1 //===- Driver.cpp ---------------------------------------------------------===// 2 // 3 // The LLVM Linker 4 // 5 // This file is distributed under the University of Illinois Open Source 6 // License. See LICENSE.TXT for details. 7 // 8 //===----------------------------------------------------------------------===// 9 // 10 // The driver drives the entire linking process. It is responsible for 11 // parsing command line options and doing whatever it is instructed to do. 12 // 13 // One notable thing in the LLD's driver when compared to other linkers is 14 // that the LLD's driver is agnostic on the host operating system. 15 // Other linkers usually have implicit default values (such as a dynamic 16 // linker path or library paths) for each host OS. 17 // 18 // I don't think implicit default values are useful because they are 19 // usually explicitly specified by the compiler driver. They can even 20 // be harmful when you are doing cross-linking. Therefore, in LLD, we 21 // simply trust the compiler driver to pass all required options and 22 // don't try to make effort on our side. 23 // 24 //===----------------------------------------------------------------------===// 25 26 #include "Driver.h" 27 #include "Config.h" 28 #include "Filesystem.h" 29 #include "ICF.h" 30 #include "InputFiles.h" 31 #include "InputSection.h" 32 #include "LinkerScript.h" 33 #include "MarkLive.h" 34 #include "OutputSections.h" 35 #include "ScriptParser.h" 36 #include "SymbolTable.h" 37 #include "Symbols.h" 38 #include "SyntheticSections.h" 39 #include "Target.h" 40 #include "Writer.h" 41 #include "lld/Common/Args.h" 42 #include "lld/Common/Driver.h" 43 #include "lld/Common/ErrorHandler.h" 44 #include "lld/Common/Memory.h" 45 #include "lld/Common/Strings.h" 46 #include "lld/Common/TargetOptionsCommandFlags.h" 47 #include "lld/Common/Threads.h" 48 #include "lld/Common/Version.h" 49 #include "llvm/ADT/SetVector.h" 50 #include "llvm/ADT/StringExtras.h" 51 #include "llvm/ADT/StringSwitch.h" 52 #include "llvm/Support/CommandLine.h" 53 #include "llvm/Support/Compression.h" 54 #include "llvm/Support/LEB128.h" 55 #include "llvm/Support/Path.h" 56 #include "llvm/Support/TarWriter.h" 57 #include "llvm/Support/TargetSelect.h" 58 #include "llvm/Support/raw_ostream.h" 59 #include <cstdlib> 60 #include <utility> 61 62 using namespace llvm; 63 using namespace llvm::ELF; 64 using namespace llvm::object; 65 using namespace llvm::sys; 66 using namespace llvm::support; 67 68 using namespace lld; 69 using namespace lld::elf; 70 71 Configuration *elf::Config; 72 LinkerDriver *elf::Driver; 73 74 static void setConfigs(opt::InputArgList &Args); 75 76 bool elf::link(ArrayRef<const char *> Args, bool CanExitEarly, 77 raw_ostream &Error) { 78 errorHandler().LogName = args::getFilenameWithoutExe(Args[0]); 79 errorHandler().ErrorLimitExceededMsg = 80 "too many errors emitted, stopping now (use " 81 "-error-limit=0 to see all errors)"; 82 errorHandler().ErrorOS = &Error; 83 errorHandler().ExitEarly = CanExitEarly; 84 errorHandler().ColorDiagnostics = Error.has_colors(); 85 86 InputSections.clear(); 87 OutputSections.clear(); 88 BinaryFiles.clear(); 89 BitcodeFiles.clear(); 90 ObjectFiles.clear(); 91 SharedFiles.clear(); 92 93 Config = make<Configuration>(); 94 Driver = make<LinkerDriver>(); 95 Script = make<LinkerScript>(); 96 Symtab = make<SymbolTable>(); 97 98 Tar = nullptr; 99 memset(&In, 0, sizeof(In)); 100 101 Config->ProgName = Args[0]; 102 103 Driver->main(Args); 104 105 // Exit immediately if we don't need to return to the caller. 106 // This saves time because the overhead of calling destructors 107 // for all globally-allocated objects is not negligible. 108 if (CanExitEarly) 109 exitLld(errorCount() ? 1 : 0); 110 111 freeArena(); 112 return !errorCount(); 113 } 114 115 // Parses a linker -m option. 116 static std::tuple<ELFKind, uint16_t, uint8_t> parseEmulation(StringRef Emul) { 117 uint8_t OSABI = 0; 118 StringRef S = Emul; 119 if (S.endswith("_fbsd")) { 120 S = S.drop_back(5); 121 OSABI = ELFOSABI_FREEBSD; 122 } 123 124 std::pair<ELFKind, uint16_t> Ret = 125 StringSwitch<std::pair<ELFKind, uint16_t>>(S) 126 .Cases("aarch64elf", "aarch64linux", "aarch64_elf64_le_vec", 127 {ELF64LEKind, EM_AARCH64}) 128 .Cases("armelf", "armelf_linux_eabi", {ELF32LEKind, EM_ARM}) 129 .Case("elf32_x86_64", {ELF32LEKind, EM_X86_64}) 130 .Cases("elf32btsmip", "elf32btsmipn32", {ELF32BEKind, EM_MIPS}) 131 .Cases("elf32ltsmip", "elf32ltsmipn32", {ELF32LEKind, EM_MIPS}) 132 .Case("elf32lriscv", {ELF32LEKind, EM_RISCV}) 133 .Case("elf32ppc", {ELF32BEKind, EM_PPC}) 134 .Case("elf64btsmip", {ELF64BEKind, EM_MIPS}) 135 .Case("elf64ltsmip", {ELF64LEKind, EM_MIPS}) 136 .Case("elf64lriscv", {ELF64LEKind, EM_RISCV}) 137 .Case("elf64ppc", {ELF64BEKind, EM_PPC64}) 138 .Case("elf64lppc", {ELF64LEKind, EM_PPC64}) 139 .Cases("elf_amd64", "elf_x86_64", {ELF64LEKind, EM_X86_64}) 140 .Case("elf_i386", {ELF32LEKind, EM_386}) 141 .Case("elf_iamcu", {ELF32LEKind, EM_IAMCU}) 142 .Default({ELFNoneKind, EM_NONE}); 143 144 if (Ret.first == ELFNoneKind) 145 error("unknown emulation: " + Emul); 146 return std::make_tuple(Ret.first, Ret.second, OSABI); 147 } 148 149 // Returns slices of MB by parsing MB as an archive file. 150 // Each slice consists of a member file in the archive. 151 std::vector<std::pair<MemoryBufferRef, uint64_t>> static getArchiveMembers( 152 MemoryBufferRef MB) { 153 std::unique_ptr<Archive> File = 154 CHECK(Archive::create(MB), 155 MB.getBufferIdentifier() + ": failed to parse archive"); 156 157 std::vector<std::pair<MemoryBufferRef, uint64_t>> V; 158 Error Err = Error::success(); 159 bool AddToTar = File->isThin() && Tar; 160 for (const ErrorOr<Archive::Child> &COrErr : File->children(Err)) { 161 Archive::Child C = 162 CHECK(COrErr, MB.getBufferIdentifier() + 163 ": could not get the child of the archive"); 164 MemoryBufferRef MBRef = 165 CHECK(C.getMemoryBufferRef(), 166 MB.getBufferIdentifier() + 167 ": could not get the buffer for a child of the archive"); 168 if (AddToTar) 169 Tar->append(relativeToRoot(check(C.getFullName())), MBRef.getBuffer()); 170 V.push_back(std::make_pair(MBRef, C.getChildOffset())); 171 } 172 if (Err) 173 fatal(MB.getBufferIdentifier() + ": Archive::children failed: " + 174 toString(std::move(Err))); 175 176 // Take ownership of memory buffers created for members of thin archives. 177 for (std::unique_ptr<MemoryBuffer> &MB : File->takeThinBuffers()) 178 make<std::unique_ptr<MemoryBuffer>>(std::move(MB)); 179 180 return V; 181 } 182 183 // Opens a file and create a file object. Path has to be resolved already. 184 void LinkerDriver::addFile(StringRef Path, bool WithLOption) { 185 using namespace sys::fs; 186 187 Optional<MemoryBufferRef> Buffer = readFile(Path); 188 if (!Buffer.hasValue()) 189 return; 190 MemoryBufferRef MBRef = *Buffer; 191 192 if (Config->FormatBinary) { 193 Files.push_back(make<BinaryFile>(MBRef)); 194 return; 195 } 196 197 switch (identify_magic(MBRef.getBuffer())) { 198 case file_magic::unknown: 199 readLinkerScript(MBRef); 200 return; 201 case file_magic::archive: { 202 // Handle -whole-archive. 203 if (InWholeArchive) { 204 for (const auto &P : getArchiveMembers(MBRef)) 205 Files.push_back(createObjectFile(P.first, Path, P.second)); 206 return; 207 } 208 209 std::unique_ptr<Archive> File = 210 CHECK(Archive::create(MBRef), Path + ": failed to parse archive"); 211 212 // If an archive file has no symbol table, it is likely that a user 213 // is attempting LTO and using a default ar command that doesn't 214 // understand the LLVM bitcode file. It is a pretty common error, so 215 // we'll handle it as if it had a symbol table. 