lld/COFF/Symbols.h

   1 //===- Symbols.h ------------------------------------------------*- C++ -*-===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8
   9 #ifndef LLD_COFF_SYMBOLS_H
  10 #define LLD_COFF_SYMBOLS_H
  11
  12 #include "Chunks.h"
  13 #include "Config.h"
  14 #include "lld/Common/LLVM.h"
  15 #include "lld/Common/Memory.h"
  16 #include "llvm/ADT/ArrayRef.h"
  17 #include "llvm/Object/Archive.h"
  18 #include "llvm/Object/COFF.h"
  19 #include <atomic>
  20 #include <memory>
  21 #include <vector>
  22
  23 namespace lld {
  24
  25 std::string toString(coff::Symbol &b);
  26
  27 // There are two different ways to convert an Archive::Symbol to a string:
  28 // One for Microsoft name mangling and one for Itanium name mangling.
  29 // Call the functions toCOFFString and toELFString, not just toString.
  30 std::string toCOFFString(const coff::Archive::Symbol &b);
  31
  32 namespace coff {
  33
  34 using llvm::object::Archive;
  35 using llvm::object::COFFSymbolRef;
  36 using llvm::object::coff_import_header;
  37 using llvm::object::coff_symbol_generic;
  38
  39 class ArchiveFile;
  40 class InputFile;
  41 class ObjFile;
  42 class SymbolTable;
  43
  44 // The base class for real symbol classes.
  45 class Symbol {
  46 public:
  47   enum Kind {
  48     // The order of these is significant. We start with the regular defined
  49     // symbols as those are the most prevalent and the zero tag is the cheapest
  50     // to set. Among the defined kinds, the lower the kind is preferred over
  51     // the higher kind when testing whether one symbol should take precedence
  52     // over another.
  53     DefinedRegularKind = 0,
  54     DefinedCommonKind,
  55     DefinedLocalImportKind,
  56     DefinedImportThunkKind,
  57     DefinedImportDataKind,
  58     DefinedAbsoluteKind,
  59     DefinedSyntheticKind,
  60
  61     UndefinedKind,
  62     LazyArchiveKind,
  63     LazyObjectKind,
  64     LazyDLLSymbolKind,
  65
  66     LastDefinedCOFFKind = DefinedCommonKind,
  67     LastDefinedKind = DefinedSyntheticKind,
  68   };
  69
  70   Kind kind() const { return static_cast<Kind>(symbolKind); }
  71
  72   // Returns the symbol name.
  73   StringRef getName() {
  74     // COFF symbol names are read lazily for a performance reason.
  75     // Non-external symbol names are never used by the linker except for logging
  76     // or debugging. Their internal references are resolved not by name but by
  77     // symbol index. And because they are not external, no one can refer them by
  78     // name. Object files contain lots of non-external symbols, and creating
  79     // StringRefs for them (which involves lots of strlen() on the string table)
  80     // is a waste of time.
  81     if (nameData == nullptr)
  82       computeName();
  83     return StringRef(nameData, nameSize);
  84   }
  85
  86   void replaceKeepingName(Symbol *other, size_t size);
  87
  88   // Returns the file from which this symbol was created.
  89   InputFile *getFile();
  90
  91   // Indicates that this symbol will be included in the final image. Only valid
  92   // after calling markLive.
  93   bool isLive() const;
  94
  95   bool isLazy() const {
  96     return symbolKind == LazyArchiveKind || symbolKind == LazyObjectKind ||
  97            symbolKind == LazyDLLSymbolKind;
  98   }
  99
 100 private:
 101   void computeName();
 102
 103 protected:
 104   friend SymbolTable;
 105   explicit Symbol(Kind k, StringRef n = "")
 106       : symbolKind(k), isExternal(true), isCOMDAT(false),
 107         writtenToSymtab(false), pendingArchiveLoad(false), isGCRoot(false),
 108         isRuntimePseudoReloc(false), deferUndefined(false), canInline(true),
 109         nameSize(n.size()), nameData(n.empty() ? nullptr : n.data()) {}
 110
 111   const unsigned symbolKind : 8;
 112   unsigned isExternal : 1;
 113
 114 public:
 115   // This bit is used by the \c DefinedRegular subclass.
