//===----------------------------------------------------------------------===//
#define DEBUG_TYPE "mips-lower"
-//#include <algorithm>
#include "MipsISelLowering.h"
#include "MipsMachineFunction.h"
#include "MipsTargetMachine.h"
case MipsISD::Hi: return "MipsISD::Hi";
case MipsISD::Lo: return "MipsISD::Lo";
case MipsISD::GPRel: return "MipsISD::GPRel";
+ case MipsISD::TlsGd: return "MipsISD::TlsGd";
+ case MipsISD::TprelHi: return "MipsISD::TprelHi";
+ case MipsISD::TprelLo: return "MipsISD::TprelLo";
+ case MipsISD::ThreadPointer: return "MipsISD::ThreadPointer";
case MipsISD::Ret: return "MipsISD::Ret";
case MipsISD::FPBrcond: return "MipsISD::FPBrcond";
case MipsISD::FPCmp: return "MipsISD::FPCmp";
case MipsISD::DivRemU: return "MipsISD::DivRemU";
case MipsISD::BuildPairF64: return "MipsISD::BuildPairF64";
case MipsISD::ExtractElementF64: return "MipsISD::ExtractElementF64";
- default: return NULL;
+ case MipsISD::WrapperPIC: return "MipsISD::WrapperPIC";
+ default: return NULL;
}
}
setOperationAction(ISD::SELECT, MVT::i32, Custom);
setOperationAction(ISD::BRCOND, MVT::Other, Custom);
setOperationAction(ISD::DYNAMIC_STACKALLOC, MVT::i32, Custom);
- setOperationAction(ISD::FP_TO_SINT, MVT::i32, Custom);
setOperationAction(ISD::VASTART, MVT::Other, Custom);
setOperationAction(ISD::SDIV, MVT::i32, Expand);
setOperationAction(ISD::FLOG10, MVT::f32, Expand);
setOperationAction(ISD::FEXP, MVT::f32, Expand);
- setOperationAction(ISD::EH_LABEL, MVT::Other, Expand);
+ setOperationAction(ISD::EXCEPTIONADDR, MVT::i32, Expand);
+ setOperationAction(ISD::EHSELECTION, MVT::i32, Expand);
setOperationAction(ISD::VAARG, MVT::Other, Expand);
setOperationAction(ISD::VACOPY, MVT::Other, Expand);
setStackPointerRegisterToSaveRestore(Mips::SP);
computeRegisterProperties();
+
+ setExceptionPointerRegister(Mips::A0);
+ setExceptionSelectorRegister(Mips::A1);
}
MVT::SimpleValueType MipsTargetLowering::getSetCCResultType(EVT VT) const {
case ISD::BRCOND: return LowerBRCOND(Op, DAG);
case ISD::ConstantPool: return LowerConstantPool(Op, DAG);
case ISD::DYNAMIC_STACKALLOC: return LowerDYNAMIC_STACKALLOC(Op, DAG);
- case ISD::FP_TO_SINT: return LowerFP_TO_SINT(Op, DAG);
case ISD::GlobalAddress: return LowerGlobalAddress(Op, DAG);
case ISD::BlockAddress: return LowerBlockAddress(Op, DAG);
case ISD::GlobalTLSAddress: return LowerGlobalTLSAddress(Op, DAG);
case ISD::SELECT: return LowerSELECT(Op, DAG);
case ISD::VASTART: return LowerVASTART(Op, DAG);
case ISD::FCOPYSIGN: return LowerFCOPYSIGN(Op, DAG);
+ case ISD::FRAMEADDR: return LowerFRAMEADDR(Op, DAG);
}
return SDValue();
}
return Mips::BRANCH_INVALID;
}
-MachineBasicBlock *
-MipsTargetLowering::EmitInstrWithCustomInserter(MachineInstr *MI,
- MachineBasicBlock *BB) const {
+static MachineBasicBlock* ExpandCondMov(MachineInstr *MI, MachineBasicBlock *BB,
+ DebugLoc dl,
+ const MipsSubtarget* Subtarget,
+ const TargetInstrInfo *TII,
+ bool isFPCmp, unsigned Opc) {
// There is no need to expand CMov instructions if target has
// conditional moves.
if (Subtarget->hasCondMov())
return BB;
- const TargetInstrInfo *TII = getTargetMachine().getInstrInfo();
- bool isFPCmp = false;
- DebugLoc dl = MI->getDebugLoc();
- unsigned Opc;
-
- switch (MI->getOpcode()) {
- default: assert(false && "Unexpected instr type to insert");
- case Mips::MOVT:
- case Mips::MOVT_S:
- case Mips::MOVT_D:
- isFPCmp = true;
- Opc = Mips::BC1F;
- break;
- case Mips::MOVF:
- case Mips::MOVF_S:
- case Mips::MOVF_D:
- isFPCmp = true;
- Opc = Mips::BC1T;
- break;
- case Mips::MOVZ_I:
- case Mips::MOVZ_S:
- case Mips::MOVZ_D:
- Opc = Mips::BNE;
- break;
- case Mips::MOVN_I:
- case Mips::MOVN_S:
- case Mips::MOVN_D:
- Opc = Mips::BEQ;
- break;
- }
-
// To "insert" a SELECT_CC instruction, we actually have to insert the
// diamond control-flow pattern. The incoming instruction knows the
// destination vreg to set, the condition code register to branch on, the
BuildMI(BB, dl, TII->get(Opc)).addReg(MI->getOperand(2).getReg())
.addReg(Mips::ZERO).addMBB(sinkMBB);
-
// copy0MBB:
// %FalseValue = ...