216 if (!File->isEmpty() && !File->hasSymbolTable()) { 217 for (const auto &P : getArchiveMembers(MBRef)) 218 Files.push_back(make<LazyObjFile>(P.first, Path, P.second)); 219 return; 220 } 221 222 // Handle the regular case. 223 Files.push_back(make<ArchiveFile>(std::move(File))); 224 return; 225 } 226 case file_magic::elf_shared_object: 227 if (Config->Static || Config->Relocatable) { 228 error("attempted static link of dynamic object " + Path); 229 return; 230 } 231 232 // DSOs usually have DT_SONAME tags in their ELF headers, and the 233 // sonames are used to identify DSOs. But if they are missing, 234 // they are identified by filenames. We don't know whether the new 235 // file has a DT_SONAME or not because we haven't parsed it yet. 236 // Here, we set the default soname for the file because we might 237 // need it later. 238 // 239 // If a file was specified by -lfoo, the directory part is not 240 // significant, as a user did not specify it. This behavior is 241 // compatible with GNU. 242 Files.push_back( 243 createSharedFile(MBRef, WithLOption ? path::filename(Path) : Path)); 244 return; 245 case file_magic::bitcode: 246 case file_magic::elf_relocatable: 247 if (InLib) 248 Files.push_back(make<LazyObjFile>(MBRef, "", 0)); 249 else 250 Files.push_back(createObjectFile(MBRef)); 251 break; 252 default: 253 error(Path + ": unknown file type"); 254 } 255 } 256 257 // Add a given library by searching it from input search paths. 258 void LinkerDriver::addLibrary(StringRef Name) { 259 if (Optional<std::string> Path = searchLibrary(Name)) 260 addFile(*Path, /*WithLOption=*/true); 261 else 262 error("unable to find library -l" + Name); 263 } 264 265 // This function is called on startup. We need this for LTO since 266 // LTO calls LLVM functions to compile bitcode files to native code. 267 // Technically this can be delayed until we read bitcode files, but 268 // we don't bother to do lazily because the initialization is fast. 269 static void initLLVM() { 270 InitializeAllTargets(); 271 InitializeAllTargetMCs(); 272 InitializeAllAsmPrinters(); 273 InitializeAllAsmParsers(); 274 } 275 276 // Some command line options or some combinations of them are not allowed. 277 // This function checks for such errors. 278 static void checkOptions() { 279 // The MIPS ABI as of 2016 does not support the GNU-style symbol lookup 280 // table which is a relatively new feature. 281 if (Config->EMachine == EM_MIPS && Config->GnuHash) 282 error("the .gnu.hash section is not compatible with the MIPS target"); 283 284 if (Config->FixCortexA53Errata843419 && Config->EMachine != EM_AARCH64) 285 error("--fix-cortex-a53-843419 is only supported on AArch64 targets"); 286 287 if (Config->TocOptimize && Config->EMachine != EM_PPC64) 288 error("--toc-optimize is only supported on the PowerPC64 target"); 289 290 if (Config->Pie && Config->Shared) 291 error("-shared and -pie may not be used together"); 292 293 if (!Config->Shared && !Config->FilterList.empty()) 294 error("-F may not be used without -shared"); 295 296 if (!Config->Shared && !Config->AuxiliaryList.empty()) 297 error("-f may not be used without -shared"); 298 299 if (!Config->Relocatable && !Config->DefineCommon) 300 error("-no-define-common not supported in non relocatable output"); 301 302 if (Config->Relocatable) { 303 if (Config->Shared) 304 error("-r and -shared may not be used together"); 305 if (Config->GcSections) 306 error("-r and --gc-sections may not be used together"); 307 if (Config->GdbIndex) 308 error("-r and --gdb-index may not be used together"); 309 if (Config->ICF != ICFLevel::None) 310 error("-r and --icf may not be used together"); 311 if (Config->Pie) 312 error("-r and -pie may not be used together"); 313 } 314 315 if (Config->ExecuteOnly) { 316 if (Config->EMachine != EM_AARCH64) 317 error("-execute-only is only supported on AArch64 targets"); 318 319 if (Config->SingleRoRx && !Script->HasSectionsCommand) 320 error("-execute-only and -no-rosegment cannot be used together"); 321 } 322 } 323 324 static const char *getReproduceOption(opt::InputArgList &Args) { 325 if (auto *Arg = Args.getLastArg(OPT_reproduce)) 326 return Arg->getValue(); 327 return getenv("LLD_REPRODUCE"); 328 } 329 330 static bool hasZOption(opt::InputArgList &Args, StringRef Key) { 331 for (auto *Arg : Args.filtered(OPT_z)) 332 if (Key == Arg->getValue()) 333 return true; 334 return false; 335 } 336 337 static bool getZFlag(opt::InputArgList &Args, StringRef K1, StringRef K2, 338 bool Default) { 339 for (auto *Arg : Args.filtered_reverse(OPT_z)) { 340 if (K1 == Arg->getValue()) 341 return true; 342 if (K2 == Arg->getValue()) 343 return false; 344 } 345 return Default; 346 } 347 348 static bool isKnownZFlag(StringRef S) { 349 return S == "combreloc" || S == "copyreloc" || S == "defs" || 350 S == "execstack" || S == "global" || S == "hazardplt" || 351 S == "initfirst" || S == "interpose" || 352 S == "keep-text-section-prefix" || S == "lazy" || S == "muldefs" || 353 S == "nocombreloc" || S == "nocopyreloc" || S == "nodefaultlib" || 354 S == "nodelete" || S == "nodlopen" || S == "noexecstack" || 355 S == "nokeep-text-section-prefix" || S == "norelro" || S == "notext" || 356 S == "now" || S == "origin" || S == "relro" || S == "retpolineplt" || 357 S == "rodynamic" || S == "text" || S == "wxneeded" || 358 S.startswith("max-page-size=") || S.startswith("stack-size="); 359 } 360 361 // Report an error for an unknown -z option. 362 static void checkZOptions(opt::InputArgList &Args) { 363 for (auto *Arg : Args.filtered(OPT_z)) 364 if (!isKnownZFlag(Arg->getValue())) 365 error("unknown -z value: " + StringRef(Arg->getValue())); 366 } 367 368 void LinkerDriver::main(ArrayRef<const char *> ArgsArr) { 369 ELFOptTable Parser; 370 opt::InputArgList Args = Parser.parse(ArgsArr.slice(1)); 371 372 // Interpret this flag early because error() depends on them. 373 errorHandler().ErrorLimit = args::getInteger(Args, OPT_error_limit, 20); 374 375 // Handle -help 376 if (Args.hasArg(OPT_help)) { 377 printHelp(); 378 return; 379 } 380 381 // Handle -v or -version. 382 // 383 // A note about "compatible with GNU linkers" message: this is a hack for 384 // scripts generated by GNU Libtool 2.4.6 (released in February 2014 and 385 // still the newest version in March 2017) or earlier to recognize LLD as 386 // a GNU compatible linker. As long as an output for the -v option 387 // contains "GNU" or "with BFD", they recognize us as GNU-compatible. 388 // 389 // This is somewhat ugly hack, but in reality, we had no choice other 390 // than doing this. Considering the very long release cycle of Libtool, 391 // it is not easy to improve it to recognize LLD as a GNU compatible 392 // linker in a timely manner. Even if we can make it, there are still a 393 // lot of "configure" scripts out there that are generated by old version 394 // of Libtool. We cannot convince every software developer to migrate to 395 // the latest version and re-generate scripts. So we have this hack. 396 if (Args.hasArg(OPT_v) || Args.hasArg(OPT_version)) 397 message(getLLDVersion() + " (compatible with GNU linkers)"); 398 399 if (const char *Path = getReproduceOption(Args)) { 400 // Note that --reproduce is a debug option so you can ignore it 401 // if you are trying to understand the whole picture of the code. 402 Expected<std::unique_ptr<TarWriter>> ErrOrWriter = 403 TarWriter::create(Path, path::stem(Path)); 404 if (ErrOrWriter) { 405 Tar = ErrOrWriter->get(); 406 Tar->append("response.txt", createResponseFile(Args)); 407 Tar->append("version.txt", getLLDVersion() + "\n"); 408 make<std::unique_ptr<TarWriter>>(std::move(*ErrOrWriter)); 409 } else { 410 error(Twine("--reproduce: failed to open ") + Path + ": " + 411 toString(ErrOrWriter.takeError())); 412 } 413 } 414 415 readConfigs(Args); 416 checkZOptions(Args); 417 418 // The behavior of -v or --version is a bit strange, but this is 419 // needed for compatibility with GNU linkers. 