1 //===- lib/Linker/IRMover.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 "llvm/Linker/IRMover.h" 10 #include "LinkDiagnosticInfo.h" 11 #include "llvm/ADT/SetVector.h" 12 #include "llvm/ADT/SmallString.h" 13 #include "llvm/ADT/Triple.h" 14 #include "llvm/IR/Constants.h" 15 #include "llvm/IR/DebugInfo.h" 16 #include "llvm/IR/DiagnosticPrinter.h" 17 #include "llvm/IR/GVMaterializer.h" 18 #include "llvm/IR/Intrinsics.h" 19 #include "llvm/IR/TypeFinder.h" 20 #include "llvm/Support/Error.h" 21 #include "llvm/Transforms/Utils/Cloning.h" 22 #include <utility> 23 using namespace llvm; 24 25 //===----------------------------------------------------------------------===// 26 // TypeMap implementation. 27 //===----------------------------------------------------------------------===// 28 29 namespace { 30 class TypeMapTy : public ValueMapTypeRemapper { 31 /// This is a mapping from a source type to a destination type to use. 32 DenseMap<Type *, Type *> MappedTypes; 33 34 /// When checking to see if two subgraphs are isomorphic, we speculatively 35 /// add types to MappedTypes, but keep track of them here in case we need to 36 /// roll back. 37 SmallVector<Type *, 16> SpeculativeTypes; 38 39 SmallVector<StructType *, 16> SpeculativeDstOpaqueTypes; 40 41 /// This is a list of non-opaque structs in the source module that are mapped 42 /// to an opaque struct in the destination module. 43 SmallVector<StructType *, 16> SrcDefinitionsToResolve; 44 45 /// This is the set of opaque types in the destination modules who are 46 /// getting a body from the source module. 47 SmallPtrSet<StructType *, 16> DstResolvedOpaqueTypes; 48 49 public: 50 TypeMapTy(IRMover::IdentifiedStructTypeSet &DstStructTypesSet) 51 : DstStructTypesSet(DstStructTypesSet) {} 52 53 IRMover::IdentifiedStructTypeSet &DstStructTypesSet; 54 /// Indicate that the specified type in the destination module is conceptually 55 /// equivalent to the specified type in the source module. 56 void addTypeMapping(Type *DstTy, Type *SrcTy); 57 58 /// Produce a body for an opaque type in the dest module from a type 59 /// definition in the source module. 60 void linkDefinedTypeBodies(); 61 62 /// Return the mapped type to use for the specified input type from the 63 /// source module. 64 Type *get(Type *SrcTy); 65 Type *get(Type *SrcTy, SmallPtrSet<StructType *, 8> &Visited); 66 67 void finishType(StructType *DTy, StructType *STy, ArrayRef<Type *> ETypes); 68 69 FunctionType *get(FunctionType *T) { 70 return cast<FunctionType>(get((Type *)T)); 71 } 72 73 private: 74 Type *remapType(Type *SrcTy) override { return get(SrcTy); } 75 76 bool areTypesIsomorphic(Type *DstTy, Type *SrcTy); 77 }; 78 } 79 80 void TypeMapTy::addTypeMapping(Type *DstTy, Type *SrcTy) { 81 assert(SpeculativeTypes.empty()); 82 assert(SpeculativeDstOpaqueTypes.empty()); 83 84 // Check to see if these types are recursively isomorphic and establish a 85 // mapping between them if so. 86 if (!areTypesIsomorphic(DstTy, SrcTy)) { 87 // Oops, they aren't isomorphic. Just discard this request by rolling out 88 // any speculative mappings we've established. 89 for (Type *Ty : SpeculativeTypes) 90 MappedTypes.erase(Ty); 91 92 SrcDefinitionsToResolve.resize(SrcDefinitionsToResolve.size() - 93 SpeculativeDstOpaqueTypes.size()); 94 for (StructType *Ty : SpeculativeDstOpaqueTypes) 95 DstResolvedOpaqueTypes.erase(Ty); 96 } else { 97 // SrcTy and DstTy are recursively ismorphic. We clear names of SrcTy 98 // and all its descendants to lower amount of renaming in LLVM context 99 // Renaming occurs because we load all source modules to the same context 100 // and declaration with existing name gets renamed (i.e Foo -> Foo.42). 101 // As a result we may get several different types in the destination 102 // module, which are in fact the same. 103 for (Type *Ty : SpeculativeTypes) 104 if (auto *STy = dyn_cast<StructType>(Ty)) 105 if (STy->hasName()) 106 STy->setName(""); 107 } 108 SpeculativeTypes.clear(); 109 SpeculativeDstOpaqueTypes.clear(); 110 } 111 112 /// Recursively walk this pair of types, returning true if they are isomorphic, 113 /// false if they are not. 114 bool TypeMapTy::areTypesIsomorphic(Type *DstTy, Type *SrcTy) { 115 // Two types with differing kinds are clearly not isomorphic. 116 if (DstTy->getTypeID() != SrcTy->getTypeID()) 117 return false; 118 119 // If we have an entry in the MappedTypes table, then we have our answer. 120 Type *&Entry = MappedTypes[SrcTy]; 121 if (Entry) 122 return Entry == DstTy; 123 124 // Two identical types are clearly isomorphic. Remember this 125 // non-speculatively. 126 if (DstTy == SrcTy) { 127 Entry = DstTy; 128 return true; 129 } 130 131 // Okay, we have two types with identical kinds that we haven't seen before. 132 133 // If this is an opaque struct type, special case it. 134 if (StructType *SSTy = dyn_cast<StructType>(SrcTy)) { 135 // Mapping an opaque type to any struct, just keep the dest struct. 136 if (SSTy->isOpaque()) { 137 Entry = DstTy; 138 SpeculativeTypes.push_back(SrcTy); 139 return true; 140 } 141 142 // Mapping a non-opaque source type to an opaque dest. If this is the first 143 // type that we're mapping onto this destination type then we succeed. Keep 144 // the dest, but fill it in later. If this is the second (different) type 145 // that we're trying to map onto the same opaque type then we fail. 146 if (cast<StructType>(DstTy)->isOpaque()) { 147 // We can only map one source type onto the opaque destination type. 148 if (!DstResolvedOpaqueTypes.insert(cast<StructType>(DstTy)).second) 149 return false; 150 SrcDefinitionsToResolve.push_back(SSTy); 151 SpeculativeTypes.push_back(SrcTy); 152 SpeculativeDstOpaqueTypes.push_back(cast<StructType>(DstTy)); 153 Entry = DstTy; 154 return true; 155 } 156 } 157 158 // If the number of subtypes disagree between the two types, then we fail. 159 if (SrcTy->getNumContainedTypes() != DstTy->getNumContainedTypes()) 160 return false; 161 162 // Fail if any of the extra properties (e.g. array size) of the type disagree. 163 if (isa<IntegerType>(DstTy)) 164 return false; // bitwidth disagrees. 165 if (PointerType *PT = dyn_cast<PointerType>(DstTy)) { 166 if (PT->getAddressSpace() != cast<PointerType>(SrcTy)->getAddressSpace()) 167 return false; 168 } else if (FunctionType *FT = dyn_cast<FunctionType>(DstTy)) { 169 if (FT->isVarArg() != cast<FunctionType>(SrcTy)->isVarArg()) 170 return false; 171 } else if (StructType *DSTy = dyn_cast<StructType>(DstTy)) { 172 StructType *SSTy = cast<StructType>(SrcTy); 173 if (DSTy->isLiteral() != SSTy->isLiteral() || 174 DSTy->isPacked() != SSTy->isPacked()) 175 return false; 176 } else if (auto *DArrTy = dyn_cast<ArrayType>(DstTy)) { 177 if (DArrTy->getNumElements() != cast<ArrayType>(SrcTy)->getNumElements()) 178 return false; 179 } else if (auto *DVecTy = dyn_cast<VectorType>(DstTy)) { 180 if (DVecTy->getElementCount() != cast<VectorType>(SrcTy)->getElementCount()) 181 return false; 182 } 183 184 // Otherwise, we speculate that these two types will line up and recursively 185 // check the subelements. 186 Entry = DstTy; 187 SpeculativeTypes.push_back(SrcTy); 188 189 for (unsigned I = 0, E = SrcTy->getNumContainedTypes(); I != E; ++I) 190 if (!areTypesIsomorphic(DstTy->getContainedType(I), 191 SrcTy->getContainedType(I))) 192 return false; 193 194 // If everything seems to have lined up, then everything is great. 