lld/ELF/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_ELF_INPUT_FILES_H
  10 #define LLD_ELF_INPUT_FILES_H
  11
  12 #include "Config.h"
  13 #include "Symbols.h"
  14 #include "lld/Common/ErrorHandler.h"
  15 #include "lld/Common/LLVM.h"
  16 #include "lld/Common/Reproduce.h"
  17 #include "llvm/ADT/DenseSet.h"
  18 #include "llvm/BinaryFormat/Magic.h"
  19 #include "llvm/Object/ELF.h"
  20 #include "llvm/Support/MemoryBufferRef.h"
  21 #include "llvm/Support/Threading.h"
  22
  23 namespace llvm {
  24 struct DILineInfo;
  25 class TarWriter;
  26 namespace lto {
  27 class InputFile;
  28 }
  29 } // namespace llvm
  30
  31 namespace lld {
  32 class DWARFCache;
  33
  34 namespace elf {
  35 class InputSection;
  36 class Symbol;
  37
  38 // Returns "<internal>", "foo.a(bar.o)" or "baz.o".
  39 std::string toStr(Ctx &, const InputFile *f);
  40 const ELFSyncStream &operator<<(const ELFSyncStream &, const InputFile *);
  41
  42 // Opens a given file.
  43 std::optional<MemoryBufferRef> readFile(Ctx &, StringRef path);
  44
  45 // Add symbols in File to the symbol table.
  46 void parseFile(Ctx &, InputFile *file);
  47 void parseFiles(Ctx &, const SmallVector<std::unique_ptr<InputFile>, 0> &);
  48
  49 // The root class of input files.
  50 class InputFile {
  51 public:
  52   Ctx &ctx;
  53
  54 protected:
  55   std::unique_ptr<Symbol *[]> symbols;
  56   uint32_t numSymbols = 0;
  57   SmallVector<InputSectionBase *, 0> sections;
  58
  59 public:
  60   enum Kind : uint8_t {
  61     ObjKind,
  62     SharedKind,
  63     BitcodeKind,
  64     BinaryKind,
  65     InternalKind,
  66   };
  67
  68   InputFile(Ctx &, Kind k, MemoryBufferRef m);
  69   virtual ~InputFile();
  70   Kind kind() const { return fileKind; }
  71
  72   bool isElf() const {
  73     Kind k = kind();
  74     return k == ObjKind || k == SharedKind;
  75   }
  76   bool isInternal() const { return kind() == InternalKind; }
  77
  78   StringRef getName() const { return mb.getBufferIdentifier(); }
  79   MemoryBufferRef mb;
  80
  81   // Returns sections. It is a runtime error to call this function
  82   // on files that don't have the notion of sections.
  83   ArrayRef<InputSectionBase *> getSections() const {
  84     assert(fileKind == ObjKind || fileKind == BinaryKind);
  85     return sections;
  86   }
  87   void cacheDecodedCrel(size_t i, InputSectionBase *s) { sections[i] = s; }
  88
  89   // Returns object file symbols. It is a runtime error to call this
  90   // function on files of other types.
  91   ArrayRef<Symbol *> getSymbols() const {
  92     assert(fileKind == BinaryKind || fileKind == ObjKind ||
  93            fileKind == BitcodeKind);
  94     return {symbols.get(), numSymbols};
  95   }
  96
  97   MutableArrayRef<Symbol *> getMutableSymbols() {
  98     assert(fileKind == BinaryKind || fileKind == ObjKind ||
  99            fileKind == BitcodeKind);
 100     return {symbols.get(), numSymbols};
 101   }
 102
 103   Symbol &getSymbol(uint32_t symbolIndex) const {
 104     assert(fileKind == ObjKind);
 105     if (symbolIndex >= numSymbols)
 106       Fatal(ctx) << this << ": invalid symbol index";
 107     return *this->symbols[symbolIndex];
 108   }
 109
 110   template <typename RelT> Symbol &getRelocTargetSym(const RelT &rel) const {
 111     uint32_t symIndex = rel.getSymbol(ctx.arg.isMips64EL);
 112     return getSymbol(symIndex);
 113   }
 114
 115   // Get filename to use for linker script processing.
