| 1 | //===----- MachOLinkGraphBuilder.h - MachO LinkGraph builder ----*- 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 | // Generic MachO LinkGraph building code. |
| 10 | // |
| 11 | //===----------------------------------------------------------------------===// |
| 12 | |
| 13 | #ifndef LIB_EXECUTIONENGINE_JITLINK_MACHOLINKGRAPHBUILDER_H |
| 14 | #define LIB_EXECUTIONENGINE_JITLINK_MACHOLINKGRAPHBUILDER_H |
| 15 | |
| 16 | #include "llvm/ADT/DenseMap.h" |
| 17 | #include "llvm/ADT/StringMap.h" |
| 18 | #include "llvm/ExecutionEngine/JITLink/JITLink.h" |
| 19 | #include "llvm/Object/MachO.h" |
| 20 | |
| 21 | #include "EHFrameSupportImpl.h" |
| 22 | #include "JITLinkGeneric.h" |
| 23 | |
| 24 | namespace llvm { |
| 25 | namespace jitlink { |
| 26 | |
| 27 | class MachOLinkGraphBuilder { |
| 28 | public: |
| 29 | virtual ~MachOLinkGraphBuilder(); |
| 30 | Expected<std::unique_ptr<LinkGraph>> buildGraph(); |
| 31 | |
| 32 | protected: |
| 33 | |
| 34 | struct NormalizedSymbol { |
| 35 | friend class MachOLinkGraphBuilder; |
| 36 | |
| 37 | private: |
| 38 | NormalizedSymbol(std::optional<StringRef> Name, uint64_t Value, |
| 39 | uint8_t Type, uint8_t Sect, uint16_t Desc, Linkage L, |
| 40 | Scope S) |
| 41 | : Name(Name), Value(Value), Type(Type), Sect(Sect), Desc(Desc), L(L), |
| 42 | S(S) { |
| 43 | assert((!Name || !Name->empty()) && "Name must be none or non-empty" ); |
| 44 | } |
| 45 | |
| 46 | public: |
| 47 | NormalizedSymbol(const NormalizedSymbol &) = delete; |
| 48 | NormalizedSymbol &operator=(const NormalizedSymbol &) = delete; |
| 49 | NormalizedSymbol(NormalizedSymbol &&) = delete; |
| 50 | NormalizedSymbol &operator=(NormalizedSymbol &&) = delete; |
| 51 | |
| 52 | std::optional<StringRef> Name; |
| 53 | uint64_t Value = 0; |
| 54 | uint8_t Type = 0; |
| 55 | uint8_t Sect = 0; |
| 56 | uint16_t Desc = 0; |
| 57 | Linkage L = Linkage::Strong; |
| 58 | Scope S = Scope::Default; |
| 59 | Symbol *GraphSymbol = nullptr; |
| 60 | }; |
| 61 | |
| 62 | // Normalized section representation. Section and segment names are guaranteed |
| 63 | // to be null-terminated, hence the extra bytes on SegName and SectName. |
| 64 | class NormalizedSection { |
| 65 | friend class MachOLinkGraphBuilder; |
| 66 | |
| 67 | private: |
| 68 | NormalizedSection() = default; |
| 69 | |
| 70 | public: |
| 71 | char SectName[17]; |
| 72 | char SegName[17]; |
| 73 | orc::ExecutorAddr Address; |
| 74 | uint64_t Size = 0; |
| 75 | uint64_t Alignment = 0; |
| 76 | uint32_t Flags = 0; |
| 77 | const char *Data = nullptr; |
| 78 | Section *GraphSection = nullptr; |
| 79 | std::map<orc::ExecutorAddr, Symbol *> CanonicalSymbols; |
| 80 | }; |
| 81 | |
| 82 | using SectionParserFunction = std::function<Error(NormalizedSection &S)>; |
| 83 | |
| 84 | MachOLinkGraphBuilder(const object::MachOObjectFile &Obj, |
| 85 | std::shared_ptr<orc::SymbolStringPool> SSP, Triple TT, |
| 86 | SubtargetFeatures Features, |
| 87 | LinkGraph::GetEdgeKindNameFunction GetEdgeKindName); |
| 88 | LinkGraph &getGraph() const { return *G; } |
| 89 | |
| 90 | const object::MachOObjectFile &getObject() const { return Obj; } |
| 91 | |
| 92 | void addCustomSectionParser(StringRef SectionName, |
| 93 | SectionParserFunction Parse); |
| 94 | |
| 95 | virtual Error addRelocations() = 0; |
