#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/StackMaps.h"
#include "llvm/IR/DerivedTypes.h"
+#include "llvm/IR/Function.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/MC/MCAsmInfo.h"
+#include "llvm/MC/MCExpr.h"
#include "llvm/MC/MCInst.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
// Pin the vtable to this file.
void X86InstrInfo::anchor() {}
-X86InstrInfo::X86InstrInfo(X86TargetMachine &tm)
- : X86GenInstrInfo((tm.getSubtarget<X86Subtarget>().is64Bit()
- ? X86::ADJCALLSTACKDOWN64
- : X86::ADJCALLSTACKDOWN32),
- (tm.getSubtarget<X86Subtarget>().is64Bit()
- ? X86::ADJCALLSTACKUP64
- : X86::ADJCALLSTACKUP32)),
- TM(tm), RI(tm) {
+X86InstrInfo::X86InstrInfo(X86Subtarget &STI)
+ : X86GenInstrInfo(
+ (STI.is64Bit() ? X86::ADJCALLSTACKDOWN64 : X86::ADJCALLSTACKDOWN32),
+ (STI.is64Bit() ? X86::ADJCALLSTACKUP64 : X86::ADJCALLSTACKUP32)),
+ Subtarget(STI), RI(STI) {
static const X86OpTblEntry OpTbl2Addr[] = {
{ X86::ADC32ri, X86::ADC32mi, 0 },
{ X86::VMOVUPDYrr, X86::VMOVUPDYmr, TB_FOLDED_STORE },
{ X86::VMOVUPSYrr, X86::VMOVUPSYmr, TB_FOLDED_STORE },
// AVX-512 foldable instructions
- { X86::VMOVPDI2DIZrr,X86::VMOVPDI2DIZmr, TB_FOLDED_STORE }
+ { X86::VMOVPDI2DIZrr, X86::VMOVPDI2DIZmr, TB_FOLDED_STORE },
+ { X86::VMOVAPDZrr, X86::VMOVAPDZmr, TB_FOLDED_STORE | TB_ALIGN_64 },
+ { X86::VMOVAPSZrr, X86::VMOVAPSZmr, TB_FOLDED_STORE | TB_ALIGN_64 },
+ { X86::VMOVDQA32Zrr, X86::VMOVDQA32Zmr, TB_FOLDED_STORE | TB_ALIGN_64 },
+ { X86::VMOVDQA64Zrr, X86::VMOVDQA64Zmr, TB_FOLDED_STORE | TB_ALIGN_64 },
+ { X86::VMOVUPDZrr, X86::VMOVUPDZmr, TB_FOLDED_STORE },
+ { X86::VMOVUPSZrr, X86::VMOVUPSZmr, TB_FOLDED_STORE },
+ { X86::VMOVDQU32Zrr, X86::VMOVDQU32Zmr, TB_FOLDED_STORE },
+ { X86::VMOVDQU64Zrr, X86::VMOVDQU64Zmr, TB_FOLDED_STORE }
};
for (unsigned i = 0, e = array_lengthof(OpTbl0); i != e; ++i) {
// AVX-512 foldable instructions
{ X86::VMOV64toPQIZrr, X86::VMOVQI2PQIZrm, 0 },
{ X86::VMOVDI2SSZrr, X86::VMOVDI2SSZrm, 0 },
- { X86::VMOVDQA32rr, X86::VMOVDQA32rm, TB_ALIGN_64 },
- { X86::VMOVDQA64rr, X86::VMOVDQA64rm, TB_ALIGN_64 },
- { X86::VMOVDQU32rr, X86::VMOVDQU32rm, 0 },
- { X86::VMOVDQU64rr, X86::VMOVDQU64rm, 0 },
+ { X86::VMOVAPDZrr, X86::VMOVAPDZrm, TB_ALIGN_64 },
+ { X86::VMOVAPSZrr, X86::VMOVAPSZrm, TB_ALIGN_64 },
+ { X86::VMOVDQA32Zrr, X86::VMOVDQA32Zrm, TB_ALIGN_64 },
+ { X86::VMOVDQA64Zrr, X86::VMOVDQA64Zrm, TB_ALIGN_64 },
+ { X86::VMOVDQU32Zrr, X86::VMOVDQU32Zrm, 0 },
+ { X86::VMOVDQU64Zrr, X86::VMOVDQU64Zrm, 0 },
+ { X86::VMOVUPDZrr, X86::VMOVUPDZrm, 0 },
+ { X86::VMOVUPSZrr, X86::VMOVUPSZrm, 0 },
{ X86::VPABSDZrr, X86::VPABSDZrm, 0 },
{ X86::VPABSQZrr, X86::VPABSQZrm, 0 },
{ X86::AESIMCrr, X86::AESIMCrm, TB_ALIGN_16 },
{ X86::AESKEYGENASSIST128rr, X86::AESKEYGENASSIST128rm, TB_ALIGN_16 },
{ X86::VAESIMCrr, X86::VAESIMCrm, TB_ALIGN_16 },
- { X86::VAESKEYGENASSIST128rr, X86::VAESKEYGENASSIST128rm, TB_ALIGN_16 },
+ { X86::VAESKEYGENASSIST128rr, X86::VAESKEYGENASSIST128rm, TB_ALIGN_16 }
};
for (unsigned i = 0, e = array_lengthof(OpTbl1); i != e; ++i) {
case X86::MOVSX32rr8:
case X86::MOVZX32rr8:
case X86::MOVSX64rr8:
- if (!TM.getSubtarget<X86Subtarget>().is64Bit())
+ if (!Subtarget.is64Bit())
// It's not always legal to reference the low 8-bit of the larger
// register in 32-bit mode.
