Driver.cpp revision 363496
1//===- Driver.cpp ---------------------------------------------------------===//
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// The driver drives the entire linking process. It is responsible for
10// parsing command line options and doing whatever it is instructed to do.
11//
12// One notable thing in the LLD's driver when compared to other linkers is
13// that the LLD's driver is agnostic on the host operating system.
14// Other linkers usually have implicit default values (such as a dynamic
15// linker path or library paths) for each host OS.
16//
17// I don't think implicit default values are useful because they are
18// usually explicitly specified by the compiler driver. They can even
19// be harmful when you are doing cross-linking. Therefore, in LLD, we
20// simply trust the compiler driver to pass all required options and
21// don't try to make effort on our side.
22//
23//===----------------------------------------------------------------------===//
24
25#include "Driver.h"
26#include "Config.h"
27#include "ICF.h"
28#include "InputFiles.h"
29#include "InputSection.h"
30#include "LinkerScript.h"
31#include "MarkLive.h"
32#include "OutputSections.h"
33#include "ScriptParser.h"
34#include "SymbolTable.h"
35#include "Symbols.h"
36#include "SyntheticSections.h"
37#include "Target.h"
38#include "Writer.h"
39#include "lld/Common/Args.h"
40#include "lld/Common/Driver.h"
41#include "lld/Common/ErrorHandler.h"
42#include "lld/Common/Filesystem.h"
43#include "lld/Common/Memory.h"
44#include "lld/Common/Strings.h"
45#include "lld/Common/TargetOptionsCommandFlags.h"
46#include "lld/Common/Threads.h"
47#include "lld/Common/Version.h"
48#include "llvm/ADT/SetVector.h"
49#include "llvm/ADT/StringExtras.h"
50#include "llvm/ADT/StringSwitch.h"
51#include "llvm/LTO/LTO.h"
52#include "llvm/Support/CommandLine.h"
53#include "llvm/Support/Compression.h"
54#include "llvm/Support/GlobPattern.h"
55#include "llvm/Support/LEB128.h"
56#include "llvm/Support/Path.h"
57#include "llvm/Support/TarWriter.h"
58#include "llvm/Support/TargetSelect.h"
59#include "llvm/Support/raw_ostream.h"
60#include <cstdlib>
61#include <utility>
62
63using namespace llvm;
64using namespace llvm::ELF;
65using namespace llvm::object;
66using namespace llvm::sys;
67using namespace llvm::support;
68
69namespace lld {
70namespace elf {
71
72Configuration *config;
73LinkerDriver *driver;
74
75static void setConfigs(opt::InputArgList &args);
76static void readConfigs(opt::InputArgList &args);
77
78bool link(ArrayRef<const char *> args, bool canExitEarly, raw_ostream &stdoutOS,
79          raw_ostream &stderrOS) {
80  lld::stdoutOS = &stdoutOS;
81  lld::stderrOS = &stderrOS;
82
83  errorHandler().logName = args::getFilenameWithoutExe(args[0]);
84  errorHandler().errorLimitExceededMsg =
85      "too many errors emitted, stopping now (use "
86      "-error-limit=0 to see all errors)";
87  errorHandler().exitEarly = canExitEarly;
88  stderrOS.enable_colors(stderrOS.has_colors());
89
90  inputSections.clear();
91  outputSections.clear();
92  binaryFiles.clear();
93  bitcodeFiles.clear();
94  objectFiles.clear();
95  sharedFiles.clear();
96
97  config = make<Configuration>();
98  driver = make<LinkerDriver>();
99  script = make<LinkerScript>();
100  symtab = make<SymbolTable>();
101
102  tar = nullptr;
103  memset(&in, 0, sizeof(in));
104
105  partitions = {Partition()};
106
107  SharedFile::vernauxNum = 0;
108
109  config->progName = args[0];
110
111  driver->main(args);
112
113  // Exit immediately if we don't need to return to the caller.
114  // This saves time because the overhead of calling destructors
115  // for all globally-allocated objects is not negligible.
116  if (canExitEarly)
117    exitLld(errorCount() ? 1 : 0);
118
119  freeArena();
120  return !errorCount();
121}
122
123// Parses a linker -m option.
124static std::tuple<ELFKind, uint16_t, uint8_t> parseEmulation(StringRef emul) {
125  uint8_t osabi = 0;
126  StringRef s = emul;
127  if (s.endswith("_fbsd")) {
128    s = s.drop_back(5);
129    osabi = ELFOSABI_FREEBSD;
130  }
131
132  std::pair<ELFKind, uint16_t> ret =
133      StringSwitch<std::pair<ELFKind, uint16_t>>(s)
134          .Cases("aarch64elf", "aarch64linux", "aarch64_elf64_le_vec",
135                 {ELF64LEKind, EM_AARCH64})
136          .Cases("armelf", "armelf_linux_eabi", {ELF32LEKind, EM_ARM})
137          .Case("elf32_x86_64", {ELF32LEKind, EM_X86_64})
138          .Cases("elf32btsmip", "elf32btsmipn32", {ELF32BEKind, EM_MIPS})
139          .Cases("elf32ltsmip", "elf32ltsmipn32", {ELF32LEKind, EM_MIPS})
140          .Case("elf32lriscv", {ELF32LEKind, EM_RISCV})
141          .Cases("elf32ppc", "elf32ppclinux", {ELF32BEKind, EM_PPC})
142          .Case("elf64btsmip", {ELF64BEKind, EM_MIPS})
143          .Case("elf64ltsmip", {ELF64LEKind, EM_MIPS})
144          .Case("elf64lriscv", {ELF64LEKind, EM_RISCV})
145          .Case("elf64ppc", {ELF64BEKind, EM_PPC64})
146          .Case("elf64lppc", {ELF64LEKind, EM_PPC64})
147          .Cases("elf_amd64", "elf_x86_64", {ELF64LEKind, EM_X86_64})
148          .Case("elf_i386", {ELF32LEKind, EM_386})
149          .Case("elf_iamcu", {ELF32LEKind, EM_IAMCU})
150          .Default({ELFNoneKind, EM_NONE});
151
152  if (ret.first == ELFNoneKind)
153    error("unknown emulation: " + emul);
154  return std::make_tuple(ret.first, ret.second, osabi);
155}
156
157// Returns slices of MB by parsing MB as an archive file.
158// Each slice consists of a member file in the archive.
159std::vector<std::pair<MemoryBufferRef, uint64_t>> static getArchiveMembers(
160    MemoryBufferRef mb) {
161  std::unique_ptr<Archive> file =
162      CHECK(Archive::create(mb),
163            mb.getBufferIdentifier() + ": failed to parse archive");
164
165  std::vector<std::pair<MemoryBufferRef, uint64_t>> v;
166  Error err = Error::success();
167  bool addToTar = file->isThin() && tar;
168  for (const Archive::Child &c : file->children(err)) {
169    MemoryBufferRef mbref =
170        CHECK(c.getMemoryBufferRef(),
171              mb.getBufferIdentifier() +
172                  ": could not get the buffer for a child of the archive");
173    if (addToTar)
174      tar->append(relativeToRoot(check(c.getFullName())), mbref.getBuffer());
175    v.push_back(std::make_pair(mbref, c.getChildOffset()));
176  }
177  if (err)
178    fatal(mb.getBufferIdentifier() + ": Archive::children failed: " +
179          toString(std::move(err)));
180
181  // Take ownership of memory buffers created for members of thin archives.
182  for (std::unique_ptr<MemoryBuffer> &mb : file->takeThinBuffers())
183    make<std::unique_ptr<MemoryBuffer>>(std::move(mb));
184
185  return v;
186}
187
188// Opens a file and create a file object. Path has to be resolved already.
189void LinkerDriver::addFile(StringRef path, bool withLOption) {
190  using namespace sys::fs;
191
192  Optional<MemoryBufferRef> buffer = readFile(path);
193  if (!buffer.hasValue())
194    return;
195  MemoryBufferRef mbref = *buffer;
196
197  if (config->formatBinary) {
198    files.push_back(make<BinaryFile>(mbref));
199    return;
200  }
201
202  switch (identify_magic(mbref.getBuffer())) {
203  case file_magic::unknown:
204    readLinkerScript(mbref);
205    return;
206  case file_magic::archive: {
207    // Handle -whole-archive.
208    if (inWholeArchive) {
209      for (const auto &p : getArchiveMembers(mbref))
210        files.push_back(createObjectFile(p.first, path, p.second));
211      return;
212    }
213
214    std::unique_ptr<Archive> file =
215        CHECK(Archive::create(mbref), path + ": failed to parse archive");
216
217    // If an archive file has no symbol table, it is likely that a user
218    // is attempting LTO and using a default ar command that doesn't
219    // understand the LLVM bitcode file. It is a pretty common error, so
220    // we'll handle it as if it had a symbol table.
221    if (!file->isEmpty() && !file->hasSymbolTable()) {
222      // Check if all members are bitcode files. If not, ignore, which is the
223      // default action without the LTO hack described above.
224      for (const std::pair<MemoryBufferRef, uint64_t> &p :
225           getArchiveMembers(mbref))
226        if (identify_magic(p.first.getBuffer()) != file_magic::bitcode) {
227          error(path + ": archive has no index; run ranlib to add one");
228          return;
229        }
230
231      for (const std::pair<MemoryBufferRef, uint64_t> &p :
232           getArchiveMembers(mbref))
233        files.push_back(make<LazyObjFile>(p.first, path, p.second));
234      return;
235    }
236
237    // Handle the regular case.
238    files.push_back(make<ArchiveFile>(std::move(file)));
239    return;
240  }
241  case file_magic::elf_shared_object:
242    if (config->isStatic || config->relocatable) {
243      error("attempted static link of dynamic object " + path);
244      return;
245    }
246
247    // DSOs usually have DT_SONAME tags in their ELF headers, and the
248    // sonames are used to identify DSOs. But if they are missing,
249    // they are identified by filenames. We don't know whether the new
250    // file has a DT_SONAME or not because we haven't parsed it yet.
251    // Here, we set the default soname for the file because we might
252    // need it later.
253    //
254    // If a file was specified by -lfoo, the directory part is not
255    // significant, as a user did not specify it. This behavior is
256    // compatible with GNU.
257    files.push_back(
258        make<SharedFile>(mbref, withLOption ? path::filename(path) : path));
259    return;
260  case file_magic::bitcode:
261  case file_magic::elf_relocatable:
262    if (inLib)
263      files.push_back(make<LazyObjFile>(mbref, "", 0));
264    else
265      files.push_back(createObjectFile(mbref));
266    break;
267  default:
268    error(path + ": unknown file type");
269  }
270}
271
272// Add a given library by searching it from input search paths.
273void LinkerDriver::addLibrary(StringRef name) {
274  if (Optional<std::string> path = searchLibrary(name))
275    addFile(*path, /*withLOption=*/true);
276  else
277    error("unable to find library -l" + name);
278}
279
280// This function is called on startup. We need this for LTO since
281// LTO calls LLVM functions to compile bitcode files to native code.
