tools/llvm-lto2/llvm-lto2.cpp

   1 //===-- llvm-lto2: test harness for the resolution-based LTO interface ----===//
   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 // This program takes in a list of bitcode files, links them and performs
  10 // link-time optimization according to the provided symbol resolutions using the
  11 // resolution-based LTO interface, and outputs one or more object files.
  12 //
  13 // This program is intended to eventually replace llvm-lto which uses the legacy
  14 // LTO interface.
  15 //
  16 //===----------------------------------------------------------------------===//
  17
  18 #include "llvm/Bitcode/BitcodeReader.h"
  19 #include "llvm/CodeGen/CommandFlags.inc"
  20 #include "llvm/IR/DiagnosticPrinter.h"
  21 #include "llvm/LTO/Caching.h"
  22 #include "llvm/LTO/LTO.h"
  23 #include "llvm/Support/CommandLine.h"
  24 #include "llvm/Support/FileSystem.h"
  25 #include "llvm/Support/InitLLVM.h"
  26 #include "llvm/Support/TargetSelect.h"
  27 #include "llvm/Support/Threading.h"
  28
  29 using namespace llvm;
  30 using namespace lto;
  31
  32 static cl::opt<char>
  33     OptLevel("O", cl::desc("Optimization level. [-O0, -O1, -O2, or -O3] "
  34                            "(default = '-O2')"),
  35              cl::Prefix, cl::ZeroOrMore, cl::init('2'));
  36
  37 static cl::opt<char> CGOptLevel(
  38     "cg-opt-level",
  39     cl::desc("Codegen optimization level (0, 1, 2 or 3, default = '2')"),
  40     cl::init('2'));
  41
  42 static cl::list<std::string> InputFilenames(cl::Positional, cl::OneOrMore,
  43                                             cl::desc("<input bitcode files>"));
  44
  45 static cl::opt<std::string> OutputFilename("o", cl::Required,
  46                                            cl::desc("Output filename"),
  47                                            cl::value_desc("filename"));
  48
  49 static cl::opt<std::string> CacheDir("cache-dir", cl::desc("Cache Directory"),
  50                                      cl::value_desc("directory"));
  51
  52 static cl::opt<std::string> OptPipeline("opt-pipeline",
  53                                         cl::desc("Optimizer Pipeline"),
  54                                         cl::value_desc("pipeline"));
  55
  56 static cl::opt<std::string> AAPipeline("aa-pipeline",
  57                                        cl::desc("Alias Analysis Pipeline"),
  58                                        cl::value_desc("aapipeline"));
  59
  60 static cl::opt<bool> SaveTemps("save-temps", cl::desc("Save temporary files"));
  61
  62 static cl::opt<bool>
  63     ThinLTODistributedIndexes("thinlto-distributed-indexes", cl::init(false),
  64                               cl::desc("Write out individual index and "
  65                                        "import files for the "
  66                                        "distributed backend case"));
  67
  68 static cl::opt<int> Threads("thinlto-threads",
  69                             cl::init(llvm::heavyweight_hardware_concurrency()));
  70
  71 static cl::list<std::string> SymbolResolutions(
  72     "r",
  73     cl::desc("Specify a symbol resolution: filename,symbolname,resolution\n"
  74              "where \"resolution\" is a sequence (which may be empty) of the\n"
  75              "following characters:\n"
  76              " p - prevailing: the linker has chosen this definition of the\n"
  77              "     symbol\n"
  78              " l - local: the definition of this symbol is unpreemptable at\n"
  79              "     runtime and is known to be in this linkage unit\n"
  80              " x - externally visible: the definition of this symbol is\n"
  81              "     visible outside of the LTO unit\n"
  82              "A resolution for each symbol must be specified."),
  83     cl::ZeroOrMore);
  84
  85 static cl::opt<std::string> OverrideTriple(
  86     "override-triple",
  87     cl::desc("Replace target triples in input files with this triple"));
  88
  89 static cl::opt<std::string> DefaultTriple(
  90     "default-triple",
  91     cl::desc(
  92         "Replace unspecified target triples in input files with this triple"));
  93
  94 static cl::opt<bool> RemarksWithHotness(
  95     "pass-remarks-with-hotness",
  96     cl::desc("With PGO, include profile count in optimization remarks"),
  97     cl::Hidden);
  98
  99 static cl::opt<std::string>
 100     RemarksFilename("pass-remarks-output",
 101                     cl::desc("Output filename for pass remarks"),
 102                     cl::value_desc("filename"));
 103
 104 static cl::opt<std::string>