// # fallthrough to sinkMBB
return BB;
}
-//===----------------------------------------------------------------------===//
-// Misc Lower Operation implementation
-//===----------------------------------------------------------------------===//
+MachineBasicBlock *
+MipsTargetLowering::EmitInstrWithCustomInserter(MachineInstr *MI,
+ MachineBasicBlock *BB) const {
+ const TargetInstrInfo *TII = getTargetMachine().getInstrInfo();
+ DebugLoc dl = MI->getDebugLoc();
-SDValue MipsTargetLowering::
-LowerFP_TO_SINT(SDValue Op, SelectionDAG &DAG) const
-{
- if (!Subtarget->isMips1())
- return Op;
+ switch (MI->getOpcode()) {
+ default:
+ assert(false && "Unexpected instr type to insert");
+ return NULL;
+ case Mips::MOVT:
+ case Mips::MOVT_S:
+ case Mips::MOVT_D:
+ return ExpandCondMov(MI, BB, dl, Subtarget, TII, true, Mips::BC1F);
+ case Mips::MOVF:
+ case Mips::MOVF_S:
+ case Mips::MOVF_D:
+ return ExpandCondMov(MI, BB, dl, Subtarget, TII, true, Mips::BC1T);
+ case Mips::MOVZ_I:
+ case Mips::MOVZ_S:
+ case Mips::MOVZ_D:
+ return ExpandCondMov(MI, BB, dl, Subtarget, TII, false, Mips::BNE);
+ case Mips::MOVN_I:
+ case Mips::MOVN_S:
+ case Mips::MOVN_D:
+ return ExpandCondMov(MI, BB, dl, Subtarget, TII, false, Mips::BEQ);
+
+ case Mips::ATOMIC_LOAD_ADD_I8:
+ return EmitAtomicBinaryPartword(MI, BB, 1, Mips::ADDu);
+ case Mips::ATOMIC_LOAD_ADD_I16:
+ return EmitAtomicBinaryPartword(MI, BB, 2, Mips::ADDu);
+ case Mips::ATOMIC_LOAD_ADD_I32:
+ return EmitAtomicBinary(MI, BB, 4, Mips::ADDu);
+
+ case Mips::ATOMIC_LOAD_AND_I8:
+ return EmitAtomicBinaryPartword(MI, BB, 1, Mips::AND);
+ case Mips::ATOMIC_LOAD_AND_I16:
+ return EmitAtomicBinaryPartword(MI, BB, 2, Mips::AND);
+ case Mips::ATOMIC_LOAD_AND_I32:
+ return EmitAtomicBinary(MI, BB, 4, Mips::AND);
+
+ case Mips::ATOMIC_LOAD_OR_I8:
+ return EmitAtomicBinaryPartword(MI, BB, 1, Mips::OR);
+ case Mips::ATOMIC_LOAD_OR_I16:
+ return EmitAtomicBinaryPartword(MI, BB, 2, Mips::OR);
+ case Mips::ATOMIC_LOAD_OR_I32:
+ return EmitAtomicBinary(MI, BB, 4, Mips::OR);
+
+ case Mips::ATOMIC_LOAD_XOR_I8:
+ return EmitAtomicBinaryPartword(MI, BB, 1, Mips::XOR);
+ case Mips::ATOMIC_LOAD_XOR_I16:
+ return EmitAtomicBinaryPartword(MI, BB, 2, Mips::XOR);
+ case Mips::ATOMIC_LOAD_XOR_I32:
+ return EmitAtomicBinary(MI, BB, 4, Mips::XOR);
+
+ case Mips::ATOMIC_LOAD_NAND_I8:
+ return EmitAtomicBinaryPartword(MI, BB, 1, 0, true);
+ case Mips::ATOMIC_LOAD_NAND_I16:
+ return EmitAtomicBinaryPartword(MI, BB, 2, 0, true);
+ case Mips::ATOMIC_LOAD_NAND_I32:
+ return EmitAtomicBinary(MI, BB, 4, 0, true);
+
+ case Mips::ATOMIC_LOAD_SUB_I8:
+ return EmitAtomicBinaryPartword(MI, BB, 1, Mips::SUBu);
+ case Mips::ATOMIC_LOAD_SUB_I16:
+ return EmitAtomicBinaryPartword(MI, BB, 2, Mips::SUBu);
+ case Mips::ATOMIC_LOAD_SUB_I32:
+ return EmitAtomicBinary(MI, BB, 4, Mips::SUBu);
+
+ case Mips::ATOMIC_SWAP_I8:
+ return EmitAtomicBinaryPartword(MI, BB, 1, 0);
+ case Mips::ATOMIC_SWAP_I16:
+ return EmitAtomicBinaryPartword(MI, BB, 2, 0);
+ case Mips::ATOMIC_SWAP_I32:
+ return EmitAtomicBinary(MI, BB, 4, 0);
+
+ case Mips::ATOMIC_CMP_SWAP_I8:
+ return EmitAtomicCmpSwapPartword(MI, BB, 1);
+ case Mips::ATOMIC_CMP_SWAP_I16:
+ return EmitAtomicCmpSwapPartword(MI, BB, 2);
+ case Mips::ATOMIC_CMP_SWAP_I32:
+ return EmitAtomicCmpSwap(MI, BB, 4);
+ }
+}
- MachineFunction &MF = DAG.getMachineFunction();
- unsigned CCReg = AddLiveIn(MF, Mips::FCR31, Mips::CCRRegisterClass);
+// This function also handles Mips::ATOMIC_SWAP_I32 (when BinOpcode == 0), and
+// Mips::ATOMIC_LOAD_NAND_I32 (when Nand == true)
+MachineBasicBlock *
+MipsTargetLowering::EmitAtomicBinary(MachineInstr *MI, MachineBasicBlock *BB,
+ unsigned Size, unsigned BinOpcode,
+ bool Nand) const {
+ assert(Size == 4 && "Unsupported size for EmitAtomicBinary.");
+
+ MachineFunction *MF = BB->getParent();
+ MachineRegisterInfo &RegInfo = MF->getRegInfo();
+ const TargetRegisterClass *RC = getRegClassFor(MVT::i32);
+ const TargetInstrInfo *TII = getTargetMachine().getInstrInfo();
+ DebugLoc dl = MI->getDebugLoc();
- SDValue Chain = DAG.getEntryNode();
- DebugLoc dl = Op.getDebugLoc();
- SDValue Src = Op.getOperand(0);
-
- // Set the condition register
- SDValue CondReg = DAG.getCopyFromReg(Chain, dl, CCReg, MVT::i32);
- CondReg = DAG.getCopyToReg(Chain, dl, Mips::AT, CondReg);
- CondReg = DAG.getCopyFromReg(CondReg, dl, Mips::AT, MVT::i32);
-
- SDValue Cst = DAG.getConstant(3, MVT::i32);
- SDValue Or = DAG.getNode(ISD::OR, dl, MVT::i32, CondReg, Cst);
- Cst = DAG.getConstant(2, MVT::i32);
- SDValue Xor = DAG.getNode(ISD::XOR, dl, MVT::i32, Or, Cst);
-
- SDValue InFlag(0, 0);
- CondReg = DAG.getCopyToReg(Chain, dl, Mips::FCR31, Xor, InFlag);
-
- // Emit the round instruction and bit convert to integer
- SDValue Trunc = DAG.getNode(MipsISD::FPRound, dl, MVT::f32,
- Src, CondReg.getValue(1));
- SDValue BitCvt = DAG.getNode(ISD::BITCAST, dl, MVT::i32, Trunc);
- return BitCvt;
+ unsigned Dest = MI->getOperand(0).getReg();
+ unsigned Ptr = MI->getOperand(1).getReg();
+ unsigned Incr = MI->getOperand(2).getReg();
+
+ unsigned Oldval = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp1 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp2 = RegInfo.createVirtualRegister(RC);
+
+ // insert new blocks after the current block
+ const BasicBlock *LLVM_BB = BB->getBasicBlock();
+ MachineBasicBlock *loopMBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineBasicBlock *exitMBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineFunction::iterator It = BB;
+ ++It;
+ MF->insert(It, loopMBB);
+ MF->insert(It, exitMBB);
+
+ // Transfer the remainder of BB and its successor edges to exitMBB.
