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::FixedVectorTyID: 309 case Type::ScalableVectorTyID: 310 return *Entry = VectorType::get(ElementTypes[0], 311 cast<VectorType>(Ty)->getNumElements()); 312 case Type::PointerTyID: 313 return *Entry = PointerType::get(ElementTypes[0], 314 cast<PointerType>(Ty)->getAddressSpace()); 315 case Type::FunctionTyID: 316 return *Entry = FunctionType::get(ElementTypes[0], 317 makeArrayRef(ElementTypes).slice(1), 318 cast<FunctionType>(Ty)->isVarArg()); 319 case Type::StructTyID: { 320 auto *STy = cast<StructType>(Ty); 321 bool IsPacked = STy->isPacked(); 322 if (IsUniqued) 323 return *Entry = StructType::get(Ty->getContext(), ElementTypes, IsPacked); 324 325 // If the type is opaque, we can just use it directly. 326 if (STy->isOpaque()) { 327 DstStructTypesSet.addOpaque(STy); 328 return *Entry = Ty; 329 } 330 331 if (StructType *OldT = 332 DstStructTypesSet.findNonOpaque(ElementTypes, IsPacked)) { 333 STy->setName(""); 334 return *Entry = OldT; 335 } 336 337 if (!AnyChange) { 338 DstStructTypesSet.addNonOpaque(STy); 339 return *Entry = Ty; 340 } 341 342 StructType *DTy = StructType::create(Ty->getContext()); 343 finishType(DTy, STy, ElementTypes); 344 return *Entry = DTy; 345 } 346 } 347 } 348 349 LinkDiagnosticInfo::LinkDiagnosticInfo(DiagnosticSeverity Severity, 350 const Twine &Msg) 351 : DiagnosticInfo(DK_Linker, Severity), Msg(Msg) {} 352 void LinkDiagnosticInfo::print(DiagnosticPrinter &DP) const { DP << Msg; } 353 354 //===----------------------------------------------------------------------===// 355 // IRLinker implementation. 356 //===----------------------------------------------------------------------===// 357 358 namespace { 359 class IRLinker; 360 361 /// Creates prototypes for functions that are lazily linked on the fly. This 362 /// speeds up linking for modules with many/ lazily linked functions of which 363 /// few get used. 364 class GlobalValueMaterializer final : public ValueMaterializer { 365 IRLinker &TheIRLinker; 366 367 public: 368 GlobalValueMaterializer(IRLinker &TheIRLinker) : TheIRLinker(TheIRLinker) {} 369 Value *materialize(Value *V) override; 370 }; 371 372 class LocalValueMaterializer final : public ValueMaterializer { 373 IRLinker &TheIRLinker; 374 375 public: 376 LocalValueMaterializer(IRLinker &TheIRLinker) : TheIRLinker(TheIRLinker) {} 377 Value *materialize(Value *V) override; 378 }; 379 380 /// Type of the Metadata map in \a ValueToValueMapTy. 381 typedef DenseMap<const Metadata *, TrackingMDRef> MDMapT; 382 383 /// This is responsible for keeping track of the state used for moving data 384 /// from SrcM to DstM. 385 class IRLinker { 386 Module &DstM; 387 std::unique_ptr<Module> SrcM; 388 389 /// See IRMover::move(). 390 std::function<void(GlobalValue &, IRMover::ValueAdder)> AddLazyFor; 391 392 TypeMapTy TypeMap; 393 GlobalValueMaterializer GValMaterializer; 394 LocalValueMaterializer LValMaterializer; 395 396 /// A metadata map that's shared between IRLinker instances. 397 MDMapT &SharedMDs; 398 399 /// Mapping of values from what they used to be in Src, to what they are now 400 /// in DstM. ValueToValueMapTy is a ValueMap, which involves some overhead 401 /// due to the use of Value handles which the Linker doesn't actually need, 402 /// but this allows us to reuse the ValueMapper code. 403 ValueToValueMapTy ValueMap; 404 ValueToValueMapTy IndirectSymbolValueMap; 405 406 DenseSet<GlobalValue *> ValuesToLink; 407 std::vector<GlobalValue *> Worklist; 408 std::vector<std::pair<GlobalValue *, Value*>> RAUWWorklist; 409 410 void maybeAdd(GlobalValue *GV) { 411 if (ValuesToLink.insert(GV).second) 412 Worklist.push_back(GV); 413 } 414 415 /// Whether we are importing globals for ThinLTO, as opposed to linking the 416 /// source module. If this flag is set, it means that we can rely on some 417 /// other object file to define any non-GlobalValue entities defined by the 418 /// source module. This currently causes us to not link retained types in 419 /// debug info metadata and module inline asm. 420 bool IsPerformingImport; 421 422 /// Set to true when all global value body linking is complete (including 423 /// lazy linking). Used to prevent metadata linking from creating new 424 /// references. 425 bool DoneLinkingBodies = false; 426 427 /// The Error encountered during materialization. We use an Optional here to 428 /// avoid needing to manage an unconsumed success value. 429 Optional<Error> FoundError; 430 void setError(Error E) { 431 if (E) 432 FoundError = std::move(E); 433 } 434 435 /// Most of the errors produced by this module are inconvertible StringErrors. 436 /// This convenience function lets us return one of those more easily. 437 Error stringErr(const Twine &T) { 438 return make_error<StringError>(T, inconvertibleErrorCode()); 439 } 440 441 /// Entry point for mapping values and alternate context for mapping aliases. 442 ValueMapper Mapper; 443 unsigned IndirectSymbolMCID; 444 445 /// Handles cloning of a global values from the source module into 446 /// the destination module, including setting the attributes and visibility. 447 GlobalValue *copyGlobalValueProto(const GlobalValue *SGV, bool ForDefinition); 448 449 void emitWarning(const Twine &Message) { 450 SrcM->getContext().diagnose(LinkDiagnosticInfo(DS_Warning, Message)); 451 } 452 453 /// Given a global in the source module, return the global in the 454 /// destination module that is being linked to, if any. 455 GlobalValue *getLinkedToGlobal(const GlobalValue *SrcGV) { 456 // If the source has no name it can't link. If it has local linkage, 457 // there is no name match-up going on. 458 if (!SrcGV->hasName() || SrcGV->hasLocalLinkage()) 459 return nullptr; 460 461 // Otherwise see if we have a match in the destination module's symtab. 462 GlobalValue *DGV = DstM.getNamedValue(SrcGV->getName()); 463 if (!DGV) 464 return nullptr; 465 466 // If we found a global with the same name in the dest module, but it has 467 // internal linkage, we are really not doing any linkage here. 468 if (DGV->hasLocalLinkage()) 469 return nullptr; 470 471 // Otherwise, we do in fact link to the destination global. 