[EVM] Do stackification of instructions in the EVMStackify pass to have valid MIR

This change moves the place where normal instructions are replaced with their
'stack' counterparts from the EVMMCInstLower to the EVMStackify pass. This
ensures we have a valid MIR after each pass.
This change also contains some updates in how labels are represented and
resolved in the BE.

Author: PavelKopyl

llvm/lib/Target/EVM/CMakeLists.txt

@@ -36,7 +36,6 @@ add_llvm_target(EVMCodeGen
   EVMStackify.cpp
   EVMSubtarget.cpp
   EVMTargetMachine.cpp
-  EVMTargetObjectFile.cpp
   EVMTargetTransformInfo.cpp
 
   LINK_COMPONENTS

llvm/lib/Target/EVM/EVMAsmPrinter.cpp

@@ -39,12 +39,36 @@ class EVMAsmPrinter : public AsmPrinter {
 
   StringRef getPassName() const override { return "EVM Assembly "; }
 
+  void SetupMachineFunction(MachineFunction &MF) override;
+
   void emitInstruction(const MachineInstr *MI) override;
 
   void emitFunctionEntryLabel() override;
+
+  /// Return true if the basic block has exactly one predecessor and the control
+  /// transfer mechanism between the predecessor and this block is a
+  /// fall-through.
+  bool isBlockOnlyReachableByFallthrough(
+      const MachineBasicBlock *MBB) const override;
 };
 } // end of anonymous namespace
 
+void EVMAsmPrinter::SetupMachineFunction(MachineFunction &MF) {
+  // Unbundle <push_label, jump> bundles.
+  for (MachineBasicBlock &MBB : MF) {
+    MachineBasicBlock::instr_iterator I = MBB.instr_begin(),
+                                      E = MBB.instr_end();
+    for (; I != E; ++I) {
+      if (I->isBundledWithPred()) {
+        assert(I->isConditionalBranch() || I->isUnconditionalBranch());
+        I->unbundleFromPred();
+      }
+    }
+  }
+
+  AsmPrinter::SetupMachineFunction(MF);
+}
+
 void EVMAsmPrinter::emitFunctionEntryLabel() {
   AsmPrinter::emitFunctionEntryLabel();
 
@@ -70,12 +94,17 @@ void EVMAsmPrinter::emitFunctionEntryLabel() {
 void EVMAsmPrinter::emitInstruction(const MachineInstr *MI) {
   EVMMCInstLower MCInstLowering(OutContext, *this, VRegMapping,
                                 MF->getRegInfo());
-
   MCInst TmpInst;
   MCInstLowering.Lower(MI, TmpInst);
   EmitToStreamer(*OutStreamer, TmpInst);
 }
 
+bool EVMAsmPrinter::isBlockOnlyReachableByFallthrough(
+    const MachineBasicBlock *MBB) const {
+  // For simplicity, always emit BB labels.
+  return false;
+}
+
 extern "C" LLVM_EXTERNAL_VISIBILITY void LLVMInitializeEVMAsmPrinter() {
   const RegisterAsmPrinter<EVMAsmPrinter> X(getTheEVMTarget());
 }

llvm/lib/Target/EVM/EVMISelLowering.cpp

@@ -387,7 +387,7 @@ SDValue EVMTargetLowering::LowerFormalArguments(
     fail(DL, DAG, "VarArg is not supported yet");
 
   MachineFunction &MF = DAG.getMachineFunction();
-  auto *MFI = MF.getInfo<EVMFunctionInfo>();
+  auto *MFI = MF.getInfo<EVMMachineFunctionInfo>();
 
   // Set up the incoming ARGUMENTS value, which serves to represent the liveness
   // of the incoming values before they're represented by virtual registers.

llvm/lib/Target/EVM/EVMInstrInfo.cpp

@@ -11,6 +11,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "EVMInstrInfo.h"
+#include "EVMMachineFunctionInfo.h"
 #include "MCTargetDesc/EVMMCTargetDesc.h"
 using namespace llvm;
 
@@ -65,6 +66,12 @@ bool EVMInstrInfo::analyzeBranch(MachineBasicBlock &MBB,
   FBB = nullptr;
   Cond.clear();
 