282// Technically this can be delayed until we read bitcode files, but
283// we don't bother to do lazily because the initialization is fast.
284static void initLLVM() {
285  InitializeAllTargets();
286  InitializeAllTargetMCs();
287  InitializeAllAsmPrinters();
288  InitializeAllAsmParsers();
289}
290
291// Some command line options or some combinations of them are not allowed.
292// This function checks for such errors.
293static void checkOptions() {
294  // The MIPS ABI as of 2016 does not support the GNU-style symbol lookup
295  // table which is a relatively new feature.
296  if (config->emachine == EM_MIPS && config->gnuHash)
297    error("the .gnu.hash section is not compatible with the MIPS target");
298
299  if (config->fixCortexA53Errata843419 && config->emachine != EM_AARCH64)
300    error("--fix-cortex-a53-843419 is only supported on AArch64 targets");
301
302  if (config->fixCortexA8 && config->emachine != EM_ARM)
303    error("--fix-cortex-a8 is only supported on ARM targets");
304
305  if (config->tocOptimize && config->emachine != EM_PPC64)
306    error("--toc-optimize is only supported on the PowerPC64 target");
307
308  if (config->pie && config->shared)
309    error("-shared and -pie may not be used together");
310
311  if (!config->shared && !config->filterList.empty())
312    error("-F may not be used without -shared");
313
314  if (!config->shared && !config->auxiliaryList.empty())
315    error("-f may not be used without -shared");
316
317  if (!config->relocatable && !config->defineCommon)
318    error("-no-define-common not supported in non relocatable output");
319
320  if (config->strip == StripPolicy::All && config->emitRelocs)
321    error("--strip-all and --emit-relocs may not be used together");
322
323  if (config->zText && config->zIfuncNoplt)
324    error("-z text and -z ifunc-noplt may not be used together");
325
326  if (config->relocatable) {
327    if (config->shared)
328      error("-r and -shared may not be used together");
329    if (config->gcSections)
330      error("-r and --gc-sections may not be used together");
331    if (config->gdbIndex)
332      error("-r and --gdb-index may not be used together");
333    if (config->icf != ICFLevel::None)
334      error("-r and --icf may not be used together");
335    if (config->pie)
336      error("-r and -pie may not be used together");
337    if (config->exportDynamic)
338      error("-r and --export-dynamic may not be used together");
339  }
340
341  if (config->executeOnly) {
342    if (config->emachine != EM_AARCH64)
343      error("-execute-only is only supported on AArch64 targets");
344
345    if (config->singleRoRx && !script->hasSectionsCommand)
346      error("-execute-only and -no-rosegment cannot be used together");
347  }
348
349  if (config->zRetpolineplt && config->zForceIbt)
350    error("-z force-ibt may not be used with -z retpolineplt");
351
352  if (config->emachine != EM_AARCH64) {
353    if (config->pacPlt)
354      error("-z pac-plt only supported on AArch64");
355    if (config->forceBTI)
356      error("-z force-bti only supported on AArch64");
357  }
358}
359
360static const char *getReproduceOption(opt::InputArgList &args) {
361  if (auto *arg = args.getLastArg(OPT_reproduce))
362    return arg->getValue();
363  return getenv("LLD_REPRODUCE");
364}
365
366static bool hasZOption(opt::InputArgList &args, StringRef key) {
367  for (auto *arg : args.filtered(OPT_z))
368    if (key == arg->getValue())
369      return true;
370  return false;
371}
372
373static bool getZFlag(opt::InputArgList &args, StringRef k1, StringRef k2,
374                     bool Default) {
375  for (auto *arg : args.filtered_reverse(OPT_z)) {
376    if (k1 == arg->getValue())
377      return true;
378    if (k2 == arg->getValue())
379      return false;
380  }
381  return Default;
382}
383
384static SeparateSegmentKind getZSeparate(opt::InputArgList &args) {
385  for (auto *arg : args.filtered_reverse(OPT_z)) {
386    StringRef v = arg->getValue();
387    if (v == "noseparate-code")
388      return SeparateSegmentKind::None;
389    if (v == "separate-code")
390      return SeparateSegmentKind::Code;
391    if (v == "separate-loadable-segments")
392      return SeparateSegmentKind::Loadable;
393  }
394  return SeparateSegmentKind::None;
395}
396
397static GnuStackKind getZGnuStack(opt::InputArgList &args) {
398  for (auto *arg : args.filtered_reverse(OPT_z)) {
399    if (StringRef("execstack") == arg->getValue())
400      return GnuStackKind::Exec;
401    if (StringRef("noexecstack") == arg->getValue())
402      return GnuStackKind::NoExec;
403    if (StringRef("nognustack") == arg->getValue())
404      return GnuStackKind::None;
405  }
406
407  return GnuStackKind::NoExec;
408}
409
410static bool isKnownZFlag(StringRef s) {
411  return s == "combreloc" || s == "copyreloc" || s == "defs" ||
412         s == "execstack" || s == "force-bti" || s == "force-ibt" ||
413         s == "global" || s == "hazardplt" || s == "ifunc-noplt" ||
414         s == "initfirst" || s == "interpose" ||
415         s == "keep-text-section-prefix" || s == "lazy" || s == "muldefs" ||
416         s == "separate-code" || s == "separate-loadable-segments" ||
417         s == "nocombreloc" || s == "nocopyreloc" || s == "nodefaultlib" ||
418         s == "nodelete" || s == "nodlopen" || s == "noexecstack" ||
419         s == "nognustack" || s == "nokeep-text-section-prefix" ||
420         s == "norelro" || s == "noseparate-code" || s == "notext" ||
421         s == "now" || s == "origin" || s == "pac-plt" || s == "relro" ||
422         s == "retpolineplt" || s == "rodynamic" || s == "shstk" ||
423         s == "text" || s == "undefs" || s == "wxneeded" ||
424         s.startswith("common-page-size=") || s.startswith("max-page-size=") ||
425         s.startswith("stack-size=");
426}
427
428// Report an error for an unknown -z option.
429static void checkZOptions(opt::InputArgList &args) {
430  for (auto *arg : args.filtered(OPT_z))
431    if (!isKnownZFlag(arg->getValue()))
432      error("unknown -z value: " + StringRef(arg->getValue()));
433}
434
435void LinkerDriver::main(ArrayRef<const char *> argsArr) {
436  ELFOptTable parser;
437  opt::InputArgList args = parser.parse(argsArr.slice(1));
438
439  // Interpret this flag early because error() depends on them.
440  errorHandler().errorLimit = args::getInteger(args, OPT_error_limit, 20);
441  checkZOptions(args);
442
443  // Handle -help
444  if (args.hasArg(OPT_help)) {
445    printHelp();
446    return;
447  }
448
449  // Handle -v or -version.
450  //
451  // A note about "compatible with GNU linkers" message: this is a hack for
452  // scripts generated by GNU Libtool 2.4.6 (released in February 2014 and
453  // still the newest version in March 2017) or earlier to recognize LLD as
454  // a GNU compatible linker. As long as an output for the -v option
455  // contains "GNU" or "with BFD", they recognize us as GNU-compatible.
456  //
457  // This is somewhat ugly hack, but in reality, we had no choice other
458  // than doing this. Considering the very long release cycle of Libtool,
459  // it is not easy to improve it to recognize LLD as a GNU compatible
460  // linker in a timely manner. Even if we can make it, there are still a
461  // lot of "configure" scripts out there that are generated by old version
462  // of Libtool. We cannot convince every software developer to migrate to
463  // the latest version and re-generate scripts. So we have this hack.
464  if (args.hasArg(OPT_v) || args.hasArg(OPT_version))
465    message(getLLDVersion() + " (compatible with GNU linkers)");
466
467  if (const char *path = getReproduceOption(args)) {
468    // Note that --reproduce is a debug option so you can ignore it
469    // if you are trying to understand the whole picture of the code.
470    Expected<std::unique_ptr<TarWriter>> errOrWriter =
471        TarWriter::create(path, path::stem(path));
472    if (errOrWriter) {
473      tar = std::move(*errOrWriter);
474      tar->append("response.txt", createResponseFile(args));
475      tar->append("version.txt", getLLDVersion() + "\n");
476    } else {
477      error("--reproduce: " + toString(errOrWriter.takeError()));
478    }
479  }
480
481  readConfigs(args);
482
483  // The behavior of -v or --version is a bit strange, but this is
484  // needed for compatibility with GNU linkers.
485  if (args.hasArg(OPT_v) && !args.hasArg(OPT_INPUT))
486    return;
487  if (args.hasArg(OPT_version))
488    return;
489
490  initLLVM();
491  createFiles(args);
492  if (errorCount())
493    return;
494
495  inferMachineType();
496  setConfigs(args);
497  checkOptions();
498  if (errorCount())
499    return;
500
501  // The Target instance handles target-specific stuff, such as applying
502  // relocations or writing a PLT section. It also contains target-dependent
503  // values such as a default image base address.
504  target = getTarget();
505
506  switch (config->ekind) {
507  case ELF32LEKind:
508    link<ELF32LE>(args);
509    return;
510  case ELF32BEKind:
511    link<ELF32BE>(args);
512    return;
513  case ELF64LEKind:
514    link<ELF64LE>(args);
515    return;
516  case ELF64BEKind:
517    link<ELF64BE>(args);
518    return;
519  default:
520    llvm_unreachable("unknown Config->EKind");
521  }
522}
523
524static std::string getRpath(opt::InputArgList &args) {
525  std::vector<StringRef> v = args::getStrings(args, OPT_rpath);
526  return llvm::join(v.begin(), v.end(), ":");
527}
528
529// Determines what we should do if there are remaining unresolved
530// symbols after the name resolution.
531static UnresolvedPolicy getUnresolvedSymbolPolicy(opt::InputArgList &args) {
532  UnresolvedPolicy errorOrWarn = args.hasFlag(OPT_error_unresolved_symbols,
533                                              OPT_warn_unresolved_symbols, true)
534                                     ? UnresolvedPolicy::ReportError
535                                     : UnresolvedPolicy::Warn;
536
537  // Process the last of -unresolved-symbols, -no-undefined or -z defs.
538  for (auto *arg : llvm::reverse(args)) {
539    switch (arg->getOption().getID()) {
540    case OPT_unresolved_symbols: {
541      StringRef s = arg->getValue();
542      if (s == "ignore-all" || s == "ignore-in-object-files")
543        return UnresolvedPolicy::Ignore;
544      if (s == "ignore-in-shared-libs" || s == "report-all")
545        return errorOrWarn;
546      error("unknown --unresolved-symbols value: " + s);
547      continue;
548    }
549    case OPT_no_undefined:
550      return errorOrWarn;
551    case OPT_z:
552      if (StringRef(arg->getValue()) == "defs")
553        return errorOrWarn;
554      if (StringRef(arg->getValue()) == "undefs")
555        return UnresolvedPolicy::Ignore;
556      continue;
557    }
558  }
559
560  // -shared implies -unresolved-symbols=ignore-all because missing
561  // symbols are likely to be resolved at runtime using other DSOs.
