1 //===- Writer.cpp ---------------------------------------------------------===// 2 // 3 // The LLVM Linker 4 // 5 // This file is distributed under the University of Illinois Open Source 6 // License. See LICENSE.TXT for details. 7 // 8 //===----------------------------------------------------------------------===// 9 10 #include "Writer.h" 11 #include "Config.h" 12 #include "Filesystem.h" 13 #include "LinkerScript.h" 14 #include "MapFile.h" 15 #include "OutputSections.h" 16 #include "Relocations.h" 17 #include "Strings.h" 18 #include "SymbolTable.h" 19 #include "SyntheticSections.h" 20 #include "Target.h" 21 #include "lld/Common/Memory.h" 22 #include "lld/Common/Threads.h" 23 #include "llvm/ADT/StringMap.h" 24 #include "llvm/ADT/StringSwitch.h" 25 #include <climits> 26 27 using namespace llvm; 28 using namespace llvm::ELF; 29 using namespace llvm::object; 30 using namespace llvm::support; 31 using namespace llvm::support::endian; 32 33 using namespace lld; 34 using namespace lld::elf; 35 36 namespace { 37 // The writer writes a SymbolTable result to a file. 38 template <class ELFT> class Writer { 39 public: 40 Writer() : Buffer(errorHandler().OutputBuffer) {} 41 typedef typename ELFT::Shdr Elf_Shdr; 42 typedef typename ELFT::Ehdr Elf_Ehdr; 43 typedef typename ELFT::Phdr Elf_Phdr; 44 45 void run(); 46 47 private: 48 void createSyntheticSections(); 49 void copyLocalSymbols(); 50 void addSectionSymbols(); 51 void addReservedSymbols(); 52 void forEachRelSec(std::function<void(InputSectionBase &)> Fn); 53 void sortSections(); 54 void sortInputSections(); 55 void finalizeSections(); 56 void addPredefinedSections(); 57 void setReservedSymbolSections(); 58 59 std::vector<PhdrEntry *> createPhdrs(); 60 void removeEmptyPTLoad(); 61 void addPtArmExid(std::vector<PhdrEntry *> &Phdrs); 62 void assignFileOffsets(); 63 void assignFileOffsetsBinary(); 64 void setPhdrs(); 65 void fixSectionAlignments(); 66 void openFile(); 67 void writeTrapInstr(); 68 void writeHeader(); 69 void writeSections(); 70 void writeSectionsBinary(); 71 void writeBuildId(); 72 73 std::unique_ptr<FileOutputBuffer> &Buffer; 74 75 void addRelIpltSymbols(); 76 void addStartEndSymbols(); 77 void addStartStopSymbols(OutputSection *Sec); 78 uint64_t getEntryAddr(); 79 OutputSection *findSection(StringRef Name); 80 81 std::vector<PhdrEntry *> Phdrs; 82 83 uint64_t FileSize; 84 uint64_t SectionHeaderOff; 85 86 bool HasGotBaseSym = false; 87 }; 88 } // anonymous namespace 89 90 StringRef elf::getOutputSectionName(InputSectionBase *S) { 91 // ".zdebug_" is a prefix for ZLIB-compressed sections. 92 // Because we decompressed input sections, we want to remove 'z'. 93 if (S->Name.startswith(".zdebug_")) 94 return Saver.save("." + S->Name.substr(2)); 95 96 if (Config->Relocatable) 97 return S->Name; 98 99 // This is for --emit-relocs. If .text.foo is emitted as .text.bar, we want 100 // to emit .rela.text.foo as .rela.text.bar for consistency (this is not 101 // technically required, but not doing it is odd). This code guarantees that. 102 if ((S->Type == SHT_REL || S->Type == SHT_RELA) && 103 !isa<SyntheticSection>(S)) { 104 OutputSection *Out = 105 cast<InputSection>(S)->getRelocatedSection()->getOutputSection(); 106 if (S->Type == SHT_RELA) 107 return Saver.save(".rela" + Out->Name); 108 return Saver.save(".rel" + Out->Name); 109 } 110 111 for (StringRef V : 112 {".text.", ".rodata.", ".data.rel.ro.", ".data.", ".bss.rel.ro.", 113 ".bss.", ".init_array.", ".fini_array.", ".ctors.", ".dtors.", ".tbss.", 114 ".gcc_except_table.", ".tdata.", ".ARM.exidx.", ".ARM.extab."}) { 115 StringRef Prefix = V.drop_back(); 116 if (S->Name.startswith(V) || S->Name == Prefix) 117 return Prefix; 118 } 119 120 // CommonSection is identified as "COMMON" in linker scripts. 121 // By default, it should go to .bss section. 122 if (S->Name == "COMMON") 123 return ".bss"; 124 125 return S->Name; 126 } 127 128 static bool needsInterpSection() { 129 return !SharedFiles.empty() && !Config->DynamicLinker.empty() && 130 Script->needsInterpSection(); 131 } 132 133 template <class ELFT> void elf::writeResult() { Writer<ELFT>().run(); } 134 135 template <class ELFT> void Writer<ELFT>::removeEmptyPTLoad() { 136 llvm::erase_if(Phdrs, [&](const PhdrEntry *P) { 137 if (P->p_type != PT_LOAD) 138 return false; 139 if (!P->FirstSec) 140 return true; 141 uint64_t Size = P->LastSec->Addr + P->LastSec->Size - P->FirstSec->Addr; 142 return Size == 0; 143 }); 144 } 145 146 template <class ELFT> static void combineEhFrameSections() { 147 for (InputSectionBase *&S : InputSections) { 148 EhInputSection *ES = dyn_cast<EhInputSection>(S); 149 if (!ES || !ES->Live) 150 continue; 151 152 InX::EhFrame->addSection<ELFT>(ES); 153 S = nullptr; 154 } 155 156 std::vector<InputSectionBase *> &V = InputSections; 157 V.erase(std::remove(V.begin(), V.end(), nullptr), V.end()); 158 } 159 160 // The main function of the writer. 161 template <class ELFT> void Writer<ELFT>::run() { 162 // Create linker-synthesized sections such as .got or .plt. 163 // Such sections are of type input section. 164 createSyntheticSections(); 165 166 if (!Config->Relocatable) 167 combineEhFrameSections<ELFT>(); 168 169 // We need to create some reserved symbols such as _end. Create them. 170 if (!Config->Relocatable) 171 addReservedSymbols(); 172 173 // We want to process linker script commands. When SECTIONS command 174 // is given we let it create sections. 175 Script->processSectionCommands(); 176 177 // Linker scripts controls how input sections are assigned to output sections. 178 // Input sections that were not handled by scripts are called "orphans", and 179 // they are assigned to output sections by the default rule. Process that. 180 Script->addOrphanSections(); 181 182 if (Config->Discard != DiscardPolicy::All) 183 copyLocalSymbols(); 184 185 if (Config->CopyRelocs) 186 addSectionSymbols(); 187 188 // Now that we have a complete set of output sections. This function 189 // completes section contents. For example, we need to add strings 190 // to the string table, and add entries to .got and .plt. 191 // finalizeSections does that. 192 finalizeSections(); 193 if (errorCount()) 194 return; 195 196 // If -compressed-debug-sections is specified, we need to compress 197 // .debug_* sections. Do it right now because it changes the size of 198 // output sections. 199 parallelForEach(OutputSections, 200 [](OutputSection *Sec) { Sec->maybeCompress<ELFT>(); }); 201 202 Script->assignAddresses(); 203 Script->allocateHeaders(Phdrs); 204 205 // Remove empty PT_LOAD to avoid causing the dynamic linker to try to mmap a 206 // 0 sized region. This has to be done late since only after assignAddresses 207 // we know the size of the sections. 208 removeEmptyPTLoad(); 209 210 if (!Config->OFormatBinary) 211 assignFileOffsets(); 212 else 213 assignFileOffsetsBinary(); 214 215 setPhdrs(); 216 217 if (Config->Relocatable) { 218 for (OutputSection *Sec : OutputSections) 219 Sec->Addr = 0; 220 } 221 222 // It does not make sense try to open the file if we have error already. 223 if (errorCount()) 224 return; 225 // Write the result down to a file. 226 openFile(); 227 if (errorCount()) 228 return; 229 230 if (!Config->OFormatBinary) { 231 writeTrapInstr(); 232 writeHeader(); 233 writeSections(); 234 } else { 235 writeSectionsBinary(); 236 } 237 238 // Backfill .note.gnu.build-id section content. This is done at last 239 // because the content is usually a hash value of the entire output file. 240 writeBuildId(); 241 if (errorCount()) 242 return; 243 244 // Handle -Map option. 245 writeMapFile(); 246 if (errorCount()) 247 return; 248 249 if (auto E = Buffer->commit()) 250 error("failed to write to the output file: " + toString(std::move(E))); 251 } 252 253 // Initialize Out members. 254 template <class ELFT> void Writer<ELFT>::createSyntheticSections() { 255 // Initialize all pointers with NULL. This is needed because 256 // you can call lld::elf::main more than once as a library. 257 memset(&Out::First, 0, sizeof(Out)); 258 259 auto Add = [](InputSectionBase *Sec) { InputSections.push_back(Sec); }; 260 261 InX::DynStrTab = make<StringTableSection>(".dynstr", true); 262 InX::Dynamic = make<DynamicSection<ELFT>>(); 263 if (Config->AndroidPackDynRelocs) { 264 In<ELFT>::RelaDyn = make<AndroidPackedRelocationSection<ELFT>>( 265 Config->IsRela ? ".rela.dyn" : ".rel.dyn"); 266 } else { 267 In<ELFT>::RelaDyn = make<RelocationSection<ELFT>>( 268 Config->IsRela ? ".rela.dyn" : ".rel.dyn", Config->ZCombreloc); 269 } 270 InX::ShStrTab = make<StringTableSection>(".shstrtab", false); 271 272 Out::ElfHeader = make<OutputSection>("", 0, SHF_ALLOC); 273 Out::ElfHeader->Size = sizeof(Elf_Ehdr); 274 Out::ProgramHeaders = make<OutputSection>("", 0, SHF_ALLOC); 275 Out::ProgramHeaders->Alignment = Config->Wordsize; 276 277 if (needsInterpSection()) { 278 InX::Interp = createInterpSection(); 279 Add(InX::Interp); 280 } else { 281 InX::Interp = nullptr; 282 } 283 284 if (Config->Strip != StripPolicy::All) { 285 InX::StrTab = make<StringTableSection>(".strtab", false); 286 InX::SymTab = make<SymbolTableSection<ELFT>>(*InX::StrTab); 287 } 288 289 if (Config->BuildId != BuildIdKind::None) { 290 InX::BuildId = make<BuildIdSection>(); 291 Add(InX::BuildId); 292 } 293 294 InX::Bss = make<BssSection>(".bss", 0, 1); 295 Add(InX::Bss); 296 297 // If there is a SECTIONS command and a .data.rel.ro section name use name 298 // .data.rel.ro.bss so that we match in the .data.rel.ro output section. 