195 return true; 196 } 197 198 void TypeMapTy::linkDefinedTypeBodies() { 199 SmallVector<Type *, 16> Elements; 200 for (StructType *SrcSTy : SrcDefinitionsToResolve) { 201 StructType *DstSTy = cast<StructType>(MappedTypes[SrcSTy]); 202 assert(DstSTy->isOpaque()); 203 204 // Map the body of the source type over to a new body for the dest type. 205 Elements.resize(SrcSTy->getNumElements()); 206 for (unsigned I = 0, E = Elements.size(); I != E; ++I) 207 Elements[I] = get(SrcSTy->getElementType(I)); 208 209 DstSTy->setBody(Elements, SrcSTy->isPacked()); 210 DstStructTypesSet.switchToNonOpaque(DstSTy); 211 } 212 SrcDefinitionsToResolve.clear(); 213 DstResolvedOpaqueTypes.clear(); 214 } 215 216 void TypeMapTy::finishType(StructType *DTy, StructType *STy, 217 ArrayRef<Type *> ETypes) { 218 DTy->setBody(ETypes, STy->isPacked()); 219 220 // Steal STy's name. 221 if (STy->hasName()) { 222 SmallString<16> TmpName = STy->getName(); 223 STy->setName(""); 224 DTy->setName(TmpName); 225 } 226 227 DstStructTypesSet.addNonOpaque(DTy); 228 } 229 230 Type *TypeMapTy::get(Type *Ty) { 231 SmallPtrSet<StructType *, 8> Visited; 232 return get(Ty, Visited); 233 } 234 235 Type *TypeMapTy::get(Type *Ty, SmallPtrSet<StructType *, 8> &Visited) { 236 // If we already have an entry for this type, return it. 237 Type **Entry = &MappedTypes[Ty]; 238 if (*Entry) 239 return *Entry; 240 241 // These are types that LLVM itself will unique. 242 bool IsUniqued = !isa<StructType>(Ty) || cast<StructType>(Ty)->isLiteral(); 243 244 if (!IsUniqued) { 245 StructType *STy = cast<StructType>(Ty); 246 // This is actually a type from the destination module, this can be reached 247 // when this type is loaded in another module, added to DstStructTypesSet, 248 // and then we reach the same type in another module where it has not been 249 // added to MappedTypes. (PR37684) 250 if (STy->getContext().isODRUniquingDebugTypes() && !STy->isOpaque() && 251 DstStructTypesSet.hasType(STy)) 252 return *Entry = STy; 253 254 #ifndef NDEBUG 255 for (auto &Pair : MappedTypes) { 256 assert(!(Pair.first != Ty && Pair.second == Ty) && 257 "mapping to a source type"); 258 } 259 #endif 260 261 if (!Visited.insert(STy).second) { 262 StructType *DTy = StructType::create(Ty->getContext()); 263 return *Entry = DTy; 264 } 265 } 266 267 // If this is not a recursive type, then just map all of the elements and 268 // then rebuild the type from inside out. 269 SmallVector<Type *, 4> ElementTypes; 270 271 // If there are no element types to map, then the type is itself. This is 272 // true for the anonymous {} struct, things like 'float', integers, etc. 273 if (Ty->getNumContainedTypes() == 0 && IsUniqued) 274 return *Entry = Ty; 275 276 // Remap all of the elements, keeping track of whether any of them change. 277 bool AnyChange = false; 278 ElementTypes.resize(Ty->getNumContainedTypes()); 279 for (unsigned I = 0, E = Ty->getNumContainedTypes(); I != E; ++I) { 280 ElementTypes[I] = get(Ty->getContainedType(I), Visited); 281 AnyChange |= ElementTypes[I] != Ty->getContainedType(I); 282 } 283 284 // If we found our type while recursively processing stuff, just use it. 285 Entry = &MappedTypes[Ty]; 286 if (*Entry) { 287 if (auto *DTy = dyn_cast<StructType>(*Entry)) { 288 if (DTy->isOpaque()) { 289 auto *STy = cast<StructType>(Ty); 290 finishType(DTy, STy, ElementTypes); 291 } 292 } 293 return *Entry; 294 } 295 296 // If all of the element types mapped directly over and the type is not 297 // a named struct, then the type is usable as-is. 298 if (!AnyChange && IsUniqued) 299 return *Entry = Ty; 300 301 // Otherwise, rebuild a modified type. 302 switch (Ty->getTypeID()) { 303 default: 304 llvm_unreachable("unknown derived type to remap"); 305 case Type::ArrayTyID: 306 return *Entry = ArrayType::get(ElementTypes[0], 307 cast<ArrayType>(Ty)->getNumElements()); 308 case Type::VectorTyID: 309 return *Entry = VectorType::get(ElementTypes[0], 310 cast<VectorType>(Ty)->getNumElements()); 311 case Type::PointerTyID: 312 return *Entry = PointerType::get(ElementTypes[0], 313 cast<PointerType>(Ty)->getAddressSpace()); 314 case Type::FunctionTyID: 315 return *Entry = FunctionType::get(ElementTypes[0], 316 makeArrayRef(ElementTypes).slice(1), 317 cast<FunctionType>(Ty)->isVarArg()); 318 case Type::StructTyID: { 319 auto *STy = cast<StructType>(Ty); 320 bool IsPacked = STy->isPacked(); 321 if (IsUniqued) 322 return *Entry = StructType::get(Ty->getContext(), ElementTypes, IsPacked); 323 324 // If the type is opaque, we can just use it directly. 325 if (STy->isOpaque()) { 326 DstStructTypesSet.addOpaque(STy); 327 return *Entry = Ty; 328 } 329 330 if (StructType *OldT = 331 DstStructTypesSet.findNonOpaque(ElementTypes, IsPacked)) { 332 STy->setName(""); 333 return *Entry = OldT; 334 } 335 336 if (!AnyChange) { 337 DstStructTypesSet.addNonOpaque(STy); 338 return *Entry = Ty; 339 } 340 341 StructType *DTy = StructType::create(Ty->getContext()); 342 finishType(DTy, STy, ElementTypes); 343 return *Entry = DTy; 344 } 345 } 346 } 347 348 LinkDiagnosticInfo::LinkDiagnosticInfo(DiagnosticSeverity Severity, 349 const Twine &Msg) 350 : DiagnosticInfo(DK_Linker, Severity), Msg(Msg) {} 351 void LinkDiagnosticInfo::print(DiagnosticPrinter &DP) const { DP << Msg; } 352 353 //===----------------------------------------------------------------------===// 354 // IRLinker implementation. 355 //===----------------------------------------------------------------------===// 356 357 namespace { 358 class IRLinker; 359 360 /// Creates prototypes for functions that are lazily linked on the fly. This 361 /// speeds up linking for modules with many/ lazily linked functions of which 362 /// few get used. 363 class GlobalValueMaterializer final : public ValueMaterializer { 364 IRLinker &TheIRLinker; 365 366 public: 367 GlobalValueMaterializer(IRLinker &TheIRLinker) : TheIRLinker(TheIRLinker) {} 368 Value *materialize(Value *V) override; 369 }; 370 371 class LocalValueMaterializer final : public ValueMaterializer { 372 IRLinker &TheIRLinker; 373 374 public: 375 LocalValueMaterializer(IRLinker &TheIRLinker) : TheIRLinker(TheIRLinker) {} 376 Value *materialize(Value *V) override; 377 }; 378 379 /// Type of the Metadata map in \a ValueToValueMapTy. 380 typedef DenseMap<const Metadata *, TrackingMDRef> MDMapT; 381 382 /// This is responsible for keeping track of the state used for moving data 383 /// from SrcM to DstM. 384 class IRLinker { 385 Module &DstM; 386 std::unique_ptr<Module> SrcM; 387 388 /// See IRMover::move(). 389 std::function<void(GlobalValue &, IRMover::ValueAdder)> AddLazyFor; 390 391 TypeMapTy TypeMap; 392 GlobalValueMaterializer GValMaterializer; 393 LocalValueMaterializer LValMaterializer; 394 395 /// A metadata map that's shared between IRLinker instances. 396 MDMapT &SharedMDs; 397 398 /// Mapping of values from what they used to be in Src, to what they are now 399 /// in DstM. ValueToValueMapTy is a ValueMap, which involves some overhead 400 /// due to the use of Value handles which the Linker doesn't actually need, 401 /// but this allows us to reuse the ValueMapper code. 402 ValueToValueMapTy ValueMap; 403 ValueToValueMapTy IndirectSymbolValueMap; 404 405 DenseSet<GlobalValue *> ValuesToLink; 406 std::vector<GlobalValue *> Worklist; 407 std::vector<std::pair<GlobalValue *, Value*>> RAUWWorklist; 408 409 void maybeAdd(GlobalValue *GV) { 410 if (ValuesToLink.insert(GV).second) 411 Worklist.push_back(GV); 412 } 413 414 /// Whether we are importing globals for ThinLTO, as opposed to linking the 415 /// source module. If this flag is set, it means that we can rely on some 416 /// other object file to define any non-GlobalValue entities defined by the 417 /// source module. This currently causes us to not link retained types in 418 /// debug info metadata and module inline asm. 419 bool IsPerformingImport; 420 421 /// Set to true when all global value body linking is complete (including 422 /// lazy linking). Used to prevent metadata linking from creating new 423 /// references. 