+  const auto *MFI = MBB.getParent()->getInfo<EVMMachineFunctionInfo>();
+  if (MFI->getIsStackified()) {
+    LLVM_DEBUG(dbgs() << "Can't analyze terminators in stackified code");
+    return true;
+  }
+
   // Iterate backwards and analyze all terminators.
   MachineBasicBlock::reverse_iterator I = MBB.rbegin(), E = MBB.rend();
   while (I != E) {

llvm/lib/Target/EVM/EVMInstrInfo.td

@@ -794,7 +794,7 @@ foreach I = {1-16} in {
 
 defm PUSH0 : I<(outs), (ins), [], "PUSH0", "", 0x5F, 2>;
 
-def PUSH8_LABEL : NI<(outs), (ins jmptarget:$dst), [], false, "", 0, 0> {
+def PUSH_LABEL : NI<(outs), (ins jmptarget:$dst), [], false, "", 0, 0> {
     let isCodeGenOnly = 1;
 }
 
@@ -1122,7 +1122,7 @@ def PUSH32_S : NI<(outs), (ins i256imm:$imm), [], true, "PUSH32 $imm",
 }
 
 // Pseudo instructions for linkage
-let isCodeGenOnly = 1 in
-defm DATA
-  : I<(outs GPR:$dst), (ins jmptarget:$reloc), [], "",
-       "", 0, 0>;
+let isCodeGenOnly = 1, BaseName = "DATA" in {
+  def DATA : NI<(outs GPR:$dst), (ins jmptarget:$reloc), [], false, "", 0, 0>;
+  def DATA_S : NI<(outs), (ins jmptarget:$reloc), [], true, "", 0, 0>;
+}

llvm/lib/Target/EVM/EVMMCInstLower.cpp

@@ -27,33 +27,23 @@ using namespace llvm;
 
 extern cl::opt<bool> EVMKeepRegisters;
 
-static void removeRegisterOperands(const MachineInstr *MI, MCInst &OutMI) {
-  // Remove all uses of stackified registers to bring the instruction format
-  // into its final stack form used throughout MC, and transition opcodes to
-  // their _S variant.
+// Stackify instruction that were not stackified before.
+// Only two instructions need to be stackified here: PUSH_LABEL and DATA_S,
+static void stackifyInstruction(const MachineInstr *MI, MCInst &OutMI) {
   if (MI->isDebugInstr() || MI->isLabel() || MI->isInlineAsm())
     return;
 
-  // Transform 'register' instruction to 'stack' one.
-  unsigned RegOpcode = OutMI.getOpcode();
-  if (RegOpcode == EVM::PUSH8_LABEL) {
-    // Replace PUSH8_LABEL with PUSH8_S opcode.
-    OutMI.setOpcode(EVM::PUSH8_S);
-  } else {
-    unsigned StackOpcode = EVM::getStackOpcode(RegOpcode);
-    OutMI.setOpcode(StackOpcode);
-  }
+  // Check there are no register operands.
+  assert(std::all_of(OutMI.begin(), OutMI.end(),
+                     [](const MCOperand &MO) { return !MO.isReg(); }));
 
-  // Remove register operands.
-  for (auto I = OutMI.getNumOperands(); I; --I) {
-    auto &MO = OutMI.getOperand(I - 1);
-    if (MO.isReg()) {
-      OutMI.erase(&MO);
-    }
-  }
-
-  if (RegOpcode == EVM::DATA)
+  // Set up final opcodes for the following codegen-only instructions.
+  unsigned Opcode = OutMI.getOpcode();
+  if (Opcode == EVM::PUSH_LABEL || Opcode == EVM::DATA_S)
     OutMI.setOpcode(EVM::PUSH4_S);
+
+  // Check that all the instructions are in the 'stack' form.
+  assert(EVM::getRegisterOpcode(OutMI.getOpcode()));
 }
 
 MCSymbol *
@@ -136,7 +126,7 @@ void EVMMCInstLower::Lower(const MachineInstr *MI, MCInst &OutMI) {
     case MachineOperand::MO_MCSymbol: {
       MCSymbolRefExpr::VariantKind Kind = MCSymbolRefExpr::VariantKind::VK_None;
       unsigned Opc = MI->getOpcode();
-      if (Opc == EVM::DATA)
+      if (Opc == EVM::DATA_S)
         Kind = MCSymbolRefExpr::VariantKind::VK_EVM_DATA;
 
       MCOp = MCOperand::createExpr(
@@ -156,7 +146,7 @@ void EVMMCInstLower::Lower(const MachineInstr *MI, MCInst &OutMI) {
     OutMI.addOperand(MCOp);
   }
   if (!EVMKeepRegisters)
-    removeRegisterOperands(MI, OutMI);
+    stackifyInstruction(MI, OutMI);
   else if (Desc.variadicOpsAreDefs())
     OutMI.insert(OutMI.begin(), MCOperand::createImm(MI->getNumExplicitDefs()));
 }

llvm/lib/Target/EVM/EVMMachineFunctionInfo.cpp

@@ -13,11 +13,26 @@
 #include "EVMMachineFunctionInfo.h"
 using namespace llvm;
 
