| 1 | //===- InputFiles.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_WASM_INPUT_FILES_H |
| 10 | #define LLD_WASM_INPUT_FILES_H |
| 11 | |
| 12 | #include "Symbols.h" |
| 13 | #include "lld/Common/LLVM.h" |
| 14 | #include "llvm/ADT/DenseMap.h" |
| 15 | #include "llvm/ADT/DenseSet.h" |
| 16 | #include "llvm/LTO/LTO.h" |
| 17 | #include "llvm/Object/Wasm.h" |
| 18 | #include "llvm/Support/MemoryBuffer.h" |
| 19 | #include "llvm/TargetParser/Triple.h" |
| 20 | #include <optional> |
| 21 | #include <vector> |
| 22 | |
| 23 | namespace llvm { |
| 24 | class TarWriter; |
| 25 | } |
| 26 | |
| 27 | namespace lld { |
| 28 | namespace wasm { |
| 29 | |
| 30 | class InputChunk; |
| 31 | class InputFunction; |
| 32 | class InputSegment; |
| 33 | class InputGlobal; |
| 34 | class InputTag; |
| 35 | class InputTable; |
| 36 | class InputSection; |
| 37 | |
| 38 | // If --reproduce option is given, all input files are written |
| 39 | // to this tar archive. |
| 40 | extern std::unique_ptr<llvm::TarWriter> tar; |
| 41 | |
| 42 | class InputFile { |
| 43 | public: |
| 44 | enum Kind { |
| 45 | ObjectKind, |
| 46 | SharedKind, |
| 47 | BitcodeKind, |
| 48 | StubKind, |
| 49 | }; |
| 50 | |
| 51 | virtual ~InputFile() {} |
| 52 | |
| 53 | // Returns the filename. |
| 54 | StringRef getName() const { return mb.getBufferIdentifier(); } |
| 55 | |
| 56 | Kind kind() const { return fileKind; } |
| 57 | |
| 58 | // An archive file name if this file is created from an archive. |
| 59 | std::string archiveName; |
| 60 | |
| 61 | ArrayRef<Symbol *> getSymbols() const { return symbols; } |
| 62 | |
| 63 | MutableArrayRef<Symbol *> getMutableSymbols() { return symbols; } |
| 64 | |
| 65 | // An InputFile is considered live if any of the symbols defined by it |
| 66 | // are live. |
| 67 | void markLive() { live = true; } |
| 68 | bool isLive() const { return live; } |
| 69 | |
| 70 | // True if this is a relocatable object file/bitcode file in an ar archive |
| 71 | // or between --start-lib and --end-lib. |
| 72 | bool lazy = false; |
| 73 | |
| 74 | protected: |
| 75 | InputFile(Kind k, MemoryBufferRef m) |
| 76 | : mb(m), fileKind(k), live(!ctx.arg.gcSections) {} |
| 77 | |
| 78 | void checkArch(llvm::Triple::ArchType arch) const; |
| 79 | |
| 80 | MemoryBufferRef mb; |
| 81 | |
| 82 | // List of all symbols referenced or defined by this file. |
| 83 | std::vector<Symbol *> symbols; |
| 84 | |
| 85 | private: |
| 86 | const Kind fileKind; |
| 87 | bool live; |
| 88 | }; |
| 89 | |
| 90 | class WasmFileBase : public InputFile { |
| 91 | public: |
| 92 | explicit WasmFileBase(Kind k, MemoryBufferRef m); |
| 93 | |
| 94 | // Returns the underlying wasm file. |
| 95 | const WasmObjectFile *getWasmObj() const { return wasmObj.get(); } |
| 96 | |
| 97 | protected: |
| 98 | std::unique_ptr<WasmObjectFile> wasmObj; |
| 99 | }; |
| 100 | |
| 101 | // .o file (wasm object file) |
| 102 | class ObjFile : public WasmFileBase { |
| 103 | public: |
| 104 | ObjFile(MemoryBufferRef m, StringRef archiveName, bool lazy = false); |
| 105 | static bool classof(const InputFile *f) { return f->kind() == ObjectKind; } |
| 106 | |
| 107 | void parse(bool ignoreComdats = false); |
| 108 | void parseLazy(); |
| 109 | |
| 110 | uint32_t calcNewIndex(const WasmRelocation &reloc) const; |
| 111 | uint64_t calcNewValue(const WasmRelocation &reloc, uint64_t tombstone, |
| 112 | const InputChunk *chunk) const; |
| 113 | int64_t calcNewAddend(const WasmRelocation &reloc) const; |
| 114 | Symbol *getSymbol(const WasmRelocation &reloc) const { |
| 115 | return symbols[reloc.Index]; |