return false;
/// operand and follow operands form a reference to the stack frame.
bool X86InstrInfo::isFrameOperand(const MachineInstr *MI, unsigned int Op,
int &FrameIndex) const {
- if (MI->getOperand(Op).isFI() && MI->getOperand(Op+1).isImm() &&
- MI->getOperand(Op+2).isReg() && MI->getOperand(Op+3).isImm() &&
- MI->getOperand(Op+1).getImm() == 1 &&
- MI->getOperand(Op+2).getReg() == 0 &&
- MI->getOperand(Op+3).getImm() == 0) {
- FrameIndex = MI->getOperand(Op).getIndex();
+ if (MI->getOperand(Op+X86::AddrBaseReg).isFI() &&
+ MI->getOperand(Op+X86::AddrScaleAmt).isImm() &&
+ MI->getOperand(Op+X86::AddrIndexReg).isReg() &&
+ MI->getOperand(Op+X86::AddrDisp).isImm() &&
+ MI->getOperand(Op+X86::AddrScaleAmt).getImm() == 1 &&
+ MI->getOperand(Op+X86::AddrIndexReg).getReg() == 0 &&
+ MI->getOperand(Op+X86::AddrDisp).getImm() == 0) {
+ FrameIndex = MI->getOperand(Op+X86::AddrBaseReg).getIndex();
return true;
}
return false;
case X86::FsMOVAPSrm:
case X86::FsMOVAPDrm: {
// Loads from constant pools are trivially rematerializable.
- if (MI->getOperand(1).isReg() &&
- MI->getOperand(2).isImm() &&
- MI->getOperand(3).isReg() && MI->getOperand(3).getReg() == 0 &&
+ if (MI->getOperand(1+X86::AddrBaseReg).isReg() &&
+ MI->getOperand(1+X86::AddrScaleAmt).isImm() &&
+ MI->getOperand(1+X86::AddrIndexReg).isReg() &&
+ MI->getOperand(1+X86::AddrIndexReg).getReg() == 0 &&
MI->isInvariantLoad(AA)) {
- unsigned BaseReg = MI->getOperand(1).getReg();
+ unsigned BaseReg = MI->getOperand(1+X86::AddrBaseReg).getReg();
if (BaseReg == 0 || BaseReg == X86::RIP)
return true;
// Allow re-materialization of PIC load.
- if (!ReMatPICStubLoad && MI->getOperand(4).isGlobal())
+ if (!ReMatPICStubLoad && MI->getOperand(1+X86::AddrDisp).isGlobal())
return false;
const MachineFunction &MF = *MI->getParent()->getParent();
const MachineRegisterInfo &MRI = MF.getRegInfo();
case X86::LEA32r:
case X86::LEA64r: {
- if (MI->getOperand(2).isImm() &&
- MI->getOperand(3).isReg() && MI->getOperand(3).getReg() == 0 &&
- !MI->getOperand(4).isReg()) {
+ if (MI->getOperand(1+X86::AddrScaleAmt).isImm() &&
+ MI->getOperand(1+X86::AddrIndexReg).isReg() &&
+ MI->getOperand(1+X86::AddrIndexReg).getReg() == 0 &&
+ !MI->getOperand(1+X86::AddrDisp).isReg()) {
// lea fi#, lea GV, etc. are all rematerializable.
- if (!MI->getOperand(1).isReg())
+ if (!MI->getOperand(1+X86::AddrBaseReg).isReg())
return true;
- unsigned BaseReg = MI->getOperand(1).getReg();
+ unsigned BaseReg = MI->getOperand(1+X86::AddrBaseReg).getReg();
if (BaseReg == 0)
return true;
// Allow re-materialization of lea PICBase + x.
return true;
}
-/// isSafeToClobberEFLAGS - Return true if it's safe insert an instruction that
-/// would clobber the EFLAGS condition register. Note the result may be
-/// conservative. If it cannot definitely determine the safety after visiting
-/// a few instructions in each direction it assumes it's not safe.