299 // This makes sure our relro is contiguous. 300 bool HasDataRelRo = 301 Script->HasSectionsCommand && findSection(".data.rel.ro"); 302 InX::BssRelRo = make<BssSection>( 303 HasDataRelRo ? ".data.rel.ro.bss" : ".bss.rel.ro", 0, 1); 304 Add(InX::BssRelRo); 305 306 // Add MIPS-specific sections. 307 if (Config->EMachine == EM_MIPS) { 308 if (!Config->Shared && Config->HasDynSymTab) { 309 InX::MipsRldMap = make<MipsRldMapSection>(); 310 Add(InX::MipsRldMap); 311 } 312 if (auto *Sec = MipsAbiFlagsSection<ELFT>::create()) 313 Add(Sec); 314 if (auto *Sec = MipsOptionsSection<ELFT>::create()) 315 Add(Sec); 316 if (auto *Sec = MipsReginfoSection<ELFT>::create()) 317 Add(Sec); 318 } 319 320 if (Config->HasDynSymTab) { 321 InX::DynSymTab = make<SymbolTableSection<ELFT>>(*InX::DynStrTab); 322 Add(InX::DynSymTab); 323 324 In<ELFT>::VerSym = make<VersionTableSection<ELFT>>(); 325 Add(In<ELFT>::VerSym); 326 327 if (!Config->VersionDefinitions.empty()) { 328 In<ELFT>::VerDef = make<VersionDefinitionSection<ELFT>>(); 329 Add(In<ELFT>::VerDef); 330 } 331 332 In<ELFT>::VerNeed = make<VersionNeedSection<ELFT>>(); 333 Add(In<ELFT>::VerNeed); 334 335 if (Config->GnuHash) { 336 InX::GnuHashTab = make<GnuHashTableSection>(); 337 Add(InX::GnuHashTab); 338 } 339 340 if (Config->SysvHash) { 341 InX::HashTab = make<HashTableSection>(); 342 Add(InX::HashTab); 343 } 344 345 Add(InX::Dynamic); 346 Add(InX::DynStrTab); 347 Add(In<ELFT>::RelaDyn); 348 } 349 350 // Add .got. MIPS' .got is so different from the other archs, 351 // it has its own class. 352 if (Config->EMachine == EM_MIPS) { 353 InX::MipsGot = make<MipsGotSection>(); 354 Add(InX::MipsGot); 355 } else { 356 InX::Got = make<GotSection>(); 357 Add(InX::Got); 358 } 359 360 InX::GotPlt = make<GotPltSection>(); 361 Add(InX::GotPlt); 362 InX::IgotPlt = make<IgotPltSection>(); 363 Add(InX::IgotPlt); 364 365 if (Config->GdbIndex) { 366 InX::GdbIndex = createGdbIndex<ELFT>(); 367 Add(InX::GdbIndex); 368 } 369 370 // We always need to add rel[a].plt to output if it has entries. 371 // Even for static linking it can contain R_[*]_IRELATIVE relocations. 372 In<ELFT>::RelaPlt = make<RelocationSection<ELFT>>( 373 Config->IsRela ? ".rela.plt" : ".rel.plt", false /*Sort*/); 374 Add(In<ELFT>::RelaPlt); 375 376 // The RelaIplt immediately follows .rel.plt (.rel.dyn for ARM) to ensure 377 // that the IRelative relocations are processed last by the dynamic loader. 378 // We cannot place the iplt section in .rel.dyn when Android relocation 379 // packing is enabled because that would cause a section type mismatch. 380 // However, because the Android dynamic loader reads .rel.plt after .rel.dyn, 381 // we can get the desired behaviour by placing the iplt section in .rel.plt. 382 In<ELFT>::RelaIplt = make<RelocationSection<ELFT>>( 383 (Config->EMachine == EM_ARM && !Config->AndroidPackDynRelocs) 384 ? ".rel.dyn" 385 : In<ELFT>::RelaPlt->Name, 386 false /*Sort*/); 387 Add(In<ELFT>::RelaIplt); 388 389 InX::Plt = make<PltSection>(Target->PltHeaderSize); 390 Add(InX::Plt); 391 InX::Iplt = make<PltSection>(0); 392 Add(InX::Iplt); 393 394 if (!Config->Relocatable) { 395 if (Config->EhFrameHdr) { 396 InX::EhFrameHdr = make<EhFrameHeader>(); 397 Add(InX::EhFrameHdr); 398 } 399 InX::EhFrame = make<EhFrameSection>(); 400 Add(InX::EhFrame); 401 } 402 403 if (InX::SymTab) 404 Add(InX::SymTab); 405 Add(InX::ShStrTab); 406 if (InX::StrTab) 407 Add(InX::StrTab); 408 } 409 410 static bool shouldKeepInSymtab(SectionBase *Sec, StringRef SymName, 411 const Symbol &B) { 412 if (B.isFile() || B.isSection()) 413 return false; 414 415 // If sym references a section in a discarded group, don't keep it. 416 if (Sec == &InputSection::Discarded) 417 return false; 418 419 if (Config->Discard == DiscardPolicy::None) 420 return true; 421 422 // In ELF assembly .L symbols are normally discarded by the assembler. 423 // If the assembler fails to do so, the linker discards them if 424 // * --discard-locals is used. 425 // * The symbol is in a SHF_MERGE section, which is normally the reason for 426 // the assembler keeping the .L symbol. 427 if (!SymName.startswith(".L") && !SymName.empty()) 428 return true; 429 430 if (Config->Discard == DiscardPolicy::Locals) 431 return false; 432 433 return !Sec || !(Sec->Flags & SHF_MERGE); 434 } 435 436 static bool includeInSymtab(const Symbol &B) { 437 if (!B.isLocal() && !B.IsUsedInRegularObj) 438 return false; 439 440 if (auto *D = dyn_cast<Defined>(&B)) { 441 // Always include absolute symbols. 442 SectionBase *Sec = D->Section; 443 if (!Sec) 444 return true; 445 if (auto *IS = dyn_cast<InputSectionBase>(Sec)) { 446 Sec = IS->Repl; 447 IS = cast<InputSectionBase>(Sec); 448 // Exclude symbols pointing to garbage-collected sections. 449 if (!IS->Live) 450 return false; 451 } 452 if (auto *S = dyn_cast<MergeInputSection>(Sec)) 453 if (!S->getSectionPiece(D->Value)->Live) 454 return false; 455 return true; 456 } 457 return B.Used; 458 } 459 460 // Local symbols are not in the linker's symbol table. This function scans 461 // each object file's symbol table to copy local symbols to the output. 462 template <class ELFT> void Writer<ELFT>::copyLocalSymbols() { 463 if (!InX::SymTab) 464 return; 465 for (InputFile *File : ObjectFiles) { 466 ObjFile<ELFT> *F = cast<ObjFile<ELFT>>(File); 467 for (Symbol *B : F->getLocalSymbols()) { 468 if (!B->isLocal()) 469 fatal(toString(F) + 470 ": broken object: getLocalSymbols returns a non-local symbol"); 471 auto *DR = dyn_cast<Defined>(B); 472 473 // No reason to keep local undefined symbol in symtab. 474 if (!DR) 475 continue; 476 if (!includeInSymtab(*B)) 477 continue; 478 479 SectionBase *Sec = DR->Section; 480 if (!shouldKeepInSymtab(Sec, B->getName(), *B)) 481 continue; 482 InX::SymTab->addSymbol(B); 483 } 484 } 485 } 486 487 template <class ELFT> void Writer<ELFT>::addSectionSymbols() { 488 // Create a section symbol for each output section so that we can represent 489 // relocations that point to the section. If we know that no relocation is 490 // referring to a section (that happens if the section is a synthetic one), we 491 // don't create a section symbol for that section. 492 for (BaseCommand *Base : Script->SectionCommands) { 493 auto *Sec = dyn_cast<OutputSection>(Base); 494 if (!Sec) 495 continue; 496 auto I = llvm::find_if(Sec->SectionCommands, [](BaseCommand *Base) { 497 if (auto *ISD = dyn_cast<InputSectionDescription>(Base)) 498 return !ISD->Sections.empty(); 499 return false; 500 }); 501 if (I == Sec->SectionCommands.end()) 502 continue; 503 InputSection *IS = cast<InputSectionDescription>(*I)->Sections[0]; 504 505 // Relocations are not using REL[A] section symbols. 506 if (IS->Type == SHT_REL || IS->Type == SHT_RELA) 507 continue; 508 509 // Unlike other synthetic sections, mergeable output sections contain data 510 // copied from input sections, and there may be a relocation pointing to its 511 // contents if -r or -emit-reloc are given. 512 if (isa<SyntheticSection>(IS) && !(IS->Flags & SHF_MERGE)) 513 continue; 514 515 auto *Sym = 516 make<Defined>(IS->File, "", STB_LOCAL, /*StOther=*/0, STT_SECTION, 517 /*Value=*/0, /*Size=*/0, IS); 518 InX::SymTab->addSymbol(Sym); 519 } 520 } 521 522 // Today's loaders have a feature to make segments read-only after 523 // processing dynamic relocations to enhance security. PT_GNU_RELRO 524 // is defined for that. 525 // 526 // This function returns true if a section needs to be put into a 527 // PT_GNU_RELRO segment. 528 static bool isRelroSection(const OutputSection *Sec) { 529 if (!Config->ZRelro) 530 return false; 531 532 uint64_t Flags = Sec->Flags; 533 534 // Non-allocatable or non-writable sections don't need RELRO because 535 // they are not writable or not even mapped to memory in the first place. 536 // RELRO is for sections that are essentially read-only but need to 537 // be writable only at process startup to allow dynamic linker to 538 // apply relocations. 539 if (!(Flags & SHF_ALLOC) || !(Flags & SHF_WRITE)) 540 return false; 541 542 // Once initialized, TLS data segments are used as data templates 543 // for a thread-local storage. For each new thread, runtime 544 // allocates memory for a TLS and copy templates there. No thread 545 // are supposed to use templates directly. Thus, it can be in RELRO. 