424 bool DoneLinkingBodies = false; 425 426 /// The Error encountered during materialization. We use an Optional here to 427 /// avoid needing to manage an unconsumed success value. 428 Optional<Error> FoundError; 429 void setError(Error E) { 430 if (E) 431 FoundError = std::move(E); 432 } 433 434 /// Most of the errors produced by this module are inconvertible StringErrors. 435 /// This convenience function lets us return one of those more easily. 436 Error stringErr(const Twine &T) { 437 return make_error<StringError>(T, inconvertibleErrorCode()); 438 } 439 440 /// Entry point for mapping values and alternate context for mapping aliases. 441 ValueMapper Mapper; 442 unsigned IndirectSymbolMCID; 443 444 /// Handles cloning of a global values from the source module into 445 /// the destination module, including setting the attributes and visibility. 446 GlobalValue *copyGlobalValueProto(const GlobalValue *SGV, bool ForDefinition); 447 448 void emitWarning(const Twine &Message) { 449 SrcM->getContext().diagnose(LinkDiagnosticInfo(DS_Warning, Message)); 450 } 451 452 /// Given a global in the source module, return the global in the 453 /// destination module that is being linked to, if any. 454 GlobalValue *getLinkedToGlobal(const GlobalValue *SrcGV) { 455 // If the source has no name it can't link. If it has local linkage, 456 // there is no name match-up going on. 457 if (!SrcGV->hasName() || SrcGV->hasLocalLinkage()) 458 return nullptr; 459 460 // Otherwise see if we have a match in the destination module's symtab. 461 GlobalValue *DGV = DstM.getNamedValue(SrcGV->getName()); 462 if (!DGV) 463 return nullptr; 464 465 // If we found a global with the same name in the dest module, but it has 466 // internal linkage, we are really not doing any linkage here. 467 if (DGV->hasLocalLinkage()) 468 return nullptr; 469 470 // Otherwise, we do in fact link to the destination global. 471 return DGV; 472 } 473 474 void computeTypeMapping(); 475 476 Expected<Constant *> linkAppendingVarProto(GlobalVariable *DstGV, 477 const GlobalVariable *SrcGV); 478 479 /// Given the GlobaValue \p SGV in the source module, and the matching 480 /// GlobalValue \p DGV (if any), return true if the linker will pull \p SGV 481 /// into the destination module. 482 /// 483 /// Note this code may call the client-provided \p AddLazyFor. 484 bool shouldLink(GlobalValue *DGV, GlobalValue &SGV); 485 Expected<Constant *> linkGlobalValueProto(GlobalValue *GV, 486 bool ForIndirectSymbol); 487 488 Error linkModuleFlagsMetadata(); 489 490 void linkGlobalVariable(GlobalVariable &Dst, GlobalVariable &Src); 491 Error linkFunctionBody(Function &Dst, Function &Src); 492 void linkIndirectSymbolBody(GlobalIndirectSymbol &Dst, 493 GlobalIndirectSymbol &Src); 494 Error linkGlobalValueBody(GlobalValue &Dst, GlobalValue &Src); 495 496 /// Replace all types in the source AttributeList with the 497 /// corresponding destination type. 498 AttributeList mapAttributeTypes(LLVMContext &C, AttributeList Attrs); 499 500 /// Functions that take care of cloning a specific global value type 501 /// into the destination module. 502 GlobalVariable *copyGlobalVariableProto(const GlobalVariable *SGVar); 503 Function *copyFunctionProto(const Function *SF); 504 GlobalValue *copyGlobalIndirectSymbolProto(const GlobalIndirectSymbol *SGIS); 505 506 /// Perform "replace all uses with" operations. These work items need to be 507 /// performed as part of materialization, but we postpone them to happen after 508 /// materialization is done. The materializer called by ValueMapper is not 509 /// expected to delete constants, as ValueMapper is holding pointers to some 510 /// of them, but constant destruction may be indirectly triggered by RAUW. 511 /// Hence, the need to move this out of the materialization call chain. 512 void flushRAUWWorklist(); 513 514 /// When importing for ThinLTO, prevent importing of types listed on 515 /// the DICompileUnit that we don't need a copy of in the importing 516 /// module. 517 void prepareCompileUnitsForImport(); 518 void linkNamedMDNodes(); 519 520 public: 521 IRLinker(Module &DstM, MDMapT &SharedMDs, 522 IRMover::IdentifiedStructTypeSet &Set, std::unique_ptr<Module> SrcM, 523 ArrayRef<GlobalValue *> ValuesToLink, 524 std::function<void(GlobalValue &, IRMover::ValueAdder)> AddLazyFor, 525 bool IsPerformingImport) 526 : DstM(DstM), SrcM(std::move(SrcM)), AddLazyFor(std::move(AddLazyFor)), 527 TypeMap(Set), GValMaterializer(*this), LValMaterializer(*this), 528 SharedMDs(SharedMDs), IsPerformingImport(IsPerformingImport), 529 Mapper(ValueMap, RF_MoveDistinctMDs | RF_IgnoreMissingLocals, &TypeMap, 530 &GValMaterializer), 531 IndirectSymbolMCID(Mapper.registerAlternateMappingContext( 532 IndirectSymbolValueMap, &LValMaterializer)) { 533 ValueMap.getMDMap() = std::move(SharedMDs); 534 for (GlobalValue *GV : ValuesToLink) 535 maybeAdd(GV); 536 if (IsPerformingImport) 537 prepareCompileUnitsForImport(); 538 } 539 ~IRLinker() { SharedMDs = std::move(*ValueMap.getMDMap()); } 540 541 Error run(); 542 Value *materialize(Value *V, bool ForIndirectSymbol); 543 }; 544 } 545 546 /// The LLVM SymbolTable class autorenames globals that conflict in the symbol 547 /// table. This is good for all clients except for us. Go through the trouble 548 /// to force this back. 549 static void forceRenaming(GlobalValue *GV, StringRef Name) { 550 // If the global doesn't force its name or if it already has the right name, 551 // there is nothing for us to do. 552 if (GV->hasLocalLinkage() || GV->getName() == Name) 553 return; 554 555 Module *M = GV->getParent(); 556 557 // If there is a conflict, rename the conflict. 558 if (GlobalValue *ConflictGV = M->getNamedValue(Name)) { 559 GV->takeName(ConflictGV); 560 ConflictGV->setName(Name); // This will cause ConflictGV to get renamed 561 assert(ConflictGV->getName() != Name && "forceRenaming didn't work"); 562 } else { 563 GV->setName(Name); // Force the name back 564 } 565 } 566 567 Value *GlobalValueMaterializer::materialize(Value *SGV) { 568 return TheIRLinker.materialize(SGV, false); 569 } 570 571 Value *LocalValueMaterializer::materialize(Value *SGV) { 572 return TheIRLinker.materialize(SGV, true); 573 } 574 575 Value *IRLinker::materialize(Value *V, bool ForIndirectSymbol) { 576 auto *SGV = dyn_cast<GlobalValue>(V); 577 if (!SGV) 578 return nullptr; 579 580 Expected<Constant *> NewProto = linkGlobalValueProto(SGV, ForIndirectSymbol); 581 if (!NewProto) { 582 setError(NewProto.takeError()); 583 return nullptr; 584 } 585 if (!*NewProto) 586 return nullptr; 587 588 GlobalValue *New = dyn_cast<GlobalValue>(*NewProto); 589 if (!New) 590 return *NewProto; 591 592 // If we already created the body, just return. 593 if (auto *F = dyn_cast<Function>(New)) { 594 if (!F->isDeclaration()) 595 return New; 596 } else if (auto *V = dyn_cast<GlobalVariable>(New)) { 597 if (V->hasInitializer() || V->hasAppendingLinkage()) 598 return New; 599 } else { 600 auto *IS = cast<GlobalIndirectSymbol>(New); 601 if (IS->getIndirectSymbol()) 602 return New; 603 } 604 605 // When linking a global for an indirect symbol, it will always be linked. 606 // However we need to check if it was not already scheduled to satisfy a 607 // reference from a regular global value initializer. We know if it has been 608 // schedule if the "New" GlobalValue that is mapped here for the indirect 609 // symbol is the same as the one already mapped. If there is an entry in the 610 // ValueMap but the value is different, it means that the value already had a 611 // definition in the destination module (linkonce for instance), but we need a 612 // new definition for the indirect symbol ("New" will be different. 