lib/AsmParser/LLParser.h

   1 //===-- LLParser.h - Parser Class -------------------------------*- C++ -*-===//
   2 //
   3 //                     The LLVM Compiler Infrastructure
   4 //
   5 // This file is distributed under the University of Illinois Open Source
   6 // License. See LICENSE.TXT for details.
   7 //
   8 //===----------------------------------------------------------------------===//
   9 //
  10 //  This file defines the parser class for .ll files.
  11 //
  12 //===----------------------------------------------------------------------===//
  13
  14 #ifndef LLVM_ASMPARSER_LLPARSER_H
  15 #define LLVM_ASMPARSER_LLPARSER_H
  16
  17 #include "LLLexer.h"
  18 #include "llvm/Module.h"
  19 #include "llvm/Type.h"
  20 #include "llvm/ADT/DenseMap.h"
  21 #include "llvm/ADT/StringMap.h"
  22 #include "llvm/Support/ValueHandle.h"
  23 #include <map>
  24
  25 namespace llvm {
  26   class Module;
  27   class OpaqueType;
  28   class Function;
  29   class Value;
  30   class BasicBlock;
  31   class Instruction;
  32   class Constant;
  33   class GlobalValue;
  34   class MDString;
  35   class MDNode;
  36   class StructType;
  37
  38   /// ValID - Represents a reference of a definition of some sort with no type.
  39   /// There are several cases where we have to parse the value but where the
  40   /// type can depend on later context.  This may either be a numeric reference
  41   /// or a symbolic (%var) reference.  This is just a discriminated union.
  42   struct ValID {
  43     enum {
  44       t_LocalID, t_GlobalID,      // ID in UIntVal.
  45       t_LocalName, t_GlobalName,  // Name in StrVal.
  46       t_APSInt, t_APFloat,        // Value in APSIntVal/APFloatVal.
  47       t_Null, t_Undef, t_Zero,    // No value.
  48       t_EmptyArray,               // No value:  []
  49       t_Constant,                 // Value in ConstantVal.
  50       t_InlineAsm,                // Value in StrVal/StrVal2/UIntVal.
  51       t_MDNode,                   // Value in MDNodeVal.
  52       t_MDString,                 // Value in MDStringVal.
  53       t_ConstantStruct,           // Value in ConstantStructElts.
  54       t_PackedConstantStruct      // Value in ConstantStructElts.
  55     } Kind;
  56
  57     LLLexer::LocTy Loc;
  58     unsigned UIntVal;
  59     std::string StrVal, StrVal2;
  60     APSInt APSIntVal;
  61     APFloat APFloatVal;
  62     Constant *ConstantVal;
  63     MDNode *MDNodeVal;
  64     MDString *MDStringVal;
  65     Constant **ConstantStructElts;
  66
  67     ValID() : Kind(t_LocalID), APFloatVal(0.0) {}
  68     ~ValID() {
  69       if (Kind == t_ConstantStruct || Kind == t_PackedConstantStruct)
  70         delete [] ConstantStructElts;
  71     }
  72
  73     bool operator<(const ValID &RHS) const {
  74       if (Kind == t_LocalID || Kind == t_GlobalID)
  75         return UIntVal < RHS.UIntVal;
  76       assert((Kind == t_LocalName || Kind == t_GlobalName ||
  77               Kind == t_ConstantStruct || Kind == t_PackedConstantStruct) &&
  78              "Ordering not defined for this ValID kind yet");
  79       return StrVal < RHS.StrVal;
  80     }
  81   };
  82
  83   class LLParser {
  84   public:
  85     typedef LLLexer::LocTy LocTy;
  86   private:
  87     LLVMContext &Context;
  88     LLLexer Lex;
  89     Module *M;
  90
  91     // Instruction metadata resolution.  Each instruction can have a list of
  92     // MDRef info associated with them.
  93     //
  94     // The simpler approach of just creating temporary MDNodes and then calling
  95     // RAUW on them when the definition is processed doesn't work because some
  96     // instruction metadata kinds, such as dbg, get stored in the IR in an
  97     // "optimized" format which doesn't participate in the normal value use
  98     // lists. This means that RAUW doesn't work, even on temporary MDNodes
  99     // which otherwise support RAUW. Instead, we defer resolving MDNode
 100     // references until the definitions have been processed.
 101     struct MDRef {
 102       SMLoc Loc;
 103       unsigned MDKind, MDSlot;
 104     };
 105     DenseMap<Instruction*, std::vector<MDRef> > ForwardRefInstMetadata;
 106
 107     // Type resolution handling data structures.  The location is set when we
 108     // have processed a use of the type but not a definition yet.
