| 1 | //===- MinGW.cpp ----------------------------------------------------------===// |
| 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 | #include "MinGW.h" |
| 10 | #include "COFFLinkerContext.h" |
| 11 | #include "Driver.h" |
| 12 | #include "InputFiles.h" |
| 13 | #include "SymbolTable.h" |
| 14 | #include "llvm/ADT/DenseMap.h" |
| 15 | #include "llvm/ADT/DenseSet.h" |
| 16 | #include "llvm/Support/Parallel.h" |
| 17 | #include "llvm/Support/Path.h" |
| 18 | #include "llvm/Support/TimeProfiler.h" |
| 19 | #include "llvm/Support/raw_ostream.h" |
| 20 | |
| 21 | using namespace llvm; |
| 22 | using namespace llvm::COFF; |
| 23 | using namespace lld; |
| 24 | using namespace lld::coff; |
| 25 | |
| 26 | AutoExporter::AutoExporter( |
| 27 | SymbolTable &symtab, const llvm::DenseSet<StringRef> &manualExcludeSymbols) |
| 28 | : manualExcludeSymbols(manualExcludeSymbols), symtab(symtab) { |
| 29 | excludeLibs = { |
| 30 | "libgcc" , |
| 31 | "libgcc_s" , |
| 32 | "libstdc++" , |
| 33 | "libmingw32" , |
| 34 | "libmingwex" , |
| 35 | "libg2c" , |
| 36 | "libsupc++" , |
| 37 | "libobjc" , |
| 38 | "libgcj" , |
| 39 | "libclang_rt.builtins" , |
| 40 | "libclang_rt.builtins-aarch64" , |
| 41 | "libclang_rt.builtins-arm" , |
| 42 | "libclang_rt.builtins-i386" , |
| 43 | "libclang_rt.builtins-x86_64" , |
| 44 | "libclang_rt.profile" , |
| 45 | "libclang_rt.profile-aarch64" , |
| 46 | "libclang_rt.profile-arm" , |
| 47 | "libclang_rt.profile-i386" , |
| 48 | "libclang_rt.profile-x86_64" , |
| 49 | "libcygwin" , |
| 50 | "libmsys-2.0" , |
| 51 | "libc++" , |
| 52 | "libc++abi" , |
| 53 | "libflang_rt.runtime" , |
| 54 | "libunwind" , |
| 55 | "libmsvcrt" , |
| 56 | "libmsvcrt-os" , |
| 57 | "libucrtbase" , |
| 58 | "libucrt" , |
| 59 | "libucrtapp" , |
| 60 | "libpthread" , |
| 61 | "libwinpthread" , |
| 62 | }; |
| 63 | |
| 64 | excludeObjects = { |
| 65 | "crt0.o" , "crt1.o" , "crt1u.o" , "crt2.o" , "crt2u.o" , "dllcrt1.o" , |
| 66 | "dllcrt2.o" , "gcrt0.o" , "gcrt1.o" , "gcrt2.o" , "crtbegin.o" , "crtend.o" , |
| 67 | }; |
| 68 | |
| 69 | excludeSymbolPrefixes = { |
| 70 | // Import symbols |
| 71 | "__imp_" , |
| 72 | "__IMPORT_DESCRIPTOR_" , |
| 73 | // Extra import symbols from GNU import libraries |
| 74 | "__nm_" , |
| 75 | // C++ symbols |
| 76 | "__rtti_" , |
| 77 | "__builtin_" , |
| 78 | // Artificial symbols such as .refptr |
| 79 | "." , |
| 80 | // profile generate symbols |
| 81 | "__profc_" , |
| 82 | "__profd_" , |
| 83 | "__profvp_" , |
| 84 | }; |
| 85 | |
| 86 | excludeSymbolSuffixes = { |
| 87 | "_iname" , |
| 88 | "_NULL_THUNK_DATA" , |
| 89 | }; |
| 90 | |
| 91 | if (symtab.machine == I386) { |
| 92 | excludeSymbols = { |
| 93 | "__NULL_IMPORT_DESCRIPTOR" , |
| 94 | "__pei386_runtime_relocator" , |
| 95 | "_do_pseudo_reloc" , |
| 96 | "_impure_ptr" , |
| 97 | "__impure_ptr" , |
| 98 | "__fmode" , |
| 99 | "_environ" , |
| 100 | "___dso_handle" , |
| 101 | "__load_config_used" , |
| 102 | // These are the MinGW names that differ from the standard |
| 103 | // ones (lacking an extra underscore). |
| 104 | "_DllMain@12" , |
| 105 | "_DllEntryPoint@12" , |
| 106 | "_DllMainCRTStartup@12" , |
| 107 | }; |
| 108 | excludeSymbolPrefixes.insert(key: "__head_" ); |
| 109 | } else { |
| 110 | excludeSymbols = { |
| 111 | "__NULL_IMPORT_DESCRIPTOR" , |
| 112 | "_pei386_runtime_relocator" , |
| 113 | "do_pseudo_reloc" , |
| 114 | "impure_ptr" , |
| 115 | "_impure_ptr" , |
| 116 | "_fmode" , |
| 117 | "environ" , |
| 118 | "__dso_handle" , |
| 119 | "_load_config_used" , |