420 if (Args.hasArg(OPT_v) && !Args.hasArg(OPT_INPUT)) 421 return; 422 if (Args.hasArg(OPT_version)) 423 return; 424 425 initLLVM(); 426 createFiles(Args); 427 if (errorCount()) 428 return; 429 430 inferMachineType(); 431 setConfigs(Args); 432 checkOptions(); 433 if (errorCount()) 434 return; 435 436 switch (Config->EKind) { 437 case ELF32LEKind: 438 link<ELF32LE>(Args); 439 return; 440 case ELF32BEKind: 441 link<ELF32BE>(Args); 442 return; 443 case ELF64LEKind: 444 link<ELF64LE>(Args); 445 return; 446 case ELF64BEKind: 447 link<ELF64BE>(Args); 448 return; 449 default: 450 llvm_unreachable("unknown Config->EKind"); 451 } 452 } 453 454 static std::string getRpath(opt::InputArgList &Args) { 455 std::vector<StringRef> V = args::getStrings(Args, OPT_rpath); 456 return llvm::join(V.begin(), V.end(), ":"); 457 } 458 459 // Determines what we should do if there are remaining unresolved 460 // symbols after the name resolution. 461 static UnresolvedPolicy getUnresolvedSymbolPolicy(opt::InputArgList &Args) { 462 UnresolvedPolicy ErrorOrWarn = Args.hasFlag(OPT_error_unresolved_symbols, 463 OPT_warn_unresolved_symbols, true) 464 ? UnresolvedPolicy::ReportError 465 : UnresolvedPolicy::Warn; 466 467 // Process the last of -unresolved-symbols, -no-undefined or -z defs. 468 for (auto *Arg : llvm::reverse(Args)) { 469 switch (Arg->getOption().getID()) { 470 case OPT_unresolved_symbols: { 471 StringRef S = Arg->getValue(); 472 if (S == "ignore-all" || S == "ignore-in-object-files") 473 return UnresolvedPolicy::Ignore; 474 if (S == "ignore-in-shared-libs" || S == "report-all") 475 return ErrorOrWarn; 476 error("unknown --unresolved-symbols value: " + S); 477 continue; 478 } 479 case OPT_no_undefined: 480 return ErrorOrWarn; 481 case OPT_z: 482 if (StringRef(Arg->getValue()) == "defs") 483 return ErrorOrWarn; 484 continue; 485 } 486 } 487 488 // -shared implies -unresolved-symbols=ignore-all because missing 489 // symbols are likely to be resolved at runtime using other DSOs. 490 if (Config->Shared) 491 return UnresolvedPolicy::Ignore; 492 return ErrorOrWarn; 493 } 494 495 static Target2Policy getTarget2(opt::InputArgList &Args) { 496 StringRef S = Args.getLastArgValue(OPT_target2, "got-rel"); 497 if (S == "rel") 498 return Target2Policy::Rel; 499 if (S == "abs") 500 return Target2Policy::Abs; 501 if (S == "got-rel") 502 return Target2Policy::GotRel; 503 error("unknown --target2 option: " + S); 504 return Target2Policy::GotRel; 505 } 506 507 static bool isOutputFormatBinary(opt::InputArgList &Args) { 508 StringRef S = Args.getLastArgValue(OPT_oformat, "elf"); 509 if (S == "binary") 510 return true; 511 if (!S.startswith("elf")) 512 error("unknown --oformat value: " + S); 513 return false; 514 } 515 516 static DiscardPolicy getDiscard(opt::InputArgList &Args) { 517 if (Args.hasArg(OPT_relocatable)) 518 return DiscardPolicy::None; 519 520 auto *Arg = 521 Args.getLastArg(OPT_discard_all, OPT_discard_locals, OPT_discard_none); 522 if (!Arg) 523 return DiscardPolicy::Default; 524 if (Arg->getOption().getID() == OPT_discard_all) 525 return DiscardPolicy::All; 526 if (Arg->getOption().getID() == OPT_discard_locals) 527 return DiscardPolicy::Locals; 528 return DiscardPolicy::None; 529 } 530 531 static StringRef getDynamicLinker(opt::InputArgList &Args) { 532 auto *Arg = Args.getLastArg(OPT_dynamic_linker, OPT_no_dynamic_linker); 533 if (!Arg || Arg->getOption().getID() == OPT_no_dynamic_linker) 534 return ""; 535 return Arg->getValue(); 536 } 537 538 static ICFLevel getICF(opt::InputArgList &Args) { 539 auto *Arg = Args.getLastArg(OPT_icf_none, OPT_icf_safe, OPT_icf_all); 540 if (!Arg || Arg->getOption().getID() == OPT_icf_none) 541 return ICFLevel::None; 542 if (Arg->getOption().getID() == OPT_icf_safe) 543 return ICFLevel::Safe; 544 return ICFLevel::All; 545 } 546 547 static StripPolicy getStrip(opt::InputArgList &Args) { 548 if (Args.hasArg(OPT_relocatable)) 549 return StripPolicy::None; 550 551 auto *Arg = Args.getLastArg(OPT_strip_all, OPT_strip_debug); 552 if (!Arg) 553 return StripPolicy::None; 554 if (Arg->getOption().getID() == OPT_strip_all) 555 return StripPolicy::All; 556 return StripPolicy::Debug; 557 } 558 559 static uint64_t parseSectionAddress(StringRef S, const opt::Arg &Arg) { 560 uint64_t VA = 0; 561 if (S.startswith("0x")) 562 S = S.drop_front(2); 563 if (!to_integer(S, VA, 16)) 564 error("invalid argument: " + toString(Arg)); 565 return VA; 566 } 567 568 static StringMap<uint64_t> getSectionStartMap(opt::InputArgList &Args) { 569 StringMap<uint64_t> Ret; 570 for (auto *Arg : Args.filtered(OPT_section_start)) { 571 StringRef Name; 572 StringRef Addr; 573 std::tie(Name, Addr) = StringRef(Arg->getValue()).split('='); 574 Ret[Name] = parseSectionAddress(Addr, *Arg); 575 } 576 577 if (auto *Arg = Args.getLastArg(OPT_Ttext)) 578 Ret[".text"] = parseSectionAddress(Arg->getValue(), *Arg); 579 if (auto *Arg = Args.getLastArg(OPT_Tdata)) 580 Ret[".data"] = parseSectionAddress(Arg->getValue(), *Arg); 581 if (auto *Arg = Args.getLastArg(OPT_Tbss)) 582 Ret[".bss"] = parseSectionAddress(Arg->getValue(), *Arg); 583 return Ret; 584 } 585 586 static SortSectionPolicy getSortSection(opt::InputArgList &Args) { 587 StringRef S = Args.getLastArgValue(OPT_sort_section); 588 if (S == "alignment") 589 return SortSectionPolicy::Alignment; 590 if (S == "name") 591 return SortSectionPolicy::Name; 592 if (!S.empty()) 593 error("unknown --sort-section rule: " + S); 594 return SortSectionPolicy::Default; 595 } 596 597 static OrphanHandlingPolicy getOrphanHandling(opt::InputArgList &Args) { 598 StringRef S = Args.getLastArgValue(OPT_orphan_handling, "place"); 599 if (S == "warn") 600 return OrphanHandlingPolicy::Warn; 601 if (S == "error") 602 return OrphanHandlingPolicy::Error; 603 if (S != "place") 604 error("unknown --orphan-handling mode: " + S); 605 return OrphanHandlingPolicy::Place; 606 } 607 608 // Parse --build-id or --build-id=<style>. We handle "tree" as a 609 // synonym for "sha1" because all our hash functions including 610 // -build-id=sha1 are actually tree hashes for performance reasons. 611 static std::pair<BuildIdKind, std::vector<uint8_t>> 612 getBuildId(opt::InputArgList &Args) { 613 auto *Arg = Args.getLastArg(OPT_build_id, OPT_build_id_eq); 614 if (!Arg) 615 return {BuildIdKind::None, {}}; 616 617 if (Arg->getOption().getID() == OPT_build_id) 618 return {BuildIdKind::Fast, {}}; 619 620 StringRef S = Arg->getValue(); 621 if (S == "fast") 622 return {BuildIdKind::Fast, {}}; 623 if (S == "md5") 624 return {BuildIdKind::Md5, {}}; 625 if (S == "sha1" || S == "tree") 626 return {BuildIdKind::Sha1, {}}; 627 if (S == "uuid") 628 return {BuildIdKind::Uuid, {}}; 629 if (S.startswith("0x")) 630 return {BuildIdKind::Hexstring, parseHex(S.substr(2))}; 631 632 if (S != "none") 633 error("unknown --build-id style: " + S); 634 return {BuildIdKind::None, {}}; 635 } 636 637 static std::pair<bool, bool> getPackDynRelocs(opt::InputArgList &Args) { 638 StringRef S = Args.getLastArgValue(OPT_pack_dyn_relocs, "none"); 639 if (S == "android") 640 return {true, false}; 641 if (S == "relr") 642 return {false, true}; 643 if (S == "android+relr") 644 return {true, true}; 645 646 if (S != "none") 647 error("unknown -pack-dyn-relocs format: " + S); 648 return {false, false}; 649 } 650 651 static void readCallGraph(MemoryBufferRef MB) { 652 // Build a map from symbol name to section 653 DenseMap<StringRef, Symbol *> Map; 654 for (InputFile *File : ObjectFiles) 655 for (Symbol *Sym : File->getSymbols()) 656 Map[Sym->getName()] = Sym; 657 658 auto FindSection = [&](StringRef Name) -> InputSectionBase * { 659 Symbol *Sym = Map.lookup(Name); 660 if (!Sym) { 661 if (Config->WarnSymbolOrdering) 662 warn(MB.getBufferIdentifier() + ": no such symbol: " + Name); 663 return nullptr; 664 } 665 maybeWarnUnorderableSymbol(Sym); 666 667 if (Defined *DR = dyn_cast_or_null<Defined>(Sym)) 668 return dyn_cast_or_null<InputSectionBase>(DR->Section); 669 return nullptr; 670 }; 671 672 for (StringRef Line : args::getLines(MB)) { 673 SmallVector<StringRef, 