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 "AArch64ErrataFix.h" 12 #include "Config.h" 13 #include "Filesystem.h" 14 #include "LinkerScript.h" 15 #include "MapFile.h" 16 #include "OutputSections.h" 17 #include "Relocations.h" 18 #include "Strings.h" 19 #include "SymbolTable.h" 20 #include "Symbols.h" 21 #include "SyntheticSections.h" 22 #include "Target.h" 23 #include "lld/Common/Memory.h" 24 #include "lld/Common/Threads.h" 25 #include "llvm/ADT/StringMap.h" 26 #include "llvm/ADT/StringSwitch.h" 27 #include <climits> 28 29 using namespace llvm; 30 using namespace llvm::ELF; 31 using namespace llvm::object; 32 using namespace llvm::support; 33 using namespace llvm::support::endian; 34 35 using namespace lld; 36 using namespace lld::elf; 37 38 namespace { 39 // The writer writes a SymbolTable result to a file. 40 template <class ELFT> class Writer { 41 public: 42 Writer() : Buffer(errorHandler().OutputBuffer) {} 43 typedef typename ELFT::Shdr Elf_Shdr; 44 typedef typename ELFT::Ehdr Elf_Ehdr; 45 typedef typename ELFT::Phdr Elf_Phdr; 46 47 void run(); 48 49 private: 50 void copyLocalSymbols(); 51 void addSectionSymbols(); 52 void forEachRelSec(std::function<void(InputSectionBase &)> Fn); 53 void sortSections(); 54 void resolveShfLinkOrder(); 55 void sortInputSections(); 56 void finalizeSections(); 57 void addPredefinedSections(); 58 void setReservedSymbolSections(); 59 60 std::vector<PhdrEntry *> createPhdrs(); 61 void removeEmptyPTLoad(); 62 void addPtArmExid(std::vector<PhdrEntry *> &Phdrs); 63 void assignFileOffsets(); 64 void assignFileOffsetsBinary(); 65 void setPhdrs(); 66 void fixSectionAlignments(); 67 void openFile(); 68 void writeTrapInstr(); 69 void writeHeader(); 70 void writeSections(); 71 void writeSectionsBinary(); 72 void writeBuildId(); 73 74 std::unique_ptr<FileOutputBuffer> &Buffer; 75 76 void addRelIpltSymbols(); 77 void addStartEndSymbols(); 78 void addStartStopSymbols(OutputSection *Sec); 79 uint64_t getEntryAddr(); 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 template <class ELFT> 161 static Defined *addOptionalRegular(StringRef Name, SectionBase *Sec, 162 uint64_t Val, uint8_t StOther = STV_HIDDEN, 163 uint8_t Binding = STB_GLOBAL) { 164 Symbol *S = Symtab->find(Name); 165 if (!S || S->isDefined()) 166 return nullptr; 167 Symbol *Sym = Symtab->addRegular<ELFT>(Name, StOther, STT_NOTYPE, Val, 168 /*Size=*/0, Binding, Sec, 169 /*File=*/nullptr); 170 return cast<Defined>(Sym); 171 } 172 173 // The linker is expected to define some symbols depending on 174 // the linking result. This function defines such symbols. 175 template <class ELFT> void elf::addReservedSymbols() { 176 if (Config->EMachine == EM_MIPS) { 177 // Define _gp for MIPS. st_value of _gp symbol will be updated by Writer 178 // so that it points to an absolute address which by default is relative 179 // to GOT. Default offset is 0x7ff0. 180 // See "Global Data Symbols" in Chapter 6 in the following document: 181 // ftp://www.linux-mips.org/pub/linux/mips/doc/ABI/mipsabi.pdf 182 ElfSym::MipsGp = Symtab->addAbsolute<ELFT>("_gp", STV_HIDDEN, STB_GLOBAL); 183 184 // On MIPS O32 ABI, _gp_disp is a magic symbol designates offset between 185 // start of function and 'gp' pointer into GOT. 186 if (Symtab->find("_gp_disp")) 187 ElfSym::MipsGpDisp = 188 Symtab->addAbsolute<ELFT>("_gp_disp", STV_HIDDEN, STB_GLOBAL); 189 190 // The __gnu_local_gp is a magic symbol equal to the current value of 'gp' 191 // pointer. This symbol is used in the code generated by .cpload pseudo-op 192 // in case of using -mno-shared option. 193 // https://sourceware.org/ml/binutils/2004-12/msg00094.html 194 if (Symtab->find("__gnu_local_gp")) 195 ElfSym::MipsLocalGp = 196 Symtab->addAbsolute<ELFT>("__gnu_local_gp", STV_HIDDEN, STB_GLOBAL); 197 } 198 199 ElfSym::GlobalOffsetTable = addOptionalRegular<ELFT>( 200 "_GLOBAL_OFFSET_TABLE_", Out::ElfHeader, Target->GotBaseSymOff); 201 202 // __ehdr_start is the location of ELF file headers. Note that we define 203 // this symbol unconditionally even when using a linker script, which 204 // differs from the behavior implemented by GNU linker which only define 205 // this symbol if ELF headers are in the memory mapped segment. 206 // __executable_start is not documented, but the expectation of at 207 // least the android libc is that it points to the elf header too. 208 // __dso_handle symbol is passed to cxa_finalize as a marker to identify 209 // each DSO. The address of the symbol doesn't matter as long as they are 210 // different in different DSOs, so we chose the start address of the DSO. 211 for (const char *Name : 212 {"__ehdr_start", "__executable_start", "__dso_handle"}) 213 addOptionalRegular<ELFT>(Name, Out::ElfHeader, 0, STV_HIDDEN); 214 215 // If linker script do layout we do not need to create any standart symbols. 216 if (Script->HasSectionsCommand) 217 return; 218 219 auto Add = [](StringRef S, int64_t Pos) { 220 return addOptionalRegular<ELFT>(S, Out::ElfHeader, Pos, STV_DEFAULT); 221 }; 222 223 ElfSym::Bss = Add("__bss_start", 0); 224 ElfSym::End1 = Add("end", -1); 225 ElfSym::End2 = Add("_end", -1); 226 ElfSym::Etext1 = Add("etext", -1); 227 ElfSym::Etext2 = Add("_etext", -1); 228 ElfSym::Edata1 = Add("edata", -1); 229 ElfSym::Edata2 = Add("_edata", -1); 230 } 231 232 static OutputSection *findSection(StringRef Name) { 233 for (BaseCommand *Base : Script->SectionCommands) 234 if (auto *Sec = dyn_cast<OutputSection>(Base)) 235 if (Sec->Name == Name) 236 return Sec; 237 return nullptr; 238 } 239 240 // Initialize Out members. 241 template <class ELFT> static void createSyntheticSections() { 242 // Initialize all pointers with NULL. This is needed because 243 // you can call lld::elf::main more than once as a library. 244 memset(&Out::First, 0, sizeof(Out)); 245 246 auto Add = [](InputSectionBase *Sec) { InputSections.push_back(Sec); }; 247 248 InX::DynStrTab = make<StringTableSection>(".dynstr", true); 249 InX::Dynamic = make<DynamicSection<ELFT>>(); 250 if (Config->AndroidPackDynRelocs) { 251 InX::RelaDyn = make<AndroidPackedRelocationSection<ELFT>>( 252 Config->IsRela ? ".rela.dyn" : ".rel.dyn"); 253 } else { 254 InX::RelaDyn = make<RelocationSection<ELFT>>( 255 Config->IsRela ? ".rela.dyn" : ".rel.dyn", Config->ZCombreloc); 256 } 257 InX::ShStrTab = make<StringTableSection>(".shstrtab", false); 258 259 Out::ProgramHeaders = make<OutputSection>("", 0, SHF_ALLOC); 260 Out::ProgramHeaders->Alignment = Config->Wordsize; 261 262 if (needsInterpSection()) { 263 InX::Interp = createInterpSection(); 264 Add(InX::Interp); 265 } else { 266 InX::Interp = nullptr; 267 } 268 269 if (Config->Strip != StripPolicy::All) { 270 InX::StrTab = make<StringTableSection>(".strtab", false); 271 InX::SymTab = make<SymbolTableSection<ELFT>>(*InX::StrTab); 272 } 273 274 if (Config->BuildId != BuildIdKind::None) { 275 InX::BuildId = make<BuildIdSection>(); 276 Add(InX::BuildId); 277 } 278 279 InX::Bss = make<BssSection>(".bss", 0, 1); 280 Add(InX::Bss); 281 282 // If there is a SECTIONS command and a .data.rel.ro section name use name 283 // .data.rel.ro.bss so that we match in the .data.rel.ro output section. 284 // This makes sure our relro is contiguous. 285 bool HasDataRelRo = 286 Script->HasSectionsCommand && findSection(".data.rel.ro"); 287 InX::BssRelRo = make<BssSection>( 288 HasDataRelRo ? ".data.rel.ro.bss" : ".bss.rel.ro", 0, 1); 289 Add(InX::BssRelRo); 290 291 // Add MIPS-specific sections. 