lld/MachO/InputSection.h

   1 //===- InputSection.h -------------------------------------------*- C++ -*-===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8
   9 #ifndef LLD_MACHO_INPUT_SECTION_H
  10 #define LLD_MACHO_INPUT_SECTION_H
  11
  12 #include "Config.h"
  13 #include "Relocations.h"
  14 #include "Symbols.h"
  15
  16 #include "lld/Common/LLVM.h"
  17 #include "lld/Common/Memory.h"
  18 #include "llvm/ADT/ArrayRef.h"
  19 #include "llvm/ADT/BitVector.h"
  20 #include "llvm/ADT/CachedHashString.h"
  21 #include "llvm/ADT/TinyPtrVector.h"
  22 #include "llvm/BinaryFormat/MachO.h"
  23
  24 namespace lld {
  25 namespace macho {
  26
  27 class InputFile;
  28 class OutputSection;
  29
  30 class InputSection {
  31 public:
  32   enum Kind : uint8_t {
  33     ConcatKind,
  34     CStringLiteralKind,
  35     WordLiteralKind,
  36   };
  37
  38   Kind kind() const { return sectionKind; }
  39   virtual ~InputSection() = default;
  40   virtual uint64_t getSize() const { return data.size(); }
  41   virtual bool empty() const { return data.empty(); }
  42   InputFile *getFile() const { return section.file; }
  43   StringRef getName() const { return section.name; }
  44   StringRef getSegName() const { return section.segname; }
  45   uint32_t getFlags() const { return section.flags; }
  46   uint64_t getFileSize() const;
  47   // Translates \p off -- an offset relative to this InputSection -- into an
  48   // offset from the beginning of its parent OutputSection.
  49   virtual uint64_t getOffset(uint64_t off) const = 0;
  50   // The offset from the beginning of the file.
  51   uint64_t getVA(uint64_t off) const;
  52   // Return a user-friendly string for use in diagnostics.
  53   // Format: /path/to/object.o:(symbol _func+0x123)
  54   std::string getLocation(uint64_t off) const;
  55   // Return the source line corresponding to an address, or the empty string.
  56   // Format: Source.cpp:123 (/path/to/Source.cpp:123)
  57   std::string getSourceLocation(uint64_t off) const;
  58   // Whether the data at \p off in this InputSection is live.
  59   virtual bool isLive(uint64_t off) const = 0;
  60   virtual void markLive(uint64_t off) = 0;
  61   virtual InputSection *canonical() { return this; }
  62   virtual const InputSection *canonical() const { return this; }
  63
  64 protected:
  65   InputSection(Kind kind, const Section &section, ArrayRef<uint8_t> data,
  66                uint32_t align)
  67       : sectionKind(kind), align(align), data(data), section(section) {}
  68
  69   InputSection(const InputSection &rhs)
  70       : sectionKind(rhs.sectionKind), align(rhs.align), data(rhs.data),
  71         section(rhs.section) {}
  72
  73   Kind sectionKind;
  74
  75 public:
  76   // is address assigned?
  77   bool isFinal = false;
  78   // keep the address of the symbol(s) in this section unique in the final
  79   // binary ?
  80   bool keepUnique = false;
  81   uint32_t align = 1;
  82
  83   OutputSection *parent = nullptr;
  84   ArrayRef<uint8_t> data;
  85   std::vector<Reloc> relocs;
  86   // The symbols that belong to this InputSection, sorted by value. With
  87   // .subsections_via_symbols, there is typically only one element here.
  88   llvm::TinyPtrVector<Defined *> symbols;
  89
  90 protected:
  91   const Section &section;
  92
  93   const Defined *getContainingSymbol(uint64_t off) const;
  94 };
  95
  96 // ConcatInputSections are combined into (Concat)OutputSections through simple
  97 // concatenation, in contrast with literal sections which may have their
  98 // contents merged before output.