| 120 | // These are the MinGW names that differ from the standard |
| 121 | // ones (lacking an extra underscore). |
| 122 | "DllMain" , |
| 123 | "DllEntryPoint" , |
| 124 | "DllMainCRTStartup" , |
| 125 | }; |
| 126 | excludeSymbolPrefixes.insert(key: "_head_" ); |
| 127 | } |
| 128 | if (symtab.isEC()) { |
| 129 | excludeSymbols.insert(key: "__chpe_metadata" ); |
| 130 | excludeSymbolPrefixes.insert(key: "__os_arm64x_" ); |
| 131 | } |
| 132 | } |
| 133 | |
| 134 | void AutoExporter::addWholeArchive(StringRef path) { |
| 135 | StringRef libName = sys::path::filename(path); |
| 136 | // Drop the file extension, to match the processing below. |
| 137 | libName = libName.substr(Start: 0, N: libName.rfind(C: '.')); |
| 138 | excludeLibs.erase(Key: libName); |
| 139 | } |
| 140 | |
| 141 | void AutoExporter::addExcludedSymbol(StringRef symbol) { |
| 142 | excludeSymbols.insert(key: symbol); |
| 143 | } |
| 144 | |
| 145 | bool AutoExporter::shouldExport(Defined *sym) const { |
| 146 | if (!sym || !sym->getChunk()) |
| 147 | return false; |
| 148 | |
| 149 | // Only allow the symbol kinds that make sense to export; in particular, |
| 150 | // disallow import symbols. |
| 151 | if (!isa<DefinedRegular>(Val: sym) && !isa<DefinedCommon>(Val: sym)) |
| 152 | return false; |
| 153 | if (excludeSymbols.count(Key: sym->getName()) || manualExcludeSymbols.count(V: sym->getName())) |
| 154 | return false; |
| 155 | |
| 156 | for (StringRef prefix : excludeSymbolPrefixes.keys()) |
| 157 | if (sym->getName().starts_with(Prefix: prefix)) |
| 158 | return false; |
| 159 | for (StringRef suffix : excludeSymbolSuffixes.keys()) |
| 160 | if (sym->getName().ends_with(Suffix: suffix)) |
| 161 | return false; |
| 162 | |
| 163 | // If a corresponding __imp_ symbol exists and is defined, don't export it. |
| 164 | if (symtab.find(name: ("__imp_" + sym->getName()).str())) |
| 165 | return false; |
| 166 | |
| 167 | // Check that file is non-null before dereferencing it, symbols not |
| 168 | // originating in regular object files probably shouldn't be exported. |
| 169 | if (!sym->getFile()) |
| 170 | return false; |
| 171 | |
| 172 | StringRef libName = sys::path::filename(path: sym->getFile()->parentName); |
| 173 | |
| 174 | // Drop the file extension. |
| 175 | libName = libName.substr(Start: 0, N: libName.rfind(C: '.')); |
| 176 | if (!libName.empty()) |
| 177 | return !excludeLibs.count(Key: libName); |
| 178 | |
| 179 | StringRef fileName = sys::path::filename(path: sym->getFile()->getName()); |
| 180 | return !excludeObjects.count(Key: fileName); |
| 181 | } |
| 182 | |
| 183 | void lld::coff::writeDefFile(COFFLinkerContext &ctx, StringRef name, |
| 184 | const std::vector<Export> &exports) { |
| 185 | llvm::TimeTraceScope timeScope("Write .def file" ); |
| 186 | std::error_code ec; |
| 187 | raw_fd_ostream os(name, ec, sys::fs::OF_None); |
| 188 | if (ec) |
| 189 | Fatal(ctx) << "cannot open " << name << ": " << ec.message(); |
| 190 | |
| 191 | os << "EXPORTS\n" ; |
| 192 | for (const Export &e : exports) { |
| 193 | os << " " << e.exportName << " " |
| 194 | << "@" << e.ordinal; |
| 195 | if (auto *def = dyn_cast_or_null<Defined>(Val: e.sym)) { |
| 196 | if (def && def->getChunk() && |
| 197 | !(def->getChunk()->getOutputCharacteristics() & IMAGE_SCN_MEM_EXECUTE)) |
| 198 | os << " DATA" ; |
| 199 | } |
| 200 | os << "\n" ; |
| 201 | } |
| 202 | } |
| 203 | |
| 204 | static StringRef mangle(Twine sym, MachineTypes machine) { |
| 205 | assert(machine != IMAGE_FILE_MACHINE_UNKNOWN); |
| 206 | if (machine == I386) |