3> Fields; 674 Line.split(Fields, ' '); 675 uint64_t Count; 676 677 if (Fields.size() != 3 || !to_integer(Fields[2], Count)) { 678 error(MB.getBufferIdentifier() + ": parse error"); 679 return; 680 } 681 682 if (InputSectionBase *From = FindSection(Fields[0])) 683 if (InputSectionBase *To = FindSection(Fields[1])) 684 Config->CallGraphProfile[std::make_pair(From, To)] += Count; 685 } 686 } 687 688 template <class ELFT> static void readCallGraphsFromObjectFiles() { 689 for (auto File : ObjectFiles) { 690 auto *Obj = cast<ObjFile<ELFT>>(File); 691 692 for (const Elf_CGProfile_Impl<ELFT> &CGPE : Obj->CGProfile) { 693 auto *FromSym = dyn_cast<Defined>(&Obj->getSymbol(CGPE.cgp_from)); 694 auto *ToSym = dyn_cast<Defined>(&Obj->getSymbol(CGPE.cgp_to)); 695 if (!FromSym || !ToSym) 696 continue; 697 698 auto *From = dyn_cast_or_null<InputSectionBase>(FromSym->Section); 699 auto *To = dyn_cast_or_null<InputSectionBase>(ToSym->Section); 700 if (From && To) 701 Config->CallGraphProfile[{From, To}] += CGPE.cgp_weight; 702 } 703 } 704 } 705 706 static bool getCompressDebugSections(opt::InputArgList &Args) { 707 StringRef S = Args.getLastArgValue(OPT_compress_debug_sections, "none"); 708 if (S == "none") 709 return false; 710 if (S != "zlib") 711 error("unknown --compress-debug-sections value: " + S); 712 if (!zlib::isAvailable()) 713 error("--compress-debug-sections: zlib is not available"); 714 return true; 715 } 716 717 static std::pair<StringRef, StringRef> getOldNewOptions(opt::InputArgList &Args, 718 unsigned Id) { 719 auto *Arg = Args.getLastArg(Id); 720 if (!Arg) 721 return {"", ""}; 722 723 StringRef S = Arg->getValue(); 724 std::pair<StringRef, StringRef> Ret = S.split(';'); 725 if (Ret.second.empty()) 726 error(Arg->getSpelling() + " expects 'old;new' format, but got " + S); 727 return Ret; 728 } 729 730 // Parse the symbol ordering file and warn for any duplicate entries. 731 static std::vector<StringRef> getSymbolOrderingFile(MemoryBufferRef MB) { 732 SetVector<StringRef> Names; 733 for (StringRef S : args::getLines(MB)) 734 if (!Names.insert(S) && Config->WarnSymbolOrdering) 735 warn(MB.getBufferIdentifier() + ": duplicate ordered symbol: " + S); 736 737 return Names.takeVector(); 738 } 739 740 static void parseClangOption(StringRef Opt, const Twine &Msg) { 741 std::string Err; 742 raw_string_ostream OS(Err); 743 744 const char *Argv[] = {Config->ProgName.data(), Opt.data()}; 745 if (cl::ParseCommandLineOptions(2, Argv, "", &OS)) 746 return; 747 OS.flush(); 748 error(Msg + ": " + StringRef(Err).trim()); 749 } 750 751 // Initializes Config members by the command line options. 752 void LinkerDriver::readConfigs(opt::InputArgList &Args) { 753 errorHandler().Verbose = Args.hasArg(OPT_verbose); 754 errorHandler().FatalWarnings = 755 Args.hasFlag(OPT_fatal_warnings, OPT_no_fatal_warnings, false); 756 ThreadsEnabled = Args.hasFlag(OPT_threads, OPT_no_threads, true); 757 758 Config->AllowMultipleDefinition = 759 Args.hasFlag(OPT_allow_multiple_definition, 760 OPT_no_allow_multiple_definition, false) || 761 hasZOption(Args, "muldefs"); 762 Config->AuxiliaryList = args::getStrings(Args, OPT_auxiliary); 763 Config->Bsymbolic = Args.hasArg(OPT_Bsymbolic); 764 Config->BsymbolicFunctions = Args.hasArg(OPT_Bsymbolic_functions); 765 Config->CheckSections = 766 Args.hasFlag(OPT_check_sections, OPT_no_check_sections, true); 767 Config->Chroot = Args.getLastArgValue(OPT_chroot); 768 Config->CompressDebugSections = getCompressDebugSections(Args); 769 Config->Cref = Args.hasFlag(OPT_cref, OPT_no_cref, false); 770 Config->DefineCommon = Args.hasFlag(OPT_define_common, OPT_no_define_common, 771 !Args.hasArg(OPT_relocatable)); 772 Config->Demangle = Args.hasFlag(OPT_demangle, OPT_no_demangle, true); 773 Config->DisableVerify = Args.hasArg(OPT_disable_verify); 774 Config->Discard = getDiscard(Args); 775 Config->DwoDir = Args.getLastArgValue(OPT_plugin_opt_dwo_dir_eq); 776 Config->DynamicLinker = getDynamicLinker(Args); 777 Config->EhFrameHdr = 778 Args.hasFlag(OPT_eh_frame_hdr, OPT_no_eh_frame_hdr, false); 779 Config->EmitLLVM = Args.hasArg(OPT_plugin_opt_emit_llvm, false); 780 Config->EmitRelocs = Args.hasArg(OPT_emit_relocs); 781 Config->CallGraphProfileSort = Args.hasFlag( 782 OPT_call_graph_profile_sort, OPT_no_call_graph_profile_sort, true); 783 Config->EnableNewDtags = 784 Args.hasFlag(OPT_enable_new_dtags, OPT_disable_new_dtags, true); 785 Config->Entry = Args.getLastArgValue(OPT_entry); 786 Config->ExecuteOnly = 787 Args.hasFlag(OPT_execute_only, OPT_no_execute_only, false); 788 Config->ExportDynamic = 789 Args.hasFlag(OPT_export_dynamic, OPT_no_export_dynamic, false); 790 Config->FilterList = args::getStrings(Args, OPT_filter); 791 Config->Fini = Args.getLastArgValue(OPT_fini, "_fini"); 792 Config->FixCortexA53Errata843419 = Args.hasArg(OPT_fix_cortex_a53_843419); 793 Config->GcSections = Args.hasFlag(OPT_gc_sections, OPT_no_gc_sections, false); 794 Config->GnuUnique = Args.hasFlag(OPT_gnu_unique, OPT_no_gnu_unique, true); 795 Config->GdbIndex = Args.hasFlag(OPT_gdb_index, OPT_no_gdb_index, false); 796 Config->ICF = getICF(Args); 797 Config->IgnoreDataAddressEquality = 798 Args.hasArg(OPT_ignore_data_address_equality); 799 Config->IgnoreFunctionAddressEquality = 800 Args.hasArg(OPT_ignore_function_address_equality); 801 Config->Init = Args.getLastArgValue(OPT_init, "_init"); 802 Config->LTOAAPipeline = Args.getLastArgValue(OPT_lto_aa_pipeline); 803 Config->LTODebugPassManager = Args.hasArg(OPT_lto_debug_pass_manager); 804 Config->LTONewPassManager = Args.hasArg(OPT_lto_new_pass_manager); 805 Config->LTONewPmPasses = Args.getLastArgValue(OPT_lto_newpm_passes); 806 Config->LTOO = args::getInteger(Args, OPT_lto_O, 2); 807 Config->LTOObjPath = Args.getLastArgValue(OPT_plugin_opt_obj_path_eq); 808 Config->LTOPartitions = args::getInteger(Args, OPT_lto_partitions, 1); 809 Config->LTOSampleProfile = Args.getLastArgValue(OPT_lto_sample_profile); 810 Config->MapFile = Args.getLastArgValue(OPT_Map); 811 Config->MipsGotSize = args::getInteger(Args, OPT_mips_got_size, 0xfff0); 812 Config->MergeArmExidx = 813 Args.hasFlag(OPT_merge_exidx_entries, OPT_no_merge_exidx_entries, true); 814 Config->NoinhibitExec = Args.hasArg(OPT_noinhibit_exec); 815 Config->Nostdlib = Args.hasArg(OPT_nostdlib); 816 Config->OFormatBinary = isOutputFormatBinary(Args); 817 Config->Omagic = Args.hasFlag(OPT_omagic, OPT_no_omagic, false); 818 Config->OptRemarksFilename = Args.getLastArgValue(OPT_opt_remarks_filename); 819 Config->OptRemarksWithHotness = Args.hasArg(OPT_opt_remarks_with_hotness); 820 Config->Optimize = args::getInteger(Args, OPT_O, 1); 821 Config->OrphanHandling = getOrphanHandling(Args); 822 Config->OutputFile = Args.getLastArgValue(OPT_o); 823 Config->Pie = Args.hasFlag(OPT_pie, OPT_no_pie, false); 824 Config->PrintIcfSections = 825 Args.hasFlag(OPT_print_icf_sections, OPT_no_print_icf_sections, false); 826 Config->PrintGcSections = 827 Args.hasFlag(OPT_print_gc_sections, OPT_no_print_gc_sections, false); 828 Config->Rpath = getRpath(Args); 829 Config->Relocatable = Args.hasArg(OPT_relocatable); 830 Config->SaveTemps = Args.hasArg(OPT_save_temps); 831 Config->SearchPaths = args::getStrings(Args, OPT_library_path); 832 Config->SectionStartMap = getSectionStartMap(Args); 833 Config->Shared = Args.hasArg(OPT_shared); 834 Config->SingleRoRx = Args.hasArg(OPT_no_rosegment); 835 Config->SoName = Args.getLastArgValue(OPT_soname); 836 Config->SortSection = getSortSection(Args); 837 Config->SplitStackAdjustSize = args::getInteger(Args, OPT_split_stack_adjust_size, 16384); 838 Config->Strip = getStrip(Args); 839 Config->Sysroot = Args.getLastArgValue(OPT_sysroot); 840 Config->Target1Rel = Args.hasFlag(OPT_target1_rel, OPT_target1_abs, false); 841 Config->Target2 = getTarget2(Args); 842 Config->ThinLTOCacheDir = Args.getLastArgValue(OPT_thinlto_cache_dir); 843 