 105     RemarksPasses("pass-remarks-filter",
 106                   cl::desc("Only record optimization remarks from passes whose "
 107                            "names match the given regular expression"),
 108                   cl::value_desc("regex"));
 109
 110 static cl::opt<std::string> RemarksFormat(
 111     "pass-remarks-format",
 112     cl::desc("The format used for serializing remarks (default: YAML)"),
 113     cl::value_desc("format"), cl::init("yaml"));
 114
 115 static cl::opt<std::string>
 116     SamplePGOFile("lto-sample-profile-file",
 117                   cl::desc("Specify a SamplePGO profile file"));
 118
 119 static cl::opt<std::string>
 120     CSPGOFile("lto-cspgo-profile-file",
 121               cl::desc("Specify a context sensitive PGO profile file"));
 122
 123 static cl::opt<bool>
 124     RunCSIRInstr("lto-cspgo-gen",
 125                  cl::desc("Run PGO context sensitive IR instrumentation"),
 126                  cl::init(false), cl::Hidden);
 127
 128 static cl::opt<bool>
 129     UseNewPM("use-new-pm",
 130              cl::desc("Run LTO passes using the new pass manager"),
 131              cl::init(false), cl::Hidden);
 132
 133 static cl::opt<bool>
 134     DebugPassManager("debug-pass-manager", cl::init(false), cl::Hidden,
 135                      cl::desc("Print pass management debugging information"));
 136
 137 static cl::opt<std::string>
 138     StatsFile("stats-file", cl::desc("Filename to write statistics to"));
 139
 140 static void check(Error E, std::string Msg) {
 141   if (!E)
 142     return;
 143   handleAllErrors(std::move(E), [&](ErrorInfoBase &EIB) {
 144     errs() << "llvm-lto2: " << Msg << ": " << EIB.message().c_str() << '\n';
 145   });
 146   exit(1);
 147 }
 148
 149 template <typename T> static T check(Expected<T> E, std::string Msg) {
 150   if (E)
 151     return std::move(*E);
 152   check(E.takeError(), Msg);
 153   return T();
 154 }
 155
 156 static void check(std::error_code EC, std::string Msg) {
 157   check(errorCodeToError(EC), Msg);
 158 }
 159
 160 template <typename T> static T check(ErrorOr<T> E, std::string Msg) {
 161   if (E)
 162     return std::move(*E);
 163   check(E.getError(), Msg);
 164   return T();
 165 }
 166
 167 static int usage() {
 168   errs() << "Available subcommands: dump-symtab run\n";
 169   return 1;
 170 }
 171
 172 static int run(int argc, char **argv) {
 173   cl::ParseCommandLineOptions(argc, argv, "Resolution-based LTO test harness");
 174
 175   // FIXME: Workaround PR30396 which means that a symbol can appear
 176   // more than once if it is defined in module-level assembly and
 177   // has a GV declaration. We allow (file, symbol) pairs to have multiple
 178   // resolutions and apply them in the order observed.
 179   std::map<std::pair<std::string, std::string>, std::list<SymbolResolution>>
 180       CommandLineResolutions;
 181   for (std::string R : SymbolResolutions) {
 182     StringRef Rest = R;
 183     StringRef FileName, SymbolName;
 184     std::tie(FileName, Rest) = Rest.split(',');
 185     if (Rest.empty()) {
 186       llvm::errs() << "invalid resolution: " << R << '\n';
 187       return 1;
 188     }
 189     std::tie(SymbolName, Rest) = Rest.split(',');
 190     SymbolResolution Res;
 191     for (char C : Rest) {
 192       if (C == 'p')
 193         Res.Prevailing = true;
 194       else if (C == 'l')
 195         Res.FinalDefinitionInLinkageUnit = true;
 196       else if (C == 'x')
 197         Res.VisibleToRegularObj = true;
 198       else if (C == 'r')
 199         Res.LinkerRedefined = true;
 200       else {
 201         llvm::errs() << "invalid character " << C << " in resolution: " << R
 202                      << '\n';
 203         return 1;
 204       }
 205     }
 206     CommandLineResolutions[{FileName, SymbolName}].push_back(Res);
 207   }
 208
 209   std::vector<std::unique_ptr<MemoryBuffer>> MBs;
 210
 211   Config Conf;
 212   Conf.DiagHandler = [](const DiagnosticInfo &DI) {
 213     DiagnosticPrinterRawOStream DP(errs());
 214     DI.print(DP);
 215     errs() << '\n';
 216     if (DI.getSeverity() == DS_Error)
 217       exit(1);
 218   };
 219
 220   Conf.CPU = MCPU;
 221   Conf.Options = InitTargetOptionsFromCodeGenFlags();
 222   Conf.MAttrs = MAttrs;
 223   if (auto RM = getRelocModel())
 224     Conf.RelocModel = *RM;
 225   Conf.CodeModel = getCodeModel();
 226
 227   Conf.DebugPassManager = DebugPassManager;
 228
 229   if (SaveTemps)
 230     check(Conf.addSaveTemps(OutputFilename + "."),
 231           "Config::addSaveTemps failed");
 232
 233   // Optimization remarks.