+ exitMBB->splice(exitMBB->begin(), BB,
+ llvm::next(MachineBasicBlock::iterator(MI)),
+ BB->end());
+ exitMBB->transferSuccessorsAndUpdatePHIs(BB);
+
+ // thisMBB:
+ // ...
+ // sw incr, fi(sp) // store incr to stack (when BinOpcode == 0)
+ // fallthrough --> loopMBB
+
+ // Note: for atomic.swap (when BinOpcode == 0), storing incr to stack before
+ // the loop and then loading it from stack in block loopMBB is necessary to
+ // prevent MachineLICM pass to hoist "or" instruction out of the block
+ // loopMBB.
+
+ int fi = 0;
+ if (BinOpcode == 0 && !Nand) {
+ // Get or create a temporary stack location.
+ MipsFunctionInfo *MipsFI = MF->getInfo<MipsFunctionInfo>();
+ fi = MipsFI->getAtomicFrameIndex();
+ if (fi == -1) {
+ fi = MF->getFrameInfo()->CreateStackObject(Size, Size, false);
+ MipsFI->setAtomicFrameIndex(fi);
+ }
+
+ BuildMI(BB, dl, TII->get(Mips::SW))
+ .addReg(Incr).addImm(0).addFrameIndex(fi);
+ }
+ BB->addSuccessor(loopMBB);
+
+ // loopMBB:
+ // ll oldval, 0(ptr)
+ // or dest, $0, oldval
+ // <binop> tmp1, oldval, incr
+ // sc tmp1, 0(ptr)
+ // beq tmp1, $0, loopMBB
+ BB = loopMBB;
+ BuildMI(BB, dl, TII->get(Mips::LL), Oldval).addImm(0).addReg(Ptr);
+ BuildMI(BB, dl, TII->get(Mips::OR), Dest).addReg(Mips::ZERO).addReg(Oldval);
+ if (Nand) {
+ // and tmp2, oldval, incr
+ // nor tmp1, $0, tmp2
+ BuildMI(BB, dl, TII->get(Mips::AND), Tmp2).addReg(Oldval).addReg(Incr);
+ BuildMI(BB, dl, TII->get(Mips::NOR), Tmp1).addReg(Mips::ZERO).addReg(Tmp2);
+ } else if (BinOpcode) {
+ // <binop> tmp1, oldval, incr
+ BuildMI(BB, dl, TII->get(BinOpcode), Tmp1).addReg(Oldval).addReg(Incr);
+ } else {
+ // lw tmp2, fi(sp) // load incr from stack
+ // or tmp1, $zero, tmp2
+ BuildMI(BB, dl, TII->get(Mips::LW), Tmp2).addImm(0).addFrameIndex(fi);;
+ BuildMI(BB, dl, TII->get(Mips::OR), Tmp1).addReg(Mips::ZERO).addReg(Tmp2);
+ }
+ BuildMI(BB, dl, TII->get(Mips::SC), Tmp1).addReg(Tmp1).addImm(0).addReg(Ptr);
+ BuildMI(BB, dl, TII->get(Mips::BEQ))
+ .addReg(Tmp1).addReg(Mips::ZERO).addMBB(loopMBB);
+ BB->addSuccessor(loopMBB);
+ BB->addSuccessor(exitMBB);
+
+ MI->eraseFromParent(); // The instruction is gone now.
+
+ return BB;
+}
+
+MachineBasicBlock *
+MipsTargetLowering::EmitAtomicBinaryPartword(MachineInstr *MI,
+ MachineBasicBlock *BB,
+ unsigned Size, unsigned BinOpcode,
+ bool Nand) const {
+ assert((Size == 1 || Size == 2) &&
+ "Unsupported size for EmitAtomicBinaryPartial.");
+
+ MachineFunction *MF = BB->getParent();
+ MachineRegisterInfo &RegInfo = MF->getRegInfo();
+ const TargetRegisterClass *RC = getRegClassFor(MVT::i32);
+ const TargetInstrInfo *TII = getTargetMachine().getInstrInfo();
+ DebugLoc dl = MI->getDebugLoc();
+
+ unsigned Dest = MI->getOperand(0).getReg();
+ unsigned Ptr = MI->getOperand(1).getReg();
+ unsigned Incr = MI->getOperand(2).getReg();
+
+ unsigned Addr = RegInfo.createVirtualRegister(RC);
+ unsigned Shift = RegInfo.createVirtualRegister(RC);
+ unsigned Mask = RegInfo.createVirtualRegister(RC);
+ unsigned Mask2 = RegInfo.createVirtualRegister(RC);
+ unsigned Newval = RegInfo.createVirtualRegister(RC);
+ unsigned Oldval = RegInfo.createVirtualRegister(RC);
+ unsigned Incr2 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp1 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp2 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp3 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp4 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp5 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp6 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp7 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp8 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp9 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp10 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp11 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp12 = RegInfo.createVirtualRegister(RC);
+
+ // insert new blocks after the current block
+ const BasicBlock *LLVM_BB = BB->getBasicBlock();
+ MachineBasicBlock *loopMBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineBasicBlock *exitMBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineFunction::iterator It = BB;
+ ++It;
+ MF->insert(It, loopMBB);
+ MF->insert(It, exitMBB);
+
+ // Transfer the remainder of BB and its successor edges to exitMBB.
+ exitMBB->splice(exitMBB->begin(), BB,
+ llvm::next(MachineBasicBlock::iterator(MI)),
+ BB->end());
+ exitMBB->transferSuccessorsAndUpdatePHIs(BB);
+
+ // thisMBB:
+ // addiu tmp1,$0,-4 # 0xfffffffc
+ // and addr,ptr,tmp1
+ // andi tmp2,ptr,3
+ // sll shift,tmp2,3
+ // ori tmp3,$0,255 # 0xff
+ // sll mask,tmp3,shift
+ // nor mask2,$0,mask
+ // andi tmp4,incr,255
+ // sll incr2,tmp4,shift
+ // sw incr2, fi(sp) // store incr2 to stack (when BinOpcode == 0)
+
+ // Note: for atomic.swap (when BinOpcode == 0), storing incr2 to stack before
+ // the loop and then loading it from stack in block loopMBB is necessary to
+ // prevent MachineLICM pass to hoist "or" instruction out of the block
+ // loopMBB.