-static bool isSafeToClobberEFLAGS(MachineBasicBlock &MBB,
- MachineBasicBlock::iterator I) {
+bool X86InstrInfo::isSafeToClobberEFLAGS(MachineBasicBlock &MBB,
+ MachineBasicBlock::iterator I) const {
MachineBasicBlock::iterator E = MBB.end();
// For compile time consideration, if we are not able to determine the
MachineRegisterInfo &RegInfo = MFI->getParent()->getRegInfo();
unsigned leaOutReg = RegInfo.createVirtualRegister(&X86::GR32RegClass);
unsigned Opc, leaInReg;
- if (TM.getSubtarget<X86Subtarget>().is64Bit()) {
+ if (Subtarget.is64Bit()) {
Opc = X86::LEA64_32r;
leaInReg = RegInfo.createVirtualRegister(&X86::GR64_NOSPRegClass);
} else {
// just a single insert_subreg.
addRegReg(MIB, leaInReg, true, leaInReg, false);
} else {
- if (TM.getSubtarget<X86Subtarget>().is64Bit())
+ if (Subtarget.is64Bit())
leaInReg2 = RegInfo.createVirtualRegister(&X86::GR64_NOSPRegClass);
else
leaInReg2 = RegInfo.createVirtualRegister(&X86::GR32_NOSPRegClass);
// we have better subtarget support, enable the 16-bit LEA generation here.
// 16-bit LEA is also slow on Core2.
bool DisableLEA16 = true;
- bool is64Bit = TM.getSubtarget<X86Subtarget>().is64Bit();
+ bool is64Bit = Subtarget.is64Bit();
unsigned MIOpc = MI->getOpcode();
switch (MIOpc) {
case X86::SHUFPSrri: {
assert(MI->getNumOperands() == 4 && "Unknown shufps instruction!");
- if (!TM.getSubtarget<X86Subtarget>().hasSSE2()) return nullptr;
+ if (!Subtarget.hasSSE2()) return nullptr;
unsigned B = MI->getOperand(1).getReg();
unsigned C = MI->getOperand(2).getReg();
}
case X86::SHUFPDrri: {
assert(MI->getNumOperands() == 4 && "Unknown shufpd instruction!");
- if (!TM.getSubtarget<X86Subtarget>().hasSSE2()) return nullptr;
+ if (!Subtarget.hasSSE2()) return nullptr;
unsigned B = MI->getOperand(1).getReg();
unsigned C = MI->getOperand(2).getReg();
/// getSETFromCond - Return a set opcode for the given condition and
/// whether it has memory operand.
-static unsigned getSETFromCond(X86::CondCode CC,
- bool HasMemoryOperand) {
+unsigned X86::getSETFromCond(CondCode CC, bool HasMemoryOperand) {
static const uint16_t Opc[16][2] = {
{ X86::SETAr, X86::SETAm },
{ X86::SETAEr, X86::SETAEm },
{ X86::SETSr, X86::SETSm }
};
- assert(CC < 16 && "Can only handle standard cond codes");
+ assert(CC <= LAST_VALID_COND && "Can only handle standard cond codes");
return Opc[CC][HasMemoryOperand ? 1 : 0];
}
/// getCMovFromCond - Return a cmov opcode for the given condition,
/// register size in bytes, and operand type.
-static unsigned getCMovFromCond(X86::CondCode CC, unsigned RegBytes,
- bool HasMemoryOperand) {
+unsigned X86::getCMovFromCond(CondCode CC, unsigned RegBytes,
+ bool HasMemoryOperand) {
static const uint16_t Opc[32][3] = {
{ X86::CMOVA16rr, X86::CMOVA32rr, X86::CMOVA64rr },
{ X86::CMOVAE16rr, X86::CMOVAE32rr, X86::CMOVAE64rr },
unsigned TrueReg, unsigned FalseReg,
int &CondCycles, int &TrueCycles, int &FalseCycles) const {
// Not all subtargets have cmov instructions.
- if (!TM.getSubtarget<X86Subtarget>().hasCMov())
+ if (!Subtarget.hasCMov())
return false;
if (Cond.size() != 1)
return false;
// Try and copy between VR128/VR64 and GR64 registers.
static unsigned CopyToFromAsymmetricReg(unsigned DestReg, unsigned SrcReg,
- const X86Subtarget& Subtarget) {
-
+ const X86Subtarget &Subtarget) {
// SrcReg(VR128) -> DestReg(GR64)
// SrcReg(VR64) -> DestReg(GR64)
inline static bool MaskRegClassContains(unsigned Reg) {
return X86::VK8RegClass.contains(Reg) ||
X86::VK16RegClass.contains(Reg) ||
+ X86::VK32RegClass.contains(Reg) ||
+ X86::VK64RegClass.contains(Reg) ||
X86::VK1RegClass.contains(Reg);
}
static
unsigned DestReg, unsigned SrcReg,
bool KillSrc) const {
// First deal with the normal symmetric copies.