613 if (ForIndirectSymbol && ValueMap.lookup(SGV) == New) 614 return New; 615 616 if (ForIndirectSymbol || shouldLink(New, *SGV)) 617 setError(linkGlobalValueBody(*New, *SGV)); 618 619 return New; 620 } 621 622 /// Loop through the global variables in the src module and merge them into the 623 /// dest module. 624 GlobalVariable *IRLinker::copyGlobalVariableProto(const GlobalVariable *SGVar) { 625 // No linking to be performed or linking from the source: simply create an 626 // identical version of the symbol over in the dest module... the 627 // initializer will be filled in later by LinkGlobalInits. 628 GlobalVariable *NewDGV = 629 new GlobalVariable(DstM, TypeMap.get(SGVar->getValueType()), 630 SGVar->isConstant(), GlobalValue::ExternalLinkage, 631 /*init*/ nullptr, SGVar->getName(), 632 /*insertbefore*/ nullptr, SGVar->getThreadLocalMode(), 633 SGVar->getAddressSpace()); 634 NewDGV->setAlignment(MaybeAlign(SGVar->getAlignment())); 635 NewDGV->copyAttributesFrom(SGVar); 636 return NewDGV; 637 } 638 639 AttributeList IRLinker::mapAttributeTypes(LLVMContext &C, AttributeList Attrs) { 640 for (unsigned i = 0; i < Attrs.getNumAttrSets(); ++i) { 641 if (Attrs.hasAttribute(i, Attribute::ByVal)) { 642 Type *Ty = Attrs.getAttribute(i, Attribute::ByVal).getValueAsType(); 643 if (!Ty) 644 continue; 645 646 Attrs = Attrs.removeAttribute(C, i, Attribute::ByVal); 647 Attrs = Attrs.addAttribute( 648 C, i, Attribute::getWithByValType(C, TypeMap.get(Ty))); 649 } 650 } 651 return Attrs; 652 } 653 654 /// Link the function in the source module into the destination module if 655 /// needed, setting up mapping information. 656 Function *IRLinker::copyFunctionProto(const Function *SF) { 657 // If there is no linkage to be performed or we are linking from the source, 658 // bring SF over. 659 auto *F = Function::Create(TypeMap.get(SF->getFunctionType()), 660 GlobalValue::ExternalLinkage, 661 SF->getAddressSpace(), SF->getName(), &DstM); 662 F->copyAttributesFrom(SF); 663 F->setAttributes(mapAttributeTypes(F->getContext(), F->getAttributes())); 664 return F; 665 } 666 667 /// Set up prototypes for any indirect symbols that come over from the source 668 /// module. 669 GlobalValue * 670 IRLinker::copyGlobalIndirectSymbolProto(const GlobalIndirectSymbol *SGIS) { 671 // If there is no linkage to be performed or we're linking from the source, 672 // bring over SGA. 673 auto *Ty = TypeMap.get(SGIS->getValueType()); 674 GlobalIndirectSymbol *GIS; 675 if (isa<GlobalAlias>(SGIS)) 676 GIS = GlobalAlias::create(Ty, SGIS->getAddressSpace(), 677 GlobalValue::ExternalLinkage, SGIS->getName(), 678 &DstM); 679 else 680 GIS = GlobalIFunc::create(Ty, SGIS->getAddressSpace(), 681 GlobalValue::ExternalLinkage, SGIS->getName(), 682 nullptr, &DstM); 683 GIS->copyAttributesFrom(SGIS); 684 return GIS; 685 } 686 687 GlobalValue *IRLinker::copyGlobalValueProto(const GlobalValue *SGV, 688 bool ForDefinition) { 689 GlobalValue *NewGV; 690 if (auto *SGVar = dyn_cast<GlobalVariable>(SGV)) { 691 NewGV = copyGlobalVariableProto(SGVar); 692 } else if (auto *SF = dyn_cast<Function>(SGV)) { 693 NewGV = copyFunctionProto(SF); 694 } else { 695 if (ForDefinition) 696 NewGV = copyGlobalIndirectSymbolProto(cast<GlobalIndirectSymbol>(SGV)); 697 else if (SGV->getValueType()->isFunctionTy()) 698 NewGV = 699 Function::Create(cast<FunctionType>(TypeMap.get(SGV->getValueType())), 700 GlobalValue::ExternalLinkage, SGV->getAddressSpace(), 701 SGV->getName(), &DstM); 702 else 703 NewGV = 704 new GlobalVariable(DstM, TypeMap.get(SGV->getValueType()), 705 /*isConstant*/ false, GlobalValue::ExternalLinkage, 706 /*init*/ nullptr, SGV->getName(), 707 /*insertbefore*/ nullptr, 708 SGV->getThreadLocalMode(), SGV->getAddressSpace()); 709 } 710 711 if (ForDefinition) 712 NewGV->setLinkage(SGV->getLinkage()); 713 else if (SGV->hasExternalWeakLinkage()) 714 NewGV->setLinkage(GlobalValue::ExternalWeakLinkage); 715 716 if (auto *NewGO = dyn_cast<GlobalObject>(NewGV)) { 717 // Metadata for global variables and function declarations is copied eagerly. 718 if (isa<GlobalVariable>(SGV) || SGV->isDeclaration()) 719 NewGO->copyMetadata(cast<GlobalObject>(SGV), 0); 720 } 721 722 // Remove these copied constants in case this stays a declaration, since 723 // they point to the source module. If the def is linked the values will 724 // be mapped in during linkFunctionBody. 725 if (auto *NewF = dyn_cast<Function>(NewGV)) { 726 NewF->setPersonalityFn(nullptr); 727 NewF->setPrefixData(nullptr); 728 NewF->setPrologueData(nullptr); 729 } 730 731 return NewGV; 732 } 733 734 static StringRef getTypeNamePrefix(StringRef Name) { 735 size_t DotPos = Name.rfind('.'); 736 return (DotPos == 0 || DotPos == StringRef::npos || Name.back() == '.' || 737 !isdigit(static_cast<unsigned char>(Name[DotPos + 1]))) 738 ? Name 739 : Name.substr(0, DotPos); 740 } 741 742 /// Loop over all of the linked values to compute type mappings. For example, 743 /// if we link "extern Foo *x" and "Foo *x = NULL", then we have two struct 744 /// types 'Foo' but one got renamed when the module was loaded into the same 745 /// LLVMContext. 746 void IRLinker::computeTypeMapping() { 747 for (GlobalValue &SGV : SrcM->globals()) { 748 GlobalValue *DGV = getLinkedToGlobal(&SGV); 749 if (!DGV) 750 continue; 751 752 if (!DGV->hasAppendingLinkage() || !SGV.hasAppendingLinkage()) { 753 TypeMap.addTypeMapping(DGV->getType(), SGV.getType()); 754 continue; 755 } 756 757 // Unify the element type of appending arrays. 758 ArrayType *DAT = cast<ArrayType>(DGV->getValueType()); 759 ArrayType *SAT = cast<ArrayType>(SGV.getValueType()); 760 TypeMap.addTypeMapping(DAT->getElementType(), SAT->getElementType()); 761 } 762 763 for (GlobalValue &SGV : *SrcM) 764 if (GlobalValue *DGV = getLinkedToGlobal(&SGV)) { 765 if (DGV->getType() == SGV.getType()) { 766 // If the types of DGV and SGV are the same, it means that DGV is from 767 // the source module and got added to DstM from a shared metadata. We 768 // shouldn't map this type to itself in case the type's components get 769 // remapped to a new type from DstM (for instance, during the loop over 770 // SrcM->getIdentifiedStructTypes() below). 771 continue; 772 } 773 774 TypeMap.addTypeMapping(DGV->getType(), SGV.getType()); 775 } 776 777 for (GlobalValue &SGV : SrcM->aliases()) 778 if (GlobalValue *DGV = getLinkedToGlobal(&SGV)) 779 TypeMap.addTypeMapping(DGV->getType(), SGV.getType()); 780 781 // Incorporate types by name, scanning all the types in the source module. 782 // At this point, the destination module may have a type "%foo = { i32 }" for 783 // example. When the source module got loaded into the same LLVMContext, if 784 // it had the same type, it would have been renamed to "%foo.42 = { i32 }". 785 std::vector<StructType *> Types = SrcM->getIdentifiedStructTypes(); 786 for (StructType *ST : Types) { 787 if (!ST->hasName()) 788 continue; 789 790 if (TypeMap.DstStructTypesSet.hasType(ST)) { 791 // This is actually a type from the destination module. 792 // getIdentifiedStructTypes() can have found it by walking debug info 793 // metadata nodes, some of which get linked by name when ODR Type Uniquing 794 // is enabled on the Context, from the source to the destination module. 