292 if (Config->EMachine == EM_MIPS) { 293 if (!Config->Shared && Config->HasDynSymTab) { 294 InX::MipsRldMap = make<MipsRldMapSection>(); 295 Add(InX::MipsRldMap); 296 } 297 if (auto *Sec = MipsAbiFlagsSection<ELFT>::create()) 298 Add(Sec); 299 if (auto *Sec = MipsOptionsSection<ELFT>::create()) 300 Add(Sec); 301 if (auto *Sec = MipsReginfoSection<ELFT>::create()) 302 Add(Sec); 303 } 304 305 if (Config->HasDynSymTab) { 306 InX::DynSymTab = make<SymbolTableSection<ELFT>>(*InX::DynStrTab); 307 Add(InX::DynSymTab); 308 309 In<ELFT>::VerSym = make<VersionTableSection<ELFT>>(); 310 Add(In<ELFT>::VerSym); 311 312 if (!Config->VersionDefinitions.empty()) { 313 In<ELFT>::VerDef = make<VersionDefinitionSection<ELFT>>(); 314 Add(In<ELFT>::VerDef); 315 } 316 317 In<ELFT>::VerNeed = make<VersionNeedSection<ELFT>>(); 318 Add(In<ELFT>::VerNeed); 319 320 if (Config->GnuHash) { 321 InX::GnuHashTab = make<GnuHashTableSection>(); 322 Add(InX::GnuHashTab); 323 } 324 325 if (Config->SysvHash) { 326 InX::HashTab = make<HashTableSection>(); 327 Add(InX::HashTab); 328 } 329 330 Add(InX::Dynamic); 331 Add(InX::DynStrTab); 332 Add(InX::RelaDyn); 333 } 334 335 // Add .got. MIPS' .got is so different from the other archs, 336 // it has its own class. 337 if (Config->EMachine == EM_MIPS) { 338 InX::MipsGot = make<MipsGotSection>(); 339 Add(InX::MipsGot); 340 } else { 341 InX::Got = make<GotSection>(); 342 Add(InX::Got); 343 } 344 345 InX::GotPlt = make<GotPltSection>(); 346 Add(InX::GotPlt); 347 InX::IgotPlt = make<IgotPltSection>(); 348 Add(InX::IgotPlt); 349 350 if (Config->GdbIndex) { 351 InX::GdbIndex = createGdbIndex<ELFT>(); 352 Add(InX::GdbIndex); 353 } 354 355 // We always need to add rel[a].plt to output if it has entries. 356 // Even for static linking it can contain R_[*]_IRELATIVE relocations. 357 InX::RelaPlt = make<RelocationSection<ELFT>>( 358 Config->IsRela ? ".rela.plt" : ".rel.plt", false /*Sort*/); 359 Add(InX::RelaPlt); 360 361 // The RelaIplt immediately follows .rel.plt (.rel.dyn for ARM) to ensure 362 // that the IRelative relocations are processed last by the dynamic loader. 363 // We cannot place the iplt section in .rel.dyn when Android relocation 364 // packing is enabled because that would cause a section type mismatch. 365 // However, because the Android dynamic loader reads .rel.plt after .rel.dyn, 366 // we can get the desired behaviour by placing the iplt section in .rel.plt. 367 InX::RelaIplt = make<RelocationSection<ELFT>>( 368 (Config->EMachine == EM_ARM && !Config->AndroidPackDynRelocs) 369 ? ".rel.dyn" 370 : InX::RelaPlt->Name, 371 false /*Sort*/); 372 Add(InX::RelaIplt); 373 374 InX::Plt = make<PltSection>(Target->PltHeaderSize); 375 Add(InX::Plt); 376 InX::Iplt = make<PltSection>(0); 377 Add(InX::Iplt); 378 379 if (!Config->Relocatable) { 380 if (Config->EhFrameHdr) { 381 InX::EhFrameHdr = make<EhFrameHeader>(); 382 Add(InX::EhFrameHdr); 383 } 384 InX::EhFrame = make<EhFrameSection>(); 385 Add(InX::EhFrame); 386 } 387 388 if (InX::SymTab) 389 Add(InX::SymTab); 390 Add(InX::ShStrTab); 391 if (InX::StrTab) 392 Add(InX::StrTab); 393 } 394 395 // The main function of the writer. 396 template <class ELFT> void Writer<ELFT>::run() { 397 // Create linker-synthesized sections such as .got or .plt. 398 // Such sections are of type input section. 399 createSyntheticSections<ELFT>(); 400 401 if (!Config->Relocatable) 402 combineEhFrameSections<ELFT>(); 403 404 // We want to process linker script commands. When SECTIONS command 405 // is given we let it create sections. 406 Script->processSectionCommands(); 407 408 // Linker scripts controls how input sections are assigned to output sections. 409 // Input sections that were not handled by scripts are called "orphans", and 410 // they are assigned to output sections by the default rule. Process that. 411 Script->addOrphanSections(); 412 413 if (Config->Discard != DiscardPolicy::All) 414 copyLocalSymbols(); 415 416 if (Config->CopyRelocs) 417 addSectionSymbols(); 418 419 // Now that we have a complete set of output sections. This function 420 // completes section contents. For example, we need to add strings 421 // to the string table, and add entries to .got and .plt. 422 // finalizeSections does that. 423 finalizeSections(); 424 if (errorCount()) 425 return; 426 427 // If -compressed-debug-sections is specified, we need to compress 428 // .debug_* sections. Do it right now because it changes the size of 429 // output sections. 430 parallelForEach(OutputSections, 431 [](OutputSection *Sec) { Sec->maybeCompress<ELFT>(); }); 432 433 Script->assignAddresses(); 434 Script->allocateHeaders(Phdrs); 435 436 // Remove empty PT_LOAD to avoid causing the dynamic linker to try to mmap a 437 // 0 sized region. This has to be done late since only after assignAddresses 438 // we know the size of the sections. 439 removeEmptyPTLoad(); 440 441 if (!Config->OFormatBinary) 442 assignFileOffsets(); 443 else 444 assignFileOffsetsBinary(); 445 446 setPhdrs(); 447 448 if (Config->Relocatable) { 449 for (OutputSection *Sec : OutputSections) 450 Sec->Addr = 0; 451 } 452 453 // It does not make sense try to open the file if we have error already. 454 if (errorCount()) 455 return; 456 // Write the result down to a file. 457 openFile(); 458 if (errorCount()) 459 return; 460 461 if (!Config->OFormatBinary) { 462 writeTrapInstr(); 463 writeHeader(); 464 writeSections(); 465 } else { 466 writeSectionsBinary(); 467 } 468 469 // Backfill .note.gnu.build-id section content. This is done at last 470 // because the content is usually a hash value of the entire output file. 471 writeBuildId(); 472 if (errorCount()) 473 return; 474 475 // Handle -Map option. 476 writeMapFile(); 477 if (errorCount()) 478 return; 479 480 if (auto E = Buffer->commit()) 481 error("failed to write to the output file: " + toString(std::move(E))); 482 } 483 484 static bool shouldKeepInSymtab(SectionBase *Sec, StringRef SymName, 485 const Symbol &B) { 486 if (B.isFile() || B.isSection()) 487 return false; 488 489 // If sym references a section in a discarded group, don't keep it. 490 if (Sec == &InputSection::Discarded) 491 return false; 492 493 if (Config->Discard == DiscardPolicy::None) 494 return true; 495 496 // In ELF assembly .L symbols are normally discarded by the assembler. 497 // If the assembler fails to do so, the linker discards them if 498 // * --discard-locals is used. 499 // * The symbol is in a SHF_MERGE section, which is normally the reason for 500 // the assembler keeping the .L symbol. 501 if (!SymName.startswith(".L") && !SymName.empty()) 502 return true; 503 504 if (Config->Discard == DiscardPolicy::Locals) 505 return false; 506 507 return !Sec || !(Sec->Flags & SHF_MERGE); 508 } 509 510 static bool includeInSymtab(const Symbol &B) { 511 if (!B.isLocal() && !B.IsUsedInRegularObj) 512 return false; 513 514 if (auto *D = dyn_cast<Defined>(&B)) { 515 // Always include absolute symbols. 516 SectionBase *Sec = D->Section; 517 if (!Sec) 518 return true; 519 if (auto *IS = dyn_cast<InputSectionBase>(Sec)) { 520 Sec = IS->Repl; 521 IS = cast<InputSectionBase>(Sec); 522 // Exclude symbols pointing to garbage-collected sections. 523 if (!IS->Live) 524 return false; 525 } 526 if (auto *S = dyn_cast<MergeInputSection>(Sec)) 527 if (!S->getSectionPiece(D->Value)->Live) 528 return false; 529 return true; 530 } 531 return B.Used; 532 } 533 534 // Local symbols are not in the linker's symbol table. This function scans 535 // each object file's symbol table to copy local symbols to the output. 