- bool HasAVX = TM.getSubtarget<X86Subtarget>().hasAVX();
- bool HasAVX512 = TM.getSubtarget<X86Subtarget>().hasAVX512();
+ bool HasAVX = Subtarget.hasAVX();
+ bool HasAVX512 = Subtarget.hasAVX512();
unsigned Opc = 0;
if (X86::GR64RegClass.contains(DestReg, SrcReg))
Opc = X86::MOV64rr;
// Copying to or from a physical H register on x86-64 requires a NOREX
// move. Otherwise use a normal move.
if ((isHReg(DestReg) || isHReg(SrcReg)) &&
- TM.getSubtarget<X86Subtarget>().is64Bit()) {
+ Subtarget.is64Bit()) {
Opc = X86::MOV8rr_NOREX;
// Both operands must be encodable without an REX prefix.
assert(X86::GR8_NOREXRegClass.contains(SrcReg, DestReg) &&
else if (X86::VR256RegClass.contains(DestReg, SrcReg))
Opc = X86::VMOVAPSYrr;
if (!Opc)
- Opc = CopyToFromAsymmetricReg(DestReg, SrcReg, TM.getSubtarget<X86Subtarget>());
+ Opc = CopyToFromAsymmetricReg(DestReg, SrcReg, Subtarget);
if (Opc) {
BuildMI(MBB, MI, DL, get(Opc), DestReg)
// Moving EFLAGS to / from another register requires a push and a pop.
// Notice that we have to adjust the stack if we don't want to clobber the
- // first frame index. See X86FrameLowering.cpp - colobbersTheStack.
+ // first frame index. See X86FrameLowering.cpp - clobbersTheStack.
if (SrcReg == X86::EFLAGS) {
if (X86::GR64RegClass.contains(DestReg)) {
BuildMI(MBB, MI, DL, get(X86::PUSHF64));
static unsigned getLoadStoreRegOpcode(unsigned Reg,
const TargetRegisterClass *RC,
bool isStackAligned,
- const TargetMachine &TM,
+ const X86Subtarget &STI,
bool load) {
- if (TM.getSubtarget<X86Subtarget>().hasAVX512()) {
+ if (STI.hasAVX512()) {
if (X86::VK8RegClass.hasSubClassEq(RC) ||
X86::VK16RegClass.hasSubClassEq(RC))
return load ? X86::KMOVWkm : X86::KMOVWmk;
return load ? X86::VMOVUPSZrm : X86::VMOVUPSZmr;
}
- bool HasAVX = TM.getSubtarget<X86Subtarget>().hasAVX();
+ bool HasAVX = STI.hasAVX();
switch (RC->getSize()) {
default:
llvm_unreachable("Unknown spill size");
case 1:
assert(X86::GR8RegClass.hasSubClassEq(RC) && "Unknown 1-byte regclass");
- if (TM.getSubtarget<X86Subtarget>().is64Bit())
+ if (STI.is64Bit())
// Copying to or from a physical H register on x86-64 requires a NOREX
// move. Otherwise use a normal move.