Config->ThinLTOCachePolicy = CHECK( 844 parseCachePruningPolicy(Args.getLastArgValue(OPT_thinlto_cache_policy)), 845 "--thinlto-cache-policy: invalid cache policy"); 846 Config->ThinLTOEmitImportsFiles = 847 Args.hasArg(OPT_plugin_opt_thinlto_emit_imports_files); 848 Config->ThinLTOIndexOnly = Args.hasArg(OPT_plugin_opt_thinlto_index_only) || 849 Args.hasArg(OPT_plugin_opt_thinlto_index_only_eq); 850 Config->ThinLTOIndexOnlyArg = 851 Args.getLastArgValue(OPT_plugin_opt_thinlto_index_only_eq); 852 Config->ThinLTOJobs = args::getInteger(Args, OPT_thinlto_jobs, -1u); 853 Config->ThinLTOObjectSuffixReplace = 854 getOldNewOptions(Args, OPT_plugin_opt_thinlto_object_suffix_replace_eq); 855 Config->ThinLTOPrefixReplace = 856 getOldNewOptions(Args, OPT_plugin_opt_thinlto_prefix_replace_eq); 857 Config->Trace = Args.hasArg(OPT_trace); 858 Config->Undefined = args::getStrings(Args, OPT_undefined); 859 Config->UndefinedVersion = 860 Args.hasFlag(OPT_undefined_version, OPT_no_undefined_version, true); 861 Config->UseAndroidRelrTags = Args.hasFlag( 862 OPT_use_android_relr_tags, OPT_no_use_android_relr_tags, false); 863 Config->UnresolvedSymbols = getUnresolvedSymbolPolicy(Args); 864 Config->WarnBackrefs = 865 Args.hasFlag(OPT_warn_backrefs, OPT_no_warn_backrefs, false); 866 Config->WarnCommon = Args.hasFlag(OPT_warn_common, OPT_no_warn_common, false); 867 Config->WarnIfuncTextrel = 868 Args.hasFlag(OPT_warn_ifunc_textrel, OPT_no_warn_ifunc_textrel, false); 869 Config->WarnSymbolOrdering = 870 Args.hasFlag(OPT_warn_symbol_ordering, OPT_no_warn_symbol_ordering, true); 871 Config->ZCombreloc = getZFlag(Args, "combreloc", "nocombreloc", true); 872 Config->ZCopyreloc = getZFlag(Args, "copyreloc", "nocopyreloc", true); 873 Config->ZExecstack = getZFlag(Args, "execstack", "noexecstack", false); 874 Config->ZGlobal = hasZOption(Args, "global"); 875 Config->ZHazardplt = hasZOption(Args, "hazardplt"); 876 Config->ZInitfirst = hasZOption(Args, "initfirst"); 877 Config->ZInterpose = hasZOption(Args, "interpose"); 878 Config->ZKeepTextSectionPrefix = getZFlag( 879 Args, "keep-text-section-prefix", "nokeep-text-section-prefix", false); 880 Config->ZNodefaultlib = hasZOption(Args, "nodefaultlib"); 881 Config->ZNodelete = hasZOption(Args, "nodelete"); 882 Config->ZNodlopen = hasZOption(Args, "nodlopen"); 883 Config->ZNow = getZFlag(Args, "now", "lazy", false); 884 Config->ZOrigin = hasZOption(Args, "origin"); 885 Config->ZRelro = getZFlag(Args, "relro", "norelro", true); 886 Config->ZRetpolineplt = hasZOption(Args, "retpolineplt"); 887 Config->ZRodynamic = hasZOption(Args, "rodynamic"); 888 Config->ZStackSize = args::getZOptionValue(Args, OPT_z, "stack-size", 0); 889 Config->ZText = getZFlag(Args, "text", "notext", true); 890 Config->ZWxneeded = hasZOption(Args, "wxneeded"); 891 892 // Parse LTO options. 893 if (auto *Arg = Args.getLastArg(OPT_plugin_opt_mcpu_eq)) 894 parseClangOption(Saver.save("-mcpu=" + StringRef(Arg->getValue())), 895 Arg->getSpelling()); 896 897 for (auto *Arg : Args.filtered(OPT_plugin_opt)) 898 parseClangOption(Arg->getValue(), Arg->getSpelling()); 899 900 // Parse -mllvm options. 901 for (auto *Arg : Args.filtered(OPT_mllvm)) 902 parseClangOption(Arg->getValue(), Arg->getSpelling()); 903 904 if (Config->LTOO > 3) 905 error("invalid optimization level for LTO: " + Twine(Config->LTOO)); 906 if (Config->LTOPartitions == 0) 907 error("--lto-partitions: number of threads must be > 0"); 908 if (Config->ThinLTOJobs == 0) 909 error("--thinlto-jobs: number of threads must be > 0"); 910 911 if (Config->SplitStackAdjustSize < 0) 912 error("--split-stack-adjust-size: size must be >= 0"); 913 914 // Parse ELF{32,64}{LE,BE} and CPU type. 915 if (auto *Arg = Args.getLastArg(OPT_m)) { 916 StringRef S = Arg->getValue(); 917 std::tie(Config->EKind, Config->EMachine, Config->OSABI) = 918 parseEmulation(S); 919 Config->MipsN32Abi = (S == "elf32btsmipn32" || S == "elf32ltsmipn32"); 920 Config->Emulation = S; 921 } 922 923 // Parse -hash-style={sysv,gnu,both}. 924 if (auto *Arg = Args.getLastArg(OPT_hash_style)) { 925 StringRef S = Arg->getValue(); 926 if (S == "sysv") 927 Config->SysvHash = true; 928 else if (S == "gnu") 929 Config->GnuHash = true; 930 else if (S == "both") 931 Config->SysvHash = Config->GnuHash = true; 932 else 933 error("unknown -hash-style: " + S); 934 } 935 936 if (Args.hasArg(OPT_print_map)) 937 Config->MapFile = "-"; 938 939 // --omagic is an option to create old-fashioned executables in which 940 // .text segments are writable. Today, the option is still in use to 941 // create special-purpose programs such as boot loaders. It doesn't 942 // make sense to create PT_GNU_RELRO for such executables. 943 if (Config->Omagic) 944 Config->ZRelro = false; 945 946 std::tie(Config->BuildId, Config->BuildIdVector) = getBuildId(Args); 947 948 std::tie(Config->AndroidPackDynRelocs, Config->RelrPackDynRelocs) = 949 getPackDynRelocs(Args); 950 951 if (auto *Arg = Args.getLastArg(OPT_symbol_ordering_file)) 952 if (Optional<MemoryBufferRef> Buffer = readFile(Arg->getValue())) 953 Config->SymbolOrderingFile = getSymbolOrderingFile(*Buffer); 954 955 // If --retain-symbol-file is used, we'll keep only the symbols listed in 956 // the file and discard all others. 957 if (auto *Arg = Args.getLastArg(OPT_retain_symbols_file)) { 958 Config->DefaultSymbolVersion = VER_NDX_LOCAL; 959 if (Optional<MemoryBufferRef> Buffer = readFile(Arg->getValue())) 960 for (StringRef S : args::getLines(*Buffer)) 961 Config->VersionScriptGlobals.push_back( 962 {S, /*IsExternCpp*/ false, /*HasWildcard*/ false}); 963 } 964 965 bool HasExportDynamic = 966 Args.hasFlag(OPT_export_dynamic, OPT_no_export_dynamic, false); 967 968 // Parses -dynamic-list and -export-dynamic-symbol. They make some 969 // symbols private. Note that -export-dynamic takes precedence over them 970 // as it says all symbols should be exported. 971 if (!HasExportDynamic) { 972 for (auto *Arg : Args.filtered(OPT_dynamic_list)) 973 if (Optional<MemoryBufferRef> Buffer = readFile(Arg->getValue())) 974 readDynamicList(*Buffer); 975 976 for (auto *Arg : Args.filtered(OPT_export_dynamic_symbol)) 977 Config->DynamicList.push_back( 978 {Arg->getValue(), /*IsExternCpp*/ false, /*HasWildcard*/ false}); 979 } 980 981 // If --export-dynamic-symbol=foo is given and symbol foo is defined in 982 // an object file in an archive file, that object file should be pulled 983 // out and linked. (It doesn't have to behave like that from technical 984 // point of view, but this is needed for compatibility with GNU.) 985 for (auto *Arg : Args.filtered(OPT_export_dynamic_symbol)) 986 Config->Undefined.push_back(Arg->getValue()); 987 988 for (auto *Arg : Args.filtered(OPT_version_script)) 989 if (Optional<std::string> Path = searchScript(Arg->getValue())) { 990 if (Optional<MemoryBufferRef> Buffer = readFile(*Path)) 991 readVersionScript(*Buffer); 992 } else { 993 error(Twine("cannot find version script ") + Arg->getValue()); 994 } 995 } 996 997 // Some Config members do not directly correspond to any particular 998 // command line options, but computed based on other Config values. 999 // This function initialize such members. See Config.h for the details 1000 // of these values. 1001 static void setConfigs(opt::InputArgList &Args) { 1002 ELFKind K = Config->EKind; 1003 uint16_t M = Config->EMachine; 1004 1005 Config->CopyRelocs = (Config->Relocatable || Config->EmitRelocs); 1006 Config->Is64 = (K == ELF64LEKind || K == ELF64BEKind); 1007 Config->IsLE = (K == ELF32LEKind || K == ELF64LEKind); 1008 Config->Endianness = Config->IsLE ? endianness::little : endianness::big; 1009 Config->IsMips64EL = (K == ELF64LEKind && M == EM_MIPS); 1010 Config->Pic = Config->Pie || Config->Shared; 1011 Config->Wordsize = Config->Is64 ? 8 : 4; 1012 1013 // ELF defines two different ways to store relocation addends as shown below: 1014 // 1015 // Rel: Addends are stored to the location where relocations are applied. 