795 continue; 796 } 797 798 auto STTypePrefix = getTypeNamePrefix(ST->getName()); 799 if (STTypePrefix.size()== ST->getName().size()) 800 continue; 801 802 // Check to see if the destination module has a struct with the prefix name. 803 StructType *DST = DstM.getTypeByName(STTypePrefix); 804 if (!DST) 805 continue; 806 807 // Don't use it if this actually came from the source module. They're in 808 // the same LLVMContext after all. Also don't use it unless the type is 809 // actually used in the destination module. This can happen in situations 810 // like this: 811 // 812 // Module A Module B 813 // -------- -------- 814 // %Z = type { %A } %B = type { %C.1 } 815 // %A = type { %B.1, [7 x i8] } %C.1 = type { i8* } 816 // %B.1 = type { %C } %A.2 = type { %B.3, [5 x i8] } 817 // %C = type { i8* } %B.3 = type { %C.1 } 818 // 819 // When we link Module B with Module A, the '%B' in Module B is 820 // used. However, that would then use '%C.1'. But when we process '%C.1', 821 // we prefer to take the '%C' version. So we are then left with both 822 // '%C.1' and '%C' being used for the same types. This leads to some 823 // variables using one type and some using the other. 824 if (TypeMap.DstStructTypesSet.hasType(DST)) 825 TypeMap.addTypeMapping(DST, ST); 826 } 827 828 // Now that we have discovered all of the type equivalences, get a body for 829 // any 'opaque' types in the dest module that are now resolved. 830 TypeMap.linkDefinedTypeBodies(); 831 } 832 833 static void getArrayElements(const Constant *C, 834 SmallVectorImpl<Constant *> &Dest) { 835 unsigned NumElements = cast<ArrayType>(C->getType())->getNumElements(); 836 837 for (unsigned i = 0; i != NumElements; ++i) 838 Dest.push_back(C->getAggregateElement(i)); 839 } 840 841 /// If there were any appending global variables, link them together now. 842 Expected<Constant *> 843 IRLinker::linkAppendingVarProto(GlobalVariable *DstGV, 844 const GlobalVariable *SrcGV) { 845 Type *EltTy = cast<ArrayType>(TypeMap.get(SrcGV->getValueType())) 846 ->getElementType(); 847 848 // FIXME: This upgrade is done during linking to support the C API. Once the 849 // old form is deprecated, we should move this upgrade to 850 // llvm::UpgradeGlobalVariable() and simplify the logic here and in 851 // Mapper::mapAppendingVariable() in ValueMapper.cpp. 852 StringRef Name = SrcGV->getName(); 853 bool IsNewStructor = false; 854 bool IsOldStructor = false; 855 if (Name == "llvm.global_ctors" || Name == "llvm.global_dtors") { 856 if (cast<StructType>(EltTy)->getNumElements() == 3) 857 IsNewStructor = true; 858 else 859 IsOldStructor = true; 860 } 861 862 PointerType *VoidPtrTy = Type::getInt8Ty(SrcGV->getContext())->getPointerTo(); 863 if (IsOldStructor) { 864 auto &ST = *cast<StructType>(EltTy); 865 Type *Tys[3] = {ST.getElementType(0), ST.getElementType(1), VoidPtrTy}; 866 EltTy = StructType::get(SrcGV->getContext(), Tys, false); 867 } 868 869 uint64_t DstNumElements = 0; 870 if (DstGV) { 871 ArrayType *DstTy = cast<ArrayType>(DstGV->getValueType()); 872 DstNumElements = DstTy->getNumElements(); 873 874 if (!SrcGV->hasAppendingLinkage() || !DstGV->hasAppendingLinkage()) 875 return stringErr( 876 "Linking globals named '" + SrcGV->getName() + 877 "': can only link appending global with another appending " 878 "global!"); 879 880 // Check to see that they two arrays agree on type. 881 if (EltTy != DstTy->getElementType()) 882 return stringErr("Appending variables with different element types!"); 883 if (DstGV->isConstant() != SrcGV->isConstant()) 884 return stringErr("Appending variables linked with different const'ness!"); 885 886 if (DstGV->getAlignment() != SrcGV->getAlignment()) 887 return stringErr( 888 "Appending variables with different alignment need to be linked!"); 889 890 if (DstGV->getVisibility() != SrcGV->getVisibility()) 891 return stringErr( 892 "Appending variables with different visibility need to be linked!"); 893 894 if (DstGV->hasGlobalUnnamedAddr() != SrcGV->hasGlobalUnnamedAddr()) 895 return stringErr( 896 "Appending variables with different unnamed_addr need to be linked!"); 897 898 if (DstGV->getSection() != SrcGV->getSection()) 899 return stringErr( 900 "Appending variables with different section name need to be linked!"); 901 } 902 903 SmallVector<Constant *, 16> SrcElements; 904 getArrayElements(SrcGV->getInitializer(), SrcElements); 905 906 if (IsNewStructor) { 907 auto It = remove_if(SrcElements, [this](Constant *E) { 908 auto *Key = 909 dyn_cast<GlobalValue>(E->getAggregateElement(2)->stripPointerCasts()); 910 if (!Key) 911 return false; 912 GlobalValue *DGV = getLinkedToGlobal(Key); 913 return !shouldLink(DGV, *Key); 914 }); 915 SrcElements.erase(It, SrcElements.end()); 916 } 917 uint64_t NewSize = DstNumElements + SrcElements.size(); 918 ArrayType *NewType = ArrayType::get(EltTy, NewSize); 919 920 // Create the new global variable. 921 GlobalVariable *NG = new GlobalVariable( 922 DstM, NewType, SrcGV->isConstant(), SrcGV->getLinkage(), 923 /*init*/ nullptr, /*name*/ "", DstGV, SrcGV->getThreadLocalMode(), 924 SrcGV->getAddressSpace()); 925 926 NG->copyAttributesFrom(SrcGV); 927 forceRenaming(NG, SrcGV->getName()); 928 929 Constant *Ret = ConstantExpr::getBitCast(NG, TypeMap.get(SrcGV->getType())); 930 931 Mapper.scheduleMapAppendingVariable(*NG, 932 DstGV ? DstGV->getInitializer() : nullptr, 933 IsOldStructor, SrcElements); 934 935 // Replace any uses of the two global variables with uses of the new 936 // global. 937 if (DstGV) { 938 RAUWWorklist.push_back( 939 std::make_pair(DstGV, ConstantExpr::getBitCast(NG, DstGV->getType()))); 940 } 941 942 return Ret; 943 } 944 945 bool IRLinker::shouldLink(GlobalValue *DGV, GlobalValue &SGV) { 946 if (ValuesToLink.count(&SGV) || SGV.hasLocalLinkage()) 947 return true; 948 949 if (DGV && !DGV->isDeclarationForLinker()) 950 return false; 951 952 if (SGV.isDeclaration() || DoneLinkingBodies) 953 return false; 954 955 // Callback to the client to give a chance to lazily add the Global to the 956 // list of value to link. 957 bool LazilyAdded = false; 958 AddLazyFor(SGV, [this, &LazilyAdded](GlobalValue &GV) { 959 maybeAdd(&GV); 960 LazilyAdded = true; 961 }); 962 return LazilyAdded; 963 } 964 965 Expected<Constant *> IRLinker::linkGlobalValueProto(GlobalValue *SGV, 966 bool ForIndirectSymbol) { 967 GlobalValue *DGV = getLinkedToGlobal(SGV); 968 969 bool ShouldLink = shouldLink(DGV, *SGV); 970 971 // just missing from map 972 if (ShouldLink) { 973 auto I = ValueMap.find(SGV); 974 if (I != ValueMap.end()) 975 return cast<Constant>(I->second); 976 977 I = IndirectSymbolValueMap.find(SGV); 978 if (I != IndirectSymbolValueMap.end()) 979 return cast<Constant>(I->second); 980 } 981 982 if (!ShouldLink && ForIndirectSymbol) 983 DGV = nullptr; 984 985 // Handle the ultra special appending linkage case first. 986 assert(!DGV || SGV->hasAppendingLinkage() == DGV->hasAppendingLinkage()); 987 if (SGV->hasAppendingLinkage()) 988 return linkAppendingVarProto(cast_or_null<GlobalVariable>(DGV), 989 cast<GlobalVariable>(SGV)); 990 991 GlobalValue *NewGV; 992 if (DGV && !ShouldLink) { 993 NewGV = DGV; 994 } else { 995 // If we are done linking global value bodies (i.e. we are performing 996 // metadata linking), don't link in the global value due to this 997 // reference, simply map it to null. 998 if (DoneLinkingBodies) 999 return nullptr; 1000 1001 NewGV = copyGlobalValueProto(SGV, ShouldLink || ForIndirectSymbol); 1002 if (ShouldLink || !ForIndirectSymbol) 1003 forceRenaming(NewGV, SGV->getName()); 1004 } 1005 1006 // Overloaded intrinsics have overloaded types names as part of their 1007 // names. If we renamed overloaded types we should rename the intrinsic 1008 // as well. 1009 if (Function *F = dyn_cast<Function>(NewGV)) 1010 if (auto Remangled = Intrinsic::remangleIntrinsicFunction(F)) 1011 NewGV = Remangled.getValue(); 1012 1013 if (ShouldLink || ForIndirectSymbol) { 1014 if (const Comdat *SC = SGV->getComdat()) { 1015 if (auto *GO = dyn_cast<GlobalObject>(NewGV)) { 1016 Comdat *DC = DstM.getOrInsertComdat(SC->getName()); 1017 DC->setSelectionKind(SC->getSelectionKind()); 1018 GO->setComdat(DC); 1019 } 1020 } 1021 } 1022 1023 if (!ShouldLink && ForIndirectSymbol) 1024 NewGV->setLinkage(GlobalValue::InternalLinkage); 1025 1026 Constant *C = NewGV; 1027 // Only create a bitcast if necessary. In particular, with 1028 // DebugTypeODRUniquing we may reach metadata in the destination module 1029 // containing a GV from the source module, in which case SGV will be 1030 // the same as DGV and NewGV, and TypeMap.get() will assert since it 1031 // assumes it is being invoked on a type in the source module. 