if (isHReg(Reg) || X86::GR8_ABCD_HRegClass.hasSubClassEq(RC))
static unsigned getStoreRegOpcode(unsigned SrcReg,
const TargetRegisterClass *RC,
bool isStackAligned,
- TargetMachine &TM) {
- return getLoadStoreRegOpcode(SrcReg, RC, isStackAligned, TM, false);
+ const X86Subtarget &STI) {
+ return getLoadStoreRegOpcode(SrcReg, RC, isStackAligned, STI, false);
}
static unsigned getLoadRegOpcode(unsigned DestReg,
const TargetRegisterClass *RC,
bool isStackAligned,
- const TargetMachine &TM) {
- return getLoadStoreRegOpcode(DestReg, RC, isStackAligned, TM, true);
+ const X86Subtarget &STI) {
+ return getLoadStoreRegOpcode(DestReg, RC, isStackAligned, STI, true);
}
void X86InstrInfo::storeRegToStackSlot(MachineBasicBlock &MBB,
assert(MF.getFrameInfo()->getObjectSize(FrameIdx) >= RC->getSize() &&
"Stack slot too small for store");
unsigned Alignment = std::max<uint32_t>(RC->getSize(), 16);
- bool isAligned = (TM.getFrameLowering()->getStackAlignment() >= Alignment) ||
- RI.canRealignStack(MF);
- unsigned Opc = getStoreRegOpcode(SrcReg, RC, isAligned, TM);
+ bool isAligned = (MF.getTarget()
+ .getSubtargetImpl()
+ ->getFrameLowering()
+ ->getStackAlignment() >= Alignment) ||
+ RI.canRealignStack(MF);
+ unsigned Opc = getStoreRegOpcode(SrcReg, RC, isAligned, Subtarget);
DebugLoc DL = MBB.findDebugLoc(MI);
addFrameReference(BuildMI(MBB, MI, DL, get(Opc)), FrameIdx)
.addReg(SrcReg, getKillRegState(isKill));
unsigned Alignment = std::max<uint32_t>(RC->getSize(), 16);
bool isAligned = MMOBegin != MMOEnd &&
(*MMOBegin)->getAlignment() >= Alignment;
- unsigned Opc = getStoreRegOpcode(SrcReg, RC, isAligned, TM);
+ unsigned Opc = getStoreRegOpcode(SrcReg, RC, isAligned, Subtarget);
DebugLoc DL;
MachineInstrBuilder MIB = BuildMI(MF, DL, get(Opc));
for (unsigned i = 0, e = Addr.size(); i != e; ++i)
const TargetRegisterInfo *TRI) const {
const MachineFunction &MF = *MBB.getParent();
unsigned Alignment = std::max<uint32_t>(RC->getSize(), 16);
- bool isAligned = (TM.getFrameLowering()->getStackAlignment() >= Alignment) ||
- RI.canRealignStack(MF);
- unsigned Opc = getLoadRegOpcode(DestReg, RC, isAligned, TM);
+ bool isAligned = (MF.getTarget()
+ .getSubtargetImpl()
+ ->getFrameLowering()
+ ->getStackAlignment() >= Alignment) ||
+ RI.canRealignStack(MF);
+ unsigned Opc = getLoadRegOpcode(DestReg, RC, isAligned, Subtarget);
DebugLoc DL = MBB.findDebugLoc(MI);
addFrameReference(BuildMI(MBB, MI, DL, get(Opc), DestReg), FrameIdx);
}
unsigned Alignment = std::max<uint32_t>(RC->getSize(), 16);
bool isAligned = MMOBegin != MMOEnd &&
(*MMOBegin)->getAlignment() >= Alignment;
- unsigned Opc = getLoadRegOpcode(DestReg, RC, isAligned, TM);
+ unsigned Opc = getLoadRegOpcode(DestReg, RC, isAligned, Subtarget);
DebugLoc DL;
MachineInstrBuilder MIB = BuildMI(MF, DL, get(Opc), DestReg);
for (unsigned i = 0, e = Addr.size(); i != e; ++i)
}
}
+/// isUseDefConvertible - check whether the use can be converted
+/// to remove a comparison against zero.
+static X86::CondCode isUseDefConvertible(MachineInstr *MI) {
+ switch (MI->getOpcode()) {
+ default: return X86::COND_INVALID;
+ case X86::LZCNT16rr: case X86::LZCNT16rm:
+ case X86::LZCNT32rr: case X86::LZCNT32rm:
+ case X86::LZCNT64rr: case X86::LZCNT64rm:
+ return X86::COND_B;
+ case X86::POPCNT16rr:case X86::POPCNT16rm:
+ case X86::POPCNT32rr:case X86::POPCNT32rm:
+ case X86::POPCNT64rr:case X86::POPCNT64rm:
+ return X86::COND_E;
+ case X86::TZCNT16rr: case X86::TZCNT16rm:
+ case X86::TZCNT32rr: case X86::TZCNT32rm:
+ case X86::TZCNT64rr: case X86::TZCNT64rm:
+ return X86::COND_B;
+ }
+}
+
/// optimizeCompareInstr - Check if there exists an earlier instruction that
/// operates on the same source operands and sets flags in the same way as
/// Compare; remove Compare if possible.
// If we are comparing against zero, check whether we can use MI to update
// EFLAGS. If MI is not in the same BB as CmpInstr, do not optimize.
bool IsCmpZero = (SrcReg2 == 0 && CmpValue == 0);
- if (IsCmpZero && (MI->getParent() != CmpInstr->getParent() ||
- !isDefConvertible(MI)))
+ if (IsCmpZero && MI->getParent() != CmpInstr->getParent())
return false;
+ // If we have a use of the source register between the def and our compare
+ // instruction we can eliminate the compare iff the use sets EFLAGS in the
+ // right way.
+ bool ShouldUpdateCC = false;
+ X86::CondCode NewCC = X86::COND_INVALID;
+ if (IsCmpZero && !isDefConvertible(MI)) {
+ // Scan forward from the use until we hit the use we're looking for or the
+ // compare instruction.
