lld/COFF/InputFiles.h

   1 //===- InputFiles.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_INPUT_FILES_H
  10 #define LLD_COFF_INPUT_FILES_H
  11
  12 #include "Config.h"
  13 #include "lld/Common/LLVM.h"
  14 #include "llvm/ADT/ArrayRef.h"
  15 #include "llvm/ADT/DenseMap.h"
  16 #include "llvm/ADT/DenseSet.h"
  17 #include "llvm/BinaryFormat/Magic.h"
  18 #include "llvm/Object/Archive.h"
  19 #include "llvm/Object/COFF.h"
  20 #include "llvm/Support/StringSaver.h"
  21 #include <memory>
  22 #include <set>
  23 #include <vector>
  24
  25 namespace llvm {
  26 struct DILineInfo;
  27 namespace pdb {
  28 class DbiModuleDescriptorBuilder;
  29 }
  30 namespace lto {
  31 class InputFile;
  32 }
  33 }
  34
  35 namespace lld {
  36 class DWARFCache;
  37
  38 namespace coff {
  39
  40 std::vector<MemoryBufferRef> getArchiveMembers(llvm::object::Archive *file);
  41
  42 using llvm::COFF::IMAGE_FILE_MACHINE_UNKNOWN;
  43 using llvm::COFF::MachineTypes;
  44 using llvm::object::Archive;
  45 using llvm::object::COFFObjectFile;
  46 using llvm::object::COFFSymbolRef;
  47 using llvm::object::coff_import_header;
  48 using llvm::object::coff_section;
  49
  50 class Chunk;
  51 class Defined;
  52 class DefinedImportData;
  53 class DefinedImportThunk;
  54 class DefinedRegular;
  55 class SectionChunk;
  56 class Symbol;
  57 class Undefined;
  58 class TpiSource;
  59
  60 // The root class of input files.
  61 class InputFile {
  62 public:
  63   enum Kind {
  64     ArchiveKind,
  65     ObjectKind,
  66     LazyObjectKind,
  67     ImportKind,
  68     BitcodeKind
  69   };
  70   Kind kind() const { return fileKind; }
  71   virtual ~InputFile() {}
  72
  73   // Returns the filename.
  74   StringRef getName() const { return mb.getBufferIdentifier(); }
  75
  76   // Reads a file (the constructor doesn't do that).
  77   virtual void parse() = 0;
  78
  79   // Returns the CPU type this file was compiled to.
  80   virtual MachineTypes getMachineType() { return IMAGE_FILE_MACHINE_UNKNOWN; }
  81
  82   MemoryBufferRef mb;
  83
  84   // An archive file name if this file is created from an archive.
  85   StringRef parentName;
  86
  87   // Returns .drectve section contents if exist.
  88   StringRef getDirectives() { return directives; }
  89
  90 protected:
  91   InputFile(Kind k, MemoryBufferRef m) : mb(m), fileKind(k) {}
  92
  93   StringRef directives;
  94
  95 private:
  96   const Kind fileKind;
  97 };
  98
  99 // .lib or .a file.
 100 class ArchiveFile : public InputFile {
 101 public:
 102   explicit ArchiveFile(MemoryBufferRef m);
 103   static bool classof(const InputFile *f) { return f->kind() == ArchiveKind; }
 104   void parse() override;
 105
 106   // Enqueues an archive member load for the given symbol. If we've already
 107   // enqueued a load for the same archive member, this function does nothing,
 108   // which ensures that we don't load the same member more than once.
 109   void addMember(const Archive::Symbol &sym);
 110
 111 private:
 112   std::unique_ptr<Archive> file;
 113   llvm::DenseSet<uint64_t> seen;
 114 };
 115
 116 // .obj or .o file between -start-lib and -end-lib.
 117 class LazyObjFile : public InputFile {
 118 public:
 119   explicit LazyObjFile(MemoryBufferRef m) : InputFile(LazyObjectKind, m) {}
 120   static bool classof(const InputFile *f) {
 121     return f->kind() == LazyObjectKind;
 122   }
 123   // Makes this object file part of the link.
 124   void fetch();
 125   // Adds the symbols in this file to the symbol table as LazyObject symbols.
 126   void parse() override;
 127
 128 private:
 129   std::vector<Symbol *> symbols;
 130 };
 131
 132 // .obj or .o file. This may be a member of an archive file.
