xref: /netbsd/external/apache2/llvm/dist/llvm/lib/Target/AMDGPU/SIPeepholeSDWA.cpp (revision da58b97a)

//===- SIPeepholeSDWA.cpp - Peephole optimization for SDWA 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
//
//===----------------------------------------------------------------------===//
//
/// \file This pass tries to apply several peephole SDWA patterns.
///
/// E.g. original:
///   V_LSHRREV_B32_e32 %0, 16, %1
///   V_ADD_CO_U32_e32 %2, %0, %3
///   V_LSHLREV_B32_e32 %4, 16, %2
///
/// Replace:
///   V_ADD_CO_U32_sdwa %4, %1, %3
///       dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD
///
//===----------------------------------------------------------------------===//

#include "AMDGPU.h"
#include "GCNSubtarget.h"
#include "MCTargetDesc/AMDGPUMCTargetDesc.h"
#include "llvm/ADT/MapVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/CodeGen/MachineFunctionPass.h"

using namespace llvm;

#define DEBUG_TYPE "si-peephole-sdwa"

STATISTIC(NumSDWAPatternsFound, "Number of SDWA patterns found.");
STATISTIC(NumSDWAInstructionsPeepholed,
          "Number of instruction converted to SDWA.");

namespace {

class SDWAOperand;
class SDWADstOperand;

class SIPeepholeSDWA : public MachineFunctionPass {
public:
  using SDWAOperandsVector = SmallVector<SDWAOperand *, 4>;

private:
  MachineRegisterInfo *MRI;
  const SIRegisterInfo *TRI;
  const SIInstrInfo *TII;

  MapVector<MachineInstr *, std::unique_ptr<SDWAOperand>> SDWAOperands;
  MapVector<MachineInstr *, SDWAOperandsVector> PotentialMatches;
  SmallVector<MachineInstr *, 8> ConvertedInstructions;

  Optional<int64_t> foldToImm(const MachineOperand &Op) const;

public:
  static char ID;

  SIPeepholeSDWA() : MachineFunctionPass(ID) {
    initializeSIPeepholeSDWAPass(*PassRegistry::getPassRegistry());
  }

  bool runOnMachineFunction(MachineFunction &MF) override;
  void matchSDWAOperands(MachineBasicBlock &MBB);
  std::unique_ptr<SDWAOperand> matchSDWAOperand(MachineInstr &MI);
  bool isConvertibleToSDWA(MachineInstr &MI, const GCNSubtarget &ST) const;
  void pseudoOpConvertToVOP2(MachineInstr &MI,
                             const GCNSubtarget &ST) const;
  bool convertToSDWA(MachineInstr &MI, const SDWAOperandsVector &SDWAOperands);
  void legalizeScalarOperands(MachineInstr &MI, const GCNSubtarget &ST) const;

  StringRef getPassName() const override { return "SI Peephole SDWA"; }

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

class SDWAOperand {
private:
  MachineOperand *Target; // Operand that would be used in converted instruction
  MachineOperand *Replaced; // Operand that would be replace by Target

public:
  SDWAOperand(MachineOperand *TargetOp, MachineOperand *ReplacedOp)
      : Target(TargetOp), Replaced(ReplacedOp) {
    assert(Target->isReg());
    assert(Replaced->isReg());
  }

  virtual ~SDWAOperand() = default;

  virtual MachineInstr *potentialToConvert(const SIInstrInfo *TII) = 0;
  virtual bool convertToSDWA(MachineInstr &MI, const SIInstrInfo *TII) = 0;

  MachineOperand *getTargetOperand() const { return Target; }
  MachineOperand *getReplacedOperand() const { return Replaced; }
  MachineInstr *getParentInst() const { return Target->getParent(); }

  MachineRegisterInfo *getMRI() const {
    return &getParentInst()->getParent()->getParent()->getRegInfo();
  }

#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
  virtual void print(raw_ostream& OS) const = 0;
  void dump() const { print(dbgs()); }
#endif
};

using namespace AMDGPU::SDWA;

class SDWASrcOperand : public SDWAOperand {
private:
  SdwaSel SrcSel;
  bool Abs;
  bool Neg;
  bool Sext;

public:
  SDWASrcOperand(MachineOperand *TargetOp, MachineOperand *ReplacedOp,
                 SdwaSel SrcSel_ = DWORD, bool Abs_ = false, bool Neg_ = false,
                 bool Sext_ = false)
      : SDWAOperand(TargetOp, ReplacedOp),
        SrcSel(SrcSel_), Abs(Abs_), Neg(Neg_), Sext(Sext_) {}

  MachineInstr *potentialToConvert(const SIInstrInfo *TII) override;
  bool convertToSDWA(MachineInstr &MI, const SIInstrInfo *TII) override;

  SdwaSel getSrcSel() const { return SrcSel; }
  bool getAbs() const { return Abs; }
  bool getNeg() const { return Neg; }
  bool getSext() const { return Sext; }

  uint64_t getSrcMods(const SIInstrInfo *TII,
                      const MachineOperand *SrcOp) const;

#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
  void print(raw_ostream& OS) const override;
#endif
};

class SDWADstOperand : public SDWAOperand {
private:
  SdwaSel DstSel;
  DstUnused DstUn;

public:

  SDWADstOperand(MachineOperand *TargetOp, MachineOperand *ReplacedOp,
                 SdwaSel DstSel_ = DWORD, DstUnused DstUn_ = UNUSED_PAD)
    : SDWAOperand(TargetOp, ReplacedOp), DstSel(DstSel_), DstUn(DstUn_) {}

  MachineInstr *potentialToConvert(const SIInstrInfo *TII) override;
  bool convertToSDWA(MachineInstr &MI, const SIInstrInfo *TII) override;

  SdwaSel getDstSel() const { return DstSel; }
  DstUnused getDstUnused() const { return DstUn; }

#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
  void print(raw_ostream& OS) const override;
#endif
};

class SDWADstPreserveOperand : public SDWADstOperand {
private:
  MachineOperand *Preserve;

public:
  SDWADstPreserveOperand(MachineOperand *TargetOp, MachineOperand *ReplacedOp,
                         MachineOperand *PreserveOp, SdwaSel DstSel_ = DWORD)
      : SDWADstOperand(TargetOp, ReplacedOp, DstSel_, UNUSED_PRESERVE),
        Preserve(PreserveOp) {}

  bool convertToSDWA(MachineInstr &MI, const SIInstrInfo *TII) override;

  MachineOperand *getPreservedOperand() const { return Preserve; }

#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
  void print(raw_ostream& OS) const override;
#endif
};