+ for (MachineBasicBlock::iterator J = MI;; ++J) {
+ // Do we have a convertible instruction?
+ NewCC = isUseDefConvertible(J);
+ if (NewCC != X86::COND_INVALID && J->getOperand(1).isReg() &&
+ J->getOperand(1).getReg() == SrcReg) {
+ assert(J->definesRegister(X86::EFLAGS) && "Must be an EFLAGS def!");
+ ShouldUpdateCC = true; // Update CC later on.
+ // This is not a def of SrcReg, but still a def of EFLAGS. Keep going
+ // with the new def.
+ MI = Def = J;
+ break;
+ }
+
+ if (J == I)
+ return false;
+ }
+ }
+
// We are searching for an earlier instruction that can make CmpInstr
// redundant and that instruction will be saved in Sub.
MachineInstr *Sub = nullptr;
continue;
// EFLAGS is used by this instruction.
- X86::CondCode OldCC;
+ X86::CondCode OldCC = X86::COND_INVALID;
bool OpcIsSET = false;
if (IsCmpZero || IsSwapped) {
// We decode the condition code from opcode.
// CF and OF are used, we can't perform this optimization.
return false;
}
+
+ // If we're updating the condition code check if we have to reverse the
+ // condition.
+ if (ShouldUpdateCC)
+ switch (OldCC) {
+ default:
+ return false;
+ case X86::COND_E:
+ break;
+ case X86::COND_NE:
+ NewCC = GetOppositeBranchCondition(NewCC);
+ break;
+ }
} else if (IsSwapped) {
// If we have SUB(r1, r2) and CMP(r2, r1), the condition code needs
// to be changed from r2 > r1 to r1 < r2, from r2 < r1 to r1 > r2, etc.
// We swap the condition code and synthesize the new opcode.
- X86::CondCode NewCC = getSwappedCondition(OldCC);
+ NewCC = getSwappedCondition(OldCC);
if (NewCC == X86::COND_INVALID) return false;
+ }
+ if ((ShouldUpdateCC || IsSwapped) && NewCC != OldCC) {
// Synthesize the new opcode.
bool HasMemoryOperand = Instr.hasOneMemOperand();
unsigned NewOpc;
return true;
}
+// LoadStackGuard has so far only been implemented for 64-bit MachO. Different
+// code sequence is needed for other targets.
+static void expandLoadStackGuard(MachineInstrBuilder &MIB,
+ const TargetInstrInfo &TII) {
+ MachineBasicBlock &MBB = *MIB->getParent();
+ DebugLoc DL = MIB->getDebugLoc();
+ unsigned Reg = MIB->getOperand(0).getReg();
+ const GlobalValue *GV =
+ cast<GlobalValue>((*MIB->memoperands_begin())->getValue());
+ unsigned Flag = MachineMemOperand::MOLoad | MachineMemOperand::MOInvariant;
+ MachineMemOperand *MMO = MBB.getParent()->
+ getMachineMemOperand(MachinePointerInfo::getGOT(), Flag, 8, 8);
+ MachineBasicBlock::iterator I = MIB;
+
+ BuildMI(MBB, I, DL, TII.get(X86::MOV64rm), Reg).addReg(X86::RIP).addImm(1)
+ .addReg(0).addGlobalAddress(GV, 0, X86II::MO_GOTPCREL).addReg(0)
+ .addMemOperand(MMO);
+ MIB->setDebugLoc(DL);
+ MIB->setDesc(TII.get(X86::MOV64rm));
+ MIB.addReg(Reg, RegState::Kill).addImm(1).addReg(0).addImm(0).addReg(0);
+}
+
bool X86InstrInfo::expandPostRAPseudo(MachineBasicBlock::iterator MI) const {
- bool HasAVX = TM.getSubtarget<X86Subtarget>().hasAVX();
+ bool HasAVX = Subtarget.hasAVX();
MachineInstrBuilder MIB(*MI->getParent()->getParent(), MI);
switch (MI->getOpcode()) {
case X86::MOV32r0:
case X86::KSET0W: return Expand2AddrUndef(MIB, get(X86::KXORWrr));
case X86::KSET1B:
case X86::KSET1W: return Expand2AddrUndef(MIB, get(X86::KXNORWrr));
+ case TargetOpcode::LOAD_STACK_GUARD:
+ expandLoadStackGuard(MIB, *this);
+ return true;
}
return false;
}
unsigned Size, unsigned Align) const {
const DenseMap<unsigned,
std::pair<unsigned,unsigned> > *OpcodeTablePtr = nullptr;
- bool isCallRegIndirect = TM.getSubtarget<X86Subtarget>().callRegIndirect();
+ bool isCallRegIndirect = Subtarget.callRegIndirect();
bool isTwoAddrFold = false;
// Atom favors register form of call. So, we do not fold loads into calls
if (X86::VR128RegClass.contains(Reg)) {
// These instructions are all floating point domain, so xorps is the best
// choice.