 116   unsigned isCOMDAT : 1;
 117
 118   // This bit is used by Writer::createSymbolAndStringTable() to prevent
 119   // symbols from being written to the symbol table more than once.
 120   unsigned writtenToSymtab : 1;
 121
 122   // True if this symbol was referenced by a regular (non-bitcode) object.
 123   unsigned isUsedInRegularObj : 1;
 124
 125   // True if we've seen both a lazy and an undefined symbol with this symbol
 126   // name, which means that we have enqueued an archive member load and should
 127   // not load any more archive members to resolve the same symbol.
 128   unsigned pendingArchiveLoad : 1;
 129
 130   /// True if we've already added this symbol to the list of GC roots.
 131   unsigned isGCRoot : 1;
 132
 133   unsigned isRuntimePseudoReloc : 1;
 134
 135   // True if we want to allow this symbol to be undefined in the early
 136   // undefined check pass in SymbolTable::reportUnresolvable(), as it
 137   // might be fixed up later.
 138   unsigned deferUndefined : 1;
 139
 140   // False if LTO shouldn't inline whatever this symbol points to. If a symbol
 141   // is overwritten after LTO, LTO shouldn't inline the symbol because it
 142   // doesn't know the final contents of the symbol.
 143   unsigned canInline : 1;
 144
 145 protected:
 146   // Symbol name length. Assume symbol lengths fit in a 32-bit integer.
 147   uint32_t nameSize;
 148
 149   const char *nameData;
 150 };
 151
 152 // The base class for any defined symbols, including absolute symbols,
 153 // etc.
 154 class Defined : public Symbol {
 155 public:
 156   Defined(Kind k, StringRef n) : Symbol(k, n) {}
 157
 158   static bool classof(const Symbol *s) { return s->kind() <= LastDefinedKind; }
 159
 160   // Returns the RVA (relative virtual address) of this symbol. The
 161   // writer sets and uses RVAs.
 162   uint64_t getRVA();
 163
 164   // Returns the chunk containing this symbol. Absolute symbols and __ImageBase
 165   // do not have chunks, so this may return null.
 166   Chunk *getChunk();
 167 };
 168
 169 // Symbols defined via a COFF object file or bitcode file.  For COFF files, this
 170 // stores a coff_symbol_generic*, and names of internal symbols are lazily
 171 // loaded through that. For bitcode files, Sym is nullptr and the name is stored
 172 // as a decomposed StringRef.
 173 class DefinedCOFF : public Defined {
 174   friend Symbol;
 175
 176 public:
 177   DefinedCOFF(Kind k, InputFile *f, StringRef n, const coff_symbol_generic *s)
 178       : Defined(k, n), file(f), sym(s) {}
 179
 180   static bool classof(const Symbol *s) {
 181     return s->kind() <= LastDefinedCOFFKind;
 182   }
 183
 184   InputFile *getFile() { return file; }
 185
 186   COFFSymbolRef getCOFFSymbol();
 187
 188   InputFile *file;
 189
 190 protected:
 191   const coff_symbol_generic *sym;
 192 };
 193
 194 // Regular defined symbols read from object file symbol tables.
 195 class DefinedRegular : public DefinedCOFF {
 196 public:
 197   DefinedRegular(InputFile *f, StringRef n, bool isCOMDAT,
 198                  bool isExternal = false,
 199                  const coff_symbol_generic *s = nullptr,
 200                  SectionChunk *c = nullptr)
 201       : DefinedCOFF(DefinedRegularKind, f, n, s), data(c ? &c->repl : nullptr) {
 202     this->isExternal = isExternal;
 203     this->isCOMDAT = isCOMDAT;
 204   }
 205
 206   static bool classof(const Symbol *s) {
 207     return s->kind() == DefinedRegularKind;
 208   }
 209
 210   uint64_t getRVA() const { return (*data)->getRVA() + sym->Value; }
 211   SectionChunk *getChunk() const { return *data; }
 212   uint32_t getValue() const { return sym->Value; }
 213
 214   SectionChunk **data;
 215 };
 216
 217 class DefinedCommon : public DefinedCOFF {
 218 public:
 219   DefinedCommon(InputFile *f, StringRef n, uint64_t size,
 220                 const coff_symbol_generic *s = nullptr,
 221                 CommonChunk *c = nullptr)
 222       : DefinedCOFF(DefinedCommonKind, f, n, s), data(c), size(size) {
 223     this->isExternal = true;
 224   }
 225
 226   static bool classof(const Symbol *s) {
 227     return s->kind() == DefinedCommonKind;
 228   }
 229
 230   uint64_t getRVA() { return data->getRVA(); }
 231   CommonChunk *getChunk() { return data; }
 232
 233 private:
 234   friend SymbolTable;
 235   uint64_t getSize() const { return size; }
 236   CommonChunk *data;
 237   uint64_t size;
 238 };
 239
 240 // Absolute symbols.