1032 if (DGV && NewGV != SGV) { 1033 C = ConstantExpr::getPointerBitCastOrAddrSpaceCast( 1034 NewGV, TypeMap.get(SGV->getType())); 1035 } 1036 1037 if (DGV && NewGV != DGV) { 1038 // Schedule "replace all uses with" to happen after materializing is 1039 // done. It is not safe to do it now, since ValueMapper may be holding 1040 // pointers to constants that will get deleted if RAUW runs. 1041 RAUWWorklist.push_back(std::make_pair( 1042 DGV, 1043 ConstantExpr::getPointerBitCastOrAddrSpaceCast(NewGV, DGV->getType()))); 1044 } 1045 1046 return C; 1047 } 1048 1049 /// Update the initializers in the Dest module now that all globals that may be 1050 /// referenced are in Dest. 1051 void IRLinker::linkGlobalVariable(GlobalVariable &Dst, GlobalVariable &Src) { 1052 // Figure out what the initializer looks like in the dest module. 1053 Mapper.scheduleMapGlobalInitializer(Dst, *Src.getInitializer()); 1054 } 1055 1056 /// Copy the source function over into the dest function and fix up references 1057 /// to values. At this point we know that Dest is an external function, and 1058 /// that Src is not. 1059 Error IRLinker::linkFunctionBody(Function &Dst, Function &Src) { 1060 assert(Dst.isDeclaration() && !Src.isDeclaration()); 1061 1062 // Materialize if needed. 1063 if (Error Err = Src.materialize()) 1064 return Err; 1065 1066 // Link in the operands without remapping. 1067 if (Src.hasPrefixData()) 1068 Dst.setPrefixData(Src.getPrefixData()); 1069 if (Src.hasPrologueData()) 1070 Dst.setPrologueData(Src.getPrologueData()); 1071 if (Src.hasPersonalityFn()) 1072 Dst.setPersonalityFn(Src.getPersonalityFn()); 1073 1074 // Copy over the metadata attachments without remapping. 1075 Dst.copyMetadata(&Src, 0); 1076 1077 // Steal arguments and splice the body of Src into Dst. 1078 Dst.stealArgumentListFrom(Src); 1079 Dst.getBasicBlockList().splice(Dst.end(), Src.getBasicBlockList()); 1080 1081 // Everything has been moved over. Remap it. 1082 Mapper.scheduleRemapFunction(Dst); 1083 return Error::success(); 1084 } 1085 1086 void IRLinker::linkIndirectSymbolBody(GlobalIndirectSymbol &Dst, 1087 GlobalIndirectSymbol &Src) { 1088 Mapper.scheduleMapGlobalIndirectSymbol(Dst, *Src.getIndirectSymbol(), 1089 IndirectSymbolMCID); 1090 } 1091 1092 Error IRLinker::linkGlobalValueBody(GlobalValue &Dst, GlobalValue &Src) { 1093 if (auto *F = dyn_cast<Function>(&Src)) 1094 return linkFunctionBody(cast<Function>(Dst), *F); 1095 if (auto *GVar = dyn_cast<GlobalVariable>(&Src)) { 1096 linkGlobalVariable(cast<GlobalVariable>(Dst), *GVar); 1097 return Error::success(); 1098 } 1099 linkIndirectSymbolBody(cast<GlobalIndirectSymbol>(Dst), cast<GlobalIndirectSymbol>(Src)); 1100 return Error::success(); 1101 } 1102 1103 void IRLinker::flushRAUWWorklist() { 1104 for (const auto &Elem : RAUWWorklist) { 1105 GlobalValue *Old; 1106 Value *New; 1107 std::tie(Old, New) = Elem; 1108 1109 Old->replaceAllUsesWith(New); 1110 Old->eraseFromParent(); 1111 } 1112 RAUWWorklist.clear(); 1113 } 1114 1115 void IRLinker::prepareCompileUnitsForImport() { 1116 NamedMDNode *SrcCompileUnits = SrcM->getNamedMetadata("llvm.dbg.cu"); 1117 if (!SrcCompileUnits) 1118 return; 1119 // When importing for ThinLTO, prevent importing of types listed on 1120 // the DICompileUnit that we don't need a copy of in the importing 1121 // module. They will be emitted by the originating module. 1122 for (unsigned I = 0, E = SrcCompileUnits->getNumOperands(); I != E; ++I) { 1123 auto *CU = cast<DICompileUnit>(SrcCompileUnits->getOperand(I)); 1124 assert(CU && "Expected valid compile unit"); 1125 // Enums, macros, and retained types don't need to be listed on the 1126 // imported DICompileUnit. This means they will only be imported 1127 // if reached from the mapped IR. Do this by setting their value map 1128 // entries to nullptr, which will automatically prevent their importing 1129 // when reached from the DICompileUnit during metadata mapping. 1130 ValueMap.MD()[CU->getRawEnumTypes()].reset(nullptr); 1131 ValueMap.MD()[CU->getRawMacros()].reset(nullptr); 1132 ValueMap.MD()[CU->getRawRetainedTypes()].reset(nullptr); 1133 // The original definition (or at least its debug info - if the variable is 1134 // internalized an optimized away) will remain in the source module, so 1135 // there's no need to import them. 1136 // If LLVM ever does more advanced optimizations on global variables 1137 // (removing/localizing write operations, for instance) that can track 1138 // through debug info, this decision may need to be revisited - but do so 1139 // with care when it comes to debug info size. Emitting small CUs containing 1140 // only a few imported entities into every destination module may be very 1141 // size inefficient. 1142 ValueMap.MD()[CU->getRawGlobalVariables()].reset(nullptr); 1143 1144 // Imported entities only need to be mapped in if they have local 1145 // scope, as those might correspond to an imported entity inside a 1146 // function being imported (any locally scoped imported entities that 1147 // don't end up referenced by an imported function will not be emitted 1148 // into the object). Imported entities not in a local scope 1149 // (e.g. on the namespace) only need to be emitted by the originating 1150 // module. Create a list of the locally scoped imported entities, and 1151 // replace the source CUs imported entity list with the new list, so 1152 // only those are mapped in. 1153 // FIXME: Locally-scoped imported entities could be moved to the 1154 // functions they are local to instead of listing them on the CU, and 1155 // we would naturally only link in those needed by function importing. 1156 SmallVector<TrackingMDNodeRef, 4> AllImportedModules; 1157 bool ReplaceImportedEntities = false; 1158 for (auto *IE : CU->getImportedEntities()) { 1159 DIScope *Scope = IE->getScope(); 1160 assert(Scope && "Invalid Scope encoding!"); 1161 if (isa<DILocalScope>(Scope)) 1162 AllImportedModules.emplace_back(IE); 1163 else 1164 ReplaceImportedEntities = true; 1165 } 1166 if (ReplaceImportedEntities) { 1167 if (!AllImportedModules.empty()) 1168 CU->replaceImportedEntities(MDTuple::get( 1169 CU->getContext(), 1170 SmallVector<Metadata *, 16>(AllImportedModules.begin(), 1171 AllImportedModules.end()))); 1172 else 1173 // If there were no local scope imported entities, we can map 1174 // the whole list to nullptr. 1175 ValueMap.MD()[CU->getRawImportedEntities()].reset(nullptr); 1176 } 1177 } 1178 } 1179 1180 /// Insert all of the named MDNodes in Src into the Dest module. 1181 void IRLinker::linkNamedMDNodes() { 1182 const NamedMDNode *SrcModFlags = SrcM->getModuleFlagsMetadata(); 1183 for (const NamedMDNode &NMD : SrcM->named_metadata()) { 1184 // Don't link module flags here. Do them separately. 