{ X86::CVTSD2SIrr, X86::CVTSD2SIrm, 0 },
{ X86::CVTSS2SI64rr, X86::CVTSS2SI64rm, 0 },
{ X86::CVTSS2SIrr, X86::CVTSS2SIrm, 0 },
- { X86::Int_CVTSD2SSrr, X86::Int_CVTSD2SSrm, 0 },
- { X86::Int_CVTSI2SD64rr,X86::Int_CVTSI2SD64rm, 0 },
- { X86::Int_CVTSI2SDrr, X86::Int_CVTSI2SDrm, 0 },
- { X86::Int_CVTSI2SS64rr,X86::Int_CVTSI2SS64rm, 0 },
- { X86::Int_CVTSI2SSrr, X86::Int_CVTSI2SSrm, 0 },
- { X86::Int_CVTSS2SDrr, X86::Int_CVTSS2SDrm, 0 },
{ X86::CVTTPD2DQrr, X86::CVTTPD2DQrm, TB_ALIGN_16 },
{ X86::CVTTPS2DQrr, X86::CVTTPS2DQrm, TB_ALIGN_16 },
{ X86::Int_CVTTSD2SI64rr,X86::Int_CVTTSD2SI64rm, 0 },
{ X86::IMUL64rr, X86::IMUL64rm, 0 },
{ X86::Int_CMPSDrr, X86::Int_CMPSDrm, 0 },
{ X86::Int_CMPSSrr, X86::Int_CMPSSrm, 0 },
+ { X86::Int_CVTSD2SSrr, X86::Int_CVTSD2SSrm, 0 },
+ { X86::Int_CVTSI2SD64rr,X86::Int_CVTSI2SD64rm, 0 },
+ { X86::Int_CVTSI2SDrr, X86::Int_CVTSI2SDrm, 0 },
+ { X86::Int_CVTSI2SS64rr,X86::Int_CVTSI2SS64rm, 0 },
+ { X86::Int_CVTSI2SSrr, X86::Int_CVTSI2SSrm, 0 },
+ { X86::Int_CVTSS2SDrr, X86::Int_CVTSS2SDrm, 0 },
{ X86::MAXPDrr, X86::MAXPDrm, TB_ALIGN_16 },
{ X86::MAXPDrr_Int, X86::MAXPDrm_Int, TB_ALIGN_16 },
{ X86::MAXPSrr, X86::MAXPSrm, TB_ALIGN_16 },
{ X86::VPUNPCKLWDYrr, X86::VPUNPCKLWDYrm, TB_ALIGN_32 },
{ X86::VPXORYrr, X86::VPXORYrm, TB_ALIGN_32 },
// FIXME: add AVX 256-bit foldable instructions
+
+ // FMA4 foldable patterns
+ { X86::VFMADDSS4rr, X86::VFMADDSS4mr, TB_ALIGN_16 },
+ { X86::VFMADDSD4rr, X86::VFMADDSD4mr, TB_ALIGN_16 },
+ { X86::VFMADDPS4rr, X86::VFMADDPS4mr, TB_ALIGN_16 },
+ { X86::VFMADDPD4rr, X86::VFMADDPD4mr, TB_ALIGN_16 },
+ { X86::VFMADDPS4rrY, X86::VFMADDPS4mrY, TB_ALIGN_32 },
+ { X86::VFMADDPD4rrY, X86::VFMADDPD4mrY, TB_ALIGN_32 },
+ { X86::VFNMADDPS4rr, X86::VFNMADDPS4mr, TB_ALIGN_16 },
+ { X86::VFNMADDPD4rr, X86::VFNMADDPD4mr, TB_ALIGN_16 },
+ { X86::VFNMADDPS4rrY, X86::VFNMADDPS4mrY, TB_ALIGN_32 },
+ { X86::VFNMADDPD4rrY, X86::VFNMADDPD4mrY, TB_ALIGN_32 },
+ { X86::VFMSUBSS4rr, X86::VFMSUBSS4mr, TB_ALIGN_16 },
+ { X86::VFMSUBSD4rr, X86::VFMSUBSD4mr, TB_ALIGN_16 },
+ { X86::VFMSUBPS4rr, X86::VFMSUBPS4mr, TB_ALIGN_16 },
+ { X86::VFMSUBPD4rr, X86::VFMSUBPD4mr, TB_ALIGN_16 },
+ { X86::VFMSUBPS4rrY, X86::VFMSUBPS4mrY, TB_ALIGN_32 },
+ { X86::VFMSUBPD4rrY, X86::VFMSUBPD4mrY, TB_ALIGN_32 },
+ { X86::VFNMSUBPS4rr, X86::VFNMSUBPS4mr, TB_ALIGN_16 },
+ { X86::VFNMSUBPD4rr, X86::VFNMSUBPD4mr, TB_ALIGN_16 },
+ { X86::VFNMSUBPS4rrY, X86::VFNMSUBPS4mrY, TB_ALIGN_32 },
+ { X86::VFNMSUBPD4rrY, X86::VFNMSUBPD4mrY, TB_ALIGN_32 },
+ { X86::VFMADDSUBPS4rr, X86::VFMADDSUBPS4mr, TB_ALIGN_16 },
+ { X86::VFMADDSUBPD4rr, X86::VFMADDSUBPD4mr, TB_ALIGN_16 },
+ { X86::VFMADDSUBPS4rrY, X86::VFMADDSUBPS4mrY, TB_ALIGN_32 },
+ { X86::VFMADDSUBPD4rrY, X86::VFMADDSUBPD4mrY, TB_ALIGN_32 },
+ { X86::VFMSUBADDPS4rr, X86::VFMSUBADDPS4mr, TB_ALIGN_16 },
+ { X86::VFMSUBADDPD4rr, X86::VFMSUBADDPD4mr, TB_ALIGN_16 },
+ { X86::VFMSUBADDPS4rrY, X86::VFMSUBADDPS4mrY, TB_ALIGN_32 },
+ { X86::VFMSUBADDPD4rrY, X86::VFMSUBADDPD4mrY, TB_ALIGN_32 },
};
for (unsigned i = 0, e = array_lengthof(OpTbl2); i != e; ++i) {
{ X86::VFMADDSDr132r, X86::VFMADDSDr132m, 0 },
{ X86::VFMADDSSr213r, X86::VFMADDSSr213m, 0 },
{ X86::VFMADDSDr213r, X86::VFMADDSDr213m, 0 },
- { X86::VFMADDSSr132r_Int, X86::VFMADDSSr132m_Int, 0 },
- { X86::VFMADDSDr132r_Int, X86::VFMADDSDr132m_Int, 0 },
+ { X86::VFMADDSSr213r_Int, X86::VFMADDSSr213m_Int, 0 },
+ { X86::VFMADDSDr213r_Int, X86::VFMADDSDr213m_Int, 0 },
{ X86::VFMADDPSr231r, X86::VFMADDPSr231m, TB_ALIGN_16 },
{ X86::VFMADDPDr231r, X86::VFMADDPDr231m, TB_ALIGN_16 },
{ X86::VFMADDPDr132rY, X86::VFMADDPDr132mY, TB_ALIGN_32 },
{ X86::VFMADDPSr213rY, X86::VFMADDPSr213mY, TB_ALIGN_32 },
{ X86::VFMADDPDr213rY, X86::VFMADDPDr213mY, TB_ALIGN_32 },
- { X86::VFMADDPSr132r_Int, X86::VFMADDPSr132m_Int, TB_ALIGN_16 },
- { X86::VFMADDPDr132r_Int, X86::VFMADDPDr132m_Int, TB_ALIGN_16 },
- { X86::VFMADDPSr132rY_Int, X86::VFMADDPSr132mY_Int, TB_ALIGN_32 },
- { X86::VFMADDPDr132rY_Int, X86::VFMADDPDr132mY_Int, TB_ALIGN_32 },
{ X86::VFNMADDSSr231r, X86::VFNMADDSSr231m, 0 },
{ X86::VFNMADDSDr231r, X86::VFNMADDSDr231m, 0 },
{ X86::VFNMADDSDr132r, X86::VFNMADDSDr132m, 0 },
{ X86::VFNMADDSSr213r, X86::VFNMADDSSr213m, 0 },
{ X86::VFNMADDSDr213r, X86::VFNMADDSDr213m, 0 },
- { X86::VFNMADDSSr132r_Int, X86::VFNMADDSSr132m_Int, 0 },
- { X86::VFNMADDSDr132r_Int, X86::VFNMADDSDr132m_Int, 0 },
+ { X86::VFNMADDSSr213r_Int, X86::VFNMADDSSr213m_Int, 0 },
+ { X86::VFNMADDSDr213r_Int, X86::VFNMADDSDr213m_Int, 0 },
{ X86::VFNMADDPSr231r, X86::VFNMADDPSr231m, TB_ALIGN_16 },