 133 class ObjFile : public InputFile {
 134 public:
 135   explicit ObjFile(MemoryBufferRef m) : InputFile(ObjectKind, m) {}
 136   explicit ObjFile(MemoryBufferRef m, std::vector<Symbol *> &&symbols)
 137       : InputFile(ObjectKind, m), symbols(std::move(symbols)) {}
 138   static bool classof(const InputFile *f) { return f->kind() == ObjectKind; }
 139   void parse() override;
 140   MachineTypes getMachineType() override;
 141   ArrayRef<Chunk *> getChunks() { return chunks; }
 142   ArrayRef<SectionChunk *> getDebugChunks() { return debugChunks; }
 143   ArrayRef<SectionChunk *> getSXDataChunks() { return sXDataChunks; }
 144   ArrayRef<SectionChunk *> getGuardFidChunks() { return guardFidChunks; }
 145   ArrayRef<SectionChunk *> getGuardLJmpChunks() { return guardLJmpChunks; }
 146   ArrayRef<Symbol *> getSymbols() { return symbols; }
 147
 148   ArrayRef<uint8_t> getDebugSection(StringRef secName);
 149
 150   // Returns a Symbol object for the symbolIndex'th symbol in the
 151   // underlying object file.
 152   Symbol *getSymbol(uint32_t symbolIndex) {
 153     return symbols[symbolIndex];
 154   }
 155
 156   // Returns the underlying COFF file.
 157   COFFObjectFile *getCOFFObj() { return coffObj.get(); }
 158
 159   // Add a symbol for a range extension thunk. Return the new symbol table
 160   // index. This index can be used to modify a relocation.
 161   uint32_t addRangeThunkSymbol(Symbol *thunk) {
 162     symbols.push_back(thunk);
 163     return symbols.size() - 1;
 164   }
 165
 166   void includeResourceChunks();
 167
 168   bool isResourceObjFile() const { return !resourceChunks.empty(); }
 169
 170   static std::vector<ObjFile *> instances;
 171
 172   // Flags in the absolute @feat.00 symbol if it is present. These usually
 173   // indicate if an object was compiled with certain security features enabled
 174   // like stack guard, safeseh, /guard:cf, or other things.
 175   uint32_t feat00Flags = 0;
 176
 177   // True if this object file is compatible with SEH.  COFF-specific and
 178   // x86-only. COFF spec 5.10.1. The .sxdata section.
 179   bool hasSafeSEH() { return feat00Flags & 0x1; }
 180
 181   // True if this file was compiled with /guard:cf.
 182   bool hasGuardCF() { return feat00Flags & 0x800; }
 183
 184   // Pointer to the PDB module descriptor builder. Various debug info records
 185   // will reference object files by "module index", which is here. Things like
 186   // source files and section contributions are also recorded here. Will be null
 187   // if we are not producing a PDB.
 188   llvm::pdb::DbiModuleDescriptorBuilder *moduleDBI = nullptr;
 189
 190   const coff_section *addrsigSec = nullptr;
 191
 192   // When using Microsoft precompiled headers, this is the PCH's key.
 193   // The same key is used by both the precompiled object, and objects using the
 194   // precompiled object. Any difference indicates out-of-date objects.
 195   llvm::Optional<uint32_t> pchSignature;
 196
 197   // Whether this file was compiled with /hotpatch.
 198   bool hotPatchable = false;
 199
 200   // Whether the object was already merged into the final PDB.
 201   bool mergedIntoPDB = false;
 202
 203   // If the OBJ has a .debug$T stream, this tells how it will be handled.
 204   TpiSource *debugTypesObj = nullptr;
 205
 206   // The .debug$P or .debug$T section data if present. Empty otherwise.
 207   ArrayRef<uint8_t> debugTypes;
 208
 209   llvm::Optional<std::pair<StringRef, uint32_t>>
 210   getVariableLocation(StringRef var);
 211
 212   llvm::Optional<llvm::DILineInfo> getDILineInfo(uint32_t offset,
 213                                                  uint32_t sectionIndex);
 214
 215 private:
 216   const coff_section* getSection(uint32_t i);
 217   const coff_section *getSection(COFFSymbolRef sym) {
 218     return getSection(sym.getSectionNumber());
 219   }
 220
 221   void initializeChunks();
 222   void initializeSymbols();
 223   void initializeFlags();
 224   void initializeDependencies();
 225
 226   SectionChunk *
 227   readSection(uint32_t sectionNumber,
 228               const llvm::object::coff_aux_section_definition *def,
 229               StringRef leaderName);
 230
 231   void readAssociativeDefinition(
 232       COFFSymbolRef coffSym,
 233       const llvm::object::coff_aux_section_definition *def);
 234
 235   void readAssociativeDefinition(
 236       COFFSymbolRef coffSym,
 237       const llvm::object::coff_aux_section_definition *def,
 238       uint32_t parentSection);
 239
 240   void recordPrevailingSymbolForMingw(
 241       COFFSymbolRef coffSym,
 242       llvm::DenseMap<StringRef, uint32_t> &prevailingSectionMap);
 243
 244   void maybeAssociateSEHForMingw(
 245       COFFSymbolRef sym, const llvm::object::coff_aux_section_definition *def,
 246       const llvm::DenseMap<StringRef, uint32_t> &prevailingSectionMap);
 247
 248   // Given a new symbol Sym with comdat selection Selection, if the new
 249   // symbol is not (yet) Prevailing and the existing comdat leader set to
 250   // Leader, emits a diagnostic if the new symbol and its selection doesn't
 251   // match the existing symbol and its selection. If either old or new
 252   // symbol have selection IMAGE_COMDAT_SELECT_LARGEST, Sym might replace
 253   // the existing leader. In that case, Prevailing is set to true.