 241 class DefinedAbsolute : public Defined {
 242 public:
 243   DefinedAbsolute(StringRef n, COFFSymbolRef s)
 244       : Defined(DefinedAbsoluteKind, n), va(s.getValue()) {
 245     isExternal = s.isExternal();
 246   }
 247
 248   DefinedAbsolute(StringRef n, uint64_t v)
 249       : Defined(DefinedAbsoluteKind, n), va(v) {}
 250
 251   static bool classof(const Symbol *s) {
 252     return s->kind() == DefinedAbsoluteKind;
 253   }
 254
 255   uint64_t getRVA() { return va - config->imageBase; }
 256   void setVA(uint64_t v) { va = v; }
 257   uint64_t getVA() const { return va; }
 258
 259   // Section index relocations against absolute symbols resolve to
 260   // this 16 bit number, and it is the largest valid section index
 261   // plus one. This variable keeps it.
 262   static uint16_t numOutputSections;
 263
 264 private:
 265   uint64_t va;
 266 };
 267
 268 // This symbol is used for linker-synthesized symbols like __ImageBase and
 269 // __safe_se_handler_table.
 270 class DefinedSynthetic : public Defined {
 271 public:
 272   explicit DefinedSynthetic(StringRef name, Chunk *c)
 273       : Defined(DefinedSyntheticKind, name), c(c) {}
 274
 275   static bool classof(const Symbol *s) {
 276     return s->kind() == DefinedSyntheticKind;
 277   }
 278
 279   // A null chunk indicates that this is __ImageBase. Otherwise, this is some
 280   // other synthesized chunk, like SEHTableChunk.
 281   uint32_t getRVA() { return c ? c->getRVA() : 0; }
 282   Chunk *getChunk() { return c; }
 283
 284 private:
 285   Chunk *c;
 286 };
 287
 288 // This class represents a symbol defined in an archive file. It is
 289 // created from an archive file header, and it knows how to load an
 290 // object file from an archive to replace itself with a defined
 291 // symbol. If the resolver finds both Undefined and LazyArchive for
 292 // the same name, it will ask the LazyArchive to load a file.
 293 class LazyArchive : public Symbol {
 294 public:
 295   LazyArchive(ArchiveFile *f, const Archive::Symbol s)
 296       : Symbol(LazyArchiveKind, s.getName()), file(f), sym(s) {}
 297
 298   static bool classof(const Symbol *s) { return s->kind() == LazyArchiveKind; }
 299
 300   MemoryBufferRef getMemberBuffer();
 301
 302   ArchiveFile *file;
 303   const Archive::Symbol sym;
 304 };
 305
 306 class LazyObject : public Symbol {
 307 public:
 308   LazyObject(LazyObjFile *f, StringRef n)
 309       : Symbol(LazyObjectKind, n), file(f) {}
 310   static bool classof(const Symbol *s) { return s->kind() == LazyObjectKind; }
 311   LazyObjFile *file;
 312 };
 313
 314 // MinGW only.
 315 class LazyDLLSymbol : public Symbol {
 316 public:
 317   LazyDLLSymbol(DLLFile *f, DLLFile::Symbol *s, StringRef n)
 318       : Symbol(LazyDLLSymbolKind, n), file(f), sym(s) {}
 319   static bool classof(const Symbol *s) {
 320     return s->kind() == LazyDLLSymbolKind;
 321   }
 322
 323   DLLFile *file;
 324   DLLFile::Symbol *sym;
 325 };
 326
 327 // Undefined symbols.