546 if (Flags & SHF_TLS) 547 return true; 548 549 // .init_array, .preinit_array and .fini_array contain pointers to 550 // functions that are executed on process startup or exit. These 551 // pointers are set by the static linker, and they are not expected 552 // to change at runtime. But if you are an attacker, you could do 553 // interesting things by manipulating pointers in .fini_array, for 554 // example. So they are put into RELRO. 555 uint32_t Type = Sec->Type; 556 if (Type == SHT_INIT_ARRAY || Type == SHT_FINI_ARRAY || 557 Type == SHT_PREINIT_ARRAY) 558 return true; 559 560 // .got contains pointers to external symbols. They are resolved by 561 // the dynamic linker when a module is loaded into memory, and after 562 // that they are not expected to change. So, it can be in RELRO. 563 if (InX::Got && Sec == InX::Got->getParent()) 564 return true; 565 566 // .got.plt contains pointers to external function symbols. They are 567 // by default resolved lazily, so we usually cannot put it into RELRO. 568 // However, if "-z now" is given, the lazy symbol resolution is 569 // disabled, which enables us to put it into RELRO. 570 if (Sec == InX::GotPlt->getParent()) 571 return Config->ZNow; 572 573 // .dynamic section contains data for the dynamic linker, and 574 // there's no need to write to it at runtime, so it's better to put 575 // it into RELRO. 576 if (Sec == InX::Dynamic->getParent()) 577 return true; 578 579 // .bss.rel.ro is used for copy relocations for read-only symbols. 580 // Since the dynamic linker needs to process copy relocations, the 581 // section cannot be read-only, but once initialized, they shouldn't 582 // change. 583 if (Sec == InX::BssRelRo->getParent()) 584 return true; 585 586 // Sections with some special names are put into RELRO. This is a 587 // bit unfortunate because section names shouldn't be significant in 588 // ELF in spirit. But in reality many linker features depend on 589 // magic section names. 590 StringRef S = Sec->Name; 591 return S == ".data.rel.ro" || S == ".ctors" || S == ".dtors" || S == ".jcr" || 592 S == ".eh_frame" || S == ".openbsd.randomdata"; 593 } 594 595 // We compute a rank for each section. The rank indicates where the 596 // section should be placed in the file. Instead of using simple 597 // numbers (0,1,2...), we use a series of flags. One for each decision 598 // point when placing the section. 599 // Using flags has two key properties: 600 // * It is easy to check if a give branch was taken. 601 // * It is easy two see how similar two ranks are (see getRankProximity). 602 enum RankFlags { 603 RF_NOT_ADDR_SET = 1 << 16, 604 RF_NOT_INTERP = 1 << 15, 605 RF_NOT_ALLOC = 1 << 14, 606 RF_WRITE = 1 << 13, 607 RF_EXEC_WRITE = 1 << 12, 608 RF_EXEC = 1 << 11, 609 RF_NON_TLS_BSS = 1 << 10, 610 RF_NON_TLS_BSS_RO = 1 << 9, 611 RF_NOT_TLS = 1 << 8, 612 RF_BSS = 1 << 7, 613 RF_PPC_NOT_TOCBSS = 1 << 6, 614 RF_PPC_OPD = 1 << 5, 615 RF_PPC_TOCL = 1 << 4, 616 RF_PPC_TOC = 1 << 3, 617 RF_PPC_BRANCH_LT = 1 << 2, 618 RF_MIPS_GPREL = 1 << 1, 619 RF_MIPS_NOT_GOT = 1 << 0 620 }; 621 622 static unsigned getSectionRank(const OutputSection *Sec) { 623 unsigned Rank = 0; 624 625 // We want to put section specified by -T option first, so we 626 // can start assigning VA starting from them later. 627 if (Config->SectionStartMap.count(Sec->Name)) 628 return Rank; 629 Rank |= RF_NOT_ADDR_SET; 630 631 // Put .interp first because some loaders want to see that section 632 // on the first page of the executable file when loaded into memory. 633 if (Sec->Name == ".interp") 634 return Rank; 635 Rank |= RF_NOT_INTERP; 636 637 // Allocatable sections go first to reduce the total PT_LOAD size and 638 // so debug info doesn't change addresses in actual code. 639 if (!(Sec->Flags & SHF_ALLOC)) 640 return Rank | RF_NOT_ALLOC; 641 642 // Sort sections based on their access permission in the following 643 // order: R, RX, RWX, RW. This order is based on the following 644 // considerations: 645 // * Read-only sections come first such that they go in the 646 // PT_LOAD covering the program headers at the start of the file. 647 // * Read-only, executable sections come next, unless the 648 // -no-rosegment option is used. 649 // * Writable, executable sections follow such that .plt on 650 // architectures where it needs to be writable will be placed 651 // between .text and .data. 652 // * Writable sections come last, such that .bss lands at the very 653 // end of the last PT_LOAD. 654 bool IsExec = Sec->Flags & SHF_EXECINSTR; 655 bool IsWrite = Sec->Flags & SHF_WRITE; 656 657 if (IsExec) { 658 if (IsWrite) 659 Rank |= RF_EXEC_WRITE; 660 else if (!Config->SingleRoRx) 661 Rank |= RF_EXEC; 662 } else { 663 if (IsWrite) 664 Rank |= RF_WRITE; 665 } 666 667 // If we got here we know that both A and B are in the same PT_LOAD. 668 669 bool IsTls = Sec->Flags & SHF_TLS; 670 bool IsNoBits = Sec->Type == SHT_NOBITS; 671 672 // The first requirement we have is to put (non-TLS) nobits sections last. The 673 // reason is that the only thing the dynamic linker will see about them is a 674 // p_memsz that is larger than p_filesz. Seeing that it zeros the end of the 675 // PT_LOAD, so that has to correspond to the nobits sections. 676 bool IsNonTlsNoBits = IsNoBits && !IsTls; 677 if (IsNonTlsNoBits) 678 Rank |= RF_NON_TLS_BSS; 679 680 // We place nobits RelRo sections before plain r/w ones, and non-nobits RelRo 681 // sections after r/w ones, so that the RelRo sections are contiguous. 682 bool IsRelRo = isRelroSection(Sec); 683 if (IsNonTlsNoBits && !IsRelRo) 684 Rank |= RF_NON_TLS_BSS_RO; 685 if (!IsNonTlsNoBits && IsRelRo) 686 Rank |= RF_NON_TLS_BSS_RO; 687 688 // The TLS initialization block needs to be a single contiguous block in a R/W 689 // PT_LOAD, so stick TLS sections directly before the other RelRo R/W 690 // sections. The TLS NOBITS sections are placed here as they don't take up 691 // virtual address space in the PT_LOAD. 692 if (!IsTls) 693 Rank |= RF_NOT_TLS; 694 695 // Within the TLS initialization block, the non-nobits sections need to appear 696 // first. 697 if (IsNoBits) 698 Rank |= RF_BSS; 699 700 // Some architectures have additional ordering restrictions for sections 701 // within the same PT_LOAD. 702 if (Config->EMachine == EM_PPC64) { 703 // PPC64 has a number of special SHT_PROGBITS+SHF_ALLOC+SHF_WRITE sections 704 // that we would like to make sure appear is a specific order to maximize 705 // their coverage by a single signed 16-bit offset from the TOC base 706 // pointer. Conversely, the special .tocbss section should be first among 707 // all SHT_NOBITS sections. This will put it next to the loaded special 708 // PPC64 sections (and, thus, within reach of the TOC base pointer). 709 StringRef Name = Sec->Name; 710 if (Name != ".tocbss") 711 Rank |= RF_PPC_NOT_TOCBSS; 712 713 if (Name == ".opd") 714 Rank |= RF_PPC_OPD; 715 716 if (Name == ".toc1") 717 Rank |= RF_PPC_TOCL; 718 719 if (Name == ".toc") 720 Rank |= RF_PPC_TOC; 721 722 if (Name == ".branch_lt") 723 Rank |= RF_PPC_BRANCH_LT; 724 } 725 if (Config->EMachine == EM_MIPS) { 726 // All sections with SHF_MIPS_GPREL flag should be grouped together 727 // because data in these sections is addressable with a gp relative address. 728 if (Sec->Flags & SHF_MIPS_GPREL) 729 Rank |= RF_MIPS_GPREL; 730 731 if (Sec->Name != ".got") 732 Rank |= RF_MIPS_NOT_GOT; 733 } 734 735 return Rank; 736 } 737 738 static bool compareSections(const BaseCommand *ACmd, const BaseCommand *BCmd) { 739 const OutputSection *A = cast<OutputSection>(ACmd); 740 const OutputSection *B = cast<OutputSection>(BCmd); 741 if (A->SortRank != B->SortRank) 742 return A->SortRank < B->SortRank; 743 if (!(A->SortRank & RF_NOT_ADDR_SET)) 744 return Config->SectionStartMap.lookup(A->Name) < 745 Config->SectionStartMap.lookup(B->Name); 746 return false; 747 } 748 749 void PhdrEntry::add(OutputSection *Sec) { 750 LastSec = Sec; 751 if (!FirstSec) 752 FirstSec = Sec; 753 p_align = std::max(p_align, Sec->Alignment); 754 if (p_type == PT_LOAD) 755 Sec->PtLoad = this; 756 } 757 758 template <class ELFT> 759 static Defined *addOptionalRegular(StringRef Name, SectionBase *Sec, 760 uint64_t Val, uint8_t StOther = STV_HIDDEN, 761 uint8_t Binding = STB_GLOBAL) { 762 Symbol *S = Symtab->find(Name); 763 if (!S || S->isDefined()) 764 return nullptr; 765 Symbol *Sym = Symtab->addRegular<ELFT>(Name, StOther, STT_NOTYPE, Val, 766 /*Size=*/0, Binding, Sec, 767 /*File=*/nullptr); 768 return cast<Defined>(Sym); 769 } 770 771 // The beginning and the ending of .rel[a].plt section are marked 772 // with __rel[a]_iplt_{start,end} symbols if it is a statically linked 773 // executable. The runtime needs these symbols in order to resolve 774 // all IRELATIVE relocs on startup. For dynamic executables, we don't 775 // need these symbols, since IRELATIVE relocs are resolved through GOT 776 // and PLT. For details, see http://www.airs.com/blog/archives/403. 