562  if (config->shared)
563    return UnresolvedPolicy::Ignore;
564  return errorOrWarn;
565}
566
567static Target2Policy getTarget2(opt::InputArgList &args) {
568  StringRef s = args.getLastArgValue(OPT_target2, "got-rel");
569  if (s == "rel")
570    return Target2Policy::Rel;
571  if (s == "abs")
572    return Target2Policy::Abs;
573  if (s == "got-rel")
574    return Target2Policy::GotRel;
575  error("unknown --target2 option: " + s);
576  return Target2Policy::GotRel;
577}
578
579static bool isOutputFormatBinary(opt::InputArgList &args) {
580  StringRef s = args.getLastArgValue(OPT_oformat, "elf");
581  if (s == "binary")
582    return true;
583  if (!s.startswith("elf"))
584    error("unknown --oformat value: " + s);
585  return false;
586}
587
588static DiscardPolicy getDiscard(opt::InputArgList &args) {
589  if (args.hasArg(OPT_relocatable))
590    return DiscardPolicy::None;
591
592  auto *arg =
593      args.getLastArg(OPT_discard_all, OPT_discard_locals, OPT_discard_none);
594  if (!arg)
595    return DiscardPolicy::Default;
596  if (arg->getOption().getID() == OPT_discard_all)
597    return DiscardPolicy::All;
598  if (arg->getOption().getID() == OPT_discard_locals)
599    return DiscardPolicy::Locals;
600  return DiscardPolicy::None;
601}
602
603static StringRef getDynamicLinker(opt::InputArgList &args) {
604  auto *arg = args.getLastArg(OPT_dynamic_linker, OPT_no_dynamic_linker);
605  if (!arg)
606    return "";
607  if (arg->getOption().getID() == OPT_no_dynamic_linker) {
608    // --no-dynamic-linker suppresses undefined weak symbols in .dynsym
609    config->noDynamicLinker = true;
610    return "";
611  }
612  return arg->getValue();
613}
614
615static ICFLevel getICF(opt::InputArgList &args) {
616  auto *arg = args.getLastArg(OPT_icf_none, OPT_icf_safe, OPT_icf_all);
617  if (!arg || arg->getOption().getID() == OPT_icf_none)
618    return ICFLevel::None;
619  if (arg->getOption().getID() == OPT_icf_safe)
620    return ICFLevel::Safe;
621  return ICFLevel::All;
622}
623
624static StripPolicy getStrip(opt::InputArgList &args) {
625  if (args.hasArg(OPT_relocatable))
626    return StripPolicy::None;
627
628  auto *arg = args.getLastArg(OPT_strip_all, OPT_strip_debug);
629  if (!arg)
630    return StripPolicy::None;
631  if (arg->getOption().getID() == OPT_strip_all)
632    return StripPolicy::All;
633  return StripPolicy::Debug;
634}
635
636static uint64_t parseSectionAddress(StringRef s, opt::InputArgList &args,
637                                    const opt::Arg &arg) {
638  uint64_t va = 0;
639  if (s.startswith("0x"))
640    s = s.drop_front(2);
641  if (!to_integer(s, va, 16))
642    error("invalid argument: " + arg.getAsString(args));
643  return va;
644}
645
646static StringMap<uint64_t> getSectionStartMap(opt::InputArgList &args) {
647  StringMap<uint64_t> ret;
648  for (auto *arg : args.filtered(OPT_section_start)) {
649    StringRef name;
650    StringRef addr;
651    std::tie(name, addr) = StringRef(arg->getValue()).split('=');
652    ret[name] = parseSectionAddress(addr, args, *arg);
653  }
654
655  if (auto *arg = args.getLastArg(OPT_Ttext))
656    ret[".text"] = parseSectionAddress(arg->getValue(), args, *arg);
657  if (auto *arg = args.getLastArg(OPT_Tdata))
658    ret[".data"] = parseSectionAddress(arg->getValue(), args, *arg);
659  if (auto *arg = args.getLastArg(OPT_Tbss))
660    ret[".bss"] = parseSectionAddress(arg->getValue(), args, *arg);
661  return ret;
662}
663
664static SortSectionPolicy getSortSection(opt::InputArgList &args) {
665  StringRef s = args.getLastArgValue(OPT_sort_section);
666  if (s == "alignment")
667    return SortSectionPolicy::Alignment;
668  if (s == "name")
669    return SortSectionPolicy::Name;
670  if (!s.empty())
671    error("unknown --sort-section rule: " + s);
672  return SortSectionPolicy::Default;
673}
674
675static OrphanHandlingPolicy getOrphanHandling(opt::InputArgList &args) {
676  StringRef s = args.getLastArgValue(OPT_orphan_handling, "place");
677  if (s == "warn")
678    return OrphanHandlingPolicy::Warn;
679  if (s == "error")
680    return OrphanHandlingPolicy::Error;
681  if (s != "place")
682    error("unknown --orphan-handling mode: " + s);
683  return OrphanHandlingPolicy::Place;
684}
685
686// Parse --build-id or --build-id=<style>. We handle "tree" as a
687// synonym for "sha1" because all our hash functions including
688// -build-id=sha1 are actually tree hashes for performance reasons.
689static std::pair<BuildIdKind, std::vector<uint8_t>>
690getBuildId(opt::InputArgList &args) {
691  auto *arg = args.getLastArg(OPT_build_id, OPT_build_id_eq);
692  if (!arg)
693    return {BuildIdKind::None, {}};
694
695  if (arg->getOption().getID() == OPT_build_id)
696    return {BuildIdKind::Fast, {}};
697
698  StringRef s = arg->getValue();
699  if (s == "fast")
700    return {BuildIdKind::Fast, {}};
701  if (s == "md5")
702    return {BuildIdKind::Md5, {}};
703  if (s == "sha1" || s == "tree")
704    return {BuildIdKind::Sha1, {}};
705  if (s == "uuid")
706    return {BuildIdKind::Uuid, {}};
707  if (s.startswith("0x"))
708    return {BuildIdKind::Hexstring, parseHex(s.substr(2))};
709
710  if (s != "none")
711    error("unknown --build-id style: " + s);
712  return {BuildIdKind::None, {}};
713}
714
715static std::pair<bool, bool> getPackDynRelocs(opt::InputArgList &args) {
716  StringRef s = args.getLastArgValue(OPT_pack_dyn_relocs, "none");
717  if (s == "android")
718    return {true, false};
719  if (s == "relr")
720    return {false, true};
721  if (s == "android+relr")
722    return {true, true};
723
724  if (s != "none")
725    error("unknown -pack-dyn-relocs format: " + s);
726  return {false, false};
727}
728
729static void readCallGraph(MemoryBufferRef mb) {
730  // Build a map from symbol name to section
731  DenseMap<StringRef, Symbol *> map;
732  for (InputFile *file : objectFiles)
733    for (Symbol *sym : file->getSymbols())
734      map[sym->getName()] = sym;
735
736  auto findSection = [&](StringRef name) -> InputSectionBase * {
737    Symbol *sym = map.lookup(name);
738    if (!sym) {
739      if (config->warnSymbolOrdering)
740        warn(mb.getBufferIdentifier() + ": no such symbol: " + name);
741      return nullptr;
742    }
743    maybeWarnUnorderableSymbol(sym);
744
745    if (Defined *dr = dyn_cast_or_null<Defined>(sym))
746      return dyn_cast_or_null<InputSectionBase>(dr->section);
747    return nullptr;
748  };
749
750  for (StringRef line : args::getLines(mb)) {
751    SmallVector<StringRef, 3> fields;
752    line.split(fields, ' ');
753    uint64_t count;
754
755    if (fields.size() != 3 || !to_integer(fields[2], count)) {
756      error(mb.getBufferIdentifier() + ": parse error");
757      return;
758    }
759
760    if (InputSectionBase *from = findSection(fields[0]))
761      if (InputSectionBase *to = findSection(fields[1]))
762        config->callGraphProfile[std::make_pair(from, to)] += count;
763  }
764}
765
766template <class ELFT> static void readCallGraphsFromObjectFiles() {
767  for (auto file : objectFiles) {
768    auto *obj = cast<ObjFile<ELFT>>(file);
769
770    for (const Elf_CGProfile_Impl<ELFT> &cgpe : obj->cgProfile) {
771      auto *fromSym = dyn_cast<Defined>(&obj->getSymbol(cgpe.cgp_from));
772      auto *toSym = dyn_cast<Defined>(&obj->getSymbol(cgpe.cgp_to));
773      if (!fromSym || !toSym)
774        continue;
775
776      auto *from = dyn_cast_or_null<InputSectionBase>(fromSym->section);
777      auto *to = dyn_cast_or_null<InputSectionBase>(toSym->section);
778      if (from && to)
779        config->callGraphProfile[{from, to}] += cgpe.cgp_weight;
780    }
781  }
782}
783
784static bool getCompressDebugSections(opt::InputArgList &args) {
785  StringRef s = args.getLastArgValue(OPT_compress_debug_sections, "none");
786  if (s == "none")
787    return false;
788  if (s != "zlib")
789    error("unknown --compress-debug-sections value: " + s);
790  if (!zlib::isAvailable())
791    error("--compress-debug-sections: zlib is not available");
792  return true;
793}
794
795static StringRef getAliasSpelling(opt::Arg *arg) {
796  if (const opt::Arg *alias = arg->getAlias())
797    return alias->getSpelling();
798  return arg->getSpelling();
799}
800
801static std::pair<StringRef, StringRef> getOldNewOptions(opt::InputArgList &args,
802                                                        unsigned id) {
803  auto *arg = args.getLastArg(id);
804  if (!arg)
805    return {"", ""};
806
807  StringRef s = arg->getValue();
808  std::pair<StringRef, StringRef> ret = s.split(';');
809  if (ret.second.empty())
810    error(getAliasSpelling(arg) + " expects 'old;new' format, but got " + s);
811  return ret;
812}
813
814// Parse the symbol ordering file and warn for any duplicate entries.
815static std::vector<StringRef> getSymbolOrderingFile(MemoryBufferRef mb) {
816  SetVector<StringRef> names;
817  for (StringRef s : args::getLines(mb))
818    if (!names.insert(s) && config->warnSymbolOrdering)
819      warn(mb.getBufferIdentifier() + ": duplicate ordered symbol: " + s);
820
821  return names.takeVector();
822}
823
824static void parseClangOption(StringRef opt, const Twine &msg) {
825  std::string err;
826  raw_string_ostream os(err);
827
828  const char *argv[] = {config->progName.data(), opt.data()};
829  if (cl::ParseCommandLineOptions(2, argv, "", &os))
830    return;
831  os.flush();
832  error(msg + ": " + StringRef(err).trim());
833}
834
835// Initializes Config members by the command line options.