472 return DGV; 473 } 474 475 void computeTypeMapping(); 476 477 Expected<Constant *> linkAppendingVarProto(GlobalVariable *DstGV, 478 const GlobalVariable *SrcGV); 479 480 /// Given the GlobaValue \p SGV in the source module, and the matching 481 /// GlobalValue \p DGV (if any), return true if the linker will pull \p SGV 482 /// into the destination module. 483 /// 484 /// Note this code may call the client-provided \p AddLazyFor. 485 bool shouldLink(GlobalValue *DGV, GlobalValue &SGV); 486 Expected<Constant *> linkGlobalValueProto(GlobalValue *GV, 487 bool ForIndirectSymbol); 488 489 Error linkModuleFlagsMetadata(); 490 491 void linkGlobalVariable(GlobalVariable &Dst, GlobalVariable &Src); 492 Error linkFunctionBody(Function &Dst, Function &Src); 493 void linkIndirectSymbolBody(GlobalIndirectSymbol &Dst, 494 GlobalIndirectSymbol &Src); 495 Error linkGlobalValueBody(GlobalValue &Dst, GlobalValue &Src); 496 497 /// Replace all types in the source AttributeList with the 498 /// corresponding destination type. 499 AttributeList mapAttributeTypes(LLVMContext &C, AttributeList Attrs); 500 501 /// Functions that take care of cloning a specific global value type 502 /// into the destination module. 503 GlobalVariable *copyGlobalVariableProto(const GlobalVariable *SGVar); 504 Function *copyFunctionProto(const Function *SF); 505 GlobalValue *copyGlobalIndirectSymbolProto(const GlobalIndirectSymbol *SGIS); 506 507 /// Perform "replace all uses with" operations. These work items need to be 508 /// performed as part of materialization, but we postpone them to happen after 509 /// materialization is done. The materializer called by ValueMapper is not 510 /// expected to delete constants, as ValueMapper is holding pointers to some 511 /// of them, but constant destruction may be indirectly triggered by RAUW. 512 /// Hence, the need to move this out of the materialization call chain. 513 void flushRAUWWorklist(); 514 515 /// When importing for ThinLTO, prevent importing of types listed on 516 /// the DICompileUnit that we don't need a copy of in the importing 517 /// module. 518 void prepareCompileUnitsForImport(); 519 void linkNamedMDNodes(); 520 521 public: 522 IRLinker(Module &DstM, MDMapT &SharedMDs, 523 IRMover::IdentifiedStructTypeSet &Set, std::unique_ptr<Module> SrcM, 524 ArrayRef<GlobalValue *> ValuesToLink, 525 std::function<void(GlobalValue &, IRMover::ValueAdder)> AddLazyFor, 526 bool IsPerformingImport) 527 : DstM(DstM), SrcM(std::move(SrcM)), AddLazyFor(std::move(AddLazyFor)), 528 TypeMap(Set), GValMaterializer(*this), LValMaterializer(*this), 529 SharedMDs(SharedMDs), IsPerformingImport(IsPerformingImport), 530 Mapper(ValueMap, RF_MoveDistinctMDs | RF_IgnoreMissingLocals, &TypeMap, 531 &GValMaterializer), 532 IndirectSymbolMCID(Mapper.registerAlternateMappingContext( 533 IndirectSymbolValueMap, &LValMaterializer)) { 534 ValueMap.getMDMap() = std::move(SharedMDs); 535 for (GlobalValue *GV : ValuesToLink) 536 maybeAdd(GV); 537 if (IsPerformingImport) 538 prepareCompileUnitsForImport(); 539 } 540 ~IRLinker() { SharedMDs = std::move(*ValueMap.getMDMap()); } 541 542 Error run(); 543 Value *materialize(Value *V, bool ForIndirectSymbol); 544 }; 545 } 546 547 /// The LLVM SymbolTable class autorenames globals that conflict in the symbol 548 /// table. This is good for all clients except for us. Go through the trouble 549 /// to force this back. 550 static void forceRenaming(GlobalValue *GV, StringRef Name) { 551 // If the global doesn't force its name or if it already has the right name, 552 // there is nothing for us to do. 553 if (GV->hasLocalLinkage() || GV->getName() == Name) 554 return; 555 556 Module *M = GV->getParent(); 557 558 // If there is a conflict, rename the conflict. 559 if (GlobalValue *ConflictGV = M->getNamedValue(Name)) { 560 GV->takeName(ConflictGV); 561 ConflictGV->setName(Name); // This will cause ConflictGV to get renamed 562 assert(ConflictGV->getName() != Name && "forceRenaming didn't work"); 563 } else { 564 GV->setName(Name); // Force the name back 565 } 566 } 567 568 Value *GlobalValueMaterializer::materialize(Value *SGV) { 569 return TheIRLinker.materialize(SGV, false); 570 } 571 572 Value *LocalValueMaterializer::materialize(Value *SGV) { 573 return TheIRLinker.materialize(SGV, true); 574 } 575 576 Value *IRLinker::materialize(Value *V, bool ForIndirectSymbol) { 577 auto *SGV = dyn_cast<GlobalValue>(V); 578 if (!SGV) 579 return nullptr; 580 581 Expected<Constant *> NewProto = linkGlobalValueProto(SGV, ForIndirectSymbol); 582 if (!NewProto) { 583 setError(NewProto.takeError()); 584 return nullptr; 585 } 586 if (!*NewProto) 587 return nullptr; 588 589 GlobalValue *New = dyn_cast<GlobalValue>(*NewProto); 590 if (!New) 591 return *NewProto; 592 593 // If we already created the body, just return. 594 if (auto *F = dyn_cast<Function>(New)) { 595 if (!F->isDeclaration()) 596 return New; 597 } else if (auto *V = dyn_cast<GlobalVariable>(New)) { 598 if (V->hasInitializer() || V->hasAppendingLinkage()) 599 return New; 600 } else { 601 auto *IS = cast<GlobalIndirectSymbol>(New); 602 if (IS->getIndirectSymbol()) 603 return New; 604 } 605 606 // When linking a global for an indirect symbol, it will always be linked. 607 // However we need to check if it was not already scheduled to satisfy a 608 // reference from a regular global value initializer. We know if it has been 609 // schedule if the "New" GlobalValue that is mapped here for the indirect 610 // symbol is the same as the one already mapped. If there is an entry in the 611 // ValueMap but the value is different, it means that the value already had a 612 // definition in the destination module (linkonce for instance), but we need a 613 // new definition for the indirect symbol ("New" will be different. 614 if (ForIndirectSymbol && ValueMap.lookup(SGV) == New) 615 return New; 616 617 if (ForIndirectSymbol || shouldLink(New, *SGV)) 618 setError(linkGlobalValueBody(*New, *SGV)); 619 620 return New; 621 } 622 623 /// Loop through the global variables in the src module and merge them into the 624 /// dest module. 625 GlobalVariable *IRLinker::copyGlobalVariableProto(const GlobalVariable *SGVar) { 626 // No linking to be performed or linking from the source: simply create an 627 // identical version of the symbol over in the dest module... the 628 // initializer will be filled in later by LinkGlobalInits. 