  99 class ConcatInputSection final : public InputSection {
 100 public:
 101   ConcatInputSection(const Section &section, ArrayRef<uint8_t> data,
 102                      uint32_t align = 1)
 103       : InputSection(ConcatKind, section, data, align) {}
 104
 105   uint64_t getOffset(uint64_t off) const override { return outSecOff + off; }
 106   uint64_t getVA() const { return InputSection::getVA(0); }
 107   // ConcatInputSections are entirely live or dead, so the offset is irrelevant.
 108   bool isLive(uint64_t off) const override { return live; }
 109   void markLive(uint64_t off) override { live = true; }
 110   bool isCoalescedWeak() const { return wasCoalesced && symbols.empty(); }
 111   bool shouldOmitFromOutput() const { return !live || isCoalescedWeak(); }
 112   void writeTo(uint8_t *buf);
 113
 114   void foldIdentical(ConcatInputSection *redundant);
 115   ConcatInputSection *canonical() override {
 116     return replacement ? replacement : this;
 117   }
 118   const InputSection *canonical() const override {
 119     return replacement ? replacement : this;
 120   }
 121
 122   static bool classof(const InputSection *isec) {
 123     return isec->kind() == ConcatKind;
 124   }
 125
 126   // Points to the surviving section after this one is folded by ICF
 127   ConcatInputSection *replacement = nullptr;
 128   // Equivalence-class ID for ICF
 129   uint32_t icfEqClass[2] = {0, 0};
 130
 131   // With subsections_via_symbols, most symbols have their own InputSection,
 132   // and for weak symbols (e.g. from inline functions), only the
 133   // InputSection from one translation unit will make it to the output,
 134   // while all copies in other translation units are coalesced into the
 135   // first and not copied to the output.
 136   bool wasCoalesced = false;
 137   bool live = !config->deadStrip;
 138   bool hasCallSites = false;
 139   // This variable has two usages. Initially, it represents the input order.
 140   // After assignAddresses is called, it represents the offset from the
 141   // beginning of the output section this section was assigned to.
 142   uint64_t outSecOff = 0;
 143 };
 144
 145 // Initialize a fake InputSection that does not belong to any InputFile.
 146 ConcatInputSection *makeSyntheticInputSection(StringRef segName,
 147                                               StringRef sectName,
 148                                               uint32_t flags = 0,
 149                                               ArrayRef<uint8_t> data = {},
 150                                               uint32_t align = 1);
 151
 152 // Helper functions to make it easy to sprinkle asserts.
 153
 154 inline bool shouldOmitFromOutput(InputSection *isec) {
 155   return isa<ConcatInputSection>(isec) &&
 156          cast<ConcatInputSection>(isec)->shouldOmitFromOutput();
 157 }
 158
 159 inline bool isCoalescedWeak(InputSection *isec) {
 160   return isa<ConcatInputSection>(isec) &&
 161          cast<ConcatInputSection>(isec)->isCoalescedWeak();
 162 }
 163
 164 // We allocate a lot of these and binary search on them, so they should be as
 165 // compact as possible. Hence the use of 31 rather than 64 bits for the hash.
 166 struct StringPiece {
 167   // Offset from the start of the containing input section.
 168   uint32_t inSecOff;
 169   uint32_t live : 1;
 170   // Only set if deduplicating literals
 171   uint32_t hash : 31;
 172   // Offset from the start of the containing output section.
 173   uint64_t outSecOff = 0;
 174
 175   StringPiece(uint64_t off, uint32_t hash)
 176       : inSecOff(off), live(!config->deadStrip), hash(hash) {}
 177 };
 178
 179 static_assert(sizeof(StringPiece) == 16, "StringPiece is too big!");
 180
 181 // CStringInputSections are composed of multiple null-terminated string
 182 // literals, which we represent using StringPieces. These literals can be
 183 // deduplicated and tail-merged, so translating offsets between the input and
 184 // outputs sections is more complicated.
 185 //
 186 // NOTE: One significant difference between LLD and ld64 is that we merge all
 187 // cstring literals, even those referenced directly by non-private symbols.
 188 // ld64 is more conservative and does not do that. This was mostly done for
 189 // implementation simplicity; if we find programs that need the more
 190 // conservative behavior we can certainly implement that.