- bool HasAVX = TM.getSubtarget<X86Subtarget>().hasAVX();
+ bool HasAVX = Subtarget.hasAVX();
unsigned Opc = HasAVX ? X86::VXORPSrr : X86::XORPSrr;
BuildMI(*MI->getParent(), MI, MI->getDebugLoc(), get(Opc), Reg)
.addReg(Reg, RegState::Undef).addReg(Reg, RegState::Undef);
// If the function stack isn't realigned we don't want to fold instructions
// that need increased alignment.
if (!RI.needsStackRealignment(MF))
- Alignment = std::min(Alignment, TM.getFrameLowering()->getStackAlignment());
+ Alignment = std::min(Alignment, MF.getTarget()
+ .getSubtargetImpl()
+ ->getFrameLowering()
+ ->getStackAlignment());
if (Ops.size() == 2 && Ops[0] == 0 && Ops[1] == 1) {
unsigned NewOpc = 0;
unsigned RCSize = 0;
// Create a constant-pool entry and operands to load from it.
// Medium and large mode can't fold loads this way.
- if (TM.getCodeModel() != CodeModel::Small &&
- TM.getCodeModel() != CodeModel::Kernel)
+ if (MF.getTarget().getCodeModel() != CodeModel::Small &&
+ MF.getTarget().getCodeModel() != CodeModel::Kernel)
return nullptr;
// x86-32 PIC requires a PIC base register for constant pools.
unsigned PICBase = 0;
- if (TM.getRelocationModel() == Reloc::PIC_) {
- if (TM.getSubtarget<X86Subtarget>().is64Bit())
+ if (MF.getTarget().getRelocationModel() == Reloc::PIC_) {
+ if (Subtarget.is64Bit())
PICBase = X86::RIP;
else
// FIXME: PICBase = getGlobalBaseReg(&MF);
const TargetRegisterClass *RC = getRegClass(MCID, Index, &RI, MF);
if (!MI->hasOneMemOperand() &&
RC == &X86::VR128RegClass &&
- !TM.getSubtarget<X86Subtarget>().isUnalignedMemAccessFast())
+ !Subtarget.isUnalignedMemAccessFast())
// Without memoperands, loadRegFromAddr and storeRegToStackSlot will
// conservatively assume the address is unaligned. That's bad for
// performance.
cast<MachineSDNode>(N)->memoperands_end());
if (!(*MMOs.first) &&
RC == &X86::VR128RegClass &&
- !TM.getSubtarget<X86Subtarget>().isUnalignedMemAccessFast())
+ !Subtarget.isUnalignedMemAccessFast())
// Do not introduce a slow unaligned load.
return false;
unsigned Alignment = RC->getSize() == 32 ? 32 : 16;
bool isAligned = (*MMOs.first) &&
(*MMOs.first)->getAlignment() >= Alignment;
- Load = DAG.getMachineNode(getLoadRegOpcode(0, RC, isAligned, TM), dl,
+ Load = DAG.getMachineNode(getLoadRegOpcode(0, RC, isAligned, Subtarget), dl,
VT, MVT::Other, AddrOps);
NewNodes.push_back(Load);
cast<MachineSDNode>(N)->memoperands_end());
if (!(*MMOs.first) &&
RC == &X86::VR128RegClass &&
- !TM.getSubtarget<X86Subtarget>().isUnalignedMemAccessFast())
+ !Subtarget.isUnalignedMemAccessFast())
// Do not introduce a slow unaligned store.
return false;
unsigned Alignment = RC->getSize() == 32 ? 32 : 16;
bool isAligned = (*MMOs.first) &&
(*MMOs.first)->getAlignment() >= Alignment;
- SDNode *Store = DAG.getMachineNode(getStoreRegOpcode(0, DstRC,
- isAligned, TM),
- dl, MVT::Other, AddrOps);
+ SDNode *Store =
+ DAG.getMachineNode(getStoreRegOpcode(0, DstRC, isAligned, Subtarget),
+ dl, MVT::Other, AddrOps);
NewNodes.push_back(Store);
// Preserve memory reference information.
default:
// XMM registers. In 64-bit mode we can be a bit more aggressive since we
// have 16 of them to play with.
- if (TM.getSubtargetImpl()->is64Bit()) {
+ if (Subtarget.is64Bit()) {
if (NumLoads >= 3)
return false;
} else if (NumLoads) {
// Check if this processor supports macro-fusion. Since this is a minor
// heuristic, we haven't specifically reserved a feature. hasAVX is a decent
// proxy for SandyBridge+.