+
+ int64_t MaskImm = (Size == 1) ? 255 : 65535;
+ BuildMI(BB, dl, TII->get(Mips::ADDiu), Tmp1).addReg(Mips::ZERO).addImm(-4);
+ BuildMI(BB, dl, TII->get(Mips::AND), Addr).addReg(Ptr).addReg(Tmp1);
+ BuildMI(BB, dl, TII->get(Mips::ANDi), Tmp2).addReg(Ptr).addImm(3);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Shift).addReg(Tmp2).addImm(3);
+ BuildMI(BB, dl, TII->get(Mips::ORi), Tmp3).addReg(Mips::ZERO).addImm(MaskImm);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Mask).addReg(Tmp3).addReg(Shift);
+ BuildMI(BB, dl, TII->get(Mips::NOR), Mask2).addReg(Mips::ZERO).addReg(Mask);
+ if (BinOpcode != Mips::SUBu) {
+ BuildMI(BB, dl, TII->get(Mips::ANDi), Tmp4).addReg(Incr).addImm(MaskImm);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Incr2).addReg(Tmp4).addReg(Shift);
+ } else {
+ BuildMI(BB, dl, TII->get(Mips::SUBu), Tmp4).addReg(Mips::ZERO).addReg(Incr);
+ BuildMI(BB, dl, TII->get(Mips::ANDi), Tmp5).addReg(Tmp4).addImm(MaskImm);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Incr2).addReg(Tmp5).addReg(Shift);
+ }
+
+ int fi = 0;
+ if (BinOpcode == 0 && !Nand) {
+ // Get or create a temporary stack location.
+ MipsFunctionInfo *MipsFI = MF->getInfo<MipsFunctionInfo>();
+ fi = MipsFI->getAtomicFrameIndex();
+ if (fi == -1) {
+ fi = MF->getFrameInfo()->CreateStackObject(Size, Size, false);
+ MipsFI->setAtomicFrameIndex(fi);
+ }
+
+ BuildMI(BB, dl, TII->get(Mips::SW))
+ .addReg(Incr2).addImm(0).addFrameIndex(fi);
+ }
+ BB->addSuccessor(loopMBB);
+
+ // loopMBB:
+ // ll oldval,0(addr)
+ // binop tmp7,oldval,incr2
+ // and newval,tmp7,mask
+ // and tmp8,oldval,mask2
+ // or tmp9,tmp8,newval
+ // sc tmp9,0(addr)
+ // beq tmp9,$0,loopMBB
+ BB = loopMBB;
+ BuildMI(BB, dl, TII->get(Mips::LL), Oldval).addImm(0).addReg(Addr);
+ if (Nand) {
+ // and tmp6, oldval, incr2
+ // nor tmp7, $0, tmp6
+ BuildMI(BB, dl, TII->get(Mips::AND), Tmp6).addReg(Oldval).addReg(Incr2);
+ BuildMI(BB, dl, TII->get(Mips::NOR), Tmp7).addReg(Mips::ZERO).addReg(Tmp6);
+ } else if (BinOpcode == Mips::SUBu) {
+ // addu tmp7, oldval, incr2
+ BuildMI(BB, dl, TII->get(Mips::ADDu), Tmp7).addReg(Oldval).addReg(Incr2);
+ } else if (BinOpcode) {
+ // <binop> tmp7, oldval, incr2
+ BuildMI(BB, dl, TII->get(BinOpcode), Tmp7).addReg(Oldval).addReg(Incr2);
+ } else {
+ // lw tmp6, fi(sp) // load incr2 from stack
+ // or tmp7, $zero, tmp6
+ BuildMI(BB, dl, TII->get(Mips::LW), Tmp6).addImm(0).addFrameIndex(fi);;
+ BuildMI(BB, dl, TII->get(Mips::OR), Tmp7).addReg(Mips::ZERO).addReg(Tmp6);
+ }
+ BuildMI(BB, dl, TII->get(Mips::AND), Newval).addReg(Tmp7).addReg(Mask);
+ BuildMI(BB, dl, TII->get(Mips::AND), Tmp8).addReg(Oldval).addReg(Mask2);
+ BuildMI(BB, dl, TII->get(Mips::OR), Tmp9).addReg(Tmp8).addReg(Newval);
+ BuildMI(BB, dl, TII->get(Mips::SC), Tmp9).addReg(Tmp9).addImm(0).addReg(Addr);
+ BuildMI(BB, dl, TII->get(Mips::BEQ))
+ .addReg(Tmp9).addReg(Mips::ZERO).addMBB(loopMBB);
+ BB->addSuccessor(loopMBB);
+ BB->addSuccessor(exitMBB);
+
+ // exitMBB:
+ // and tmp10,oldval,mask
+ // srl tmp11,tmp10,shift
+ // sll tmp12,tmp11,24
+ // sra dest,tmp12,24
+ BB = exitMBB;
+ int64_t ShiftImm = (Size == 1) ? 24 : 16;
+ // reverse order
+ BuildMI(*BB, BB->begin(), dl, TII->get(Mips::SRA), Dest)
+ .addReg(Tmp12).addImm(ShiftImm);
+ BuildMI(*BB, BB->begin(), dl, TII->get(Mips::SLL), Tmp12)
+ .addReg(Tmp11).addImm(ShiftImm);
+ BuildMI(*BB, BB->begin(), dl, TII->get(Mips::SRL), Tmp11)
+ .addReg(Tmp10).addReg(Shift);
+ BuildMI(*BB, BB->begin(), dl, TII->get(Mips::AND), Tmp10)
+ .addReg(Oldval).addReg(Mask);
+
+ MI->eraseFromParent(); // The instruction is gone now.
+
+ return BB;
}
+MachineBasicBlock *
+MipsTargetLowering::EmitAtomicCmpSwap(MachineInstr *MI,
+ MachineBasicBlock *BB,
+ unsigned Size) const {
+ assert(Size == 4 && "Unsupported size for EmitAtomicCmpSwap.");
+
+ MachineFunction *MF = BB->getParent();
+ MachineRegisterInfo &RegInfo = MF->getRegInfo();
+ const TargetRegisterClass *RC = getRegClassFor(MVT::i32);
+ const TargetInstrInfo *TII = getTargetMachine().getInstrInfo();
+ DebugLoc dl = MI->getDebugLoc();
+
+ unsigned Dest = MI->getOperand(0).getReg();
+ unsigned Ptr = MI->getOperand(1).getReg();
+ unsigned Oldval = MI->getOperand(2).getReg();
+ unsigned Newval = MI->getOperand(3).getReg();
+
+ unsigned Tmp1 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp2 = RegInfo.createVirtualRegister(RC);
+
+ // insert new blocks after the current block
+ const BasicBlock *LLVM_BB = BB->getBasicBlock();
+ MachineBasicBlock *loop1MBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineBasicBlock *loop2MBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineBasicBlock *exitMBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineFunction::iterator It = BB;
+ ++It;
+ MF->insert(It, loop1MBB);
+ MF->insert(It, loop2MBB);
+ MF->insert(It, exitMBB);
+
+ // Transfer the remainder of BB and its successor edges to exitMBB.
+ exitMBB->splice(exitMBB->begin(), BB,
+ llvm::next(MachineBasicBlock::iterator(MI)),
+ BB->end());
+ exitMBB->transferSuccessorsAndUpdatePHIs(BB);
+
+ // Get or create a temporary stack location.