536 template <class ELFT> void Writer<ELFT>::copyLocalSymbols() { 537 if (!InX::SymTab) 538 return; 539 for (InputFile *File : ObjectFiles) { 540 ObjFile<ELFT> *F = cast<ObjFile<ELFT>>(File); 541 for (Symbol *B : F->getLocalSymbols()) { 542 if (!B->isLocal()) 543 fatal(toString(F) + 544 ": broken object: getLocalSymbols returns a non-local symbol"); 545 auto *DR = dyn_cast<Defined>(B); 546 547 // No reason to keep local undefined symbol in symtab. 548 if (!DR) 549 continue; 550 if (!includeInSymtab(*B)) 551 continue; 552 553 SectionBase *Sec = DR->Section; 554 if (!shouldKeepInSymtab(Sec, B->getName(), *B)) 555 continue; 556 InX::SymTab->addSymbol(B); 557 } 558 } 559 } 560 561 template <class ELFT> void Writer<ELFT>::addSectionSymbols() { 562 // Create a section symbol for each output section so that we can represent 563 // relocations that point to the section. If we know that no relocation is 564 // referring to a section (that happens if the section is a synthetic one), we 565 // don't create a section symbol for that section. 566 for (BaseCommand *Base : Script->SectionCommands) { 567 auto *Sec = dyn_cast<OutputSection>(Base); 568 if (!Sec) 569 continue; 570 auto I = llvm::find_if(Sec->SectionCommands, [](BaseCommand *Base) { 571 if (auto *ISD = dyn_cast<InputSectionDescription>(Base)) 572 return !ISD->Sections.empty(); 573 return false; 574 }); 575 if (I == Sec->SectionCommands.end()) 576 continue; 577 InputSection *IS = cast<InputSectionDescription>(*I)->Sections[0]; 578 579 // Relocations are not using REL[A] section symbols. 580 if (IS->Type == SHT_REL || IS->Type == SHT_RELA) 581 continue; 582 583 // Unlike other synthetic sections, mergeable output sections contain data 584 // copied from input sections, and there may be a relocation pointing to its 585 // contents if -r or -emit-reloc are given. 586 if (isa<SyntheticSection>(IS) && !(IS->Flags & SHF_MERGE)) 587 continue; 588 589 auto *Sym = 590 make<Defined>(IS->File, "", STB_LOCAL, /*StOther=*/0, STT_SECTION, 591 /*Value=*/0, /*Size=*/0, IS); 592 InX::SymTab->addSymbol(Sym); 593 } 594 } 595 596 // Today's loaders have a feature to make segments read-only after 597 // processing dynamic relocations to enhance security. PT_GNU_RELRO 598 // is defined for that. 599 // 600 // This function returns true if a section needs to be put into a 601 // PT_GNU_RELRO segment. 602 static bool isRelroSection(const OutputSection *Sec) { 603 if (!Config->ZRelro) 604 return false; 605 606 uint64_t Flags = Sec->Flags; 607 608 // Non-allocatable or non-writable sections don't need RELRO because 609 // they are not writable or not even mapped to memory in the first place. 610 // RELRO is for sections that are essentially read-only but need to 611 // be writable only at process startup to allow dynamic linker to 612 // apply relocations. 613 if (!(Flags & SHF_ALLOC) || !(Flags & SHF_WRITE)) 614 return false; 615 616 // Once initialized, TLS data segments are used as data templates 617 // for a thread-local storage. For each new thread, runtime 618 // allocates memory for a TLS and copy templates there. No thread 619 // are supposed to use templates directly. Thus, it can be in RELRO. 620 if (Flags & SHF_TLS) 621 return true; 622 623 // .init_array, .preinit_array and .fini_array contain pointers to 624 // functions that are executed on process startup or exit. These 625 // pointers are set by the static linker, and they are not expected 626 // to change at runtime. But if you are an attacker, you could do 627 // interesting things by manipulating pointers in .fini_array, for 628 // example. So they are put into RELRO. 629 uint32_t Type = Sec->Type; 630 if (Type == SHT_INIT_ARRAY || Type == SHT_FINI_ARRAY || 631 Type == SHT_PREINIT_ARRAY) 632 return true; 633 634 // .got contains pointers to external symbols. They are resolved by 635 // the dynamic linker when a module is loaded into memory, and after 636 // that they are not expected to change. So, it can be in RELRO. 637 if (InX::Got && Sec == InX::Got->getParent()) 638 return true; 639 640 // .got.plt contains pointers to external function symbols. They are 641 // by default resolved lazily, so we usually cannot put it into RELRO. 642 // However, if "-z now" is given, the lazy symbol resolution is 643 // disabled, which enables us to put it into RELRO. 644 if (Sec == InX::GotPlt->getParent()) 645 return Config->ZNow; 646 647 // .dynamic section contains data for the dynamic linker, and 648 // there's no need to write to it at runtime, so it's better to put 649 // it into RELRO. 650 if (Sec == InX::Dynamic->getParent()) 651 return true; 652 653 // Sections with some special names are put into RELRO. This is a 654 // bit unfortunate because section names shouldn't be significant in 655 // ELF in spirit. But in reality many linker features depend on 656 // magic section names. 657 StringRef S = Sec->Name; 658 return S == ".data.rel.ro" || S == ".bss.rel.ro" || S == ".ctors" || 659 S == ".dtors" || S == ".jcr" || S == ".eh_frame" || 660 S == ".openbsd.randomdata"; 661 } 662 663 // We compute a rank for each section. The rank indicates where the 664 // section should be placed in the file. Instead of using simple 665 // numbers (0,1,2...), we use a series of flags. One for each decision 666 // point when placing the section. 667 // Using flags has two key properties: 668 // * It is easy to check if a give branch was taken. 669 // * It is easy two see how similar two ranks are (see getRankProximity). 670 enum RankFlags { 671 RF_NOT_ADDR_SET = 1 << 16, 672 RF_NOT_INTERP = 1 << 15, 673 RF_NOT_ALLOC = 1 << 14, 674 RF_WRITE = 1 << 13, 675 RF_EXEC_WRITE = 1 << 12, 676 RF_EXEC = 1 << 11, 677 RF_NON_TLS_BSS = 1 << 10, 678 RF_NON_TLS_BSS_RO = 1 << 9, 679 RF_NOT_TLS = 1 << 8, 680 RF_BSS = 1 << 7, 681 RF_PPC_NOT_TOCBSS = 1 << 6, 682 RF_PPC_OPD = 1 << 5, 683 RF_PPC_TOCL = 1 << 4, 684 RF_PPC_TOC = 1 << 3, 685 RF_PPC_BRANCH_LT = 1 << 2, 686 RF_MIPS_GPREL = 1 << 1, 687 RF_MIPS_NOT_GOT = 1 << 0 688 }; 689 690 static unsigned getSectionRank(const OutputSection *Sec) { 691 unsigned Rank = 0; 692 693 // We want to put section specified by -T option first, so we 694 // can start assigning VA starting from them later. 695 if (Config->SectionStartMap.count(Sec->Name)) 696 return Rank; 697 Rank |= RF_NOT_ADDR_SET; 698 699 // Put .interp first because some loaders want to see that section 700 // on the first page of the executable file when loaded into memory. 701 if (Sec->Name == ".interp") 702 return Rank; 703 Rank |= RF_NOT_INTERP; 704 705 // Allocatable sections go first to reduce the total PT_LOAD size and 706 // so debug info doesn't change addresses in actual code. 707 if (!(Sec->Flags & SHF_ALLOC)) 708 return Rank | RF_NOT_ALLOC; 709 710 // Sort sections based on their access permission in the following 711 // order: R, RX, RWX, RW. This order is based on the following 712 // considerations: 713 // * Read-only sections come first such that they go in the 714 // PT_LOAD covering the program headers at the start of the file. 715 // * Read-only, executable sections come next, unless the 716 // -no-rosegment option is used. 717 // * Writable, executable sections follow such that .plt on 718 // architectures where it needs to be writable will be placed 719 // between .text and .data. 720 // * Writable sections come last, such that .bss lands at the very 721 // end of the last PT_LOAD. 722 bool IsExec = Sec->Flags & SHF_EXECINSTR; 723 bool IsWrite = Sec->Flags & SHF_WRITE; 724 725 if (IsExec) { 726 if (IsWrite) 727 Rank |= RF_EXEC_WRITE; 728 else if (!Config->SingleRoRx) 729 Rank |= RF_EXEC; 730 } else { 731 if (IsWrite) 732 Rank |= RF_WRITE; 733 } 734 735 // If we got here we know that both A and B are in the same PT_LOAD. 736 737 bool IsTls = Sec->Flags & SHF_TLS; 738 bool IsNoBits = Sec->Type == SHT_NOBITS; 739 740 // The first requirement we have is to put (non-TLS) nobits sections last. The 741 // reason is that the only thing the dynamic linker will see about them is a 742 // p_memsz that is larger than p_filesz. Seeing that it zeros the end of the 743 // PT_LOAD, so that has to correspond to the nobits sections. 