| 96 | |
| 97 | /// Create a symbol. |
| 98 | template <typename... ArgTs> |
| 99 | NormalizedSymbol &createNormalizedSymbol(ArgTs &&... Args) { |
| 100 | NormalizedSymbol *Sym = reinterpret_cast<NormalizedSymbol *>( |
| 101 | Allocator.Allocate<NormalizedSymbol>()); |
| 102 | new (Sym) NormalizedSymbol(std::forward<ArgTs>(Args)...); |
| 103 | return *Sym; |
| 104 | } |
| 105 | |
| 106 | /// Index is zero-based (MachO section indexes are usually one-based) and |
| 107 | /// assumed to be in-range. Client is responsible for checking. |
| 108 | NormalizedSection &getSectionByIndex(unsigned Index) { |
| 109 | auto I = IndexToSection.find(Val: Index); |
| 110 | assert(I != IndexToSection.end() && "No section recorded at index" ); |
| 111 | return I->second; |
| 112 | } |
| 113 | |
| 114 | /// Try to get the section at the given index. Will return an error if the |
| 115 | /// given index is out of range, or if no section has been added for the given |
| 116 | /// index. |
| 117 | Expected<NormalizedSection &> findSectionByIndex(unsigned Index) { |
| 118 | auto I = IndexToSection.find(Val: Index); |
| 119 | if (I == IndexToSection.end()) |
| 120 | return make_error<JITLinkError>(Args: "No section recorded for index " + |
| 121 | formatv(Fmt: "{0:d}" , Vals&: Index)); |
| 122 | return I->second; |
| 123 | } |
| 124 | |
| 125 | /// Try to get the symbol at the given index. Will return an error if the |
| 126 | /// given index is out of range, or if no symbol has been added for the given |
| 127 | /// index. |
| 128 | Expected<NormalizedSymbol &> findSymbolByIndex(uint64_t Index) { |
| 129 | auto I = IndexToSymbol.find(Val: Index); |
| 130 | if (I == IndexToSymbol.end()) |
| 131 | return make_error<JITLinkError>(Args: "No symbol at index " + |
| 132 | formatv(Fmt: "{0:d}" , Vals&: Index)); |
| 133 | assert(I->second && "Null symbol at index" ); |
| 134 | return *I->second; |
| 135 | } |
| 136 | |
| 137 | /// Returns the symbol with the highest address not greater than the search |
| 138 | /// address, or null if no such symbol exists. |
| 139 | Symbol *getSymbolByAddress(NormalizedSection &NSec, |
| 140 | orc::ExecutorAddr Address) { |
| 141 | auto I = NSec.CanonicalSymbols.upper_bound(x: Address); |
| 142 | if (I == NSec.CanonicalSymbols.begin()) |
| 143 | return nullptr; |
| 144 | return std::prev(x: I)->second; |
| 145 | } |
| 146 | |
| 147 | /// Returns the symbol with the highest address not greater than the search |
| 148 | /// address, or an error if no such symbol exists. |
| 149 | Expected<Symbol &> findSymbolByAddress(NormalizedSection &NSec, |
| 150 | orc::ExecutorAddr Address) { |
| 151 | auto *Sym = getSymbolByAddress(NSec, Address); |
| 152 | if (Sym) |
| 153 | if (Address <= Sym->getAddress() + Sym->getSize()) |
| 154 | return *Sym; |
| 155 | return make_error<JITLinkError>(Args: "No symbol covering address " + |
| 156 | formatv(Fmt: "{0:x16}" , Vals&: Address)); |
| 157 | } |
| 158 | |
| 159 | static Linkage getLinkage(uint16_t Desc); |
| 160 | static Scope getScope(StringRef Name, uint8_t Type); |
| 161 | static bool isAltEntry(const NormalizedSymbol &NSym); |
| 162 | |
| 163 | static bool isDebugSection(const NormalizedSection &NSec); |
| 164 | static bool isZeroFillSection(const NormalizedSection &NSec); |
| 165 | |
| 166 | MachO::relocation_info |
