llvm.org GIT mirror llvm / f726aa1 lib / Target / X86 / X86ExpandPseudo.cpp
f726aa1

Tree @f726aa1 (Download .tar.gz)

X86ExpandPseudo.cpp @f726aa1raw · history · blame

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
//===------- X86ExpandPseudo.cpp - Expand pseudo instructions -------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This file contains a pass that expands pseudo instructions into target
// instructions to allow proper scheduling, if-conversion, other late
// optimizations, or simply the encoding of the instructions.
//
//===----------------------------------------------------------------------===//

#include "X86.h"
#include "X86FrameLowering.h"
#include "X86InstrBuilder.h"
#include "X86InstrInfo.h"
#include "X86MachineFunctionInfo.h"
#include "X86Subtarget.h"
#include "llvm/Analysis/EHPersonalities.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/Passes.h" // For IDs of passes that are preserved.
#include "llvm/IR/GlobalValue.h"
using namespace llvm;

#define DEBUG_TYPE "x86-pseudo"
#define X86_EXPAND_PSEUDO_NAME "X86 pseudo instruction expansion pass"

namespace {
class X86ExpandPseudo : public MachineFunctionPass {
public:
  static char ID;
  X86ExpandPseudo() : MachineFunctionPass(ID) {}

  void getAnalysisUsage(AnalysisUsage &AU) const override {
    AU.setPreservesCFG();
    AU.addPreservedID(MachineLoopInfoID);
    AU.addPreservedID(MachineDominatorsID);
    MachineFunctionPass::getAnalysisUsage(AU);
  }

  const X86Subtarget *STI;
  const X86InstrInfo *TII;
  const X86RegisterInfo *TRI;
  const X86MachineFunctionInfo *X86FI;
  const X86FrameLowering *X86FL;

  bool runOnMachineFunction(MachineFunction &Fn) override;

  MachineFunctionProperties getRequiredProperties() const override {
    return MachineFunctionProperties().set(
        MachineFunctionProperties::Property::NoVRegs);
  }

  StringRef getPassName() const override {
    return "X86 pseudo instruction expansion pass";
  }

private:
  void ExpandICallBranchFunnel(MachineBasicBlock *MBB,
                               MachineBasicBlock::iterator MBBI);

  bool ExpandMI(MachineBasicBlock &MBB, MachineBasicBlock::iterator MBBI);
  bool ExpandMBB(MachineBasicBlock &MBB);
};
char X86ExpandPseudo::ID = 0;

} // End anonymous namespace.

INITIALIZE_PASS(X86ExpandPseudo, DEBUG_TYPE, X86_EXPAND_PSEUDO_NAME, false,
                false)

void X86ExpandPseudo::ExpandICallBranchFunnel(
    MachineBasicBlock *MBB, MachineBasicBlock::iterator MBBI) {
  MachineBasicBlock *JTMBB = MBB;
  MachineInstr *JTInst = &*MBBI;
  MachineFunction *MF = MBB->getParent();
  const BasicBlock *BB = MBB->getBasicBlock();
  auto InsPt = MachineFunction::iterator(MBB);
  ++InsPt;

  std::vector<std::pair<MachineBasicBlock *, unsigned>> TargetMBBs;
  DebugLoc DL = JTInst->getDebugLoc();
  MachineOperand Selector = JTInst->getOperand(0);
  const GlobalValue *CombinedGlobal = JTInst->getOperand(1).getGlobal();

  auto CmpTarget = [&](unsigned Target) {
    if (Selector.isReg())
      MBB->addLiveIn(Selector.getReg());
    BuildMI(*MBB, MBBI, DL, TII->get(X86::LEA64r), X86::R11)
        .addReg(X86::RIP)
        .addImm(1)
        .addReg(0)
        .addGlobalAddress(CombinedGlobal,
                          JTInst->getOperand(2 + 2 * Target).getImm())
        .addReg(0);
    BuildMI(*MBB, MBBI, DL, TII->get(X86::CMP64rr))
        .add(Selector)
        .addReg(X86::R11);
  };

  auto CreateMBB = [&]() {
    auto *NewMBB = MF->CreateMachineBasicBlock(BB);
    MBB->addSuccessor(NewMBB);
    if (!MBB->isLiveIn(X86::EFLAGS))
      MBB->addLiveIn(X86::EFLAGS);
    return NewMBB;
  };

  auto EmitCondJump = [&](unsigned CC, MachineBasicBlock *ThenMBB) {
    BuildMI(*MBB, MBBI, DL, TII->get(X86::JCC_1)).addMBB(ThenMBB).addImm(CC);