 191 class CStringInputSection final : public InputSection {
 192 public:
 193   CStringInputSection(const Section &section, ArrayRef<uint8_t> data,
 194                       uint32_t align)
 195       : InputSection(CStringLiteralKind, section, data, align) {}
 196   uint64_t getOffset(uint64_t off) const override;
 197   bool isLive(uint64_t off) const override { return getStringPiece(off).live; }
 198   void markLive(uint64_t off) override { getStringPiece(off).live = true; }
 199   // Find the StringPiece that contains this offset.
 200   StringPiece &getStringPiece(uint64_t off);
 201   const StringPiece &getStringPiece(uint64_t off) const;
 202   // Split at each null byte.
 203   void splitIntoPieces();
 204
 205   LLVM_ATTRIBUTE_ALWAYS_INLINE
 206   StringRef getStringRef(size_t i) const {
 207     size_t begin = pieces[i].inSecOff;
 208     size_t end =
 209         (pieces.size() - 1 == i) ? data.size() : pieces[i + 1].inSecOff;
 210     return toStringRef(data.slice(begin, end - begin));
 211   }
 212
 213   // Returns i'th piece as a CachedHashStringRef. This function is very hot when
 214   // string merging is enabled, so we want to inline.
 215   LLVM_ATTRIBUTE_ALWAYS_INLINE
 216   llvm::CachedHashStringRef getCachedHashStringRef(size_t i) const {
 217     assert(config->dedupLiterals);
 218     return {getStringRef(i), pieces[i].hash};
 219   }
 220
 221   static bool classof(const InputSection *isec) {
 222     return isec->kind() == CStringLiteralKind;
 223   }
 224
 225   std::vector<StringPiece> pieces;
 226 };
 227
 228 class WordLiteralInputSection final : public InputSection {
 229 public:
 230   WordLiteralInputSection(const Section &section, ArrayRef<uint8_t> data,
 231                           uint32_t align);
 232   uint64_t getOffset(uint64_t off) const override;
 233   bool isLive(uint64_t off) const override {
 234     return live[off >> power2LiteralSize];
 235   }
 236   void markLive(uint64_t off) override {
 237     live[off >> power2LiteralSize] = true;
 238   }
 239
 240   static bool classof(const InputSection *isec) {
 241     return isec->kind() == WordLiteralKind;
 242   }
 243
 244 private:
 245   unsigned power2LiteralSize;
 246   // The liveness of data[off] is tracked by live[off >> power2LiteralSize].
 247   llvm::BitVector live;
 248 };
 249
 250 inline uint8_t sectionType(uint32_t flags) {
 251   return flags & llvm::MachO::SECTION_TYPE;
 252 }
 253
 254 inline bool isZeroFill(uint32_t flags) {
 255   return llvm::MachO::isVirtualSection(sectionType(flags));
 256 }
 257
 258 inline bool isThreadLocalVariables(uint32_t flags) {
 259   return sectionType(flags) == llvm::MachO::S_THREAD_LOCAL_VARIABLES;
 260 }
 261
 262 // These sections contain the data for initializing thread-local variables.