+ MipsFunctionInfo *MipsFI = MF->getInfo<MipsFunctionInfo>();
+ int fi = MipsFI->getAtomicFrameIndex();
+ if (fi == -1) {
+ fi = MF->getFrameInfo()->CreateStackObject(Size, Size, false);
+ MipsFI->setAtomicFrameIndex(fi);
+ }
+
+ // thisMBB:
+ // ...
+ // sw newval, fi(sp) // store newval to stack
+ // fallthrough --> loop1MBB
+
+ // Note: storing newval to stack before the loop and then loading it from
+ // stack in block loop2MBB is necessary to prevent MachineLICM pass to
+ // hoist "or" instruction out of the block loop2MBB.
+
+ BuildMI(BB, dl, TII->get(Mips::SW))
+ .addReg(Newval).addImm(0).addFrameIndex(fi);
+ BB->addSuccessor(loop1MBB);
+
+ // loop1MBB:
+ // ll dest, 0(ptr)
+ // bne dest, oldval, exitMBB
+ BB = loop1MBB;
+ BuildMI(BB, dl, TII->get(Mips::LL), Dest).addImm(0).addReg(Ptr);
+ BuildMI(BB, dl, TII->get(Mips::BNE))
+ .addReg(Dest).addReg(Oldval).addMBB(exitMBB);
+ BB->addSuccessor(exitMBB);
+ BB->addSuccessor(loop2MBB);
+
+ // loop2MBB:
+ // lw tmp2, fi(sp) // load newval from stack
+ // or tmp1, $0, tmp2
+ // sc tmp1, 0(ptr)
+ // beq tmp1, $0, loop1MBB
+ BB = loop2MBB;
+ BuildMI(BB, dl, TII->get(Mips::LW), Tmp2).addImm(0).addFrameIndex(fi);;
+ BuildMI(BB, dl, TII->get(Mips::OR), Tmp1).addReg(Mips::ZERO).addReg(Tmp2);
+ BuildMI(BB, dl, TII->get(Mips::SC), Tmp1).addReg(Tmp1).addImm(0).addReg(Ptr);
+ BuildMI(BB, dl, TII->get(Mips::BEQ))
+ .addReg(Tmp1).addReg(Mips::ZERO).addMBB(loop1MBB);
+ BB->addSuccessor(loop1MBB);
+ BB->addSuccessor(exitMBB);
+
+ MI->eraseFromParent(); // The instruction is gone now.
+
+ return BB;
+}
+
+MachineBasicBlock *
+MipsTargetLowering::EmitAtomicCmpSwapPartword(MachineInstr *MI,
+ MachineBasicBlock *BB,
+ unsigned Size) const {
+ assert((Size == 1 || Size == 2) &&
+ "Unsupported size for EmitAtomicCmpSwapPartial.");
+
+ MachineFunction *MF = BB->getParent();
+ MachineRegisterInfo &RegInfo = MF->getRegInfo();
+ const TargetRegisterClass *RC = getRegClassFor(MVT::i32);
+ const TargetInstrInfo *TII = getTargetMachine().getInstrInfo();
+ DebugLoc dl = MI->getDebugLoc();
+
+ unsigned Dest = MI->getOperand(0).getReg();
+ unsigned Ptr = MI->getOperand(1).getReg();
+ unsigned Oldval = MI->getOperand(2).getReg();
+ unsigned Newval = MI->getOperand(3).getReg();
+
+ unsigned Addr = RegInfo.createVirtualRegister(RC);
+ unsigned Shift = RegInfo.createVirtualRegister(RC);
+ unsigned Mask = RegInfo.createVirtualRegister(RC);
+ unsigned Mask2 = RegInfo.createVirtualRegister(RC);
+ unsigned Oldval2 = RegInfo.createVirtualRegister(RC);
+ unsigned Oldval3 = RegInfo.createVirtualRegister(RC);
+ unsigned Oldval4 = RegInfo.createVirtualRegister(RC);
+ unsigned Newval2 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp1 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp2 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp3 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp4 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp5 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp6 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp7 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp8 = RegInfo.createVirtualRegister(RC);
+ unsigned Tmp9 = RegInfo.createVirtualRegister(RC);
+
+ // insert new blocks after the current block
+ const BasicBlock *LLVM_BB = BB->getBasicBlock();
+ MachineBasicBlock *loop1MBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineBasicBlock *loop2MBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineBasicBlock *exitMBB = MF->CreateMachineBasicBlock(LLVM_BB);
+ MachineFunction::iterator It = BB;
+ ++It;
+ MF->insert(It, loop1MBB);
+ MF->insert(It, loop2MBB);
+ MF->insert(It, exitMBB);
+
+ // Transfer the remainder of BB and its successor edges to exitMBB.
+ exitMBB->splice(exitMBB->begin(), BB,
+ llvm::next(MachineBasicBlock::iterator(MI)),
+ BB->end());
+ exitMBB->transferSuccessorsAndUpdatePHIs(BB);
+
+ // thisMBB:
+ // addiu tmp1,$0,-4 # 0xfffffffc
+ // and addr,ptr,tmp1
+ // andi tmp2,ptr,3
+ // sll shift,tmp2,3
+ // ori tmp3,$0,255 # 0xff
+ // sll mask,tmp3,shift
+ // nor mask2,$0,mask
+ // andi tmp4,oldval,255
+ // sll oldval2,tmp4,shift
+ // andi tmp5,newval,255
+ // sll newval2,tmp5,shift
+ int64_t MaskImm = (Size == 1) ? 255 : 65535;
+ BuildMI(BB, dl, TII->get(Mips::ADDiu), Tmp1).addReg(Mips::ZERO).addImm(-4);
+ BuildMI(BB, dl, TII->get(Mips::AND), Addr).addReg(Ptr).addReg(Tmp1);
+ BuildMI(BB, dl, TII->get(Mips::ANDi), Tmp2).addReg(Ptr).addImm(3);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Shift).addReg(Tmp2).addImm(3);
+ BuildMI(BB, dl, TII->get(Mips::ORi), Tmp3).addReg(Mips::ZERO).addImm(MaskImm);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Mask).addReg(Tmp3).addReg(Shift);
+ BuildMI(BB, dl, TII->get(Mips::NOR), Mask2).addReg(Mips::ZERO).addReg(Mask);