} // end anonymous namespace

INITIALIZE_PASS(SIPeepholeSDWA, DEBUG_TYPE, "SI Peephole SDWA", false, false)

char SIPeepholeSDWA::ID = 0;

char &llvm::SIPeepholeSDWAID = SIPeepholeSDWA::ID;

FunctionPass *llvm::createSIPeepholeSDWAPass() {
  return new SIPeepholeSDWA();
}


#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
static raw_ostream& operator<<(raw_ostream &OS, SdwaSel Sel) {
  switch(Sel) {
  case BYTE_0: OS << "BYTE_0"; break;
  case BYTE_1: OS << "BYTE_1"; break;
  case BYTE_2: OS << "BYTE_2"; break;
  case BYTE_3: OS << "BYTE_3"; break;
  case WORD_0: OS << "WORD_0"; break;
  case WORD_1: OS << "WORD_1"; break;
  case DWORD:  OS << "DWORD"; break;
  }
  return OS;
}

static raw_ostream& operator<<(raw_ostream &OS, const DstUnused &Un) {
  switch(Un) {
  case UNUSED_PAD: OS << "UNUSED_PAD"; break;
  case UNUSED_SEXT: OS << "UNUSED_SEXT"; break;
  case UNUSED_PRESERVE: OS << "UNUSED_PRESERVE"; break;
  }
  return OS;
}

LLVM_DUMP_METHOD
void SDWASrcOperand::print(raw_ostream& OS) const {
  OS << "SDWA src: " << *getTargetOperand()
    << " src_sel:" << getSrcSel()
    << " abs:" << getAbs() << " neg:" << getNeg()
    << " sext:" << getSext() << '\n';
}

LLVM_DUMP_METHOD
void SDWADstOperand::print(raw_ostream& OS) const {
  OS << "SDWA dst: " << *getTargetOperand()
    << " dst_sel:" << getDstSel()
    << " dst_unused:" << getDstUnused() << '\n';
}

LLVM_DUMP_METHOD
void SDWADstPreserveOperand::print(raw_ostream& OS) const {
  OS << "SDWA preserve dst: " << *getTargetOperand()
    << " dst_sel:" << getDstSel()
    << " preserve:" << *getPreservedOperand() << '\n';
}

#endif

static void copyRegOperand(MachineOperand &To, const MachineOperand &From) {
  assert(To.isReg() && From.isReg());
  To.setReg(From.getReg());
  To.setSubReg(From.getSubReg());
  To.setIsUndef(From.isUndef());
  if (To.isUse()) {
    To.setIsKill(From.isKill());
  } else {
    To.setIsDead(From.isDead());
  }
}

static bool isSameReg(const MachineOperand &LHS, const MachineOperand &RHS) {
  return LHS.isReg() &&
         RHS.isReg() &&
         LHS.getReg() == RHS.getReg() &&
         LHS.getSubReg() == RHS.getSubReg();
}

static MachineOperand *findSingleRegUse(const MachineOperand *Reg,
                                        const MachineRegisterInfo *MRI) {
  if (!Reg->isReg() || !Reg->isDef())
    return nullptr;

  MachineOperand *ResMO = nullptr;
  for (MachineOperand &UseMO : MRI->use_nodbg_operands(Reg->getReg())) {
    // If there exist use of subreg of Reg then return nullptr
    if (!isSameReg(UseMO, *Reg))
      return nullptr;

    // Check that there is only one instruction that uses Reg
    if (!ResMO) {
      ResMO = &UseMO;
    } else if (ResMO->getParent() != UseMO.getParent()) {
      return nullptr;
    }
  }

  return ResMO;
}

static MachineOperand *findSingleRegDef(const MachineOperand *Reg,
                                        const MachineRegisterInfo *MRI) {
  if (!Reg->isReg())
    return nullptr;

  MachineInstr *DefInstr = MRI->getUniqueVRegDef(Reg->getReg());
  if (!DefInstr)
    return nullptr;

  for (auto &DefMO : DefInstr->defs()) {
    if (DefMO.isReg() && DefMO.getReg() == Reg->getReg())
      return &DefMO;
  }

  // Ignore implicit defs.
  return nullptr;
}

uint64_t SDWASrcOperand::getSrcMods(const SIInstrInfo *TII,
                                    const MachineOperand *SrcOp) const {
  uint64_t Mods = 0;
  const auto *MI = SrcOp->getParent();
  if (TII->getNamedOperand(*MI, AMDGPU::OpName::src0) == SrcOp) {
    if (auto *Mod = TII->getNamedOperand(*MI, AMDGPU::OpName::src0_modifiers)) {
      Mods = Mod->getImm();
    }
  } else if (TII->getNamedOperand(*MI, AMDGPU::OpName::src1) == SrcOp) {
    if (auto *Mod = TII->getNamedOperand(*MI, AMDGPU::OpName::src1_modifiers)) {
      Mods = Mod->getImm();
    }
  }
  if (Abs || Neg) {
    assert(!Sext &&
           "Float and integer src modifiers can't be set simulteniously");
    Mods |= Abs ? SISrcMods::ABS : 0u;
    Mods ^= Neg ? SISrcMods::NEG : 0u;
  } else if (Sext) {
    Mods |= SISrcMods::SEXT;
  }

  return Mods;
}

MachineInstr *SDWASrcOperand::potentialToConvert(const SIInstrInfo *TII) {
  // For SDWA src operand potential instruction is one that use register
  // defined by parent instruction
  MachineOperand *PotentialMO = findSingleRegUse(getReplacedOperand(), getMRI());
  if (!PotentialMO)
    return nullptr;

  return PotentialMO->getParent();
}

bool SDWASrcOperand::convertToSDWA(MachineInstr &MI, const SIInstrInfo *TII) {
  // Find operand in instruction that matches source operand and replace it with
  // target operand. Set corresponding src_sel
  bool IsPreserveSrc = false;
  MachineOperand *Src = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
  MachineOperand *SrcSel = TII->getNamedOperand(MI, AMDGPU::OpName::src0_sel);
  MachineOperand *SrcMods =
      TII->getNamedOperand(MI, AMDGPU::OpName::src0_modifiers);
  assert(Src && (Src->isReg() || Src->isImm()));
  if (!isSameReg(*Src, *getReplacedOperand())) {
    // If this is not src0 then it could be src1
    Src = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
    SrcSel = TII->getNamedOperand(MI, AMDGPU::OpName::src1_sel);
    SrcMods = TII->getNamedOperand(MI, AMDGPU::OpName::src1_modifiers);