    auto *ElseMBB = CreateMBB();
    MF->insert(InsPt, ElseMBB);
    MBB = ElseMBB;
    MBBI = MBB->end();
  };

  auto EmitCondJumpTarget = [&](unsigned CC, unsigned Target) {
    auto *ThenMBB = CreateMBB();
    TargetMBBs.push_back({ThenMBB, Target});
    EmitCondJump(CC, ThenMBB);
  };

  auto EmitTailCall = [&](unsigned Target) {
    BuildMI(*MBB, MBBI, DL, TII->get(X86::TAILJMPd64))
        .add(JTInst->getOperand(3 + 2 * Target));
  };

  std::function<void(unsigned, unsigned)> EmitBranchFunnel =
      [&](unsigned FirstTarget, unsigned NumTargets) {
    if (NumTargets == 1) {
      EmitTailCall(FirstTarget);
      return;
    }

    if (NumTargets == 2) {
      CmpTarget(FirstTarget + 1);
      EmitCondJumpTarget(X86::COND_B, FirstTarget);
      EmitTailCall(FirstTarget + 1);
      return;
    }

    if (NumTargets < 6) {
      CmpTarget(FirstTarget + 1);
      EmitCondJumpTarget(X86::COND_B, FirstTarget);
      EmitCondJumpTarget(X86::COND_E, FirstTarget + 1);
      EmitBranchFunnel(FirstTarget + 2, NumTargets - 2);
      return;
    }

    auto *ThenMBB = CreateMBB();
    CmpTarget(FirstTarget + (NumTargets / 2));
    EmitCondJump(X86::COND_B, ThenMBB);
    EmitCondJumpTarget(X86::COND_E, FirstTarget + (NumTargets / 2));
    EmitBranchFunnel(FirstTarget + (NumTargets / 2) + 1,
                  NumTargets - (NumTargets / 2) - 1);

    MF->insert(InsPt, ThenMBB);
    MBB = ThenMBB;
    MBBI = MBB->end();
    EmitBranchFunnel(FirstTarget, NumTargets / 2);
  };

  EmitBranchFunnel(0, (JTInst->getNumOperands() - 2) / 2);
  for (auto P : TargetMBBs) {
    MF->insert(InsPt, P.first);
    BuildMI(P.first, DL, TII->get(X86::TAILJMPd64))
        .add(JTInst->getOperand(3 + 2 * P.second));
  }
  JTMBB->erase(JTInst);
}

/// If \p MBBI is a pseudo instruction, this method expands
/// it to the corresponding (sequence of) actual instruction(s).
/// \returns true if \p MBBI has been expanded.
bool X86ExpandPseudo::ExpandMI(MachineBasicBlock &MBB,
                               MachineBasicBlock::iterator MBBI) {
  MachineInstr &MI = *MBBI;
  unsigned Opcode = MI.getOpcode();
  DebugLoc DL = MBBI->getDebugLoc();
  switch (Opcode) {
  default:
    return false;
  case X86::TCRETURNdi:
  case X86::TCRETURNdicc:
  case X86::TCRETURNri:
  case X86::TCRETURNmi:
  case X86::TCRETURNdi64:
  case X86::TCRETURNdi64cc:
  case X86::TCRETURNri64:
  case X86::TCRETURNmi64: {
    bool isMem = Opcode == X86::TCRETURNmi || Opcode == X86::TCRETURNmi64;
    MachineOperand &JumpTarget = MBBI->getOperand(0);
    MachineOperand &StackAdjust = MBBI->getOperand(isMem ? X86::AddrNumOperands
                                                         : 1);
    assert(StackAdjust.isImm() && "Expecting immediate value.");

    // Adjust stack pointer.
    int StackAdj = StackAdjust.getImm();
    int MaxTCDelta = X86FI->getTCReturnAddrDelta();
    int Offset = 0;
    assert(MaxTCDelta <= 0 && "MaxTCDelta should never be positive");

    // Incoporate the retaddr area.
    Offset = StackAdj - MaxTCDelta;
    assert(Offset >= 0 && "Offset should never be negative");

    if (Opcode == X86::TCRETURNdicc || Opcode == X86::TCRETURNdi64cc) {
      assert(Offset == 0 && "Conditional tail call cannot adjust the stack.");
    }

    if (Offset) {
      // Check for possible merge with preceding ADD instruction.
      Offset += X86FL->mergeSPUpdates(MBB, MBBI, true);
      X86FL->emitSPUpdate(MBB, MBBI, DL, Offset, /*InEpilogue=*/true);
    }