 328 class Undefined : public Symbol {
 329 public:
 330   explicit Undefined(StringRef n) : Symbol(UndefinedKind, n) {}
 331
 332   static bool classof(const Symbol *s) { return s->kind() == UndefinedKind; }
 333
 334   // An undefined symbol can have a fallback symbol which gives an
 335   // undefined symbol a second chance if it would remain undefined.
 336   // If it remains undefined, it'll be replaced with whatever the
 337   // Alias pointer points to.
 338   Symbol *weakAlias = nullptr;
 339
 340   // If this symbol is external weak, try to resolve it to a defined
 341   // symbol by searching the chain of fallback symbols. Returns the symbol if
 342   // successful, otherwise returns null.
 343   Defined *getWeakAlias();
 344 };
 345
 346 // Windows-specific classes.
 347
 348 // This class represents a symbol imported from a DLL. This has two
 349 // names for internal use and external use. The former is used for
 350 // name resolution, and the latter is used for the import descriptor
 351 // table in an output. The former has "__imp_" prefix.
 352 class DefinedImportData : public Defined {
 353 public:
 354   DefinedImportData(StringRef n, ImportFile *f)
 355       : Defined(DefinedImportDataKind, n), file(f) {
 356   }
 357
 358   static bool classof(const Symbol *s) {
 359     return s->kind() == DefinedImportDataKind;
 360   }
 361
 362   uint64_t getRVA() { return file->location->getRVA(); }
 363   Chunk *getChunk() { return file->location; }
 364   void setLocation(Chunk *addressTable) { file->location = addressTable; }
 365
 366   StringRef getDLLName() { return file->dllName; }
 367   StringRef getExternalName() { return file->externalName; }
 368   uint16_t getOrdinal() { return file->hdr->OrdinalHint; }
 369
 370   ImportFile *file;
 371
 372   // This is a pointer to the synthetic symbol associated with the load thunk
 373   // for this symbol that will be called if the DLL is delay-loaded. This is
 374   // needed for Control Flow Guard because if this DefinedImportData symbol is a
 375   // valid call target, the corresponding load thunk must also be marked as a
 376   // valid call target.
 377   DefinedSynthetic *loadThunkSym = nullptr;
 378 };
 379
 380 // This class represents a symbol for a jump table entry which jumps
 381 // to a function in a DLL. Linker are supposed to create such symbols
 382 // without "__imp_" prefix for all function symbols exported from
 383 // DLLs, so that you can call DLL functions as regular functions with
 384 // a regular name. A function pointer is given as a DefinedImportData.
 385 class DefinedImportThunk : public Defined {
 386 public:
 387   DefinedImportThunk(StringRef name, DefinedImportData *s, uint16_t machine);
 388
 389   static bool classof(const Symbol *s) {
 390     return s->kind() == DefinedImportThunkKind;
 391   }
 392
 393   uint64_t getRVA() { return data->getRVA(); }
 394   Chunk *getChunk() { return data; }
 395
 396   DefinedImportData *wrappedSym;
 397
 398 private:
 399   Chunk *data;
 400 };
 401
 402 // If you have a symbol "foo" in your object file, a symbol name
 403 // "__imp_foo" becomes automatically available as a pointer to "foo".
 404 // This class is for such automatically-created symbols.
 405 // Yes, this is an odd feature. We didn't intend to implement that.
 406 // This is here just for compatibility with MSVC.