| 167 | getRelocationInfo(const object::relocation_iterator RelItr) { |
| 168 | MachO::any_relocation_info ARI = |
| 169 | getObject().getRelocation(Rel: RelItr->getRawDataRefImpl()); |
| 170 | MachO::relocation_info RI; |
| 171 | RI.r_address = ARI.r_word0; |
| 172 | RI.r_symbolnum = ARI.r_word1 & 0xffffff; |
| 173 | RI.r_pcrel = (ARI.r_word1 >> 24) & 1; |
| 174 | RI.r_length = (ARI.r_word1 >> 25) & 3; |
| 175 | RI.r_extern = (ARI.r_word1 >> 27) & 1; |
| 176 | RI.r_type = (ARI.r_word1 >> 28); |
| 177 | return RI; |
| 178 | } |
| 179 | |
| 180 | private: |
| 181 | static unsigned getPointerSize(const object::MachOObjectFile &Obj); |
| 182 | static llvm::endianness getEndianness(const object::MachOObjectFile &Obj); |
| 183 | |
| 184 | void setCanonicalSymbol(NormalizedSection &NSec, Symbol &Sym) { |
| 185 | auto *&CanonicalSymEntry = NSec.CanonicalSymbols[Sym.getAddress()]; |
| 186 | // There should be no symbol at this address, or, if there is, |
| 187 | // it should be a zero-sized symbol from an empty section (which |
| 188 | // we can safely override). |
| 189 | assert((!CanonicalSymEntry || CanonicalSymEntry->getSize() == 0) && |
| 190 | "Duplicate canonical symbol at address" ); |
| 191 | CanonicalSymEntry = &Sym; |
| 192 | } |
| 193 | |
| 194 | Section &getCommonSection(); |
| 195 | void addSectionStartSymAndBlock(unsigned SecIndex, Section &GraphSec, |
| 196 | orc::ExecutorAddr Address, const char *Data, |
| 197 | orc::ExecutorAddrDiff Size, |
| 198 | uint32_t Alignment, bool IsLive); |
| 199 | |
| 200 | Error createNormalizedSections(); |
| 201 | Error createNormalizedSymbols(); |
| 202 | |
| 203 | /// Create graph blocks and symbols for externals, absolutes, commons and |
| 204 | /// all defined symbols in sections without custom parsers. |
| 205 | Error graphifyRegularSymbols(); |
| 206 | |
| 207 | /// Create and return a graph symbol for the given normalized symbol. |
| 208 | /// |
| 209 | /// NSym's GraphSymbol member will be updated to point at the newly created |
| 210 | /// symbol. |
| 211 | Symbol &createStandardGraphSymbol(NormalizedSymbol &Sym, Block &B, |
| 212 | size_t Size, bool IsText, |
| 213 | bool IsNoDeadStrip, bool IsCanonical); |
| 214 | |
| 215 | /// Create graph blocks and symbols for all sections. |
| 216 | Error graphifySectionsWithCustomParsers(); |
| 217 | |
| 218 | /// Graphify cstring section. |
| 219 | Error graphifyCStringSection(NormalizedSection &NSec, |
| 220 | std::vector<NormalizedSymbol *> NSyms); |
| 221 | |
| 222 | // Put the BumpPtrAllocator first so that we don't free any of the underlying |
| 223 | // memory until the Symbol/Addressable destructors have been run. |
| 224 | BumpPtrAllocator Allocator; |
| 225 | |
| 226 | const object::MachOObjectFile &Obj; |
| 227 | std::unique_ptr<LinkGraph> G; |
| 228 | |
| 229 | bool SubsectionsViaSymbols = false; |
| 230 | DenseMap<unsigned, NormalizedSection> IndexToSection; |
| 231 | Section *CommonSection = nullptr; |
| 232 | |
| 233 | DenseMap<uint32_t, NormalizedSymbol *> IndexToSymbol; |
| 234 | StringMap<SectionParserFunction> CustomSectionParserFunctions; |
| 235 | }; |
| 236 | |
| 237 | } // end namespace jitlink |
| 238 | } // end namespace llvm |
| 239 | |
| 240 | #endif // LIB_EXECUTIONENGINE_JITLINK_MACHOLINKGRAPHBUILDER_H |
| 241 | |