    // Jump to label or value in register.
    bool IsWin64 = STI->isTargetWin64();
    if (Opcode == X86::TCRETURNdi || Opcode == X86::TCRETURNdicc ||
        Opcode == X86::TCRETURNdi64 || Opcode == X86::TCRETURNdi64cc) {
      unsigned Op;
      switch (Opcode) {
      case X86::TCRETURNdi:
        Op = X86::TAILJMPd;
        break;
      case X86::TCRETURNdicc:
        Op = X86::TAILJMPd_CC;
        break;
      case X86::TCRETURNdi64cc:
        assert(!MBB.getParent()->hasWinCFI() &&
               "Conditional tail calls confuse "
               "the Win64 unwinder.");
        Op = X86::TAILJMPd64_CC;
        break;
      default:
        // Note: Win64 uses REX prefixes indirect jumps out of functions, but
        // not direct ones.
        Op = X86::TAILJMPd64;
        break;
      }
      MachineInstrBuilder MIB = BuildMI(MBB, MBBI, DL, TII->get(Op));
      if (JumpTarget.isGlobal()) {
        MIB.addGlobalAddress(JumpTarget.getGlobal(), JumpTarget.getOffset(),
                             JumpTarget.getTargetFlags());
      } else {
        assert(JumpTarget.isSymbol());
        MIB.addExternalSymbol(JumpTarget.getSymbolName(),
                              JumpTarget.getTargetFlags());
      }
      if (Op == X86::TAILJMPd_CC || Op == X86::TAILJMPd64_CC) {
        MIB.addImm(MBBI->getOperand(2).getImm());
      }

    } else if (Opcode == X86::TCRETURNmi || Opcode == X86::TCRETURNmi64) {
      unsigned Op = (Opcode == X86::TCRETURNmi)
                        ? X86::TAILJMPm
                        : (IsWin64 ? X86::TAILJMPm64_REX : X86::TAILJMPm64);
      MachineInstrBuilder MIB = BuildMI(MBB, MBBI, DL, TII->get(Op));
      for (unsigned i = 0; i != X86::AddrNumOperands; ++i)
        MIB.add(MBBI->getOperand(i));
    } else if (Opcode == X86::TCRETURNri64) {
      JumpTarget.setIsKill();
      BuildMI(MBB, MBBI, DL,
              TII->get(IsWin64 ? X86::TAILJMPr64_REX : X86::TAILJMPr64))
          .add(JumpTarget);
    } else {
      JumpTarget.setIsKill();
      BuildMI(MBB, MBBI, DL, TII->get(X86::TAILJMPr))
          .add(JumpTarget);
    }

    MachineInstr &NewMI = *std::prev(MBBI);
    NewMI.copyImplicitOps(*MBBI->getParent()->getParent(), *MBBI);
    MBB.getParent()->updateCallSiteInfo(&*MBBI, &NewMI);

    // Delete the pseudo instruction TCRETURN.
    MBB.erase(MBBI);