{ X86::VFNMADDPDr231r, X86::VFNMADDPDr231m, TB_ALIGN_16 },
{ X86::VFNMADDPDr132rY, X86::VFNMADDPDr132mY, TB_ALIGN_32 },
{ X86::VFNMADDPSr213rY, X86::VFNMADDPSr213mY, TB_ALIGN_32 },
{ X86::VFNMADDPDr213rY, X86::VFNMADDPDr213mY, TB_ALIGN_32 },
- { X86::VFNMADDPSr132r_Int, X86::VFNMADDPSr132m_Int, TB_ALIGN_16 },
- { X86::VFNMADDPDr132r_Int, X86::VFNMADDPDr132m_Int, TB_ALIGN_16 },
- { X86::VFNMADDPSr132rY_Int, X86::VFNMADDPSr132mY_Int, TB_ALIGN_32 },
- { X86::VFNMADDPDr132rY_Int, X86::VFNMADDPDr132mY_Int, TB_ALIGN_32 },
{ X86::VFMSUBSSr231r, X86::VFMSUBSSr231m, 0 },
{ X86::VFMSUBSDr231r, X86::VFMSUBSDr231m, 0 },
{ X86::VFMSUBSDr132r, X86::VFMSUBSDr132m, 0 },
{ X86::VFMSUBSSr213r, X86::VFMSUBSSr213m, 0 },
{ X86::VFMSUBSDr213r, X86::VFMSUBSDr213m, 0 },
- { X86::VFMSUBSSr132r_Int, X86::VFMSUBSSr132m_Int, 0 },
- { X86::VFMSUBSDr132r_Int, X86::VFMSUBSDr132m_Int, 0 },
+ { X86::VFMSUBSSr213r_Int, X86::VFMSUBSSr213m_Int, 0 },
+ { X86::VFMSUBSDr213r_Int, X86::VFMSUBSDr213m_Int, 0 },
{ X86::VFMSUBPSr231r, X86::VFMSUBPSr231m, TB_ALIGN_16 },
{ X86::VFMSUBPDr231r, X86::VFMSUBPDr231m, TB_ALIGN_16 },
{ X86::VFMSUBPDr132rY, X86::VFMSUBPDr132mY, TB_ALIGN_32 },
{ X86::VFMSUBPSr213rY, X86::VFMSUBPSr213mY, TB_ALIGN_32 },
{ X86::VFMSUBPDr213rY, X86::VFMSUBPDr213mY, TB_ALIGN_32 },
- { X86::VFMSUBPSr132r_Int, X86::VFMSUBPSr132m_Int, TB_ALIGN_16 },
- { X86::VFMSUBPDr132r_Int, X86::VFMSUBPDr132m_Int, TB_ALIGN_16 },
- { X86::VFMSUBPSr132rY_Int, X86::VFMSUBPSr132mY_Int, TB_ALIGN_32 },
- { X86::VFMSUBPDr132rY_Int, X86::VFMSUBPDr132mY_Int, TB_ALIGN_32 },
{ X86::VFNMSUBSSr231r, X86::VFNMSUBSSr231m, 0 },
{ X86::VFNMSUBSDr231r, X86::VFNMSUBSDr231m, 0 },
{ X86::VFNMSUBSDr132r, X86::VFNMSUBSDr132m, 0 },
{ X86::VFNMSUBSSr213r, X86::VFNMSUBSSr213m, 0 },
{ X86::VFNMSUBSDr213r, X86::VFNMSUBSDr213m, 0 },
- { X86::VFNMSUBSSr132r_Int, X86::VFNMSUBSSr132m_Int, 0 },
- { X86::VFNMSUBSDr132r_Int, X86::VFNMSUBSDr132m_Int, 0 },
+ { X86::VFNMSUBSSr213r_Int, X86::VFNMSUBSSr213m_Int, 0 },
+ { X86::VFNMSUBSDr213r_Int, X86::VFNMSUBSDr213m_Int, 0 },
{ X86::VFNMSUBPSr231r, X86::VFNMSUBPSr231m, TB_ALIGN_16 },
{ X86::VFNMSUBPDr231r, X86::VFNMSUBPDr231m, TB_ALIGN_16 },
{ X86::VFNMSUBPDr132rY, X86::VFNMSUBPDr132mY, TB_ALIGN_32 },
{ X86::VFNMSUBPSr213rY, X86::VFNMSUBPSr213mY, TB_ALIGN_32 },
{ X86::VFNMSUBPDr213rY, X86::VFNMSUBPDr213mY, TB_ALIGN_32 },
- { X86::VFNMSUBPSr132r_Int, X86::VFNMSUBPSr132m_Int, TB_ALIGN_16 },
- { X86::VFNMSUBPDr132r_Int, X86::VFNMSUBPDr132m_Int, TB_ALIGN_16 },
- { X86::VFNMSUBPSr132rY_Int, X86::VFNMSUBPSr132mY_Int, TB_ALIGN_32 },
- { X86::VFNMSUBPDr132rY_Int, X86::VFNMSUBPDr132mY_Int, TB_ALIGN_32 },
{ X86::VFMADDSUBPSr231r, X86::VFMADDSUBPSr231m, TB_ALIGN_16 },
{ X86::VFMADDSUBPDr231r, X86::VFMADDSUBPDr231m, TB_ALIGN_16 },
{ X86::VFMADDSUBPDr132rY, X86::VFMADDSUBPDr132mY, TB_ALIGN_32 },
{ X86::VFMADDSUBPSr213rY, X86::VFMADDSUBPSr213mY, TB_ALIGN_32 },
{ X86::VFMADDSUBPDr213rY, X86::VFMADDSUBPDr213mY, TB_ALIGN_32 },
- { X86::VFMADDSUBPSr132r_Int, X86::VFMADDSUBPSr132m_Int, TB_ALIGN_16 },
- { X86::VFMADDSUBPDr132r_Int, X86::VFMADDSUBPDr132m_Int, TB_ALIGN_16 },
- { X86::VFMADDSUBPSr132rY_Int, X86::VFMADDSUBPSr132mY_Int, TB_ALIGN_32 },
- { X86::VFMADDSUBPDr132rY_Int, X86::VFMADDSUBPDr132mY_Int, TB_ALIGN_32 },
{ X86::VFMSUBADDPSr231r, X86::VFMSUBADDPSr231m, TB_ALIGN_16 },
{ X86::VFMSUBADDPDr231r, X86::VFMSUBADDPDr231m, TB_ALIGN_16 },
{ X86::VFMSUBADDPDr132rY, X86::VFMSUBADDPDr132mY, TB_ALIGN_32 },
{ X86::VFMSUBADDPSr213rY, X86::VFMSUBADDPSr213mY, TB_ALIGN_32 },
{ X86::VFMSUBADDPDr213rY, X86::VFMSUBADDPDr213mY, TB_ALIGN_32 },
- { X86::VFMSUBADDPSr132r_Int, X86::VFMSUBADDPSr132m_Int, TB_ALIGN_16 },
- { X86::VFMSUBADDPDr132r_Int, X86::VFMSUBADDPDr132m_Int, TB_ALIGN_16 },
- { X86::VFMSUBADDPSr132rY_Int, X86::VFMSUBADDPSr132mY_Int, TB_ALIGN_32 },
- { X86::VFMSUBADDPDr132rY_Int, X86::VFMSUBADDPDr132mY_Int, TB_ALIGN_32 },
+
+ // FMA4 foldable patterns
+ { X86::VFMADDSS4rr, X86::VFMADDSS4rm, TB_ALIGN_16 },
+ { X86::VFMADDSD4rr, X86::VFMADDSD4rm, TB_ALIGN_16 },
+ { X86::VFMADDPS4rr, X86::VFMADDPS4rm, TB_ALIGN_16 },
+ { X86::VFMADDPD4rr, X86::VFMADDPD4rm, TB_ALIGN_16 },
+ { X86::VFMADDPS4rrY, X86::VFMADDPS4rmY, TB_ALIGN_32 },
+ { X86::VFMADDPD4rrY, X86::VFMADDPD4rmY, TB_ALIGN_32 },
+ { X86::VFNMADDPS4rr, X86::VFNMADDPS4rm, TB_ALIGN_16 },
+ { X86::VFNMADDPD4rr, X86::VFNMADDPD4rm, TB_ALIGN_16 },
+ { X86::VFNMADDPS4rrY, X86::VFNMADDPS4rmY, TB_ALIGN_32 },
+ { X86::VFNMADDPD4rrY, X86::VFNMADDPD4rmY, TB_ALIGN_32 },
+ { X86::VFMSUBSS4rr, X86::VFMSUBSS4rm, TB_ALIGN_16 },
+ { X86::VFMSUBSD4rr, X86::VFMSUBSD4rm, TB_ALIGN_16 },