-EVMFunctionInfo::~EVMFunctionInfo() = default; // anchor.
+EVMMachineFunctionInfo::~EVMMachineFunctionInfo() = default;
 
-MachineFunctionInfo *
-EVMFunctionInfo::clone(BumpPtrAllocator &Allocator, MachineFunction &DestMF,
-                       const DenseMap<MachineBasicBlock *, MachineBasicBlock *>
-                           &Src2DstMBB) const {
-  return DestMF.cloneInfo<EVMFunctionInfo>(*this);
+MachineFunctionInfo *EVMMachineFunctionInfo::clone(
+    BumpPtrAllocator &Allocator, MachineFunction &DestMF,
+    const DenseMap<MachineBasicBlock *, MachineBasicBlock *> &Src2DstMBB)
+    const {
+  return DestMF.cloneInfo<EVMMachineFunctionInfo>(*this);
+}
+
+yaml::EVMMachineFunctionInfo::~EVMMachineFunctionInfo() = default;
+
+yaml::EVMMachineFunctionInfo::EVMMachineFunctionInfo(
+    const llvm::EVMMachineFunctionInfo &MFI)
+    : IsStackified(MFI.getIsStackified()) {}
+
+void yaml::EVMMachineFunctionInfo::mappingImpl(yaml::IO &YamlIO) {
+  MappingTraits<EVMMachineFunctionInfo>::mapping(YamlIO, *this);
+}
+
+void EVMMachineFunctionInfo::initializeBaseYamlFields(
+    const yaml::EVMMachineFunctionInfo &YamlMFI) {
+  IsStackified = YamlMFI.IsStackified;
 }

llvm/lib/Target/EVM/EVMMachineFunctionInfo.h

@@ -16,13 +16,33 @@
 #include "MCTargetDesc/EVMMCTargetDesc.h"
 #include "llvm/CodeGen/MIRYamlMapping.h"
 #include "llvm/CodeGen/MachineRegisterInfo.h"
-#include "llvm/MC/MCSymbolWasm.h"
 
 namespace llvm {
 
+class EVMMachineFunctionInfo;
+
+namespace yaml {
+
+struct EVMMachineFunctionInfo final : public yaml::MachineFunctionInfo {
+  bool IsStackified = false;
+
+  EVMMachineFunctionInfo() = default;
+  explicit EVMMachineFunctionInfo(const llvm::EVMMachineFunctionInfo &MFI);
+  ~EVMMachineFunctionInfo() override;
+
+  void mappingImpl(yaml::IO &YamlIO) override;
+};
+
+template <> struct MappingTraits<EVMMachineFunctionInfo> {
+  static void mapping(IO &YamlIO, EVMMachineFunctionInfo &MFI) {
+    YamlIO.mapOptional("isStackified", MFI.IsStackified, false);
+  }
+};
+} // end namespace yaml
+
 /// This class is derived from MachineFunctionInfo and contains private
 /// EVM-specific information for each MachineFunction.
-class EVMFunctionInfo final : public MachineFunctionInfo {
+class EVMMachineFunctionInfo final : public MachineFunctionInfo {
   /// A mapping from CodeGen vreg index to a boolean value indicating whether
   /// the given register is considered to be "stackified", meaning it has been
   /// determined or made to meet the stack requirements:
@@ -34,16 +54,21 @@ class EVMFunctionInfo final : public MachineFunctionInfo {
   /// Number of parameters. Their type doesn't matter as it always is i256.
   unsigned NumberOfParameters = 0;
 
+  /// If the MF's instructions are in 'stack' form.
+  bool IsStackified = false;
+
 public:
-  explicit EVMFunctionInfo(MachineFunction &MF) {}
-  EVMFunctionInfo(const Function &F, const TargetSubtargetInfo *STI) {}
-  ~EVMFunctionInfo() override;
+  explicit EVMMachineFunctionInfo(MachineFunction &MF) {}
+  EVMMachineFunctionInfo(const Function &F, const TargetSubtargetInfo *STI) {}
+  ~EVMMachineFunctionInfo() override;
 
   MachineFunctionInfo *
   clone(BumpPtrAllocator &Allocator, MachineFunction &DestMF,
         const DenseMap<MachineBasicBlock *, MachineBasicBlock *> &Src2DstMBB)
       const override;
 