    if (!Src ||
        !isSameReg(*Src, *getReplacedOperand())) {
      // It's possible this Src is a tied operand for
      // UNUSED_PRESERVE, in which case we can either
      // abandon the peephole attempt, or if legal we can
      // copy the target operand into the tied slot
      // if the preserve operation will effectively cause the same
      // result by overwriting the rest of the dst.
      MachineOperand *Dst = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);
      MachineOperand *DstUnused =
        TII->getNamedOperand(MI, AMDGPU::OpName::dst_unused);

      if (Dst &&
          DstUnused->getImm() == AMDGPU::SDWA::DstUnused::UNUSED_PRESERVE) {
        // This will work if the tied src is acessing WORD_0, and the dst is
        // writing WORD_1. Modifiers don't matter because all the bits that
        // would be impacted are being overwritten by the dst.
        // Any other case will not work.
        SdwaSel DstSel = static_cast<SdwaSel>(
            TII->getNamedImmOperand(MI, AMDGPU::OpName::dst_sel));
        if (DstSel == AMDGPU::SDWA::SdwaSel::WORD_1 &&
            getSrcSel() == AMDGPU::SDWA::SdwaSel::WORD_0) {
          IsPreserveSrc = true;
          auto DstIdx = AMDGPU::getNamedOperandIdx(MI.getOpcode(),
                                                   AMDGPU::OpName::vdst);
          auto TiedIdx = MI.findTiedOperandIdx(DstIdx);
          Src = &MI.getOperand(TiedIdx);
          SrcSel = nullptr;
          SrcMods = nullptr;
        } else {
          // Not legal to convert this src
          return false;
        }
      }
    }
    assert(Src && Src->isReg());

    if ((MI.getOpcode() == AMDGPU::V_FMAC_F16_sdwa ||
         MI.getOpcode() == AMDGPU::V_FMAC_F32_sdwa ||
         MI.getOpcode() == AMDGPU::V_MAC_F16_sdwa ||
         MI.getOpcode() == AMDGPU::V_MAC_F32_sdwa) &&
         !isSameReg(*Src, *getReplacedOperand())) {
      // In case of v_mac_f16/32_sdwa this pass can try to apply src operand to
      // src2. This is not allowed.
      return false;
    }

    assert(isSameReg(*Src, *getReplacedOperand()) &&
           (IsPreserveSrc || (SrcSel && SrcMods)));
  }
  copyRegOperand(*Src, *getTargetOperand());
  if (!IsPreserveSrc) {
    SrcSel->setImm(getSrcSel());
    SrcMods->setImm(getSrcMods(TII, Src));
  }
  getTargetOperand()->setIsKill(false);
  return true;
}

MachineInstr *SDWADstOperand::potentialToConvert(const SIInstrInfo *TII) {
  // For SDWA dst operand potential instruction is one that defines register
  // that this operand uses
  MachineRegisterInfo *MRI = getMRI();
  MachineInstr *ParentMI = getParentInst();

  MachineOperand *PotentialMO = findSingleRegDef(getReplacedOperand(), MRI);
  if (!PotentialMO)
    return nullptr;

  // Check that ParentMI is the only instruction that uses replaced register
  for (MachineInstr &UseInst : MRI->use_nodbg_instructions(PotentialMO->getReg())) {
    if (&UseInst != ParentMI)
      return nullptr;
  }

  return PotentialMO->getParent();
}

bool SDWADstOperand::convertToSDWA(MachineInstr &MI, const SIInstrInfo *TII) {
  // Replace vdst operand in MI with target operand. Set dst_sel and dst_unused

  if ((MI.getOpcode() == AMDGPU::V_FMAC_F16_sdwa ||
       MI.getOpcode() == AMDGPU::V_FMAC_F32_sdwa ||
       MI.getOpcode() == AMDGPU::V_MAC_F16_sdwa ||
       MI.getOpcode() == AMDGPU::V_MAC_F32_sdwa) &&
      getDstSel() != AMDGPU::SDWA::DWORD) {
    // v_mac_f16/32_sdwa allow dst_sel to be equal only to DWORD
    return false;
  }

  MachineOperand *Operand = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);
  assert(Operand &&
         Operand->isReg() &&
         isSameReg(*Operand, *getReplacedOperand()));
  copyRegOperand(*Operand, *getTargetOperand());
  MachineOperand *DstSel= TII->getNamedOperand(MI, AMDGPU::OpName::dst_sel);
  assert(DstSel);
  DstSel->setImm(getDstSel());
  MachineOperand *DstUnused= TII->getNamedOperand(MI, AMDGPU::OpName::dst_unused);
  assert(DstUnused);
  DstUnused->setImm(getDstUnused());

  // Remove original instruction  because it would conflict with our new
  // instruction by register definition
  getParentInst()->eraseFromParent();
  return true;
}

bool SDWADstPreserveOperand::convertToSDWA(MachineInstr &MI,
                                           const SIInstrInfo *TII) {
  // MI should be moved right before v_or_b32.
  // For this we should clear all kill flags on uses of MI src-operands or else
  // we can encounter problem with use of killed operand.
  for (MachineOperand &MO : MI.uses()) {
    if (!MO.isReg())
      continue;
    getMRI()->clearKillFlags(MO.getReg());
  }

  // Move MI before v_or_b32
  auto MBB = MI.getParent();
  MBB->remove(&MI);
  MBB->insert(getParentInst(), &MI);

  // Add Implicit use of preserved register
  MachineInstrBuilder MIB(*MBB->getParent(), MI);
  MIB.addReg(getPreservedOperand()->getReg(),
             RegState::ImplicitKill,
             getPreservedOperand()->getSubReg());

  // Tie dst to implicit use
  MI.tieOperands(AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::vdst),
                 MI.getNumOperands() - 1);

  // Convert MI as any other SDWADstOperand and remove v_or_b32
  return SDWADstOperand::convertToSDWA(MI, TII);
}

Optional<int64_t> SIPeepholeSDWA::foldToImm(const MachineOperand &Op) const {
  if (Op.isImm()) {
    return Op.getImm();
  }