777 template <class ELFT> void Writer<ELFT>::addRelIpltSymbols() { 778 if (!Config->Static) 779 return; 780 StringRef S = Config->IsRela ? "__rela_iplt_start" : "__rel_iplt_start"; 781 addOptionalRegular<ELFT>(S, In<ELFT>::RelaIplt, 0, STV_HIDDEN, STB_WEAK); 782 783 S = Config->IsRela ? "__rela_iplt_end" : "__rel_iplt_end"; 784 addOptionalRegular<ELFT>(S, In<ELFT>::RelaIplt, -1, STV_HIDDEN, STB_WEAK); 785 } 786 787 // The linker is expected to define some symbols depending on 788 // the linking result. This function defines such symbols. 789 template <class ELFT> void Writer<ELFT>::addReservedSymbols() { 790 if (Config->EMachine == EM_MIPS) { 791 // Define _gp for MIPS. st_value of _gp symbol will be updated by Writer 792 // so that it points to an absolute address which by default is relative 793 // to GOT. Default offset is 0x7ff0. 794 // See "Global Data Symbols" in Chapter 6 in the following document: 795 // ftp://www.linux-mips.org/pub/linux/mips/doc/ABI/mipsabi.pdf 796 ElfSym::MipsGp = Symtab->addAbsolute<ELFT>("_gp", STV_HIDDEN, STB_GLOBAL); 797 798 // On MIPS O32 ABI, _gp_disp is a magic symbol designates offset between 799 // start of function and 'gp' pointer into GOT. 800 if (Symtab->find("_gp_disp")) 801 ElfSym::MipsGpDisp = 802 Symtab->addAbsolute<ELFT>("_gp_disp", STV_HIDDEN, STB_GLOBAL); 803 804 // The __gnu_local_gp is a magic symbol equal to the current value of 'gp' 805 // pointer. This symbol is used in the code generated by .cpload pseudo-op 806 // in case of using -mno-shared option. 807 // https://sourceware.org/ml/binutils/2004-12/msg00094.html 808 if (Symtab->find("__gnu_local_gp")) 809 ElfSym::MipsLocalGp = 810 Symtab->addAbsolute<ELFT>("__gnu_local_gp", STV_HIDDEN, STB_GLOBAL); 811 } 812 813 // The _GLOBAL_OFFSET_TABLE_ symbol is defined by target convention to 814 // be at some offset from the base of the .got section, usually 0 or the end 815 // of the .got 816 InputSection *GotSection = InX::MipsGot ? cast<InputSection>(InX::MipsGot) 817 : cast<InputSection>(InX::Got); 818 ElfSym::GlobalOffsetTable = addOptionalRegular<ELFT>( 819 "_GLOBAL_OFFSET_TABLE_", GotSection, Target->GotBaseSymOff); 820 821 // __ehdr_start is the location of ELF file headers. Note that we define 822 // this symbol unconditionally even when using a linker script, which 823 // differs from the behavior implemented by GNU linker which only define 824 // this symbol if ELF headers are in the memory mapped segment. 825 // __executable_start is not documented, but the expectation of at 826 // least the android libc is that it points to the elf header too. 827 // __dso_handle symbol is passed to cxa_finalize as a marker to identify 828 // each DSO. The address of the symbol doesn't matter as long as they are 829 // different in different DSOs, so we chose the start address of the DSO. 830 for (const char *Name : 831 {"__ehdr_start", "__executable_start", "__dso_handle"}) 832 addOptionalRegular<ELFT>(Name, Out::ElfHeader, 0, STV_HIDDEN); 833 834 // If linker script do layout we do not need to create any standart symbols. 835 if (Script->HasSectionsCommand) 836 return; 837 838 auto Add = [](StringRef S, int64_t Pos) { 839 return addOptionalRegular<ELFT>(S, Out::ElfHeader, Pos, STV_DEFAULT); 840 }; 841 842 ElfSym::Bss = Add("__bss_start", 0); 843 ElfSym::End1 = Add("end", -1); 844 ElfSym::End2 = Add("_end", -1); 845 ElfSym::Etext1 = Add("etext", -1); 846 ElfSym::Etext2 = Add("_etext", -1); 847 ElfSym::Edata1 = Add("edata", -1); 848 ElfSym::Edata2 = Add("_edata", -1); 849 } 850 851 template <class ELFT> 852 void Writer<ELFT>::forEachRelSec(std::function<void(InputSectionBase &)> Fn) { 853 // Scan all relocations. Each relocation goes through a series 854 // of tests to determine if it needs special treatment, such as 855 // creating GOT, PLT, copy relocations, etc. 856 // Note that relocations for non-alloc sections are directly 857 // processed by InputSection::relocateNonAlloc. 858 for (InputSectionBase *IS : InputSections) 859 if (IS->Live && isa<InputSection>(IS) && (IS->Flags & SHF_ALLOC)) 860 Fn(*IS); 861 for (EhInputSection *ES : InX::EhFrame->Sections) 862 Fn(*ES); 863 } 864 865 // This function generates assignments for predefined symbols (e.g. _end or 866 // _etext) and inserts them into the commands sequence to be processed at the 867 // appropriate time. This ensures that the value is going to be correct by the 868 // time any references to these symbols are processed and is equivalent to 869 // defining these symbols explicitly in the linker script. 870 template <class ELFT> void Writer<ELFT>::setReservedSymbolSections() { 871 PhdrEntry *Last = nullptr; 872 PhdrEntry *LastRO = nullptr; 873 874 for (PhdrEntry *P : Phdrs) { 875 if (P->p_type != PT_LOAD) 876 continue; 877 Last = P; 878 if (!(P->p_flags & PF_W)) 879 LastRO = P; 880 } 881 882 if (LastRO) { 883 // _etext is the first location after the last read-only loadable segment. 884 if (ElfSym::Etext1) 885 ElfSym::Etext1->Section = LastRO->LastSec; 886 if (ElfSym::Etext2) 887 ElfSym::Etext2->Section = LastRO->LastSec; 888 } 889 890 if (Last) { 891 // _edata points to the end of the last mapped initialized section. 892 OutputSection *Edata = nullptr; 893 for (OutputSection *OS : OutputSections) { 894 if (OS->Type != SHT_NOBITS) 895 Edata = OS; 896 if (OS == Last->LastSec) 897 break; 898 } 899 900 if (ElfSym::Edata1) 901 ElfSym::Edata1->Section = Edata; 902 if (ElfSym::Edata2) 903 ElfSym::Edata2->Section = Edata; 904 905 // _end is the first location after the uninitialized data region. 906 if (ElfSym::End1) 907 ElfSym::End1->Section = Last->LastSec; 908 if (ElfSym::End2) 909 ElfSym::End2->Section = Last->LastSec; 910 } 911 912 if (ElfSym::Bss) 913 ElfSym::Bss->Section = findSection(".bss"); 914 915 // Setup MIPS _gp_disp/__gnu_local_gp symbols which should 916 // be equal to the _gp symbol's value. 917 if (ElfSym::MipsGp) { 918 // Find GP-relative section with the lowest address 919 // and use this address to calculate default _gp value. 920 for (OutputSection *OS : OutputSections) { 921 if (OS->Flags & SHF_MIPS_GPREL) { 922 ElfSym::MipsGp->Section = OS; 923 ElfSym::MipsGp->Value = 0x7ff0; 924 break; 925 } 926 } 927 } 928 } 929 930 // We want to find how similar two ranks are. 931 // The more branches in getSectionRank that match, the more similar they are. 932 // Since each branch corresponds to a bit flag, we can just use 933 // countLeadingZeros. 934 static int getRankProximityAux(OutputSection *A, OutputSection *B) { 935 return countLeadingZeros(A->SortRank ^ B->SortRank); 936 } 937 938 static int getRankProximity(OutputSection *A, BaseCommand *B) { 939 if (auto *Sec = dyn_cast<OutputSection>(B)) 940 if (Sec->Live) 941 return getRankProximityAux(A, Sec); 942 return -1; 943 } 944 945 // When placing orphan sections, we want to place them after symbol assignments 946 // so that an orphan after 947 // begin_foo = .; 948 // foo : { *(foo) } 949 // end_foo = .; 950 // doesn't break the intended meaning of the begin/end symbols. 951 // We don't want to go over sections since findOrphanPos is the 952 // one in charge of deciding the order of the sections. 953 // We don't want to go over changes to '.', since doing so in 954 // rx_sec : { *(rx_sec) } 955 // . = ALIGN(0x1000); 956 // /* The RW PT_LOAD starts here*/ 957 // rw_sec : { *(rw_sec) } 958 // would mean that the RW PT_LOAD would become unaligned. 959 static bool shouldSkip(BaseCommand *Cmd) { 960 if (isa<OutputSection>(Cmd)) 961 return false; 962 if (auto *Assign = dyn_cast<SymbolAssignment>(Cmd)) 963 return Assign->Name != "."; 964 return true; 965 } 966 967 // We want to place orphan sections so that they share as much 968 // characteristics with their neighbors as possible. For example, if 969 // both are rw, or both are tls. 970 template <typename ELFT> 971 static std::vector<BaseCommand *>::iterator 972 findOrphanPos(std::vector<BaseCommand *>::iterator B, 973 std::vector<BaseCommand *>::iterator E) { 974 OutputSection *Sec = cast<OutputSection>(*E); 975 976 // Find the first element that has as close a rank as possible. 977 auto I = std::max_element(B, E, [=](BaseCommand *A, BaseCommand *B) { 978 return getRankProximity(Sec, A) < getRankProximity(Sec, B); 979 }); 980 if (I == E) 981 return E; 982 983 // Consider all existing sections with the same proximity. 984 int Proximity = getRankProximity(Sec, *I); 985 for (; I != E; ++I) { 986 auto *CurSec = dyn_cast<OutputSection>(*I); 987 if (!CurSec || !CurSec->Live) 988 continue; 989 if (getRankProximity(Sec, CurSec) != Proximity || 990 Sec->SortRank < CurSec->SortRank) 991 break; 992 } 993 994 auto IsLiveSection = [](BaseCommand *Cmd) { 995 auto *OS = dyn_cast<OutputSection>(Cmd); 996 return OS && OS->Live; 997 }; 998 999 auto J = std::find_if(llvm::make_reverse_iterator(I), 1000 llvm::make_reverse_iterator(B), IsLiveSection); 1001 I = J.base(); 1002 1003 // As a special case, if the orphan section is the last section, put 1004 // it at the very end, past any other commands. 1005 // This matches bfd's behavior and is convenient when the linker script fully 1006 // specifies the start of the file, but doesn't care about the end (the non 1007 // alloc sections for example). 