744 bool IsNonTlsNoBits = IsNoBits && !IsTls; 745 if (IsNonTlsNoBits) 746 Rank |= RF_NON_TLS_BSS; 747 748 // We place nobits RelRo sections before plain r/w ones, and non-nobits RelRo 749 // sections after r/w ones, so that the RelRo sections are contiguous. 750 bool IsRelRo = isRelroSection(Sec); 751 if (IsNonTlsNoBits && !IsRelRo) 752 Rank |= RF_NON_TLS_BSS_RO; 753 if (!IsNonTlsNoBits && IsRelRo) 754 Rank |= RF_NON_TLS_BSS_RO; 755 756 // The TLS initialization block needs to be a single contiguous block in a R/W 757 // PT_LOAD, so stick TLS sections directly before the other RelRo R/W 758 // sections. The TLS NOBITS sections are placed here as they don't take up 759 // virtual address space in the PT_LOAD. 760 if (!IsTls) 761 Rank |= RF_NOT_TLS; 762 763 // Within the TLS initialization block, the non-nobits sections need to appear 764 // first. 765 if (IsNoBits) 766 Rank |= RF_BSS; 767 768 // Some architectures have additional ordering restrictions for sections 769 // within the same PT_LOAD. 770 if (Config->EMachine == EM_PPC64) { 771 // PPC64 has a number of special SHT_PROGBITS+SHF_ALLOC+SHF_WRITE sections 772 // that we would like to make sure appear is a specific order to maximize 773 // their coverage by a single signed 16-bit offset from the TOC base 774 // pointer. Conversely, the special .tocbss section should be first among 775 // all SHT_NOBITS sections. This will put it next to the loaded special 776 // PPC64 sections (and, thus, within reach of the TOC base pointer). 777 StringRef Name = Sec->Name; 778 if (Name != ".tocbss") 779 Rank |= RF_PPC_NOT_TOCBSS; 780 781 if (Name == ".opd") 782 Rank |= RF_PPC_OPD; 783 784 if (Name == ".toc1") 785 Rank |= RF_PPC_TOCL; 786 787 if (Name == ".toc") 788 Rank |= RF_PPC_TOC; 789 790 if (Name == ".branch_lt") 791 Rank |= RF_PPC_BRANCH_LT; 792 } 793 if (Config->EMachine == EM_MIPS) { 794 // All sections with SHF_MIPS_GPREL flag should be grouped together 795 // because data in these sections is addressable with a gp relative address. 796 if (Sec->Flags & SHF_MIPS_GPREL) 797 Rank |= RF_MIPS_GPREL; 798 799 if (Sec->Name != ".got") 800 Rank |= RF_MIPS_NOT_GOT; 801 } 802 803 return Rank; 804 } 805 806 static bool compareSections(const BaseCommand *ACmd, const BaseCommand *BCmd) { 807 const OutputSection *A = cast<OutputSection>(ACmd); 808 const OutputSection *B = cast<OutputSection>(BCmd); 809 if (A->SortRank != B->SortRank) 810 return A->SortRank < B->SortRank; 811 if (!(A->SortRank & RF_NOT_ADDR_SET)) 812 return Config->SectionStartMap.lookup(A->Name) < 813 Config->SectionStartMap.lookup(B->Name); 814 return false; 815 } 816 817 void PhdrEntry::add(OutputSection *Sec) { 818 LastSec = Sec; 819 if (!FirstSec) 820 FirstSec = Sec; 821 p_align = std::max(p_align, Sec->Alignment); 822 if (p_type == PT_LOAD) 823 Sec->PtLoad = this; 824 } 825 826 // The beginning and the ending of .rel[a].plt section are marked 827 // with __rel[a]_iplt_{start,end} symbols if it is a statically linked 828 // executable. The runtime needs these symbols in order to resolve 829 // all IRELATIVE relocs on startup. For dynamic executables, we don't 830 // need these symbols, since IRELATIVE relocs are resolved through GOT 831 // and PLT. For details, see http://www.airs.com/blog/archives/403. 832 template <class ELFT> void Writer<ELFT>::addRelIpltSymbols() { 833 if (!Config->Static) 834 return; 835 StringRef S = Config->IsRela ? "__rela_iplt_start" : "__rel_iplt_start"; 836 addOptionalRegular<ELFT>(S, InX::RelaIplt, 0, STV_HIDDEN, STB_WEAK); 837 838 S = Config->IsRela ? "__rela_iplt_end" : "__rel_iplt_end"; 839 addOptionalRegular<ELFT>(S, InX::RelaIplt, -1, STV_HIDDEN, STB_WEAK); 840 } 841 842 template <class ELFT> 843 void Writer<ELFT>::forEachRelSec(std::function<void(InputSectionBase &)> Fn) { 844 // Scan all relocations. Each relocation goes through a series 845 // of tests to determine if it needs special treatment, such as 846 // creating GOT, PLT, copy relocations, etc. 847 // Note that relocations for non-alloc sections are directly 848 // processed by InputSection::relocateNonAlloc. 849 for (InputSectionBase *IS : InputSections) 850 if (IS->Live && isa<InputSection>(IS) && (IS->Flags & SHF_ALLOC)) 851 Fn(*IS); 852 for (EhInputSection *ES : InX::EhFrame->Sections) 853 Fn(*ES); 854 } 855 856 // This function generates assignments for predefined symbols (e.g. _end or 857 // _etext) and inserts them into the commands sequence to be processed at the 858 // appropriate time. This ensures that the value is going to be correct by the 859 // time any references to these symbols are processed and is equivalent to 860 // defining these symbols explicitly in the linker script. 861 template <class ELFT> void Writer<ELFT>::setReservedSymbolSections() { 862 if (ElfSym::GlobalOffsetTable) { 863 // The _GLOBAL_OFFSET_TABLE_ symbol is defined by target convention to 864 // be at some offset from the base of the .got section, usually 0 or the end 865 // of the .got 866 InputSection *GotSection = InX::MipsGot ? cast<InputSection>(InX::MipsGot) 867 : cast<InputSection>(InX::Got); 868 ElfSym::GlobalOffsetTable->Section = GotSection; 869 } 870 871 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 bool compareByFilePosition(InputSection *A, InputSection *B) { 1151 // Synthetic doesn't have link order dependecy, stable_sort will keep it last 1152 if (A->kind() == InputSectionBase::Synthetic || 1153 B->kind() == InputSectionBase::Synthetic) 1154 return false; 1155 InputSection *LA = A->getLinkOrderDep(); 1156 InputSection *LB = B->getLinkOrderDep(); 1157 OutputSection *AOut = LA->getParent(); 1158 OutputSection *BOut = LB->getParent(); 1159 if (AOut != BOut) 1160 return AOut->SectionIndex < BOut->SectionIndex; 1161 return LA->OutSecOff < LB->OutSecOff; 1162 } 1163 1164 template <class ELFT> void Writer<ELFT>::resolveShfLinkOrder() { 1165 for (OutputSection *Sec : OutputSections) { 1166 if (!(Sec->Flags & SHF_LINK_ORDER)) 1167 continue; 1168 1169 // Link order may be distributed across several InputSectionDescriptions 1170 // but sort must consider them all at once. 1171 std::vector<InputSection **> ScriptSections; 1172 std::vector<InputSection *> Sections; 1173 for (BaseCommand *Base : Sec->SectionCommands) { 1174 if (auto *ISD = dyn_cast<InputSectionDescription>(Base)) { 1175 for (InputSection *&IS : ISD->Sections) { 1176 ScriptSections.push_back(&IS); 1177 Sections.push_back(IS); 1178 } 1179 } 1180 } 1181 std::stable_sort(Sections.begin(), Sections.end(), compareByFilePosition); 1182 for (int I = 0, N = Sections.size(); I < N; ++I) 1183 *ScriptSections[I] = Sections[I]; 1184 } 1185 } 1186 1187 static void applySynthetic(const std::vector<SyntheticSection *> &Sections, 1188 std::function<void(SyntheticSection *)> Fn) { 1189 for (SyntheticSection *SS : Sections) 1190 if (SS && SS->getParent() && !SS->empty()) 1191 Fn(SS); 1192 } 1193 1194 // In order to allow users to manipulate linker-synthesized sections, 1195 // we had to add synthetic sections to the input section list early, 1196 // even before we make decisions whether they are needed. This allows 1197 // users to write scripts like this: ".mygot : { .got }". 