  // If this is not immediate then it can be copy of immediate value, e.g.:
  // %1 = S_MOV_B32 255;
  if (Op.isReg()) {
    for (const MachineOperand &Def : MRI->def_operands(Op.getReg())) {
      if (!isSameReg(Op, Def))
        continue;

      const MachineInstr *DefInst = Def.getParent();
      if (!TII->isFoldableCopy(*DefInst))
        return None;

      const MachineOperand &Copied = DefInst->getOperand(1);
      if (!Copied.isImm())
        return None;

      return Copied.getImm();
    }
  }

  return None;
}

std::unique_ptr<SDWAOperand>
SIPeepholeSDWA::matchSDWAOperand(MachineInstr &MI) {
  unsigned Opcode = MI.getOpcode();
  switch (Opcode) {
  case AMDGPU::V_LSHRREV_B32_e32:
  case AMDGPU::V_ASHRREV_I32_e32:
  case AMDGPU::V_LSHLREV_B32_e32:
  case AMDGPU::V_LSHRREV_B32_e64:
  case AMDGPU::V_ASHRREV_I32_e64:
  case AMDGPU::V_LSHLREV_B32_e64: {
    // from: v_lshrrev_b32_e32 v1, 16/24, v0
    // to SDWA src:v0 src_sel:WORD_1/BYTE_3

    // from: v_ashrrev_i32_e32 v1, 16/24, v0
    // to SDWA src:v0 src_sel:WORD_1/BYTE_3 sext:1

    // from: v_lshlrev_b32_e32 v1, 16/24, v0
    // to SDWA dst:v1 dst_sel:WORD_1/BYTE_3 dst_unused:UNUSED_PAD
    MachineOperand *Src0 = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
    auto Imm = foldToImm(*Src0);
    if (!Imm)
      break;

    if (*Imm != 16 && *Imm != 24)
      break;

    MachineOperand *Src1 = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
    MachineOperand *Dst = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);
    if (Src1->getReg().isPhysical() || Dst->getReg().isPhysical())
      break;

    if (Opcode == AMDGPU::V_LSHLREV_B32_e32 ||
        Opcode == AMDGPU::V_LSHLREV_B32_e64) {
      return std::make_unique<SDWADstOperand>(
          Dst, Src1, *Imm == 16 ? WORD_1 : BYTE_3, UNUSED_PAD);
    } else {
      return std::make_unique<SDWASrcOperand>(
          Src1, Dst, *Imm == 16 ? WORD_1 : BYTE_3, false, false,
          Opcode != AMDGPU::V_LSHRREV_B32_e32 &&
          Opcode != AMDGPU::V_LSHRREV_B32_e64);
    }
    break;
  }

  case AMDGPU::V_LSHRREV_B16_e32:
  case AMDGPU::V_ASHRREV_I16_e32:
  case AMDGPU::V_LSHLREV_B16_e32:
  case AMDGPU::V_LSHRREV_B16_e64:
  case AMDGPU::V_ASHRREV_I16_e64:
  case AMDGPU::V_LSHLREV_B16_e64: {
    // from: v_lshrrev_b16_e32 v1, 8, v0
    // to SDWA src:v0 src_sel:BYTE_1

    // from: v_ashrrev_i16_e32 v1, 8, v0
    // to SDWA src:v0 src_sel:BYTE_1 sext:1

    // from: v_lshlrev_b16_e32 v1, 8, v0
    // to SDWA dst:v1 dst_sel:BYTE_1 dst_unused:UNUSED_PAD
    MachineOperand *Src0 = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
    auto Imm = foldToImm(*Src0);
    if (!Imm || *Imm != 8)
      break;

    MachineOperand *Src1 = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
    MachineOperand *Dst = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);

    if (Src1->getReg().isPhysical() || Dst->getReg().isPhysical())
      break;

    if (Opcode == AMDGPU::V_LSHLREV_B16_e32 ||
        Opcode == AMDGPU::V_LSHLREV_B16_e64) {
      return std::make_unique<SDWADstOperand>(Dst, Src1, BYTE_1, UNUSED_PAD);
    } else {
      return std::make_unique<SDWASrcOperand>(
            Src1, Dst, BYTE_1, false, false,
            Opcode != AMDGPU::V_LSHRREV_B16_e32 &&
            Opcode != AMDGPU::V_LSHRREV_B16_e64);
    }
    break;
  }

  case AMDGPU::V_BFE_I32_e64:
  case AMDGPU::V_BFE_U32_e64: {
    // e.g.:
    // from: v_bfe_u32 v1, v0, 8, 8
    // to SDWA src:v0 src_sel:BYTE_1

    // offset | width | src_sel
    // ------------------------
    // 0      | 8     | BYTE_0
    // 0      | 16    | WORD_0
    // 0      | 32    | DWORD ?
    // 8      | 8     | BYTE_1
    // 16     | 8     | BYTE_2
    // 16     | 16    | WORD_1
    // 24     | 8     | BYTE_3

    MachineOperand *Src1 = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
    auto Offset = foldToImm(*Src1);
    if (!Offset)
      break;

    MachineOperand *Src2 = TII->getNamedOperand(MI, AMDGPU::OpName::src2);
    auto Width = foldToImm(*Src2);
    if (!Width)
      break;

    SdwaSel SrcSel = DWORD;

    if (*Offset == 0 && *Width == 8)
      SrcSel = BYTE_0;
    else if (*Offset == 0 && *Width == 16)
      SrcSel = WORD_0;
    else if (*Offset == 0 && *Width == 32)
      SrcSel = DWORD;
    else if (*Offset == 8 && *Width == 8)
      SrcSel = BYTE_1;
    else if (*Offset == 16 && *Width == 8)
      SrcSel = BYTE_2;
    else if (*Offset == 16 && *Width == 16)
      SrcSel = WORD_1;
    else if (*Offset == 24 && *Width == 8)
      SrcSel = BYTE_3;
    else
      break;

    MachineOperand *Src0 = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
    MachineOperand *Dst = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);

    if (Src0->getReg().isPhysical() || Dst->getReg().isPhysical())
      break;

    return std::make_unique<SDWASrcOperand>(
          Src0, Dst, SrcSel, false, false, Opcode != AMDGPU::V_BFE_U32_e64);
  }

  case AMDGPU::V_AND_B32_e32:
  case AMDGPU::V_AND_B32_e64: {
    // e.g.:
    // from: v_and_b32_e32 v1, 0x0000ffff/0x000000ff, v0
    // to SDWA src:v0 src_sel:WORD_0/BYTE_0

    MachineOperand *Src0 = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
    MachineOperand *Src1 = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
    auto ValSrc = Src1;
    auto Imm = foldToImm(*Src0);

    if (!Imm) {
      Imm = foldToImm(*Src1);
      ValSrc = Src0;
    }

    if (!Imm || (*Imm != 0x0000ffff && *Imm != 0x000000ff))
      break;

