1 //===----- CompileOnDemandLayer.cpp - Lazily emit IR on first call --------===//
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/ExecutionEngine/Orc/CompileOnDemandLayer.h"
10 
11 #include "llvm/ADT/Hashing.h"
12 #include "llvm/ExecutionEngine/Orc/ExecutionUtils.h"
13 #include "llvm/IR/Mangler.h"
14 #include "llvm/IR/Module.h"
15 #include "llvm/Support/FormatVariadic.h"
16 
17 using namespace llvm;
18 using namespace llvm::orc;
19 
20 static ThreadSafeModule extractSubModule(ThreadSafeModule &TSM,
21                                          StringRef Suffix,
22                                          GVPredicate ShouldExtract) {
23 
24   auto DeleteExtractedDefs = [](GlobalValue &GV) {
25     // Bump the linkage: this global will be provided by the external module.
26     GV.setLinkage(GlobalValue::ExternalLinkage);
27 
28     // Delete the definition in the source module.
29     if (isa<Function>(GV)) {
30       auto &F = cast<Function>(GV);
31       F.deleteBody();
32       F.setPersonalityFn(nullptr);
33     } else if (isa<GlobalVariable>(GV)) {
34       cast<GlobalVariable>(GV).setInitializer(nullptr);
35     } else if (isa<GlobalAlias>(GV)) {
36       // We need to turn deleted aliases into function or variable decls based
37       // on the type of their aliasee.
38       auto &A = cast<GlobalAlias>(GV);
39       Constant *Aliasee = A.getAliasee();
40       assert(A.hasName() && "Anonymous alias?");
41       assert(Aliasee->hasName() && "Anonymous aliasee");
42       std::string AliasName = std::string(A.getName());
43 
44       if (isa<Function>(Aliasee)) {
45         auto *F = cloneFunctionDecl(*A.getParent(), *cast<Function>(Aliasee));
46         A.replaceAllUsesWith(F);
47         A.eraseFromParent();
48         F->setName(AliasName);
49       } else if (isa<GlobalVariable>(Aliasee)) {
50         auto *G = cloneGlobalVariableDecl(*A.getParent(),
51                                           *cast<GlobalVariable>(Aliasee));
52         A.replaceAllUsesWith(G);
53         A.eraseFromParent();
54         G->setName(AliasName);
55       } else
56         llvm_unreachable("Alias to unsupported type");
57     } else
58       llvm_unreachable("Unsupported global type");
59   };
60 
61   auto NewTSM = cloneToNewContext(TSM, ShouldExtract, DeleteExtractedDefs);
62   NewTSM.withModuleDo([&](Module &M) {
63     M.setModuleIdentifier((M.getModuleIdentifier() + Suffix).str());
64   });
65 
66   return NewTSM;
67 }
68 
69 namespace llvm {
70 namespace orc {
71 
72 class PartitioningIRMaterializationUnit : public IRMaterializationUnit {
73 public:
74   PartitioningIRMaterializationUnit(ExecutionSession &ES,
75                                     const IRSymbolMapper::ManglingOptions &MO,
76                                     ThreadSafeModule TSM,
77                                     CompileOnDemandLayer &Parent)
78       : IRMaterializationUnit(ES, MO, std::move(TSM)), Parent(Parent) {}
79 
80   PartitioningIRMaterializationUnit(
81       ThreadSafeModule TSM, SymbolFlagsMap SymbolFlags,
82       SymbolStringPtr InitSymbol, SymbolNameToDefinitionMap SymbolToDefinition,
83       CompileOnDemandLayer &Parent)
84       : IRMaterializationUnit(std::move(TSM), std::move(SymbolFlags),
85                               std::move(InitSymbol),
86                               std::move(SymbolToDefinition)),
87         Parent(Parent) {}
88 
89 private:
90   void materialize(std::unique_ptr<MaterializationResponsibility> R) override {
91     Parent.emitPartition(std::move(R), std::move(TSM),
92                          std::move(SymbolToDefinition));
93   }
94 
95   void discard(const JITDylib &V, const SymbolStringPtr &Name) override {
96     // All original symbols were materialized by the CODLayer and should be
97     // final. The function bodies provided by M should never be overridden.
98     llvm_unreachable("Discard should never be called on an "
99                      "ExtractingIRMaterializationUnit");
100   }
101 
102   mutable std::mutex SourceModuleMutex;
103   CompileOnDemandLayer &Parent;
104 };
105 
106 Optional<CompileOnDemandLayer::GlobalValueSet>
107 CompileOnDemandLayer::compileRequested(GlobalValueSet Requested) {
108   return std::move(Requested);
109 }
110 
111 Optional<CompileOnDemandLayer::GlobalValueSet>
112 CompileOnDemandLayer::compileWholeModule(GlobalValueSet Requested) {
113   return None;
114 }
115 
116 CompileOnDemandLayer::CompileOnDemandLayer(
117     ExecutionSession &ES, IRLayer &BaseLayer, LazyCallThroughManager &LCTMgr,
118     IndirectStubsManagerBuilder BuildIndirectStubsManager)
119     : IRLayer(ES, BaseLayer.getManglingOptions()), BaseLayer(BaseLayer),
120       LCTMgr(LCTMgr),
121       BuildIndirectStubsManager(std::move(BuildIndirectStubsManager)) {}
122 
123 void CompileOnDemandLayer::setPartitionFunction(PartitionFunction Partition) {
124   this->Partition = std::move(Partition);
125 }
126 
127 void CompileOnDemandLayer::setImplMap(ImplSymbolMap *Imp) {
128   this->AliaseeImpls = Imp;
129 }
130 void CompileOnDemandLayer::emit(
131     std::unique_ptr<MaterializationResponsibility> R, ThreadSafeModule TSM) {
132   assert(TSM && "Null module");
133 
134   auto &ES = getExecutionSession();
135 
136   // Sort the callables and non-callables, build re-exports and lodge the
137   // actual module with the implementation dylib.