1198 // 1199 // Doing it has an unintended side effects. If it turns out that we 1200 // don't need a .got (for example) at all because there's no 1201 // relocation that needs a .got, we don't want to emit .got. 1202 // 1203 // To deal with the above problem, this function is called after 1204 // scanRelocations is called to remove synthetic sections that turn 1205 // out to be empty. 1206 static void removeUnusedSyntheticSections() { 1207 // All input synthetic sections that can be empty are placed after 1208 // all regular ones. We iterate over them all and exit at first 1209 // non-synthetic. 1210 for (InputSectionBase *S : llvm::reverse(InputSections)) { 1211 SyntheticSection *SS = dyn_cast<SyntheticSection>(S); 1212 if (!SS) 1213 return; 1214 OutputSection *OS = SS->getParent(); 1215 if (!SS->empty() || !OS) 1216 continue; 1217 1218 std::vector<BaseCommand *>::iterator Empty = OS->SectionCommands.end(); 1219 for (auto I = OS->SectionCommands.begin(), E = OS->SectionCommands.end(); 1220 I != E; ++I) { 1221 BaseCommand *B = *I; 1222 if (auto *ISD = dyn_cast<InputSectionDescription>(B)) { 1223 llvm::erase_if(ISD->Sections, 1224 [=](InputSection *IS) { return IS == SS; }); 1225 if (ISD->Sections.empty()) 1226 Empty = I; 1227 } 1228 } 1229 if (Empty != OS->SectionCommands.end()) 1230 OS->SectionCommands.erase(Empty); 1231 1232 // If there are no other sections in the output section, remove it from the 1233 // output. 1234 if (OS->SectionCommands.empty()) 1235 OS->Live = false; 1236 } 1237 } 1238 1239 // Returns true if a symbol can be replaced at load-time by a symbol 1240 // with the same name defined in other ELF executable or DSO. 1241 static bool computeIsPreemptible(const Symbol &B) { 1242 assert(!B.isLocal()); 1243 // Only symbols that appear in dynsym can be preempted. 1244 if (!B.includeInDynsym()) 1245 return false; 1246 1247 // Only default visibility symbols can be preempted. 1248 if (B.Visibility != STV_DEFAULT) 1249 return false; 1250 1251 // At this point copy relocations have not been created yet, so any 1252 // symbol that is not defined locally is preemptible. 1253 if (!B.isDefined()) 1254 return true; 1255 1256 // If we have a dynamic list it specifies which local symbols are preemptible. 1257 if (Config->HasDynamicList) 1258 return false; 1259 1260 if (!Config->Shared) 1261 return false; 1262 1263 // -Bsymbolic means that definitions are not preempted. 1264 if (Config->Bsymbolic || (Config->BsymbolicFunctions && B.isFunc())) 1265 return false; 1266 return true; 1267 } 1268 1269 // Create output section objects and add them to OutputSections. 1270 template <class ELFT> void Writer<ELFT>::finalizeSections() { 1271 Out::DebugInfo = findSection(".debug_info"); 1272 Out::PreinitArray = findSection(".preinit_array"); 1273 Out::InitArray = findSection(".init_array"); 1274 Out::FiniArray = findSection(".fini_array"); 1275 1276 // The linker needs to define SECNAME_start, SECNAME_end and SECNAME_stop 1277 // symbols for sections, so that the runtime can get the start and end 1278 // addresses of each section by section name. Add such symbols. 1279 if (!Config->Relocatable) { 1280 addStartEndSymbols(); 1281 for (BaseCommand *Base : Script->SectionCommands) 1282 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1283 addStartStopSymbols(Sec); 1284 } 1285 1286 // Add _DYNAMIC symbol. Unlike GNU gold, our _DYNAMIC symbol has no type. 1287 // It should be okay as no one seems to care about the type. 1288 // Even the author of gold doesn't remember why gold behaves that way. 1289 // https://sourceware.org/ml/binutils/2002-03/msg00360.html 1290 if (InX::DynSymTab) 1291 Symtab->addRegular<ELFT>("_DYNAMIC", STV_HIDDEN, STT_NOTYPE, 0 /*Value*/, 1292 /*Size=*/0, STB_WEAK, InX::Dynamic, 1293 /*File=*/nullptr); 1294 1295 // Define __rel[a]_iplt_{start,end} symbols if needed. 1296 addRelIpltSymbols(); 1297 1298 // This responsible for splitting up .eh_frame section into 1299 // pieces. The relocation scan uses those pieces, so this has to be 1300 // earlier. 1301 applySynthetic({InX::EhFrame}, 1302 [](SyntheticSection *SS) { SS->finalizeContents(); }); 1303 1304 for (Symbol *S : Symtab->getSymbols()) 1305 S->IsPreemptible |= computeIsPreemptible(*S); 1306 1307 // Scan relocations. This must be done after every symbol is declared so that 1308 // we can correctly decide if a dynamic relocation is needed. 1309 if (!Config->Relocatable) 1310 forEachRelSec(scanRelocations<ELFT>); 1311 1312 if (InX::Plt && !InX::Plt->empty()) 1313 InX::Plt->addSymbols(); 1314 if (InX::Iplt && !InX::Iplt->empty()) 1315 InX::Iplt->addSymbols(); 1316 1317 // Now that we have defined all possible global symbols including linker- 1318 // synthesized ones. Visit all symbols to give the finishing touches. 1319 for (Symbol *Sym : Symtab->getSymbols()) { 1320 if (!includeInSymtab(*Sym)) 1321 continue; 1322 if (InX::SymTab) 1323 InX::SymTab->addSymbol(Sym); 1324 1325 if (InX::DynSymTab && Sym->includeInDynsym()) { 1326 InX::DynSymTab->addSymbol(Sym); 1327 if (auto *SS = dyn_cast<SharedSymbol>(Sym)) 1328 if (cast<SharedFile<ELFT>>(Sym->File)->IsNeeded) 1329 In<ELFT>::VerNeed->addSymbol(SS); 1330 } 1331 } 1332 1333 // Do not proceed if there was an undefined symbol. 1334 if (errorCount()) 1335 return; 1336 1337 addPredefinedSections(); 1338 removeUnusedSyntheticSections(); 1339 1340 sortSections(); 1341 Script->removeEmptyCommands(); 1342 1343 // Now that we have the final list, create a list of all the 1344 // OutputSections for convenience. 1345 for (BaseCommand *Base : Script->SectionCommands) 1346 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1347 OutputSections.push_back(Sec); 1348 1349 // Prefer command line supplied address over other constraints. 1350 for (OutputSection *Sec : OutputSections) { 1351 auto I = Config->SectionStartMap.find(Sec->Name); 1352 if (I != Config->SectionStartMap.end()) 1353 Sec->AddrExpr = [=] { return I->second; }; 1354 } 1355 1356 // This is a bit of a hack. A value of 0 means undef, so we set it 1357 // to 1 t make __ehdr_start defined. The section number is not 1358 // particularly relevant. 1359 Out::ElfHeader->SectionIndex = 1; 1360 1361 unsigned I = 1; 1362 for (OutputSection *Sec : OutputSections) { 1363 Sec->SectionIndex = I++; 1364 Sec->ShName = InX::ShStrTab->addString(Sec->Name); 1365 } 1366 1367 // Binary and relocatable output does not have PHDRS. 1368 // The headers have to be created before finalize as that can influence the 1369 // image base and the dynamic section on mips includes the image base. 1370 if (!Config->Relocatable && !Config->OFormatBinary) { 1371 Phdrs = Script->hasPhdrsCommands() ? Script->createPhdrs() : createPhdrs(); 1372 addPtArmExid(Phdrs); 1373 Out::ProgramHeaders->Size = sizeof(Elf_Phdr) * Phdrs.size(); 1374 } 1375 1376 // Some symbols are defined in term of program headers. Now that we 1377 // have the headers, we can find out which sections they point to. 1378 setReservedSymbolSections(); 1379 1380 // Dynamic section must be the last one in this list and dynamic 1381 // symbol table section (DynSymTab) must be the first one. 1382 applySynthetic( 1383 {InX::DynSymTab, InX::Bss, InX::BssRelRo, InX::GnuHashTab, 1384 InX::HashTab, InX::SymTab, InX::ShStrTab, InX::StrTab, 1385 In<ELFT>::VerDef, InX::DynStrTab, InX::Got, InX::MipsGot, 1386 InX::IgotPlt, InX::GotPlt, InX::RelaDyn, InX::RelaIplt, 1387 InX::RelaPlt, InX::Plt, InX::Iplt, InX::EhFrameHdr, 1388 In<ELFT>::VerSym, In<ELFT>::VerNeed, InX::Dynamic}, 1389 [](SyntheticSection *SS) { SS->finalizeContents(); }); 1390 1391 if (!Script->HasSectionsCommand && !Config->Relocatable) 1392 fixSectionAlignments(); 1393 1394 // After link order processing .ARM.exidx sections can be deduplicated, which 1395 // needs to be resolved before any other address dependent operation. 