    MachineOperand *Dst = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);

    if (ValSrc->getReg().isPhysical() || Dst->getReg().isPhysical())
      break;

    return std::make_unique<SDWASrcOperand>(
        ValSrc, Dst, *Imm == 0x0000ffff ? WORD_0 : BYTE_0);
  }

  case AMDGPU::V_OR_B32_e32:
  case AMDGPU::V_OR_B32_e64: {
    // Patterns for dst_unused:UNUSED_PRESERVE.
    // e.g., from:
    // v_add_f16_sdwa v0, v1, v2 dst_sel:WORD_1 dst_unused:UNUSED_PAD
    //                           src1_sel:WORD_1 src2_sel:WORD1
    // v_add_f16_e32 v3, v1, v2
    // v_or_b32_e32 v4, v0, v3
    // to SDWA preserve dst:v4 dst_sel:WORD_1 dst_unused:UNUSED_PRESERVE preserve:v3

    // Check if one of operands of v_or_b32 is SDWA instruction
    using CheckRetType = Optional<std::pair<MachineOperand *, MachineOperand *>>;
    auto CheckOROperandsForSDWA =
      [&](const MachineOperand *Op1, const MachineOperand *Op2) -> CheckRetType {
        if (!Op1 || !Op1->isReg() || !Op2 || !Op2->isReg())
          return CheckRetType(None);

        MachineOperand *Op1Def = findSingleRegDef(Op1, MRI);
        if (!Op1Def)
          return CheckRetType(None);

        MachineInstr *Op1Inst = Op1Def->getParent();
        if (!TII->isSDWA(*Op1Inst))
          return CheckRetType(None);

        MachineOperand *Op2Def = findSingleRegDef(Op2, MRI);
        if (!Op2Def)
          return CheckRetType(None);

        return CheckRetType(std::make_pair(Op1Def, Op2Def));
      };

    MachineOperand *OrSDWA = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
    MachineOperand *OrOther = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
    assert(OrSDWA && OrOther);
    auto Res = CheckOROperandsForSDWA(OrSDWA, OrOther);
    if (!Res) {
      OrSDWA = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
      OrOther = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
      assert(OrSDWA && OrOther);
      Res = CheckOROperandsForSDWA(OrSDWA, OrOther);
      if (!Res)
        break;
    }

    MachineOperand *OrSDWADef = Res->first;
    MachineOperand *OrOtherDef = Res->second;
    assert(OrSDWADef && OrOtherDef);

    MachineInstr *SDWAInst = OrSDWADef->getParent();
    MachineInstr *OtherInst = OrOtherDef->getParent();

    // Check that OtherInstr is actually bitwise compatible with SDWAInst = their
    // destination patterns don't overlap. Compatible instruction can be either
    // regular instruction with compatible bitness or SDWA instruction with
    // correct dst_sel
    // SDWAInst | OtherInst bitness / OtherInst dst_sel
    // -----------------------------------------------------
    // DWORD    | no                    / no
    // WORD_0   | no                    / BYTE_2/3, WORD_1
    // WORD_1   | 8/16-bit instructions / BYTE_0/1, WORD_0
    // BYTE_0   | no                    / BYTE_1/2/3, WORD_1
    // BYTE_1   | 8-bit                 / BYTE_0/2/3, WORD_1
    // BYTE_2   | 8/16-bit              / BYTE_0/1/3. WORD_0
    // BYTE_3   | 8/16/24-bit           / BYTE_0/1/2, WORD_0
    // E.g. if SDWAInst is v_add_f16_sdwa dst_sel:WORD_1 then v_add_f16 is OK
    // but v_add_f32 is not.

    // TODO: add support for non-SDWA instructions as OtherInst.
    // For now this only works with SDWA instructions. For regular instructions
    // there is no way to determine if the instruction writes only 8/16/24-bit
    // out of full register size and all registers are at min 32-bit wide.
    if (!TII->isSDWA(*OtherInst))
      break;

    SdwaSel DstSel = static_cast<SdwaSel>(
      TII->getNamedImmOperand(*SDWAInst, AMDGPU::OpName::dst_sel));;
    SdwaSel OtherDstSel = static_cast<SdwaSel>(
      TII->getNamedImmOperand(*OtherInst, AMDGPU::OpName::dst_sel));

    bool DstSelAgree = false;
    switch (DstSel) {
    case WORD_0: DstSelAgree = ((OtherDstSel == BYTE_2) ||
                                (OtherDstSel == BYTE_3) ||
                                (OtherDstSel == WORD_1));
      break;
    case WORD_1: DstSelAgree = ((OtherDstSel == BYTE_0) ||
                                (OtherDstSel == BYTE_1) ||
                                (OtherDstSel == WORD_0));
      break;
    case BYTE_0: DstSelAgree = ((OtherDstSel == BYTE_1) ||
                                (OtherDstSel == BYTE_2) ||
                                (OtherDstSel == BYTE_3) ||
                                (OtherDstSel == WORD_1));
      break;
    case BYTE_1: DstSelAgree = ((OtherDstSel == BYTE_0) ||
                                (OtherDstSel == BYTE_2) ||
                                (OtherDstSel == BYTE_3) ||
                                (OtherDstSel == WORD_1));
      break;
    case BYTE_2: DstSelAgree = ((OtherDstSel == BYTE_0) ||
                                (OtherDstSel == BYTE_1) ||
                                (OtherDstSel == BYTE_3) ||
                                (OtherDstSel == WORD_0));
      break;
    case BYTE_3: DstSelAgree = ((OtherDstSel == BYTE_0) ||
                                (OtherDstSel == BYTE_1) ||
                                (OtherDstSel == BYTE_2) ||
                                (OtherDstSel == WORD_0));
      break;
    default: DstSelAgree = false;
    }

    if (!DstSelAgree)
      break;

    // Also OtherInst dst_unused should be UNUSED_PAD
    DstUnused OtherDstUnused = static_cast<DstUnused>(
      TII->getNamedImmOperand(*OtherInst, AMDGPU::OpName::dst_unused));
    if (OtherDstUnused != DstUnused::UNUSED_PAD)
      break;

    // Create DstPreserveOperand
    MachineOperand *OrDst = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);
    assert(OrDst && OrDst->isReg());

    return std::make_unique<SDWADstPreserveOperand>(
      OrDst, OrSDWADef, OrOtherDef, DstSel);