 234   Conf.RemarksFilename = RemarksFilename;
 235   Conf.RemarksPasses = RemarksPasses;
 236   Conf.RemarksWithHotness = RemarksWithHotness;
 237   Conf.RemarksFormat = RemarksFormat;
 238
 239   Conf.SampleProfile = SamplePGOFile;
 240   Conf.CSIRProfile = CSPGOFile;
 241   Conf.RunCSIRInstr = RunCSIRInstr;
 242
 243   // Run a custom pipeline, if asked for.
 244   Conf.OptPipeline = OptPipeline;
 245   Conf.AAPipeline = AAPipeline;
 246
 247   Conf.OptLevel = OptLevel - '0';
 248   Conf.UseNewPM = UseNewPM;
 249   switch (CGOptLevel) {
 250   case '0':
 251     Conf.CGOptLevel = CodeGenOpt::None;
 252     break;
 253   case '1':
 254     Conf.CGOptLevel = CodeGenOpt::Less;
 255     break;
 256   case '2':
 257     Conf.CGOptLevel = CodeGenOpt::Default;
 258     break;
 259   case '3':
 260     Conf.CGOptLevel = CodeGenOpt::Aggressive;
 261     break;
 262   default:
 263     llvm::errs() << "invalid cg optimization level: " << CGOptLevel << '\n';
 264     return 1;
 265   }
 266
 267   if (FileType.getNumOccurrences())
 268     Conf.CGFileType = FileType;
 269
 270   Conf.OverrideTriple = OverrideTriple;
 271   Conf.DefaultTriple = DefaultTriple;
 272   Conf.StatsFile = StatsFile;
 273
 274   ThinBackend Backend;
 275   if (ThinLTODistributedIndexes)
 276     Backend = createWriteIndexesThinBackend(/* OldPrefix */ "",
 277                                             /* NewPrefix */ "",
 278                                             /* ShouldEmitImportsFiles */ true,
 279                                             /* LinkedObjectsFile */ nullptr,
 280                                             /* OnWrite */ {});
 281   else
 282     Backend = createInProcessThinBackend(Threads);
 283   LTO Lto(std::move(Conf), std::move(Backend));
 284
 285   bool HasErrors = false;
 286   for (std::string F : InputFilenames) {
 287     std::unique_ptr<MemoryBuffer> MB = check(MemoryBuffer::getFile(F), F);
 288     std::unique_ptr<InputFile> Input =
 289         check(InputFile::create(MB->getMemBufferRef()), F);
 290
 291     std::vector<SymbolResolution> Res;
 292     for (const InputFile::Symbol &Sym : Input->symbols()) {
 293       auto I = CommandLineResolutions.find({F, Sym.getName()});
 294       // If it isn't found, look for "$", which would have been added
 295       // (followed by a hash) when the symbol was promoted during module
 296       // splitting if it was defined in one part and used in the other.
 297       // Try looking up the symbol name before the "$".
 298       if (I == CommandLineResolutions.end()) {
 299         auto SplitName = Sym.getName().rsplit("$");
 300         I = CommandLineResolutions.find({F, SplitName.first});
 301       }
 302       if (I == CommandLineResolutions.end()) {
 303         llvm::errs() << argv[0] << ": missing symbol resolution for " << F
 304                      << ',' << Sym.getName() << '\n';
 305         HasErrors = true;
 306       } else {
 307         Res.push_back(I->second.front());
 308         I->second.pop_front();
 309         if (I->second.empty())
 310           CommandLineResolutions.erase(I);
 311       }
 312     }
 313
 314     if (HasErrors)
 315       continue;
 316
 317     MBs.push_back(std::move(MB));
 318     check(Lto.add(std::move(Input), Res), F);
 319   }
 320
 321   if (!CommandLineResolutions.empty()) {
 322     HasErrors = true;
 323     for (auto UnusedRes : CommandLineResolutions)
 324       llvm::errs() << argv[0] << ": unused symbol resolution for "
 325                    << UnusedRes.first.first << ',' << UnusedRes.first.second
 326                    << '\n';
 327   }
 328   if (HasErrors)
 329     return 1;
 330
 331   auto AddStream =
 332       [&](size_t Task) -> std::unique_ptr<lto::NativeObjectStream> {
 333     std::string Path = OutputFilename + "." + utostr(Task);
 334
 335     std::error_code EC;
 336     auto S = llvm::make_unique<raw_fd_ostream>(Path, EC, sys::fs::OF_None);
 337     check(EC, Path);
 338     return llvm::make_unique<lto::NativeObjectStream>(std::move(S));
 339   };
 340
 341   auto AddBuffer = [&](size_t Task, std::unique_ptr<MemoryBuffer> MB) {