138   auto &PDR = getPerDylibResources(R->getTargetJITDylib());
139 
140   SymbolAliasMap NonCallables;
141   SymbolAliasMap Callables;
142   TSM.withModuleDo([&](Module &M) {
143     // First, do some cleanup on the module:
144     cleanUpModule(M);
145   });
146 
147   for (auto &KV : R->getSymbols()) {
148     auto &Name = KV.first;
149     auto &Flags = KV.second;
150     if (Flags.isCallable())
151       Callables[Name] = SymbolAliasMapEntry(Name, Flags);
152     else
153       NonCallables[Name] = SymbolAliasMapEntry(Name, Flags);
154   }
155 
156   // Create a partitioning materialization unit and lodge it with the
157   // implementation dylib.
158   if (auto Err = PDR.getImplDylib().define(
159           std::make_unique<PartitioningIRMaterializationUnit>(
160               ES, *getManglingOptions(), std::move(TSM), *this))) {
161     ES.reportError(std::move(Err));
162     R->failMaterialization();
163     return;
164   }
165 
166   if (!NonCallables.empty())
167     if (auto Err =
168             R->replace(reexports(PDR.getImplDylib(), std::move(NonCallables),
169                                  JITDylibLookupFlags::MatchAllSymbols))) {
170       getExecutionSession().reportError(std::move(Err));
171       R->failMaterialization();
172       return;
173     }
174   if (!Callables.empty()) {
175     if (auto Err = R->replace(
176             lazyReexports(LCTMgr, PDR.getISManager(), PDR.getImplDylib(),
177                           std::move(Callables), AliaseeImpls))) {
178       getExecutionSession().reportError(std::move(Err));
179       R->failMaterialization();
180       return;
181     }
182   }
183 }
184 
185 CompileOnDemandLayer::PerDylibResources &
186 CompileOnDemandLayer::getPerDylibResources(JITDylib &TargetD) {
187   auto I = DylibResources.find(&TargetD);
188   if (I == DylibResources.end()) {
189     auto &ImplD =
190         getExecutionSession().createBareJITDylib(TargetD.getName() + ".impl");
191     JITDylibSearchOrder NewLinkOrder;
192     TargetD.withLinkOrderDo([&](const JITDylibSearchOrder &TargetLinkOrder) {
193       NewLinkOrder = TargetLinkOrder;
194     });
195 
196     assert(!NewLinkOrder.empty() && NewLinkOrder.front().first == &TargetD &&
197            NewLinkOrder.front().second ==
198                JITDylibLookupFlags::MatchAllSymbols &&
199            "TargetD must be at the front of its own search order and match "
200            "non-exported symbol");
201     NewLinkOrder.insert(std::next(NewLinkOrder.begin()),
202                         {&ImplD, JITDylibLookupFlags::MatchAllSymbols});
203     ImplD.setLinkOrder(NewLinkOrder, false);
204     TargetD.setLinkOrder(std::move(NewLinkOrder), false);
205 
206     PerDylibResources PDR(ImplD, BuildIndirectStubsManager());
207     I = DylibResources.insert(std::make_pair(&TargetD, std::move(PDR))).first;
208   }
209 
210   return I->second;
211 }
212 
213 void CompileOnDemandLayer::cleanUpModule(Module &M) {
214   for (auto &F : M.functions()) {
215     if (F.isDeclaration())
216       continue;
217 
218     if (F.hasAvailableExternallyLinkage()) {
219       F.deleteBody();
220       F.setPersonalityFn(nullptr);
221       continue;
222     }
223   }
224 }
225 
226 void CompileOnDemandLayer::expandPartition(GlobalValueSet &Partition) {
227   // Expands the partition to ensure the following rules hold:
228   // (1) If any alias is in the partition, its aliasee is also in the partition.
229   // (2) If any aliasee is in the partition, its aliases are also in the
230   //     partiton.
231   // (3) If any global variable is in the partition then all global variables
232   //     are in the partition.