 263 inline bool isThreadLocalData(uint32_t flags) {
 264   return sectionType(flags) == llvm::MachO::S_THREAD_LOCAL_REGULAR ||
 265          sectionType(flags) == llvm::MachO::S_THREAD_LOCAL_ZEROFILL;
 266 }
 267
 268 inline bool isDebugSection(uint32_t flags) {
 269   return (flags & llvm::MachO::SECTION_ATTRIBUTES_USR) ==
 270          llvm::MachO::S_ATTR_DEBUG;
 271 }
 272
 273 inline bool isWordLiteralSection(uint32_t flags) {
 274   return sectionType(flags) == llvm::MachO::S_4BYTE_LITERALS ||
 275          sectionType(flags) == llvm::MachO::S_8BYTE_LITERALS ||
 276          sectionType(flags) == llvm::MachO::S_16BYTE_LITERALS;
 277 }
 278
 279 bool isCodeSection(const InputSection *);
 280 bool isCfStringSection(const InputSection *);
 281 bool isClassRefsSection(const InputSection *);
 282 bool isEhFrameSection(const InputSection *);
 283
 284 extern std::vector<ConcatInputSection *> inputSections;
 285
 286 namespace section_names {
 287
 288 constexpr const char authGot[] = "__auth_got";
 289 constexpr const char authPtr[] = "__auth_ptr";
 290 constexpr const char binding[] = "__binding";
 291 constexpr const char bitcodeBundle[] = "__bundle";
 292 constexpr const char cString[] = "__cstring";
 293 constexpr const char cfString[] = "__cfstring";
 294 constexpr const char cgProfile[] = "__cg_profile";
 295 constexpr const char codeSignature[] = "__code_signature";
 296 constexpr const char common[] = "__common";
 297 constexpr const char compactUnwind[] = "__compact_unwind";
 298 constexpr const char data[] = "__data";
 299 constexpr const char debugAbbrev[] = "__debug_abbrev";
 300 constexpr const char debugInfo[] = "__debug_info";
 301 constexpr const char debugLine[] = "__debug_line";
 302 constexpr const char debugStr[] = "__debug_str";
 303 constexpr const char ehFrame[] = "__eh_frame";
 304 constexpr const char gccExceptTab[] = "__gcc_except_tab";
 305 constexpr const char export_[] = "__export";
 306 constexpr const char dataInCode[] = "__data_in_code";
 307 constexpr const char functionStarts[] = "__func_starts";
 308 constexpr const char got[] = "__got";
 309 constexpr const char header[] = "__mach_header";
 310 constexpr const char indirectSymbolTable[] = "__ind_sym_tab";
 311 constexpr const char const_[] = "__const";
 312 constexpr const char lazySymbolPtr[] = "__la_symbol_ptr";
 313 constexpr const char lazyBinding[] = "__lazy_binding";
 314 constexpr const char literals[] = "__literals";
 315 constexpr const char moduleInitFunc[] = "__mod_init_func";
 316 constexpr const char moduleTermFunc[] = "__mod_term_func";
 317 constexpr const char nonLazySymbolPtr[] = "__nl_symbol_ptr";
 318 constexpr const char objcCatList[] = "__objc_catlist";
 319 constexpr const char objcClassList[] = "__objc_classlist";
 320 constexpr const char objcClassRefs[] = "__objc_classrefs";
 321 constexpr const char objcConst[] = "__objc_const";
 322 constexpr const char objcImageInfo[] = "__objc_imageinfo";
 323 constexpr const char objcNonLazyCatList[] = "__objc_nlcatlist";
 324 constexpr const char objcNonLazyClassList[] = "__objc_nlclslist";
 325 constexpr const char objcProtoList[] = "__objc_protolist";
 326 constexpr const char pageZero[] = "__pagezero";
 327 constexpr const char pointers[] = "__pointers";
 328 constexpr const char rebase[] = "__rebase";
 329 constexpr const char staticInit[] = "__StaticInit";
 330 constexpr const char stringTable[] = "__string_table";
 331 constexpr const char stubHelper[] = "__stub_helper";
 332 constexpr const char stubs[] = "__stubs";
 333 constexpr const char swift[] = "__swift";
 334 constexpr const char symbolTable[] = "__symbol_table";
 335 constexpr const char textCoalNt[] = "__textcoal_nt";
 336 constexpr const char text[] = "__text";
 337 constexpr const char threadPtrs[] = "__thread_ptrs";
 338 constexpr const char threadVars[] = "__thread_vars";
 339 constexpr const char unwindInfo[] = "__unwind_info";
 340 constexpr const char weakBinding[] = "__weak_binding";
 341 constexpr const char zeroFill[] = "__zerofill";
 342 constexpr const char addrSig[] = "__llvm_addrsig";
 343
 344 } // namespace section_names
 345
 346 } // namespace macho
 347
 348 std::string toString(const macho::InputSection *);
 349
 350 } // namespace lld
 351
 352 #endif