReg(SrcReg, getKillRegState(KillSrc)); + } + return; + } + + // Copy a DDDD register quad by copying the individual sub-registers. + if (AArch64::DDDDRegClass.contains(DestReg) && + AArch64::DDDDRegClass.contains(SrcReg)) { + static const unsigned Indices[] = { AArch64::dsub0, AArch64::dsub1, + AArch64::dsub2, AArch64::dsub3 }; + copyPhysRegTuple(MBB, I, DL, DestReg, SrcReg, KillSrc, AArch64::ORRv8i8, + Indices); + return; + } + + // Copy a DDD register triple by copying the individual sub-registers. + if (AArch64::DDDRegClass.contains(DestReg) && + AArch64::DDDRegClass.contains(SrcReg)) { + static const unsigned Indices[] = { AArch64::dsub0, AArch64::dsub1, + AArch64::dsub2 }; + copyPhysRegTuple(MBB, I, DL, DestReg, SrcReg, KillSrc, AArch64::ORRv8i8, + Indices); + return; + } + + // Copy a DD register pair by copying the individual sub-registers. + if (AArch64::DDRegClass.contains(DestReg) && + AArch64::DDRegClass.contains(SrcReg)) { + static const unsigned Indices[] = { AArch64::dsub0, AArch64::dsub1 }; + copyPhysRegTuple(MBB, I, DL, DestReg, SrcReg, KillSrc, AArch64::ORRv8i8, + Indices); + return; + } + + // Copy a QQQQ register quad by copying the individual sub-registers. + if (AArch64::QQQQRegClass.contains(DestReg) && + AArch64::QQQQRegClass.contains(SrcReg)) { + static const unsigned Indices[] = { AArch64::qsub0, AArch64::qsub1, + AArch64::qsub2, AArch64::qsub3 }; + copyPhysRegTuple(MBB, I, DL, DestReg, SrcReg, KillSrc, AArch64::ORRv16i8, + Indices); + return; + } + + // Copy a QQQ register triple by copying the individual sub-registers. + if (AArch64::QQQRegClass.contains(DestReg) && + AArch64::QQQRegClass.contains(SrcReg)) { + static const unsigned Indices[] = { AArch64::qsub0, AArch64::qsub1, + AArch64::qsub2 }; + copyPhysRegTuple(MBB, I, DL, DestReg, SrcReg, KillSrc, AArch64::ORRv16i8, + Indices); + return; + } + + // Copy a QQ register pair by copying the individual sub-registers. + if (AArch64::QQRegClass.contains(DestReg) && + AArch64::QQRegClass.contains(SrcReg)) { + static const unsigned Indices[] = { AArch64::qsub0, AArch64::qsub1 }; + copyPhysRegTuple(MBB, I, DL, DestReg, SrcReg, KillSrc, AArch64::ORRv16i8, + Indices); + return; + } + + if (AArch64::FPR128RegClass.contains(DestReg) && + AArch64::FPR128RegClass.contains(SrcReg)) { + if(Subtarget.hasNEON()) { + BuildMI(MBB, I, DL, get(AArch64::ORRv16i8), DestReg) + .addReg(SrcReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } else { + BuildMI(MBB, I, DL, get(AArch64::STRQpre)) + .addReg(AArch64::SP, RegState::Define) + .addReg(SrcReg, getKillRegState(KillSrc)) + .addReg(AArch64::SP) + .addImm(-16); + BuildMI(MBB, I, DL, get(AArch64::LDRQpre)) + .addReg(AArch64::SP, RegState::Define) + .addReg(DestReg, RegState::Define) + .addReg(AArch64::SP) + .addImm(16); + } + return; + } + + if (AArch64::FPR64RegClass.contains(DestReg) && + AArch64::FPR64RegClass.contains(SrcReg)) { + if(Subtarget.hasNEON()) { + DestReg = RI.getMatchingSuperReg(DestReg, AArch64::dsub, + &AArch64::FPR128RegClass); + SrcReg = RI.getMatchingSuperReg(SrcReg, AArch64::dsub, + &AArch64::FPR128RegClass); + BuildMI(MBB, I, DL, get(AArch64::ORRv16i8), DestReg) + .addReg(SrcReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } else { + BuildMI(MBB, I, DL, get(AArch64::FMOVDr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } + return; + } + + if (AArch64::FPR32RegClass.contains(DestReg) && + AArch64::FPR32RegClass.contains(SrcReg)) { + if(Subtarget.hasNEON()) { + DestReg = RI.getMatchingSuperReg(DestReg, AArch64::ssub, + &AArch64::FPR128RegClass); + SrcReg = RI.getMatchingSuperReg(SrcReg, AArch64::ssub, + &AArch64::FPR128RegClass); + BuildMI(MBB, I, DL, get(AArch64::ORRv16i8), DestReg) + .addReg(SrcReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } else { + BuildMI(MBB, I, DL, get(AArch64::FMOVSr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } + return; + } + + if (AArch64::FPR16RegClass.contains(DestReg) && + AArch64::FPR16RegClass.contains(SrcReg)) { + if(Subtarget.hasNEON()) { + DestReg = RI.getMatchingSuperReg(DestReg, AArch64::hsub, + &AArch64::FPR128RegClass); + SrcReg = RI.getMatchingSuperReg(SrcReg, AArch64::hsub, + &AArch64::FPR128RegClass); + BuildMI(MBB, I, DL, get(AArch64::ORRv16i8), DestReg) + .addReg(SrcReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } else { + DestReg = RI.getMatchingSuperReg(DestReg, AArch64::hsub, + &AArch64::FPR32RegClass); + SrcReg = RI.getMatchingSuperReg(SrcReg, AArch64::hsub, + &AArch64::FPR32RegClass); + BuildMI(MBB, I, DL, get(AArch64::FMOVSr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } + return; + } + + if (AArch64::FPR8RegClass.contains(DestReg) && + AArch64::FPR8RegClass.contains(SrcReg)) { + if(Subtarget.hasNEON()) { + DestReg = RI.getMatchingSuperReg(DestReg, AArch64::bsub, + &AArch64::FPR128RegClass); + SrcReg = RI.getMatchingSuperReg(SrcReg, AArch64::bsub, + &AArch64::FPR128RegClass); + BuildMI(MBB, I, DL, get(AArch64::ORRv16i8), DestReg) + .addReg(SrcReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } else { + DestReg = RI.getMatchingSuperReg(DestReg, AArch64::bsub, + &AArch64::FPR32RegClass); + SrcReg = RI.getMatchingSuperReg(SrcReg, AArch64::bsub, + &AArch64::FPR32RegClass); + BuildMI(MBB, I, DL, get(AArch64::FMOVSr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + } + return; + } + + // Copies between GPR64 and FPR64. + if (AArch64::FPR64RegClass.contains(DestReg) && + AArch64::GPR64RegClass.contains(SrcReg)) { + BuildMI(MBB, I, DL, get(AArch64::FMOVXDr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + return; + } + if (AArch64::GPR64RegClass.contains(DestReg) && + AArch64::FPR64RegClass.contains(SrcReg)) { + BuildMI(MBB, I, DL, get(AArch64::FMOVDXr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + return; + } + // Copies between GPR32 and FPR32. + if (AArch64::FPR32RegClass.contains(DestReg) && + AArch64::GPR32RegClass.contains(SrcReg)) { + BuildMI(MBB, I, DL, get(AArch64::FMOVWSr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + return; + } + if (AArch64::GPR32RegClass.contains(DestReg) && + AArch64::FPR32RegClass.contains(SrcReg)) { + BuildMI(MBB, I, DL, get(AArch64::FMOVSWr), DestReg) + .addReg(SrcReg, getKillRegState(KillSrc)); + return; + } + + if (DestReg == AArch64::NZCV) { + assert(AArch64::GPR64RegClass.contains(SrcReg) && "Invalid NZCV copy"); + BuildMI(MBB, I, DL, get(AArch64::MSR)) + .addImm(AArch64SysReg::NZCV) + .addReg(SrcReg, getKillRegState(KillSrc)) + .addReg(AArch64::NZCV, RegState::Implicit | RegState::Define); + return; + } + + if (SrcReg == AArch64::NZCV) { + assert(AArch64::GPR64RegClass.contains(DestReg) && "Invalid NZCV copy"); + BuildMI(MBB, I, DL, get(AArch64::MRS)) + .addReg(DestReg) + .addImm(AArch64SysReg::NZCV) + .addReg(AArch64::NZCV, RegState::Implicit | getKillRegState(KillSrc)); + return; + } + + llvm_unreachable("unimplemented reg-to-reg copy"); +} + +void AArch64InstrInfo::storeRegToStackSlot( + MachineBasicBlock &MBB, MachineBasicBlock::iterator MBBI, unsigned SrcReg, + bool isKill, int FI, const TargetRegisterClass *RC, + const TargetRegisterInfo *TRI) const { + DebugLoc DL; + if (MBBI != MBB.end()) + DL = MBBI->getDebugLoc(); + MachineFunction &MF = *MBB.getParent(); + MachineFrameInfo &MFI = *MF.getFrameInfo(); + unsigned Align = MFI.getObjectAlignment(FI); + + MachinePointerInfo PtrInfo = MachinePointerInfo::getFixedStack(MF, FI); + MachineMemOperand *MMO = MF.getMachineMemOperand( + PtrInfo, MachineMemOperand::MOStore, MFI.getObjectSize(FI), Align); + unsigned Opc = 0; + bool Offset = true; + switch (RC->getSize()) { + case 1: + if (AArch64::FPR8RegClass.hasSubClassEq(RC)) + Opc = AArch64::STRBui; + break; + case 2: + if (AArch64::FPR16RegClass.hasSubClassEq(RC)) + Opc = AArch64::STRHui; + break; + case 4: + if (AArch64::GPR32allRegClass.hasSubClassEq(RC)) { + Opc = AArch64::STRWui; + if (TargetRegisterInfo::isVirtualRegister(SrcReg)) + MF.getRegInfo().constrainRegClass(SrcReg, &AArch64::GPR32RegClass); + else + assert(SrcReg != AArch64::WSP); + } else if (AArch64::FPR32RegClass.hasSubClassEq(RC)) + Opc = AArch64::STRSui; + break; + case 8: + if (AArch64::GPR64allRegClass.hasSubClassEq(RC)) { + Opc = AArch64::STRXui; + if (TargetRegisterInfo::isVirtualRegister(SrcReg)) + MF.getRegInfo().constrainRegClass(SrcReg, &AArch64::GPR64RegClass); + else + assert(SrcReg != AArch64::SP); + } else if (AArch64::FPR64RegClass.hasSubClassEq(RC)) + Opc = AArch64::STRDui; + break; + case 16: + if (AArch64::FPR128RegClass.hasSubClassEq(RC)) + Opc = AArch64::STRQui; + else if (AArch64::DDRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register store without NEON"); + Opc = AArch64::ST1Twov1d, Offset = false; + } + break; + case 24: + if (AArch64::DDDRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register store without NEON"); + Opc = AArch64::ST1Threev1d, Offset = false; + } + break; + case 32: + if (AArch64::DDDDRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register store without NEON"); + Opc = AArch64::ST1Fourv1d, Offset = false; + } else if (AArch64::QQRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register store without NEON"); + Opc = AArch64::ST1Twov2d, Offset = false; + } + break; + case 48: + if (AArch64::QQQRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register store without NEON"); + Opc = AArch64::ST1Threev2d, Offset = false; + } + break; + case 64: + if (AArch64::QQQQRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register store without NEON"); + Opc = AArch64::ST1Fourv2d, Offset = false; + } + break; + } + assert(Opc && "Unknown register class"); + + const MachineInstrBuilder MI = BuildMI(MBB, MBBI, DL, get(Opc)) + .addReg(SrcReg, getKillRegState(isKill)) + .addFrameIndex(FI); + + if (Offset) + MI.addImm(0); + MI.addMemOperand(MMO); +} + +void AArch64InstrInfo::loadRegFromStackSlot( + MachineBasicBlock &MBB, MachineBasicBlock::iterator MBBI, unsigned DestReg, + int FI, const TargetRegisterClass *RC, + const TargetRegisterInfo *TRI) const { + DebugLoc DL; + if (MBBI != MBB.end()) + DL = MBBI->getDebugLoc(); + MachineFunction &MF = *MBB.getParent(); + MachineFrameInfo &MFI = *MF.getFrameInfo(); + unsigned Align = MFI.getObjectAlignment(FI); + MachinePointerInfo PtrInfo = MachinePointerInfo::getFixedStack(MF, FI); + MachineMemOperand *MMO = MF.getMachineMemOperand( + PtrInfo, MachineMemOperand::MOLoad, MFI.getObjectSize(FI), Align); + + unsigned Opc = 0; + bool Offset = true; + switch (RC->getSize()) { + case 1: + if (AArch64::FPR8RegClass.hasSubClassEq(RC)) + Opc = AArch64::LDRBui; + break; + case 2: + if (AArch64::FPR16RegClass.hasSubClassEq(RC)) + Opc = AArch64::LDRHui; + break; + case 4: + if (AArch64::GPR32allRegClass.hasSubClassEq(RC)) { + Opc = AArch64::LDRWui; + if (TargetRegisterInfo::isVirtualRegister(DestReg)) + MF.getRegInfo().constrainRegClass(DestReg, &AArch64::GPR32RegClass); + else + assert(DestReg != AArch64::WSP); + } else if (AArch64::FPR32RegClass.hasSubClassEq(RC)) + Opc = AArch64::LDRSui; + break; + case 8: + if (AArch64::GPR64allRegClass.hasSubClassEq(RC)) { + Opc = AArch64::LDRXui; + if (TargetRegisterInfo::isVirtualRegister(DestReg)) + MF.getRegInfo().constrainRegClass(DestReg, &AArch64::GPR64RegClass); + else + assert(DestReg != AArch64::SP); + } else if (AArch64::FPR64RegClass.hasSubClassEq(RC)) + Opc = AArch64::LDRDui; + break; + case 16: + if (AArch64::FPR128RegClass.hasSubClassEq(RC)) + Opc = AArch64::LDRQui; + else if (AArch64::DDRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register load without NEON"); + Opc = AArch64::LD1Twov1d, Offset = false; + } + break; + case 24: + if (AArch64::DDDRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register load without NEON"); + Opc = AArch64::LD1Threev1d, Offset = false; + } + break; + case 32: + if (AArch64::DDDDRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register load without NEON"); + Opc = AArch64::LD1Fourv1d, Offset = false; + } else if (AArch64::QQRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register load without NEON"); + Opc = AArch64::LD1Twov2d, Offset = false; + } + break; + case 48: + if (AArch64::QQQRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register load without NEON"); + Opc = AArch64::LD1Threev2d, Offset = false; + } + break; + case 64: + if (AArch64::QQQQRegClass.hasSubClassEq(RC)) { + assert(Subtarget.hasNEON() && + "Unexpected register load without NEON"); + Opc = AArch64::LD1Fourv2d, Offset = false; + } + break; + } + assert(Opc && "Unknown register class"); + + const MachineInstrBuilder MI = BuildMI(MBB, MBBI, DL, get(Opc)) + .addReg(DestReg, getDefRegState(true)) + .addFrameIndex(FI); + if (Offset) + MI.addImm(0); + MI.addMemOperand(MMO); +} + +void llvm::emitFrameOffset(MachineBasicBlock &MBB, + MachineBasicBlock::iterator MBBI, DebugLoc DL, + unsigned DestReg, unsigned SrcReg, int Offset, + const TargetInstrInfo *TII, + MachineInstr::MIFlag Flag, bool SetNZCV) { + if (DestReg == SrcReg && Offset == 0) + return; + + bool isSub = Offset < 0; + if (isSub) + Offset = -Offset; + + // FIXME: If the offset won't fit in 24-bits, compute the offset into a + // scratch register. If DestReg is a virtual register, use it as the + // scratch register; otherwise, create a new virtual register (to be + // replaced by the scavenger at the end of PEI). That case can be optimized + // slightly if DestReg is SP which is always 16-byte aligned, so the scratch + // register can be loaded with offset%8 and the add/sub can use an extending + // instruction with LSL#3. + // Currently the function handles any offsets but generates a poor sequence + // of code. + // assert(Offset < (1 << 24) && "unimplemented reg plus immediate"); + + unsigned Opc; + if (SetNZCV) + Opc = isSub ? AArch64::SUBSXri : AArch64::ADDSXri; + else + Opc = isSub ? AArch64::SUBXri : AArch64::ADDXri; + const unsigned MaxEncoding = 0xfff; + const unsigned ShiftSize = 12; + const unsigned MaxEncodableValue = MaxEncoding << ShiftSize; + while (((unsigned)Offset) >= (1 << ShiftSize)) { + unsigned ThisVal; + if (((unsigned)Offset) > MaxEncodableValue) { + ThisVal = MaxEncodableValue; + } else { + ThisVal = Offset & MaxEncodableValue; + } + assert((ThisVal >> ShiftSize) <= MaxEncoding && + "Encoding cannot handle value that big"); + BuildMI(MBB, MBBI, DL, TII->get(Opc), DestReg) + .addReg(SrcReg) + .addImm(ThisVal >> ShiftSize) + .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, ShiftSize)) + .setMIFlag(Flag); + + SrcReg = DestReg; + Offset -= ThisVal; + if (Offset == 0) + return; + } + BuildMI(MBB, MBBI, DL, TII->get(Opc), DestReg) + .addReg(SrcReg) + .addImm(Offset) + .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, 0)) + .setMIFlag(Flag); +} + +MachineInstr *AArch64InstrInfo::foldMemoryOperandImpl( + MachineFunction &MF, MachineInstr *MI, ArrayRef Ops, + MachineBasicBlock::iterator InsertPt, int FrameIndex) const { + // This is a bit of a hack. Consider this instruction: + // + // %vreg0 = COPY %SP; GPR64all:%vreg0 + // + // We explicitly chose GPR64all for the virtual register so such a copy might + // be eliminated by RegisterCoalescer. However, that may not be possible, and + // %vreg0 may even spill. We can't spill %SP, and since it is in the GPR64all + // register class, TargetInstrInfo::foldMemoryOperand() is going to try. + // + // To prevent that, we are going to constrain the %vreg0 register class here. + // + // + // + if (MI->isCopy()) { + unsigned DstReg = MI->getOperand(0).getReg(); + unsigned SrcReg = MI->getOperand(1).getReg(); + if (SrcReg == AArch64::SP && + TargetRegisterInfo::isVirtualRegister(DstReg)) { + MF.getRegInfo().constrainRegClass(DstReg, &AArch64::GPR64RegClass); + return nullptr; + } + if (DstReg == AArch64::SP && + TargetRegisterInfo::isVirtualRegister(SrcReg)) { + MF.getRegInfo().constrainRegClass(SrcReg, &AArch64::GPR64RegClass); + return nullptr; + } + } + + // Cannot fold. + return nullptr; +} + +int llvm::isAArch64FrameOffsetLegal(const MachineInstr &MI, int &Offset, + bool *OutUseUnscaledOp, + unsigned *OutUnscaledOp, + int *EmittableOffset) { + int Scale = 1; + bool IsSigned = false; + // The ImmIdx should be changed case by case if it is not 2. + unsigned ImmIdx = 2; + unsigned UnscaledOp = 0; + // Set output values in case of early exit. + if (EmittableOffset) + *EmittableOffset = 0; + if (OutUseUnscaledOp) + *OutUseUnscaledOp = false; + if (OutUnscaledOp) + *OutUnscaledOp = 0; + switch (MI.getOpcode()) { + default: + llvm_unreachable("unhandled opcode in rewriteAArch64FrameIndex"); + // Vector spills/fills can't take an immediate offset. + case AArch64::LD1Twov2d: + case AArch64::LD1Threev2d: + case AArch64::LD1Fourv2d: + case AArch64::LD1Twov1d: + case AArch64::LD1Threev1d: + case AArch64::LD1Fourv1d: + case AArch64::ST1Twov2d: + case AArch64::ST1Threev2d: + case AArch64::ST1Fourv2d: + case AArch64::ST1Twov1d: + case AArch64::ST1Threev1d: + case AArch64::ST1Fourv1d: + return AArch64FrameOffsetCannotUpdate; + case AArch64::PRFMui: + Scale = 8; + UnscaledOp = AArch64::PRFUMi; + break; + case AArch64::LDRXui: + Scale = 8; + UnscaledOp = AArch64::LDURXi; + break; + case AArch64::LDRWui: + Scale = 4; + UnscaledOp = AArch64::LDURWi; + break; + case AArch64::LDRBui: + Scale = 1; + UnscaledOp = AArch64::LDURBi; + break; + case AArch64::LDRHui: + Scale = 2; + UnscaledOp = AArch64::LDURHi; + break; + case AArch64::LDRSui: + Scale = 4; + UnscaledOp = AArch64::LDURSi; + break; + case AArch64::LDRDui: + Scale = 8; + UnscaledOp = AArch64::LDURDi; + break; + case AArch64::LDRQui: + Scale = 16; + UnscaledOp = AArch64::LDURQi; + break; + case AArch64::LDRBBui: + Scale = 1; + UnscaledOp = AArch64::LDURBBi; + break; + case AArch64::LDRHHui: + Scale = 2; + UnscaledOp = AArch64::LDURHHi; + break; + case AArch64::LDRSBXui: + Scale = 1; + UnscaledOp = AArch64::LDURSBXi; + break; + case AArch64::LDRSBWui: + Scale = 1; + UnscaledOp = AArch64::LDURSBWi; + break; + case AArch64::LDRSHXui: + Scale = 2; + UnscaledOp = AArch64::LDURSHXi; + break; + case AArch64::LDRSHWui: + Scale = 2; + UnscaledOp = AArch64::LDURSHWi; + break; + case AArch64::LDRSWui: + Scale = 4; + UnscaledOp = AArch64::LDURSWi; + break; + + case AArch64::STRXui: + Scale = 8; + UnscaledOp = AArch64::STURXi; + break; + case AArch64::STRWui: + Scale = 4; + UnscaledOp = AArch64::STURWi; + break; + case AArch64::STRBui: + Scale = 1; + UnscaledOp = AArch64::STURBi; + break; + case AArch64::STRHui: + Scale = 2; + UnscaledOp = AArch64::STURHi; + break; + case AArch64::STRSui: + Scale = 4; + UnscaledOp = AArch64::STURSi; + break; + case AArch64::STRDui: + Scale = 8; + UnscaledOp = AArch64::STURDi; + break; + case AArch64::STRQui: + Scale = 16; + UnscaledOp = AArch64::STURQi; + break; + case AArch64::STRBBui: + Scale = 1; + UnscaledOp = AArch64::STURBBi; + break; + case AArch64::STRHHui: + Scale = 2; + UnscaledOp = AArch64::STURHHi; + break; + + case AArch64::LDPXi: + case AArch64::LDPDi: + case AArch64::STPXi: + case AArch64::STPDi: + case AArch64::LDNPXi: + case AArch64::LDNPDi: + case AArch64::STNPXi: + case AArch64::STNPDi: + ImmIdx = 3; + IsSigned = true; + Scale = 8; + break; + case AArch64::LDPQi: + case AArch64::STPQi: + case AArch64::LDNPQi: + case AArch64::STNPQi: + ImmIdx = 3; + IsSigned = true; + Scale = 16; + break; + case AArch64::LDPWi: + case AArch64::LDPSi: + case AArch64::STPWi: + case AArch64::STPSi: + case AArch64::LDNPWi: + case AArch64::LDNPSi: + case AArch64::STNPWi: + case AArch64::STNPSi: + ImmIdx = 3; + IsSigned = true; + Scale = 4; + break; + + case AArch64::LDURXi: + case AArch64::LDURWi: + case AArch64::LDURBi: + case AArch64::LDURHi: + case AArch64::LDURSi: + case AArch64::LDURDi: + case AArch64::LDURQi: + case AArch64::LDURHHi: + case AArch64::LDURBBi: + case AArch64::LDURSBXi: + case AArch64::LDURSBWi: + case AArch64::LDURSHXi: + case AArch64::LDURSHWi: + case AArch64::LDURSWi: + case AArch64::STURXi: + case AArch64::STURWi: + case AArch64::STURBi: + case AArch64::STURHi: + case AArch64::STURSi: + case AArch64::STURDi: + case AArch64::STURQi: + case AArch64::STURBBi: + case AArch64::STURHHi: + Scale = 1; + break; + } + + Offset += MI.getOperand(ImmIdx).getImm() * Scale; + + bool useUnscaledOp = false; + // If the offset doesn't match the scale, we rewrite the instruction to + // use the unscaled instruction instead. Likewise, if we have a negative + // offset (and have an unscaled op to use). + if ((Offset & (Scale - 1)) != 0 || (Offset < 0 && UnscaledOp != 0)) + useUnscaledOp = true; + + // Use an unscaled addressing mode if the instruction has a negative offset + // (or if the instruction is already using an unscaled addressing mode). + unsigned MaskBits; + if (IsSigned) { + // ldp/stp instructions. + MaskBits = 7; + Offset /= Scale; + } else if (UnscaledOp == 0 || useUnscaledOp) { + MaskBits = 9; + IsSigned = true; + Scale = 1; + } else { + MaskBits = 12; + IsSigned = false; + Offset /= Scale; + } + + // Attempt to fold address computation. + int MaxOff = (1 << (MaskBits - IsSigned)) - 1; + int MinOff = (IsSigned ? (-MaxOff - 1) : 0); + if (Offset >= MinOff && Offset <= MaxOff) { + if (EmittableOffset) + *EmittableOffset = Offset; + Offset = 0; + } else { + int NewOff = Offset < 0 ? MinOff : MaxOff; + if (EmittableOffset) + *EmittableOffset = NewOff; + Offset = (Offset - NewOff) * Scale; + } + if (OutUseUnscaledOp) + *OutUseUnscaledOp = useUnscaledOp; + if (OutUnscaledOp) + *OutUnscaledOp = UnscaledOp; + return AArch64FrameOffsetCanUpdate | + (Offset == 0 ? AArch64FrameOffsetIsLegal : 0); +} + +bool llvm::rewriteAArch64FrameIndex(MachineInstr &MI, unsigned FrameRegIdx, + unsigned FrameReg, int &Offset, + const AArch64InstrInfo *TII) { + unsigned Opcode = MI.getOpcode(); + unsigned ImmIdx = FrameRegIdx + 1; + + if (Opcode == AArch64::ADDSXri || Opcode == AArch64::ADDXri) { + Offset += MI.getOperand(ImmIdx).getImm(); + emitFrameOffset(*MI.getParent(), MI, MI.getDebugLoc(), + MI.getOperand(0).getReg(), FrameReg, Offset, TII, + MachineInstr::NoFlags, (Opcode == AArch64::ADDSXri)); + MI.eraseFromParent(); + Offset = 0; + return true; + } + + int NewOffset; + unsigned UnscaledOp; + bool UseUnscaledOp; + int Status = isAArch64FrameOffsetLegal(MI, Offset, &UseUnscaledOp, + &UnscaledOp, &NewOffset); + if (Status & AArch64FrameOffsetCanUpdate) { + if (Status & AArch64FrameOffsetIsLegal) + // Replace the FrameIndex with FrameReg. + MI.getOperand(FrameRegIdx).ChangeToRegister(FrameReg, false); + if (UseUnscaledOp) + MI.setDesc(TII->get(UnscaledOp)); + + MI.getOperand(ImmIdx).ChangeToImmediate(NewOffset); + return Offset == 0; + } + + return false; +} + +void AArch64InstrInfo::getNoopForMachoTarget(MCInst &NopInst) const { + NopInst.setOpcode(AArch64::HINT); + NopInst.addOperand(MCOperand::createImm(0)); +} +/// useMachineCombiner - return true when a target supports MachineCombiner +bool AArch64InstrInfo::useMachineCombiner() const { + // AArch64 supports the combiner + return true; +} +// +// True when Opc sets flag +static bool isCombineInstrSettingFlag(unsigned Opc) { + switch (Opc) { + case AArch64::ADDSWrr: + case AArch64::ADDSWri: + case AArch64::ADDSXrr: + case AArch64::ADDSXri: + case AArch64::SUBSWrr: + case AArch64::SUBSXrr: + // Note: MSUB Wd,Wn,Wm,Wi -> Wd = Wi - WnxWm, not Wd=WnxWm - Wi. + case AArch64::SUBSWri: + case AArch64::SUBSXri: + return true; + default: + break; + } + return false; +} +// +// 32b Opcodes that can be combined with a MUL +static bool isCombineInstrCandidate32(unsigned Opc) { + switch (Opc) { + case AArch64::ADDWrr: + case AArch64::ADDWri: + case AArch64::SUBWrr: + case AArch64::ADDSWrr: + case AArch64::ADDSWri: + case AArch64::SUBSWrr: + // Note: MSUB Wd,Wn,Wm,Wi -> Wd = Wi - WnxWm, not Wd=WnxWm - Wi. + case AArch64::SUBWri: + case AArch64::SUBSWri: + return true; + default: + break; + } + return false; +} +// +// 64b Opcodes that can be combined with a MUL +static bool isCombineInstrCandidate64(unsigned Opc) { + switch (Opc) { + case AArch64::ADDXrr: + case AArch64::ADDXri: + case AArch64::SUBXrr: + case AArch64::ADDSXrr: + case AArch64::ADDSXri: + case AArch64::SUBSXrr: + // Note: MSUB Wd,Wn,Wm,Wi -> Wd = Wi - WnxWm, not Wd=WnxWm - Wi. + case AArch64::SUBXri: + case AArch64::SUBSXri: + return true; + default: + break; + } + return false; +} +// +// Opcodes that can be combined with a MUL +static bool isCombineInstrCandidate(unsigned Opc) { + return (isCombineInstrCandidate32(Opc) || isCombineInstrCandidate64(Opc)); +} + +static bool canCombineWithMUL(MachineBasicBlock &MBB, MachineOperand &MO, + unsigned MulOpc, unsigned ZeroReg) { + MachineRegisterInfo &MRI = MBB.getParent()->getRegInfo(); + MachineInstr *MI = nullptr; + // We need a virtual register definition. + if (MO.isReg() && TargetRegisterInfo::isVirtualRegister(MO.getReg())) + MI = MRI.getUniqueVRegDef(MO.getReg()); + // And it needs to be in the trace (otherwise, it won't have a depth). + if (!MI || MI->getParent() != &MBB || (unsigned)MI->getOpcode() != MulOpc) + return false; + + assert(MI->getNumOperands() >= 4 && MI->getOperand(0).isReg() && + MI->getOperand(1).isReg() && MI->getOperand(2).isReg() && + MI->getOperand(3).isReg() && "MAdd/MSub must have a least 4 regs"); + + // The third input reg must be zero. + if (MI->getOperand(3).getReg() != ZeroReg) + return false; + + // Must only used by the user we combine with. + if (!MRI.hasOneNonDBGUse(MI->getOperand(0).getReg())) + return false; + + return true; +} + +// TODO: There are many more machine instruction opcodes to match: +// 1. Other data types (integer, vectors) +// 2. Other math / logic operations (xor, or) +// 3. Other forms of the same operation (intrinsics and other variants) +bool AArch64InstrInfo::isAssociativeAndCommutative(const MachineInstr &Inst) const { + switch (Inst.getOpcode()) { + case AArch64::FADDDrr: + case AArch64::FADDSrr: + case AArch64::FADDv2f32: + case AArch64::FADDv2f64: + case AArch64::FADDv4f32: + case AArch64::FMULDrr: + case AArch64::FMULSrr: + case AArch64::FMULX32: + case AArch64::FMULX64: + case AArch64::FMULXv2f32: + case AArch64::FMULXv2f64: + case AArch64::FMULXv4f32: + case AArch64::FMULv2f32: + case AArch64::FMULv2f64: + case AArch64::FMULv4f32: + return Inst.getParent()->getParent()->getTarget().Options.UnsafeFPMath; + default: + return false; + } +} + +/// Find instructions that can be turned into madd. +static bool getMaddPatterns(MachineInstr &Root, + SmallVectorImpl &Patterns) { + unsigned Opc = Root.getOpcode(); + MachineBasicBlock &MBB = *Root.getParent(); + bool Found = false; + + if (!isCombineInstrCandidate(Opc)) + return 0; + if (isCombineInstrSettingFlag(Opc)) { + int Cmp_NZCV = Root.findRegisterDefOperandIdx(AArch64::NZCV, true); + // When NZCV is live bail out. + if (Cmp_NZCV == -1) + return 0; + unsigned NewOpc = convertFlagSettingOpcode(&Root); + // When opcode can't change bail out. + // CHECKME: do we miss any cases for opcode conversion? + if (NewOpc == Opc) + return 0; + Opc = NewOpc; + } + + switch (Opc) { + default: + break; + case AArch64::ADDWrr: + assert(Root.getOperand(1).isReg() && Root.getOperand(2).isReg() && + "ADDWrr does not have register