 109     StringMap<std::pair<Type*, LocTy> > NamedTypes;
 110     std::vector<std::pair<Type*, LocTy> > NumberedTypes;
 111
 112     std::vector<TrackingVH<MDNode> > NumberedMetadata;
 113     std::map<unsigned, std::pair<TrackingVH<MDNode>, LocTy> > ForwardRefMDNodes;
 114
 115     // Global Value reference information.
 116     std::map<std::string, std::pair<GlobalValue*, LocTy> > ForwardRefVals;
 117     std::map<unsigned, std::pair<GlobalValue*, LocTy> > ForwardRefValIDs;
 118     std::vector<GlobalValue*> NumberedVals;
 119
 120     // References to blockaddress.  The key is the function ValID, the value is
 121     // a list of references to blocks in that function.
 122     std::map<ValID, std::vector<std::pair<ValID, GlobalValue*> > >
 123       ForwardRefBlockAddresses;
 124
 125   public:
 126     LLParser(MemoryBuffer *F, SourceMgr &SM, SMDiagnostic &Err, Module *m) :
 127       Context(m->getContext()), Lex(F, SM, Err, m->getContext()),
 128       M(m) {}
 129     bool Run();
 130
 131     LLVMContext &getContext() { return Context; }
 132
 133   private:
 134
 135     bool Error(LocTy L, const Twine &Msg) const {
 136       return Lex.Error(L, Msg);
 137     }
 138     bool TokError(const Twine &Msg) const {
 139       return Error(Lex.getLoc(), Msg);
 140     }
 141
 142     /// GetGlobalVal - Get a value with the specified name or ID, creating a
 143     /// forward reference record if needed.  This can return null if the value
 144     /// exists but does not have the right type.
 145     GlobalValue *GetGlobalVal(const std::string &N, const Type *Ty, LocTy Loc);
 146     GlobalValue *GetGlobalVal(unsigned ID, const Type *Ty, LocTy Loc);
 147
 148     // Helper Routines.
 149     bool ParseToken(lltok::Kind T, const char *ErrMsg);
 150     bool EatIfPresent(lltok::Kind T) {
 151       if (Lex.getKind() != T) return false;
 152       Lex.Lex();
 153       return true;
 154     }
 155     bool ParseOptionalToken(lltok::Kind T, bool &Present, LocTy *Loc = 0) {
 156       if (Lex.getKind() != T) {
 157         Present = false;
 158       } else {
 159         if (Loc)
 160           *Loc = Lex.getLoc();
 161         Lex.Lex();
 162         Present = true;
 163       }
 164       return false;
 165     }
 166     bool ParseStringConstant(std::string &Result);
 167     bool ParseUInt32(unsigned &Val);
 168     bool ParseUInt32(unsigned &Val, LocTy &Loc) {
 169       Loc = Lex.getLoc();
 170       return ParseUInt32(Val);
 171     }
 172     bool ParseOptionalAddrSpace(unsigned &AddrSpace);
 173     bool ParseOptionalAttrs(unsigned &Attrs, unsigned AttrKind);
 174     bool ParseOptionalLinkage(unsigned &Linkage, bool &HasLinkage);
 175     bool ParseOptionalLinkage(unsigned &Linkage) {
 176       bool HasLinkage; return ParseOptionalLinkage(Linkage, HasLinkage);
 177     }
 178     bool ParseOptionalVisibility(unsigned &Visibility);
 179     bool ParseOptionalCallingConv(CallingConv::ID &CC);
 180     bool ParseOptionalAlignment(unsigned &Alignment);
 181     bool ParseOptionalStackAlignment(unsigned &Alignment);
 182     bool ParseOptionalCommaAlign(unsigned &Alignment, bool &AteExtraComma);
 183     bool ParseIndexList(SmallVectorImpl<unsigned> &Indices,bool &AteExtraComma);
 184     bool ParseIndexList(SmallVectorImpl<unsigned> &Indices) {
 185       bool AteExtraComma;
 186       if (ParseIndexList(Indices, AteExtraComma)) return true;
 187       if (AteExtraComma)
 188         return TokError("expected index");
 189       return false;
 190     }
 191
 192     // Top-Level Entities
 193     bool ParseTopLevelEntities();
 194     bool ValidateEndOfModule();
 195     bool ParseTargetDefinition();
 196     bool ParseDepLibs();
 197     bool ParseModuleAsm();
 198     bool ParseUnnamedType();
 199     bool ParseNamedType();
 200     bool ParseDeclare();
 201     bool ParseDefine();
 202
 203     bool ParseGlobalType(bool &IsConstant);
 204     bool ParseUnnamedGlobal();
 205     bool ParseNamedGlobal();
 206     bool ParseGlobal(const std::string &Name, LocTy Loc, unsigned Linkage,
 207                      bool HasLinkage, unsigned Visibility);
 208     bool ParseAlias(const std::string &Name, LocTy Loc, unsigned Visibility);
 209     bool ParseStandaloneMetadata();
 210     bool ParseNamedMetadata();
 211     bool ParseMDString(MDString *&Result);
 212     bool ParseMDNodeID(MDNode *&Result);
 213     bool ParseMDNodeID(MDNode *&Result, unsigned &SlotNo);
 214
 215     // Type Parsing.