+ BuildMI(BB, dl, TII->get(Mips::ANDi), Tmp4).addReg(Oldval).addImm(MaskImm);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Oldval2).addReg(Tmp4).addReg(Shift);
+ BuildMI(BB, dl, TII->get(Mips::ANDi), Tmp5).addReg(Newval).addImm(MaskImm);
+ BuildMI(BB, dl, TII->get(Mips::SLL), Newval2).addReg(Tmp5).addReg(Shift);
+ BB->addSuccessor(loop1MBB);
+
+ // loop1MBB:
+ // ll oldval3,0(addr)
+ // and oldval4,oldval3,mask
+ // bne oldval4,oldval2,exitMBB
+ BB = loop1MBB;
+ BuildMI(BB, dl, TII->get(Mips::LL), Oldval3).addImm(0).addReg(Addr);
+ BuildMI(BB, dl, TII->get(Mips::AND), Oldval4).addReg(Oldval3).addReg(Mask);
+ BuildMI(BB, dl, TII->get(Mips::BNE))
+ .addReg(Oldval4).addReg(Oldval2).addMBB(exitMBB);
+ BB->addSuccessor(exitMBB);
+ BB->addSuccessor(loop2MBB);
+
+ // loop2MBB:
+ // and tmp6,oldval3,mask2
+ // or tmp7,tmp6,newval2
+ // sc tmp7,0(addr)
+ // beq tmp7,$0,loop1MBB
+ BB = loop2MBB;
+ BuildMI(BB, dl, TII->get(Mips::AND), Tmp6).addReg(Oldval3).addReg(Mask2);
+ BuildMI(BB, dl, TII->get(Mips::OR), Tmp7).addReg(Tmp6).addReg(Newval2);
+ BuildMI(BB, dl, TII->get(Mips::SC), Tmp7)
+ .addReg(Tmp7).addImm(0).addReg(Addr);
+ BuildMI(BB, dl, TII->get(Mips::BEQ))
+ .addReg(Tmp7).addReg(Mips::ZERO).addMBB(loop1MBB);
+ BB->addSuccessor(loop1MBB);
+ BB->addSuccessor(exitMBB);
+
+ // exitMBB:
+ // srl tmp8,oldval4,shift
+ // sll tmp9,tmp8,24
+ // sra dest,tmp9,24
+ BB = exitMBB;
+ int64_t ShiftImm = (Size == 1) ? 24 : 16;
+ // reverse order
+ BuildMI(*BB, BB->begin(), dl, TII->get(Mips::SRA), Dest)
+ .addReg(Tmp9).addImm(ShiftImm);
+ BuildMI(*BB, BB->begin(), dl, TII->get(Mips::SLL), Tmp9)
+ .addReg(Tmp8).addImm(ShiftImm);
+ BuildMI(*BB, BB->begin(), dl, TII->get(Mips::SRL), Tmp8)
+ .addReg(Oldval4).addReg(Shift);
+
+ MI->eraseFromParent(); // The instruction is gone now.
+
+ return BB;
+}
+
+//===----------------------------------------------------------------------===//
+// Misc Lower Operation implementation
+//===----------------------------------------------------------------------===//
SDValue MipsTargetLowering::
LowerDYNAMIC_STACKALLOC(SDValue Op, SelectionDAG &DAG) const
{
// clobber $sp from being inserted between copytoreg and copyfromreg.
SDValue NewSP = DAG.getCopyFromReg(Chain, dl, Mips::SP, MVT::i32,
Chain.getValue(1));
-
+
// The stack space reserved by alloca is located right above the argument
// area. It is aligned on a boundary that is a multiple of StackAlignment.
MachineFunction &MF = DAG.getMachineFunction();
SDValue HiPart = DAG.getNode(MipsISD::Hi, dl, VTs, &GAHi, 1);
SDValue Lo = DAG.getNode(MipsISD::Lo, dl, MVT::i32, GALo);
return DAG.getNode(ISD::ADD, dl, MVT::i32, HiPart, Lo);
- } else {
- SDValue GA = DAG.getTargetGlobalAddress(GV, dl, MVT::i32, 0,
- MipsII::MO_GOT);
- SDValue ResNode = DAG.getLoad(MVT::i32, dl,
- DAG.getEntryNode(), GA, MachinePointerInfo(),
- false, false, 0);
- // On functions and global targets not internal linked only
- // a load from got/GP is necessary for PIC to work.
- if (!GV->hasInternalLinkage() &&
- (!GV->hasLocalLinkage() || isa<Function>(GV)))
- return ResNode;
- SDValue GALo = DAG.getTargetGlobalAddress(GV, dl, MVT::i32, 0,
- MipsII::MO_ABS_LO);
- SDValue Lo = DAG.getNode(MipsISD::Lo, dl, MVT::i32, GALo);
- return DAG.getNode(ISD::ADD, dl, MVT::i32, ResNode, Lo);
}
- llvm_unreachable("Dont know how to handle GlobalAddress");
- return SDValue(0,0);
+ SDValue GA = DAG.getTargetGlobalAddress(GV, dl, MVT::i32, 0,
+ MipsII::MO_GOT);
+ GA = DAG.getNode(MipsISD::WrapperPIC, dl, MVT::i32, GA);
+ SDValue ResNode = DAG.getLoad(MVT::i32, dl,
+ DAG.getEntryNode(), GA, MachinePointerInfo(),
+ false, false, 0);
+ // On functions and global targets not internal linked only
+ // a load from got/GP is necessary for PIC to work.
+ if (!GV->hasInternalLinkage() &&
+ (!GV->hasLocalLinkage() || isa<Function>(GV)))
+ return ResNode;
+ SDValue GALo = DAG.getTargetGlobalAddress(GV, dl, MVT::i32, 0,
+ MipsII::MO_ABS_LO);
+ SDValue Lo = DAG.getNode(MipsISD::Lo, dl, MVT::i32, GALo);
+ return DAG.getNode(ISD::ADD, dl, MVT::i32, ResNode, Lo);
}
SDValue MipsTargetLowering::LowerBlockAddress(SDValue Op,
SDValue BAGOTOffset = DAG.getBlockAddress(BA, MVT::i32, true,
MipsII::MO_GOT);
+ BAGOTOffset = DAG.getNode(MipsISD::WrapperPIC, dl, MVT::i32, BAGOTOffset);
SDValue BALOOffset = DAG.getBlockAddress(BA, MVT::i32, true,
MipsII::MO_ABS_LO);
SDValue Load = DAG.getLoad(MVT::i32, dl,
SDValue MipsTargetLowering::
LowerGlobalTLSAddress(SDValue Op, SelectionDAG &DAG) const
{
- llvm_unreachable("TLS not implemented for MIPS.");
- return SDValue(); // Not reached
+ // If the relocation model is PIC, use the General Dynamic TLS Model,
+ // otherwise use the Initial Exec or Local Exec TLS Model.