 254   void handleComdatSelection(COFFSymbolRef sym,
 255                              llvm::COFF::COMDATType &selection,
 256                              bool &prevailing, DefinedRegular *leader);
 257
 258   llvm::Optional<Symbol *>
 259   createDefined(COFFSymbolRef sym,
 260                 std::vector<const llvm::object::coff_aux_section_definition *>
 261                     &comdatDefs,
 262                 bool &prevailingComdat);
 263   Symbol *createRegular(COFFSymbolRef sym);
 264   Symbol *createUndefined(COFFSymbolRef sym);
 265
 266   std::unique_ptr<COFFObjectFile> coffObj;
 267
 268   // List of all chunks defined by this file. This includes both section
 269   // chunks and non-section chunks for common symbols.
 270   std::vector<Chunk *> chunks;
 271
 272   std::vector<SectionChunk *> resourceChunks;
 273
 274   // CodeView debug info sections.
 275   std::vector<SectionChunk *> debugChunks;
 276
 277   // Chunks containing symbol table indices of exception handlers. Only used for
 278   // 32-bit x86.
 279   std::vector<SectionChunk *> sXDataChunks;
 280
 281   // Chunks containing symbol table indices of address taken symbols and longjmp
 282   // targets.  These are not linked into the final binary when /guard:cf is set.
 283   std::vector<SectionChunk *> guardFidChunks;
 284   std::vector<SectionChunk *> guardLJmpChunks;
 285
 286   // This vector contains the same chunks as Chunks, but they are
 287   // indexed such that you can get a SectionChunk by section index.
 288   // Nonexistent section indices are filled with null pointers.
 289   // (Because section number is 1-based, the first slot is always a
 290   // null pointer.)
 291   std::vector<SectionChunk *> sparseChunks;
 292
 293   // This vector contains a list of all symbols defined or referenced by this
 294   // file. They are indexed such that you can get a Symbol by symbol
 295   // index. Nonexistent indices (which are occupied by auxiliary
 296   // symbols in the real symbol table) are filled with null pointers.
 297   std::vector<Symbol *> symbols;
 298
 299   DWARFCache *dwarf = nullptr;
 300 };
 301
 302 // This type represents import library members that contain DLL names
 303 // and symbols exported from the DLLs. See Microsoft PE/COFF spec. 7
 304 // for details about the format.
 305 class ImportFile : public InputFile {
 306 public:
 307   explicit ImportFile(MemoryBufferRef m) : InputFile(ImportKind, m) {}
 308
 309   static bool classof(const InputFile *f) { return f->kind() == ImportKind; }
 310
 311   static std::vector<ImportFile *> instances;
 312
 313   Symbol *impSym = nullptr;
 314   Symbol *thunkSym = nullptr;
 315   std::string dllName;
 316
 317 private:
 318   void parse() override;
 319
 320 public:
 321   StringRef externalName;
 322   const coff_import_header *hdr;
 323   Chunk *location = nullptr;
 324
 325   // We want to eliminate dllimported symbols if no one actually refers them.
 326   // These "Live" bits are used to keep track of which import library members
 327   // are actually in use.
 328   //
 329   // If the Live bit is turned off by MarkLive, Writer will ignore dllimported
 330   // symbols provided by this import library member. We also track whether the
 331   // imported symbol is used separately from whether the thunk is used in order
 332   // to avoid creating unnecessary thunks.
 333   bool live = !config->doGC;
 334   bool thunkLive = !config->doGC;
 335 };
 336
 337 // Used for LTO.
 338 class BitcodeFile : public InputFile {
 339 public:
 340   BitcodeFile(MemoryBufferRef mb, StringRef archiveName,
 341               uint64_t offsetInArchive);
 342   explicit BitcodeFile(MemoryBufferRef m, StringRef archiveName,
 343                        uint64_t offsetInArchive,
 344                        std::vector<Symbol *> &&symbols);
 345   ~BitcodeFile();
 346   static bool classof(const InputFile *f) { return f->kind() == BitcodeKind; }
 347   ArrayRef<Symbol *> getSymbols() { return symbols; }
 348   MachineTypes getMachineType() override;
 349   static std::vector<BitcodeFile *> instances;
 350   std::unique_ptr<llvm::lto::InputFile> obj;
 351
 352 private:
 353   void parse() override;
 354
 355   std::vector<Symbol *> symbols;
 356 };
 357
 358 inline bool isBitcode(MemoryBufferRef mb) {
 359   return identify_magic(mb.getBuffer()) == llvm::file_magic::bitcode;
 360 }
 361
 362 std::string replaceThinLTOSuffix(StringRef path);
 363 } // namespace coff
 364
 365 std::string toString(const coff::InputFile *file);
 366 } // namespace lld
 367
 368 #endif