+ { X86::VFMSUBPS4rr, X86::VFMSUBPS4rm, TB_ALIGN_16 },
+ { X86::VFMSUBPD4rr, X86::VFMSUBPD4rm, TB_ALIGN_16 },
+ { X86::VFMSUBPS4rrY, X86::VFMSUBPS4rmY, TB_ALIGN_32 },
+ { X86::VFMSUBPD4rrY, X86::VFMSUBPD4rmY, TB_ALIGN_32 },
+ { X86::VFNMSUBPS4rr, X86::VFNMSUBPS4rm, TB_ALIGN_16 },
+ { X86::VFNMSUBPD4rr, X86::VFNMSUBPD4rm, TB_ALIGN_16 },
+ { X86::VFNMSUBPS4rrY, X86::VFNMSUBPS4rmY, TB_ALIGN_32 },
+ { X86::VFNMSUBPD4rrY, X86::VFNMSUBPD4rmY, TB_ALIGN_32 },
+ { X86::VFMADDSUBPS4rr, X86::VFMADDSUBPS4rm, TB_ALIGN_16 },
+ { X86::VFMADDSUBPD4rr, X86::VFMADDSUBPD4rm, TB_ALIGN_16 },
+ { X86::VFMADDSUBPS4rrY, X86::VFMADDSUBPS4rmY, TB_ALIGN_32 },
+ { X86::VFMADDSUBPD4rrY, X86::VFMADDSUBPD4rmY, TB_ALIGN_32 },
+ { X86::VFMSUBADDPS4rr, X86::VFMSUBADDPS4rm, TB_ALIGN_16 },
+ { X86::VFMSUBADDPD4rr, X86::VFMSUBADDPD4rm, TB_ALIGN_16 },
+ { X86::VFMSUBADDPS4rrY, X86::VFMSUBADDPS4rmY, TB_ALIGN_32 },
+ { X86::VFMSUBADDPD4rrY, X86::VFMSUBADDPD4rmY, TB_ALIGN_32 },
};
for (unsigned i = 0, e = array_lengthof(OpTbl3); i != e; ++i) {
SrcReg = MI.getOperand(1).getReg();
DstReg = MI.getOperand(0).getReg();
switch (MI.getOpcode()) {
- default:
- llvm_unreachable(0);
+ default: llvm_unreachable("Unreachable!");
case X86::MOVSX16rr8:
case X86::MOVZX16rr8:
case X86::MOVSX32rr8:
/// regIsPICBase - Return true if register is PIC base (i.e.g defined by
/// X86::MOVPC32r.
static bool regIsPICBase(unsigned BaseReg, const MachineRegisterInfo &MRI) {
+ // Don't waste compile time scanning use-def chains of physregs.
+ if (!TargetRegisterInfo::isVirtualRegister(BaseReg))
+ return false;
bool isPICBase = false;
for (MachineRegisterInfo::def_iterator I = MRI.def_begin(BaseReg),
E = MRI.def_end(); I != E; ++I) {
AliasAnalysis *AA) const {
switch (MI->getOpcode()) {
default: break;
- case X86::MOV8rm:
- case X86::MOV16rm:
- case X86::MOV32rm:
- case X86::MOV64rm:
- case X86::LD_Fp64m:
- case X86::MOVSSrm:
- case X86::MOVSDrm:
- case X86::MOVAPSrm:
- case X86::MOVUPSrm:
- case X86::MOVAPDrm:
- case X86::MOVDQArm:
- case X86::VMOVSSrm:
- case X86::VMOVSDrm:
- case X86::VMOVAPSrm:
- case X86::VMOVUPSrm:
- case X86::VMOVAPDrm:
- case X86::VMOVDQArm:
- case X86::VMOVAPSYrm:
- case X86::VMOVUPSYrm:
- case X86::VMOVAPDYrm:
- case X86::VMOVDQAYrm:
- case X86::MMX_MOVD64rm:
- case X86::MMX_MOVQ64rm:
- case X86::FsVMOVAPSrm:
- case X86::FsVMOVAPDrm:
- 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 &&
- MI->isInvariantLoad(AA)) {
- unsigned BaseReg = MI->getOperand(1).getReg();
- if (BaseReg == 0 || BaseReg == X86::RIP)
- return true;
- // Allow re-materialization of PIC load.
- if (!ReMatPICStubLoad && MI->getOperand(4).isGlobal())
- return false;
- const MachineFunction &MF = *MI->getParent()->getParent();
- const MachineRegisterInfo &MRI = MF.getRegInfo();
- bool isPICBase = false;
- for (MachineRegisterInfo::def_iterator I = MRI.def_begin(BaseReg),
- E = MRI.def_end(); I != E; ++I) {
- MachineInstr *DefMI = I.getOperand().getParent();
- if (DefMI->getOpcode() != X86::MOVPC32r)
- return false;
- assert(!isPICBase && "More than one PIC base?");
- isPICBase = true;
- }
- return isPICBase;
- }
- return false;
+ case X86::MOV8rm:
+ case X86::MOV16rm:
+ case X86::MOV32rm:
+ case X86::MOV64rm:
+ case X86::LD_Fp64m:
+ case X86::MOVSSrm:
+ case X86::MOVSDrm:
+ case X86::MOVAPSrm:
+ case X86::MOVUPSrm:
+ case X86::MOVAPDrm:
+ case X86::MOVDQArm:
+ case X86::VMOVSSrm:
+ case X86::VMOVSDrm:
+ case X86::VMOVAPSrm:
+ case X86::VMOVUPSrm:
+ case X86::VMOVAPDrm:
+ case X86::VMOVDQArm:
+ case X86::VMOVAPSYrm:
+ case X86::VMOVUPSYrm:
+ case X86::VMOVAPDYrm:
+ case X86::VMOVDQAYrm:
+ case X86::MMX_MOVD64rm:
+ case X86::MMX_MOVQ64rm:
+ case X86::FsVMOVAPSrm:
+ case X86::FsVMOVAPDrm:
+ 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 &&
+ MI->isInvariantLoad(AA)) {
+ unsigned BaseReg = MI->getOperand(1).getReg();
+ if (BaseReg == 0 || BaseReg == X86::RIP)
+ return true;
+ // Allow re-materialization of PIC load.
+ if (!ReMatPICStubLoad && MI->getOperand(4).isGlobal())
+ return false;
+ const MachineFunction &MF = *MI->getParent()->getParent();
+ const MachineRegisterInfo &MRI = MF.getRegInfo();
+ return regIsPICBase(BaseReg, MRI);
}
+ return false;
+ }
- case X86::LEA32r:
- case X86::LEA64r: {
- if (MI->getOperand(2).isImm() &&
- MI->getOperand(3).isReg() && MI->getOperand(3).getReg() == 0 &&
- !MI->getOperand(4).isReg()) {
- // lea fi#, lea GV, etc. are all rematerializable.
- if (!MI->getOperand(1).isReg())
- return true;
- unsigned BaseReg = MI->getOperand(1).getReg();
- if (BaseReg == 0)
- return true;
- // Allow re-materialization of lea PICBase + x.