 407 class DefinedLocalImport : public Defined {
 408 public:
 409   DefinedLocalImport(StringRef n, Defined *s)
 410       : Defined(DefinedLocalImportKind, n), data(make<LocalImportChunk>(s)) {}
 411
 412   static bool classof(const Symbol *s) {
 413     return s->kind() == DefinedLocalImportKind;
 414   }
 415
 416   uint64_t getRVA() { return data->getRVA(); }
 417   Chunk *getChunk() { return data; }
 418
 419 private:
 420   LocalImportChunk *data;
 421 };
 422
 423 inline uint64_t Defined::getRVA() {
 424   switch (kind()) {
 425   case DefinedAbsoluteKind:
 426     return cast<DefinedAbsolute>(this)->getRVA();
 427   case DefinedSyntheticKind:
 428     return cast<DefinedSynthetic>(this)->getRVA();
 429   case DefinedImportDataKind:
 430     return cast<DefinedImportData>(this)->getRVA();
 431   case DefinedImportThunkKind:
 432     return cast<DefinedImportThunk>(this)->getRVA();
 433   case DefinedLocalImportKind:
 434     return cast<DefinedLocalImport>(this)->getRVA();
 435   case DefinedCommonKind:
 436     return cast<DefinedCommon>(this)->getRVA();
 437   case DefinedRegularKind:
 438     return cast<DefinedRegular>(this)->getRVA();
 439   case LazyArchiveKind:
 440   case LazyObjectKind:
 441   case LazyDLLSymbolKind:
 442   case UndefinedKind:
 443     llvm_unreachable("Cannot get the address for an undefined symbol.");
 444   }
 445   llvm_unreachable("unknown symbol kind");
 446 }
 447
 448 inline Chunk *Defined::getChunk() {
 449   switch (kind()) {
 450   case DefinedRegularKind:
 451     return cast<DefinedRegular>(this)->getChunk();
 452   case DefinedAbsoluteKind:
 453     return nullptr;
 454   case DefinedSyntheticKind:
 455     return cast<DefinedSynthetic>(this)->getChunk();
 456   case DefinedImportDataKind:
 457     return cast<DefinedImportData>(this)->getChunk();
 458   case DefinedImportThunkKind:
 459     return cast<DefinedImportThunk>(this)->getChunk();
 460   case DefinedLocalImportKind:
 461     return cast<DefinedLocalImport>(this)->getChunk();
 462   case DefinedCommonKind:
 463     return cast<DefinedCommon>(this)->getChunk();
 464   case LazyArchiveKind:
 465   case LazyObjectKind:
 466   case LazyDLLSymbolKind:
 467   case UndefinedKind:
 468     llvm_unreachable("Cannot get the chunk of an undefined symbol.");
 469   }
 470   llvm_unreachable("unknown symbol kind");
 471 }
 472
 473 // A buffer class that is large enough to hold any Symbol-derived
 474 // object. We allocate memory using this class and instantiate a symbol
 475 // using the placement new.
 476 union SymbolUnion {
 477   alignas(DefinedRegular) char a[sizeof(DefinedRegular)];
 478   alignas(DefinedCommon) char b[sizeof(DefinedCommon)];
 479   alignas(DefinedAbsolute) char c[sizeof(DefinedAbsolute)];
 480   alignas(DefinedSynthetic) char d[sizeof(DefinedSynthetic)];
 481   alignas(LazyArchive) char e[sizeof(LazyArchive)];
 482   alignas(Undefined) char f[sizeof(Undefined)];
 483   alignas(DefinedImportData) char g[sizeof(DefinedImportData)];
 484   alignas(DefinedImportThunk) char h[sizeof(DefinedImportThunk)];
 485   alignas(DefinedLocalImport) char i[sizeof(DefinedLocalImport)];
 486   alignas(LazyObject) char j[sizeof(LazyObject)];
 487   alignas(LazyDLLSymbol) char k[sizeof(LazyDLLSymbol)];
 488 };
 489
 490 template <typename T, typename... ArgT>
 491 void replaceSymbol(Symbol *s, ArgT &&... arg) {
 492   static_assert(std::is_trivially_destructible<T>(),
 493                 "Symbol types must be trivially destructible");
 494   static_assert(sizeof(T) <= sizeof(SymbolUnion), "Symbol too small");
 495   static_assert(alignof(T) <= alignof(SymbolUnion),
 496                 "SymbolUnion not aligned enough");
 497   assert(static_cast<Symbol *>(static_cast<T *>(nullptr)) == nullptr &&
 498          "Not a Symbol");
 499   bool canInline = s->canInline;
 500   new (s) T(std::forward<ArgT>(arg)...);
 501   s->canInline = canInline;
 502 }
 503 } // namespace coff
 504
 505 } // namespace lld
 506
 507 #endif