 342     *AddStream(Task)->OS << MB->getBuffer();
 343   };
 344
 345   NativeObjectCache Cache;
 346   if (!CacheDir.empty())
 347     Cache = check(localCache(CacheDir, AddBuffer), "failed to create cache");
 348
 349   check(Lto.run(AddStream, Cache), "LTO::run failed");
 350   return 0;
 351 }
 352
 353 static int dumpSymtab(int argc, char **argv) {
 354   for (StringRef F : make_range(argv + 1, argv + argc)) {
 355     std::unique_ptr<MemoryBuffer> MB = check(MemoryBuffer::getFile(F), F);
 356     BitcodeFileContents BFC = check(getBitcodeFileContents(*MB), F);
 357
 358     if (BFC.Symtab.size() >= sizeof(irsymtab::storage::Header)) {
 359       auto *Hdr = reinterpret_cast<const irsymtab::storage::Header *>(
 360           BFC.Symtab.data());
 361       outs() << "version: " << Hdr->Version << '\n';
 362       if (Hdr->Version == irsymtab::storage::Header::kCurrentVersion)
 363         outs() << "producer: " << Hdr->Producer.get(BFC.StrtabForSymtab)
 364                << '\n';
 365     }
 366
 367     std::unique_ptr<InputFile> Input =
 368         check(InputFile::create(MB->getMemBufferRef()), F);
 369
 370     outs() << "target triple: " << Input->getTargetTriple() << '\n';
 371     Triple TT(Input->getTargetTriple());
 372
 373     outs() << "source filename: " << Input->getSourceFileName() << '\n';
 374
 375     if (TT.isOSBinFormatCOFF())
 376       outs() << "linker opts: " << Input->getCOFFLinkerOpts() << '\n';
 377
 378     if (TT.isOSBinFormatELF()) {
 379       outs() << "dependent libraries:";
 380       for (auto L : Input->getDependentLibraries())
 381         outs() << " \"" << L << "\"";
 382       outs() << '\n';
 383     }
 384
 385     std::vector<StringRef> ComdatTable = Input->getComdatTable();
 386     for (const InputFile::Symbol &Sym : Input->symbols()) {
 387       switch (Sym.getVisibility()) {
 388       case GlobalValue::HiddenVisibility:
 389         outs() << 'H';
 390         break;
 391       case GlobalValue::ProtectedVisibility:
 392         outs() << 'P';
 393         break;
 394       case GlobalValue::DefaultVisibility:
 395         outs() << 'D';
 396         break;
 397       }
 398
 399       auto PrintBool = [&](char C, bool B) { outs() << (B ? C : '-'); };
 400       PrintBool('U', Sym.isUndefined());
 401       PrintBool('C', Sym.isCommon());
 402       PrintBool('W', Sym.isWeak());
 403       PrintBool('I', Sym.isIndirect());
 404       PrintBool('O', Sym.canBeOmittedFromSymbolTable());
 405       PrintBool('T', Sym.isTLS());
 406       PrintBool('X', Sym.isExecutable());
 407       outs() << ' ' << Sym.getName() << '\n';
 408
 409       if (Sym.isCommon())
 410         outs() << "         size " << Sym.getCommonSize() << " align "
 411                << Sym.getCommonAlignment() << '\n';
 412
 413       int Comdat = Sym.getComdatIndex();
 414       if (Comdat != -1)
 415         outs() << "         comdat " << ComdatTable[Comdat] << '\n';
 416
 417       if (TT.isOSBinFormatCOFF() && Sym.isWeak() && Sym.isIndirect())
 418         outs() << "         fallback " << Sym.getCOFFWeakExternalFallback() << '\n';
 419
 420       if (!Sym.getSectionName().empty())
 421         outs() << "         section " << Sym.getSectionName() << "\n";
 422     }
 423
 424     outs() << '\n';
 425   }
 426
 427   return 0;
 428 }
 429
 430 int main(int argc, char **argv) {
 431   InitLLVM X(argc, argv);
 432   InitializeAllTargets();
 433   InitializeAllTargetMCs();
 434   InitializeAllAsmPrinters();
 435   InitializeAllAsmParsers();
 436
 437   // FIXME: This should use llvm::cl subcommands, but it isn't currently
 438   // possible to pass an argument not associated with a subcommand to a
 439   // subcommand (e.g. -use-new-pm).
 440   if (argc < 2)
 441     return usage();
 442
 443   StringRef Subcommand = argv[1];
 444   // Ensure that argv[0] is correct after adjusting argv/argc.
 445   argv[1] = argv[0];
 446   if (Subcommand == "dump-symtab")
 447     return dumpSymtab(argc - 1, argv + 1);
 448   if (Subcommand == "run")
 449     return run(argc - 1, argv + 1);
 450   return usage();
 451 }