836static void readConfigs(opt::InputArgList &args) {
837  errorHandler().verbose = args.hasArg(OPT_verbose);
838  errorHandler().fatalWarnings =
839      args.hasFlag(OPT_fatal_warnings, OPT_no_fatal_warnings, false);
840  errorHandler().vsDiagnostics =
841      args.hasArg(OPT_visual_studio_diagnostics_format, false);
842  threadsEnabled = args.hasFlag(OPT_threads, OPT_no_threads, true);
843
844  config->allowMultipleDefinition =
845      args.hasFlag(OPT_allow_multiple_definition,
846                   OPT_no_allow_multiple_definition, false) ||
847      hasZOption(args, "muldefs");
848  config->allowShlibUndefined =
849      args.hasFlag(OPT_allow_shlib_undefined, OPT_no_allow_shlib_undefined,
850                   args.hasArg(OPT_shared));
851  config->auxiliaryList = args::getStrings(args, OPT_auxiliary);
852  config->bsymbolic = args.hasArg(OPT_Bsymbolic);
853  config->bsymbolicFunctions = args.hasArg(OPT_Bsymbolic_functions);
854  config->checkSections =
855      args.hasFlag(OPT_check_sections, OPT_no_check_sections, true);
856  config->chroot = args.getLastArgValue(OPT_chroot);
857  config->compressDebugSections = getCompressDebugSections(args);
858  config->cref = args.hasFlag(OPT_cref, OPT_no_cref, false);
859  config->defineCommon = args.hasFlag(OPT_define_common, OPT_no_define_common,
860                                      !args.hasArg(OPT_relocatable));
861  config->demangle = args.hasFlag(OPT_demangle, OPT_no_demangle, true);
862  config->dependentLibraries = args.hasFlag(OPT_dependent_libraries, OPT_no_dependent_libraries, true);
863  config->disableVerify = args.hasArg(OPT_disable_verify);
864  config->discard = getDiscard(args);
865  config->dwoDir = args.getLastArgValue(OPT_plugin_opt_dwo_dir_eq);
866  config->dynamicLinker = getDynamicLinker(args);
867  config->ehFrameHdr =
868      args.hasFlag(OPT_eh_frame_hdr, OPT_no_eh_frame_hdr, false);
869  config->emitLLVM = args.hasArg(OPT_plugin_opt_emit_llvm, false);
870  config->emitRelocs = args.hasArg(OPT_emit_relocs);
871  config->callGraphProfileSort = args.hasFlag(
872      OPT_call_graph_profile_sort, OPT_no_call_graph_profile_sort, true);
873  config->enableNewDtags =
874      args.hasFlag(OPT_enable_new_dtags, OPT_disable_new_dtags, true);
875  config->entry = args.getLastArgValue(OPT_entry);
876  config->executeOnly =
877      args.hasFlag(OPT_execute_only, OPT_no_execute_only, false);
878  config->exportDynamic =
879      args.hasFlag(OPT_export_dynamic, OPT_no_export_dynamic, false);
880  config->filterList = args::getStrings(args, OPT_filter);
881  config->fini = args.getLastArgValue(OPT_fini, "_fini");
882  config->fixCortexA53Errata843419 = args.hasArg(OPT_fix_cortex_a53_843419);
883  config->fixCortexA8 = args.hasArg(OPT_fix_cortex_a8);
884  config->forceBTI = hasZOption(args, "force-bti");
885  config->gcSections = args.hasFlag(OPT_gc_sections, OPT_no_gc_sections, false);
886  config->gnuUnique = args.hasFlag(OPT_gnu_unique, OPT_no_gnu_unique, true);
887  config->gdbIndex = args.hasFlag(OPT_gdb_index, OPT_no_gdb_index, false);
888  config->icf = getICF(args);
889  config->ignoreDataAddressEquality =
890      args.hasArg(OPT_ignore_data_address_equality);
891  config->ignoreFunctionAddressEquality =
892      args.hasArg(OPT_ignore_function_address_equality);
893  config->init = args.getLastArgValue(OPT_init, "_init");
894  config->ltoAAPipeline = args.getLastArgValue(OPT_lto_aa_pipeline);
895  config->ltoCSProfileGenerate = args.hasArg(OPT_lto_cs_profile_generate);
896  config->ltoCSProfileFile = args.getLastArgValue(OPT_lto_cs_profile_file);
897  config->ltoDebugPassManager = args.hasArg(OPT_lto_debug_pass_manager);
898  config->ltoNewPassManager = args.hasArg(OPT_lto_new_pass_manager);
899  config->ltoNewPmPasses = args.getLastArgValue(OPT_lto_newpm_passes);
900  config->ltoo = args::getInteger(args, OPT_lto_O, 2);
901  config->ltoObjPath = args.getLastArgValue(OPT_lto_obj_path_eq);
902  config->ltoPartitions = args::getInteger(args, OPT_lto_partitions, 1);
903  config->ltoSampleProfile = args.getLastArgValue(OPT_lto_sample_profile);
904  config->mapFile = args.getLastArgValue(OPT_Map);
905  config->mipsGotSize = args::getInteger(args, OPT_mips_got_size, 0xfff0);
906  config->mergeArmExidx =
907      args.hasFlag(OPT_merge_exidx_entries, OPT_no_merge_exidx_entries, true);
908  config->mmapOutputFile =
909      args.hasFlag(OPT_mmap_output_file, OPT_no_mmap_output_file, true);
910  config->nmagic = args.hasFlag(OPT_nmagic, OPT_no_nmagic, false);
911  config->noinhibitExec = args.hasArg(OPT_noinhibit_exec);
912  config->nostdlib = args.hasArg(OPT_nostdlib);
913  config->oFormatBinary = isOutputFormatBinary(args);
914  config->omagic = args.hasFlag(OPT_omagic, OPT_no_omagic, false);
915  config->optRemarksFilename = args.getLastArgValue(OPT_opt_remarks_filename);
916  config->optRemarksPasses = args.getLastArgValue(OPT_opt_remarks_passes);
917  config->optRemarksWithHotness = args.hasArg(OPT_opt_remarks_with_hotness);
918  config->optRemarksFormat = args.getLastArgValue(OPT_opt_remarks_format);
919  config->optimize = args::getInteger(args, OPT_O, 1);
920  config->orphanHandling = getOrphanHandling(args);
921  config->outputFile = args.getLastArgValue(OPT_o);
922  config->pacPlt = hasZOption(args, "pac-plt");
923  config->pie = args.hasFlag(OPT_pie, OPT_no_pie, false);
924  config->printIcfSections =
925      args.hasFlag(OPT_print_icf_sections, OPT_no_print_icf_sections, false);
926  config->printGcSections =
927      args.hasFlag(OPT_print_gc_sections, OPT_no_print_gc_sections, false);
928  config->printSymbolOrder =
929      args.getLastArgValue(OPT_print_symbol_order);
930  config->rpath = getRpath(args);
931  config->relocatable = args.hasArg(OPT_relocatable);
932  config->saveTemps = args.hasArg(OPT_save_temps);
933  config->searchPaths = args::getStrings(args, OPT_library_path);
934  config->sectionStartMap = getSectionStartMap(args);
935  config->shared = args.hasArg(OPT_shared);
936  config->singleRoRx = args.hasArg(OPT_no_rosegment);
937  config->soName = args.getLastArgValue(OPT_soname);
938  config->sortSection = getSortSection(args);
939  config->splitStackAdjustSize = args::getInteger(args, OPT_split_stack_adjust_size, 16384);
940  config->strip = getStrip(args);
941  config->sysroot = args.getLastArgValue(OPT_sysroot);
942  config->target1Rel = args.hasFlag(OPT_target1_rel, OPT_target1_abs, false);
943  config->target2 = getTarget2(args);
944  config->thinLTOCacheDir = args.getLastArgValue(OPT_thinlto_cache_dir);
945  config->thinLTOCachePolicy = CHECK(
946      parseCachePruningPolicy(args.getLastArgValue(OPT_thinlto_cache_policy)),
947      "--thinlto-cache-policy: invalid cache policy");
948  config->thinLTOEmitImportsFiles = args.hasArg(OPT_thinlto_emit_imports_files);
949  config->thinLTOIndexOnly = args.hasArg(OPT_thinlto_index_only) ||
950                             args.hasArg(OPT_thinlto_index_only_eq);
951  config->thinLTOIndexOnlyArg = args.getLastArgValue(OPT_thinlto_index_only_eq);
952  config->thinLTOJobs = args::getInteger(args, OPT_thinlto_jobs, -1u);
953  config->thinLTOObjectSuffixReplace =
954      getOldNewOptions(args, OPT_thinlto_object_suffix_replace_eq);
955  config->thinLTOPrefixReplace =
956      getOldNewOptions(args, OPT_thinlto_prefix_replace_eq);
957  config->trace = args.hasArg(OPT_trace);
958  config->undefined = args::getStrings(args, OPT_undefined);
959  config->undefinedVersion =
960      args.hasFlag(OPT_undefined_version, OPT_no_undefined_version, true);
961  config->useAndroidRelrTags = args.hasFlag(
962      OPT_use_android_relr_tags, OPT_no_use_android_relr_tags, false);
963  config->unresolvedSymbols = getUnresolvedSymbolPolicy(args);
964  config->warnBackrefs =
965      args.hasFlag(OPT_warn_backrefs, OPT_no_warn_backrefs, false);
966  config->warnCommon = args.hasFlag(OPT_warn_common, OPT_no_warn_common, false);
967  config->warnIfuncTextrel =
968      args.hasFlag(OPT_warn_ifunc_textrel, OPT_no_warn_ifunc_textrel, false);
969  config->warnSymbolOrdering =
970      args.hasFlag(OPT_warn_symbol_ordering, OPT_no_warn_symbol_ordering, true);
971  config->zCombreloc = getZFlag(args, "combreloc", "nocombreloc", true);
972  config->zCopyreloc = getZFlag(args, "copyreloc", "nocopyreloc", true);
973  config->zForceIbt = hasZOption(args, "force-ibt");
974  config->zGlobal = hasZOption(args, "global");
975  config->zGnustack = getZGnuStack(args);
976  config->zHazardplt = hasZOption(args, "hazardplt");
977  config->zIfuncNoplt = hasZOption(args, "ifunc-noplt");
978  config->zInitfirst = hasZOption(args, "initfirst");
979  config->zInterpose = hasZOption(args, "interpose");
980  config->zKeepTextSectionPrefix = getZFlag(
981      args, "keep-text-section-prefix", "nokeep-text-section-prefix", false);
982  config->zNodefaultlib = hasZOption(args, "nodefaultlib");
983  config->zNodelete = hasZOption(args, "nodelete");
984  config->zNodlopen = hasZOption(args, "nodlopen");
985  config->zNow = getZFlag(args, "now", "lazy", false);
986  config->zOrigin = hasZOption(args, "origin");
987  config->zRelro = getZFlag(args, "relro", "norelro", true);
988  config->zRetpolineplt = hasZOption(args, "retpolineplt");
989  config->zRodynamic = hasZOption(args, "rodynamic");
990  config->zSeparate = getZSeparate(args);
991  config->zShstk = hasZOption(args, "shstk");
992  config->zStackSize = args::getZOptionValue(args, OPT_z, "stack-size", 0);
993  config->zText = getZFlag(args, "text", "notext", true);
994  config->zWxneeded = hasZOption(args, "wxneeded");
995
996  // Parse LTO options.