- const MachineFunction &MF = *MI->getParent()->getParent();
- const MachineRegisterInfo &MRI = MF.getRegInfo();
- return regIsPICBase(BaseReg, MRI);
- }
- return false;
- }
+ case X86::LEA32r:
+ case X86::LEA64r: {
+ if (MI->getOperand(2).isImm() &&
+ MI->getOperand(3).isReg() && MI->getOperand(3).getReg() == 0 &&
+ !MI->getOperand(4).isReg()) {
+ // lea fi#, lea GV, etc. are all rematerializable.
+ if (!MI->getOperand(1).isReg())
+ return true;
+ unsigned BaseReg = MI->getOperand(1).getReg();
+ if (BaseReg == 0)
+ return true;
+ // Allow re-materialization of lea PICBase + x.
+ const MachineFunction &MF = *MI->getParent()->getParent();
+ const MachineRegisterInfo &MRI = MF.getRegInfo();
+ return regIsPICBase(BaseReg, MRI);
+ }
+ return false;
+ }
}
// All other instructions marked M_REMATERIALIZABLE are always trivially
case X86::MOV64r0: {
if (!isSafeToClobberEFLAGS(MBB, I)) {
switch (Opc) {
- default: break;
+ default: llvm_unreachable("Unreachable!");
case X86::MOV8r0: Opc = X86::MOV8ri; break;
case X86::MOV16r0: Opc = X86::MOV16ri; break;
case X86::MOV32r0: Opc = X86::MOV32ri; break;
MachineInstrBuilder MIB = BuildMI(*MFI, MBBI, MI->getDebugLoc(),
get(Opc), leaOutReg);
switch (MIOpc) {
- default:
- llvm_unreachable(0);
+ default: llvm_unreachable("Unreachable!");
case X86::SHL16ri: {
unsigned ShAmt = MI->getOperand(2).getImm();
MIB.addReg(0).addImm(1 << ShAmt)
MachineInstr *MI = MBBI;
MachineFunction &MF = *MI->getParent()->getParent();
// All instructions input are two-addr instructions. Get the known operands.
- unsigned Dest = MI->getOperand(0).getReg();
- unsigned Src = MI->getOperand(1).getReg();
- bool isDead = MI->getOperand(0).isDead();
- bool isKill = MI->getOperand(1).isKill();
+ const MachineOperand &Dest = MI->getOperand(0);
+ const MachineOperand &Src = MI->getOperand(1);
MachineInstr *NewMI = NULL;
// FIXME: 16-bit LEA's are really slow on Athlons, but not bad on P4's. When
unsigned B = MI->getOperand(1).getReg();
unsigned C = MI->getOperand(2).getReg();
if (B != C) return 0;
- unsigned A = MI->getOperand(0).getReg();
unsigned M = MI->getOperand(3).getImm();
NewMI = BuildMI(MF, MI->getDebugLoc(), get(X86::PSHUFDri))
- .addReg(A, RegState::Define | getDeadRegState(isDead))
- .addReg(B, getKillRegState(isKill)).addImm(M);
+ .addOperand(Dest).addOperand(Src).addImm(M);
break;
}
case X86::SHUFPDrri: {
unsigned B = MI->getOperand(1).getReg();
unsigned C = MI->getOperand(2).getReg();
if (B != C) return 0;
- unsigned A = MI->getOperand(0).getReg();
unsigned M = MI->getOperand(3).getImm();
// Convert to PSHUFD mask.
M = ((M & 1) << 1) | ((M & 1) << 3) | ((M & 2) << 4) | ((M & 2) << 6)| 0x44;
NewMI = BuildMI(MF, MI->getDebugLoc(), get(X86::PSHUFDri))
- .addReg(A, RegState::Define | getDeadRegState(isDead))
- .addReg(B, getKillRegState(isKill)).addImm(M);
+ .addOperand(Dest).addOperand(Src).addImm(M);
break;
}
case X86::SHL64ri: {
if (ShAmt == 0 || ShAmt >= 4) return 0;
// LEA can't handle RSP.
- if (TargetRegisterInfo::isVirtualRegister(Src) &&
- !MF.getRegInfo().constrainRegClass(Src, &X86::GR64_NOSPRegClass))
+ if (TargetRegisterInfo::isVirtualRegister(Src.getReg()) &&
+ !MF.getRegInfo().constrainRegClass(Src.getReg(),
+ &X86::GR64_NOSPRegClass))
return 0;
NewMI = BuildMI(MF, MI->getDebugLoc(), get(X86::LEA64r))
- .addReg(Dest, RegState::Define | getDeadRegState(isDead))
- .addReg(0).addImm(1 << ShAmt)
- .addReg(Src, getKillRegState(isKill))
- .addImm(0).addReg(0);
+ .addOperand(Dest)
+ .addReg(0).addImm(1 << ShAmt).addOperand(Src).addImm(0).addReg(0);
break;
}
case X86::SHL32ri: {
if (ShAmt == 0 || ShAmt >= 4) return 0;
// LEA can't handle ESP.
- if (TargetRegisterInfo::isVirtualRegister(Src) &&
- !MF.getRegInfo().constrainRegClass(Src, &X86::GR32_NOSPRegClass))
+ if (TargetRegisterInfo::isVirtualRegister(Src.getReg()) &&
+ !MF.getRegInfo().constrainRegClass(Src.getReg(),
+ &X86::GR32_NOSPRegClass))
return 0;
unsigned Opc = is64Bit ? X86::LEA64_32r : X86::LEA32r;
NewMI = BuildMI(MF, MI->getDebugLoc(), get(Opc))
- .addReg(Dest, RegState::Define | getDeadRegState(isDead))
- .addReg(0).addImm(1 << ShAmt)
- .addReg(Src, getKillRegState(isKill)).addImm(0).addReg(0);
+ .addOperand(Dest)
+ .addReg(0).addImm(1 << ShAmt).addOperand(Src).addImm(0).addReg(0);
break;
}
case X86::SHL16ri: {
if (DisableLEA16)
return is64Bit ? convertToThreeAddressWithLEA(MIOpc, MFI, MBBI, LV) : 0;
NewMI = BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
- .addReg(Dest, RegState::Define | getDeadRegState(isDead))
- .addReg(0).addImm(1 << ShAmt)
- .addReg(Src, getKillRegState(isKill))
- .addImm(0).addReg(0);
+ .addOperand(Dest)
+ .addReg(0).addImm(1 << ShAmt).addOperand(Src).addImm(0).addReg(0);
break;
}
default: {
(const TargetRegisterClass*)&X86::GR32_NOSPRegClass;
// LEA can't handle RSP.
- if (TargetRegisterInfo::isVirtualRegister(Src) &&
- !MF.getRegInfo().constrainRegClass(Src, RC))
+ if (TargetRegisterInfo::isVirtualRegister(Src.getReg()) &&
+ !MF.getRegInfo().constrainRegClass(Src.getReg(), RC))
return 0;
- NewMI = addRegOffset(BuildMI(MF, MI->getDebugLoc(), get(Opc))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, 1);
+ NewMI = addOffset(BuildMI(MF, MI->getDebugLoc(), get(Opc))
+ .addOperand(Dest).addOperand(Src), 1);
break;
}
case X86::INC16r:
if (DisableLEA16)
return is64Bit ? convertToThreeAddressWithLEA(MIOpc, MFI, MBBI, LV) : 0;
assert(MI->getNumOperands() >= 2 && "Unknown inc instruction!");
- NewMI = addRegOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, 1);
+ NewMI = addOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
+ .addOperand(Dest).addOperand(Src), 1);
break;
case X86::DEC64r:
case X86::DEC32r:
(const TargetRegisterClass*)&X86::GR64_NOSPRegClass :
(const TargetRegisterClass*)&X86::GR32_NOSPRegClass;
// LEA can't handle RSP.