+ // TODO: implement Local Dynamic TLS model
+
+ GlobalAddressSDNode *GA = cast<GlobalAddressSDNode>(Op);
+ DebugLoc dl = GA->getDebugLoc();
+ const GlobalValue *GV = GA->getGlobal();
+ EVT PtrVT = getPointerTy();
+
+ if (getTargetMachine().getRelocationModel() == Reloc::PIC_) {
+ // General Dynamic TLS Model
+ SDValue TGA = DAG.getTargetGlobalAddress(GV, dl, MVT::i32,
+ 0, MipsII::MO_TLSGD);
+ SDValue Tlsgd = DAG.getNode(MipsISD::TlsGd, dl, MVT::i32, TGA);
+ SDValue GP = DAG.getRegister(Mips::GP, MVT::i32);
+ SDValue Argument = DAG.getNode(ISD::ADD, dl, MVT::i32, GP, Tlsgd);
+
+ ArgListTy Args;
+ ArgListEntry Entry;
+ Entry.Node = Argument;
+ Entry.Ty = (const Type *) Type::getInt32Ty(*DAG.getContext());
+ Args.push_back(Entry);
+ std::pair<SDValue, SDValue> CallResult =
+ LowerCallTo(DAG.getEntryNode(),
+ (const Type *) Type::getInt32Ty(*DAG.getContext()),
+ false, false, false, false,
+ 0, CallingConv::C, false, true,
+ DAG.getExternalSymbol("__tls_get_addr", PtrVT), Args, DAG, dl);
+
+ return CallResult.first;
+ } else {
+ SDValue Offset;
+ if (GV->isDeclaration()) {
+ // Initial Exec TLS Model
+ SDValue TGA = DAG.getTargetGlobalAddress(GV, dl, MVT::i32, 0,
+ MipsII::MO_GOTTPREL);
+ Offset = DAG.getLoad(MVT::i32, dl,
+ DAG.getEntryNode(), TGA, MachinePointerInfo(),
+ false, false, 0);
+ } else {
+ // Local Exec TLS Model
+ SDVTList VTs = DAG.getVTList(MVT::i32);
+ SDValue TGAHi = DAG.getTargetGlobalAddress(GV, dl, MVT::i32, 0,
+ MipsII::MO_TPREL_HI);
+ SDValue TGALo = DAG.getTargetGlobalAddress(GV, dl, MVT::i32, 0,
+ MipsII::MO_TPREL_LO);
+ SDValue Hi = DAG.getNode(MipsISD::TprelHi, dl, VTs, &TGAHi, 1);
+ SDValue Lo = DAG.getNode(MipsISD::TprelLo, dl, MVT::i32, TGALo);
+ Offset = DAG.getNode(ISD::ADD, dl, MVT::i32, Hi, Lo);
+ }
+
+ SDValue ThreadPointer = DAG.getNode(MipsISD::ThreadPointer, dl, PtrVT);
+ return DAG.getNode(ISD::ADD, dl, PtrVT, ThreadPointer, Offset);
+ }
}
SDValue MipsTargetLowering::
if (!IsPIC) {
SDValue Ops[] = { JTI };
HiPart = DAG.getNode(MipsISD::Hi, dl, DAG.getVTList(MVT::i32), Ops, 1);
- } else // Emit Load from Global Pointer
+ } else {// Emit Load from Global Pointer
+ JTI = DAG.getNode(MipsISD::WrapperPIC, dl, MVT::i32, JTI);
HiPart = DAG.getLoad(MVT::i32, dl, DAG.getEntryNode(), JTI,
MachinePointerInfo(),
false, false, 0);
+ }
SDValue JTILo = DAG.getTargetJumpTable(JT->getIndex(), PtrVT,
MipsII::MO_ABS_LO);
} else {
SDValue CP = DAG.getTargetConstantPool(C, MVT::i32, N->getAlignment(),
N->getOffset(), MipsII::MO_GOT);
+ CP = DAG.getNode(MipsISD::WrapperPIC, dl, MVT::i32, CP);
SDValue Load = DAG.getLoad(MVT::i32, dl, DAG.getEntryNode(),
CP, MachinePointerInfo::getConstantPool(),
false, false, 0);
}
static SDValue LowerFCOPYSIGN64(SDValue Op, SelectionDAG &DAG, bool isLittle) {
- // FIXME:
+ // FIXME:
// Use ext/ins instructions if target architecture is Mips32r2.
// Eliminate redundant mfc1 and mtc1 instructions.
unsigned LoIdx = 0, HiIdx = 1;
-
+
if (!isLittle)
std::swap(LoIdx, HiIdx);
return LowerFCOPYSIGN64(Op, DAG, Subtarget->isLittle());
}
+SDValue MipsTargetLowering::
+LowerFRAMEADDR(SDValue Op, SelectionDAG &DAG) const {
+ // check the depth
+ assert((cast<ConstantSDNode>(Op.getOperand(0))->getZExtValue() == 0) &&
+ "Frame address can only be determined for current frame.");
+
+ MachineFrameInfo *MFI = DAG.getMachineFunction().getFrameInfo();
+ MFI->setFrameAddressIsTaken(true);
+ EVT VT = Op.getValueType();
+ DebugLoc dl = Op.getDebugLoc();
+ SDValue FrameAddr = DAG.getCopyFromReg(DAG.getEntryNode(), dl, Mips::FP, VT);
+ return FrameAddr;
+}
+
//===----------------------------------------------------------------------===//
// Calling Convention Implementation
//===----------------------------------------------------------------------===//
// copy remaining part of byval arg to stack.
if (CurWord < LastWord) {
- unsigned SizeInBytes = (LastWord - CurWord) * 4;
+ unsigned SizeInBytes = (LastWord - CurWord) * 4;
SDValue Src = DAG.getNode(ISD::ADD, dl, MVT::i32, Arg,
DAG.getConstant((CurWord - FirstWord) * 4,
MVT::i32));
// Analyze operands of the call, assigning locations to each operand.
SmallVector<CCValAssign, 16> ArgLocs;
- CCState CCInfo(CallConv, isVarArg, getTargetMachine(), ArgLocs,
- *DAG.getContext());
+ CCState CCInfo(CallConv, isVarArg, DAG.getMachineFunction(),
+ getTargetMachine(), ArgLocs, *DAG.getContext());
if (Subtarget->isABI_O32())
CCInfo.AnalyzeCallOperands(Outs, CC_MipsO32);
CCInfo.AnalyzeCallOperands(Outs, CC_Mips);
// Get a count of how many bytes are to be pushed on the stack.
- unsigned NumBytes = CCInfo.getNextStackOffset();
- Chain = DAG.getCALLSEQ_START(Chain, DAG.getIntPtrConstant(NumBytes, true));
-
- // With EABI is it possible to have 16 args on registers.
- SmallVector<std::pair<unsigned, SDValue>, 16> RegsToPass;
- SmallVector<SDValue, 8> MemOpChains;
+ unsigned NextStackOffset = CCInfo.getNextStackOffset();
- MipsFI->setHasCall();
+ Chain = DAG.getCALLSEQ_START(Chain, DAG.getIntPtrConstant(NextStackOffset,
+ true));
// If this is the first call, create a stack frame object that points to
- // a location to which .cprestore saves $gp. The offset of this frame object
- // is set to 0, since we know nothing about the size of the argument area at
- // this point.