1396 resolveShfLinkOrder(); 1397 1398 // Some architectures need to generate content that depends on the address 1399 // of InputSections. For example some architectures use small displacements 1400 // for jump instructions that is is the linker's responsibility for creating 1401 // range extension thunks for. As the generation of the content may also 1402 // alter InputSection addresses we must converge to a fixed point. 1403 if (Target->NeedsThunks || Config->AndroidPackDynRelocs) { 1404 ThunkCreator TC; 1405 bool Changed; 1406 do { 1407 Script->assignAddresses(); 1408 Changed = false; 1409 if (Target->NeedsThunks) 1410 Changed |= TC.createThunks(OutputSections); 1411 if (Config->FixCortexA53Errata843419) { 1412 if (Changed) 1413 Script->assignAddresses(); 1414 reportA53Errata843419Fixes(); 1415 } 1416 if (InX::MipsGot) 1417 InX::MipsGot->updateAllocSize(); 1418 Changed |= InX::RelaDyn->updateAllocSize(); 1419 } while (Changed); 1420 } 1421 1422 // Fill other section headers. The dynamic table is finalized 1423 // at the end because some tags like RELSZ depend on result 1424 // of finalizing other sections. 1425 for (OutputSection *Sec : OutputSections) 1426 Sec->finalize<ELFT>(); 1427 1428 // createThunks may have added local symbols to the static symbol table 1429 applySynthetic({InX::SymTab}, 1430 [](SyntheticSection *SS) { SS->postThunkContents(); }); 1431 } 1432 1433 template <class ELFT> void Writer<ELFT>::addPredefinedSections() { 1434 // ARM ABI requires .ARM.exidx to be terminated by some piece of data. 1435 // We have the terminater synthetic section class. Add that at the end. 1436 OutputSection *Cmd = findSection(".ARM.exidx"); 1437 if (!Cmd || !Cmd->Live || Config->Relocatable) 1438 return; 1439 1440 auto *Sentinel = make<ARMExidxSentinelSection>(); 1441 Cmd->addSection(Sentinel); 1442 } 1443 1444 // The linker is expected to define SECNAME_start and SECNAME_end 1445 // symbols for a few sections. This function defines them. 1446 template <class ELFT> void Writer<ELFT>::addStartEndSymbols() { 1447 auto Define = [&](StringRef Start, StringRef End, OutputSection *OS) { 1448 // These symbols resolve to the image base if the section does not exist. 1449 // A special value -1 indicates end of the section. 1450 if (OS) { 1451 addOptionalRegular<ELFT>(Start, OS, 0); 1452 addOptionalRegular<ELFT>(End, OS, -1); 1453 } else { 1454 if (Config->Pic) 1455 OS = Out::ElfHeader; 1456 addOptionalRegular<ELFT>(Start, OS, 0); 1457 addOptionalRegular<ELFT>(End, OS, 0); 1458 } 1459 }; 1460 1461 Define("__preinit_array_start", "__preinit_array_end", Out::PreinitArray); 1462 Define("__init_array_start", "__init_array_end", Out::InitArray); 1463 Define("__fini_array_start", "__fini_array_end", Out::FiniArray); 1464 1465 if (OutputSection *Sec = findSection(".ARM.exidx")) 1466 Define("__exidx_start", "__exidx_end", Sec); 1467 } 1468 1469 // If a section name is valid as a C identifier (which is rare because of 1470 // the leading '.'), linkers are expected to define __start_<secname> and 1471 // __stop_<secname> symbols. They are at beginning and end of the section, 1472 // respectively. This is not requested by the ELF standard, but GNU ld and 1473 // gold provide the feature, and used by many programs. 1474 template <class ELFT> 1475 void Writer<ELFT>::addStartStopSymbols(OutputSection *Sec) { 1476 StringRef S = Sec->Name; 1477 if (!isValidCIdentifier(S)) 1478 return; 1479 addOptionalRegular<ELFT>(Saver.save("__start_" + S), Sec, 0, STV_DEFAULT); 1480 addOptionalRegular<ELFT>(Saver.save("__stop_" + S), Sec, -1, STV_DEFAULT); 1481 } 1482 1483 static bool needsPtLoad(OutputSection *Sec) { 1484 if (!(Sec->Flags & SHF_ALLOC)) 1485 return false; 1486 1487 // Don't allocate VA space for TLS NOBITS sections. The PT_TLS PHDR is 1488 // responsible for allocating space for them, not the PT_LOAD that 1489 // contains the TLS initialization image. 1490 if (Sec->Flags & SHF_TLS && Sec->Type == SHT_NOBITS) 1491 return false; 1492 return true; 1493 } 1494 1495 // Linker scripts are responsible for aligning addresses. Unfortunately, most 1496 // linker scripts are designed for creating two PT_LOADs only, one RX and one 1497 // RW. This means that there is no alignment in the RO to RX transition and we 1498 // cannot create a PT_LOAD there. 1499 static uint64_t computeFlags(uint64_t Flags) { 1500 if (Config->Omagic) 1501 return PF_R | PF_W | PF_X; 1502 if (Config->SingleRoRx && !(Flags & PF_W)) 1503 return Flags | PF_X; 1504 return Flags; 1505 } 1506 1507 // Decide which program headers to create and which sections to include in each 1508 // one. 1509 template <class ELFT> std::vector<PhdrEntry *> Writer<ELFT>::createPhdrs() { 1510 std::vector<PhdrEntry *> Ret; 1511 auto AddHdr = [&](unsigned Type, unsigned Flags) -> PhdrEntry * { 1512 Ret.push_back(make<PhdrEntry>(Type, Flags)); 1513 return Ret.back(); 1514 }; 1515 1516 // The first phdr entry is PT_PHDR which describes the program header itself. 1517 AddHdr(PT_PHDR, PF_R)->add(Out::ProgramHeaders); 1518 1519 // PT_INTERP must be the second entry if exists. 1520 if (OutputSection *Cmd = findSection(".interp")) 1521 AddHdr(PT_INTERP, Cmd->getPhdrFlags())->add(Cmd); 1522 1523 // Add the first PT_LOAD segment for regular output sections. 1524 uint64_t Flags = computeFlags(PF_R); 1525 PhdrEntry *Load = AddHdr(PT_LOAD, Flags); 1526 1527 // Add the headers. We will remove them if they don't fit. 1528 Load->add(Out::ElfHeader); 1529 Load->add(Out::ProgramHeaders); 1530 1531 for (OutputSection *Sec : OutputSections) { 1532 if (!(Sec->Flags & SHF_ALLOC)) 1533 break; 1534 if (!needsPtLoad(Sec)) 1535 continue; 1536 1537 // Segments are contiguous memory regions that has the same attributes 1538 // (e.g. executable or writable). There is one phdr for each segment. 1539 // Therefore, we need to create a new phdr when the next section has 1540 // different flags or is loaded at a discontiguous address using AT linker 1541 // script command. 1542 uint64_t NewFlags = computeFlags(Sec->getPhdrFlags()); 1543 if (Sec->LMAExpr || Flags != NewFlags) { 1544 Load = AddHdr(PT_LOAD, NewFlags); 1545 Flags = NewFlags; 1546 } 1547 1548 Load->add(Sec); 1549 } 1550 1551 // Add a TLS segment if any. 1552 PhdrEntry *TlsHdr = make<PhdrEntry>(PT_TLS, PF_R); 1553 for (OutputSection *Sec : OutputSections) 1554 if (Sec->Flags & SHF_TLS) 1555 TlsHdr->add(Sec); 1556 if (TlsHdr->FirstSec) 1557 Ret.push_back(TlsHdr); 1558 1559 // Add an entry for .dynamic. 1560 if (InX::DynSymTab) 1561 AddHdr(PT_DYNAMIC, InX::Dynamic->getParent()->getPhdrFlags()) 1562 ->add(InX::Dynamic->getParent()); 1563 1564 // PT_GNU_RELRO includes all sections that should be marked as 1565 // read-only by dynamic linker after proccessing relocations. 1566 // Current dynamic loaders only support one PT_GNU_RELRO PHDR, give 1567 // an error message if more than one PT_GNU_RELRO PHDR is required. 