- if (TargetRegisterInfo::isVirtualRegister(Src) &&
- !MF.getRegInfo().constrainRegClass(Src, RC))
+ if (TargetRegisterInfo::isVirtualRegister(Src.getReg()) &&
+ !MF.getRegInfo().constrainRegClass(Src.getReg(), RC))
return 0;
- NewMI = addRegOffset(BuildMI(MF, MI->getDebugLoc(), get(Opc))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, -1);
+ NewMI = addOffset(BuildMI(MF, MI->getDebugLoc(), get(Opc))
+ .addOperand(Dest).addOperand(Src), -1);
break;
}
case X86::DEC16r:
if (DisableLEA16)
return is64Bit ? convertToThreeAddressWithLEA(MIOpc, MFI, MBBI, LV) : 0;
assert(MI->getNumOperands() >= 2 && "Unknown dec instruction!");
- NewMI = addRegOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, -1);
+ NewMI = addOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
+ .addOperand(Dest).addOperand(Src), -1);
break;
case X86::ADD64rr:
case X86::ADD64rr_DB:
return 0;
NewMI = addRegReg(BuildMI(MF, MI->getDebugLoc(), get(Opc))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, Src2, isKill2);
+ .addOperand(Dest),
+ Src.getReg(), Src.isKill(), Src2, isKill2);
// Preserve undefness of the operands.
bool isUndef = MI->getOperand(1).isUndef();
unsigned Src2 = MI->getOperand(2).getReg();
bool isKill2 = MI->getOperand(2).isKill();
NewMI = addRegReg(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, Src2, isKill2);
+ .addOperand(Dest),
+ Src.getReg(), Src.isKill(), Src2, isKill2);
+
+ // Preserve undefness of the operands.
+ bool isUndef = MI->getOperand(1).isUndef();
+ bool isUndef2 = MI->getOperand(2).isUndef();
+ NewMI->getOperand(1).setIsUndef(isUndef);
+ NewMI->getOperand(3).setIsUndef(isUndef2);
+
if (LV && isKill2)
LV->replaceKillInstruction(Src2, MI, NewMI);
break;
case X86::ADD64ri32_DB:
case X86::ADD64ri8_DB:
assert(MI->getNumOperands() >= 3 && "Unknown add instruction!");
- NewMI = addRegOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA64r))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, MI->getOperand(2).getImm());
+ NewMI = addOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA64r))
+ .addOperand(Dest).addOperand(Src),
+ MI->getOperand(2).getImm());
break;
case X86::ADD32ri:
case X86::ADD32ri8:
case X86::ADD32ri8_DB: {
assert(MI->getNumOperands() >= 3 && "Unknown add instruction!");
unsigned Opc = is64Bit ? X86::LEA64_32r : X86::LEA32r;
- NewMI = addRegOffset(BuildMI(MF, MI->getDebugLoc(), get(Opc))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, MI->getOperand(2).getImm());
+ NewMI = addOffset(BuildMI(MF, MI->getDebugLoc(), get(Opc))
+ .addOperand(Dest).addOperand(Src),
+ MI->getOperand(2).getImm());
break;
}
case X86::ADD16ri:
if (DisableLEA16)
return is64Bit ? convertToThreeAddressWithLEA(MIOpc, MFI, MBBI, LV) : 0;
assert(MI->getNumOperands() >= 3 && "Unknown add instruction!");
- NewMI = addRegOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
- .addReg(Dest, RegState::Define |
- getDeadRegState(isDead)),
- Src, isKill, MI->getOperand(2).getImm());
+ NewMI = addOffset(BuildMI(MF, MI->getDebugLoc(), get(X86::LEA16r))
+ .addOperand(Dest).addOperand(Src),
+ MI->getOperand(2).getImm());
break;
}
}
if (!NewMI) return 0;
if (LV) { // Update live variables
- if (isKill)
- LV->replaceKillInstruction(Src, MI, NewMI);
- if (isDead)
- LV->replaceKillInstruction(Dest, MI, NewMI);
+ if (Src.isKill())
+ LV->replaceKillInstruction(Src.getReg(), MI, NewMI);
+ if (Dest.isDead())
+ LV->replaceKillInstruction(Dest.getReg(), MI, NewMI);
}
MFI->insert(MBBI, NewMI); // Insert the new inst
MI->getOperand(3).setImm(Size-Amt);
return TargetInstrInfoImpl::commuteInstruction(MI, NewMI);
}
- case X86::CMOVB16rr:
- case X86::CMOVB32rr:
- case X86::CMOVB64rr:
- case X86::CMOVAE16rr:
- case X86::CMOVAE32rr:
- case X86::CMOVAE64rr:
- case X86::CMOVE16rr:
- case X86::CMOVE32rr:
- case X86::CMOVE64rr:
- case X86::CMOVNE16rr:
- case X86::CMOVNE32rr:
- case X86::CMOVNE64rr:
- case X86::CMOVBE16rr:
- case X86::CMOVBE32rr:
- case X86::CMOVBE64rr:
- case X86::CMOVA16rr:
- case X86::CMOVA32rr:
- case X86::CMOVA64rr:
- case X86::CMOVL16rr:
- case X86::CMOVL32rr:
- case X86::CMOVL64rr:
- case X86::CMOVGE16rr:
- case X86::CMOVGE32rr:
- case X86::CMOVGE64rr:
- case X86::CMOVLE16rr:
- case X86::CMOVLE32rr:
- case X86::CMOVLE64rr:
- case X86::CMOVG16rr:
- case X86::CMOVG32rr:
- case X86::CMOVG64rr:
- case X86::CMOVS16rr:
- case X86::CMOVS32rr:
- case X86::CMOVS64rr:
- case X86::CMOVNS16rr:
- case X86::CMOVNS32rr:
- case X86::CMOVNS64rr:
- case X86::CMOVP16rr:
- case X86::CMOVP32rr:
- case X86::CMOVP64rr:
- case X86::CMOVNP16rr:
- case X86::CMOVNP32rr:
- case X86::CMOVNP64rr:
- case X86::CMOVO16rr:
- case X86::CMOVO32rr:
- case X86::CMOVO64rr:
- case X86::CMOVNO16rr:
- case X86::CMOVNO32rr:
- case X86::CMOVNO64rr: {
- unsigned Opc = 0;
+ case X86::CMOVB16rr: case X86::CMOVB32rr: case X86::CMOVB64rr:
+ case X86::CMOVAE16rr: case X86::CMOVAE32rr: case X86::CMOVAE64rr:
+ case X86::CMOVE16rr: case X86::CMOVE32rr: case X86::CMOVE64rr:
+ case X86::CMOVNE16rr: case X86::CMOVNE32rr: case X86::CMOVNE64rr:
+ case X86::CMOVBE16rr: case X86::CMOVBE32rr: case X86::CMOVBE64rr:
+ case X86::CMOVA16rr: case X86::CMOVA32rr: case X86::CMOVA64rr:
+ case X86::CMOVL16rr: case X86::CMOVL32rr: case X86::CMOVL64rr:
+ case X86::CMOVGE16rr: case X86::CMOVGE32rr: case X86::CMOVGE64rr:
+ case X86::CMOVLE16rr: case X86::CMOVLE32rr: case X86::CMOVLE64rr:
+ case X86::CMOVG16rr: case X86::CMOVG32rr: case X86::CMOVG64rr:
+ case X86::CMOVS16rr: case X86::CMOVS32rr: case X86::CMOVS64rr:
+ case X86::CMOVNS16rr: case X86::CMOVNS32rr: case X86::CMOVNS64rr:
+ case X86::CMOVP16rr: case X86::CMOVP32rr: case X86::CMOVP64rr:
+ case X86::CMOVNP16rr: case X86::CMOVNP32rr: case X86::CMOVNP64rr:
+ case X86::CMOVO16rr: case X86::CMOVO32rr: case X86::CMOVO64rr:
+ case X86::CMOVNO16rr: case X86::CMOVNO32rr: case X86::CMOVNO64rr: {
+ unsigned Opc;
switch (MI->getOpcode()) {
- default: break;
+ default: llvm_unreachable("Unreachable!");
case X86::CMOVB16rr: Opc = X86::CMOVAE16rr; break;
case X86::CMOVB32rr: Opc = X86::CMOVAE32rr; break;
case X86::CMOVB64rr: Opc = X86::CMOVAE64rr; break;
/// whether it has memory operand.
static unsigned getSETFromCond(X86::CondCode CC,
bool HasMemoryOperand) {
- static const unsigned Opc[16][2] = {
+ static const uint16_t Opc[16][2] = {
{ X86::SETAr, X86::SETAm },
{ X86::SETAEr, X86::SETAEm },
{ X86::SETBr, X86::SETBm },
/// register size in bytes, and operand type.