629 GlobalVariable *NewDGV = 630 new GlobalVariable(DstM, TypeMap.get(SGVar->getValueType()), 631 SGVar->isConstant(), GlobalValue::ExternalLinkage, 632 /*init*/ nullptr, SGVar->getName(), 633 /*insertbefore*/ nullptr, SGVar->getThreadLocalMode(), 634 SGVar->getAddressSpace()); 635 NewDGV->setAlignment(MaybeAlign(SGVar->getAlignment())); 636 NewDGV->copyAttributesFrom(SGVar); 637 return NewDGV; 638 } 639 640 AttributeList IRLinker::mapAttributeTypes(LLVMContext &C, AttributeList Attrs) { 641 for (unsigned i = 0; i < Attrs.getNumAttrSets(); ++i) { 642 if (Attrs.hasAttribute(i, Attribute::ByVal)) { 643 Type *Ty = Attrs.getAttribute(i, Attribute::ByVal).getValueAsType(); 644 if (!Ty) 645 continue; 646 647 Attrs = Attrs.removeAttribute(C, i, Attribute::ByVal); 648 Attrs = Attrs.addAttribute( 649 C, i, Attribute::getWithByValType(C, TypeMap.get(Ty))); 650 } 651 } 652 return Attrs; 653 } 654 655 /// Link the function in the source module into the destination module if 656 /// needed, setting up mapping information. 657 Function *IRLinker::copyFunctionProto(const Function *SF) { 658 // If there is no linkage to be performed or we are linking from the source, 659 // bring SF over. 660 auto *F = Function::Create(TypeMap.get(SF->getFunctionType()), 661 GlobalValue::ExternalLinkage, 662 SF->getAddressSpace(), SF->getName(), &DstM); 663 F->copyAttributesFrom(SF); 664 F->setAttributes(mapAttributeTypes(F->getContext(), F->getAttributes())); 665 return F; 666 } 667 668 /// Set up prototypes for any indirect symbols that come over from the source 669 /// module. 670 GlobalValue * 671 IRLinker::copyGlobalIndirectSymbolProto(const GlobalIndirectSymbol *SGIS) { 672 // If there is no linkage to be performed or we're linking from the source, 673 // bring over SGA. 674 auto *Ty = TypeMap.get(SGIS->getValueType()); 675 GlobalIndirectSymbol *GIS; 676 if (isa<GlobalAlias>(SGIS)) 677 GIS = GlobalAlias::create(Ty, SGIS->getAddressSpace(), 678 GlobalValue::ExternalLinkage, SGIS->getName(), 679 &DstM); 680 else 681 GIS = GlobalIFunc::create(Ty, SGIS->getAddressSpace(), 682 GlobalValue::ExternalLinkage, SGIS->getName(), 683 nullptr, &DstM); 684 GIS->copyAttributesFrom(SGIS); 685 return GIS; 686 } 687 688 GlobalValue *IRLinker::copyGlobalValueProto(const GlobalValue *SGV, 689 bool ForDefinition) { 690 GlobalValue *NewGV; 691 if (auto *SGVar = dyn_cast<GlobalVariable>(SGV)) { 692 NewGV = copyGlobalVariableProto(SGVar); 693 } else if (auto *SF = dyn_cast<Function>(SGV)) { 694 NewGV = copyFunctionProto(SF); 695 } else { 696 if (ForDefinition) 697 NewGV = copyGlobalIndirectSymbolProto(cast<GlobalIndirectSymbol>(SGV)); 698 else if (SGV->getValueType()->isFunctionTy()) 699 NewGV = 700 Function::Create(cast<FunctionType>(TypeMap.get(SGV->getValueType())), 701 GlobalValue::ExternalLinkage, SGV->getAddressSpace(), 702 SGV->getName(), &DstM); 703 else 704 NewGV = 705 new GlobalVariable(DstM, TypeMap.get(SGV->getValueType()), 706 /*isConstant*/ false, GlobalValue::ExternalLinkage, 707 /*init*/ nullptr, SGV->getName(), 708 /*insertbefore*/ nullptr, 709 SGV->getThreadLocalMode(), SGV->getAddressSpace()); 710 } 711 712 if (ForDefinition) 713 NewGV->setLinkage(SGV->getLinkage()); 714 else if (SGV->hasExternalWeakLinkage()) 715 NewGV->setLinkage(GlobalValue::ExternalWeakLinkage); 716 717 if (auto *NewGO = dyn_cast<GlobalObject>(NewGV)) { 718 // Metadata for global variables and function declarations is copied eagerly. 719 if (isa<GlobalVariable>(SGV) || SGV->isDeclaration()) 720 NewGO->copyMetadata(cast<GlobalObject>(SGV), 0); 721 } 722 723 // Remove these copied constants in case this stays a declaration, since 724 // they point to the source module. If the def is linked the values will 725 // be mapped in during linkFunctionBody. 726 if (auto *NewF = dyn_cast<Function>(NewGV)) { 727 NewF->setPersonalityFn(nullptr); 728 NewF->setPrefixData(nullptr); 729 NewF->setPrologueData(nullptr); 730 } 731 732 return NewGV; 733 } 734 735 static StringRef getTypeNamePrefix(StringRef Name) { 736 size_t DotPos = Name.rfind('.'); 737 return (DotPos == 0 || DotPos == StringRef::npos || Name.back() == '.' || 738 !isdigit(static_cast<unsigned char>(Name[DotPos + 1]))) 739 ? Name 740 : Name.substr(0, DotPos); 741 } 742 743 /// Loop over all of the linked values to compute type mappings. For example, 744 /// if we link "extern Foo *x" and "Foo *x = NULL", then we have two struct 745 /// types 'Foo' but one got renamed when the module was loaded into the same 746 /// LLVMContext. 747 void IRLinker::computeTypeMapping() { 748 for (GlobalValue &SGV : SrcM->globals()) { 749 GlobalValue *DGV = getLinkedToGlobal(&SGV); 750 if (!DGV) 751 continue; 752 753 if (!DGV->hasAppendingLinkage() || !SGV.hasAppendingLinkage()) { 754 TypeMap.addTypeMapping(DGV->getType(), SGV.getType()); 755 continue; 756 } 757 758 // Unify the element type of appending arrays. 759 ArrayType *DAT = cast<ArrayType>(DGV->getValueType()); 760 ArrayType *SAT = cast<ArrayType>(SGV.getValueType()); 761 TypeMap.addTypeMapping(DAT->getElementType(), SAT->getElementType()); 762 } 763 764 for (GlobalValue &SGV : *SrcM) 765 if (GlobalValue *DGV = getLinkedToGlobal(&SGV)) { 766 if (DGV->getType() == SGV.getType()) { 767 // If the types of DGV and SGV are the same, it means that DGV is from 768 // the source module and got added to DstM from a shared metadata. We 769 // shouldn't map this type to itself in case the type's components get 770 // remapped to a new type from DstM (for instance, during the loop over 771 // SrcM->getIdentifiedStructTypes() below). 772 continue; 773 } 774 775 TypeMap.addTypeMapping(DGV->getType(), SGV.getType()); 776 } 777 778 for (GlobalValue &SGV : SrcM->aliases()) 779 if (GlobalValue *DGV = getLinkedToGlobal(&SGV)) 780 TypeMap.addTypeMapping(DGV->getType(), SGV.getType()); 781 782 // Incorporate types by name, scanning all the types in the source module. 783 // At this point, the destination module may have a type "%foo = { i32 }" for 784 // example. When the source module got loaded into the same LLVMContext, if 785 // it had the same type, it would have been renamed to "%foo.42 = { i32 }". 