+ // a location to which .cprestore saves $gp.
if (IsPIC && !MipsFI->getGPFI())
MipsFI->setGPFI(MFI->CreateFixedObject(4, 0, true));
- int FirstFI = -MFI->getNumFixedObjects() - 1, LastFI = 0;
+ // Update size of the maximum argument space.
+ // For O32, a minimum of four words (16 bytes) of argument space is
+ // allocated.
+ if (Subtarget->isABI_O32())
+ NextStackOffset = std::max(NextStackOffset, (unsigned)16);
+
+ unsigned MaxCallFrameSize = MipsFI->getMaxCallFrameSize();
+
+ if (MaxCallFrameSize < NextStackOffset) {
+ MipsFI->setMaxCallFrameSize(NextStackOffset);
+
+ if (IsPIC) {
+ // $gp restore slot must be aligned.
+ unsigned StackAlignment = TFL->getStackAlignment();
+ NextStackOffset = (NextStackOffset + StackAlignment - 1) /
+ StackAlignment * StackAlignment;
+ int GPFI = MipsFI->getGPFI();
+ MFI->setObjectOffset(GPFI, NextStackOffset);
+ }
+ }
+
+ // With EABI is it possible to have 16 args on registers.
+ SmallVector<std::pair<unsigned, SDValue>, 16> RegsToPass;
+ SmallVector<SDValue, 8> MemOpChains;
+
+ int FirstFI = -MFI->getNumFixedObjects() - 1, LastFI = 0;
// Walk the register/memloc assignments, inserting copies/loads.
for (unsigned i = 0, e = ArgLocs.size(); i != e; ++i) {
// Register can't get to this point...
assert(VA.isMemLoc());
- // ByVal Arg.
+ // ByVal Arg.
ISD::ArgFlagsTy Flags = Outs[i].Flags;
if (Flags.isByVal()) {
assert(Subtarget->isABI_O32() &&
}
// Create the frame index object for this incoming parameter
- LastFI = MFI->CreateFixedObject(VA.getValVT().getSizeInBits()/8,
+ LastFI = MFI->CreateFixedObject(VA.getValVT().getSizeInBits()/8,
VA.getLocMemOffset(), true);
SDValue PtrOff = DAG.getFrameIndex(LastFI, getPointerTy());
false, false, 0));
}
+ // Extend range of indices of frame objects for outgoing arguments that were
+ // created during this function call. Skip this step if no such objects were
+ // created.
+ if (LastFI)
+ MipsFI->extendOutArgFIRange(FirstFI, LastFI);
+
// Transform all store nodes into one single node because all store
// nodes are independent of each other.
if (!MemOpChains.empty())
if (IsPIC) {
if (LoadSymAddr) {
// Load callee address
+ Callee = DAG.getNode(MipsISD::WrapperPIC, dl, MVT::i32, Callee);
SDValue LoadValue = DAG.getLoad(MVT::i32, dl, Chain, Callee,
MachinePointerInfo::getGOT(),
false, false, 0);
} else
Callee = LoadValue;
- // Use chain output from LoadValue
+ // Use chain output from LoadValue
Chain = LoadValue.getValue(1);
}
Chain = DAG.getNode(MipsISD::JmpLink, dl, NodeTys, &Ops[0], Ops.size());
InFlag = Chain.getValue(1);
- // Function can have an arbitrary number of calls, so
- // hold the LastArgStackLoc with the biggest offset.
- unsigned MaxCallFrameSize = MipsFI->getMaxCallFrameSize();
- unsigned NextStackOffset = CCInfo.getNextStackOffset();
-
- // For O32, a minimum of four words (16 bytes) of argument space is
- // allocated.
- if (Subtarget->isABI_O32())
- NextStackOffset = std::max(NextStackOffset, (unsigned)16);
-
- if (MaxCallFrameSize < NextStackOffset) {
- MipsFI->setMaxCallFrameSize(NextStackOffset);
-
- if (IsPIC) {
- // $gp restore slot must be aligned.
- unsigned StackAlignment = TFL->getStackAlignment();
- NextStackOffset = (NextStackOffset + StackAlignment - 1) /
- StackAlignment * StackAlignment;
- int GPFI = MipsFI->getGPFI();
- MFI->setObjectOffset(GPFI, NextStackOffset);
- }
- }
-
- // Extend range of indices of frame objects for outgoing arguments that were
- // created during this function call. Skip this step if no such objects were
- // created.
- if (LastFI)
- MipsFI->extendOutArgFIRange(FirstFI, LastFI);
-
// Create the CALLSEQ_END node.
- Chain = DAG.getCALLSEQ_END(Chain, DAG.getIntPtrConstant(NumBytes, true),
+ Chain = DAG.getCALLSEQ_END(Chain, DAG.getIntPtrConstant(NextStackOffset, true),
DAG.getIntPtrConstant(0, true), InFlag);
InFlag = Chain.getValue(1);
const SmallVectorImpl<ISD::InputArg> &Ins,
DebugLoc dl, SelectionDAG &DAG,
SmallVectorImpl<SDValue> &InVals) const {
-
// Assign locations to each value returned by this call.
SmallVector<CCValAssign, 16> RVLocs;
- CCState CCInfo(CallConv, isVarArg, getTargetMachine(),
- RVLocs, *DAG.getContext());
+ CCState CCInfo(CallConv, isVarArg, DAG.getMachineFunction(),
+ getTargetMachine(), RVLocs, *DAG.getContext());
CCInfo.AnalyzeCallResult(Ins, RetCC_Mips);
// Assign locations to all of the incoming arguments.
SmallVector<CCValAssign, 16> ArgLocs;
- CCState CCInfo(CallConv, isVarArg, getTargetMachine(),
- ArgLocs, *DAG.getContext());
+ CCState CCInfo(CallConv, isVarArg, DAG.getMachineFunction(),
+ getTargetMachine(), ArgLocs, *DAG.getContext());
if (Subtarget->isABI_O32())
CCInfo.AnalyzeFormalArguments(Ins, CC_MipsO32);
if (isVarArg && Subtarget->isABI_O32()) {
// Record the frame index of the first variable argument
- // which is a value necessary to VASTART.
+ // which is a value necessary to VASTART.
unsigned NextStackOffset = CCInfo.getNextStackOffset();
assert(NextStackOffset % 4 == 0 &&
"NextStackOffset must be aligned to 4-byte boundaries.");
SmallVector<CCValAssign, 16> RVLocs;
// CCState - Info about the registers and stack slot.
- CCState CCInfo(CallConv, isVarArg, getTargetMachine(),
- RVLocs, *DAG.getContext());
+ CCState CCInfo(CallConv, isVarArg, DAG.getMachineFunction(),
+ getTargetMachine(), RVLocs, *DAG.getContext());
// Analize return values.
CCInfo.AnalyzeReturn(Outs, RetCC_Mips);