- if (!TM.getSubtarget<X86Subtarget>().hasAVX())
+ if (!Subtarget.hasAVX())
return false;
enum {
case X86::TEST16rm:
case X86::TEST32rm:
case X86::TEST64rm:
+ case X86::TEST8ri_NOREX:
case X86::AND16i16:
case X86::AND16ri:
case X86::AND16ri8:
/// TODO: Eliminate this and move the code to X86MachineFunctionInfo.
///
unsigned X86InstrInfo::getGlobalBaseReg(MachineFunction *MF) const {
- assert(!TM.getSubtarget<X86Subtarget>().is64Bit() &&
+ assert(!Subtarget.is64Bit() &&
"X86-64 PIC uses RIP relative addressing");
X86MachineFunctionInfo *X86FI = MF->getInfo<X86MachineFunctionInfo>();
std::pair<uint16_t, uint16_t>
X86InstrInfo::getExecutionDomain(const MachineInstr *MI) const {
uint16_t domain = (MI->getDesc().TSFlags >> X86II::SSEDomainShift) & 3;
- bool hasAVX2 = TM.getSubtarget<X86Subtarget>().hasAVX2();
+ bool hasAVX2 = Subtarget.hasAVX2();
uint16_t validDomains = 0;
if (domain && lookup(MI->getOpcode(), domain))
validDomains = 0xe;
assert(dom && "Not an SSE instruction");
const uint16_t *table = lookup(MI->getOpcode(), dom);
if (!table) { // try the other table
- assert((TM.getSubtarget<X86Subtarget>().hasAVX2() || Domain < 3) &&
+ assert((Subtarget.hasAVX2() || Domain < 3) &&
"256-bit vector operations only available in AVX2");
table = lookupAVX2(MI->getOpcode(), dom);
}
NopInst.setOpcode(X86::NOOP);
}
+void X86InstrInfo::getUnconditionalBranch(
+ MCInst &Branch, const MCSymbolRefExpr *BranchTarget) const {
+ Branch.setOpcode(X86::JMP_4);
+ Branch.addOperand(MCOperand::CreateExpr(BranchTarget));
+}
+
+void X86InstrInfo::getTrap(MCInst &MI) const {
+ MI.setOpcode(X86::TRAP);
+}
+
bool X86InstrInfo::isHighLatencyDef(int opc) const {
switch (opc) {
default: return false;
const X86TargetMachine *TM =
static_cast<const X86TargetMachine *>(&MF.getTarget());
- assert(!TM->getSubtarget<X86Subtarget>().is64Bit() &&
- "X86-64 PIC uses RIP relative addressing");
+ // Don't do anything if this is 64-bit as 64-bit PIC
+ // uses RIP relative addressing.
+ if (TM->getSubtarget<X86Subtarget>().is64Bit())
+ return false;
// Only emit a global base reg in PIC mode.
if (TM->getRelocationModel() != Reloc::PIC_)
MachineBasicBlock::iterator MBBI = FirstMBB.begin();
DebugLoc DL = FirstMBB.findDebugLoc(MBBI);
MachineRegisterInfo &RegInfo = MF.getRegInfo();
- const X86InstrInfo *TII = TM->getInstrInfo();
+ const X86InstrInfo *TII = TM->getSubtargetImpl()->getInstrInfo();
unsigned PC;
if (TM->getSubtarget<X86Subtarget>().isPICStyleGOT())
char CGBR::ID = 0;
FunctionPass*
-llvm::createGlobalBaseRegPass() { return new CGBR(); }
+llvm::createX86GlobalBaseRegPass() { return new CGBR(); }
namespace {
struct LDTLSCleanup : public MachineFunctionPass {
const X86TargetMachine *TM =
static_cast<const X86TargetMachine *>(&MF->getTarget());
const bool is64Bit = TM->getSubtarget<X86Subtarget>().is64Bit();
- const X86InstrInfo *TII = TM->getInstrInfo();
+ const X86InstrInfo *TII = TM->getSubtargetImpl()->getInstrInfo();
// Insert a Copy from TLSBaseAddrReg to RAX/EAX.
MachineInstr *Copy = BuildMI(*I->getParent(), I, I->getDebugLoc(),
const X86TargetMachine *TM =
static_cast<const X86TargetMachine *>(&MF->getTarget());
const bool is64Bit = TM->getSubtarget<X86Subtarget>().is64Bit();
- const X86InstrInfo *TII = TM->getInstrInfo();
+ const X86InstrInfo *TII = TM->getSubtargetImpl()->getInstrInfo();
// Create a virtual register for the TLS base address.
MachineRegisterInfo &RegInfo = MF->getRegInfo();