+  void initializeBaseYamlFields(const yaml::EVMMachineFunctionInfo &YamlMFI);
+
   void stackifyVReg(MachineRegisterInfo &MRI, unsigned VReg) {
     assert(MRI.getUniqueVRegDef(VReg));
     auto I = Register::virtReg2Index(VReg);
@@ -72,7 +97,12 @@ class EVMFunctionInfo final : public MachineFunctionInfo {
   unsigned getNumParams() const {
     return NumberOfParameters;
   }
+
+  void setIsStackified(bool Val = true) { IsStackified = Val; }
+
+  bool getIsStackified() const { return IsStackified; }
 };
+
 } // end namespace llvm
 
 #endif // LLVM_LIB_TARGET_EVM_EVMMACHINEFUNCTIONINFO_H

llvm/lib/Target/EVM/EVMRegColoring.cpp

@@ -80,7 +80,7 @@ bool EVMRegColoring::runOnMachineFunction(MachineFunction &MF) {
   LiveIntervals *Liveness = &getAnalysis<LiveIntervals>();
   const MachineBlockFrequencyInfo *MBFI =
       &getAnalysis<MachineBlockFrequencyInfo>();
-  EVMFunctionInfo &MFI = *MF.getInfo<EVMFunctionInfo>();
+  EVMMachineFunctionInfo &MFI = *MF.getInfo<EVMMachineFunctionInfo>();
 
   // Gather all register intervals into a list and sort them.
   unsigned NumVRegs = MRI->getNumVirtRegs();

llvm/lib/Target/EVM/EVMSingleUseExpression.cpp

@@ -268,7 +268,8 @@ static bool shouldRematerialize(const MachineInstr &Def,
 // TODO: Compute memory dependencies in a way that uses AliasAnalysis to be
 // more precise.
 static bool isSafeToMove(const MachineOperand *Def, const MachineOperand *Use,
-                         const MachineInstr *Insert, const EVMFunctionInfo &MFI,
+                         const MachineInstr *Insert,
+                         const EVMMachineFunctionInfo &MFI,
                          const MachineRegisterInfo &MRI) {
   const MachineInstr *DefI = Def->getParent();
   const MachineInstr *UseI = Use->getParent();
@@ -390,7 +391,7 @@ static void shrinkToUses(LiveInterval &LI, LiveIntervals &LIS) {
 static MachineInstr *moveForSingleUse(unsigned Reg, MachineOperand &Op,
                                       MachineInstr *Def, MachineBasicBlock &MBB,
                                       MachineInstr *Insert, LiveIntervals &LIS,
-                                      EVMFunctionInfo &MFI,
+                                      EVMMachineFunctionInfo &MFI,
                                       MachineRegisterInfo &MRI) {
   LLVM_DEBUG(dbgs() << "Move for single use: "; Def->dump());
 
@@ -430,8 +431,8 @@ static MachineInstr *moveForSingleUse(unsigned Reg, MachineOperand &Op,
 static MachineInstr *rematerializeCheapDef(
     unsigned Reg, MachineOperand &Op, MachineInstr &Def, MachineBasicBlock &MBB,
     MachineBasicBlock::instr_iterator Insert, LiveIntervals &LIS,
-    EVMFunctionInfo &MFI, MachineRegisterInfo &MRI, const EVMInstrInfo *TII,
-    const EVMRegisterInfo *TRI) {
+    EVMMachineFunctionInfo &MFI, MachineRegisterInfo &MRI,
+    const EVMInstrInfo *TII, const EVMRegisterInfo *TRI) {
   LLVM_DEBUG(dbgs() << "Rematerializing cheap def: "; Def.dump());
   LLVM_DEBUG(dbgs() << " - for use in "; Op.getParent()->dump());
 
@@ -593,7 +594,7 @@ bool EVMSingleUseExpression::runOnMachineFunction(MachineFunction &MF) {
 
   bool Changed = false;
   MachineRegisterInfo &MRI = MF.getRegInfo();
-  EVMFunctionInfo &MFI = *MF.getInfo<EVMFunctionInfo>();
+  EVMMachineFunctionInfo &MFI = *MF.getInfo<EVMMachineFunctionInfo>();
   const auto *TII = MF.getSubtarget<EVMSubtarget>().getInstrInfo();
   const auto *TRI = MF.getSubtarget<EVMSubtarget>().getRegisterInfo();
   auto &MDT = getAnalysis<MachineDominatorTree>();