static unsigned getCMovFromCond(X86::CondCode CC, unsigned RegBytes,
bool HasMemoryOperand) {
- static const unsigned Opc[32][3] = {
+ static const uint16_t Opc[32][3] = {
{ X86::CMOVA16rr, X86::CMOVA32rr, X86::CMOVA64rr },
{ X86::CMOVAE16rr, X86::CMOVAE32rr, X86::CMOVAE64rr },
{ X86::CMOVB16rr, X86::CMOVB32rr, X86::CMOVB64rr },
// SrcReg(GR64) -> DestReg(VR64)
if (X86::GR64RegClass.contains(DestReg)) {
- if (X86::VR128RegClass.contains(SrcReg)) {
+ if (X86::VR128RegClass.contains(SrcReg))
// Copy from a VR128 register to a GR64 register.
return HasAVX ? X86::VMOVPQIto64rr : X86::MOVPQIto64rr;
- } else if (X86::VR64RegClass.contains(SrcReg)) {
+ if (X86::VR64RegClass.contains(SrcReg))
// Copy from a VR64 register to a GR64 register.
return X86::MOVSDto64rr;
- }
} else if (X86::GR64RegClass.contains(SrcReg)) {
// Copy from a GR64 register to a VR128 register.
if (X86::VR128RegClass.contains(DestReg))
return HasAVX ? X86::VMOV64toPQIrr : X86::MOV64toPQIrr;
// Copy from a GR64 register to a VR64 register.
- else if (X86::VR64RegClass.contains(DestReg))
+ if (X86::VR64RegClass.contains(DestReg))
return X86::MOV64toSDrr;
}
// SrcReg(GR32) -> DestReg(FR32)
if (X86::GR32RegClass.contains(DestReg) && X86::FR32RegClass.contains(SrcReg))
- // Copy from a FR32 register to a GR32 register.
- return HasAVX ? X86::VMOVSS2DIrr : X86::MOVSS2DIrr;
+ // Copy from a FR32 register to a GR32 register.
+ return HasAVX ? X86::VMOVSS2DIrr : X86::MOVSS2DIrr;
if (X86::FR32RegClass.contains(DestReg) && X86::GR32RegClass.contains(SrcReg))
- // Copy from a GR32 register to a FR32 register.
- return HasAVX ? X86::VMOVDI2SSrr : X86::MOVDI2SSrr;
+ // Copy from a GR32 register to a FR32 register.
+ return HasAVX ? X86::VMOVDI2SSrr : X86::MOVDI2SSrr;
return 0;
}
bool KillSrc) const {
// First deal with the normal symmetric copies.
bool HasAVX = TM.getSubtarget<X86Subtarget>().hasAVX();
- unsigned Opc = 0;
+ unsigned Opc;
if (X86::GR64RegClass.contains(DestReg, SrcReg))
Opc = X86::MOV64rr;
else if (X86::GR32RegClass.contains(DestReg, SrcReg))
BuildMI(MBB, MI, DL, get(X86::PUSHF64));
BuildMI(MBB, MI, DL, get(X86::POP64r), DestReg);
return;
- } else if (X86::GR32RegClass.contains(DestReg)) {
+ }
+ if (X86::GR32RegClass.contains(DestReg)) {
BuildMI(MBB, MI, DL, get(X86::PUSHF32));
BuildMI(MBB, MI, DL, get(X86::POP32r), DestReg);
return;
.addReg(SrcReg, getKillRegState(KillSrc));
BuildMI(MBB, MI, DL, get(X86::POPF64));
return;
- } else if (X86::GR32RegClass.contains(SrcReg)) {
+ }
+ if (X86::GR32RegClass.contains(SrcReg)) {
BuildMI(MBB, MI, DL, get(X86::PUSH32r))
.addReg(SrcReg, getKillRegState(KillSrc));
BuildMI(MBB, MI, DL, get(X86::POPF32));
CmpMask = ~0;
CmpValue = MI->getOperand(1).getImm();
return true;
+ // A SUB can be used to perform comparison.
+ case X86::SUB64rm:
+ case X86::SUB32rm:
+ case X86::SUB16rm:
+ case X86::SUB8rm:
+ SrcReg = MI->getOperand(1).getReg();
+ SrcReg2 = 0;
+ CmpMask = ~0;
+ CmpValue = 0;
+ return true;
+ case X86::SUB64rr:
+ case X86::SUB32rr:
+ case X86::SUB16rr:
+ case X86::SUB8rr:
+ SrcReg = MI->getOperand(1).getReg();
+ SrcReg2 = MI->getOperand(2).getReg();
+ CmpMask = ~0;
+ CmpValue = 0;
+ return true;
+ case X86::SUB64ri32:
+ case X86::SUB64ri8:
+ case X86::SUB32ri:
+ case X86::SUB32ri8:
+ case X86::SUB16ri:
+ case X86::SUB16ri8:
+ case X86::SUB8ri:
+ SrcReg = MI->getOperand(1).getReg();
+ SrcReg2 = 0;
+ CmpMask = ~0;
+ CmpValue = MI->getOperand(2).getImm();
+ return true;
case X86::CMP64rr:
case X86::CMP32rr:
case X86::CMP16rr:
CmpMask = ~0;
CmpValue = 0;
return true;
+ case X86::TEST8rr:
+ case X86::TEST16rr:
+ case X86::TEST32rr:
+ case X86::TEST64rr:
+ SrcReg = MI->getOperand(0).getReg();
+ if (MI->getOperand(1).getReg() != SrcReg) return false;
+ // Compare against zero.
+ SrcReg2 = 0;
+ CmpMask = ~0;
+ CmpValue = 0;
+ return true;
}
return false;
}
return false;
}
+/// isDefConvertible - check whether the definition can be converted
+/// to remove a comparison against zero.
+inline static bool isDefConvertible(MachineInstr *MI) {
+ switch (MI->getOpcode()) {
+ default: return false;
+ case X86::SUB64ri32: case X86::SUB64ri8: case X86::SUB32ri:
+ case X86::SUB32ri8: case X86::SUB16ri: case X86::SUB16ri8:
+ case X86::SUB8ri: case X86::SUB64rr: case X86::SUB32rr:
+ case X86::SUB16rr: case X86::SUB8rr: case X86::SUB64rm:
+ case X86::SUB32rm: case X86::SUB16rm: case X86::SUB8rm:
+ case X86::ADD64ri32: case X86::ADD64ri8: case X86::ADD32ri:
+ case X86::ADD32ri8: case X86::ADD16ri: case X86::ADD16ri8:
+ case X86::ADD8ri: case X86::ADD64rr: case X86::ADD32rr:
+ case X86::ADD16rr: case X86::ADD8rr: case X86::ADD64rm:
+ case X86::ADD32rm: case X86::ADD16rm: case X86::ADD8rm:
+ case X86::AND64ri32: case X86::AND64ri8: case X86::AND32ri:
+ case X86::AND32ri8: case X86::AND16ri: case X86::AND16ri8:
+ case X86::AND8ri: case X86::AND64rr: case X86::AND32rr:
+ case X86::AND16rr: case X86::AND8rr: case X86::AND64rm:
+ case X86::AND32rm: case X86::AND16rm: case X86::AND8rm:
+ case X86::XOR64ri32: case X86::XOR64ri8: case X86::XOR32ri:
+ case X86::XOR32ri8: case X86::XOR16ri: case X86::XOR16ri8:
+ case X86::XOR8ri: case X86::XOR64rr: case X86::XOR32rr:
+ case X86::XOR16rr: case X86::XOR8rr: case X86::XOR64rm:
+ case X86::XOR32rm: case X86::XOR16rm: case X86::XOR8rm:
+ case X86::OR64ri32: case X86::OR64ri8: case X86::OR32ri:
+ case X86::OR32ri8: case X86::OR16ri: case X86::OR16ri8:
+ case X86::OR8ri: case X86::OR64rr: case X86::OR32rr:
+ case X86::OR16rr: case X86::OR8rr: case X86::OR64rm:
+ case X86::OR32rm: case X86::OR16rm: case X86::OR8rm:
+ return true;
+ }
+}
+
/// 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.
optimizeCompareInstr(MachineInstr *CmpInstr, unsigned SrcReg, unsigned SrcReg2,
int CmpMask, int CmpValue,
const MachineRegisterInfo *MRI) const {
+ // Check whether we can replace SUB with CMP.