  }
  }

  return std::unique_ptr<SDWAOperand>(nullptr);
}

#if !defined(NDEBUG)
static raw_ostream& operator<<(raw_ostream &OS, const SDWAOperand &Operand) {
  Operand.print(OS);
  return OS;
}
#endif

void SIPeepholeSDWA::matchSDWAOperands(MachineBasicBlock &MBB) {
  for (MachineInstr &MI : MBB) {
    if (auto Operand = matchSDWAOperand(MI)) {
      LLVM_DEBUG(dbgs() << "Match: " << MI << "To: " << *Operand << '\n');
      SDWAOperands[&MI] = std::move(Operand);
      ++NumSDWAPatternsFound;
    }
  }
}

// Convert the V_ADDC_U32_e64 into V_ADDC_U32_e32, and
// V_ADD_CO_U32_e64 into V_ADD_CO_U32_e32. This allows isConvertibleToSDWA
// to perform its transformation on V_ADD_CO_U32_e32 into V_ADD_CO_U32_sdwa.
//
// We are transforming from a VOP3 into a VOP2 form of the instruction.
//   %19:vgpr_32 = V_AND_B32_e32 255,
//       killed %16:vgpr_32, implicit $exec
//   %47:vgpr_32, %49:sreg_64_xexec = V_ADD_CO_U32_e64
//       %26.sub0:vreg_64, %19:vgpr_32, implicit $exec
//  %48:vgpr_32, dead %50:sreg_64_xexec = V_ADDC_U32_e64
//       %26.sub1:vreg_64, %54:vgpr_32, killed %49:sreg_64_xexec, implicit $exec
//
// becomes
//   %47:vgpr_32 = V_ADD_CO_U32_sdwa
//       0, %26.sub0:vreg_64, 0, killed %16:vgpr_32, 0, 6, 0, 6, 0,
//       implicit-def $vcc, implicit $exec
//  %48:vgpr_32 = V_ADDC_U32_e32
//       0, %26.sub1:vreg_64, implicit-def $vcc, implicit $vcc, implicit $exec
void SIPeepholeSDWA::pseudoOpConvertToVOP2(MachineInstr &MI,
                                           const GCNSubtarget &ST) const {
  int Opc = MI.getOpcode();
  assert((Opc == AMDGPU::V_ADD_CO_U32_e64 || Opc == AMDGPU::V_SUB_CO_U32_e64) &&
         "Currently only handles V_ADD_CO_U32_e64 or V_SUB_CO_U32_e64");

  // Can the candidate MI be shrunk?
  if (!TII->canShrink(MI, *MRI))
    return;
  Opc = AMDGPU::getVOPe32(Opc);
  // Find the related ADD instruction.
  const MachineOperand *Sdst = TII->getNamedOperand(MI, AMDGPU::OpName::sdst);
  if (!Sdst)
    return;
  MachineOperand *NextOp = findSingleRegUse(Sdst, MRI);
  if (!NextOp)
    return;
  MachineInstr &MISucc = *NextOp->getParent();
  // Can the successor be shrunk?
  if (!TII->canShrink(MISucc, *MRI))
    return;
  int SuccOpc = AMDGPU::getVOPe32(MISucc.getOpcode());
  // Make sure the carry in/out are subsequently unused.
  MachineOperand *CarryIn = TII->getNamedOperand(MISucc, AMDGPU::OpName::src2);
  if (!CarryIn)
    return;
  MachineOperand *CarryOut = TII->getNamedOperand(MISucc, AMDGPU::OpName::sdst);
  if (!CarryOut)
    return;
  if (!MRI->hasOneUse(CarryIn->getReg()) || !MRI->use_empty(CarryOut->getReg()))
    return;
  // Make sure VCC or its subregs are dead before MI.
  MachineBasicBlock &MBB = *MI.getParent();
  auto Liveness = MBB.computeRegisterLiveness(TRI, AMDGPU::VCC, MI, 25);
  if (Liveness != MachineBasicBlock::LQR_Dead)
    return;
  // Check if VCC is referenced in range of (MI,MISucc].
  for (auto I = std::next(MI.getIterator()), E = MISucc.getIterator();
       I != E; ++I) {
    if (I->modifiesRegister(AMDGPU::VCC, TRI))
      return;
  }

  // Make the two new e32 instruction variants.
  // Replace MI with V_{SUB|ADD}_I32_e32
  BuildMI(MBB, MI, MI.getDebugLoc(), TII->get(Opc))
    .add(*TII->getNamedOperand(MI, AMDGPU::OpName::vdst))
    .add(*TII->getNamedOperand(MI, AMDGPU::OpName::src0))
    .add(*TII->getNamedOperand(MI, AMDGPU::OpName::src1))
    .setMIFlags(MI.getFlags());

  MI.eraseFromParent();

  // Replace MISucc with V_{SUBB|ADDC}_U32_e32
  BuildMI(MBB, MISucc, MISucc.getDebugLoc(), TII->get(SuccOpc))
    .add(*TII->getNamedOperand(MISucc, AMDGPU::OpName::vdst))
    .add(*TII->getNamedOperand(MISucc, AMDGPU::OpName::src0))
    .add(*TII->getNamedOperand(MISucc, AMDGPU::OpName::src1))
    .setMIFlags(MISucc.getFlags());

  MISucc.eraseFromParent();
}

bool SIPeepholeSDWA::isConvertibleToSDWA(MachineInstr &MI,
                                         const GCNSubtarget &ST) const {
  // Check if this is already an SDWA instruction
  unsigned Opc = MI.getOpcode();
  if (TII->isSDWA(Opc))
    return true;

  // Check if this instruction has opcode that supports SDWA
  if (AMDGPU::getSDWAOp(Opc) == -1)
    Opc = AMDGPU::getVOPe32(Opc);

  if (AMDGPU::getSDWAOp(Opc) == -1)
    return false;

  if (!ST.hasSDWAOmod() && TII->hasModifiersSet(MI, AMDGPU::OpName::omod))
    return false;

  if (TII->isVOPC(Opc)) {
    if (!ST.hasSDWASdst()) {
      const MachineOperand *SDst = TII->getNamedOperand(MI, AMDGPU::OpName::sdst);
      if (SDst && (SDst->getReg() != AMDGPU::VCC &&
                   SDst->getReg() != AMDGPU::VCC_LO))
        return false;
    }

    if (!ST.hasSDWAOutModsVOPC() &&
        (TII->hasModifiersSet(MI, AMDGPU::OpName::clamp) ||
         TII->hasModifiersSet(MI, AMDGPU::OpName::omod)))
      return false;