786 std::vector<StructType *> Types = SrcM->getIdentifiedStructTypes(); 787 for (StructType *ST : Types) { 788 if (!ST->hasName()) 789 continue; 790 791 if (TypeMap.DstStructTypesSet.hasType(ST)) { 792 // This is actually a type from the destination module. 793 // getIdentifiedStructTypes() can have found it by walking debug info 794 // metadata nodes, some of which get linked by name when ODR Type Uniquing 795 // is enabled on the Context, from the source to the destination module. 796 continue; 797 } 798 799 auto STTypePrefix = getTypeNamePrefix(ST->getName()); 800 if (STTypePrefix.size()== ST->getName().size()) 801 continue; 802 803 // Check to see if the destination module has a struct with the prefix name. 804 StructType *DST = DstM.getTypeByName(STTypePrefix); 805 if (!DST) 806 continue; 807 808 // Don't use it if this actually came from the source module. They're in 809 // the same LLVMContext after all. Also don't use it unless the type is 810 // actually used in the destination module. This can happen in situations 811 // like this: 812 // 813 // Module A Module B 814 // -------- -------- 815 // %Z = type { %A } %B = type { %C.1 } 816 // %A = type { %B.1, [7 x i8] } %C.1 = type { i8* } 817 // %B.1 = type { %C } %A.2 = type { %B.3, [5 x i8] } 818 // %C = type { i8* } %B.3 = type { %C.1 } 819 // 820 // When we link Module B with Module A, the '%B' in Module B is 821 // used. However, that would then use '%C.1'. But when we process '%C.1', 822 // we prefer to take the '%C' version. So we are then left with both 823 // '%C.1' and '%C' being used for the same types. This leads to some 824 // variables using one type and some using the other. 825 if (TypeMap.DstStructTypesSet.hasType(DST)) 826 TypeMap.addTypeMapping(DST, ST); 827 } 828 829 // Now that we have discovered all of the type equivalences, get a body for 830 // any 'opaque' types in the dest module that are now resolved. 831 TypeMap.linkDefinedTypeBodies(); 832 } 833 834 static void getArrayElements(const Constant *C, 835 SmallVectorImpl<Constant *> &Dest) { 836 unsigned NumElements = cast<ArrayType>(C->getType())->getNumElements(); 837 838 for (unsigned i = 0; i != NumElements; ++i) 839 Dest.push_back(C->getAggregateElement(i)); 840 } 841 842 /// If there were any appending global variables, link them together now. 843 Expected<Constant *> 844 IRLinker::linkAppendingVarProto(GlobalVariable *DstGV, 845 const GlobalVariable *SrcGV) { 846 Type *EltTy = cast<ArrayType>(TypeMap.get(SrcGV->getValueType())) 847 ->getElementType(); 848 849 // FIXME: This upgrade is done during linking to support the C API. Once the 850 // old form is deprecated, we should move this upgrade to 851 // llvm::UpgradeGlobalVariable() and simplify the logic here and in 852 // Mapper::mapAppendingVariable() in ValueMapper.cpp. 853 StringRef Name = SrcGV->getName(); 854 bool IsNewStructor = false; 855 bool IsOldStructor = false; 856 if (Name == "llvm.global_ctors" || Name == "llvm.global_dtors") { 857 if (cast<StructType>(EltTy)->getNumElements() == 3) 858 IsNewStructor = true; 859 else 860 IsOldStructor = true; 861 } 862 863 PointerType *VoidPtrTy = Type::getInt8Ty(SrcGV->getContext())->getPointerTo(); 864 if (IsOldStructor) { 865 auto &ST = *cast<StructType>(EltTy); 866 Type *Tys[3] = {ST.getElementType(0), ST.getElementType(1), VoidPtrTy}; 867 EltTy = StructType::get(SrcGV->getContext(), Tys, false); 868 } 869 870 uint64_t DstNumElements = 0; 871 if (DstGV) { 872 ArrayType *DstTy = cast<ArrayType>(DstGV->getValueType()); 873 DstNumElements = DstTy->getNumElements(); 874 875 if (!SrcGV->hasAppendingLinkage() || !DstGV->hasAppendingLinkage()) 876 return stringErr( 877 "Linking globals named '" + SrcGV->getName() + 878 "': can only link appending global with another appending " 879 "global!"); 880 881 // Check to see that they two arrays agree on type. 882 if (EltTy != DstTy->getElementType()) 883 return stringErr("Appending variables with different element types!"); 884 if (DstGV->isConstant() != SrcGV->isConstant()) 885 return stringErr("Appending variables linked with different const'ness!"); 886 887 if (DstGV->getAlignment() != SrcGV->getAlignment()) 888 return stringErr( 889 "Appending variables with different alignment need to be linked!"); 890 891 if (DstGV->getVisibility() != SrcGV->getVisibility()) 892 return stringErr( 893 "Appending variables with different visibility need to be linked!"); 894 895 if (DstGV->hasGlobalUnnamedAddr() != SrcGV->hasGlobalUnnamedAddr()) 896 return stringErr( 897 "Appending variables with different unnamed_addr need to be linked!"); 898 899 if (DstGV->getSection() != SrcGV->getSection()) 900 return stringErr( 901 "Appending variables with different section name need to be linked!"); 902 } 903 904 SmallVector<Constant *, 16> SrcElements; 905 getArrayElements(SrcGV->getInitializer(), SrcElements); 906 907 if (IsNewStructor) { 908 auto It = remove_if(SrcElements, [this](Constant *E) { 909 auto *Key = 910 dyn_cast<GlobalValue>(E->getAggregateElement(2)->stripPointerCasts()); 911 if (!Key) 912 return false; 913 GlobalValue *DGV = getLinkedToGlobal(Key); 914 return !shouldLink(DGV, *Key); 915 }); 916 SrcElements.erase(It, SrcElements.end()); 917 } 918 uint64_t NewSize = DstNumElements + SrcElements.size(); 919 ArrayType *NewType = ArrayType::get(EltTy, NewSize); 920 921 // Create the new global variable. 922 GlobalVariable *NG = new GlobalVariable( 923 DstM, NewType, SrcGV->isConstant(), SrcGV->getLinkage(), 924 /*init*/ nullptr, /*name*/ "", DstGV, SrcGV->getThreadLocalMode(), 925 SrcGV->getAddressSpace()); 926 927 NG->copyAttributesFrom(SrcGV); 928 forceRenaming(NG, SrcGV->getName()); 929 930 Constant *Ret = ConstantExpr::getBitCast(NG, TypeMap.get(SrcGV->getType())); 931 932 Mapper.scheduleMapAppendingVariable(*NG, 933 DstGV ? DstGV->getInitializer() : nullptr, 934 IsOldStructor, SrcElements); 935 936 // Replace any uses of the two global variables with uses of the new 937 // global. 