+ unsigned NewOpcode = 0;
+ switch (CmpInstr->getOpcode()) {
+ default: break;
+ case X86::SUB64ri32:
+ case X86::SUB64ri8:
+ case X86::SUB32ri:
+ case X86::SUB32ri8:
+ case X86::SUB16ri:
+ case X86::SUB16ri8:
+ case X86::SUB8ri:
+ case X86::SUB64rm:
+ case X86::SUB32rm:
+ case X86::SUB16rm:
+ case X86::SUB8rm:
+ case X86::SUB64rr:
+ case X86::SUB32rr:
+ case X86::SUB16rr:
+ case X86::SUB8rr: {
+ if (!MRI->use_nodbg_empty(CmpInstr->getOperand(0).getReg()))
+ return false;
+ // There is no use of the destination register, we can replace SUB with CMP.
+ switch (CmpInstr->getOpcode()) {
+ default: llvm_unreachable("Unreachable!");
+ case X86::SUB64rm: NewOpcode = X86::CMP64rm; break;
+ case X86::SUB32rm: NewOpcode = X86::CMP32rm; break;
+ case X86::SUB16rm: NewOpcode = X86::CMP16rm; break;
+ case X86::SUB8rm: NewOpcode = X86::CMP8rm; break;
+ case X86::SUB64rr: NewOpcode = X86::CMP64rr; break;
+ case X86::SUB32rr: NewOpcode = X86::CMP32rr; break;
+ case X86::SUB16rr: NewOpcode = X86::CMP16rr; break;
+ case X86::SUB8rr: NewOpcode = X86::CMP8rr; break;
+ case X86::SUB64ri32: NewOpcode = X86::CMP64ri32; break;
+ case X86::SUB64ri8: NewOpcode = X86::CMP64ri8; break;
+ case X86::SUB32ri: NewOpcode = X86::CMP32ri; break;
+ case X86::SUB32ri8: NewOpcode = X86::CMP32ri8; break;
+ case X86::SUB16ri: NewOpcode = X86::CMP16ri; break;
+ case X86::SUB16ri8: NewOpcode = X86::CMP16ri8; break;
+ case X86::SUB8ri: NewOpcode = X86::CMP8ri; break;
+ }
+ CmpInstr->setDesc(get(NewOpcode));
+ CmpInstr->RemoveOperand(0);
+ // Fall through to optimize Cmp if Cmp is CMPrr or CMPri.
+ if (NewOpcode == X86::CMP64rm || NewOpcode == X86::CMP32rm ||
+ NewOpcode == X86::CMP16rm || NewOpcode == X86::CMP8rm)
+ return false;
+ }
+ }
+
// Get the unique definition of SrcReg.
MachineInstr *MI = MRI->getUniqueVRegDef(SrcReg);
if (!MI) return false;
// CmpInstr is the first instruction of the BB.
MachineBasicBlock::iterator I = CmpInstr, Def = MI;
+ // 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)))
+ return false;
+
// We are searching for an earlier instruction that can make CmpInstr
// redundant and that instruction will be saved in Sub.
MachineInstr *Sub = NULL;
for (; RI != RE; ++RI) {
MachineInstr *Instr = &*RI;
// Check whether CmpInstr can be made redundant by the current instruction.
- if (isRedundantFlagInstr(CmpInstr, SrcReg, SrcReg2, CmpValue, Instr)) {
+ if (!IsCmpZero &&
+ isRedundantFlagInstr(CmpInstr, SrcReg, SrcReg2, CmpValue, Instr)) {
Sub = Instr;
break;
}
}
// Return false if no candidates exist.
- if (!Sub)
+ if (!IsCmpZero && !Sub)
return false;
bool IsSwapped = (SrcReg2 != 0 && Sub->getOperand(1).getReg() == SrcReg2 &&
MachineBasicBlock::iterator E = CmpInstr->getParent()->end();
for (++I; I != E; ++I) {
const MachineInstr &Instr = *I;
- if (Instr.modifiesRegister(X86::EFLAGS, TRI)) {
+ bool ModifyEFLAGS = Instr.modifiesRegister(X86::EFLAGS, TRI);
+ bool UseEFLAGS = Instr.readsRegister(X86::EFLAGS, TRI);
+ // We should check the usage if this instruction uses and updates EFLAGS.
+ if (!UseEFLAGS && ModifyEFLAGS) {
// It is safe to remove CmpInstr if EFLAGS is updated again.
IsSafe = true;
break;
}
- if (!Instr.readsRegister(X86::EFLAGS, TRI))
+ if (!UseEFLAGS && !ModifyEFLAGS)
continue;
// EFLAGS is used by this instruction.
- 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 decode the condition code from opcode, swap the condition code,
- // and synthesize the new opcode.
- bool OpcIsSET = false;
- X86::CondCode OldCC;
+ X86::CondCode OldCC;
+ bool OpcIsSET = false;
+ if (IsCmpZero || IsSwapped) {
+ // We decode the condition code from opcode.
if (Instr.isBranch())
OldCC = getCondFromBranchOpc(Instr.getOpcode());
else {
OldCC = getCondFromCMovOpc(Instr.getOpcode());
}
if (OldCC == X86::COND_INVALID) return false;
+ }
+ if (IsCmpZero) {
+ switch (OldCC) {
+ default: break;
+ case X86::COND_A: case X86::COND_AE:
+ case X86::COND_B: case X86::COND_BE:
+ case X86::COND_G: case X86::COND_GE:
+ case X86::COND_L: case X86::COND_LE:
+ case X86::COND_O: case X86::COND_NO:
+ // CF and OF are used, we can't perform this optimization.
+ return false;
+ }
+ } 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);
if (NewCC == X86::COND_INVALID) return false;
// instructions will be modified.
OpsToUpdate.push_back(std::make_pair(&*I, NewOpc));
}
- if (Instr.killsRegister(X86::EFLAGS, TRI)) {
+ if (ModifyEFLAGS || Instr.killsRegister(X86::EFLAGS, TRI)) {
+ // It is safe to remove CmpInstr if EFLAGS is updated again or killed.
IsSafe = true;
break;
}
// If EFLAGS is not killed nor re-defined, we should check whether it is
// live-out. If it is live-out, do not optimize.
- if (IsSwapped && !IsSafe) {
+ if ((IsCmpZero || IsSwapped) && !IsSafe) {
MachineBasicBlock *MBB = CmpInstr->getParent();
for (MachineBasicBlock::succ_iterator SI = MBB->succ_begin(),
SE = MBB->succ_end(); SI != SE; ++SI)
return false;
}
+ // The instruction to be updated is either Sub or MI.
+ Sub = IsCmpZero ? MI : Sub;
// Move Movr0Inst to the place right before Sub.
if (Movr0Inst) {
Sub->getParent()->remove(Movr0Inst);
}
// Make sure Sub instruction defines EFLAGS.
- assert(Sub->getNumOperands() >= 4 && Sub->getOperand(3).isReg() &&
- Sub->getOperand(3).getReg() == X86::EFLAGS &&
- "EFLAGS should be the 4th operand of SUBrr or SUBri.");
- Sub->getOperand(3).setIsDef(true);
+ assert(Sub->getNumOperands() >= 2 &&
+ Sub->getOperand(Sub->getNumOperands()-1).isReg() &&
+ Sub->getOperand(Sub->getNumOperands()-1).getReg() == X86::EFLAGS &&
+ "EFLAGS should be the last operand of SUB, ADD, OR, XOR, AND");
+ Sub->getOperand(Sub->getNumOperands()-1).setIsDef(true);
CmpInstr->eraseFromParent();
// Modify the condition code of instructions in OpsToUpdate.
return true;
}
+/// optimizeLoadInstr - Try to remove the load by folding it to a register
+/// operand at the use. We fold the load instructions if load defines a virtual
+/// register, the virtual register is used once in the same BB, and the
+/// instructions in-between do not load or store, and have no side effects.
+MachineInstr* X86InstrInfo::
+optimizeLoadInstr(MachineInstr *MI, const MachineRegisterInfo *MRI,
+ unsigned &FoldAsLoadDefReg,
+ MachineInstr *&DefMI) const {
+ if (FoldAsLoadDefReg == 0)
+ return 0;
+ // To be conservative, if there exists another load, clear the load candidate.