1008 auto NextSec = std::find_if(I, E, IsLiveSection); 1009 if (NextSec == E) 1010 return E; 1011 1012 while (I != E && shouldSkip(*I)) 1013 ++I; 1014 return I; 1015 } 1016 1017 // If no layout was provided by linker script, we want to apply default 1018 // sorting for special input sections and handle --symbol-ordering-file. 1019 template <class ELFT> void Writer<ELFT>::sortInputSections() { 1020 assert(!Script->HasSectionsCommand); 1021 1022 // Sort input sections by priority using the list provided 1023 // by --symbol-ordering-file. 1024 DenseMap<SectionBase *, int> Order = buildSectionOrder(); 1025 if (!Order.empty()) 1026 for (BaseCommand *Base : Script->SectionCommands) 1027 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1028 if (Sec->Live) 1029 Sec->sort([&](InputSectionBase *S) { return Order.lookup(S); }); 1030 1031 // Sort input sections by section name suffixes for 1032 // __attribute__((init_priority(N))). 1033 if (OutputSection *Sec = findSection(".init_array")) 1034 Sec->sortInitFini(); 1035 if (OutputSection *Sec = findSection(".fini_array")) 1036 Sec->sortInitFini(); 1037 1038 // Sort input sections by the special rule for .ctors and .dtors. 1039 if (OutputSection *Sec = findSection(".ctors")) 1040 Sec->sortCtorsDtors(); 1041 if (OutputSection *Sec = findSection(".dtors")) 1042 Sec->sortCtorsDtors(); 1043 } 1044 1045 template <class ELFT> void Writer<ELFT>::sortSections() { 1046 Script->adjustSectionsBeforeSorting(); 1047 1048 // Don't sort if using -r. It is not necessary and we want to preserve the 1049 // relative order for SHF_LINK_ORDER sections. 1050 if (Config->Relocatable) 1051 return; 1052 1053 for (BaseCommand *Base : Script->SectionCommands) 1054 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1055 Sec->SortRank = getSectionRank(Sec); 1056 1057 if (!Script->HasSectionsCommand) { 1058 sortInputSections(); 1059 1060 // We know that all the OutputSections are contiguous in this case. 1061 auto E = Script->SectionCommands.end(); 1062 auto I = Script->SectionCommands.begin(); 1063 auto IsSection = [](BaseCommand *Base) { return isa<OutputSection>(Base); }; 1064 I = std::find_if(I, E, IsSection); 1065 E = std::find_if(llvm::make_reverse_iterator(E), 1066 llvm::make_reverse_iterator(I), IsSection) 1067 .base(); 1068 std::stable_sort(I, E, compareSections); 1069 return; 1070 } 1071 1072 // Orphan sections are sections present in the input files which are 1073 // not explicitly placed into the output file by the linker script. 1074 // 1075 // The sections in the linker script are already in the correct 1076 // order. We have to figuere out where to insert the orphan 1077 // sections. 1078 // 1079 // The order of the sections in the script is arbitrary and may not agree with 1080 // compareSections. This means that we cannot easily define a strict weak 1081 // ordering. To see why, consider a comparison of a section in the script and 1082 // one not in the script. We have a two simple options: 1083 // * Make them equivalent (a is not less than b, and b is not less than a). 1084 // The problem is then that equivalence has to be transitive and we can 1085 // have sections a, b and c with only b in a script and a less than c 1086 // which breaks this property. 1087 // * Use compareSectionsNonScript. Given that the script order doesn't have 1088 // to match, we can end up with sections a, b, c, d where b and c are in the 1089 // script and c is compareSectionsNonScript less than b. In which case d 1090 // can be equivalent to c, a to b and d < a. As a concrete example: 1091 // .a (rx) # not in script 1092 // .b (rx) # in script 1093 // .c (ro) # in script 1094 // .d (ro) # not in script 1095 // 1096 // The way we define an order then is: 1097 // * Sort only the orphan sections. They are in the end right now. 1098 // * Move each orphan section to its preferred position. We try 1099 // to put each section in the last position where it it can share 1100 // a PT_LOAD. 1101 // 1102 // There is some ambiguity as to where exactly a new entry should be 1103 // inserted, because Commands contains not only output section 1104 // commands but also other types of commands such as symbol assignment 1105 // expressions. There's no correct answer here due to the lack of the 1106 // formal specification of the linker script. We use heuristics to 1107 // determine whether a new output command should be added before or 1108 // after another commands. For the details, look at shouldSkip 1109 // function. 1110 1111 auto I = Script->SectionCommands.begin(); 1112 auto E = Script->SectionCommands.end(); 1113 auto NonScriptI = std::find_if(I, E, [](BaseCommand *Base) { 1114 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1115 return Sec->Live && Sec->SectionIndex == INT_MAX; 1116 return false; 1117 }); 1118 1119 // Sort the orphan sections. 1120 std::stable_sort(NonScriptI, E, compareSections); 1121 1122 // As a horrible special case, skip the first . assignment if it is before any 1123 // section. We do this because it is common to set a load address by starting 1124 // the script with ". = 0xabcd" and the expectation is that every section is 1125 // after that. 1126 auto FirstSectionOrDotAssignment = 1127 std::find_if(I, E, [](BaseCommand *Cmd) { return !shouldSkip(Cmd); }); 1128 if (FirstSectionOrDotAssignment != E && 1129 isa<SymbolAssignment>(**FirstSectionOrDotAssignment)) 1130 ++FirstSectionOrDotAssignment; 1131 I = FirstSectionOrDotAssignment; 1132 1133 while (NonScriptI != E) { 1134 auto Pos = findOrphanPos<ELFT>(I, NonScriptI); 1135 OutputSection *Orphan = cast<OutputSection>(*NonScriptI); 1136 1137 // As an optimization, find all sections with the same sort rank 1138 // and insert them with one rotate. 1139 unsigned Rank = Orphan->SortRank; 1140 auto End = std::find_if(NonScriptI + 1, E, [=](BaseCommand *Cmd) { 1141 return cast<OutputSection>(Cmd)->SortRank != Rank; 1142 }); 1143 std::rotate(Pos, NonScriptI, End); 1144 NonScriptI = End; 1145 } 1146 1147 Script->adjustSectionsAfterSorting(); 1148 } 1149 1150 static void applySynthetic(const std::vector<SyntheticSection *> &Sections, 1151 std::function<void(SyntheticSection *)> Fn) { 1152 for (SyntheticSection *SS : Sections) 1153 if (SS && SS->getParent() && !SS->empty()) 1154 Fn(SS); 1155 } 1156 1157 // In order to allow users to manipulate linker-synthesized sections, 1158 // we had to add synthetic sections to the input section list early, 1159 // even before we make decisions whether they are needed. This allows 1160 // users to write scripts like this: ".mygot : { .got }". 1161 // 1162 // Doing it has an unintended side effects. If it turns out that we 1163 // don't need a .got (for example) at all because there's no 1164 // relocation that needs a .got, we don't want to emit .got. 1165 // 1166 // To deal with the above problem, this function is called after 1167 // scanRelocations is called to remove synthetic sections that turn 1168 // out to be empty. 1169 static void removeUnusedSyntheticSections() { 1170 // All input synthetic sections that can be empty are placed after 1171 // all regular ones. We iterate over them all and exit at first 1172 // non-synthetic. 1173 for (InputSectionBase *S : llvm::reverse(InputSections)) { 1174 SyntheticSection *SS = dyn_cast<SyntheticSection>(S); 1175 if (!SS) 1176 return; 1177 OutputSection *OS = SS->getParent(); 1178 if (!SS->empty() || !OS) 1179 continue; 1180 1181 std::vector<BaseCommand *>::iterator Empty = OS->SectionCommands.end(); 1182 for (auto I = OS->SectionCommands.begin(), E = OS->SectionCommands.end(); 1183 I != E; ++I) { 1184 BaseCommand *B = *I; 1185 if (auto *ISD = dyn_cast<InputSectionDescription>(B)) { 1186 llvm::erase_if(ISD->Sections, 1187 [=](InputSection *IS) { return IS == SS; }); 1188 if (ISD->Sections.empty()) 1189 Empty = I; 1190 } 1191 } 1192 if (Empty != OS->SectionCommands.end()) 1193 OS->SectionCommands.erase(Empty); 1194 1195 // If there are no other sections in the output section, remove it from the 1196 // output. 1197 if (OS->SectionCommands.empty()) 1198 OS->Live = false; 1199 } 1200 } 1201 1202 // Returns true if a symbol can be replaced at load-time by a symbol 1203 // with the same name defined in other ELF executable or DSO. 1204 static bool computeIsPreemptible(const Symbol &B) { 1205 assert(!B.isLocal()); 1206 // Only symbols that appear in dynsym can be preempted. 1207 if (!B.includeInDynsym()) 1208 return false; 1209 1210 // Only default visibility symbols can be preempted. 1211 if (B.Visibility != STV_DEFAULT) 1212 return false; 1213 1214 // At this point copy relocations have not been created yet, so any 1215 // symbol that is not defined locally is preemptible. 1216 if (!B.isDefined()) 1217 return true; 1218 1219 // If we have a dynamic list it specifies which local symbols are preemptible. 1220 if (Config->HasDynamicList) 1221 return false; 1222 1223 if (!Config->Shared) 1224 return false; 1225 1226 // -Bsymbolic means that definitions are not preempted. 