938 if (DstGV) { 939 RAUWWorklist.push_back( 940 std::make_pair(DstGV, ConstantExpr::getBitCast(NG, DstGV->getType()))); 941 } 942 943 return Ret; 944 } 945 946 bool IRLinker::shouldLink(GlobalValue *DGV, GlobalValue &SGV) { 947 if (ValuesToLink.count(&SGV) || SGV.hasLocalLinkage()) 948 return true; 949 950 if (DGV && !DGV->isDeclarationForLinker()) 951 return false; 952 953 if (SGV.isDeclaration() || DoneLinkingBodies) 954 return false; 955 956 // Callback to the client to give a chance to lazily add the Global to the 957 // list of value to link. 958 bool LazilyAdded = false; 959 AddLazyFor(SGV, [this, &LazilyAdded](GlobalValue &GV) { 960 maybeAdd(&GV); 961 LazilyAdded = true; 962 }); 963 return LazilyAdded; 964 } 965 966 Expected<Constant *> IRLinker::linkGlobalValueProto(GlobalValue *SGV, 967 bool ForIndirectSymbol) { 968 GlobalValue *DGV = getLinkedToGlobal(SGV); 969 970 bool ShouldLink = shouldLink(DGV, *SGV); 971 972 // just missing from map 973 if (ShouldLink) { 974 auto I = ValueMap.find(SGV); 975 if (I != ValueMap.end()) 976 return cast<Constant>(I->second); 977 978 I = IndirectSymbolValueMap.find(SGV); 979 if (I != IndirectSymbolValueMap.end()) 980 return cast<Constant>(I->second); 981 } 982 983 if (!ShouldLink && ForIndirectSymbol) 984 DGV = nullptr; 985 986 // Handle the ultra special appending linkage case first. 987 assert(!DGV || SGV->hasAppendingLinkage() == DGV->hasAppendingLinkage()); 988 if (SGV->hasAppendingLinkage()) 989 return linkAppendingVarProto(cast_or_null<GlobalVariable>(DGV), 990 cast<GlobalVariable>(SGV)); 991 992 GlobalValue *NewGV; 993 if (DGV && !ShouldLink) { 994 NewGV = DGV; 995 } else { 996 // If we are done linking global value bodies (i.e. we are performing 997 // metadata linking), don't link in the global value due to this 998 // reference, simply map it to null. 999 if (DoneLinkingBodies) 1000 return nullptr; 1001 1002 NewGV = copyGlobalValueProto(SGV, ShouldLink || ForIndirectSymbol); 1003 if (ShouldLink || !ForIndirectSymbol) 1004 forceRenaming(NewGV, SGV->getName()); 1005 } 1006 1007 // Overloaded intrinsics have overloaded types names as part of their 1008 // names. If we renamed overloaded types we should rename the intrinsic 1009 // as well. 1010 if (Function *F = dyn_cast<Function>(NewGV)) 1011 if (auto Remangled = Intrinsic::remangleIntrinsicFunction(F)) 1012 NewGV = Remangled.getValue(); 1013 1014 if (ShouldLink || ForIndirectSymbol) { 1015 if (const Comdat *SC = SGV->getComdat()) { 1016 if (auto *GO = dyn_cast<GlobalObject>(NewGV)) { 1017 Comdat *DC = DstM.getOrInsertComdat(SC->getName()); 1018 DC->setSelectionKind(SC->getSelectionKind()); 1019 GO->setComdat(DC); 1020 } 1021 } 1022 } 1023 1024 if (!ShouldLink && ForIndirectSymbol) 1025 NewGV->setLinkage(GlobalValue::InternalLinkage); 1026 1027 Constant *C = NewGV; 1028 // Only create a bitcast if necessary. In particular, with 1029 // DebugTypeODRUniquing we may reach metadata in the destination module 1030 // containing a GV from the source module, in which case SGV will be 1031 // the same as DGV and NewGV, and TypeMap.get() will assert since it 1032 // assumes it is being invoked on a type in the source module. 1033 if (DGV && NewGV != SGV) { 1034 C = ConstantExpr::getPointerBitCastOrAddrSpaceCast( 1035 NewGV, TypeMap.get(SGV->getType())); 1036 } 1037 1038 if (DGV && NewGV != DGV) { 1039 // Schedule "replace all uses with" to happen after materializing is 1040 // done. It is not safe to do it now, since ValueMapper may be holding 1041 // pointers to constants that will get deleted if RAUW runs. 1042 RAUWWorklist.push_back(std::make_pair( 1043 DGV, 1044 ConstantExpr::getPointerBitCastOrAddrSpaceCast(NewGV, DGV->getType()))); 1045 } 1046 1047 return C; 1048 } 1049 1050 /// Update the initializers in the Dest module now that all globals that may be 1051 /// referenced are in Dest. 1052 void IRLinker::linkGlobalVariable(GlobalVariable &Dst, GlobalVariable &Src) { 1053 // Figure out what the initializer looks like in the dest module. 1054 Mapper.scheduleMapGlobalInitializer(Dst, *Src.getInitializer()); 1055 } 1056 1057 /// Copy the source function over into the dest function and fix up references 1058 /// to values. At this point we know that Dest is an external function, and 1059 /// that Src is not. 1060 Error IRLinker::linkFunctionBody(Function &Dst, Function &Src) { 1061 assert(Dst.isDeclaration() && !Src.isDeclaration()); 1062 1063 // Materialize if needed. 1064 if (Error Err = Src.materialize()) 1065 return Err; 1066 1067 // Link in the operands without remapping. 1068 if (Src.hasPrefixData()) 1069 Dst.setPrefixData(Src.getPrefixData()); 1070 if (Src.hasPrologueData()) 1071 Dst.setPrologueData(Src.getPrologueData()); 1072 if (Src.hasPersonalityFn()) 1073 Dst.setPersonalityFn(Src.getPersonalityFn()); 1074 1075 // Copy over the metadata attachments without remapping. 1076 Dst.copyMetadata(&Src, 0); 1077 1078 // Steal arguments and splice the body of Src into Dst. 1079 Dst.stealArgumentListFrom(Src); 1080 Dst.getBasicBlockList().splice(Dst.end(), Src.getBasicBlockList()); 1081 1082 // Everything has been moved over. Remap it. 1083 Mapper.scheduleRemapFunction(Dst); 1084 return Error::success(); 1085 } 1086 1087 void IRLinker::linkIndirectSymbolBody(GlobalIndirectSymbol &Dst, 1088 GlobalIndirectSymbol &Src) { 1089 Mapper.scheduleMapGlobalIndirectSymbol(Dst, *Src.getIndirectSymbol(), 1090 IndirectSymbolMCID); 1091 } 1092 1093 Error IRLinker::linkGlobalValueBody(GlobalValue &Dst, GlobalValue &Src) { 1094 if (auto *F = dyn_cast<Function>(&Src)) 1095 return linkFunctionBody(cast<Function>(Dst), *F); 1096 if (auto *GVar = dyn_cast<GlobalVariable>(&Src)) { 1097 linkGlobalVariable(cast<GlobalVariable>(Dst), *GVar); 1098 return Error::success(); 1099 } 1100 linkIndirectSymbolBody(cast<GlobalIndirectSymbol>(Dst), cast<GlobalIndirectSymbol>(Src)); 1101 return Error::success(); 1102 } 1103 1104 void IRLinker::flushRAUWWorklist() { 1105 for (const auto &Elem : RAUWWorklist) { 1106 GlobalValue *Old; 1107 Value *New; 1108 std::tie(Old, New) = Elem; 1109 1110 Old->replaceAllUsesWith(New); 1111 Old->eraseFromParent(); 1112 } 1113 RAUWWorklist.clear(); 1114 } 1115 1116 void IRLinker::prepareCompileUnitsForImport() { 1117 NamedMDNode *SrcCompileUnits = SrcM->getNamedMetadata("llvm.dbg.cu"); 1118 if (!SrcCompileUnits) 1119 return; 1120 // When importing for ThinLTO, prevent importing of types listed on 1121 // the DICompileUnit that we don't need a copy of in the importing 1122 // module. They will be emitted by the originating module. 