997  if (auto *arg = args.getLastArg(OPT_plugin_opt_mcpu_eq))
998    parseClangOption(saver.save("-mcpu=" + StringRef(arg->getValue())),
999                     arg->getSpelling());
1000
1001  for (auto *arg : args.filtered(OPT_plugin_opt))
1002    parseClangOption(arg->getValue(), arg->getSpelling());
1003
1004  // Parse -mllvm options.
1005  for (auto *arg : args.filtered(OPT_mllvm))
1006    parseClangOption(arg->getValue(), arg->getSpelling());
1007
1008  if (config->ltoo > 3)
1009    error("invalid optimization level for LTO: " + Twine(config->ltoo));
1010  if (config->ltoPartitions == 0)
1011    error("--lto-partitions: number of threads must be > 0");
1012  if (config->thinLTOJobs == 0)
1013    error("--thinlto-jobs: number of threads must be > 0");
1014
1015  if (config->splitStackAdjustSize < 0)
1016    error("--split-stack-adjust-size: size must be >= 0");
1017
1018  // The text segment is traditionally the first segment, whose address equals
1019  // the base address. However, lld places the R PT_LOAD first. -Ttext-segment
1020  // is an old-fashioned option that does not play well with lld's layout.
1021  // Suggest --image-base as a likely alternative.
1022  if (args.hasArg(OPT_Ttext_segment))
1023    error("-Ttext-segment is not supported. Use --image-base if you "
1024          "intend to set the base address");
1025
1026  // Parse ELF{32,64}{LE,BE} and CPU type.
1027  if (auto *arg = args.getLastArg(OPT_m)) {
1028    StringRef s = arg->getValue();
1029    std::tie(config->ekind, config->emachine, config->osabi) =
1030        parseEmulation(s);
1031    config->mipsN32Abi =
1032        (s.startswith("elf32btsmipn32") || s.startswith("elf32ltsmipn32"));
1033    config->emulation = s;
1034  }
1035
1036  // Parse -hash-style={sysv,gnu,both}.
1037  if (auto *arg = args.getLastArg(OPT_hash_style)) {
1038    StringRef s = arg->getValue();
1039    if (s == "sysv")
1040      config->sysvHash = true;
1041    else if (s == "gnu")
1042      config->gnuHash = true;
1043    else if (s == "both")
1044      config->sysvHash = config->gnuHash = true;
1045    else
1046      error("unknown -hash-style: " + s);
1047  }
1048
1049  if (args.hasArg(OPT_print_map))
1050    config->mapFile = "-";
1051
1052  // Page alignment can be disabled by the -n (--nmagic) and -N (--omagic).
1053  // As PT_GNU_RELRO relies on Paging, do not create it when we have disabled
1054  // it.
1055  if (config->nmagic || config->omagic)
1056    config->zRelro = false;
1057
1058  std::tie(config->buildId, config->buildIdVector) = getBuildId(args);
1059
1060  std::tie(config->androidPackDynRelocs, config->relrPackDynRelocs) =
1061      getPackDynRelocs(args);
1062
1063  if (auto *arg = args.getLastArg(OPT_symbol_ordering_file)){
1064    if (args.hasArg(OPT_call_graph_ordering_file))
1065      error("--symbol-ordering-file and --call-graph-order-file "
1066            "may not be used together");
1067    if (Optional<MemoryBufferRef> buffer = readFile(arg->getValue())){
1068      config->symbolOrderingFile = getSymbolOrderingFile(*buffer);
1069      // Also need to disable CallGraphProfileSort to prevent
1070      // LLD order symbols with CGProfile
1071      config->callGraphProfileSort = false;
1072    }
1073  }
1074
1075  assert(config->versionDefinitions.empty());
1076  config->versionDefinitions.push_back({"local", (uint16_t)VER_NDX_LOCAL, {}});
1077  config->versionDefinitions.push_back(
1078      {"global", (uint16_t)VER_NDX_GLOBAL, {}});
1079
1080  // If --retain-symbol-file is used, we'll keep only the symbols listed in
1081  // the file and discard all others.
1082  if (auto *arg = args.getLastArg(OPT_retain_symbols_file)) {
1083    config->versionDefinitions[VER_NDX_LOCAL].patterns.push_back(
1084        {"*", /*isExternCpp=*/false, /*hasWildcard=*/true});
1085    if (Optional<MemoryBufferRef> buffer = readFile(arg->getValue()))
1086      for (StringRef s : args::getLines(*buffer))
1087        config->versionDefinitions[VER_NDX_GLOBAL].patterns.push_back(
1088            {s, /*isExternCpp=*/false, /*hasWildcard=*/false});
1089  }
1090
1091  // Parses -dynamic-list and -export-dynamic-symbol. They make some
1092  // symbols private. Note that -export-dynamic takes precedence over them
1093  // as it says all symbols should be exported.
1094  if (!config->exportDynamic) {
1095    for (auto *arg : args.filtered(OPT_dynamic_list))
1096      if (Optional<MemoryBufferRef> buffer = readFile(arg->getValue()))
1097        readDynamicList(*buffer);
1098
1099    for (auto *arg : args.filtered(OPT_export_dynamic_symbol))
1100      config->dynamicList.push_back(
1101          {arg->getValue(), /*isExternCpp=*/false, /*hasWildcard=*/false});
1102  }
1103
1104  // If --export-dynamic-symbol=foo is given and symbol foo is defined in
1105  // an object file in an archive file, that object file should be pulled
1106  // out and linked. (It doesn't have to behave like that from technical
1107  // point of view, but this is needed for compatibility with GNU.)
1108  for (auto *arg : args.filtered(OPT_export_dynamic_symbol))
1109    config->undefined.push_back(arg->getValue());
1110
1111  for (auto *arg : args.filtered(OPT_version_script))
1112    if (Optional<std::string> path = searchScript(arg->getValue())) {
1113      if (Optional<MemoryBufferRef> buffer = readFile(*path))
1114        readVersionScript(*buffer);
1115    } else {
1116      error(Twine("cannot find version script ") + arg->getValue());
1117    }
1118}
1119
1120// Some Config members do not directly correspond to any particular
1121// command line options, but computed based on other Config values.
1122// This function initialize such members. See Config.h for the details
1123// of these values.
1124static void setConfigs(opt::InputArgList &args) {
1125  ELFKind k = config->ekind;
1126  uint16_t m = config->emachine;
1127
1128  config->copyRelocs = (config->relocatable || config->emitRelocs);
1129  config->is64 = (k == ELF64LEKind || k == ELF64BEKind);
1130  config->isLE = (k == ELF32LEKind || k == ELF64LEKind);
1131  config->endianness = config->isLE ? endianness::little : endianness::big;
1132  config->isMips64EL = (k == ELF64LEKind && m == EM_MIPS);
1133  config->isPic = config->pie || config->shared;
1134  config->picThunk = args.hasArg(OPT_pic_veneer, config->isPic);
1135  config->wordsize = config->is64 ? 8 : 4;
1136
1137  // ELF defines two different ways to store relocation addends as shown below:
1138  //
1139  //  Rel:  Addends are stored to the location where relocations are applied.
1140  //  Rela: Addends are stored as part of relocation entry.
1141  //
1142  // In other words, Rela makes it easy to read addends at the price of extra
1143  // 4 or 8 byte for each relocation entry. We don't know why ELF defined two
1144  // different mechanisms in the first place, but this is how the spec is
1145  // defined.
1146  //
1147  // You cannot choose which one, Rel or Rela, you want to use. Instead each
1148  // ABI defines which one you need to use. The following expression expresses
1149  // that.
1150  config->isRela = m == EM_AARCH64 || m == EM_AMDGPU || m == EM_HEXAGON ||
1151                   m == EM_PPC || m == EM_PPC64 || m == EM_RISCV ||
1152                   m == EM_X86_64;
1153
1154  // If the output uses REL relocations we must store the dynamic relocation
1155  // addends to the output sections. We also store addends for RELA relocations
1156  // if --apply-dynamic-relocs is used.
1157  // We default to not writing the addends when using RELA relocations since
1158  // any standard conforming tool can find it in r_addend.
1159  config->writeAddends = args.hasFlag(OPT_apply_dynamic_relocs,
1160                                      OPT_no_apply_dynamic_relocs, false) ||
1161                         !config->isRela;
1162
1163  config->tocOptimize =
1164      args.hasFlag(OPT_toc_optimize, OPT_no_toc_optimize, m == EM_PPC64);
1165}
1166
1167// Returns a value of "-format" option.
1168static bool isFormatBinary(StringRef s) {
1169  if (s == "binary")
1170    return true;
1171  if (s == "elf" || s == "default")
1172    return false;
1173  error("unknown -format value: " + s +
1174        " (supported formats: elf, default, binary)");
1175  return false;
1176}
1177
1178void LinkerDriver::createFiles(opt::InputArgList &args) {
1179  // For --{push,pop}-state.
1180  std::vector<std::tuple<bool, bool, bool>> stack;
1181
1182  // Iterate over argv to process input files and positional arguments.
1183  for (auto *arg : args) {
1184    switch (arg->getOption().getID()) {
1185    case OPT_library:
1186      addLibrary(arg->getValue());
1187      break;
1188    case OPT_INPUT:
1189      addFile(arg->getValue(), /*withLOption=*/false);
1190      break;
1191    case OPT_defsym: {
1192      StringRef from;
1193      StringRef to;
1194      std::tie(from, to) = StringRef(arg->getValue()).split('=');
1195      if (from.empty() || to.empty())
1196        error("-defsym: syntax error: " + StringRef(arg->getValue()));
1197      else
1198        readDefsym(from, MemoryBufferRef(to, "-defsym"));
1199      break;
1200    }
1201    case OPT_script:
1202      if (Optional<std::string> path = searchScript(arg->getValue())) {
1203        if (Optional<MemoryBufferRef> mb = readFile(*path))
1204          readLinkerScript(*mb);
1205        break;
1206      }
1207      error(Twine("cannot find linker script ") + arg->getValue());
1208      break;
1209    case OPT_as_needed:
1210      config->asNeeded = true;
1211      break;
1212    case OPT_format:
1213      config->formatBinary = isFormatBinary(arg->getValue());
1214      break;
1215    case OPT_no_as_needed:
1216      config->asNeeded = false;
1217      break;
1218    case OPT_Bstatic:
1219    case OPT_omagic:
1220    case OPT_nmagic:
1221      config->isStatic = true;
1222      break;
1223    case OPT_Bdynamic:
1224      config->isStatic = false;
1225      break;
1226    case OPT_whole_archive:
1227      inWholeArchive = true;
1228      break;
1229    case OPT_no_whole_archive:
1230      inWholeArchive = false;
1231      break;
1232    case OPT_just_symbols:
1233      if (Optional<MemoryBufferRef> mb = readFile(arg->getValue())) {
1234        files.push_back(createObjectFile(*mb));
1235        files.back()->justSymbols = true;
1236      }
1237      break;
1238    case OPT_start_group:
1239      if (InputFile::isInGroup)
1240        error("nested --start-group");
1241      InputFile::isInGroup = true;
1242      break;
1243    case OPT_end_group:
1244      if (!InputFile::isInGroup)
1245        error("stray --end-group");
1246      InputFile::isInGroup = false;
1247      ++InputFile::nextGroupId;
1248      break;
1249    case OPT_start_lib:
1250      if (inLib)
1251        error("nested --start-lib");
1252      if (InputFile::isInGroup)
1253        error("may not nest --start-lib in --start-group");
1254      inLib = true;
1255      InputFile::isInGroup = true;
1256      break;
1257    case OPT_end_lib:
1258      if (!inLib)
1259        error("stray --end-lib");
1260      inLib = false;
1261      InputFile::isInGroup = false;
1262      ++InputFile::nextGroupId;
1263      break;
1264    case OPT_push_state:
1265      stack.emplace_back(config->asNeeded, config->isStatic, inWholeArchive);
1266      break;
1267    case OPT_pop_state:
1268      if (stack.empty()) {
1269        error("unbalanced --push-state/--pop-state");
1270        break;
1271      }
1272      std::tie(config->asNeeded, config->isStatic, inWholeArchive) = stack.back();
1273      stack.pop_back();
1274      break;
1275    }
1276  }
1277
1278  if (files.empty() && errorCount() == 0)
1279    error("no input files");
1280}
1281
1282// If -m <machine_type> was not given, infer it from object files.