| 207 | return saver().save(S: "_" + sym); |
| 208 | return saver().save(S: sym); |
| 209 | } |
| 210 | |
| 211 | // Handles -wrap option. |
| 212 | // |
| 213 | // This function instantiates wrapper symbols. At this point, they seem |
| 214 | // like they are not being used at all, so we explicitly set some flags so |
| 215 | // that LTO won't eliminate them. |
| 216 | void lld::coff::addWrappedSymbols(SymbolTable &symtab, |
| 217 | opt::InputArgList &args) { |
| 218 | std::vector<WrappedSymbol> v; |
| 219 | DenseSet<StringRef> seen; |
| 220 | |
| 221 | for (auto *arg : args.filtered(Ids: OPT_wrap)) { |
| 222 | StringRef name = arg->getValue(); |
| 223 | if (!seen.insert(V: name).second) |
| 224 | continue; |
| 225 | |
| 226 | Symbol *sym = symtab.findUnderscore(name); |
| 227 | if (!sym) |
| 228 | continue; |
| 229 | |
| 230 | Symbol *real = |
| 231 | symtab.addUndefined(name: mangle(sym: "__real_" + name, machine: symtab.machine)); |
| 232 | Symbol *wrap = |
| 233 | symtab.addUndefined(name: mangle(sym: "__wrap_" + name, machine: symtab.machine)); |
| 234 | v.push_back(x: {.sym: sym, .real: real, .wrap: wrap}); |
| 235 | |
| 236 | // These symbols may seem undefined initially, but don't bail out |
| 237 | // at symtab.reportUnresolvable() due to them, but let wrapSymbols |
| 238 | // below sort things out before checking finally with |
| 239 | // symtab.resolveRemainingUndefines(). |
| 240 | sym->deferUndefined = true; |
| 241 | real->deferUndefined = true; |
| 242 | // We want to tell LTO not to inline symbols to be overwritten |
| 243 | // because LTO doesn't know the final symbol contents after renaming. |
| 244 | real->canInline = false; |
| 245 | sym->canInline = false; |
| 246 | |
| 247 | // Tell LTO not to eliminate these symbols. |
| 248 | sym->isUsedInRegularObj = true; |
| 249 | if (!isa<Undefined>(Val: wrap)) |
| 250 | wrap->isUsedInRegularObj = true; |
| 251 | } |
| 252 | symtab.wrapped = std::move(v); |
| 253 | } |
| 254 | |
| 255 | // Do renaming for -wrap by updating pointers to symbols. |
| 256 | // |
| 257 | // When this function is executed, only InputFiles and symbol table |
| 258 | // contain pointers to symbol objects. We visit them to replace pointers, |
| 259 | // so that wrapped symbols are swapped as instructed by the command line. |
| 260 | void lld::coff::wrapSymbols(SymbolTable &symtab) { |
| 261 | DenseMap<Symbol *, Symbol *> map; |
| 262 | for (const WrappedSymbol &w : symtab.wrapped) { |
| 263 | map[w.sym] = w.wrap; |
| 264 | map[w.real] = w.sym; |
| 265 | if (Defined *d = dyn_cast<Defined>(Val: w.wrap)) { |
| 266 | Symbol *imp = symtab.find(name: ("__imp_" + w.sym->getName()).str()); |
| 267 | // Create a new defined local import for the wrap symbol. If |
| 268 | // no imp prefixed symbol existed, there's no need for it. |
| 269 | // (We can't easily distinguish whether any object file actually |
| 270 | // referenced it or not, though.) |
| 271 | if (imp) { |
| 272 | DefinedLocalImport *wrapimp = make<DefinedLocalImport>( |
| 273 | args&: symtab.ctx, args: saver().save(S: "__imp_" + w.wrap->getName()), args&: d); |
| 274 | symtab.localImportChunks.push_back(x: wrapimp->getChunk()); |
| 275 | map[imp] = wrapimp; |
| 276 | } |
| 277 | } |
| 278 | } |
| 279 | |
| 280 | // Update pointers in input files. |
| 281 | parallelForEach(R&: symtab.ctx.objFileInstances, Fn: [&](ObjFile *file) { |
| 282 | MutableArrayRef<Symbol *> syms = file->getMutableSymbols(); |
| 283 | for (auto &sym : syms) |
| 284 | if (Symbol *s = map.lookup(Val: sym)) |
| 285 | sym = s; |
| 286 | }); |
| 287 | } |
| 288 | |