1123 for (unsigned I = 0, E = SrcCompileUnits->getNumOperands(); I != E; ++I) { 1124 auto *CU = cast<DICompileUnit>(SrcCompileUnits->getOperand(I)); 1125 assert(CU && "Expected valid compile unit"); 1126 // Enums, macros, and retained types don't need to be listed on the 1127 // imported DICompileUnit. This means they will only be imported 1128 // if reached from the mapped IR. Do this by setting their value map 1129 // entries to nullptr, which will automatically prevent their importing 1130 // when reached from the DICompileUnit during metadata mapping. 1131 ValueMap.MD()[CU->getRawEnumTypes()].reset(nullptr); 1132 ValueMap.MD()[CU->getRawMacros()].reset(nullptr); 1133 ValueMap.MD()[CU->getRawRetainedTypes()].reset(nullptr); 1134 // The original definition (or at least its debug info - if the variable is 1135 // internalized an optimized away) will remain in the source module, so 1136 // there's no need to import them. 1137 // If LLVM ever does more advanced optimizations on global variables 1138 // (removing/localizing write operations, for instance) that can track 1139 // through debug info, this decision may need to be revisited - but do so 1140 // with care when it comes to debug info size. Emitting small CUs containing 1141 // only a few imported entities into every destination module may be very 1142 // size inefficient. 1143 ValueMap.MD()[CU->getRawGlobalVariables()].reset(nullptr); 1144 1145 // Imported entities only need to be mapped in if they have local 1146 // scope, as those might correspond to an imported entity inside a 1147 // function being imported (any locally scoped imported entities that 1148 // don't end up referenced by an imported function will not be emitted 1149 // into the object). Imported entities not in a local scope 1150 // (e.g. on the namespace) only need to be emitted by the originating 1151 // module. Create a list of the locally scoped imported entities, and 1152 // replace the source CUs imported entity list with the new list, so 1153 // only those are mapped in. 1154 // FIXME: Locally-scoped imported entities could be moved to the 1155 // functions they are local to instead of listing them on the CU, and 1156 // we would naturally only link in those needed by function importing. 1157 SmallVector<TrackingMDNodeRef, 4> AllImportedModules; 1158 bool ReplaceImportedEntities = false; 1159 for (auto *IE : CU->getImportedEntities()) { 1160 DIScope *Scope = IE->getScope(); 1161 assert(Scope && "Invalid Scope encoding!"); 1162 if (isa<DILocalScope>(Scope)) 1163 AllImportedModules.emplace_back(IE); 1164 else 1165 ReplaceImportedEntities = true; 1166 } 1167 if (ReplaceImportedEntities) { 1168 if (!AllImportedModules.empty()) 1169 CU->replaceImportedEntities(MDTuple::get( 1170 CU->getContext(), 1171 SmallVector<Metadata *, 16>(AllImportedModules.begin(), 1172 AllImportedModules.end()))); 1173 else 1174 // If there were no local scope imported entities, we can map 1175 // the whole list to nullptr. 1176 ValueMap.MD()[CU->getRawImportedEntities()].reset(nullptr); 1177 } 1178 } 1179 } 1180 1181 /// Insert all of the named MDNodes in Src into the Dest module. 1182 void IRLinker::linkNamedMDNodes() { 1183 const NamedMDNode *SrcModFlags = SrcM->getModuleFlagsMetadata(); 1184 for (const NamedMDNode &NMD : SrcM->named_metadata()) { 1185 // Don't link module flags here. Do them separately. 1186 if (&NMD == SrcModFlags) 1187 continue; 1188 NamedMDNode *DestNMD = DstM.getOrInsertNamedMetadata(NMD.getName()); 1189 // Add Src elements into Dest node. 1190 for (const MDNode *Op : NMD.operands()) 1191 DestNMD->addOperand(Mapper.mapMDNode(*Op)); 1192 } 1193 } 1194 1195 /// Merge the linker flags in Src into the Dest module. 1196 Error IRLinker::linkModuleFlagsMetadata() { 1197 // If the source module has no module flags, we are done. 1198 const NamedMDNode *SrcModFlags = SrcM->getModuleFlagsMetadata(); 1199 if (!SrcModFlags) 1200 return Error::success(); 1201 1202 // If the destination module doesn't have module flags yet, then just copy 1203 // over the source module's flags. 1204 NamedMDNode *DstModFlags = DstM.getOrInsertModuleFlagsMetadata(); 1205 if (DstModFlags->getNumOperands() == 0) { 1206 for (unsigned I = 0, E = SrcModFlags->getNumOperands(); I != E; ++I) 1207 DstModFlags->addOperand(SrcModFlags->getOperand(I)); 1208 1209 return Error::success(); 1210 } 1211 1212 // First build a map of the existing module flags and requirements. 1213 DenseMap<MDString *, std::pair<MDNode *, unsigned>> Flags; 1214 SmallSetVector<MDNode *, 16> Requirements; 1215 for (unsigned I = 0, E = DstModFlags->getNumOperands(); I != E; ++I) { 1216 MDNode *Op = DstModFlags->getOperand(I); 1217 ConstantInt *Behavior = mdconst::extract<ConstantInt>(Op->getOperand(0)); 1218 MDString *ID = cast<MDString>(Op->getOperand(1)); 1219 1220 if (Behavior->getZExtValue() == Module::Require) { 1221 Requirements.insert(cast<MDNode>(Op->getOperand(2))); 1222 } else { 1223 Flags[ID] = std::make_pair(Op, I); 1224 } 1225 } 1226 1227 // Merge in the flags from the source module, and also collect its set of 1228 // requirements. 1229 for (unsigned I = 0, E = SrcModFlags->getNumOperands(); I != E; ++I) { 1230 MDNode *SrcOp = SrcModFlags->getOperand(I); 1231 ConstantInt *SrcBehavior = 1232 mdconst::extract<ConstantInt>(SrcOp->getOperand(0)); 1233 MDString *ID = cast<MDString>(SrcOp->getOperand(1)); 1234 MDNode *DstOp; 1235 unsigned DstIndex; 1236 std::tie(DstOp, DstIndex) = Flags.lookup(ID); 1237 unsigned SrcBehaviorValue = SrcBehavior->getZExtValue(); 1238 1239 // If this is a requirement, add it and continue. 1240 if (SrcBehaviorValue == Module::Require) { 1241 // If the destination module does not already have this requirement, add 1242 // it. 1243 if (Requirements.insert(cast<MDNode>(SrcOp->getOperand(2)))) { 1244 DstModFlags->addOperand(SrcOp); 1245 } 1246 continue; 1247 } 1248 1249 // If there is no existing flag with this ID, just add it. 1250 if (!DstOp) { 1251 Flags[ID] = std::make_pair(SrcOp, DstModFlags->getNumOperands()); 1252 DstModFlags->addOperand(SrcOp); 1253 continue; 1254 } 1255 1256 // Otherwise, perform a merge. 1257 ConstantInt *DstBehavior = 1258 mdconst::extract<ConstantInt>(DstOp->getOperand(0)); 1259 unsigned DstBehaviorValue = DstBehavior->getZExtValue(); 1260 1261 auto overrideDstValue = [&]() { 1262 DstModFlags->setOperand(DstIndex, SrcOp); 1263 Flags[ID].first = SrcOp; 1264 }; 1265 1266 // If either flag has override behavior, handle it first. 