1227 if (Config->Bsymbolic || (Config->BsymbolicFunctions && B.isFunc())) 1228 return false; 1229 return true; 1230 } 1231 1232 // Create output section objects and add them to OutputSections. 1233 template <class ELFT> void Writer<ELFT>::finalizeSections() { 1234 Out::DebugInfo = findSection(".debug_info"); 1235 Out::PreinitArray = findSection(".preinit_array"); 1236 Out::InitArray = findSection(".init_array"); 1237 Out::FiniArray = findSection(".fini_array"); 1238 1239 // The linker needs to define SECNAME_start, SECNAME_end and SECNAME_stop 1240 // symbols for sections, so that the runtime can get the start and end 1241 // addresses of each section by section name. Add such symbols. 1242 if (!Config->Relocatable) { 1243 addStartEndSymbols(); 1244 for (BaseCommand *Base : Script->SectionCommands) 1245 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1246 addStartStopSymbols(Sec); 1247 } 1248 1249 // Add _DYNAMIC symbol. Unlike GNU gold, our _DYNAMIC symbol has no type. 1250 // It should be okay as no one seems to care about the type. 1251 // Even the author of gold doesn't remember why gold behaves that way. 1252 // https://sourceware.org/ml/binutils/2002-03/msg00360.html 1253 if (InX::DynSymTab) 1254 Symtab->addRegular<ELFT>("_DYNAMIC", STV_HIDDEN, STT_NOTYPE, 0 /*Value*/, 1255 /*Size=*/0, STB_WEAK, InX::Dynamic, 1256 /*File=*/nullptr); 1257 1258 // Define __rel[a]_iplt_{start,end} symbols if needed. 1259 addRelIpltSymbols(); 1260 1261 // This responsible for splitting up .eh_frame section into 1262 // pieces. The relocation scan uses those pieces, so this has to be 1263 // earlier. 1264 applySynthetic({InX::EhFrame}, 1265 [](SyntheticSection *SS) { SS->finalizeContents(); }); 1266 1267 for (Symbol *S : Symtab->getSymbols()) 1268 S->IsPreemptible |= computeIsPreemptible(*S); 1269 1270 // Scan relocations. This must be done after every symbol is declared so that 1271 // we can correctly decide if a dynamic relocation is needed. 1272 if (!Config->Relocatable) 1273 forEachRelSec(scanRelocations<ELFT>); 1274 1275 if (InX::Plt && !InX::Plt->empty()) 1276 InX::Plt->addSymbols(); 1277 if (InX::Iplt && !InX::Iplt->empty()) 1278 InX::Iplt->addSymbols(); 1279 1280 // Now that we have defined all possible global symbols including linker- 1281 // synthesized ones. Visit all symbols to give the finishing touches. 1282 for (Symbol *Sym : Symtab->getSymbols()) { 1283 if (!includeInSymtab(*Sym)) 1284 continue; 1285 if (InX::SymTab) 1286 InX::SymTab->addSymbol(Sym); 1287 1288 if (InX::DynSymTab && Sym->includeInDynsym()) { 1289 InX::DynSymTab->addSymbol(Sym); 1290 if (auto *SS = dyn_cast<SharedSymbol>(Sym)) 1291 if (cast<SharedFile<ELFT>>(Sym->File)->IsNeeded) 1292 In<ELFT>::VerNeed->addSymbol(SS); 1293 } 1294 } 1295 1296 // Do not proceed if there was an undefined symbol. 1297 if (errorCount()) 1298 return; 1299 1300 addPredefinedSections(); 1301 removeUnusedSyntheticSections(); 1302 1303 sortSections(); 1304 Script->removeEmptyCommands(); 1305 1306 // Now that we have the final list, create a list of all the 1307 // OutputSections for convenience. 1308 for (BaseCommand *Base : Script->SectionCommands) 1309 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1310 OutputSections.push_back(Sec); 1311 1312 // Prefer command line supplied address over other constraints. 1313 for (OutputSection *Sec : OutputSections) { 1314 auto I = Config->SectionStartMap.find(Sec->Name); 1315 if (I != Config->SectionStartMap.end()) 1316 Sec->AddrExpr = [=] { return I->second; }; 1317 } 1318 1319 // This is a bit of a hack. A value of 0 means undef, so we set it 1320 // to 1 t make __ehdr_start defined. The section number is not 1321 // particularly relevant. 1322 Out::ElfHeader->SectionIndex = 1; 1323 1324 unsigned I = 1; 1325 for (OutputSection *Sec : OutputSections) { 1326 Sec->SectionIndex = I++; 1327 Sec->ShName = InX::ShStrTab->addString(Sec->Name); 1328 } 1329 1330 // Binary and relocatable output does not have PHDRS. 1331 // The headers have to be created before finalize as that can influence the 1332 // image base and the dynamic section on mips includes the image base. 1333 if (!Config->Relocatable && !Config->OFormatBinary) { 1334 Phdrs = Script->hasPhdrsCommands() ? Script->createPhdrs() : createPhdrs(); 1335 addPtArmExid(Phdrs); 1336 Out::ProgramHeaders->Size = sizeof(Elf_Phdr) * Phdrs.size(); 1337 } 1338 1339 // Some symbols are defined in term of program headers. Now that we 1340 // have the headers, we can find out which sections they point to. 1341 setReservedSymbolSections(); 1342 1343 // Dynamic section must be the last one in this list and dynamic 1344 // symbol table section (DynSymTab) must be the first one. 1345 applySynthetic({InX::DynSymTab, InX::Bss, InX::BssRelRo, 1346 InX::GnuHashTab, InX::HashTab, InX::SymTab, 1347 InX::ShStrTab, InX::StrTab, In<ELFT>::VerDef, 1348 InX::DynStrTab, InX::Got, InX::MipsGot, 1349 InX::IgotPlt, InX::GotPlt, In<ELFT>::RelaDyn, 1350 In<ELFT>::RelaIplt, In<ELFT>::RelaPlt, InX::Plt, 1351 InX::Iplt, InX::EhFrameHdr, In<ELFT>::VerSym, 1352 In<ELFT>::VerNeed, InX::Dynamic}, 1353 [](SyntheticSection *SS) { SS->finalizeContents(); }); 1354 1355 if (!Script->HasSectionsCommand && !Config->Relocatable) 1356 fixSectionAlignments(); 1357 1358 // Some architectures use small displacements for jump instructions. 1359 // It is linker's responsibility to create thunks containing long 1360 // jump instructions if jump targets are too far. Create thunks. 1361 if (Target->NeedsThunks || Config->AndroidPackDynRelocs) { 1362 ThunkCreator TC; 1363 bool Changed; 1364 do { 1365 Script->assignAddresses(); 1366 Changed = false; 1367 if (Target->NeedsThunks) 1368 Changed |= TC.createThunks(OutputSections); 1369 if (InX::MipsGot) 1370 InX::MipsGot->updateAllocSize(); 1371 Changed |= In<ELFT>::RelaDyn->updateAllocSize(); 1372 } while (Changed); 1373 } 1374 1375 // Fill other section headers. The dynamic table is finalized 1376 // at the end because some tags like RELSZ depend on result 1377 // of finalizing other sections. 1378 for (OutputSection *Sec : OutputSections) 1379 Sec->finalize<ELFT>(); 1380 1381 // createThunks may have added local symbols to the static symbol table 1382 applySynthetic({InX::SymTab}, 1383 [](SyntheticSection *SS) { SS->postThunkContents(); }); 1384 } 1385 1386 template <class ELFT> void Writer<ELFT>::addPredefinedSections() { 1387 // ARM ABI requires .ARM.exidx to be terminated by some piece of data. 1388 // We have the terminater synthetic section class. Add that at the end. 1389 OutputSection *Cmd = findSection(".ARM.exidx"); 1390 if (!Cmd || !Cmd->Live || Config->Relocatable) 1391 return; 1392 1393 auto *Sentinel = make<ARMExidxSentinelSection>(); 1394 Cmd->addSection(Sentinel); 1395 } 1396 1397 // The linker is expected to define SECNAME_start and SECNAME_end 1398 // symbols for a few sections. This function defines them. 1399 template <class ELFT> void Writer<ELFT>::addStartEndSymbols() { 1400 auto Define = [&](StringRef Start, StringRef End, OutputSection *OS) { 1401 // These symbols resolve to the image base if the section does not exist. 1402 // A special value -1 indicates end of the section. 1403 if (OS) { 1404 addOptionalRegular<ELFT>(Start, OS, 0); 1405 addOptionalRegular<ELFT>(End, OS, -1); 1406 } else { 1407 if (Config->Pic) 1408 OS = Out::ElfHeader; 1409 addOptionalRegular<ELFT>(Start, OS, 0); 1410 addOptionalRegular<ELFT>(End, OS, 0); 1411 } 1412 }; 1413 1414 Define("__preinit_array_start", "__preinit_array_end", Out::PreinitArray); 1415 Define("__init_array_start", "__init_array_end", Out::InitArray); 1416 Define("__fini_array_start", "__fini_array_end", Out::FiniArray); 1417 1418 if (OutputSection *Sec = findSection(".ARM.exidx")) 1419 Define("__exidx_start", "__exidx_end", Sec); 1420 } 1421 1422 // If a section name is valid as a C identifier (which is rare because of 1423 // the leading '.'), linkers are expected to define __start_<secname> and 1424 // __stop_<secname> symbols. They are at beginning and end of the section, 1425 // respectively. This is not requested by the ELF standard, but GNU ld and 1426 // gold provide the feature, and used by many programs. 1427 template <class ELFT> 1428 void Writer<ELFT>::addStartStopSymbols(OutputSection *Sec) { 1429 StringRef S = Sec->Name; 1430 if (!isValidCIdentifier(S)) 1431 return; 1432 addOptionalRegular<ELFT>(Saver.save("__start_" + S), Sec, 0, STV_DEFAULT); 1433 addOptionalRegular<ELFT>(Saver.save("__stop_" + S), Sec, -1, STV_DEFAULT); 1434 } 1435 1436 template <class ELFT> OutputSection *Writer<ELFT>::findSection(StringRef Name) { 1437 for (BaseCommand *Base : Script->SectionCommands) 1438 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1439 if (Sec->Name == Name) 1440 return Sec; 1441 return nullptr; 1442 } 1443 1444 static bool needsPtLoad(OutputSection *Sec) { 1445 if (!(Sec->Flags & SHF_ALLOC)) 1446 return false; 1447 1448 // Don't allocate VA space for TLS NOBITS sections. The PT_TLS PHDR is 1449 // responsible for allocating space for them, not the PT_LOAD that 1450 // contains the TLS initialization image. 