1016 // Rela: Addends are stored as part of relocation entry. 1017 // 1018 // In other words, Rela makes it easy to read addends at the price of extra 1019 // 4 or 8 byte for each relocation entry. We don't know why ELF defined two 1020 // different mechanisms in the first place, but this is how the spec is 1021 // defined. 1022 // 1023 // You cannot choose which one, Rel or Rela, you want to use. Instead each 1024 // ABI defines which one you need to use. The following expression expresses 1025 // that. 1026 Config->IsRela = M == EM_AARCH64 || M == EM_AMDGPU || M == EM_HEXAGON || 1027 M == EM_PPC || M == EM_PPC64 || M == EM_RISCV || 1028 M == EM_X86_64; 1029 1030 // If the output uses REL relocations we must store the dynamic relocation 1031 // addends to the output sections. We also store addends for RELA relocations 1032 // if --apply-dynamic-relocs is used. 1033 // We default to not writing the addends when using RELA relocations since 1034 // any standard conforming tool can find it in r_addend. 1035 Config->WriteAddends = Args.hasFlag(OPT_apply_dynamic_relocs, 1036 OPT_no_apply_dynamic_relocs, false) || 1037 !Config->IsRela; 1038 1039 Config->TocOptimize = 1040 Args.hasFlag(OPT_toc_optimize, OPT_no_toc_optimize, M == EM_PPC64); 1041 } 1042 1043 // Returns a value of "-format" option. 1044 static bool isFormatBinary(StringRef S) { 1045 if (S == "binary") 1046 return true; 1047 if (S == "elf" || S == "default") 1048 return false; 1049 error("unknown -format value: " + S + 1050 " (supported formats: elf, default, binary)"); 1051 return false; 1052 } 1053 1054 void LinkerDriver::createFiles(opt::InputArgList &Args) { 1055 // For --{push,pop}-state. 1056 std::vector<std::tuple<bool, bool, bool>> Stack; 1057 1058 // Iterate over argv to process input files and positional arguments. 1059 for (auto *Arg : Args) { 1060 switch (Arg->getOption().getUnaliasedOption().getID()) { 1061 case OPT_library: 1062 addLibrary(Arg->getValue()); 1063 break; 1064 case OPT_INPUT: 1065 addFile(Arg->getValue(), /*WithLOption=*/false); 1066 break; 1067 case OPT_defsym: { 1068 StringRef From; 1069 StringRef To; 1070 std::tie(From, To) = StringRef(Arg->getValue()).split('='); 1071 if (From.empty() || To.empty()) 1072 error("-defsym: syntax error: " + StringRef(Arg->getValue())); 1073 else 1074 readDefsym(From, MemoryBufferRef(To, "-defsym")); 1075 break; 1076 } 1077 case OPT_script: 1078 if (Optional<std::string> Path = searchScript(Arg->getValue())) { 1079 if (Optional<MemoryBufferRef> MB = readFile(*Path)) 1080 readLinkerScript(*MB); 1081 break; 1082 } 1083 error(Twine("cannot find linker script ") + Arg->getValue()); 1084 break; 1085 case OPT_as_needed: 1086 Config->AsNeeded = true; 1087 break; 1088 case OPT_format: 1089 Config->FormatBinary = isFormatBinary(Arg->getValue()); 1090 break; 1091 case OPT_no_as_needed: 1092 Config->AsNeeded = false; 1093 break; 1094 case OPT_Bstatic: 1095 Config->Static = true; 1096 break; 1097 case OPT_Bdynamic: 1098 Config->Static = false; 1099 break; 1100 case OPT_whole_archive: 1101 InWholeArchive = true; 1102 break; 1103 case OPT_no_whole_archive: 1104 InWholeArchive = false; 1105 break; 1106 case OPT_just_symbols: 1107 if (Optional<MemoryBufferRef> MB = readFile(Arg->getValue())) { 1108 Files.push_back(createObjectFile(*MB)); 1109 Files.back()->JustSymbols = true; 1110 } 1111 break; 1112 case OPT_start_group: 1113 if (InputFile::IsInGroup) 1114 error("nested --start-group"); 1115 InputFile::IsInGroup = true; 1116 break; 1117 case OPT_end_group: 1118 if (!InputFile::IsInGroup) 1119 error("stray --end-group"); 1120 InputFile::IsInGroup = false; 1121 ++InputFile::NextGroupId; 1122 break; 1123 case OPT_start_lib: 1124 if (InLib) 1125 error("nested --start-lib"); 1126 if (InputFile::IsInGroup) 1127 error("may not nest --start-lib in --start-group"); 1128 InLib = true; 1129 InputFile::IsInGroup = true; 1130 break; 1131 case OPT_end_lib: 1132 if (!InLib) 1133 error("stray --end-lib"); 1134 InLib = false; 1135 InputFile::IsInGroup = false; 1136 ++InputFile::NextGroupId; 1137 break; 1138 case OPT_push_state: 1139 Stack.emplace_back(Config->AsNeeded, Config->Static, InWholeArchive); 1140 break; 1141 case OPT_pop_state: 1142 if (Stack.empty()) { 1143 error("unbalanced --push-state/--pop-state"); 1144 break; 1145 } 1146 std::tie(Config->AsNeeded, Config->Static, InWholeArchive) = Stack.back(); 1147 Stack.pop_back(); 1148 break; 1149 } 1150 } 1151 1152 if (Files.empty() && errorCount() == 0) 1153 error("no input files"); 1154 } 1155 1156 // If -m <machine_type> was not given, infer it from object files. 1157 void LinkerDriver::inferMachineType() { 1158 if (Config->EKind != ELFNoneKind) 1159 return; 1160 1161 for (InputFile *F : Files) { 1162 if (F->EKind == ELFNoneKind) 1163 continue; 1164 Config->EKind = F->EKind; 1165 Config->EMachine = F->EMachine; 1166 Config->OSABI = F->OSABI; 1167 Config->MipsN32Abi = Config->EMachine == EM_MIPS && isMipsN32Abi(F); 1168 return; 1169 } 1170 error("target emulation unknown: -m or at least one .o file required"); 1171 } 1172 1173 // Parse -z max-page-size=<value>. The default value is defined by 1174 // each target. 1175 static uint64_t getMaxPageSize(opt::InputArgList &Args) { 1176 uint64_t Val = args::getZOptionValue(Args, OPT_z, "max-page-size", 1177 Target->DefaultMaxPageSize); 1178 if (!isPowerOf2_64(Val)) 1179 error("max-page-size: value isn't a power of 2"); 1180 return Val; 1181 } 1182 1183 // Parses -image-base option. 1184 static Optional<uint64_t> getImageBase(opt::InputArgList &Args) { 1185 // Because we are using "Config->MaxPageSize" here, this function has to be 1186 // called after the variable is initialized. 1187 auto *Arg = Args.getLastArg(OPT_image_base); 1188 if (!Arg) 1189 return None; 1190 1191 StringRef S = Arg->getValue(); 1192 uint64_t V; 1193 if (!to_integer(S, V)) { 1194 error("-image-base: number expected, but got " + S); 1195 return 0; 1196 } 1197 if ((V % Config->MaxPageSize) != 0) 1198 warn("-image-base: address isn't multiple of page size: " + S); 1199 return V; 1200 } 1201 1202 // Parses `--exclude-libs=lib,lib,...`. 1203 // The library names may be delimited by commas or colons. 1204 static DenseSet<StringRef> getExcludeLibs(opt::InputArgList &Args) { 1205 DenseSet<StringRef> Ret; 1206 for (auto *Arg : Args.filtered(OPT_exclude_libs)) { 1207 StringRef S = Arg->getValue(); 1208 for (;;) { 1209 size_t Pos = S.find_first_of(",:"); 1210 if (Pos == StringRef::npos) 1211 break; 1212 Ret.insert(S.substr(0, Pos)); 1213 S = S.substr(Pos + 1); 1214 } 1215 Ret.insert(S); 1216 } 1217 return Ret; 1218 } 1219 1220 // Handles the -exclude-libs option. If a static library file is specified 1221 // by the -exclude-libs option, all public symbols from the archive become 1222 // private unless otherwise specified by version scripts or something. 1223 // A special library name "ALL" means all archive files. 1224 // 1225 // This is not a popular option, but some programs such as bionic libc use it. 1226 template <class ELFT> 1227 static void excludeLibs(opt::InputArgList &Args) { 1228 DenseSet<StringRef> Libs = getExcludeLibs(Args); 1229 bool All = Libs.count("ALL"); 1230 1231 auto Visit = [&](InputFile *File) { 1232 if (!File->ArchiveName.empty()) 1233 if (All || Libs.count(path::filename(File->ArchiveName))) 1234 for (Symbol *Sym : File->getSymbols()) 1235 if (!Sym->isLocal() && Sym->File == File) 1236 Sym->VersionId = VER_NDX_LOCAL; 1237 }; 1238 1239 for (InputFile *File : ObjectFiles) 1240 Visit(File); 1241 1242 for (BitcodeFile *File : BitcodeFiles) 1243 Visit(File); 1244 } 1245 1246 // Force Sym to be entered in the output. Used for -u or equivalent. 