 216     bool ParseType(Type *&Result, bool AllowVoid = false);
 217     bool ParseType(Type *&Result, LocTy &Loc, bool AllowVoid = false) {
 218       Loc = Lex.getLoc();
 219       return ParseType(Result, AllowVoid);
 220     }
 221     bool ParseAnonStructType(Type *&Result, bool Packed);
 222     bool ParseStructBody(SmallVectorImpl<Type*> &Body);
 223     bool ParseStructDefinition(SMLoc TypeLoc, StringRef Name,
 224                                std::pair<Type*, LocTy> &Entry,
 225                                Type *&ResultTy);
 226
 227     bool ParseArrayVectorType(Type *&Result, bool isVector);
 228     bool ParseFunctionType(Type *&Result);
 229
 230     // Function Semantic Analysis.
 231     class PerFunctionState {
 232       LLParser &P;
 233       Function &F;
 234       std::map<std::string, std::pair<Value*, LocTy> > ForwardRefVals;
 235       std::map<unsigned, std::pair<Value*, LocTy> > ForwardRefValIDs;
 236       std::vector<Value*> NumberedVals;
 237
 238       /// FunctionNumber - If this is an unnamed function, this is the slot
 239       /// number of it, otherwise it is -1.
 240       int FunctionNumber;
 241     public:
 242       PerFunctionState(LLParser &p, Function &f, int FunctionNumber);
 243       ~PerFunctionState();
 244
 245       Function &getFunction() const { return F; }
 246
 247       bool FinishFunction();
 248
 249       /// GetVal - Get a value with the specified name or ID, creating a
 250       /// forward reference record if needed.  This can return null if the value
 251       /// exists but does not have the right type.
 252       Value *GetVal(const std::string &Name, const Type *Ty, LocTy Loc);
 253       Value *GetVal(unsigned ID, const Type *Ty, LocTy Loc);
 254
 255       /// SetInstName - After an instruction is parsed and inserted into its
 256       /// basic block, this installs its name.
 257       bool SetInstName(int NameID, const std::string &NameStr, LocTy NameLoc,
 258                        Instruction *Inst);
 259
 260       /// GetBB - Get a basic block with the specified name or ID, creating a
 261       /// forward reference record if needed.  This can return null if the value
 262       /// is not a BasicBlock.
 263       BasicBlock *GetBB(const std::string &Name, LocTy Loc);
 264       BasicBlock *GetBB(unsigned ID, LocTy Loc);
 265
 266       /// DefineBB - Define the specified basic block, which is either named or
 267       /// unnamed.  If there is an error, this returns null otherwise it returns
 268       /// the block being defined.
 269       BasicBlock *DefineBB(const std::string &Name, LocTy Loc);
 270     };
 271
 272     bool ConvertValIDToValue(const Type *Ty, ValID &ID, Value *&V,
 273                              PerFunctionState *PFS);
 274
 275     bool ParseValue(const Type *Ty, Value *&V, PerFunctionState *PFS);
 276     bool ParseValue(const Type *Ty, Value *&V, PerFunctionState &PFS) {
 277       return ParseValue(Ty, V, &PFS);
 278     }
 279     bool ParseValue(const Type *Ty, Value *&V, LocTy &Loc,
 280                     PerFunctionState &PFS) {
 281       Loc = Lex.getLoc();
 282       return ParseValue(Ty, V, &PFS);
 283     }
 284
 285     bool ParseTypeAndValue(Value *&V, PerFunctionState *PFS);
 286     bool ParseTypeAndValue(Value *&V, PerFunctionState &PFS) {
 287       return ParseTypeAndValue(V, &PFS);
 288     }
 289     bool ParseTypeAndValue(Value *&V, LocTy &Loc, PerFunctionState &PFS) {
 290       Loc = Lex.getLoc();
 291       return ParseTypeAndValue(V, PFS);
 292     }
 293     bool ParseTypeAndBasicBlock(BasicBlock *&BB, LocTy &Loc,
 294                                 PerFunctionState &PFS);
 295     bool ParseTypeAndBasicBlock(BasicBlock *&BB, PerFunctionState &PFS) {
 296       LocTy Loc;
 297       return ParseTypeAndBasicBlock(BB, Loc, PFS);
 298     }
 299
 300
 301     struct ParamInfo {
 302       LocTy Loc;
 303       Value *V;
 304       unsigned Attrs;
 305       ParamInfo(LocTy loc, Value *v, unsigned attrs)
 306         : Loc(loc), V(v), Attrs(attrs) {}
 307     };
 308     bool ParseParameterList(SmallVectorImpl<ParamInfo> &ArgList,
 309                             PerFunctionState &PFS);
 310
 311     // Constant Parsing.