1451 if (Sec->Flags & SHF_TLS && Sec->Type == SHT_NOBITS) 1452 return false; 1453 return true; 1454 } 1455 1456 // Linker scripts are responsible for aligning addresses. Unfortunately, most 1457 // linker scripts are designed for creating two PT_LOADs only, one RX and one 1458 // RW. This means that there is no alignment in the RO to RX transition and we 1459 // cannot create a PT_LOAD there. 1460 static uint64_t computeFlags(uint64_t Flags) { 1461 if (Config->Omagic) 1462 return PF_R | PF_W | PF_X; 1463 if (Config->SingleRoRx && !(Flags & PF_W)) 1464 return Flags | PF_X; 1465 return Flags; 1466 } 1467 1468 // Decide which program headers to create and which sections to include in each 1469 // one. 1470 template <class ELFT> std::vector<PhdrEntry *> Writer<ELFT>::createPhdrs() { 1471 std::vector<PhdrEntry *> Ret; 1472 auto AddHdr = [&](unsigned Type, unsigned Flags) -> PhdrEntry * { 1473 Ret.push_back(make<PhdrEntry>(Type, Flags)); 1474 return Ret.back(); 1475 }; 1476 1477 // The first phdr entry is PT_PHDR which describes the program header itself. 1478 AddHdr(PT_PHDR, PF_R)->add(Out::ProgramHeaders); 1479 1480 // PT_INTERP must be the second entry if exists. 1481 if (OutputSection *Cmd = findSection(".interp")) 1482 AddHdr(PT_INTERP, Cmd->getPhdrFlags())->add(Cmd); 1483 1484 // Add the first PT_LOAD segment for regular output sections. 1485 uint64_t Flags = computeFlags(PF_R); 1486 PhdrEntry *Load = AddHdr(PT_LOAD, Flags); 1487 1488 // Add the headers. We will remove them if they don't fit. 1489 Load->add(Out::ElfHeader); 1490 Load->add(Out::ProgramHeaders); 1491 1492 for (OutputSection *Sec : OutputSections) { 1493 if (!(Sec->Flags & SHF_ALLOC)) 1494 break; 1495 if (!needsPtLoad(Sec)) 1496 continue; 1497 1498 // Segments are contiguous memory regions that has the same attributes 1499 // (e.g. executable or writable). There is one phdr for each segment. 1500 // Therefore, we need to create a new phdr when the next section has 1501 // different flags or is loaded at a discontiguous address using AT linker 1502 // script command. 1503 uint64_t NewFlags = computeFlags(Sec->getPhdrFlags()); 1504 if (Sec->LMAExpr || Flags != NewFlags) { 1505 Load = AddHdr(PT_LOAD, NewFlags); 1506 Flags = NewFlags; 1507 } 1508 1509 Load->add(Sec); 1510 } 1511 1512 // Add a TLS segment if any. 1513 PhdrEntry *TlsHdr = make<PhdrEntry>(PT_TLS, PF_R); 1514 for (OutputSection *Sec : OutputSections) 1515 if (Sec->Flags & SHF_TLS) 1516 TlsHdr->add(Sec); 1517 if (TlsHdr->FirstSec) 1518 Ret.push_back(TlsHdr); 1519 1520 // Add an entry for .dynamic. 1521 if (InX::DynSymTab) 1522 AddHdr(PT_DYNAMIC, InX::Dynamic->getParent()->getPhdrFlags()) 1523 ->add(InX::Dynamic->getParent()); 1524 1525 // PT_GNU_RELRO includes all sections that should be marked as 1526 // read-only by dynamic linker after proccessing relocations. 1527 // Current dynamic loaders only support one PT_GNU_RELRO PHDR, give 1528 // an error message if more than one PT_GNU_RELRO PHDR is required. 1529 PhdrEntry *RelRo = make<PhdrEntry>(PT_GNU_RELRO, PF_R); 1530 bool InRelroPhdr = false; 1531 bool IsRelroFinished = false; 1532 for (OutputSection *Sec : OutputSections) { 1533 if (!needsPtLoad(Sec)) 1534 continue; 1535 if (isRelroSection(Sec)) { 1536 InRelroPhdr = true; 1537 if (!IsRelroFinished) 1538 RelRo->add(Sec); 1539 else 1540 error("section: " + Sec->Name + " is not contiguous with other relro" + 1541 " sections"); 1542 } else if (InRelroPhdr) { 1543 InRelroPhdr = false; 1544 IsRelroFinished = true; 1545 } 1546 } 1547 if (RelRo->FirstSec) 1548 Ret.push_back(RelRo); 1549 1550 // PT_GNU_EH_FRAME is a special section pointing on .eh_frame_hdr. 1551 if (!InX::EhFrame->empty() && InX::EhFrameHdr && InX::EhFrame->getParent() && 1552 InX::EhFrameHdr->getParent()) 1553 AddHdr(PT_GNU_EH_FRAME, InX::EhFrameHdr->getParent()->getPhdrFlags()) 1554 ->add(InX::EhFrameHdr->getParent()); 1555 1556 // PT_OPENBSD_RANDOMIZE is an OpenBSD-specific feature. That makes 1557 // the dynamic linker fill the segment with random data. 1558 if (OutputSection *Cmd = findSection(".openbsd.randomdata")) 1559 AddHdr(PT_OPENBSD_RANDOMIZE, Cmd->getPhdrFlags())->add(Cmd); 1560 1561 // PT_GNU_STACK is a special section to tell the loader to make the 1562 // pages for the stack non-executable. If you really want an executable 1563 // stack, you can pass -z execstack, but that's not recommended for 1564 // security reasons. 1565 unsigned Perm; 1566 if (Config->ZExecstack) 1567 Perm = PF_R | PF_W | PF_X; 1568 else 1569 Perm = PF_R | PF_W; 1570 AddHdr(PT_GNU_STACK, Perm)->p_memsz = Config->ZStackSize; 1571 1572 // PT_OPENBSD_WXNEEDED is a OpenBSD-specific header to mark the executable 1573 // is expected to perform W^X violations, such as calling mprotect(2) or 1574 // mmap(2) with PROT_WRITE | PROT_EXEC, which is prohibited by default on 1575 // OpenBSD. 1576 if (Config->ZWxneeded) 1577 AddHdr(PT_OPENBSD_WXNEEDED, PF_X); 1578 1579 // Create one PT_NOTE per a group of contiguous .note sections. 1580 PhdrEntry *Note = nullptr; 1581 for (OutputSection *Sec : OutputSections) { 1582 if (Sec->Type == SHT_NOTE) { 1583 if (!Note || Sec->LMAExpr) 1584 Note = AddHdr(PT_NOTE, PF_R); 1585 Note->add(Sec); 1586 } else { 1587 Note = nullptr; 1588 } 1589 } 1590 return Ret; 1591 } 1592 1593 template <class ELFT> 1594 void Writer<ELFT>::addPtArmExid(std::vector<PhdrEntry *> &Phdrs) { 1595 if (Config->EMachine != EM_ARM) 1596 return; 1597 auto I = llvm::find_if(OutputSections, [](OutputSection *Cmd) { 1598 return Cmd->Type == SHT_ARM_EXIDX; 1599 }); 1600 if (I == OutputSections.end()) 1601 return; 1602 1603 // PT_ARM_EXIDX is the ARM EHABI equivalent of PT_GNU_EH_FRAME 1604 PhdrEntry *ARMExidx = make<PhdrEntry>(PT_ARM_EXIDX, PF_R); 1605 ARMExidx->add(*I); 1606 Phdrs.push_back(ARMExidx); 1607 } 1608 1609 // The first section of each PT_LOAD, the first section in PT_GNU_RELRO and the 1610 // first section after PT_GNU_RELRO have to be page aligned so that the dynamic 1611 // linker can set the permissions. 1612 template <class ELFT> void Writer<ELFT>::fixSectionAlignments() { 1613 auto PageAlign = [](OutputSection *Cmd) { 1614 if (Cmd && !Cmd->AddrExpr) 1615 Cmd->AddrExpr = [=] { 1616 return alignTo(Script->getDot(), Config->MaxPageSize); 1617 }; 1618 }; 1619 1620 for (const PhdrEntry *P : Phdrs) 1621 if (P->p_type == PT_LOAD && P->FirstSec) 1622 PageAlign(P->FirstSec); 1623 1624 for (const PhdrEntry *P : Phdrs) { 1625 if (P->p_type != PT_GNU_RELRO) 1626 continue; 1627 if (P->FirstSec) 1628 PageAlign(P->FirstSec); 1629 // Find the first section after PT_GNU_RELRO. If it is in a PT_LOAD we 1630 // have to align it to a page. 1631 auto End = OutputSections.end(); 1632 auto I = std::find(OutputSections.begin(), End, P->LastSec); 1633 if (I == End || (I + 1) == End) 1634 continue; 1635 OutputSection *Cmd = (*(I + 1)); 1636 if (needsPtLoad(Cmd)) 1637 PageAlign(Cmd); 1638 } 1639 } 1640 1641 // Adjusts the file alignment for a given output section and returns 1642 // its new file offset. The file offset must be the same with its 1643 // virtual address (modulo the page size) so that the loader can load 1644 // executables without any address adjustment. 1645 static uint64_t getFileAlignment(uint64_t Off, OutputSection *Cmd) { 1646 // If the section is not in a PT_LOAD, we just have to align it. 1647 if (!Cmd->PtLoad) 1648 return alignTo(Off, Cmd->Alignment); 1649 1650 OutputSection *First = Cmd->PtLoad->FirstSec; 1651 // The first section in a PT_LOAD has to have congruent offset and address 1652 // module the page size. 1653 if (Cmd == First) 1654 return alignTo(Off, std::max<uint64_t>(Cmd->Alignment, Config->MaxPageSize), 1655 Cmd->Addr); 1656 1657 // If two sections share the same PT_LOAD the file offset is calculated 1658 // using this formula: Off2 = Off1 + (VA2 - VA1). 1659 return First->Offset + Cmd->Addr - First->Addr; 1660 } 1661 1662 static uint64_t setOffset(OutputSection *Cmd, uint64_t Off) { 1663 if (Cmd->Type == SHT_NOBITS) { 1664 Cmd->Offset = Off; 1665 return Off; 1666 } 1667 1668 Off = getFileAlignment(Off, Cmd); 1669 Cmd->Offset = Off; 1670 return Off + Cmd->Size; 1671 } 1672 1673 template <class ELFT> void Writer<ELFT>::assignFileOffsetsBinary() { 1674 uint64_t Off = 0; 1675 for (OutputSection *Sec : OutputSections) 1676 if (Sec->Flags & SHF_ALLOC) 1677 Off = setOffset(Sec, Off); 1678 FileSize = alignTo(Off, Config->Wordsize); 1679 } 1680 1681 // Assign file offsets to output sections. 