1568 PhdrEntry *RelRo = make<PhdrEntry>(PT_GNU_RELRO, PF_R); 1569 bool InRelroPhdr = false; 1570 bool IsRelroFinished = false; 1571 for (OutputSection *Sec : OutputSections) { 1572 if (!needsPtLoad(Sec)) 1573 continue; 1574 if (isRelroSection(Sec)) { 1575 InRelroPhdr = true; 1576 if (!IsRelroFinished) 1577 RelRo->add(Sec); 1578 else 1579 error("section: " + Sec->Name + " is not contiguous with other relro" + 1580 " sections"); 1581 } else if (InRelroPhdr) { 1582 InRelroPhdr = false; 1583 IsRelroFinished = true; 1584 } 1585 } 1586 if (RelRo->FirstSec) 1587 Ret.push_back(RelRo); 1588 1589 // PT_GNU_EH_FRAME is a special section pointing on .eh_frame_hdr. 1590 if (!InX::EhFrame->empty() && InX::EhFrameHdr && InX::EhFrame->getParent() && 1591 InX::EhFrameHdr->getParent()) 1592 AddHdr(PT_GNU_EH_FRAME, InX::EhFrameHdr->getParent()->getPhdrFlags()) 1593 ->add(InX::EhFrameHdr->getParent()); 1594 1595 // PT_OPENBSD_RANDOMIZE is an OpenBSD-specific feature. That makes 1596 // the dynamic linker fill the segment with random data. 1597 if (OutputSection *Cmd = findSection(".openbsd.randomdata")) 1598 AddHdr(PT_OPENBSD_RANDOMIZE, Cmd->getPhdrFlags())->add(Cmd); 1599 1600 // PT_GNU_STACK is a special section to tell the loader to make the 1601 // pages for the stack non-executable. If you really want an executable 1602 // stack, you can pass -z execstack, but that's not recommended for 1603 // security reasons. 1604 unsigned Perm; 1605 if (Config->ZExecstack) 1606 Perm = PF_R | PF_W | PF_X; 1607 else 1608 Perm = PF_R | PF_W; 1609 AddHdr(PT_GNU_STACK, Perm)->p_memsz = Config->ZStackSize; 1610 1611 // PT_OPENBSD_WXNEEDED is a OpenBSD-specific header to mark the executable 1612 // is expected to perform W^X violations, such as calling mprotect(2) or 1613 // mmap(2) with PROT_WRITE | PROT_EXEC, which is prohibited by default on 1614 // OpenBSD. 1615 if (Config->ZWxneeded) 1616 AddHdr(PT_OPENBSD_WXNEEDED, PF_X); 1617 1618 // Create one PT_NOTE per a group of contiguous .note sections. 1619 PhdrEntry *Note = nullptr; 1620 for (OutputSection *Sec : OutputSections) { 1621 if (Sec->Type == SHT_NOTE) { 1622 if (!Note || Sec->LMAExpr) 1623 Note = AddHdr(PT_NOTE, PF_R); 1624 Note->add(Sec); 1625 } else { 1626 Note = nullptr; 1627 } 1628 } 1629 return Ret; 1630 } 1631 1632 template <class ELFT> 1633 void Writer<ELFT>::addPtArmExid(std::vector<PhdrEntry *> &Phdrs) { 1634 if (Config->EMachine != EM_ARM) 1635 return; 1636 auto I = llvm::find_if(OutputSections, [](OutputSection *Cmd) { 1637 return Cmd->Type == SHT_ARM_EXIDX; 1638 }); 1639 if (I == OutputSections.end()) 1640 return; 1641 1642 // PT_ARM_EXIDX is the ARM EHABI equivalent of PT_GNU_EH_FRAME 1643 PhdrEntry *ARMExidx = make<PhdrEntry>(PT_ARM_EXIDX, PF_R); 1644 ARMExidx->add(*I); 1645 Phdrs.push_back(ARMExidx); 1646 } 1647 1648 // The first section of each PT_LOAD, the first section in PT_GNU_RELRO and the 1649 // first section after PT_GNU_RELRO have to be page aligned so that the dynamic 1650 // linker can set the permissions. 1651 template <class ELFT> void Writer<ELFT>::fixSectionAlignments() { 1652 auto PageAlign = [](OutputSection *Cmd) { 1653 if (Cmd && !Cmd->AddrExpr) 1654 Cmd->AddrExpr = [=] { 1655 return alignTo(Script->getDot(), Config->MaxPageSize); 1656 }; 1657 }; 1658 1659 for (const PhdrEntry *P : Phdrs) 1660 if (P->p_type == PT_LOAD && P->FirstSec) 1661 PageAlign(P->FirstSec); 1662 1663 for (const PhdrEntry *P : Phdrs) { 1664 if (P->p_type != PT_GNU_RELRO) 1665 continue; 1666 if (P->FirstSec) 1667 PageAlign(P->FirstSec); 1668 // Find the first section after PT_GNU_RELRO. If it is in a PT_LOAD we 1669 // have to align it to a page. 1670 auto End = OutputSections.end(); 1671 auto I = std::find(OutputSections.begin(), End, P->LastSec); 1672 if (I == End || (I + 1) == End) 1673 continue; 1674 OutputSection *Cmd = (*(I + 1)); 1675 if (needsPtLoad(Cmd)) 1676 PageAlign(Cmd); 1677 } 1678 } 1679 1680 // Adjusts the file alignment for a given output section and returns 1681 // its new file offset. The file offset must be the same with its 1682 // virtual address (modulo the page size) so that the loader can load 1683 // executables without any address adjustment. 1684 static uint64_t getFileAlignment(uint64_t Off, OutputSection *Cmd) { 1685 // If the section is not in a PT_LOAD, we just have to align it. 1686 if (!Cmd->PtLoad) 1687 return alignTo(Off, Cmd->Alignment); 1688 1689 OutputSection *First = Cmd->PtLoad->FirstSec; 1690 // The first section in a PT_LOAD has to have congruent offset and address 1691 // module the page size. 1692 if (Cmd == First) 1693 return alignTo(Off, std::max<uint64_t>(Cmd->Alignment, Config->MaxPageSize), 1694 Cmd->Addr); 1695 1696 // If two sections share the same PT_LOAD the file offset is calculated 1697 // using this formula: Off2 = Off1 + (VA2 - VA1). 1698 return First->Offset + Cmd->Addr - First->Addr; 1699 } 1700 1701 static uint64_t setOffset(OutputSection *Cmd, uint64_t Off) { 1702 if (Cmd->Type == SHT_NOBITS) { 1703 Cmd->Offset = Off; 1704 return Off; 1705 } 1706 1707 Off = getFileAlignment(Off, Cmd); 1708 Cmd->Offset = Off; 1709 return Off + Cmd->Size; 1710 } 1711 1712 template <class ELFT> void Writer<ELFT>::assignFileOffsetsBinary() { 1713 uint64_t Off = 0; 1714 for (OutputSection *Sec : OutputSections) 1715 if (Sec->Flags & SHF_ALLOC) 1716 Off = setOffset(Sec, Off); 1717 FileSize = alignTo(Off, Config->Wordsize); 1718 } 1719 1720 // Assign file offsets to output sections. 1721 template <class ELFT> void Writer<ELFT>::assignFileOffsets() { 1722 uint64_t Off = 0; 1723 Off = setOffset(Out::ElfHeader, Off); 1724 Off = setOffset(Out::ProgramHeaders, Off); 1725 1726 PhdrEntry *LastRX = nullptr; 1727 for (PhdrEntry *P : Phdrs) 1728 if (P->p_type == PT_LOAD && (P->p_flags & PF_X)) 1729 LastRX = P; 1730 1731 for (OutputSection *Sec : OutputSections) { 1732 Off = setOffset(Sec, Off); 1733 if (Script->HasSectionsCommand) 1734 continue; 1735 // If this is a last section of the last executable segment and that 1736 // segment is the last loadable segment, align the offset of the 1737 // following section to avoid loading non-segments parts of the file. 1738 if (LastRX && LastRX->LastSec == Sec) 1739 Off = alignTo(Off, Target->PageSize); 1740 } 1741 1742 SectionHeaderOff = alignTo(Off, Config->Wordsize); 1743 FileSize = SectionHeaderOff + (OutputSections.size() + 1) * sizeof(Elf_Shdr); 1744 } 1745 1746 // Finalize the program headers. We call this function after we assign 1747 // file offsets and VAs to all sections. 1748 template <class ELFT> void Writer<ELFT>::setPhdrs() { 1749 for (PhdrEntry *P : Phdrs) { 1750 OutputSection *First = P->FirstSec; 1751 OutputSection *Last = P->LastSec; 1752 if (First) { 1753 P->p_filesz = Last->Offset - First->Offset; 1754 if (Last->Type != SHT_NOBITS) 1755 P->p_filesz += Last->Size; 1756 P->p_memsz = Last->Addr + Last->Size - First->Addr; 1757 P->p_offset = First->Offset; 1758 P->p_vaddr = First->Addr; 1759 if (!P->HasLMA) 1760 P->p_paddr = First->getLMA(); 1761 } 1762 if (P->p_type == PT_LOAD) 1763 P->p_align = std::max<uint64_t>(P->p_align, Config->MaxPageSize); 1764 else if (P->p_type == PT_GNU_RELRO) { 1765 P->p_align = 1; 1766 // The glibc dynamic loader rounds the size down, so we need to round up 1767 // to protect the last page. This is a no-op on FreeBSD which always 1768 // rounds up. 1769 P->p_memsz = alignTo(P->p_memsz, Target->PageSize); 1770 } 1771 1772 // The TLS pointer goes after PT_TLS. At least glibc will align it, 1773 // so round up the size to make sure the offsets are correct. 