+ if (MI->mayLoad()) {
+ FoldAsLoadDefReg = 0;
+ return 0;
+ }
+
+ // Check whether we can move DefMI here.
+ DefMI = MRI->getVRegDef(FoldAsLoadDefReg);
+ assert(DefMI);
+ bool SawStore = false;
+ if (!DefMI->isSafeToMove(this, 0, SawStore))
+ return 0;
+
+ // We try to commute MI if possible.
+ unsigned IdxEnd = (MI->isCommutable()) ? 2 : 1;
+ for (unsigned Idx = 0; Idx < IdxEnd; Idx++) {
+ // Collect information about virtual register operands of MI.
+ unsigned SrcOperandId = 0;
+ bool FoundSrcOperand = false;
+ for (unsigned i = 0, e = MI->getDesc().getNumOperands(); i != e; ++i) {
+ MachineOperand &MO = MI->getOperand(i);
+ if (!MO.isReg())
+ continue;
+ unsigned Reg = MO.getReg();
+ if (Reg != FoldAsLoadDefReg)
+ continue;
+ // Do not fold if we have a subreg use or a def or multiple uses.
+ if (MO.getSubReg() || MO.isDef() || FoundSrcOperand)
+ return 0;
+
+ SrcOperandId = i;
+ FoundSrcOperand = true;
+ }
+ if (!FoundSrcOperand) return 0;
+
+ // Check whether we can fold the def into SrcOperandId.
+ SmallVector<unsigned, 8> Ops;
+ Ops.push_back(SrcOperandId);
+ MachineInstr *FoldMI = foldMemoryOperand(MI, Ops, DefMI);
+ if (FoldMI) {
+ FoldAsLoadDefReg = 0;
+ return FoldMI;
+ }
+
+ if (Idx == 1) {
+ // MI was changed but it didn't help, commute it back!
+ commuteInstruction(MI, false);
+ return 0;
+ }
+
+ // Check whether we can commute MI and enable folding.
+ if (MI->isCommutable()) {
+ MachineInstr *NewMI = commuteInstruction(MI, false);
+ // Unable to commute.
+ if (!NewMI) return 0;
+ if (NewMI != MI) {
+ // New instruction. It doesn't need to be kept.
+ NewMI->eraseFromParent();
+ return 0;
+ }
+ }
+ }
+ return 0;
+}
+
/// Expand2AddrUndef - Expand a single-def pseudo instruction to a two-addr
/// instruction with two undef reads of the register being defined. This is
/// used for mapping:
case X86::FsFLD0SS:
case X86::FsFLD0SD:
return Expand2AddrUndef(MI, get(HasAVX ? X86::VXORPSrr : X86::XORPSrr));
+ case X86::AVX_SET0:
+ assert(HasAVX && "AVX not supported");
+ return Expand2AddrUndef(MI, get(X86::VXORPSYrr));
+ case X86::V_SETALLONES:
+ return Expand2AddrUndef(MI, get(HasAVX ? X86::VPCMPEQDrr : X86::PCMPEQDrr));
+ case X86::AVX2_SETALLONES:
+ return Expand2AddrUndef(MI, get(X86::VPCMPEQDYrr));
case X86::TEST8ri_NOREX:
MI->setDesc(get(X86::TEST8ri));
return true;
OpcodeTablePtr = &RegOp2MemOpTable2Addr;
isTwoAddrFold = true;
} else if (i == 0) { // If operand 0
- if (MI->getOpcode() == X86::MOV64r0)
- NewMI = MakeM0Inst(*this, X86::MOV64mi32, MOs, MI);
- else if (MI->getOpcode() == X86::MOV32r0)
- NewMI = MakeM0Inst(*this, X86::MOV32mi, MOs, MI);
- else if (MI->getOpcode() == X86::MOV16r0)
- NewMI = MakeM0Inst(*this, X86::MOV16mi, MOs, MI);
- else if (MI->getOpcode() == X86::MOV8r0)
- NewMI = MakeM0Inst(*this, X86::MOV8mi, MOs, MI);
+ unsigned Opc = 0;
+ switch (MI->getOpcode()) {
+ default: break;
+ case X86::MOV64r0: Opc = X86::MOV64mi32; break;
+ case X86::MOV32r0: Opc = X86::MOV32mi; break;
+ case X86::MOV16r0: Opc = X86::MOV16mi; break;
+ case X86::MOV8r0: Opc = X86::MOV8mi; break;
+ }
+ if (Opc)
+ NewMI = MakeM0Inst(*this, Opc, MOs, MI);
if (NewMI)
return NewMI;
OpcodeTablePtr = &RegOp2MemOpTable1;
} else if (i == 2) {
OpcodeTablePtr = &RegOp2MemOpTable2;
+ } else if (i == 3) {
+ OpcodeTablePtr = &RegOp2MemOpTable3;
}
// If table selected...
Alignment = (*LoadMI->memoperands_begin())->getAlignment();
else
switch (LoadMI->getOpcode()) {
- case X86::AVX_SET0PSY:
- case X86::AVX_SET0PDY:
case X86::AVX2_SETALLONES:
- case X86::AVX2_SET0:
+ case X86::AVX_SET0:
Alignment = 32;
break;
case X86::V_SET0:
case X86::V_SETALLONES:
- case X86::AVX_SETALLONES:
Alignment = 16;
break;
case X86::FsFLD0SD:
switch (LoadMI->getOpcode()) {
case X86::V_SET0:
case X86::V_SETALLONES:
- case X86::AVX_SET0PSY:
- case X86::AVX_SET0PDY:
- case X86::AVX_SETALLONES:
case X86::AVX2_SETALLONES:
- case X86::AVX2_SET0:
+ case X86::AVX_SET0:
case X86::FsFLD0SD:
case X86::FsFLD0SS: {
// Folding a V_SET0 or V_SETALLONES as a load, to ease register pressure.
Ty = Type::getFloatTy(MF.getFunction()->getContext());
else if (Opc == X86::FsFLD0SD)
Ty = Type::getDoubleTy(MF.getFunction()->getContext());
- else if (Opc == X86::AVX_SET0PSY || Opc == X86::AVX_SET0PDY)
- Ty = VectorType::get(Type::getFloatTy(MF.getFunction()->getContext()), 8);
- else if (Opc == X86::AVX2_SETALLONES || Opc == X86::AVX2_SET0)
+ else if (Opc == X86::AVX2_SETALLONES || Opc == X86::AVX_SET0)
Ty = VectorType::get(Type::getInt32Ty(MF.getFunction()->getContext()), 8);
else
Ty = VectorType::get(Type::getInt32Ty(MF.getFunction()->getContext()), 4);
- bool IsAllOnes = (Opc == X86::V_SETALLONES || Opc == X86::AVX_SETALLONES ||
- Opc == X86::AVX2_SETALLONES);
+ bool IsAllOnes = (Opc == X86::V_SETALLONES || Opc == X86::AVX2_SETALLONES);
const Constant *C = IsAllOnes ? Constant::getAllOnesValue(Ty) :
Constant::getNullValue(Ty);
unsigned CPI = MCP.getConstantPoolIndex(C, Alignment);
OpcodeTablePtr = &RegOp2MemOpTable1;
} else if (OpNum == 2) {
OpcodeTablePtr = &RegOp2MemOpTable2;
+ } else if (OpNum == 3) {
+ OpcodeTablePtr = &RegOp2MemOpTable3;
}
if (OpcodeTablePtr && OpcodeTablePtr->count(Opc))
getUndefRegState(MO.isUndef()));
}
// Change CMP32ri r, 0 back to TEST32rr r, r, etc.
- unsigned NewOpc = 0;
switch (DataMI->getOpcode()) {
default: break;
case X86::CMP64ri32:
MachineOperand &MO0 = DataMI->getOperand(0);
MachineOperand &MO1 = DataMI->getOperand(1);
if (MO1.getImm() == 0) {
+ unsigned NewOpc;
switch (DataMI->getOpcode()) {
- default: break;
+ default: llvm_unreachable("Unreachable!");
case X86::CMP64ri8:
case X86::CMP64ri32: NewOpc = X86::TEST64rr; break;
case X86::CMP32ri8:
projects / oota-llvm.git / blobdiff