 312     bool ParseValID(ValID &ID, PerFunctionState *PFS = NULL);
 313     bool ParseGlobalValue(const Type *Ty, Constant *&V);
 314     bool ParseGlobalTypeAndValue(Constant *&V);
 315     bool ParseGlobalValueVector(SmallVectorImpl<Constant*> &Elts);
 316     bool ParseMetadataListValue(ValID &ID, PerFunctionState *PFS);
 317     bool ParseMetadataValue(ValID &ID, PerFunctionState *PFS);
 318     bool ParseMDNodeVector(SmallVectorImpl<Value*> &, PerFunctionState *PFS);
 319     bool ParseInstructionMetadata(Instruction *Inst, PerFunctionState *PFS);
 320
 321     // Function Parsing.
 322     struct ArgInfo {
 323       LocTy Loc;
 324       Type *Ty;
 325       unsigned Attrs;
 326       std::string Name;
 327       ArgInfo(LocTy L, Type *ty, unsigned Attr, const std::string &N)
 328         : Loc(L), Ty(ty), Attrs(Attr), Name(N) {}
 329     };
 330     bool ParseArgumentList(SmallVectorImpl<ArgInfo> &ArgList, bool &isVarArg);
 331     bool ParseFunctionHeader(Function *&Fn, bool isDefine);
 332     bool ParseFunctionBody(Function &Fn);
 333     bool ParseBasicBlock(PerFunctionState &PFS);
 334
 335     // Instruction Parsing.  Each instruction parsing routine can return with a
 336     // normal result, an error result, or return having eaten an extra comma.
 337     enum InstResult { InstNormal = 0, InstError = 1, InstExtraComma = 2 };
 338     int ParseInstruction(Instruction *&Inst, BasicBlock *BB,
 339                          PerFunctionState &PFS);
 340     bool ParseCmpPredicate(unsigned &Pred, unsigned Opc);
 341
 342     bool ParseRet(Instruction *&Inst, BasicBlock *BB, PerFunctionState &PFS);
 343     bool ParseBr(Instruction *&Inst, PerFunctionState &PFS);
 344     bool ParseSwitch(Instruction *&Inst, PerFunctionState &PFS);
 345     bool ParseIndirectBr(Instruction *&Inst, PerFunctionState &PFS);
 346     bool ParseInvoke(Instruction *&Inst, PerFunctionState &PFS);
 347
 348     bool ParseArithmetic(Instruction *&I, PerFunctionState &PFS, unsigned Opc,
 349                          unsigned OperandType);
 350     bool ParseLogical(Instruction *&I, PerFunctionState &PFS, unsigned Opc);
 351     bool ParseCompare(Instruction *&I, PerFunctionState &PFS, unsigned Opc);
 352     bool ParseCast(Instruction *&I, PerFunctionState &PFS, unsigned Opc);
 353     bool ParseSelect(Instruction *&I, PerFunctionState &PFS);
 354     bool ParseVA_Arg(Instruction *&I, PerFunctionState &PFS);
 355     bool ParseExtractElement(Instruction *&I, PerFunctionState &PFS);
 356     bool ParseInsertElement(Instruction *&I, PerFunctionState &PFS);
 357     bool ParseShuffleVector(Instruction *&I, PerFunctionState &PFS);
 358     int ParsePHI(Instruction *&I, PerFunctionState &PFS);
 359     bool ParseCall(Instruction *&I, PerFunctionState &PFS, bool isTail);
 360     int ParseAlloc(Instruction *&I, PerFunctionState &PFS);
 361     int ParseLoad(Instruction *&I, PerFunctionState &PFS, bool isVolatile);
 362     int ParseStore(Instruction *&I, PerFunctionState &PFS, bool isVolatile);
 363     int ParseGetElementPtr(Instruction *&I, PerFunctionState &PFS);
 364     int ParseExtractValue(Instruction *&I, PerFunctionState &PFS);
 365     int ParseInsertValue(Instruction *&I, PerFunctionState &PFS);
 366
 367     bool ResolveForwardRefBlockAddresses(Function *TheFn,
 368                              std::vector<std::pair<ValID, GlobalValue*> > &Refs,
 369                                          PerFunctionState *PFS);
 370   };
 371 } // End llvm namespace
 372
 373 #endif