1774 if (P->p_type == PT_TLS) { 1775 Out::TlsPhdr = P; 1776 if (P->p_memsz) 1777 P->p_memsz = alignTo(P->p_memsz, P->p_align); 1778 } 1779 } 1780 } 1781 1782 // The entry point address is chosen in the following ways. 1783 // 1784 // 1. the '-e' entry command-line option; 1785 // 2. the ENTRY(symbol) command in a linker control script; 1786 // 3. the value of the symbol _start, if present; 1787 // 4. the number represented by the entry symbol, if it is a number; 1788 // 5. the address of the first byte of the .text section, if present; 1789 // 6. the address 0. 1790 template <class ELFT> uint64_t Writer<ELFT>::getEntryAddr() { 1791 // Case 1, 2 or 3 1792 if (Symbol *B = Symtab->find(Config->Entry)) 1793 return B->getVA(); 1794 1795 // Case 4 1796 uint64_t Addr; 1797 if (to_integer(Config->Entry, Addr)) 1798 return Addr; 1799 1800 // Case 5 1801 if (OutputSection *Sec = findSection(".text")) { 1802 if (Config->WarnMissingEntry) 1803 warn("cannot find entry symbol " + Config->Entry + "; defaulting to 0x" + 1804 utohexstr(Sec->Addr)); 1805 return Sec->Addr; 1806 } 1807 1808 // Case 6 1809 if (Config->WarnMissingEntry) 1810 warn("cannot find entry symbol " + Config->Entry + 1811 "; not setting start address"); 1812 return 0; 1813 } 1814 1815 static uint16_t getELFType() { 1816 if (Config->Pic) 1817 return ET_DYN; 1818 if (Config->Relocatable) 1819 return ET_REL; 1820 return ET_EXEC; 1821 } 1822 1823 template <class ELFT> void Writer<ELFT>::writeHeader() { 1824 uint8_t *Buf = Buffer->getBufferStart(); 1825 memcpy(Buf, "\177ELF", 4); 1826 1827 // Write the ELF header. 1828 auto *EHdr = reinterpret_cast<Elf_Ehdr *>(Buf); 1829 EHdr->e_ident[EI_CLASS] = Config->Is64 ? ELFCLASS64 : ELFCLASS32; 1830 EHdr->e_ident[EI_DATA] = Config->IsLE ? ELFDATA2LSB : ELFDATA2MSB; 1831 EHdr->e_ident[EI_VERSION] = EV_CURRENT; 1832 EHdr->e_ident[EI_OSABI] = Config->OSABI; 1833 EHdr->e_type = getELFType(); 1834 EHdr->e_machine = Config->EMachine; 1835 EHdr->e_version = EV_CURRENT; 1836 EHdr->e_entry = getEntryAddr(); 1837 EHdr->e_shoff = SectionHeaderOff; 1838 EHdr->e_flags = Config->EFlags; 1839 EHdr->e_ehsize = sizeof(Elf_Ehdr); 1840 EHdr->e_phnum = Phdrs.size(); 1841 EHdr->e_shentsize = sizeof(Elf_Shdr); 1842 EHdr->e_shnum = OutputSections.size() + 1; 1843 EHdr->e_shstrndx = InX::ShStrTab->getParent()->SectionIndex; 1844 1845 if (!Config->Relocatable) { 1846 EHdr->e_phoff = sizeof(Elf_Ehdr); 1847 EHdr->e_phentsize = sizeof(Elf_Phdr); 1848 } 1849 1850 // Write the program header table. 1851 auto *HBuf = reinterpret_cast<Elf_Phdr *>(Buf + EHdr->e_phoff); 1852 for (PhdrEntry *P : Phdrs) { 1853 HBuf->p_type = P->p_type; 1854 HBuf->p_flags = P->p_flags; 1855 HBuf->p_offset = P->p_offset; 1856 HBuf->p_vaddr = P->p_vaddr; 1857 HBuf->p_paddr = P->p_paddr; 1858 HBuf->p_filesz = P->p_filesz; 1859 HBuf->p_memsz = P->p_memsz; 1860 HBuf->p_align = P->p_align; 1861 ++HBuf; 1862 } 1863 1864 // Write the section header table. Note that the first table entry is null. 1865 auto *SHdrs = reinterpret_cast<Elf_Shdr *>(Buf + EHdr->e_shoff); 1866 for (OutputSection *Sec : OutputSections) 1867 Sec->writeHeaderTo<ELFT>(++SHdrs); 1868 } 1869 1870 // Open a result file. 1871 template <class ELFT> void Writer<ELFT>::openFile() { 1872 if (!Config->Is64 && FileSize > UINT32_MAX) { 1873 error("output file too large: " + Twine(FileSize) + " bytes"); 1874 return; 1875 } 1876 1877 unlinkAsync(Config->OutputFile); 1878 unsigned Flags = 0; 1879 if (!Config->Relocatable) 1880 Flags = FileOutputBuffer::F_executable; 1881 Expected<std::unique_ptr<FileOutputBuffer>> BufferOrErr = 1882 FileOutputBuffer::create(Config->OutputFile, FileSize, Flags); 1883 1884 if (!BufferOrErr) 1885 error("failed to open " + Config->OutputFile + ": " + 1886 llvm::toString(BufferOrErr.takeError())); 1887 else 1888 Buffer = std::move(*BufferOrErr); 1889 } 1890 1891 template <class ELFT> void Writer<ELFT>::writeSectionsBinary() { 1892 uint8_t *Buf = Buffer->getBufferStart(); 1893 for (OutputSection *Sec : OutputSections) 1894 if (Sec->Flags & SHF_ALLOC) 1895 Sec->writeTo<ELFT>(Buf + Sec->Offset); 1896 } 1897 1898 static void fillTrap(uint8_t *I, uint8_t *End) { 1899 for (; I + 4 <= End; I += 4) 1900 memcpy(I, &Target->TrapInstr, 4); 1901 } 1902 1903 // Fill the last page of executable segments with trap instructions 1904 // instead of leaving them as zero. Even though it is not required by any 1905 // standard, it is in general a good thing to do for security reasons. 1906 // 1907 // We'll leave other pages in segments as-is because the rest will be 1908 // overwritten by output sections. 1909 template <class ELFT> void Writer<ELFT>::writeTrapInstr() { 1910 if (Script->HasSectionsCommand) 1911 return; 1912 1913 // Fill the last page. 1914 uint8_t *Buf = Buffer->getBufferStart(); 1915 for (PhdrEntry *P : Phdrs) 1916 if (P->p_type == PT_LOAD && (P->p_flags & PF_X)) 1917 fillTrap(Buf + alignDown(P->p_offset + P->p_filesz, Target->PageSize), 1918 Buf + alignTo(P->p_offset + P->p_filesz, Target->PageSize)); 1919 1920 // Round up the file size of the last segment to the page boundary iff it is 1921 // an executable segment to ensure that other tools don't accidentally 1922 // trim the instruction padding (e.g. when stripping the file). 1923 PhdrEntry *Last = nullptr; 1924 for (PhdrEntry *P : Phdrs) 1925 if (P->p_type == PT_LOAD) 1926 Last = P; 1927 1928 if (Last && (Last->p_flags & PF_X)) 1929 Last->p_memsz = Last->p_filesz = alignTo(Last->p_filesz, Target->PageSize); 1930 } 1931 1932 // Write section contents to a mmap'ed file. 1933 template <class ELFT> void Writer<ELFT>::writeSections() { 1934 uint8_t *Buf = Buffer->getBufferStart(); 1935 1936 // PPC64 needs to process relocations in the .opd section 1937 // before processing relocations in code-containing sections. 1938 if (auto *OpdCmd = findSection(".opd")) { 1939 Out::Opd = OpdCmd; 1940 Out::OpdBuf = Buf + Out::Opd->Offset; 1941 OpdCmd->template writeTo<ELFT>(Buf + Out::Opd->Offset); 1942 } 1943 1944 OutputSection *EhFrameHdr = nullptr; 1945 if (InX::EhFrameHdr && !InX::EhFrameHdr->empty()) 1946 EhFrameHdr = InX::EhFrameHdr->getParent(); 1947 1948 // In -r or -emit-relocs mode, write the relocation sections first as in 1949 // ELf_Rel targets we might find out that we need to modify the relocated 1950 // section while doing it. 1951 for (OutputSection *Sec : OutputSections) 1952 if (Sec->Type == SHT_REL || Sec->Type == SHT_RELA) 1953 Sec->writeTo<ELFT>(Buf + Sec->Offset); 1954 1955 for (OutputSection *Sec : OutputSections) 1956 if (Sec != Out::Opd && Sec != EhFrameHdr && Sec->Type != SHT_REL && 1957 Sec->Type != SHT_RELA) 1958 Sec->writeTo<ELFT>(Buf + Sec->Offset); 1959 1960 // The .eh_frame_hdr depends on .eh_frame section contents, therefore 1961 // it should be written after .eh_frame is written. 1962 if (EhFrameHdr) 1963 EhFrameHdr->writeTo<ELFT>(Buf + EhFrameHdr->Offset); 1964 } 1965 1966 template <class ELFT> void Writer<ELFT>::writeBuildId() { 1967 if (!InX::BuildId || !InX::BuildId->getParent()) 1968 return; 1969 1970 // Compute a hash of all sections of the output file. 1971 uint8_t *Start = Buffer->getBufferStart(); 1972 uint8_t *End = Start + FileSize; 1973 InX::BuildId->writeBuildId({Start, End}); 1974 } 1975 1976 template void elf::writeResult<ELF32LE>(); 1977 template void elf::writeResult<ELF32BE>(); 1978 template void elf::writeResult<ELF64LE>(); 1979 template void elf::writeResult<ELF64BE>(); 1980 1981 template void elf::addReservedSymbols<ELF32LE>(); 1982 template void elf::addReservedSymbols<ELF32BE>(); 1983 template void elf::addReservedSymbols<ELF64LE>(); 1984 template void elf::addReservedSymbols<ELF64BE>(); 1985