  } else if (TII->getNamedOperand(MI, AMDGPU::OpName::sdst) ||
             !TII->getNamedOperand(MI, AMDGPU::OpName::vdst)) {
    return false;
  }

  if (!ST.hasSDWAMac() && (Opc == AMDGPU::V_FMAC_F16_e32 ||
                           Opc == AMDGPU::V_FMAC_F32_e32 ||
                           Opc == AMDGPU::V_MAC_F16_e32 ||
                           Opc == AMDGPU::V_MAC_F32_e32))
    return false;

  // Check if target supports this SDWA opcode
  if (TII->pseudoToMCOpcode(Opc) == -1)
    return false;

  // FIXME: has SDWA but require handling of implicit VCC use
  if (Opc == AMDGPU::V_CNDMASK_B32_e32)
    return false;

  if (MachineOperand *Src0 = TII->getNamedOperand(MI, AMDGPU::OpName::src0)) {
    if (!Src0->isReg() && !Src0->isImm())
      return false;
  }

  if (MachineOperand *Src1 = TII->getNamedOperand(MI, AMDGPU::OpName::src1)) {
    if (!Src1->isReg() && !Src1->isImm())
      return false;
  }

  return true;
}

bool SIPeepholeSDWA::convertToSDWA(MachineInstr &MI,
                                   const SDWAOperandsVector &SDWAOperands) {

  LLVM_DEBUG(dbgs() << "Convert instruction:" << MI);

  // Convert to sdwa
  int SDWAOpcode;
  unsigned Opcode = MI.getOpcode();
  if (TII->isSDWA(Opcode)) {
    SDWAOpcode = Opcode;
  } else {
    SDWAOpcode = AMDGPU::getSDWAOp(Opcode);
    if (SDWAOpcode == -1)
      SDWAOpcode = AMDGPU::getSDWAOp(AMDGPU::getVOPe32(Opcode));
  }
  assert(SDWAOpcode != -1);

  const MCInstrDesc &SDWADesc = TII->get(SDWAOpcode);

  // Create SDWA version of instruction MI and initialize its operands
  MachineInstrBuilder SDWAInst =
    BuildMI(*MI.getParent(), MI, MI.getDebugLoc(), SDWADesc)
    .setMIFlags(MI.getFlags());

  // Copy dst, if it is present in original then should also be present in SDWA
  MachineOperand *Dst = TII->getNamedOperand(MI, AMDGPU::OpName::vdst);
  if (Dst) {
    assert(AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::vdst) != -1);
    SDWAInst.add(*Dst);
  } else if ((Dst = TII->getNamedOperand(MI, AMDGPU::OpName::sdst))) {
    assert(Dst &&
           AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::sdst) != -1);
    SDWAInst.add(*Dst);
  } else {
    assert(AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::sdst) != -1);
    SDWAInst.addReg(TRI->getVCC(), RegState::Define);
  }

  // Copy src0, initialize src0_modifiers. All sdwa instructions has src0 and
  // src0_modifiers (except for v_nop_sdwa, but it can't get here)
  MachineOperand *Src0 = TII->getNamedOperand(MI, AMDGPU::OpName::src0);
  assert(
    Src0 &&
    AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::src0) != -1 &&
    AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::src0_modifiers) != -1);
  if (auto *Mod = TII->getNamedOperand(MI, AMDGPU::OpName::src0_modifiers))
    SDWAInst.addImm(Mod->getImm());
  else
    SDWAInst.addImm(0);
  SDWAInst.add(*Src0);

  // Copy src1 if present, initialize src1_modifiers.
  MachineOperand *Src1 = TII->getNamedOperand(MI, AMDGPU::OpName::src1);
  if (Src1) {
    assert(
      AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::src1) != -1 &&
      AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::src1_modifiers) != -1);
    if (auto *Mod = TII->getNamedOperand(MI, AMDGPU::OpName::src1_modifiers))
      SDWAInst.addImm(Mod->getImm());
    else
      SDWAInst.addImm(0);
    SDWAInst.add(*Src1);
  }

  if (SDWAOpcode == AMDGPU::V_FMAC_F16_sdwa ||
      SDWAOpcode == AMDGPU::V_FMAC_F32_sdwa ||
      SDWAOpcode == AMDGPU::V_MAC_F16_sdwa ||
      SDWAOpcode == AMDGPU::V_MAC_F32_sdwa) {
    // v_mac_f16/32 has additional src2 operand tied to vdst
    MachineOperand *Src2 = TII->getNamedOperand(MI, AMDGPU::OpName::src2);
    assert(Src2);
    SDWAInst.add(*Src2);
  }

  // Copy clamp if present, initialize otherwise
  assert(AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::clamp) != -1);
  MachineOperand *Clamp = TII->getNamedOperand(MI, AMDGPU::OpName::clamp);
  if (Clamp) {
    SDWAInst.add(*Clamp);
  } else {
    SDWAInst.addImm(0);
  }

  // Copy omod if present, initialize otherwise if needed
  if (AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::omod) != -1) {
    MachineOperand *OMod = TII->getNamedOperand(MI, AMDGPU::OpName::omod);
    if (OMod) {
      SDWAInst.add(*OMod);
    } else {
      SDWAInst.addImm(0);
    }
  }

  // Copy dst_sel if present, initialize otherwise if needed
  if (AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::dst_sel) != -1) {
    MachineOperand *DstSel = TII->getNamedOperand(MI, AMDGPU::OpName::dst_sel);
    if (DstSel) {
      SDWAInst.add(*DstSel);
    } else {
      SDWAInst.addImm(AMDGPU::SDWA::SdwaSel::DWORD);
    }
  }

  // Copy dst_unused if present, initialize otherwise if needed
  if (AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::dst_unused) != -1) {
    MachineOperand *DstUnused = TII->getNamedOperand(MI, AMDGPU::OpName::dst_unused);
    if (DstUnused) {
      SDWAInst.add(*DstUnused);
    } else {
      SDWAInst.addImm(AMDGPU::SDWA::DstUnused::UNUSED_PAD);
    }
  }

  // Copy src0_sel if present, initialize otherwise
  assert(AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::src0_sel) != -1);
  MachineOperand *Src0Sel = TII->getNamedOperand(MI, AMDGPU::OpName::src0_sel);
  if (Src0Sel) {
    SDWAInst.add(*Src0Sel);
  } else {
    SDWAInst.addImm(AMDGPU::SDWA::SdwaSel::DWORD);
  }

  // Copy src1_sel if present, initialize otherwise if needed
  if (Src1) {
    assert(AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::src1_sel) != -1);
    MachineOperand *Src1Sel = TII->getNamedOperand(MI, AMDGPU::OpName::src1_sel);
    if (Src1Sel) {
      SDWAInst.add(*Src1Sel);
    } else {
      SDWAInst.addImm(AMDGPU::SDWA::SdwaSel::DWORD);
    }
  }