    return true;
  }
  case X86::EH_RETURN:
  case X86::EH_RETURN64: {
    MachineOperand &DestAddr = MBBI->getOperand(0);
    assert(DestAddr.isReg() && "Offset should be in register!");
    const bool Uses64BitFramePtr =
        STI->isTarget64BitLP64() || STI->isTargetNaCl64();
    Register StackPtr = TRI->getStackRegister();
    BuildMI(MBB, MBBI, DL,
            TII->get(Uses64BitFramePtr ? X86::MOV64rr : X86::MOV32rr), StackPtr)
        .addReg(DestAddr.getReg());
    // The EH_RETURN pseudo is really removed during the MC Lowering.
    return true;
  }
  case X86::IRET: {
    // Adjust stack to erase error code
    int64_t StackAdj = MBBI->getOperand(0).getImm();
    X86FL->emitSPUpdate(MBB, MBBI, DL, StackAdj, true);
    // Replace pseudo with machine iret
    BuildMI(MBB, MBBI, DL,
            TII->get(STI->is64Bit() ? X86::IRET64 : X86::IRET32));
    MBB.erase(MBBI);
    return true;
  }
  case X86::RET: {
    // Adjust stack to erase error code
    int64_t StackAdj = MBBI->getOperand(0).getImm();
    MachineInstrBuilder MIB;
    if (StackAdj == 0) {
      MIB = BuildMI(MBB, MBBI, DL,
                    TII->get(STI->is64Bit() ? X86::RETQ : X86::RETL));
    } else if (isUInt<16>(StackAdj)) {
      MIB = BuildMI(MBB, MBBI, DL,
                    TII->get(STI->is64Bit() ? X86::RETIQ : X86::RETIL))
                .addImm(StackAdj);
    } else {
      assert(!STI->is64Bit() &&
             "shouldn't need to do this for x86_64 targets!");
      // A ret can only handle immediates as big as 2**16-1.  If we need to pop
      // off bytes before the return address, we must do it manually.
      BuildMI(MBB, MBBI, DL, TII->get(X86::POP32r)).addReg(X86::ECX, RegState::Define);
      X86FL->emitSPUpdate(MBB, MBBI, DL, StackAdj, /*InEpilogue=*/true);
      BuildMI(MBB, MBBI, DL, TII->get(X86::PUSH32r)).addReg(X86::ECX);
      MIB = BuildMI(MBB, MBBI, DL, TII->get(X86::RETL));
    }
    for (unsigned I = 1, E = MBBI->getNumOperands(); I != E; ++I)
      MIB.add(MBBI->getOperand(I));
    MBB.erase(MBBI);
    return true;
  }
  case X86::EH_RESTORE: {
    // Restore ESP and EBP, and optionally ESI if required.
    bool IsSEH = isAsynchronousEHPersonality(classifyEHPersonality(
        MBB.getParent()->getFunction().getPersonalityFn()));
    X86FL->restoreWin32EHStackPointers(MBB, MBBI, DL, /*RestoreSP=*/IsSEH);
    MBBI->eraseFromParent();
    return true;
  }
  case X86::LCMPXCHG8B_SAVE_EBX:
  case X86::LCMPXCHG16B_SAVE_RBX: {
    // Perform the following transformation.
    // SaveRbx = pseudocmpxchg Addr, <4 opds for the address>, InArg, SaveRbx
    // =>
    // [E|R]BX = InArg
    // actualcmpxchg Addr
    // [E|R]BX = SaveRbx
    const MachineOperand &InArg = MBBI->getOperand(6);
    Register SaveRbx = MBBI->getOperand(7).getReg();

    unsigned ActualInArg =
        Opcode == X86::LCMPXCHG8B_SAVE_EBX ? X86::EBX : X86::RBX;
    // Copy the input argument of the pseudo into the argument of the
    // actual instruction.
    TII->copyPhysReg(MBB, MBBI, DL, ActualInArg, InArg.getReg(),
                     InArg.isKill());
    // Create the actual instruction.
    unsigned ActualOpc =
        Opcode == X86::LCMPXCHG8B_SAVE_EBX ? X86::LCMPXCHG8B : X86::LCMPXCHG16B;
    MachineInstr *NewInstr = BuildMI(MBB, MBBI, DL, TII->get(ActualOpc));
    // Copy the operands related to the address.
    for (unsigned Idx = 1; Idx < 6; ++Idx)
      NewInstr->addOperand(MBBI->getOperand(Idx));
    // Finally, restore the value of RBX.
    TII->copyPhysReg(MBB, MBBI, DL, ActualInArg, SaveRbx,
                     /*SrcIsKill*/ true);

    // Delete the pseudo.
    MBBI->eraseFromParent();
    return true;
  }
  case TargetOpcode::ICALL_BRANCH_FUNNEL:
    ExpandICallBranchFunnel(&MBB, MBBI);
    return true;
  }
  llvm_unreachable("Previous switch has a fallthrough?");
}

/// Expand all pseudo instructions contained in \p MBB.
/// \returns true if any expansion occurred for \p MBB.
bool X86ExpandPseudo::ExpandMBB(MachineBasicBlock &MBB) {
  bool Modified = false;

  // MBBI may be invalidated by the expansion.
  MachineBasicBlock::iterator MBBI = MBB.begin(), E = MBB.end();
  while (MBBI != E) {
    MachineBasicBlock::iterator NMBBI = std::next(MBBI);
    Modified |= ExpandMI(MBB, MBBI);
    MBBI = NMBBI;
  }

  return Modified;
}

bool X86ExpandPseudo::runOnMachineFunction(MachineFunction &MF) {
  STI = &static_cast<const X86Subtarget &>(MF.getSubtarget());
  TII = STI->getInstrInfo();
  TRI = STI->getRegisterInfo();
  X86FI = MF.getInfo<X86MachineFunctionInfo>();
  X86FL = STI->getFrameLowering();

  bool Modified = false;
  for (MachineBasicBlock &MBB : MF)
    Modified |= ExpandMBB(MBB);
  return Modified;
}

/// Returns an instance of the pseudo instruction expansion pass.
FunctionPass *llvm::createX86ExpandPseudoPass() {
  return new X86ExpandPseudo();
}