1267 if (DstBehaviorValue == Module::Override) { 1268 // Diagnose inconsistent flags which both have override behavior. 1269 if (SrcBehaviorValue == Module::Override && 1270 SrcOp->getOperand(2) != DstOp->getOperand(2)) 1271 return stringErr("linking module flags '" + ID->getString() + 1272 "': IDs have conflicting override values in '" + 1273 SrcM->getModuleIdentifier() + "' and '" + 1274 DstM.getModuleIdentifier() + "'"); 1275 continue; 1276 } else if (SrcBehaviorValue == Module::Override) { 1277 // Update the destination flag to that of the source. 1278 overrideDstValue(); 1279 continue; 1280 } 1281 1282 // Diagnose inconsistent merge behavior types. 1283 if (SrcBehaviorValue != DstBehaviorValue) { 1284 bool MaxAndWarn = (SrcBehaviorValue == Module::Max && 1285 DstBehaviorValue == Module::Warning) || 1286 (DstBehaviorValue == Module::Max && 1287 SrcBehaviorValue == Module::Warning); 1288 if (!MaxAndWarn) 1289 return stringErr("linking module flags '" + ID->getString() + 1290 "': IDs have conflicting behaviors in '" + 1291 SrcM->getModuleIdentifier() + "' and '" + 1292 DstM.getModuleIdentifier() + "'"); 1293 } 1294 1295 auto replaceDstValue = [&](MDNode *New) { 1296 Metadata *FlagOps[] = {DstOp->getOperand(0), ID, New}; 1297 MDNode *Flag = MDNode::get(DstM.getContext(), FlagOps); 1298 DstModFlags->setOperand(DstIndex, Flag); 1299 Flags[ID].first = Flag; 1300 }; 1301 1302 // Emit a warning if the values differ and either source or destination 1303 // request Warning behavior. 1304 if ((DstBehaviorValue == Module::Warning || 1305 SrcBehaviorValue == Module::Warning) && 1306 SrcOp->getOperand(2) != DstOp->getOperand(2)) { 1307 std::string Str; 1308 raw_string_ostream(Str) 1309 << "linking module flags '" << ID->getString() 1310 << "': IDs have conflicting values ('" << *SrcOp->getOperand(2) 1311 << "' from " << SrcM->getModuleIdentifier() << " with '" 1312 << *DstOp->getOperand(2) << "' from " << DstM.getModuleIdentifier() 1313 << ')'; 1314 emitWarning(Str); 1315 } 1316 1317 // Choose the maximum if either source or destination request Max behavior. 1318 if (DstBehaviorValue == Module::Max || SrcBehaviorValue == Module::Max) { 1319 ConstantInt *DstValue = 1320 mdconst::extract<ConstantInt>(DstOp->getOperand(2)); 1321 ConstantInt *SrcValue = 1322 mdconst::extract<ConstantInt>(SrcOp->getOperand(2)); 1323 1324 // The resulting flag should have a Max behavior, and contain the maximum 1325 // value from between the source and destination values. 1326 Metadata *FlagOps[] = { 1327 (DstBehaviorValue != Module::Max ? SrcOp : DstOp)->getOperand(0), ID, 1328 (SrcValue->getZExtValue() > DstValue->getZExtValue() ? SrcOp : DstOp) 1329 ->getOperand(2)}; 1330 MDNode *Flag = MDNode::get(DstM.getContext(), FlagOps); 1331 DstModFlags->setOperand(DstIndex, Flag); 1332 Flags[ID].first = Flag; 1333 continue; 1334 } 1335 1336 // Perform the merge for standard behavior types. 1337 switch (SrcBehaviorValue) { 1338 case Module::Require: 1339 case Module::Override: 1340 llvm_unreachable("not possible"); 1341 case Module::Error: { 1342 // Emit an error if the values differ. 1343 if (SrcOp->getOperand(2) != DstOp->getOperand(2)) 1344 return stringErr("linking module flags '" + ID->getString() + 1345 "': IDs have conflicting values in '" + 1346 SrcM->getModuleIdentifier() + "' and '" + 1347 DstM.getModuleIdentifier() + "'"); 1348 continue; 1349 } 1350 case Module::Warning: { 1351 break; 1352 } 1353 case Module::Max: { 1354 break; 1355 } 1356 case Module::Append: { 1357 MDNode *DstValue = cast<MDNode>(DstOp->getOperand(2)); 1358 MDNode *SrcValue = cast<MDNode>(SrcOp->getOperand(2)); 1359 SmallVector<Metadata *, 8> MDs; 1360 MDs.reserve(DstValue->getNumOperands() + SrcValue->getNumOperands()); 1361 MDs.append(DstValue->op_begin(), DstValue->op_end()); 1362 MDs.append(SrcValue->op_begin(), SrcValue->op_end()); 1363 1364 replaceDstValue(MDNode::get(DstM.getContext(), MDs)); 1365 break; 1366 } 1367 case Module::AppendUnique: { 1368 SmallSetVector<Metadata *, 16> Elts; 1369 MDNode *DstValue = cast<MDNode>(DstOp->getOperand(2)); 1370 MDNode *SrcValue = cast<MDNode>(SrcOp->getOperand(2)); 1371 Elts.insert(DstValue->op_begin(), DstValue->op_end()); 1372 Elts.insert(SrcValue->op_begin(), SrcValue->op_end()); 1373 1374 replaceDstValue(MDNode::get(DstM.getContext(), 1375 makeArrayRef(Elts.begin(), Elts.end()))); 1376 break; 1377 } 1378 } 1379 1380 } 1381 1382 // Check all of the requirements. 1383 for (unsigned I = 0, E = Requirements.size(); I != E; ++I) { 1384 MDNode *Requirement = Requirements[I]; 1385 MDString *Flag = cast<MDString>(Requirement->getOperand(0)); 1386 Metadata *ReqValue = Requirement->getOperand(1); 1387 1388 MDNode *Op = Flags[Flag].first; 1389 if (!Op || Op->getOperand(2) != ReqValue) 1390 return stringErr("linking module flags '" + Flag->getString() + 1391 "': does not have the required value"); 1392 } 1393 return Error::success(); 1394 } 1395 1396 /// Return InlineAsm adjusted with target-specific directives if required. 1397 /// For ARM and Thumb, we have to add directives to select the appropriate ISA 1398 /// to support mixing module-level inline assembly from ARM and Thumb modules. 1399 static std::string adjustInlineAsm(const std::string &InlineAsm, 1400 const Triple &Triple) { 1401 if (Triple.getArch() == Triple::thumb || Triple.getArch() == Triple::thumbeb) 1402 return ".text\n.balign 2\n.thumb\n" + InlineAsm; 1403 if (Triple.getArch() == Triple::arm || Triple.getArch() == Triple::armeb) 1404 return ".text\n.balign 4\n.arm\n" + InlineAsm; 1405 return InlineAsm; 1406 } 1407 1408 Error IRLinker::run() { 1409 // Ensure metadata materialized before value mapping. 1410 if (SrcM->getMaterializer()) 1411 if (Error Err = SrcM->getMaterializer()->materializeMetadata()) 1412 return Err; 1413 1414 // Inherit the target data from the source module if the destination module 1415 // doesn't have one already. 