1283void LinkerDriver::inferMachineType() {
1284  if (config->ekind != ELFNoneKind)
1285    return;
1286
1287  for (InputFile *f : files) {
1288    if (f->ekind == ELFNoneKind)
1289      continue;
1290    config->ekind = f->ekind;
1291    config->emachine = f->emachine;
1292    config->osabi = f->osabi;
1293    config->mipsN32Abi = config->emachine == EM_MIPS && isMipsN32Abi(f);
1294    return;
1295  }
1296  error("target emulation unknown: -m or at least one .o file required");
1297}
1298
1299// Parse -z max-page-size=<value>. The default value is defined by
1300// each target.
1301static uint64_t getMaxPageSize(opt::InputArgList &args) {
1302  uint64_t val = args::getZOptionValue(args, OPT_z, "max-page-size",
1303                                       target->defaultMaxPageSize);
1304  if (!isPowerOf2_64(val))
1305    error("max-page-size: value isn't a power of 2");
1306  if (config->nmagic || config->omagic) {
1307    if (val != target->defaultMaxPageSize)
1308      warn("-z max-page-size set, but paging disabled by omagic or nmagic");
1309    return 1;
1310  }
1311  return val;
1312}
1313
1314// Parse -z common-page-size=<value>. The default value is defined by
1315// each target.
1316static uint64_t getCommonPageSize(opt::InputArgList &args) {
1317  uint64_t val = args::getZOptionValue(args, OPT_z, "common-page-size",
1318                                       target->defaultCommonPageSize);
1319  if (!isPowerOf2_64(val))
1320    error("common-page-size: value isn't a power of 2");
1321  if (config->nmagic || config->omagic) {
1322    if (val != target->defaultCommonPageSize)
1323      warn("-z common-page-size set, but paging disabled by omagic or nmagic");
1324    return 1;
1325  }
1326  // commonPageSize can't be larger than maxPageSize.
1327  if (val > config->maxPageSize)
1328    val = config->maxPageSize;
1329  return val;
1330}
1331
1332// Parses -image-base option.
1333static Optional<uint64_t> getImageBase(opt::InputArgList &args) {
1334  // Because we are using "Config->maxPageSize" here, this function has to be
1335  // called after the variable is initialized.
1336  auto *arg = args.getLastArg(OPT_image_base);
1337  if (!arg)
1338    return None;
1339
1340  StringRef s = arg->getValue();
1341  uint64_t v;
1342  if (!to_integer(s, v)) {
1343    error("-image-base: number expected, but got " + s);
1344    return 0;
1345  }
1346  if ((v % config->maxPageSize) != 0)
1347    warn("-image-base: address isn't multiple of page size: " + s);
1348  return v;
1349}
1350
1351// Parses `--exclude-libs=lib,lib,...`.
1352// The library names may be delimited by commas or colons.
1353static DenseSet<StringRef> getExcludeLibs(opt::InputArgList &args) {
1354  DenseSet<StringRef> ret;
1355  for (auto *arg : args.filtered(OPT_exclude_libs)) {
1356    StringRef s = arg->getValue();
1357    for (;;) {
1358      size_t pos = s.find_first_of(",:");
1359      if (pos == StringRef::npos)
1360        break;
1361      ret.insert(s.substr(0, pos));
1362      s = s.substr(pos + 1);
1363    }
1364    ret.insert(s);
1365  }
1366  return ret;
1367}
1368
1369// Handles the -exclude-libs option. If a static library file is specified
1370// by the -exclude-libs option, all public symbols from the archive become
1371// private unless otherwise specified by version scripts or something.
1372// A special library name "ALL" means all archive files.
1373//
1374// This is not a popular option, but some programs such as bionic libc use it.
1375static void excludeLibs(opt::InputArgList &args) {
1376  DenseSet<StringRef> libs = getExcludeLibs(args);
1377  bool all = libs.count("ALL");
1378
1379  auto visit = [&](InputFile *file) {
1380    if (!file->archiveName.empty())
1381      if (all || libs.count(path::filename(file->archiveName)))
1382        for (Symbol *sym : file->getSymbols())
1383          if (!sym->isUndefined() && !sym->isLocal() && sym->file == file)
1384            sym->versionId = VER_NDX_LOCAL;
1385  };
1386
1387  for (InputFile *file : objectFiles)
1388    visit(file);
1389
1390  for (BitcodeFile *file : bitcodeFiles)
1391    visit(file);
1392}
1393
1394// Force Sym to be entered in the output. Used for -u or equivalent.
1395static void handleUndefined(Symbol *sym) {
1396  // Since a symbol may not be used inside the program, LTO may
1397  // eliminate it. Mark the symbol as "used" to prevent it.
1398  sym->isUsedInRegularObj = true;
1399
1400  if (sym->isLazy())
1401    sym->fetch();
1402}
1403
1404// As an extension to GNU linkers, lld supports a variant of `-u`
1405// which accepts wildcard patterns. All symbols that match a given
1406// pattern are handled as if they were given by `-u`.
1407static void handleUndefinedGlob(StringRef arg) {
1408  Expected<GlobPattern> pat = GlobPattern::create(arg);
1409  if (!pat) {
1410    error("--undefined-glob: " + toString(pat.takeError()));
1411    return;
1412  }
1413
1414  std::vector<Symbol *> syms;
1415  for (Symbol *sym : symtab->symbols()) {
1416    // Calling Sym->fetch() from here is not safe because it may
1417    // add new symbols to the symbol table, invalidating the
1418    // current iterator. So we just keep a note.
1419    if (pat->match(sym->getName()))
1420      syms.push_back(sym);
1421  }
1422
1423  for (Symbol *sym : syms)
1424    handleUndefined(sym);
1425}
1426
1427static void handleLibcall(StringRef name) {
1428  Symbol *sym = symtab->find(name);
1429  if (!sym || !sym->isLazy())
1430    return;
1431
1432  MemoryBufferRef mb;
1433  if (auto *lo = dyn_cast<LazyObject>(sym))
1434    mb = lo->file->mb;
1435  else
1436    mb = cast<LazyArchive>(sym)->getMemberBuffer();
1437
1438  if (isBitcode(mb))
1439    sym->fetch();
1440}
1441
1442// Replaces common symbols with defined symbols reside in .bss sections.
1443// This function is called after all symbol names are resolved. As a
1444// result, the passes after the symbol resolution won't see any
1445// symbols of type CommonSymbol.
1446static void replaceCommonSymbols() {
1447  for (Symbol *sym : symtab->symbols()) {
1448    auto *s = dyn_cast<CommonSymbol>(sym);
1449    if (!s)
1450      continue;
1451
1452    auto *bss = make<BssSection>("COMMON", s->size, s->alignment);
1453    bss->file = s->file;
1454    bss->markDead();
1455    inputSections.push_back(bss);
1456    s->replace(Defined{s->file, s->getName(), s->binding, s->stOther, s->type,
1457                       /*value=*/0, s->size, bss});
1458  }
1459}
1460
1461// If all references to a DSO happen to be weak, the DSO is not added
1462// to DT_NEEDED. If that happens, we need to eliminate shared symbols
1463// created from the DSO. Otherwise, they become dangling references
1464// that point to a non-existent DSO.
1465static void demoteSharedSymbols() {
1466  for (Symbol *sym : symtab->symbols()) {
1467    auto *s = dyn_cast<SharedSymbol>(sym);
1468    if (!s || s->getFile().isNeeded)
1469      continue;
1470
1471    bool used = s->used;
1472    s->replace(Undefined{nullptr, s->getName(), STB_WEAK, s->stOther, s->type});
1473    s->used = used;
1474  }
1475}
1476
1477// The section referred to by `s` is considered address-significant. Set the
1478// keepUnique flag on the section if appropriate.
1479static void markAddrsig(Symbol *s) {
1480  if (auto *d = dyn_cast_or_null<Defined>(s))
1481    if (d->section)
1482      // We don't need to keep text sections unique under --icf=all even if they
1483      // are address-significant.
1484      if (config->icf == ICFLevel::Safe || !(d->section->flags & SHF_EXECINSTR))
1485        d->section->keepUnique = true;
1486}
1487
1488// Record sections that define symbols mentioned in --keep-unique <symbol>
1489// and symbols referred to by address-significance tables. These sections are
1490// ineligible for ICF.
1491template <class ELFT>
1492static void findKeepUniqueSections(opt::InputArgList &args) {
1493  for (auto *arg : args.filtered(OPT_keep_unique)) {
1494    StringRef name = arg->getValue();
1495    auto *d = dyn_cast_or_null<Defined>(symtab->find(name));
1496    if (!d || !d->section) {
1497      warn("could not find symbol " + name + " to keep unique");
1498      continue;
1499    }
1500    d->section->keepUnique = true;
1501  }
1502
1503  // --icf=all --ignore-data-address-equality means that we can ignore
1504  // the dynsym and address-significance tables entirely.
1505  if (config->icf == ICFLevel::All && config->ignoreDataAddressEquality)
1506    return;
1507
1508  // Symbols in the dynsym could be address-significant in other executables
1509  // or DSOs, so we conservatively mark them as address-significant.
1510  for (Symbol *sym : symtab->symbols())
1511    if (sym->includeInDynsym())
1512      markAddrsig(sym);
1513
1514  // Visit the address-significance table in each object file and mark each
1515  // referenced symbol as address-significant.