operands"); + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDWrrr, + AArch64::WZR)) { + Patterns.push_back(MachineCombinerPattern::MULADDW_OP1); + Found = true; + } + if (canCombineWithMUL(MBB, Root.getOperand(2), AArch64::MADDWrrr, + AArch64::WZR)) { + Patterns.push_back(MachineCombinerPattern::MULADDW_OP2); + Found = true; + } + break; + case AArch64::ADDXrr: + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDXrrr, + AArch64::XZR)) { + Patterns.push_back(MachineCombinerPattern::MULADDX_OP1); + Found = true; + } + if (canCombineWithMUL(MBB, Root.getOperand(2), AArch64::MADDXrrr, + AArch64::XZR)) { + Patterns.push_back(MachineCombinerPattern::MULADDX_OP2); + Found = true; + } + break; + case AArch64::SUBWrr: + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDWrrr, + AArch64::WZR)) { + Patterns.push_back(MachineCombinerPattern::MULSUBW_OP1); + Found = true; + } + if (canCombineWithMUL(MBB, Root.getOperand(2), AArch64::MADDWrrr, + AArch64::WZR)) { + Patterns.push_back(MachineCombinerPattern::MULSUBW_OP2); + Found = true; + } + break; + case AArch64::SUBXrr: + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDXrrr, + AArch64::XZR)) { + Patterns.push_back(MachineCombinerPattern::MULSUBX_OP1); + Found = true; + } + if (canCombineWithMUL(MBB, Root.getOperand(2), AArch64::MADDXrrr, + AArch64::XZR)) { + Patterns.push_back(MachineCombinerPattern::MULSUBX_OP2); + Found = true; + } + break; + case AArch64::ADDWri: + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDWrrr, + AArch64::WZR)) { + Patterns.push_back(MachineCombinerPattern::MULADDWI_OP1); + Found = true; + } + break; + case AArch64::ADDXri: + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDXrrr, + AArch64::XZR)) { + Patterns.push_back(MachineCombinerPattern::MULADDXI_OP1); + Found = true; + } + break; + case AArch64::SUBWri: + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDWrrr, + AArch64::WZR)) { + Patterns.push_back(MachineCombinerPattern::MULSUBWI_OP1); + Found = true; + } + break; + case AArch64::SUBXri: + if (canCombineWithMUL(MBB, Root.getOperand(1), AArch64::MADDXrrr, + AArch64::XZR)) { + Patterns.push_back(MachineCombinerPattern::MULSUBXI_OP1); + Found = true; + } + break; + } + return Found; +} + +/// Return true when there is potentially a faster code sequence for an +/// instruction chain ending in \p Root. All potential patterns are listed in +/// the \p Pattern vector. Pattern should be sorted in priority order since the +/// pattern evaluator stops checking as soon as it finds a faster sequence. + +bool AArch64InstrInfo::getMachineCombinerPatterns( + MachineInstr &Root, + SmallVectorImpl &Patterns) const { + if (getMaddPatterns(Root, Patterns)) + return true; + + return TargetInstrInfo::getMachineCombinerPatterns(Root, Patterns); +} + +/// genMadd - Generate madd instruction and combine mul and add. +/// Example: +/// MUL I=A,B,0 +/// ADD R,I,C +/// ==> MADD R,A,B,C +/// \param Root is the ADD instruction +/// \param [out] InsInstrs is a vector of machine instructions and will +/// contain the generated madd instruction +/// \param IdxMulOpd is index of operand in Root that is the result of +/// the MUL. In the example above IdxMulOpd is 1. +/// \param MaddOpc the opcode fo the madd instruction +static MachineInstr *genMadd(MachineFunction &MF, MachineRegisterInfo &MRI, + const TargetInstrInfo *TII, MachineInstr &Root, + SmallVectorImpl &InsInstrs, + unsigned IdxMulOpd, unsigned MaddOpc, + const TargetRegisterClass *RC) { + assert(IdxMulOpd == 1 || IdxMulOpd == 2); + + unsigned IdxOtherOpd = IdxMulOpd == 1 ? 2 : 1; + MachineInstr *MUL = MRI.getUniqueVRegDef(Root.getOperand(IdxMulOpd).getReg()); + unsigned ResultReg = Root.getOperand(0).getReg(); + unsigned SrcReg0 = MUL->getOperand(1).getReg(); + bool Src0IsKill = MUL->getOperand(1).isKill(); + unsigned SrcReg1 = MUL->getOperand(2).getReg(); + bool Src1IsKill = MUL->getOperand(2).isKill(); + unsigned SrcReg2 = Root.getOperand(IdxOtherOpd).getReg(); + bool Src2IsKill = Root.getOperand(IdxOtherOpd).isKill(); + + if (TargetRegisterInfo::isVirtualRegister(ResultReg)) + MRI.constrainRegClass(ResultReg, RC); + if (TargetRegisterInfo::isVirtualRegister(SrcReg0)) + MRI.constrainRegClass(SrcReg0, RC); + if (TargetRegisterInfo::isVirtualRegister(SrcReg1)) + MRI.constrainRegClass(SrcReg1, RC); + if (TargetRegisterInfo::isVirtualRegister(SrcReg2)) + MRI.constrainRegClass(SrcReg2, RC); + + MachineInstrBuilder MIB = BuildMI(MF, Root.getDebugLoc(), TII->get(MaddOpc), + ResultReg) + .addReg(SrcReg0, getKillRegState(Src0IsKill)) + .addReg(SrcReg1, getKillRegState(Src1IsKill)) + .addReg(SrcReg2, getKillRegState(Src2IsKill)); + // Insert the MADD + InsInstrs.push_back(MIB); + return MUL; +} + +/// genMaddR - Generate madd instruction and combine mul and add using +/// an extra virtual register +/// Example - an ADD intermediate needs to be stored in a register: +/// MUL I=A,B,0 +/// ADD R,I,Imm +/// ==> ORR V, ZR, Imm +/// ==> MADD R,A,B,V +/// \param Root is the ADD instruction +/// \param [out] InsInstrs is a vector of machine instructions and will +/// contain the generated madd instruction +/// \param IdxMulOpd is index of operand in Root that is the result of +/// the MUL. In the example above IdxMulOpd is 1. +/// \param MaddOpc the opcode fo the madd instruction +/// \param VR is a virtual register that holds the value of an ADD operand +/// (V in the example above). +static MachineInstr *genMaddR(MachineFunction &MF, MachineRegisterInfo &MRI, + const TargetInstrInfo *TII, MachineInstr &Root, + SmallVectorImpl &InsInstrs, + unsigned IdxMulOpd, unsigned MaddOpc, + unsigned VR, const TargetRegisterClass *RC) { + assert(IdxMulOpd == 1 || IdxMulOpd == 2); + + MachineInstr *MUL = MRI.getUniqueVRegDef(Root.getOperand(IdxMulOpd).getReg()); + unsigned ResultReg = Root.getOperand(0).getReg(); + unsigned SrcReg0 = MUL->getOperand(1).getReg(); + bool Src0IsKill = MUL->getOperand(1).isKill(); + unsigned SrcReg1 = MUL->getOperand(2).getReg(); + bool Src1IsKill = MUL->getOperand(2).isKill(); + + if (TargetRegisterInfo::isVirtualRegister(ResultReg)) + MRI.constrainRegClass(ResultReg, RC); + if (TargetRegisterInfo::isVirtualRegister(SrcReg0)) + MRI.constrainRegClass(SrcReg0, RC); + if (TargetRegisterInfo::isVirtualRegister(SrcReg1)) + MRI.constrainRegClass(SrcReg1, RC); + if (TargetRegisterInfo::isVirtualRegister(VR)) + MRI.constrainRegClass(VR, RC); + + MachineInstrBuilder MIB = BuildMI(MF, Root.getDebugLoc(), TII->get(MaddOpc), + ResultReg) + .addReg(SrcReg0, getKillRegState(Src0IsKill)) + .addReg(SrcReg1, getKillRegState(Src1IsKill)) + .addReg(VR); + // Insert the MADD + InsInstrs.push_back(MIB); + return MUL; +} + +/// When getMachineCombinerPatterns() finds potential patterns, +/// this function generates the instructions that could replace the +/// original code sequence +void AArch64InstrInfo::genAlternativeCodeSequence( + MachineInstr &Root, MachineCombinerPattern Pattern, + SmallVectorImpl &InsInstrs, + SmallVectorImpl &DelInstrs, + DenseMap &InstrIdxForVirtReg) const { + MachineBasicBlock &MBB = *Root.getParent(); + MachineRegisterInfo &MRI = MBB.getParent()->getRegInfo(); + MachineFunction &MF = *MBB.getParent(); + const TargetInstrInfo *TII = MF.getSubtarget().getInstrInfo(); + + MachineInstr *MUL; + const TargetRegisterClass *RC; + unsigned Opc; + switch (Pattern) { + default: + // Reassociate instructions. + TargetInstrInfo::genAlternativeCodeSequence(Root, Pattern, InsInstrs, + DelInstrs, InstrIdxForVirtReg); + return; + case MachineCombinerPattern::MULADDW_OP1: + case MachineCombinerPattern::MULADDX_OP1: + // MUL I=A,B,0 + // ADD R,I,C + // ==> MADD R,A,B,C + // --- Create(MADD); + if (Pattern == MachineCombinerPattern::MULADDW_OP1) { + Opc = AArch64::MADDWrrr; + RC = &AArch64::GPR32RegClass; + } else { + Opc = AArch64::MADDXrrr; + RC = &AArch64::GPR64RegClass; + } + MUL = genMadd(MF, MRI, TII, Root, InsInstrs, 1, Opc, RC); + break; + case MachineCombinerPattern::MULADDW_OP2: + case MachineCombinerPattern::MULADDX_OP2: + // MUL I=A,B,0 + // ADD R,C,I + // ==> MADD R,A,B,C + // --- Create(MADD); + if (Pattern == MachineCombinerPattern::MULADDW_OP2) { + Opc = AArch64::MADDWrrr; + RC = &AArch64::GPR32RegClass; + } else { + Opc = AArch64::MADDXrrr; + RC = &AArch64::GPR64RegClass; + } + MUL = genMadd(MF, MRI, TII, Root, InsInstrs, 2, Opc, RC); + break; + case MachineCombinerPattern::MULADDWI_OP1: + case MachineCombinerPattern::MULADDXI_OP1: { + // MUL I=A,B,0 + // ADD R,I,Imm + // ==> ORR V, ZR, Imm + // ==> MADD R,A,B,V + // --- Create(MADD); + const TargetRegisterClass *OrrRC; + unsigned BitSize, OrrOpc, ZeroReg; + if (Pattern == MachineCombinerPattern::MULADDWI_OP1) { + OrrOpc = AArch64::ORRWri; + OrrRC = &AArch64::GPR32spRegClass; + BitSize = 32; + ZeroReg = AArch64::WZR; + Opc = AArch64::MADDWrrr; + RC = &AArch64::GPR32RegClass; + } else { + OrrOpc = AArch64::ORRXri; + OrrRC = &AArch64::GPR64spRegClass; + BitSize = 64; + ZeroReg = AArch64::XZR; + Opc = AArch64::MADDXrrr; + RC = &AArch64::GPR64RegClass; + } + unsigned NewVR = MRI.createVirtualRegister(OrrRC); + uint64_t Imm = Root.getOperand(2).getImm(); + + if (Root.getOperand(3).isImm()) { + unsigned Val = Root.getOperand(3).getImm(); + Imm = Imm << Val; + } + uint64_t UImm = Imm << (64 - BitSize) >> (64 - BitSize); + uint64_t Encoding; + if (AArch64_AM::processLogicalImmediate(UImm, BitSize, Encoding)) { + MachineInstrBuilder MIB1 = + BuildMI(MF, Root.getDebugLoc(), TII->get(OrrOpc), NewVR) + .addReg(ZeroReg) + .addImm(Encoding); + InsInstrs.push_back(MIB1); + InstrIdxForVirtReg.insert(std::make_pair(NewVR, 0)); + MUL = genMaddR(MF, MRI, TII, Root, InsInstrs, 1, Opc, NewVR, RC); + } + break; + } + case MachineCombinerPattern::MULSUBW_OP1: + case MachineCombinerPattern::MULSUBX_OP1: { + // MUL I=A,B,0 + // SUB R,I, C + // ==> SUB V, 0, C + // ==> MADD R,A,B,V // = -C + A*B + // --- Create(MADD); + const TargetRegisterClass *SubRC; + unsigned SubOpc, ZeroReg; + if (Pattern == MachineCombinerPattern::MULSUBW_OP1) { + SubOpc = AArch64::SUBWrr; + SubRC = &AArch64::GPR32spRegClass; + ZeroReg = AArch64::WZR; + Opc = AArch64::MADDWrrr; + RC = &AArch64::GPR32RegClass; + } else { + SubOpc = AArch64::SUBXrr; + SubRC = &AArch64::GPR64spRegClass; + ZeroReg = AArch64::XZR; + Opc = AArch64::MADDXrrr; + RC = &AArch64::GPR64RegClass; + } + unsigned NewVR = MRI.createVirtualRegister(SubRC); + // SUB NewVR, 0, C + MachineInstrBuilder MIB1 = + BuildMI(MF, Root.getDebugLoc(), TII->get(SubOpc), NewVR) + .addReg(ZeroReg) + .addOperand(Root.getOperand(2)); + InsInstrs.push_back(MIB1); + InstrIdxForVirtReg.insert(std::make_pair(NewVR, 0)); + MUL = genMaddR(MF, MRI, TII, Root, InsInstrs, 1, Opc, NewVR, RC); + break; + } + case MachineCombinerPattern::MULSUBW_OP2: + case MachineCombinerPattern::MULSUBX_OP2: + // MUL I=A,B,0 + // SUB R,C,I + // ==> MSUB R,A,B,C (computes C - A*B) + // --- Create(MSUB); + if (Pattern == MachineCombinerPattern::MULSUBW_OP2) { + Opc = AArch64::MSUBWrrr; + RC = &AArch64::GPR32RegClass; + } else { + Opc = AArch64::MSUBXrrr; + RC = &AArch64::GPR64RegClass; + } + MUL = genMadd(MF, MRI, TII, Root, InsInstrs, 2, Opc, RC); + break; + case MachineCombinerPattern::MULSUBWI_OP1: + case MachineCombinerPattern::MULSUBXI_OP1: { + // MUL I=A,B,0 + // SUB R,I, Imm + // ==> ORR V, ZR, -Imm + // ==> MADD R,A,B,V // = -Imm + A*B + // --- Create(MADD); + const TargetRegisterClass *OrrRC; + unsigned BitSize, OrrOpc, ZeroReg; + if (Pattern == MachineCombinerPattern::MULSUBWI_OP1) { + OrrOpc = AArch64::ORRWri; + OrrRC = &AArch64::GPR32spRegClass; + BitSize = 32; + ZeroReg = AArch64::WZR; + Opc = AArch64::MADDWrrr; + RC = &AArch64::GPR32RegClass; + } else { + OrrOpc = AArch64::ORRXri; + OrrRC = &AArch64::GPR64spRegClass; + BitSize = 64; + ZeroReg = AArch64::XZR; + Opc = AArch64::MADDXrrr; + RC = &AArch64::GPR64RegClass; + } + unsigned NewVR = MRI.createVirtualRegister(OrrRC); + int Imm = Root.getOperand(2).getImm(); + if (Root.getOperand(3).isImm()) { + unsigned Val = Root.getOperand(3).getImm(); + Imm = Imm << Val; + } + uint64_t UImm = -Imm << (64 - BitSize) >> (64 - BitSize); + uint64_t Encoding; + if (AArch64_AM::processLogicalImmediate(UImm, BitSize, Encoding)) { + MachineInstrBuilder MIB1 = + BuildMI(MF, Root.getDebugLoc(), TII->get(OrrOpc), NewVR) + .addReg(ZeroReg) + .addImm(Encoding); + InsInstrs.push_back(MIB1); + InstrIdxForVirtReg.insert(std::make_pair(NewVR, 0)); + MUL = genMaddR(MF, MRI, TII, Root, InsInstrs, 1, Opc, NewVR, RC); + } + break; + } + } // end switch (Pattern) + // Record MUL and ADD/SUB for deletion + DelInstrs.push_back(MUL); + DelInstrs.push_back(&Root); + + return; +} + +/// \brief Replace csincr-branch sequence by simple conditional branch +/// +/// Examples: +/// 1. +/// csinc w9, wzr, wzr, +/// tbnz w9, #0, 0x44 +/// to +/// b. +/// +/// 2. +/// csinc w9, wzr, wzr, +/// tbz w9, #0, 0x44 +/// to +/// b. +/// +/// \param MI Conditional Branch +/// \return True when the simple conditional branch is generated +/// +bool AArch64InstrInfo::optimizeCondBranch(MachineInstr *MI) const { + bool IsNegativeBranch = false; + bool IsTestAndBranch = false; + unsigned TargetBBInMI = 0; + switch (MI->getOpcode()) { + default: + llvm_unreachable("Unknown branch instruction?"); + case AArch64::Bcc: + return false; + case AArch64::CBZW: + case AArch64::CBZX: + TargetBBInMI = 1; + break; + case AArch64::CBNZW: + case AArch64::CBNZX: + TargetBBInMI = 1; + IsNegativeBranch = true; + break; + case AArch64::TBZW: + case AArch64::TBZX: + TargetBBInMI = 2; + IsTestAndBranch = true; + break; + case AArch64::TBNZW: + case AArch64::TBNZX: + TargetBBInMI = 2; + IsNegativeBranch = true; + IsTestAndBranch = true; + break; + } + // So we increment a zero register and test for bits other + // than bit 0? Conservatively bail out in case the verifier + // missed this case. + if (IsTestAndBranch && MI->getOperand(1).getImm()) + return false; + + // Find Definition. + assert(MI->getParent() && "Incomplete machine instruciton\n"); + MachineBasicBlock *MBB = MI->getParent(); + MachineFunction *MF = MBB->getParent(); + MachineRegisterInfo *MRI = &MF->getRegInfo(); + unsigned VReg = MI->getOperand(0).getReg(); + if (!TargetRegisterInfo::isVirtualRegister(VReg)) + return false; + + MachineInstr *DefMI = MRI->getVRegDef(VReg); + + // Look for CSINC + if (!(DefMI->getOpcode() == AArch64::CSINCWr && + DefMI->getOperand(1).getReg() == AArch64::WZR && + DefMI->getOperand(2).getReg() == AArch64::WZR) && + !(DefMI->getOpcode() == AArch64::CSINCXr && + DefMI->getOperand(1).getReg() == AArch64::XZR && + DefMI->getOperand(2).getReg() == AArch64::XZR)) + return false; + + if (DefMI->findRegisterDefOperandIdx(AArch64::NZCV, true) != -1) + return false; + + AArch64CC::CondCode CC = + (AArch64CC::CondCode)DefMI->getOperand(3).getImm(); + bool CheckOnlyCCWrites = true; + // Convert only when the condition code is not modified between + // the CSINC and the branch. The CC may be used by other + // instructions in between. + if (modifiesConditionCode(DefMI, MI, CheckOnlyCCWrites, &getRegisterInfo())) + return false; + MachineBasicBlock &RefToMBB = *MBB; + MachineBasicBlock *TBB = MI->getOperand(TargetBBInMI).getMBB(); + DebugLoc DL = MI->getDebugLoc(); + if (IsNegativeBranch) + CC = AArch64CC::getInvertedCondCode(CC); + BuildMI(RefToMBB, MI, DL, get(AArch64::Bcc)).addImm(CC).addMBB(TBB); + MI->eraseFromParent(); + return true; +} + +std::pair +AArch64InstrInfo::decomposeMachineOperandsTargetFlags(unsigned TF) const { + const unsigned Mask = AArch64II::MO_FRAGMENT; + return std::make_pair(TF & Mask, TF & ~Mask); +} + +ArrayRef> +AArch64InstrInfo::getSerializableDirectMachineOperandTargetFlags() const { + using namespace AArch64II; + static const std::pair TargetFlags[] = { + {MO_PAGE, "aarch64-page"}, + {MO_PAGEOFF, "aarch64-pageoff"}, + {MO_G3, "aarch64-g3"}, + {MO_G2, "aarch64-g2"}, + {MO_G1, "aarch64-g1"}, + {MO_G0, "aarch64-g0"}, + {MO_HI12, "aarch64-hi12"}}; + return makeArrayRef(TargetFlags); +} + +ArrayRef> +AArch64InstrInfo::getSerializableBitmaskMachineOperandTargetFlags() const { + using namespace AArch64II; + static const std::pair TargetFlags[] = { + {MO_GOT, "aarch64-got"}, + {MO_NC, "aarch64-nc"}, + {MO_TLS, "aarch64-tls"}, + {MO_CONSTPOOL, "aarch64-constant-pool"}}; + return makeArrayRef(TargetFlags); +}