1416 if (DstM.getDataLayout().isDefault()) 1417 DstM.setDataLayout(SrcM->getDataLayout()); 1418 1419 if (SrcM->getDataLayout() != DstM.getDataLayout()) { 1420 emitWarning("Linking two modules of different data layouts: '" + 1421 SrcM->getModuleIdentifier() + "' is '" + 1422 SrcM->getDataLayoutStr() + "' whereas '" + 1423 DstM.getModuleIdentifier() + "' is '" + 1424 DstM.getDataLayoutStr() + "'\n"); 1425 } 1426 1427 // Copy the target triple from the source to dest if the dest's is empty. 1428 if (DstM.getTargetTriple().empty() && !SrcM->getTargetTriple().empty()) 1429 DstM.setTargetTriple(SrcM->getTargetTriple()); 1430 1431 Triple SrcTriple(SrcM->getTargetTriple()), DstTriple(DstM.getTargetTriple()); 1432 1433 if (!SrcM->getTargetTriple().empty()&& 1434 !SrcTriple.isCompatibleWith(DstTriple)) 1435 emitWarning("Linking two modules of different target triples: " + 1436 SrcM->getModuleIdentifier() + "' is '" + 1437 SrcM->getTargetTriple() + "' whereas '" + 1438 DstM.getModuleIdentifier() + "' is '" + DstM.getTargetTriple() + 1439 "'\n"); 1440 1441 DstM.setTargetTriple(SrcTriple.merge(DstTriple)); 1442 1443 // Append the module inline asm string. 1444 if (!IsPerformingImport && !SrcM->getModuleInlineAsm().empty()) { 1445 std::string SrcModuleInlineAsm = adjustInlineAsm(SrcM->getModuleInlineAsm(), 1446 SrcTriple); 1447 if (DstM.getModuleInlineAsm().empty()) 1448 DstM.setModuleInlineAsm(SrcModuleInlineAsm); 1449 else 1450 DstM.setModuleInlineAsm(DstM.getModuleInlineAsm() + "\n" + 1451 SrcModuleInlineAsm); 1452 } 1453 1454 // Loop over all of the linked values to compute type mappings. 1455 computeTypeMapping(); 1456 1457 std::reverse(Worklist.begin(), Worklist.end()); 1458 while (!Worklist.empty()) { 1459 GlobalValue *GV = Worklist.back(); 1460 Worklist.pop_back(); 1461 1462 // Already mapped. 1463 if (ValueMap.find(GV) != ValueMap.end() || 1464 IndirectSymbolValueMap.find(GV) != IndirectSymbolValueMap.end()) 1465 continue; 1466 1467 assert(!GV->isDeclaration()); 1468 Mapper.mapValue(*GV); 1469 if (FoundError) 1470 return std::move(*FoundError); 1471 flushRAUWWorklist(); 1472 } 1473 1474 // Note that we are done linking global value bodies. This prevents 1475 // metadata linking from creating new references. 1476 DoneLinkingBodies = true; 1477 Mapper.addFlags(RF_NullMapMissingGlobalValues); 1478 1479 // Remap all of the named MDNodes in Src into the DstM module. We do this 1480 // after linking GlobalValues so that MDNodes that reference GlobalValues 1481 // are properly remapped. 1482 linkNamedMDNodes(); 1483 1484 // Merge the module flags into the DstM module. 1485 return linkModuleFlagsMetadata(); 1486 } 1487 1488 IRMover::StructTypeKeyInfo::KeyTy::KeyTy(ArrayRef<Type *> E, bool P) 1489 : ETypes(E), IsPacked(P) {} 1490 1491 IRMover::StructTypeKeyInfo::KeyTy::KeyTy(const StructType *ST) 1492 : ETypes(ST->elements()), IsPacked(ST->isPacked()) {} 1493 1494 bool IRMover::StructTypeKeyInfo::KeyTy::operator==(const KeyTy &That) const { 1495 return IsPacked == That.IsPacked && ETypes == That.ETypes; 1496 } 1497 1498 bool IRMover::StructTypeKeyInfo::KeyTy::operator!=(const KeyTy &That) const { 1499 return !this->operator==(That); 1500 } 1501 1502 StructType *IRMover::StructTypeKeyInfo::getEmptyKey() { 1503 return DenseMapInfo<StructType *>::getEmptyKey(); 1504 } 1505 1506 StructType *IRMover::StructTypeKeyInfo::getTombstoneKey() { 1507 return DenseMapInfo<StructType *>::getTombstoneKey(); 1508 } 1509 1510 unsigned IRMover::StructTypeKeyInfo::getHashValue(const KeyTy &Key) { 1511 return hash_combine(hash_combine_range(Key.ETypes.begin(), Key.ETypes.end()), 1512 Key.IsPacked); 1513 } 1514 1515 unsigned IRMover::StructTypeKeyInfo::getHashValue(const StructType *ST) { 1516 return getHashValue(KeyTy(ST)); 1517 } 1518 1519 bool IRMover::StructTypeKeyInfo::isEqual(const KeyTy &LHS, 1520 const StructType *RHS) { 1521 if (RHS == getEmptyKey() || RHS == getTombstoneKey()) 1522 return false; 1523 return LHS == KeyTy(RHS); 1524 } 1525 1526 bool IRMover::StructTypeKeyInfo::isEqual(const StructType *LHS, 1527 const StructType *RHS) { 1528 if (RHS == getEmptyKey() || RHS == getTombstoneKey()) 1529 return LHS == RHS; 1530 return KeyTy(LHS) == KeyTy(RHS); 1531 } 1532 1533 void IRMover::IdentifiedStructTypeSet::addNonOpaque(StructType *Ty) { 1534 assert(!Ty->isOpaque()); 1535 NonOpaqueStructTypes.insert(Ty); 1536 } 1537 1538 void IRMover::IdentifiedStructTypeSet::switchToNonOpaque(StructType *Ty) { 1539 assert(!Ty->isOpaque()); 1540 NonOpaqueStructTypes.insert(Ty); 1541 bool Removed = OpaqueStructTypes.erase(Ty); 1542 (void)Removed; 1543 assert(Removed); 1544 } 1545 1546 void IRMover::IdentifiedStructTypeSet::addOpaque(StructType *Ty) { 1547 assert(Ty->isOpaque()); 1548 OpaqueStructTypes.insert(Ty); 1549 } 1550 1551 StructType * 1552 IRMover::IdentifiedStructTypeSet::findNonOpaque(ArrayRef<Type *> ETypes, 1553 bool IsPacked) { 1554 IRMover::StructTypeKeyInfo::KeyTy Key(ETypes, IsPacked); 1555 auto I = NonOpaqueStructTypes.find_as(Key); 1556 return I == NonOpaqueStructTypes.end() ? nullptr : *I; 1557 } 1558 1559 bool IRMover::IdentifiedStructTypeSet::hasType(StructType *Ty) { 1560 if (Ty->isOpaque()) 1561 return OpaqueStructTypes.count(Ty); 1562 auto I = NonOpaqueStructTypes.find(Ty); 1563 return I == NonOpaqueStructTypes.end() ? false : *I == Ty; 1564 } 1565 1566 IRMover::IRMover(Module &M) : Composite(M) { 1567 TypeFinder StructTypes; 1568 StructTypes.run(M, /* OnlyNamed */ false); 1569 for (StructType *Ty : StructTypes) { 1570 if (Ty->isOpaque()) 1571 IdentifiedStructTypes.addOpaque(Ty); 1572 else 1573 IdentifiedStructTypes.addNonOpaque(Ty); 1574 } 1575 // Self-map metadatas in the destination module. This is needed when 1576 // DebugTypeODRUniquing is enabled on the LLVMContext, since metadata in the 1577 // destination module may be reached from the source module. 1578 for (auto *MD : StructTypes.getVisitedMetadata()) { 1579 SharedMDs[MD].reset(const_cast<MDNode *>(MD)); 1580 } 1581 } 1582 1583 Error IRMover::move( 1584 std::unique_ptr<Module> Src, ArrayRef<GlobalValue *> ValuesToLink, 1585 std::function<void(GlobalValue &, ValueAdder Add)> AddLazyFor, 1586 bool IsPerformingImport) { 1587 IRLinker TheIRLinker(Composite, SharedMDs, IdentifiedStructTypes, 1588 std::move(Src), ValuesToLink, std::move(AddLazyFor), 1589 IsPerformingImport); 1590 Error E = TheIRLinker.run(); 1591 Composite.dropTriviallyDeadConstantArrays(); 1592 return E; 1593 } 1594