1247 template <class ELFT> static void handleUndefined(StringRef Name) { 1248 Symbol *Sym = Symtab->find(Name); 1249 if (!Sym) 1250 return; 1251 1252 // Since symbol S may not be used inside the program, LTO may 1253 // eliminate it. Mark the symbol as "used" to prevent it. 1254 Sym->IsUsedInRegularObj = true; 1255 1256 if (Sym->isLazy()) 1257 Symtab->fetchLazy<ELFT>(Sym); 1258 } 1259 1260 template <class ELFT> static void handleLibcall(StringRef Name) { 1261 Symbol *Sym = Symtab->find(Name); 1262 if (!Sym || !Sym->isLazy()) 1263 return; 1264 1265 MemoryBufferRef MB; 1266 if (auto *LO = dyn_cast<LazyObject>(Sym)) 1267 MB = LO->File->MB; 1268 else 1269 MB = cast<LazyArchive>(Sym)->getMemberBuffer(); 1270 1271 if (isBitcode(MB)) 1272 Symtab->fetchLazy<ELFT>(Sym); 1273 } 1274 1275 // If all references to a DSO happen to be weak, the DSO is not added 1276 // to DT_NEEDED. If that happens, we need to eliminate shared symbols 1277 // created from the DSO. Otherwise, they become dangling references 1278 // that point to a non-existent DSO. 1279 template <class ELFT> static void demoteSharedSymbols() { 1280 for (Symbol *Sym : Symtab->getSymbols()) { 1281 if (auto *S = dyn_cast<SharedSymbol>(Sym)) { 1282 if (!S->getFile<ELFT>().IsNeeded) { 1283 bool Used = S->Used; 1284 replaceSymbol<Undefined>(S, nullptr, S->getName(), STB_WEAK, S->StOther, 1285 S->Type); 1286 S->Used = Used; 1287 } 1288 } 1289 } 1290 } 1291 1292 // The section referred to by S is considered address-significant. Set the 1293 // KeepUnique flag on the section if appropriate. 1294 static void markAddrsig(Symbol *S) { 1295 if (auto *D = dyn_cast_or_null<Defined>(S)) 1296 if (D->Section) 1297 // We don't need to keep text sections unique under --icf=all even if they 1298 // are address-significant. 1299 if (Config->ICF == ICFLevel::Safe || !(D->Section->Flags & SHF_EXECINSTR)) 1300 D->Section->KeepUnique = true; 1301 } 1302 1303 // Record sections that define symbols mentioned in --keep-unique <symbol> 1304 // and symbols referred to by address-significance tables. These sections are 1305 // ineligible for ICF. 1306 template <class ELFT> 1307 static void findKeepUniqueSections(opt::InputArgList &Args) { 1308 for (auto *Arg : Args.filtered(OPT_keep_unique)) { 1309 StringRef Name = Arg->getValue(); 1310 auto *D = dyn_cast_or_null<Defined>(Symtab->find(Name)); 1311 if (!D || !D->Section) { 1312 warn("could not find symbol " + Name + " to keep unique"); 1313 continue; 1314 } 1315 D->Section->KeepUnique = true; 1316 } 1317 1318 // --icf=all --ignore-data-address-equality means that we can ignore 1319 // the dynsym and address-significance tables entirely. 1320 if (Config->ICF == ICFLevel::All && Config->IgnoreDataAddressEquality) 1321 return; 1322 1323 // Symbols in the dynsym could be address-significant in other executables 1324 // or DSOs, so we conservatively mark them as address-significant. 1325 for (Symbol *S : Symtab->getSymbols()) 1326 if (S->includeInDynsym()) 1327 markAddrsig(S); 1328 1329 // Visit the address-significance table in each object file and mark each 1330 // referenced symbol as address-significant. 1331 for (InputFile *F : ObjectFiles) { 1332 auto *Obj = cast<ObjFile<ELFT>>(F); 1333 ArrayRef<Symbol *> Syms = Obj->getSymbols(); 1334 if (Obj->AddrsigSec) { 1335 ArrayRef<uint8_t> Contents = 1336 check(Obj->getObj().getSectionContents(Obj->AddrsigSec)); 1337 const uint8_t *Cur = Contents.begin(); 1338 while (Cur != Contents.end()) { 1339 unsigned Size; 1340 const char *Err; 1341 uint64_t SymIndex = decodeULEB128(Cur, &Size, Contents.end(), &Err); 1342 if (Err) 1343 fatal(toString(F) + ": could not decode addrsig section: " + Err); 1344 markAddrsig(Syms[SymIndex]); 1345 Cur += Size; 1346 } 1347 } else { 1348 // If an object file does not have an address-significance table, 1349 // conservatively mark all of its symbols as address-significant. 1350 for (Symbol *S : Syms) 1351 markAddrsig(S); 1352 } 1353 } 1354 } 1355 1356 template <class ELFT> static Symbol *addUndefined(StringRef Name) { 1357 return Symtab->addUndefined<ELFT>(Name, STB_GLOBAL, STV_DEFAULT, 0, false, 1358 nullptr); 1359 } 1360 1361 // The --wrap option is a feature to rename symbols so that you can write 1362 // wrappers for existing functions. If you pass `-wrap=foo`, all 1363 // occurrences of symbol `foo` are resolved to `wrap_foo` (so, you are 1364 // expected to write `wrap_foo` function as a wrapper). The original 1365 // symbol becomes accessible as `real_foo`, so you can call that from your 1366 // wrapper. 1367 // 1368 // This data structure is instantiated for each -wrap option. 1369 struct WrappedSymbol { 1370 Symbol *Sym; 1371 Symbol *Real; 1372 Symbol *Wrap; 1373 }; 1374 1375 // Handles -wrap option. 1376 // 1377 // This function instantiates wrapper symbols. At this point, they seem 1378 // like they are not being used at all, so we explicitly set some flags so 1379 // that LTO won't eliminate them. 1380 template <class ELFT> 1381 static std::vector<WrappedSymbol> addWrappedSymbols(opt::InputArgList &Args) { 1382 std::vector<WrappedSymbol> V; 1383 DenseSet<StringRef> Seen; 1384 1385 for (auto *Arg : Args.filtered(OPT_wrap)) { 1386 StringRef Name = Arg->getValue(); 1387 if (!Seen.insert(Name).second) 1388 continue; 1389 1390 Symbol *Sym = Symtab->find(Name); 1391 if (!Sym) 1392 continue; 1393 1394 Symbol *Real = addUndefined<ELFT>(Saver.save("__real_" + Name)); 1395 Symbol *Wrap = addUndefined<ELFT>(Saver.save("__wrap_" + Name)); 1396 V.push_back({Sym, Real, Wrap}); 1397 1398 // We want to tell LTO not to inline symbols to be overwritten 1399 // because LTO doesn't know the final symbol contents after renaming. 1400 Real->CanInline = false; 1401 Sym->CanInline = false; 1402 1403 // Tell LTO not to eliminate these symbols. 1404 Sym->IsUsedInRegularObj = true; 1405 Wrap->IsUsedInRegularObj = true; 1406 } 1407 return V; 1408 } 1409 1410 // Do renaming for -wrap by updating pointers to symbols. 1411 // 1412 // When this function is executed, only InputFiles and symbol table 1413 // contain pointers to symbol objects. We visit them to replace pointers, 1414 // so that wrapped symbols are swapped as instructed by the command line. 1415 template <class ELFT> static void wrapSymbols(ArrayRef<WrappedSymbol> Wrapped) { 1416 DenseMap<Symbol *, Symbol *> Map; 1417 for (const WrappedSymbol &W : Wrapped) { 1418 Map[W.Sym] = W.Wrap; 1419 Map[W.Real] = W.Sym; 1420 } 1421 1422 // Update pointers in input files. 1423 parallelForEach(ObjectFiles, [&](InputFile *File) { 1424 std::vector<Symbol *> &Syms = File->getMutableSymbols(); 1425 for (size_t I = 0, E = Syms.size(); I != E; ++I) 1426 if (Symbol *S = Map.lookup(Syms[I])) 1427 Syms[I] = S; 1428 }); 1429 1430 // Update pointers in the symbol table. 1431 for (const WrappedSymbol &W : Wrapped) 1432 Symtab->wrap(W.Sym, W.Real, W.Wrap); 1433 } 1434 1435 static const char *LibcallRoutineNames[] = { 1436 #define HANDLE_LIBCALL(code, name) name, 1437 #include "llvm/IR/RuntimeLibcalls.def" 1438 #undef HANDLE_LIBCALL 1439 }; 1440 1441 // Do actual linking. Note that when this function is called, 1442 // all linker scripts have already been parsed. 1443 template <class ELFT> void LinkerDriver::link(opt::InputArgList &Args) { 1444 Target = getTarget(); 1445 InX<ELFT>::VerSym = nullptr; 1446 InX<ELFT>::VerNeed = nullptr; 1447 1448 Config->MaxPageSize = getMaxPageSize(Args); 1449 Config->ImageBase = getImageBase(Args); 1450 1451 // If a -hash-style option was not given, set to a default value, 1452 // which varies depending on the target. 1453 if (!Args.hasArg(OPT_hash_style)) { 1454 if (Config->EMachine == EM_MIPS) 1455 Config->SysvHash = true; 1456 else 1457 Config->SysvHash = Config->GnuHash = true; 1458 } 1459 1460 // Default output filename is "a.out" by the Unix tradition. 