1516  for (InputFile *f : objectFiles) {
1517    auto *obj = cast<ObjFile<ELFT>>(f);
1518    ArrayRef<Symbol *> syms = obj->getSymbols();
1519    if (obj->addrsigSec) {
1520      ArrayRef<uint8_t> contents =
1521          check(obj->getObj().getSectionContents(obj->addrsigSec));
1522      const uint8_t *cur = contents.begin();
1523      while (cur != contents.end()) {
1524        unsigned size;
1525        const char *err;
1526        uint64_t symIndex = decodeULEB128(cur, &size, contents.end(), &err);
1527        if (err)
1528          fatal(toString(f) + ": could not decode addrsig section: " + err);
1529        markAddrsig(syms[symIndex]);
1530        cur += size;
1531      }
1532    } else {
1533      // If an object file does not have an address-significance table,
1534      // conservatively mark all of its symbols as address-significant.
1535      for (Symbol *s : syms)
1536        markAddrsig(s);
1537    }
1538  }
1539}
1540
1541// This function reads a symbol partition specification section. These sections
1542// are used to control which partition a symbol is allocated to. See
1543// https://lld.llvm.org/Partitions.html for more details on partitions.
1544template <typename ELFT>
1545static void readSymbolPartitionSection(InputSectionBase *s) {
1546  // Read the relocation that refers to the partition's entry point symbol.
1547  Symbol *sym;
1548  if (s->areRelocsRela)
1549    sym = &s->getFile<ELFT>()->getRelocTargetSym(s->template relas<ELFT>()[0]);
1550  else
1551    sym = &s->getFile<ELFT>()->getRelocTargetSym(s->template rels<ELFT>()[0]);
1552  if (!isa<Defined>(sym) || !sym->includeInDynsym())
1553    return;
1554
1555  StringRef partName = reinterpret_cast<const char *>(s->data().data());
1556  for (Partition &part : partitions) {
1557    if (part.name == partName) {
1558      sym->partition = part.getNumber();
1559      return;
1560    }
1561  }
1562
1563  // Forbid partitions from being used on incompatible targets, and forbid them
1564  // from being used together with various linker features that assume a single
1565  // set of output sections.
1566  if (script->hasSectionsCommand)
1567    error(toString(s->file) +
1568          ": partitions cannot be used with the SECTIONS command");
1569  if (script->hasPhdrsCommands())
1570    error(toString(s->file) +
1571          ": partitions cannot be used with the PHDRS command");
1572  if (!config->sectionStartMap.empty())
1573    error(toString(s->file) + ": partitions cannot be used with "
1574                              "--section-start, -Ttext, -Tdata or -Tbss");
1575  if (config->emachine == EM_MIPS)
1576    error(toString(s->file) + ": partitions cannot be used on this target");
1577
1578  // Impose a limit of no more than 254 partitions. This limit comes from the
1579  // sizes of the Partition fields in InputSectionBase and Symbol, as well as
1580  // the amount of space devoted to the partition number in RankFlags.
1581  if (partitions.size() == 254)
1582    fatal("may not have more than 254 partitions");
1583
1584  partitions.emplace_back();
1585  Partition &newPart = partitions.back();
1586  newPart.name = partName;
1587  sym->partition = newPart.getNumber();
1588}
1589
1590static Symbol *addUndefined(StringRef name) {
1591  return symtab->addSymbol(
1592      Undefined{nullptr, name, STB_GLOBAL, STV_DEFAULT, 0});
1593}
1594
1595// This function is where all the optimizations of link-time
1596// optimization takes place. When LTO is in use, some input files are
1597// not in native object file format but in the LLVM bitcode format.
1598// This function compiles bitcode files into a few big native files
1599// using LLVM functions and replaces bitcode symbols with the results.
1600// Because all bitcode files that the program consists of are passed to
1601// the compiler at once, it can do a whole-program optimization.
1602template <class ELFT> void LinkerDriver::compileBitcodeFiles() {
1603  // Compile bitcode files and replace bitcode symbols.
1604  lto.reset(new BitcodeCompiler);
1605  for (BitcodeFile *file : bitcodeFiles)
1606    lto->add(*file);
1607
1608  for (InputFile *file : lto->compile()) {
1609    auto *obj = cast<ObjFile<ELFT>>(file);
1610    obj->parse(/*ignoreComdats=*/true);
1611    for (Symbol *sym : obj->getGlobalSymbols())
1612      sym->parseSymbolVersion();
1613    objectFiles.push_back(file);
1614  }
1615}
1616
1617// The --wrap option is a feature to rename symbols so that you can write
1618// wrappers for existing functions. If you pass `-wrap=foo`, all
1619// occurrences of symbol `foo` are resolved to `wrap_foo` (so, you are
1620// expected to write `wrap_foo` function as a wrapper). The original
1621// symbol becomes accessible as `real_foo`, so you can call that from your
1622// wrapper.
1623//
1624// This data structure is instantiated for each -wrap option.
1625struct WrappedSymbol {
1626  Symbol *sym;
1627  Symbol *real;
1628  Symbol *wrap;
1629};
1630
1631// Handles -wrap option.
1632//
1633// This function instantiates wrapper symbols. At this point, they seem
1634// like they are not being used at all, so we explicitly set some flags so
1635// that LTO won't eliminate them.
1636static std::vector<WrappedSymbol> addWrappedSymbols(opt::InputArgList &args) {
1637  std::vector<WrappedSymbol> v;
1638  DenseSet<StringRef> seen;
1639
1640  for (auto *arg : args.filtered(OPT_wrap)) {
1641    StringRef name = arg->getValue();
1642    if (!seen.insert(name).second)
1643      continue;
1644
1645    Symbol *sym = symtab->find(name);
1646    if (!sym)
1647      continue;
1648
1649    Symbol *real = addUndefined(saver.save("__real_" + name));
1650    Symbol *wrap = addUndefined(saver.save("__wrap_" + name));
1651    v.push_back({sym, real, wrap});
1652
1653    // We want to tell LTO not to inline symbols to be overwritten
1654    // because LTO doesn't know the final symbol contents after renaming.
1655    real->canInline = false;
1656    sym->canInline = false;
1657
1658    // Tell LTO not to eliminate these symbols.
1659    sym->isUsedInRegularObj = true;
1660    wrap->isUsedInRegularObj = true;
1661  }
1662  return v;
1663}
1664
1665// Do renaming for -wrap by updating pointers to symbols.
1666//
1667// When this function is executed, only InputFiles and symbol table
1668// contain pointers to symbol objects. We visit them to replace pointers,
1669// so that wrapped symbols are swapped as instructed by the command line.
1670static void wrapSymbols(ArrayRef<WrappedSymbol> wrapped) {
1671  DenseMap<Symbol *, Symbol *> map;
1672  for (const WrappedSymbol &w : wrapped) {
1673    map[w.sym] = w.wrap;
1674    map[w.real] = w.sym;
1675  }
1676
1677  // Update pointers in input files.
1678  parallelForEach(objectFiles, [&](InputFile *file) {
1679    MutableArrayRef<Symbol *> syms = file->getMutableSymbols();
1680    for (size_t i = 0, e = syms.size(); i != e; ++i)
1681      if (Symbol *s = map.lookup(syms[i]))
1682        syms[i] = s;
1683  });
1684
1685  // Update pointers in the symbol table.
1686  for (const WrappedSymbol &w : wrapped)
1687    symtab->wrap(w.sym, w.real, w.wrap);
1688}
1689
1690// To enable CET (x86's hardware-assited control flow enforcement), each
1691// source file must be compiled with -fcf-protection. Object files compiled
1692// with the flag contain feature flags indicating that they are compatible
1693// with CET. We enable the feature only when all object files are compatible
1694// with CET.
1695//
1696// This is also the case with AARCH64's BTI and PAC which use the similar
1697// GNU_PROPERTY_AARCH64_FEATURE_1_AND mechanism.
1698template <class ELFT> static uint32_t getAndFeatures() {
1699  if (config->emachine != EM_386 && config->emachine != EM_X86_64 &&
1700      config->emachine != EM_AARCH64)
1701    return 0;
1702
1703  uint32_t ret = -1;
1704  for (InputFile *f : objectFiles) {
1705    uint32_t features = cast<ObjFile<ELFT>>(f)->andFeatures;
1706    if (config->forceBTI && !(features & GNU_PROPERTY_AARCH64_FEATURE_1_BTI)) {
1707      warn(toString(f) + ": -z force-bti: file does not have BTI property");
1708      features |= GNU_PROPERTY_AARCH64_FEATURE_1_BTI;
1709    } else if (config->zForceIbt &&
1710               !(features & GNU_PROPERTY_X86_FEATURE_1_IBT)) {
1711      warn(toString(f) + ": -z force-ibt: file does not have "
1712                         "GNU_PROPERTY_X86_FEATURE_1_IBT property");
1713      features |= GNU_PROPERTY_X86_FEATURE_1_IBT;
1714    }
1715    ret &= features;
1716  }
1717
1718  // Force enable pointer authentication Plt, we don't warn in this case as
1719  // this does not require support in the object for correctness.
1720  if (config->pacPlt)
1721    ret |= GNU_PROPERTY_AARCH64_FEATURE_1_PAC;
1722  // Force enable Shadow Stack.
1723  if (config->zShstk)
1724    ret |= GNU_PROPERTY_X86_FEATURE_1_SHSTK;
1725
1726  return ret;
1727}
1728
1729// Do actual linking. Note that when this function is called,
1730// all linker scripts have already been parsed.
1731template <class ELFT> void LinkerDriver::link(opt::InputArgList &args) {
1732  // If a -hash-style option was not given, set to a default value,
1733  // which varies depending on the target.
1734  if (!args.hasArg(OPT_hash_style)) {
1735    if (config->emachine == EM_MIPS)
1736      config->sysvHash = true;
1737    else
1738      config->sysvHash = config->gnuHash = true;
1739  }
1740
1741  // Default output filename is "a.out" by the Unix tradition.
1742  if (config->outputFile.empty())
1743    config->outputFile = "a.out";
1744
1745  // Fail early if the output file or map file is not writable. If a user has a
1746  // long link, e.g. due to a large LTO link, they do not wish to run it and
1747  // find that it failed because there was a mistake in their command-line.
1748  if (auto e = tryCreateFile(config->outputFile))
1749    error("cannot open output file " + config->outputFile + ": " + e.message());
1750  if (auto e = tryCreateFile(config->mapFile))
1751    error("cannot open map file " + config->mapFile + ": " + e.message());
1752  if (errorCount())
1753    return;
1754
1755  // Use default entry point name if no name was given via the command
1756  // line nor linker scripts. For some reason, MIPS entry point name is
1757  // different from others.
1758  config->warnMissingEntry =
1759      (!config->entry.empty() || (!config->shared && !config->relocatable));
1760  if (config->entry.empty() && !config->relocatable)
1761    config->entry = (config->emachine == EM_MIPS) ? "__start" : "_start";
1762
1763  // Handle --trace-symbol.
1764  for (auto *arg : args.filtered(OPT_trace_symbol))
1765    symtab->insert(arg->getValue())->traced = true;
1766
1767  // Add all files to the symbol table. This will add almost all
1768  // symbols that we need to the symbol table. This process might
1769  // add files to the link, via autolinking, these files are always
1770  // appended to the Files vector.