| 116 | }; |
| 117 | |
| 118 | const WasmSection *codeSection = nullptr; |
| 119 | const WasmSection *dataSection = nullptr; |
| 120 | |
| 121 | // Maps input type indices to output type indices |
| 122 | std::vector<uint32_t> typeMap; |
| 123 | std::vector<bool> typeIsUsed; |
| 124 | // Maps function indices to table indices |
| 125 | std::vector<uint32_t> tableEntries; |
| 126 | std::vector<uint32_t> tableEntriesRel; |
| 127 | std::vector<bool> keptComdats; |
| 128 | std::vector<InputChunk *> segments; |
| 129 | std::vector<InputFunction *> functions; |
| 130 | std::vector<InputGlobal *> globals; |
| 131 | std::vector<InputTag *> tags; |
| 132 | std::vector<InputTable *> tables; |
| 133 | std::vector<InputChunk *> customSections; |
| 134 | llvm::DenseMap<uint32_t, InputChunk *> customSectionsByIndex; |
| 135 | |
| 136 | Symbol *getSymbol(uint32_t index) const { return symbols[index]; } |
| 137 | FunctionSymbol *getFunctionSymbol(uint32_t index) const; |
| 138 | DataSymbol *getDataSymbol(uint32_t index) const; |
| 139 | GlobalSymbol *getGlobalSymbol(uint32_t index) const; |
| 140 | SectionSymbol *getSectionSymbol(uint32_t index) const; |
| 141 | TagSymbol *getTagSymbol(uint32_t index) const; |
| 142 | TableSymbol *getTableSymbol(uint32_t index) const; |
| 143 | |
| 144 | private: |
| 145 | Symbol *createDefined(const WasmSymbol &sym); |
| 146 | Symbol *createUndefined(const WasmSymbol &sym, bool isCalledDirectly); |
| 147 | |
| 148 | bool isExcludedByComdat(const InputChunk *chunk) const; |
| 149 | void addLegacyIndirectFunctionTableIfNeeded(uint32_t tableSymbolCount); |
| 150 | }; |
| 151 | |
| 152 | // .so file. |
| 153 | class SharedFile : public WasmFileBase { |
| 154 | public: |
| 155 | explicit SharedFile(MemoryBufferRef m) : WasmFileBase(SharedKind, m) {} |
| 156 | |
| 157 | void parse(); |
| 158 | |
| 159 | static bool classof(const InputFile *f) { return f->kind() == SharedKind; } |
| 160 | }; |
| 161 | |
| 162 | // .bc file |
| 163 | class BitcodeFile : public InputFile { |
| 164 | public: |
| 165 | BitcodeFile(MemoryBufferRef m, StringRef archiveName, |
| 166 | uint64_t offsetInArchive, bool lazy); |
| 167 | static bool classof(const InputFile *f) { return f->kind() == BitcodeKind; } |
| 168 | |
| 169 | void parse(StringRef symName); |
| 170 | void parseLazy(); |
| 171 | std::unique_ptr<llvm::lto::InputFile> obj; |
| 172 | |
| 173 | // Set to true once LTO is complete in order prevent further bitcode objects |
| 174 | // being added. |
| 175 | static bool doneLTO; |
| 176 | }; |
| 177 | |
| 178 | // Stub library (See docs/WebAssembly.rst) |
| 179 | class StubFile : public InputFile { |
| 180 | public: |
| 181 | explicit StubFile(MemoryBufferRef m) : InputFile(StubKind, m) {} |
| 182 | |
| 183 | static bool classof(const InputFile *f) { return f->kind() == StubKind; } |
| 184 | |
| 185 | void parse(); |
| 186 | |
| 187 | llvm::DenseMap<StringRef, std::vector<StringRef>> symbolDependencies; |
| 188 | }; |
| 189 | |
| 190 | // Will report a fatal() error if the input buffer is not a valid bitcode |
| 191 | // or wasm object file. |
| 192 | InputFile *createObjectFile(MemoryBufferRef mb, StringRef archiveName = "" , |
| 193 | uint64_t offsetInArchive = 0, bool lazy = false); |
| 194 | |
| 195 | // Opens a given file. |
| 196 | std::optional<MemoryBufferRef> readFile(StringRef path); |
| 197 | |
| 198 | std::string replaceThinLTOSuffix(StringRef path); |
| 199 | |
| 200 | } // namespace wasm |
| 201 | |
| 202 | std::string toString(const wasm::InputFile *file); |
| 203 | |
| 204 | } // namespace lld |
| 205 | |
| 206 | #endif |
| 207 | |