1461 if (Config->OutputFile.empty()) 1462 Config->OutputFile = "a.out"; 1463 1464 // Fail early if the output file or map file is not writable. If a user has a 1465 // long link, e.g. due to a large LTO link, they do not wish to run it and 1466 // find that it failed because there was a mistake in their command-line. 1467 if (auto E = tryCreateFile(Config->OutputFile)) 1468 error("cannot open output file " + Config->OutputFile + ": " + E.message()); 1469 if (auto E = tryCreateFile(Config->MapFile)) 1470 error("cannot open map file " + Config->MapFile + ": " + E.message()); 1471 if (errorCount()) 1472 return; 1473 1474 // Use default entry point name if no name was given via the command 1475 // line nor linker scripts. For some reason, MIPS entry point name is 1476 // different from others. 1477 Config->WarnMissingEntry = 1478 (!Config->Entry.empty() || (!Config->Shared && !Config->Relocatable)); 1479 if (Config->Entry.empty() && !Config->Relocatable) 1480 Config->Entry = (Config->EMachine == EM_MIPS) ? "__start" : "_start"; 1481 1482 // Handle --trace-symbol. 1483 for (auto *Arg : Args.filtered(OPT_trace_symbol)) 1484 Symtab->trace(Arg->getValue()); 1485 1486 // Add all files to the symbol table. This will add almost all 1487 // symbols that we need to the symbol table. 1488 for (InputFile *F : Files) 1489 Symtab->addFile<ELFT>(F); 1490 1491 // Now that we have every file, we can decide if we will need a 1492 // dynamic symbol table. 1493 // We need one if we were asked to export dynamic symbols or if we are 1494 // producing a shared library. 1495 // We also need one if any shared libraries are used and for pie executables 1496 // (probably because the dynamic linker needs it). 1497 Config->HasDynSymTab = 1498 !SharedFiles.empty() || Config->Pic || Config->ExportDynamic; 1499 1500 // Some symbols (such as __ehdr_start) are defined lazily only when there 1501 // are undefined symbols for them, so we add these to trigger that logic. 1502 for (StringRef Name : Script->ReferencedSymbols) 1503 addUndefined<ELFT>(Name); 1504 1505 // Handle the `--undefined <sym>` options. 1506 for (StringRef S : Config->Undefined) 1507 handleUndefined<ELFT>(S); 1508 1509 // If an entry symbol is in a static archive, pull out that file now. 1510 handleUndefined<ELFT>(Config->Entry); 1511 1512 // If any of our inputs are bitcode files, the LTO code generator may create 1513 // references to certain library functions that might not be explicit in the 1514 // bitcode file's symbol table. If any of those library functions are defined 1515 // in a bitcode file in an archive member, we need to arrange to use LTO to 1516 // compile those archive members by adding them to the link beforehand. 1517 // 1518 // However, adding all libcall symbols to the link can have undesired 1519 // consequences. For example, the libgcc implementation of 1520 // __sync_val_compare_and_swap_8 on 32-bit ARM pulls in an .init_array entry 1521 // that aborts the program if the Linux kernel does not support 64-bit 1522 // atomics, which would prevent the program from running even if it does not 1523 // use 64-bit atomics. 1524 // 1525 // Therefore, we only add libcall symbols to the link before LTO if we have 1526 // to, i.e. if the symbol's definition is in bitcode. Any other required 1527 // libcall symbols will be added to the link after LTO when we add the LTO 1528 // object file to the link. 1529 if (!BitcodeFiles.empty()) 1530 for (const char *S : LibcallRoutineNames) 1531 handleLibcall<ELFT>(S); 1532 1533 // Return if there were name resolution errors. 1534 if (errorCount()) 1535 return; 1536 1537 // Now when we read all script files, we want to finalize order of linker 1538 // script commands, which can be not yet final because of INSERT commands. 1539 Script->processInsertCommands(); 1540 1541 // We want to declare linker script's symbols early, 1542 // so that we can version them. 1543 // They also might be exported if referenced by DSOs. 1544 Script->declareSymbols(); 1545 1546 // Handle the -exclude-libs option. 1547 if (Args.hasArg(OPT_exclude_libs)) 1548 excludeLibs<ELFT>(Args); 1549 1550 // Create ElfHeader early. We need a dummy section in 1551 // addReservedSymbols to mark the created symbols as not absolute. 1552 Out::ElfHeader = make<OutputSection>("", 0, SHF_ALLOC); 1553 Out::ElfHeader->Size = sizeof(typename ELFT::Ehdr); 1554 1555 // We need to create some reserved symbols such as _end. Create them. 1556 if (!Config->Relocatable) 1557 addReservedSymbols(); 1558 1559 // Apply version scripts. 1560 // 1561 // For a relocatable output, version scripts don't make sense, and 1562 // parsing a symbol version string (e.g. dropping "@ver1" from a symbol 1563 // name "foo@ver1") rather do harm, so we don't call this if -r is given. 1564 if (!Config->Relocatable) 1565 Symtab->scanVersionScript(); 1566 1567 // Create wrapped symbols for -wrap option. 1568 std::vector<WrappedSymbol> Wrapped = addWrappedSymbols<ELFT>(Args); 1569 1570 // Do link-time optimization if given files are LLVM bitcode files. 1571 // This compiles bitcode files into real object files. 1572 // 1573 // With this the symbol table should be complete. After this, no new names 1574 // except a few linker-synthesized ones will be added to the symbol table. 1575 Symtab->addCombinedLTOObject<ELFT>(); 1576 if (errorCount()) 1577 return; 1578 1579 // If -thinlto-index-only is given, we should create only "index 1580 // files" and not object files. Index file creation is already done 1581 // in addCombinedLTOObject, so we are done if that's the case. 1582 if (Config->ThinLTOIndexOnly) 1583 return; 1584 1585 // Likewise, --plugin-opt=emit-llvm is an option to make LTO create 1586 // an output file in bitcode and exit, so that you can just get a 1587 // combined bitcode file. 1588 if (Config->EmitLLVM) 1589 return; 1590 1591 // Apply symbol renames for -wrap. 1592 if (!Wrapped.empty()) 1593 wrapSymbols<ELFT>(Wrapped); 1594 1595 // Now that we have a complete list of input files. 1596 // Beyond this point, no new files are added. 1597 // Aggregate all input sections into one place. 1598 for (InputFile *F : ObjectFiles) 1599 for (InputSectionBase *S : F->getSections()) 1600 if (S && S != &InputSection::Discarded) 1601 InputSections.push_back(S); 1602 for (BinaryFile *F : BinaryFiles) 1603 for (InputSectionBase *S : F->getSections()) 1604 InputSections.push_back(cast<InputSection>(S)); 1605 1606 // We do not want to emit debug sections if --strip-all 1607 // or -strip-debug are given. 1608 if (Config->Strip != StripPolicy::None) 1609 llvm::erase_if(InputSections, [](InputSectionBase *S) { 1610 return S->Name.startswith(".debug") || S->Name.startswith(".zdebug"); 1611 }); 1612 1613 Config->EFlags = Target->calcEFlags(); 1614 1615 if (Config->EMachine == EM_ARM) { 1616 // FIXME: These warnings can be removed when lld only uses these features 1617 // when the input objects have been compiled with an architecture that 1618 // supports them. 1619 if (Config->ARMHasBlx == false) 1620 warn("lld uses blx instruction, no object with architecture supporting " 1621 "feature detected"); 1622 } 1623 1624 // This adds a .comment section containing a version string. We have to add it 1625 // before mergeSections because the .comment section is a mergeable section. 1626 if (!Config->Relocatable) 1627 InputSections.push_back(createCommentSection()); 1628 1629 // Do size optimizations: garbage collection, merging of SHF_MERGE sections 1630 // and identical code folding. 1631 splitSections<ELFT>(); 1632 markLive<ELFT>(); 1633 demoteSharedSymbols<ELFT>(); 1634 mergeSections(); 1635 if (Config->ICF != ICFLevel::None) { 1636 findKeepUniqueSections<ELFT>(Args); 1637 doIcf<ELFT>(); 1638 } 1639 1640 // Read the callgraph now that we know what was gced or icfed 1641 if (Config->CallGraphProfileSort) { 1642 if (auto *Arg = Args.getLastArg(OPT_call_graph_ordering_file)) 1643 if (Optional<MemoryBufferRef> Buffer = readFile(Arg->getValue())) 1644 readCallGraph(*Buffer); 1645 readCallGraphsFromObjectFiles<ELFT>(); 1646 } 1647 1648 // Write the result to the file. 1649 writeResult<ELFT>(); 1650 } 1651