1185 if (&NMD == SrcModFlags) 1186 continue; 1187 NamedMDNode *DestNMD = DstM.getOrInsertNamedMetadata(NMD.getName()); 1188 // Add Src elements into Dest node. 1189 for (const MDNode *Op : NMD.operands()) 1190 DestNMD->addOperand(Mapper.mapMDNode(*Op)); 1191 } 1192 } 1193 1194 /// Merge the linker flags in Src into the Dest module. 1195 Error IRLinker::linkModuleFlagsMetadata() { 1196 // If the source module has no module flags, we are done. 1197 const NamedMDNode *SrcModFlags = SrcM->getModuleFlagsMetadata(); 1198 if (!SrcModFlags) 1199 return Error::success(); 1200 1201 // If the destination module doesn't have module flags yet, then just copy 1202 // over the source module's flags. 1203 NamedMDNode *DstModFlags = DstM.getOrInsertModuleFlagsMetadata(); 1204 if (DstModFlags->getNumOperands() == 0) { 1205 for (unsigned I = 0, E = SrcModFlags->getNumOperands(); I != E; ++I) 1206 DstModFlags->addOperand(SrcModFlags->getOperand(I)); 1207 1208 return Error::success(); 1209 } 1210 1211 // First build a map of the existing module flags and requirements. 1212 DenseMap<MDString *, std::pair<MDNode *, unsigned>> Flags; 1213 SmallSetVector<MDNode *, 16> Requirements; 1214 for (unsigned I = 0, E = DstModFlags->getNumOperands(); I != E; ++I) { 1215 MDNode *Op = DstModFlags->getOperand(I); 1216 ConstantInt *Behavior = mdconst::extract<ConstantInt>(Op->getOperand(0)); 1217 MDString *ID = cast<MDString>(Op->getOperand(1)); 1218 1219 if (Behavior->getZExtValue() == Module::Require) { 1220 Requirements.insert(cast<MDNode>(Op->getOperand(2))); 1221 } else { 1222 Flags[ID] = std::make_pair(Op, I); 1223 } 1224 } 1225 1226 // Merge in the flags from the source module, and also collect its set of 1227 // requirements. 1228 for (unsigned I = 0, E = SrcModFlags->getNumOperands(); I != E; ++I) { 1229 MDNode *SrcOp = SrcModFlags->getOperand(I); 1230 ConstantInt *SrcBehavior = 1231 mdconst::extract<ConstantInt>(SrcOp->getOperand(0)); 1232 MDString *ID = cast<MDString>(SrcOp->getOperand(1)); 1233 MDNode *DstOp; 1234 unsigned DstIndex; 1235 std::tie(DstOp, DstIndex) = Flags.lookup(ID); 1236 unsigned SrcBehaviorValue = SrcBehavior->getZExtValue(); 1237 1238 // If this is a requirement, add it and continue. 1239 if (SrcBehaviorValue == Module::Require) { 1240 // If the destination module does not already have this requirement, add 1241 // it. 1242 if (Requirements.insert(cast<MDNode>(SrcOp->getOperand(2)))) { 1243 DstModFlags->addOperand(SrcOp); 1244 } 1245 continue; 1246 } 1247 1248 // If there is no existing flag with this ID, just add it. 1249 if (!DstOp) { 1250 Flags[ID] = std::make_pair(SrcOp, DstModFlags->getNumOperands()); 1251 DstModFlags->addOperand(SrcOp); 1252 continue; 1253 } 1254 1255 // Otherwise, perform a merge. 1256 ConstantInt *DstBehavior = 1257 mdconst::extract<ConstantInt>(DstOp->getOperand(0)); 1258 unsigned DstBehaviorValue = DstBehavior->getZExtValue(); 1259 1260 auto overrideDstValue = [&]() { 1261 DstModFlags->setOperand(DstIndex, SrcOp); 1262 Flags[ID].first = SrcOp; 1263 }; 1264 1265 // If either flag has override behavior, handle it first. 1266 if (DstBehaviorValue == Module::Override) { 1267 // Diagnose inconsistent flags which both have override behavior. 1268 if (SrcBehaviorValue == Module::Override && 1269 SrcOp->getOperand(2) != DstOp->getOperand(2)) 1270 return stringErr("linking module flags '" + ID->getString() + 1271 "': IDs have conflicting override values in '" + 1272 SrcM->getModuleIdentifier() + "' and '" + 1273 DstM.getModuleIdentifier() + "'"); 1274 continue; 1275 } else if (SrcBehaviorValue == Module::Override) { 1276 // Update the destination flag to that of the source. 1277 overrideDstValue(); 1278 continue; 1279 } 1280 1281 // Diagnose inconsistent merge behavior types. 1282 if (SrcBehaviorValue != DstBehaviorValue) { 1283 bool MaxAndWarn = (SrcBehaviorValue == Module::Max && 1284 DstBehaviorValue == Module::Warning) || 1285 (DstBehaviorValue == Module::Max && 1286 SrcBehaviorValue == Module::Warning); 1287 if (!MaxAndWarn) 1288 return stringErr("linking module flags '" + ID->getString() + 1289 "': IDs have conflicting behaviors in '" + 1290 SrcM->getModuleIdentifier() + "' and '" + 1291 DstM.getModuleIdentifier() + "'"); 1292 } 1293 1294 auto replaceDstValue = [&](MDNode *New) { 1295 Metadata *FlagOps[] = {DstOp->getOperand(0), ID, New}; 1296 MDNode *Flag = MDNode::get(DstM.getContext(), FlagOps); 1297 DstModFlags->setOperand(DstIndex, Flag); 1298 Flags[ID].first = Flag; 1299 }; 1300 1301 // Emit a warning if the values differ and either source or destination 1302 // request Warning behavior. 1303 if ((DstBehaviorValue == Module::Warning || 1304 SrcBehaviorValue == Module::Warning) && 1305 SrcOp->getOperand(2) != DstOp->getOperand(2)) { 1306 std::string Str; 1307 raw_string_ostream(Str) 1308 << "linking module flags '" << ID->getString() 1309 << "': IDs have conflicting values ('" << *SrcOp->getOperand(2) 1310 << "' from " << SrcM->getModuleIdentifier() << " with '" 1311 << *DstOp->getOperand(2) << "' from " << DstM.getModuleIdentifier() 1312 << ')'; 1313 emitWarning(Str); 1314 } 1315 1316 // Choose the maximum if either source or destination request Max behavior. 1317 if (DstBehaviorValue == Module::Max || SrcBehaviorValue == Module::Max) { 1318 ConstantInt *DstValue = 1319 mdconst::extract<ConstantInt>(DstOp->getOperand(2)); 1320 ConstantInt *SrcValue = 1321 mdconst::extract<ConstantInt>(SrcOp->getOperand(2)); 1322 1323 // The resulting flag should have a Max behavior, and contain the maximum 1324 // value from between the source and destination values. 1325 Metadata *FlagOps[] = { 1326 (DstBehaviorValue != Module::Max ? SrcOp : DstOp)->getOperand(0), ID, 1327 (SrcValue->getZExtValue() > DstValue->getZExtValue() ? SrcOp : DstOp) 1328 ->getOperand(2)}; 1329 MDNode *Flag = MDNode::get(DstM.getContext(), FlagOps); 1330 DstModFlags->setOperand(DstIndex, Flag); 1331 Flags[ID].first = Flag; 1332 continue; 1333 } 1334 1335 // Perform the merge for standard behavior types. 1336 switch (SrcBehaviorValue) { 1337 case Module::Require: 1338 case Module::Override: 1339 llvm_unreachable("not possible"); 1340 case Module::Error: { 1341 // Emit an error if the values differ. 1342 if (SrcOp->getOperand(2) != DstOp->getOperand(2)) 1343 return stringErr("linking module flags '" + ID->getString() + 1344 "': IDs have conflicting values in '" + 1345 SrcM->getModuleIdentifier() + "' and '" + 1346 DstM.getModuleIdentifier() + "'"); 1347 continue; 1348 } 1349 case Module::Warning: { 1350 break; 1351 } 1352 case Module::Max: { 1353 break; 1354 } 1355 case Module::Append: { 1356 MDNode *DstValue = cast<MDNode>(DstOp->getOperand(2)); 1357 MDNode *SrcValue = cast<MDNode>(SrcOp->getOperand(2)); 1358 SmallVector<Metadata *, 8> MDs; 1359 MDs.reserve(DstValue->getNumOperands() + SrcValue->getNumOperands()); 1360 MDs.append(DstValue->op_begin(), DstValue->op_end()); 1361 MDs.append(SrcValue->op_begin(), SrcValue->op_end()); 1362 1363 replaceDstValue(MDNode::get(DstM.getContext(), MDs)); 1364 break; 1365 } 1366 case Module::AppendUnique: { 1367 SmallSetVector<Metadata *, 16> Elts; 1368 MDNode *DstValue = cast<MDNode>(DstOp->getOperand(2)); 1369 MDNode *SrcValue = cast<MDNode>(SrcOp->getOperand(2)); 1370 Elts.insert(DstValue->op_begin(), DstValue->op_end()); 1371 Elts.insert(SrcValue->op_begin(), SrcValue->op_end()); 1372 1373 replaceDstValue(MDNode::get(DstM.getContext(), 1374 makeArrayRef(Elts.begin(), Elts.end()))); 1375 break; 1376 } 1377 } 1378 1379 } 1380 1381 // Check all of the requirements. 