 116   StringRef getNameForScript() const;
 117
 118   // Check if a non-common symbol should be extracted to override a common
 119   // definition.
 120   bool shouldExtractForCommon(StringRef name) const;
 121
 122   // .got2 in the current file. This is used by PPC32 -fPIC/-fPIE to compute
 123   // offsets in PLT call stubs.
 124   InputSection *ppc32Got2 = nullptr;
 125
 126   // Index of MIPS GOT built for this file.
 127   uint32_t mipsGotIndex = -1;
 128
 129   // groupId is used for --warn-backrefs which is an optional error
 130   // checking feature. All files within the same --{start,end}-group or
 131   // --{start,end}-lib get the same group ID. Otherwise, each file gets a new
 132   // group ID. For more info, see checkDependency() in SymbolTable.cpp.
 133   uint32_t groupId;
 134
 135   // If this is an architecture-specific file, the following members
 136   // have ELF type (i.e. ELF{32,64}{LE,BE}) and target machine type.
 137   uint16_t emachine = llvm::ELF::EM_NONE;
 138   const Kind fileKind;
 139   ELFKind ekind = ELFNoneKind;
 140   uint8_t osabi = 0;
 141   uint8_t abiVersion = 0;
 142
 143   // True if this is a relocatable object file/bitcode file in an ar archive
 144   // or between --start-lib and --end-lib.
 145   bool lazy = false;
 146
 147   // True if this is an argument for --just-symbols. Usually false.
 148   bool justSymbols = false;
 149
 150   std::string getSrcMsg(const Symbol &sym, const InputSectionBase &sec,
 151                         uint64_t offset);
 152
 153   // On PPC64 we need to keep track of which files contain small code model
 154   // relocations that access the .toc section. To minimize the chance of a
 155   // relocation overflow, files that do contain said relocations should have
 156   // their .toc sections sorted closer to the .got section than files that do
 157   // not contain any small code model relocations. Thats because the toc-pointer
 158   // is defined to point at .got + 0x8000 and the instructions used with small
 159   // code model relocations support immediates in the range [-0x8000, 0x7FFC],
 160   // making the addressable range relative to the toc pointer
 161   // [.got, .got + 0xFFFC].
 162   bool ppc64SmallCodeModelTocRelocs = false;
 163
 164   // True if the file has TLSGD/TLSLD GOT relocations without R_PPC64_TLSGD or
 165   // R_PPC64_TLSLD. Disable TLS relaxation to avoid bad code generation.
 166   bool ppc64DisableTLSRelax = false;
 167
 168 public:
 169   // If not empty, this stores the name of the archive containing this file.
 170   // We use this string for creating error messages.
 171   SmallString<0> archiveName;
 172   // Cache for toStr(Ctx &, const InputFile *). Only toStr should use this
 173   // member.
 174   mutable SmallString<0> toStringCache;
 175
 176 private:
 177   // Cache for getNameForScript().
 178   mutable SmallString<0> nameForScriptCache;
 179 };
 180
 181 class ELFFileBase : public InputFile {
 182 public:
 183   ELFFileBase(Ctx &ctx, Kind k, ELFKind ekind, MemoryBufferRef m);
 184   static bool classof(const InputFile *f) { return f->isElf(); }
 185
 186   void init();
 187   template <typename ELFT> llvm::object::ELFFile<ELFT> getObj() const {
 188     return check(llvm::object::ELFFile<ELFT>::create(mb.getBuffer()));
 189   }
 190
 191   StringRef getStringTable() const { return stringTable; }
 192
 193   ArrayRef<Symbol *> getLocalSymbols() {
 194     if (numSymbols == 0)
 195       return {};
 196     return llvm::ArrayRef(symbols.get() + 1, firstGlobal - 1);
 197   }
 198   ArrayRef<Symbol *> getGlobalSymbols() {
 199     return llvm::ArrayRef(symbols.get() + firstGlobal,
 200                           numSymbols - firstGlobal);
 201   }
 202   MutableArrayRef<Symbol *> getMutableGlobalSymbols() {
 203     return llvm::MutableArrayRef(symbols.get() + firstGlobal,
 204                                      numSymbols - firstGlobal);
 205   }
 206
 207   template <typename ELFT> typename ELFT::ShdrRange getELFShdrs() const {
 208     return typename ELFT::ShdrRange(
 209         reinterpret_cast<const typename ELFT::Shdr *>(elfShdrs), numELFShdrs);
 210   }
 211   template <typename ELFT> typename ELFT::SymRange getELFSyms() const {
 212     return typename ELFT::SymRange(
 213         reinterpret_cast<const typename ELFT::Sym *>(elfSyms), numELFSyms);
 214   }
 215   template <typename ELFT> typename ELFT::SymRange getGlobalELFSyms() const {
 216     return getELFSyms<ELFT>().slice(firstGlobal);
 217   }
 218
 219 protected:
 220   // Initializes this class's member variables.