1771  for (size_t i = 0; i < files.size(); ++i)
1772    parseFile(files[i]);
1773
1774  // Now that we have every file, we can decide if we will need a
1775  // dynamic symbol table.
1776  // We need one if we were asked to export dynamic symbols or if we are
1777  // producing a shared library.
1778  // We also need one if any shared libraries are used and for pie executables
1779  // (probably because the dynamic linker needs it).
1780  config->hasDynSymTab =
1781      !sharedFiles.empty() || config->isPic || config->exportDynamic;
1782
1783  // Some symbols (such as __ehdr_start) are defined lazily only when there
1784  // are undefined symbols for them, so we add these to trigger that logic.
1785  for (StringRef name : script->referencedSymbols)
1786    addUndefined(name);
1787
1788  // Handle the `--undefined <sym>` options.
1789  for (StringRef arg : config->undefined)
1790    if (Symbol *sym = symtab->find(arg))
1791      handleUndefined(sym);
1792
1793  // If an entry symbol is in a static archive, pull out that file now.
1794  if (Symbol *sym = symtab->find(config->entry))
1795    handleUndefined(sym);
1796
1797  // Handle the `--undefined-glob <pattern>` options.
1798  for (StringRef pat : args::getStrings(args, OPT_undefined_glob))
1799    handleUndefinedGlob(pat);
1800
1801  // Mark -init and -fini symbols so that the LTO doesn't eliminate them.
1802  if (Symbol *sym = symtab->find(config->init))
1803    sym->isUsedInRegularObj = true;
1804  if (Symbol *sym = symtab->find(config->fini))
1805    sym->isUsedInRegularObj = true;
1806
1807  // If any of our inputs are bitcode files, the LTO code generator may create
1808  // references to certain library functions that might not be explicit in the
1809  // bitcode file's symbol table. If any of those library functions are defined
1810  // in a bitcode file in an archive member, we need to arrange to use LTO to
1811  // compile those archive members by adding them to the link beforehand.
1812  //
1813  // However, adding all libcall symbols to the link can have undesired
1814  // consequences. For example, the libgcc implementation of
1815  // __sync_val_compare_and_swap_8 on 32-bit ARM pulls in an .init_array entry
1816  // that aborts the program if the Linux kernel does not support 64-bit
1817  // atomics, which would prevent the program from running even if it does not
1818  // use 64-bit atomics.
1819  //
1820  // Therefore, we only add libcall symbols to the link before LTO if we have
1821  // to, i.e. if the symbol's definition is in bitcode. Any other required
1822  // libcall symbols will be added to the link after LTO when we add the LTO
1823  // object file to the link.
1824  if (!bitcodeFiles.empty())
1825    for (auto *s : lto::LTO::getRuntimeLibcallSymbols())
1826      handleLibcall(s);
1827
1828  // Return if there were name resolution errors.
1829  if (errorCount())
1830    return;
1831
1832  // Now when we read all script files, we want to finalize order of linker
1833  // script commands, which can be not yet final because of INSERT commands.
1834  script->processInsertCommands();
1835
1836  // We want to declare linker script's symbols early,
1837  // so that we can version them.
1838  // They also might be exported if referenced by DSOs.
1839  script->declareSymbols();
1840
1841  // Handle the -exclude-libs option.
1842  if (args.hasArg(OPT_exclude_libs))
1843    excludeLibs(args);
1844
1845  // Create elfHeader early. We need a dummy section in
1846  // addReservedSymbols to mark the created symbols as not absolute.
1847  Out::elfHeader = make<OutputSection>("", 0, SHF_ALLOC);
1848  Out::elfHeader->size = sizeof(typename ELFT::Ehdr);
1849
1850  // Create wrapped symbols for -wrap option.
1851  std::vector<WrappedSymbol> wrapped = addWrappedSymbols(args);
1852
1853  // We need to create some reserved symbols such as _end. Create them.
1854  if (!config->relocatable)
1855    addReservedSymbols();
1856
1857  // Apply version scripts.
1858  //
1859  // For a relocatable output, version scripts don't make sense, and
1860  // parsing a symbol version string (e.g. dropping "@ver1" from a symbol
1861  // name "foo@ver1") rather do harm, so we don't call this if -r is given.
1862  if (!config->relocatable)
1863    symtab->scanVersionScript();
1864
1865  // Do link-time optimization if given files are LLVM bitcode files.
1866  // This compiles bitcode files into real object files.
1867  //
1868  // With this the symbol table should be complete. After this, no new names
1869  // except a few linker-synthesized ones will be added to the symbol table.
1870  compileBitcodeFiles<ELFT>();
1871  if (errorCount())
1872    return;
1873
1874  // If -thinlto-index-only is given, we should create only "index
1875  // files" and not object files. Index file creation is already done
1876  // in addCombinedLTOObject, so we are done if that's the case.
1877  if (config->thinLTOIndexOnly)
1878    return;
1879
1880  // Likewise, --plugin-opt=emit-llvm is an option to make LTO create
1881  // an output file in bitcode and exit, so that you can just get a
1882  // combined bitcode file.
1883  if (config->emitLLVM)
1884    return;
1885
1886  // Apply symbol renames for -wrap.
1887  if (!wrapped.empty())
1888    wrapSymbols(wrapped);
1889
1890  // Now that we have a complete list of input files.
1891  // Beyond this point, no new files are added.
1892  // Aggregate all input sections into one place.
1893  for (InputFile *f : objectFiles)
1894    for (InputSectionBase *s : f->getSections())
1895      if (s && s != &InputSection::discarded)
1896        inputSections.push_back(s);
1897  for (BinaryFile *f : binaryFiles)
1898    for (InputSectionBase *s : f->getSections())
1899      inputSections.push_back(cast<InputSection>(s));
1900
1901  llvm::erase_if(inputSections, [](InputSectionBase *s) {
1902    if (s->type == SHT_LLVM_SYMPART) {
1903      readSymbolPartitionSection<ELFT>(s);
1904      return true;
1905    }
1906
1907    // We do not want to emit debug sections if --strip-all
1908    // or -strip-debug are given.
1909    if (config->strip == StripPolicy::None)
1910      return false;
1911
1912    if (isDebugSection(*s))
1913      return true;
1914    if (auto *isec = dyn_cast<InputSection>(s))
1915      if (InputSectionBase *rel = isec->getRelocatedSection())
1916        if (isDebugSection(*rel))
1917          return true;
1918
1919    return false;
1920  });
1921
1922  // Now that the number of partitions is fixed, save a pointer to the main
1923  // partition.
1924  mainPart = &partitions[0];
1925
1926  // Read .note.gnu.property sections from input object files which
1927  // contain a hint to tweak linker's and loader's behaviors.
1928  config->andFeatures = getAndFeatures<ELFT>();
1929
1930  // The Target instance handles target-specific stuff, such as applying
1931  // relocations or writing a PLT section. It also contains target-dependent
1932  // values such as a default image base address.
1933  target = getTarget();
1934
1935  config->eflags = target->calcEFlags();
1936  // maxPageSize (sometimes called abi page size) is the maximum page size that
1937  // the output can be run on. For example if the OS can use 4k or 64k page
1938  // sizes then maxPageSize must be 64k for the output to be useable on both.
1939  // All important alignment decisions must use this value.
1940  config->maxPageSize = getMaxPageSize(args);
1941  // commonPageSize is the most common page size that the output will be run on.
1942  // For example if an OS can use 4k or 64k page sizes and 4k is more common
1943  // than 64k then commonPageSize is set to 4k. commonPageSize can be used for
1944  // optimizations such as DATA_SEGMENT_ALIGN in linker scripts. LLD's use of it
1945  // is limited to writing trap instructions on the last executable segment.
1946  config->commonPageSize = getCommonPageSize(args);
1947
1948  config->imageBase = getImageBase(args);
1949
1950  if (config->emachine == EM_ARM) {
1951    // FIXME: These warnings can be removed when lld only uses these features
1952    // when the input objects have been compiled with an architecture that
1953    // supports them.
1954    if (config->armHasBlx == false)
1955      warn("lld uses blx instruction, no object with architecture supporting "
1956           "feature detected");
1957  }
1958
1959  // This adds a .comment section containing a version string.
1960  if (!config->relocatable)
1961    inputSections.push_back(createCommentSection());
1962
1963  // Replace common symbols with regular symbols.
1964  replaceCommonSymbols();
1965
1966  // Split SHF_MERGE and .eh_frame sections into pieces in preparation for garbage collection.
1967  splitSections<ELFT>();
1968
1969  // Garbage collection and removal of shared symbols from unused shared objects.
1970  markLive<ELFT>();
1971  demoteSharedSymbols();
1972
1973  // Make copies of any input sections that need to be copied into each
1974  // partition.
1975  copySectionsIntoPartitions();
1976
1977  // Create synthesized sections such as .got and .plt. This is called before
1978  // processSectionCommands() so that they can be placed by SECTIONS commands.
1979  createSyntheticSections<ELFT>();
1980
1981  // Some input sections that are used for exception handling need to be moved
1982  // into synthetic sections. Do that now so that they aren't assigned to
1983  // output sections in the usual way.
1984  if (!config->relocatable)
1985    combineEhSections();
1986
1987  // Create output sections described by SECTIONS commands.
1988  script->processSectionCommands();
1989
1990  // Linker scripts control how input sections are assigned to output sections.
1991  // Input sections that were not handled by scripts are called "orphans", and
1992  // they are assigned to output sections by the default rule. Process that.
1993  script->addOrphanSections();
1994
1995  // Migrate InputSectionDescription::sectionBases to sections. This includes
1996  // merging MergeInputSections into a single MergeSyntheticSection. From this
1997  // point onwards InputSectionDescription::sections should be used instead of
1998  // sectionBases.
1999  for (BaseCommand *base : script->sectionCommands)
2000    if (auto *sec = dyn_cast<OutputSection>(base))
2001      sec->finalizeInputSections();
2002  llvm::erase_if(inputSections,
2003                 [](InputSectionBase *s) { return isa<MergeInputSection>(s); });
2004
2005  // Two input sections with different output sections should not be folded.
2006  // ICF runs after processSectionCommands() so that we know the output sections.
2007  if (config->icf != ICFLevel::None) {
2008    findKeepUniqueSections<ELFT>(args);
2009    doIcf<ELFT>();
2010  }
2011
2012  // Read the callgraph now that we know what was gced or icfed
2013  if (config->callGraphProfileSort) {
2014    if (auto *arg = args.getLastArg(OPT_call_graph_ordering_file))
2015      if (Optional<MemoryBufferRef> buffer = readFile(arg->getValue()))
2016        readCallGraph(*buffer);
2017    readCallGraphsFromObjectFiles<ELFT>();
2018  }
2019
2020  // Write the result to the file.
2021  writeResult<ELFT>();
2022}
2023
2024} // namespace elf
2025} // namespace lld
2026