1382 for (unsigned I = 0, E = Requirements.size(); I != E; ++I) { 1383 MDNode *Requirement = Requirements[I]; 1384 MDString *Flag = cast<MDString>(Requirement->getOperand(0)); 1385 Metadata *ReqValue = Requirement->getOperand(1); 1386 1387 MDNode *Op = Flags[Flag].first; 1388 if (!Op || Op->getOperand(2) != ReqValue) 1389 return stringErr("linking module flags '" + Flag->getString() + 1390 "': does not have the required value"); 1391 } 1392 return Error::success(); 1393 } 1394 1395 /// Return InlineAsm adjusted with target-specific directives if required. 1396 /// For ARM and Thumb, we have to add directives to select the appropriate ISA 1397 /// to support mixing module-level inline assembly from ARM and Thumb modules. 1398 static std::string adjustInlineAsm(const std::string &InlineAsm, 1399 const Triple &Triple) { 1400 if (Triple.getArch() == Triple::thumb || Triple.getArch() == Triple::thumbeb) 1401 return ".text\n.balign 2\n.thumb\n" + InlineAsm; 1402 if (Triple.getArch() == Triple::arm || Triple.getArch() == Triple::armeb) 1403 return ".text\n.balign 4\n.arm\n" + InlineAsm; 1404 return InlineAsm; 1405 } 1406 1407 Error IRLinker::run() { 1408 // Ensure metadata materialized before value mapping. 1409 if (SrcM->getMaterializer()) 1410 if (Error Err = SrcM->getMaterializer()->materializeMetadata()) 1411 return Err; 1412 1413 // Inherit the target data from the source module if the destination module 1414 // doesn't have one already. 1415 if (DstM.getDataLayout().isDefault()) 1416 DstM.setDataLayout(SrcM->getDataLayout()); 1417 1418 if (SrcM->getDataLayout() != DstM.getDataLayout()) { 1419 emitWarning("Linking two modules of different data layouts: '" + 1420 SrcM->getModuleIdentifier() + "' is '" + 1421 SrcM->getDataLayoutStr() + "' whereas '" + 1422 DstM.getModuleIdentifier() + "' is '" + 1423 DstM.getDataLayoutStr() + "'\n"); 1424 } 1425 1426 // Copy the target triple from the source to dest if the dest's is empty. 1427 if (DstM.getTargetTriple().empty() && !SrcM->getTargetTriple().empty()) 1428 DstM.setTargetTriple(SrcM->getTargetTriple()); 1429 1430 Triple SrcTriple(SrcM->getTargetTriple()), DstTriple(DstM.getTargetTriple()); 1431 1432 if (!SrcM->getTargetTriple().empty()&& 1433 !SrcTriple.isCompatibleWith(DstTriple)) 1434 emitWarning("Linking two modules of different target triples: " + 1435 SrcM->getModuleIdentifier() + "' is '" + 1436 SrcM->getTargetTriple() + "' whereas '" + 1437 DstM.getModuleIdentifier() + "' is '" + DstM.getTargetTriple() + 1438 "'\n"); 1439 1440 DstM.setTargetTriple(SrcTriple.merge(DstTriple)); 1441 1442 // Append the module inline asm string. 1443 if (!IsPerformingImport && !SrcM->getModuleInlineAsm().empty()) { 1444 std::string SrcModuleInlineAsm = adjustInlineAsm(SrcM->getModuleInlineAsm(), 1445 SrcTriple); 1446 if (DstM.getModuleInlineAsm().empty()) 1447 DstM.setModuleInlineAsm(SrcModuleInlineAsm); 1448 else 1449 DstM.setModuleInlineAsm(DstM.getModuleInlineAsm() + "\n" + 1450 SrcModuleInlineAsm); 1451 } 1452 1453 // Loop over all of the linked values to compute type mappings. 1454 computeTypeMapping(); 1455 1456 std::reverse(Worklist.begin(), Worklist.end()); 1457 while (!Worklist.empty()) { 1458 GlobalValue *GV = Worklist.back(); 1459 Worklist.pop_back(); 1460 1461 // Already mapped. 1462 if (ValueMap.find(GV) != ValueMap.end() || 1463 IndirectSymbolValueMap.find(GV) != IndirectSymbolValueMap.end()) 1464 continue; 1465 1466 assert(!GV->isDeclaration()); 1467 Mapper.mapValue(*GV); 1468 if (FoundError) 1469 return std::move(*FoundError); 1470 flushRAUWWorklist(); 1471 } 1472 1473 // Note that we are done linking global value bodies. This prevents 1474 // metadata linking from creating new references. 1475 DoneLinkingBodies = true; 1476 Mapper.addFlags(RF_NullMapMissingGlobalValues); 1477 1478 // Remap all of the named MDNodes in Src into the DstM module. We do this 1479 // after linking GlobalValues so that MDNodes that reference GlobalValues 1480 // are properly remapped. 1481 linkNamedMDNodes(); 1482 1483 // Merge the module flags into the DstM module. 1484 return linkModuleFlagsMetadata(); 1485 } 1486 1487 IRMover::StructTypeKeyInfo::KeyTy::KeyTy(ArrayRef<Type *> E, bool P) 1488 : ETypes(E), IsPacked(P) {} 1489 1490 IRMover::StructTypeKeyInfo::KeyTy::KeyTy(const StructType *ST) 1491 : ETypes(ST->elements()), IsPacked(ST->isPacked()) {} 1492 1493 bool IRMover::StructTypeKeyInfo::KeyTy::operator==(const KeyTy &That) const { 1494 return IsPacked == That.IsPacked && ETypes == That.ETypes; 1495 } 1496 1497 bool IRMover::StructTypeKeyInfo::KeyTy::operator!=(const KeyTy &That) const { 1498 return !this->operator==(That); 1499 } 1500 1501 StructType *IRMover::StructTypeKeyInfo::getEmptyKey() { 1502 return DenseMapInfo<StructType *>::getEmptyKey(); 1503 } 1504 1505 StructType *IRMover::StructTypeKeyInfo::getTombstoneKey() { 1506 return DenseMapInfo<StructType *>::getTombstoneKey(); 1507 } 1508 1509 unsigned IRMover::StructTypeKeyInfo::getHashValue(const KeyTy &Key) { 1510 return hash_combine(hash_combine_range(Key.ETypes.begin(), Key.ETypes.end()), 1511 Key.IsPacked); 1512 } 1513 1514 unsigned IRMover::StructTypeKeyInfo::getHashValue(const StructType *ST) { 1515 return getHashValue(KeyTy(ST)); 1516 } 1517 1518 bool IRMover::StructTypeKeyInfo::isEqual(const KeyTy &LHS, 1519 const StructType *RHS) { 1520 if (RHS == getEmptyKey() || RHS == getTombstoneKey()) 1521 return false; 1522 return LHS == KeyTy(RHS); 1523 } 1524 1525 bool IRMover::StructTypeKeyInfo::isEqual(const StructType *LHS, 1526 const StructType *RHS) { 1527 if (RHS == getEmptyKey() || RHS == getTombstoneKey()) 1528 return LHS == RHS; 1529 return KeyTy(LHS) == KeyTy(RHS); 1530 } 1531 1532 void IRMover::IdentifiedStructTypeSet::addNonOpaque(StructType *Ty) { 1533 assert(!Ty->isOpaque()); 1534 NonOpaqueStructTypes.insert(Ty); 1535 } 1536 1537 void IRMover::IdentifiedStructTypeSet::switchToNonOpaque(StructType *Ty) { 1538 assert(!Ty->isOpaque()); 1539 NonOpaqueStructTypes.insert(Ty); 1540 bool Removed = OpaqueStructTypes.erase(Ty); 1541 (void)Removed; 1542 assert(Removed); 1543 } 1544 1545 void IRMover::IdentifiedStructTypeSet::addOpaque(StructType *Ty) { 1546 assert(Ty->isOpaque()); 1547 OpaqueStructTypes.insert(Ty); 1548 } 1549 1550 StructType * 1551 IRMover::IdentifiedStructTypeSet::findNonOpaque(ArrayRef<Type *> ETypes, 1552 bool IsPacked) { 1553 IRMover::StructTypeKeyInfo::KeyTy Key(ETypes, IsPacked); 1554 auto I = NonOpaqueStructTypes.find_as(Key); 1555 return I == NonOpaqueStructTypes.end() ? nullptr : *I; 1556 } 1557 1558 bool IRMover::IdentifiedStructTypeSet::hasType(StructType *Ty) { 1559 if (Ty->isOpaque()) 1560 return OpaqueStructTypes.count(Ty); 1561 auto I = NonOpaqueStructTypes.find(Ty); 1562 return I == NonOpaqueStructTypes.end() ? false : *I == Ty; 1563 } 1564 1565 IRMover::IRMover(Module &M) : Composite(M) { 1566 TypeFinder StructTypes; 1567 StructTypes.run(M, /* OnlyNamed */ false); 1568 for (StructType *Ty : StructTypes) { 1569 if (Ty->isOpaque()) 1570 IdentifiedStructTypes.addOpaque(Ty); 1571 else 1572 IdentifiedStructTypes.addNonOpaque(Ty); 1573 } 1574 // Self-map metadatas in the destination module. This is needed when 1575 // DebugTypeODRUniquing is enabled on the LLVMContext, since metadata in the 1576 // destination module may be reached from the source module. 1577 for (auto *MD : StructTypes.getVisitedMetadata()) { 1578 SharedMDs[MD].reset(const_cast<MDNode *>(MD)); 1579 } 1580 } 1581 1582 Error IRMover::move( 1583 std::unique_ptr<Module> Src, ArrayRef<GlobalValue *> ValuesToLink, 1584 std::function<void(GlobalValue &, ValueAdder Add)> AddLazyFor, 1585 bool IsPerformingImport) { 1586 IRLinker TheIRLinker(Composite, SharedMDs, IdentifiedStructTypes, 1587 std::move(Src), ValuesToLink, std::move(AddLazyFor), 1588 IsPerformingImport); 1589 Error E = TheIRLinker.run(); 1590 Composite.dropTriviallyDeadConstantArrays(); 1591 return E; 1592 } 1593