 221   template <typename ELFT> void init(InputFile::Kind k);
 222
 223   StringRef stringTable;
 224   const void *elfShdrs = nullptr;
 225   const void *elfSyms = nullptr;
 226   uint32_t numELFShdrs = 0;
 227   uint32_t numELFSyms = 0;
 228   uint32_t firstGlobal = 0;
 229
 230 public:
 231   uint32_t andFeatures = 0;
 232   bool hasCommonSyms = false;
 233   ArrayRef<uint8_t> aarch64PauthAbiCoreInfo;
 234 };
 235
 236 // .o file.
 237 template <class ELFT> class ObjFile : public ELFFileBase {
 238   LLVM_ELF_IMPORT_TYPES_ELFT(ELFT)
 239
 240 public:
 241   static bool classof(const InputFile *f) { return f->kind() == ObjKind; }
 242
 243   llvm::object::ELFFile<ELFT> getObj() const {
 244     return this->ELFFileBase::getObj<ELFT>();
 245   }
 246
 247   ObjFile(Ctx &ctx, ELFKind ekind, MemoryBufferRef m, StringRef archiveName)
 248       : ELFFileBase(ctx, ObjKind, ekind, m) {
 249     this->archiveName = archiveName;
 250   }
 251
 252   void parse(bool ignoreComdats = false);
 253   void parseLazy();
 254
 255   StringRef getShtGroupSignature(ArrayRef<Elf_Shdr> sections,
 256                                  const Elf_Shdr &sec);
 257
 258   uint32_t getSectionIndex(const Elf_Sym &sym) const;
 259
 260   std::optional<llvm::DILineInfo> getDILineInfo(const InputSectionBase *,
 261                                                 uint64_t);
 262   std::optional<std::pair<std::string, unsigned>>
 263   getVariableLoc(StringRef name);
 264
 265   // Name of source file obtained from STT_FILE symbol value,
 266   // or empty string if there is no such symbol in object file
 267   // symbol table.
 268   StringRef sourceFile;
 269
 270   // Pointer to this input file's .llvm_addrsig section, if it has one.
 271   const Elf_Shdr *addrsigSec = nullptr;
 272
 273   // SHT_LLVM_CALL_GRAPH_PROFILE section index.
 274   uint32_t cgProfileSectionIndex = 0;
 275
 276   // MIPS GP0 value defined by this file. This value represents the gp value
 277   // used to create the relocatable object and required to support
 278   // R_MIPS_GPREL16 / R_MIPS_GPREL32 relocations.
 279   uint32_t mipsGp0 = 0;
 280
 281   // True if the file defines functions compiled with
 282   // -fsplit-stack. Usually false.
 283   bool splitStack = false;
 284
 285   // True if the file defines functions compiled with -fsplit-stack,
 286   // but had one or more functions with the no_split_stack attribute.
 287   bool someNoSplitStack = false;
 288
 289   // Get cached DWARF information.