1682 template <class ELFT> void Writer<ELFT>::assignFileOffsets() { 1683 uint64_t Off = 0; 1684 Off = setOffset(Out::ElfHeader, Off); 1685 Off = setOffset(Out::ProgramHeaders, Off); 1686 1687 PhdrEntry *LastRX = nullptr; 1688 for (PhdrEntry *P : Phdrs) 1689 if (P->p_type == PT_LOAD && (P->p_flags & PF_X)) 1690 LastRX = P; 1691 1692 for (OutputSection *Sec : OutputSections) { 1693 Off = setOffset(Sec, Off); 1694 if (Script->HasSectionsCommand) 1695 continue; 1696 // If this is a last section of the last executable segment and that 1697 // segment is the last loadable segment, align the offset of the 1698 // following section to avoid loading non-segments parts of the file. 1699 if (LastRX && LastRX->LastSec == Sec) 1700 Off = alignTo(Off, Target->PageSize); 1701 } 1702 1703 SectionHeaderOff = alignTo(Off, Config->Wordsize); 1704 FileSize = SectionHeaderOff + (OutputSections.size() + 1) * sizeof(Elf_Shdr); 1705 } 1706 1707 // Finalize the program headers. We call this function after we assign 1708 // file offsets and VAs to all sections. 1709 template <class ELFT> void Writer<ELFT>::setPhdrs() { 1710 for (PhdrEntry *P : Phdrs) { 1711 OutputSection *First = P->FirstSec; 1712 OutputSection *Last = P->LastSec; 1713 if (First) { 1714 P->p_filesz = Last->Offset - First->Offset; 1715 if (Last->Type != SHT_NOBITS) 1716 P->p_filesz += Last->Size; 1717 P->p_memsz = Last->Addr + Last->Size - First->Addr; 1718 P->p_offset = First->Offset; 1719 P->p_vaddr = First->Addr; 1720 if (!P->HasLMA) 1721 P->p_paddr = First->getLMA(); 1722 } 1723 if (P->p_type == PT_LOAD) 1724 P->p_align = std::max<uint64_t>(P->p_align, Config->MaxPageSize); 1725 else if (P->p_type == PT_GNU_RELRO) { 1726 P->p_align = 1; 1727 // The glibc dynamic loader rounds the size down, so we need to round up 1728 // to protect the last page. This is a no-op on FreeBSD which always 1729 // rounds up. 1730 P->p_memsz = alignTo(P->p_memsz, Target->PageSize); 1731 } 1732 1733 // The TLS pointer goes after PT_TLS. At least glibc will align it, 1734 // so round up the size to make sure the offsets are correct. 1735 if (P->p_type == PT_TLS) { 1736 Out::TlsPhdr = P; 1737 if (P->p_memsz) 1738 P->p_memsz = alignTo(P->p_memsz, P->p_align); 1739 } 1740 } 1741 } 1742 1743 // The entry point address is chosen in the following ways. 1744 // 1745 // 1. the '-e' entry command-line option; 1746 // 2. the ENTRY(symbol) command in a linker control script; 1747 // 3. the value of the symbol _start, if present; 1748 // 4. the number represented by the entry symbol, if it is a number; 1749 // 5. the address of the first byte of the .text section, if present; 1750 // 6. the address 0. 1751 template <class ELFT> uint64_t Writer<ELFT>::getEntryAddr() { 1752 // Case 1, 2 or 3 1753 if (Symbol *B = Symtab->find(Config->Entry)) 1754 return B->getVA(); 1755 1756 // Case 4 1757 uint64_t Addr; 1758 if (to_integer(Config->Entry, Addr)) 1759 return Addr; 1760 1761 // Case 5 1762 if (OutputSection *Sec = findSection(".text")) { 1763 if (Config->WarnMissingEntry) 1764 warn("cannot find entry symbol " + Config->Entry + "; defaulting to 0x" + 1765 utohexstr(Sec->Addr)); 1766 return Sec->Addr; 1767 } 1768 1769 // Case 6 1770 if (Config->WarnMissingEntry) 1771 warn("cannot find entry symbol " + Config->Entry + 1772 "; not setting start address"); 1773 return 0; 1774 } 1775 1776 static uint16_t getELFType() { 1777 if (Config->Pic) 1778 return ET_DYN; 1779 if (Config->Relocatable) 1780 return ET_REL; 1781 return ET_EXEC; 1782 } 1783 1784 template <class ELFT> void Writer<ELFT>::writeHeader() { 1785 uint8_t *Buf = Buffer->getBufferStart(); 1786 memcpy(Buf, "\177ELF", 4); 1787 1788 // Write the ELF header. 1789 auto *EHdr = reinterpret_cast<Elf_Ehdr *>(Buf); 1790 EHdr->e_ident[EI_CLASS] = Config->Is64 ? ELFCLASS64 : ELFCLASS32; 1791 EHdr->e_ident[EI_DATA] = Config->IsLE ? ELFDATA2LSB : ELFDATA2MSB; 1792 EHdr->e_ident[EI_VERSION] = EV_CURRENT; 1793 EHdr->e_ident[EI_OSABI] = Config->OSABI; 1794 EHdr->e_type = getELFType(); 1795 EHdr->e_machine = Config->EMachine; 1796 EHdr->e_version = EV_CURRENT; 1797 EHdr->e_entry = getEntryAddr(); 1798 EHdr->e_shoff = SectionHeaderOff; 1799 EHdr->e_flags = Config->EFlags; 1800 EHdr->e_ehsize = sizeof(Elf_Ehdr); 1801 EHdr->e_phnum = Phdrs.size(); 1802 EHdr->e_shentsize = sizeof(Elf_Shdr); 1803 EHdr->e_shnum = OutputSections.size() + 1; 1804 EHdr->e_shstrndx = InX::ShStrTab->getParent()->SectionIndex; 1805 1806 if (!Config->Relocatable) { 1807 EHdr->e_phoff = sizeof(Elf_Ehdr); 1808 EHdr->e_phentsize = sizeof(Elf_Phdr); 1809 } 1810 1811 // Write the program header table. 1812 auto *HBuf = reinterpret_cast<Elf_Phdr *>(Buf + EHdr->e_phoff); 1813 for (PhdrEntry *P : Phdrs) { 1814 HBuf->p_type = P->p_type; 1815 HBuf->p_flags = P->p_flags; 1816 HBuf->p_offset = P->p_offset; 1817 HBuf->p_vaddr = P->p_vaddr; 1818 HBuf->p_paddr = P->p_paddr; 1819 HBuf->p_filesz = P->p_filesz; 1820 HBuf->p_memsz = P->p_memsz; 1821 HBuf->p_align = P->p_align; 1822 ++HBuf; 1823 } 1824 1825 // Write the section header table. Note that the first table entry is null. 1826 auto *SHdrs = reinterpret_cast<Elf_Shdr *>(Buf + EHdr->e_shoff); 1827 for (OutputSection *Sec : OutputSections) 1828 Sec->writeHeaderTo<ELFT>(++SHdrs); 1829 } 1830 1831 // Open a result file. 1832 template <class ELFT> void Writer<ELFT>::openFile() { 1833 if (!Config->Is64 && FileSize > UINT32_MAX) { 1834 error("output file too large: " + Twine(FileSize) + " bytes"); 1835 return; 1836 } 1837 1838 unlinkAsync(Config->OutputFile); 1839 Expected<std::unique_ptr<FileOutputBuffer>> BufferOrErr = 1840 FileOutputBuffer::create(Config->OutputFile, FileSize, 1841 FileOutputBuffer::F_executable); 1842 1843 if (!BufferOrErr) 1844 error("failed to open " + Config->OutputFile + ": " + 1845 llvm::toString(BufferOrErr.takeError())); 1846 else 1847 Buffer = std::move(*BufferOrErr); 1848 } 1849 1850 template <class ELFT> void Writer<ELFT>::writeSectionsBinary() { 1851 uint8_t *Buf = Buffer->getBufferStart(); 1852 for (OutputSection *Sec : OutputSections) 1853 if (Sec->Flags & SHF_ALLOC) 1854 Sec->writeTo<ELFT>(Buf + Sec->Offset); 1855 } 1856 1857 static void fillTrap(uint8_t *I, uint8_t *End) { 1858 for (; I + 4 <= End; I += 4) 1859 memcpy(I, &Target->TrapInstr, 4); 1860 } 1861 1862 // Fill the last page of executable segments with trap instructions 1863 // instead of leaving them as zero. Even though it is not required by any 1864 // standard, it is in general a good thing to do for security reasons. 1865 // 1866 // We'll leave other pages in segments as-is because the rest will be 1867 // overwritten by output sections. 1868 template <class ELFT> void Writer<ELFT>::writeTrapInstr() { 1869 if (Script->HasSectionsCommand) 1870 return; 1871 1872 // Fill the last page. 1873 uint8_t *Buf = Buffer->getBufferStart(); 1874 for (PhdrEntry *P : Phdrs) 1875 if (P->p_type == PT_LOAD && (P->p_flags & PF_X)) 1876 fillTrap(Buf + alignDown(P->p_offset + P->p_filesz, Target->PageSize), 1877 Buf + alignTo(P->p_offset + P->p_filesz, Target->PageSize)); 1878 1879 // Round up the file size of the last segment to the page boundary iff it is 1880 // an executable segment to ensure that other tools don't accidentally 1881 // trim the instruction padding (e.g. when stripping the file). 1882 PhdrEntry *Last = nullptr; 1883 for (PhdrEntry *P : Phdrs) 1884 if (P->p_type == PT_LOAD) 1885 Last = P; 1886 1887 if (Last && (Last->p_flags & PF_X)) 1888 Last->p_memsz = Last->p_filesz = alignTo(Last->p_filesz, Target->PageSize); 1889 } 1890 1891 // Write section contents to a mmap'ed file. 1892 template <class ELFT> void Writer<ELFT>::writeSections() { 1893 uint8_t *Buf = Buffer->getBufferStart(); 1894 1895 // PPC64 needs to process relocations in the .opd section 1896 // before processing relocations in code-containing sections. 1897 if (auto *OpdCmd = findSection(".opd")) { 1898 Out::Opd = OpdCmd; 1899 Out::OpdBuf = Buf + Out::Opd->Offset; 1900 OpdCmd->template writeTo<ELFT>(Buf + Out::Opd->Offset); 1901 } 1902 1903 OutputSection *EhFrameHdr = nullptr; 1904 if (InX::EhFrameHdr && !InX::EhFrameHdr->empty()) 1905 EhFrameHdr = InX::EhFrameHdr->getParent(); 1906 1907 // In -r or -emit-relocs mode, write the relocation sections first as in 1908 // ELf_Rel targets we might find out that we need to modify the relocated 1909 // section while doing it. 1910 for (OutputSection *Sec : OutputSections) 1911 if (Sec->Type == SHT_REL || Sec->Type == SHT_RELA) 1912 Sec->writeTo<ELFT>(Buf + Sec->Offset); 1913 1914 for (OutputSection *Sec : OutputSections) 1915 if (Sec != Out::Opd && Sec != EhFrameHdr && Sec->Type != SHT_REL && 1916 Sec->Type != SHT_RELA) 1917 Sec->writeTo<ELFT>(Buf + Sec->Offset); 1918 1919 // The .eh_frame_hdr depends on .eh_frame section contents, therefore 1920 // it should be written after .eh_frame is written. 1921 if (EhFrameHdr) 1922 EhFrameHdr->writeTo<ELFT>(Buf + EhFrameHdr->Offset); 1923 } 1924 1925 template <class ELFT> void Writer<ELFT>::writeBuildId() { 1926 if (!InX::BuildId || !InX::BuildId->getParent()) 1927 return; 1928 1929 // Compute a hash of all sections of the output file. 1930 uint8_t *Start = Buffer->getBufferStart(); 1931 uint8_t *End = Start + FileSize; 1932 InX::BuildId->writeBuildId({Start, End}); 1933 } 1934 1935 template void elf::writeResult<ELF32LE>(); 1936 template void elf::writeResult<ELF32BE>(); 1937 template void elf::writeResult<ELF64LE>(); 1938 template void elf::writeResult<ELF64BE>(); 1939