  // Check for a preserved register that needs to be copied.
  auto DstUnused = TII->getNamedOperand(MI, AMDGPU::OpName::dst_unused);
  if (DstUnused &&
      DstUnused->getImm() == AMDGPU::SDWA::DstUnused::UNUSED_PRESERVE) {
    // We expect, if we are here, that the instruction was already in it's SDWA form,
    // with a tied operand.
    assert(Dst && Dst->isTied());
    assert(Opcode == static_cast<unsigned int>(SDWAOpcode));
    // We also expect a vdst, since sdst can't preserve.
    auto PreserveDstIdx = AMDGPU::getNamedOperandIdx(SDWAOpcode, AMDGPU::OpName::vdst);
    assert(PreserveDstIdx != -1);

    auto TiedIdx = MI.findTiedOperandIdx(PreserveDstIdx);
    auto Tied = MI.getOperand(TiedIdx);

    SDWAInst.add(Tied);
    SDWAInst->tieOperands(PreserveDstIdx, SDWAInst->getNumOperands() - 1);
  }

  // Apply all sdwa operand patterns.
  bool Converted = false;
  for (auto &Operand : SDWAOperands) {
    LLVM_DEBUG(dbgs() << *SDWAInst << "\nOperand: " << *Operand);
    // There should be no intesection between SDWA operands and potential MIs
    // e.g.:
    // v_and_b32 v0, 0xff, v1 -> src:v1 sel:BYTE_0
    // v_and_b32 v2, 0xff, v0 -> src:v0 sel:BYTE_0
    // v_add_u32 v3, v4, v2
    //
    // In that example it is possible that we would fold 2nd instruction into 3rd
    // (v_add_u32_sdwa) and then try to fold 1st instruction into 2nd (that was
    // already destroyed). So if SDWAOperand is also a potential MI then do not
    // apply it.
    if (PotentialMatches.count(Operand->getParentInst()) == 0)
      Converted |= Operand->convertToSDWA(*SDWAInst, TII);
  }
  if (Converted) {
    ConvertedInstructions.push_back(SDWAInst);
  } else {
    SDWAInst->eraseFromParent();
    return false;
  }

  LLVM_DEBUG(dbgs() << "\nInto:" << *SDWAInst << '\n');
  ++NumSDWAInstructionsPeepholed;

  MI.eraseFromParent();
  return true;
}

// If an instruction was converted to SDWA it should not have immediates or SGPR
// operands (allowed one SGPR on GFX9). Copy its scalar operands into VGPRs.
void SIPeepholeSDWA::legalizeScalarOperands(MachineInstr &MI,
                                            const GCNSubtarget &ST) const {
  const MCInstrDesc &Desc = TII->get(MI.getOpcode());
  unsigned ConstantBusCount = 0;
  for (MachineOperand &Op : MI.explicit_uses()) {
    if (!Op.isImm() && !(Op.isReg() && !TRI->isVGPR(*MRI, Op.getReg())))
      continue;

    unsigned I = MI.getOperandNo(&Op);
    if (Desc.OpInfo[I].RegClass == -1 ||
       !TRI->hasVGPRs(TRI->getRegClass(Desc.OpInfo[I].RegClass)))
      continue;

    if (ST.hasSDWAScalar() && ConstantBusCount == 0 && Op.isReg() &&
        TRI->isSGPRReg(*MRI, Op.getReg())) {
      ++ConstantBusCount;
      continue;
    }

    Register VGPR = MRI->createVirtualRegister(&AMDGPU::VGPR_32RegClass);
    auto Copy = BuildMI(*MI.getParent(), MI.getIterator(), MI.getDebugLoc(),
                        TII->get(AMDGPU::V_MOV_B32_e32), VGPR);
    if (Op.isImm())
      Copy.addImm(Op.getImm());
    else if (Op.isReg())
      Copy.addReg(Op.getReg(), Op.isKill() ? RegState::Kill : 0,
                  Op.getSubReg());
    Op.ChangeToRegister(VGPR, false);
  }
}

bool SIPeepholeSDWA::runOnMachineFunction(MachineFunction &MF) {
  const GCNSubtarget &ST = MF.getSubtarget<GCNSubtarget>();

  if (!ST.hasSDWA() || skipFunction(MF.getFunction()))
    return false;

  MRI = &MF.getRegInfo();
  TRI = ST.getRegisterInfo();
  TII = ST.getInstrInfo();

  // Find all SDWA operands in MF.
  bool Ret = false;
  for (MachineBasicBlock &MBB : MF) {
    bool Changed = false;
    do {
      // Preprocess the ADD/SUB pairs so they could be SDWA'ed.
      // Look for a possible ADD or SUB that resulted from a previously lowered
      // V_{ADD|SUB}_U64_PSEUDO. The function pseudoOpConvertToVOP2
      // lowers the pair of instructions into e32 form.
      matchSDWAOperands(MBB);
      for (const auto &OperandPair : SDWAOperands) {
        const auto &Operand = OperandPair.second;
        MachineInstr *PotentialMI = Operand->potentialToConvert(TII);
        if (PotentialMI &&
           (PotentialMI->getOpcode() == AMDGPU::V_ADD_CO_U32_e64 ||
            PotentialMI->getOpcode() == AMDGPU::V_SUB_CO_U32_e64))
          pseudoOpConvertToVOP2(*PotentialMI, ST);
      }
      SDWAOperands.clear();

      // Generate potential match list.
      matchSDWAOperands(MBB);

      for (const auto &OperandPair : SDWAOperands) {
        const auto &Operand = OperandPair.second;
        MachineInstr *PotentialMI = Operand->potentialToConvert(TII);
        if (PotentialMI && isConvertibleToSDWA(*PotentialMI, ST)) {
          PotentialMatches[PotentialMI].push_back(Operand.get());
        }
      }

      for (auto &PotentialPair : PotentialMatches) {
        MachineInstr &PotentialMI = *PotentialPair.first;
        convertToSDWA(PotentialMI, PotentialPair.second);
      }

      PotentialMatches.clear();
      SDWAOperands.clear();

      Changed = !ConvertedInstructions.empty();

      if (Changed)
        Ret = true;
      while (!ConvertedInstructions.empty())
        legalizeScalarOperands(*ConvertedInstructions.pop_back_val(), ST);
    } while (Changed);
  }

  return Ret;
}