 290   DWARFCache *getDwarf();
 291
 292   void initSectionsAndLocalSyms(bool ignoreComdats);
 293   void postParse();
 294   void importCmseSymbols();
 295
 296 private:
 297   void initializeSections(bool ignoreComdats,
 298                           const llvm::object::ELFFile<ELFT> &obj);
 299   void initializeSymbols(const llvm::object::ELFFile<ELFT> &obj);
 300   void initializeJustSymbols();
 301
 302   InputSectionBase *getRelocTarget(uint32_t idx, uint32_t info);
 303   InputSectionBase *createInputSection(uint32_t idx, const Elf_Shdr &sec,
 304                                        StringRef name);
 305
 306   bool shouldMerge(const Elf_Shdr &sec, StringRef name);
 307
 308   // Each ELF symbol contains a section index which the symbol belongs to.
 309   // However, because the number of bits dedicated for that is limited, a
 310   // symbol can directly point to a section only when the section index is
 311   // equal to or smaller than 65280.
 312   //
 313   // If an object file contains more than 65280 sections, the file must
 314   // contain .symtab_shndx section. The section contains an array of
 315   // 32-bit integers whose size is the same as the number of symbols.
 316   // Nth symbol's section index is in the Nth entry of .symtab_shndx.
 317   //
 318   // The following variable contains the contents of .symtab_shndx.
 319   // If the section does not exist (which is common), the array is empty.
 320   ArrayRef<Elf_Word> shndxTable;
 321
 322   // Debugging information to retrieve source file and line for error
 323   // reporting. Linker may find reasonable number of errors in a
 324   // single object file, so we cache debugging information in order to
 325   // parse it only once for each object file we link.
 326   std::unique_ptr<DWARFCache> dwarf;
 327   llvm::once_flag initDwarf;
 328 };
 329
 330 class BitcodeFile : public InputFile {
 331 public:
 332   BitcodeFile(Ctx &, MemoryBufferRef m, StringRef archiveName,
 333               uint64_t offsetInArchive, bool lazy);
 334   static bool classof(const InputFile *f) { return f->kind() == BitcodeKind; }
 335   void parse();
 336   void parseLazy();
 337   void postParse();
 338   std::unique_ptr<llvm::lto::InputFile> obj;
 339   std::vector<bool> keptComdats;
 340 };
 341
 342 // .so file.
 343 class SharedFile : public ELFFileBase {
 344 public:
 345   SharedFile(Ctx &, MemoryBufferRef m, StringRef defaultSoName);
 346
 347   // This is actually a vector of Elf_Verdef pointers.
 348   SmallVector<const void *, 0> verdefs;
 349
 350   // If the output file needs Elf_Verneed data structures for this file, this is
 351   // a vector of Elf_Vernaux version identifiers that map onto the entries in
 352   // Verdefs, otherwise it is empty.
 353   SmallVector<uint32_t, 0> vernauxs;
 354
 355   SmallVector<StringRef, 0> dtNeeded;
 356   StringRef soName;
 357
 358   static bool classof(const InputFile *f) { return f->kind() == SharedKind; }
 359
 360   template <typename ELFT> void parse();
 361
 362   // Used for --as-needed
 363   bool isNeeded;
 364
 365   // Non-weak undefined symbols which are not yet resolved when the SO is
 366   // parsed. Only filled for `--no-allow-shlib-undefined`.
 367   SmallVector<Symbol *, 0> requiredSymbols;
 368
 369 private:
 370   template <typename ELFT>
 371   std::vector<uint32_t> parseVerneed(const llvm::object::ELFFile<ELFT> &obj,
 372                                      const typename ELFT::Shdr *sec);
 373 };
 374
 375 class BinaryFile : public InputFile {
 376 public:
 377   explicit BinaryFile(Ctx &ctx, MemoryBufferRef m)
 378       : InputFile(ctx, BinaryKind, m) {}
 379   static bool classof(const InputFile *f) { return f->kind() == BinaryKind; }
 380   void parse();
 381 };
 382
 383 InputFile *createInternalFile(Ctx &, StringRef name);
 384 std::unique_ptr<ELFFileBase> createObjFile(Ctx &, MemoryBufferRef mb,
 385                                            StringRef archiveName = "",
 386                                            bool lazy = false);
 387
 388 std::string replaceThinLTOSuffix(Ctx &, StringRef path);
 389
 390 } // namespace elf
 391 } // namespace lld
 392
 393 #endif