233   assert(!Partition.empty() && "Unexpected empty partition");
234 
235   const Module &M = *(*Partition.begin())->getParent();
236   bool ContainsGlobalVariables = false;
237   std::vector<const GlobalValue *> GVsToAdd;
238 
239   for (auto *GV : Partition)
240     if (isa<GlobalAlias>(GV))
241       GVsToAdd.push_back(
242           cast<GlobalValue>(cast<GlobalAlias>(GV)->getAliasee()));
243     else if (isa<GlobalVariable>(GV))
244       ContainsGlobalVariables = true;
245 
246   for (auto &A : M.aliases())
247     if (Partition.count(cast<GlobalValue>(A.getAliasee())))
248       GVsToAdd.push_back(&A);
249 
250   if (ContainsGlobalVariables)
251     for (auto &G : M.globals())
252       GVsToAdd.push_back(&G);
253 
254   for (auto *GV : GVsToAdd)
255     Partition.insert(GV);
256 }
257 
258 void CompileOnDemandLayer::emitPartition(
259     std::unique_ptr<MaterializationResponsibility> R, ThreadSafeModule TSM,
260     IRMaterializationUnit::SymbolNameToDefinitionMap Defs) {
261 
262   // FIXME: Need a 'notify lazy-extracting/emitting' callback to tie the
263   //        extracted module key, extracted module, and source module key
264   //        together. This could be used, for example, to provide a specific
265   //        memory manager instance to the linking layer.
266 
267   auto &ES = getExecutionSession();
268   GlobalValueSet RequestedGVs;
269   for (auto &Name : R->getRequestedSymbols()) {
270     if (Name == R->getInitializerSymbol())
271       TSM.withModuleDo([&](Module &M) {
272         for (auto &GV : getStaticInitGVs(M))
273           RequestedGVs.insert(&GV);
274       });
275     else {
276       assert(Defs.count(Name) && "No definition for symbol");
277       RequestedGVs.insert(Defs[Name]);
278     }
279   }
280 
281   /// Perform partitioning with the context lock held, since the partition
282   /// function is allowed to access the globals to compute the partition.
283   auto GVsToExtract =
284       TSM.withModuleDo([&](Module &M) { return Partition(RequestedGVs); });
285 
286   // Take a 'None' partition to mean the whole module (as opposed to an empty
287   // partition, which means "materialize nothing"). Emit the whole module
288   // unmodified to the base layer.
289   if (GVsToExtract == None) {
290     Defs.clear();
291     BaseLayer.emit(std::move(R), std::move(TSM));
292     return;
293   }
294 
295   // If the partition is empty, return the whole module to the symbol table.
296   if (GVsToExtract->empty()) {
297     if (auto Err =
298             R->replace(std::make_unique<PartitioningIRMaterializationUnit>(
299                 std::move(TSM), R->getSymbols(), R->getInitializerSymbol(),
300                 std::move(Defs), *this))) {
301       getExecutionSession().reportError(std::move(Err));
302       R->failMaterialization();
303       return;
304     }
305     return;
306   }
307 
308   // Ok -- we actually need to partition the symbols. Promote the symbol
309   // linkages/names, expand the partition to include any required symbols
310   // (i.e. symbols that can't be separated from our partition), and
311   // then extract the partition.
312   //
313   // FIXME: We apply this promotion once per partitioning. It's safe, but
314   // overkill.
315   auto ExtractedTSM =
316       TSM.withModuleDo([&](Module &M) -> Expected<ThreadSafeModule> {
317         auto PromotedGlobals = PromoteSymbols(M);
318         if (!PromotedGlobals.empty()) {
319 
320           MangleAndInterner Mangle(ES, M.getDataLayout());
321           SymbolFlagsMap SymbolFlags;
322           IRSymbolMapper::add(ES, *getManglingOptions(),
323                               PromotedGlobals, SymbolFlags);
324 
325           if (auto Err = R->defineMaterializing(SymbolFlags))
326             return std::move(Err);
327         }
328 
329         expandPartition(*GVsToExtract);
330 
331         // Submodule name is given by hashing the names of the globals.
332         std::string SubModuleName;
333         {
334           std::vector<const GlobalValue*> HashGVs;
335           HashGVs.reserve(GVsToExtract->size());
336           for (auto *GV : *GVsToExtract)
337             HashGVs.push_back(GV);
338           llvm::sort(HashGVs, [](const GlobalValue *LHS, const GlobalValue *RHS) {
339               return LHS->getName() < RHS->getName();
340             });
341           hash_code HC(0);
342           for (auto *GV : HashGVs) {
343             assert(GV->hasName() && "All GVs to extract should be named by now");
344             auto GVName = GV->getName();
345             HC = hash_combine(HC, hash_combine_range(GVName.begin(), GVName.end()));
346           }
347           raw_string_ostream(SubModuleName)
348             << ".submodule."
349             << formatv(sizeof(size_t) == 8 ? "{0:x16}" : "{0:x8}",
350                        static_cast<size_t>(HC))
351             << ".ll";
352         }
353 
354         // Extract the requested partiton (plus any necessary aliases) and
355         // put the rest back into the impl dylib.
356         auto ShouldExtract = [&](const GlobalValue &GV) -> bool {
357           return GVsToExtract->count(&GV);
358         };
359 
360         return extractSubModule(TSM, SubModuleName , ShouldExtract);
361       });
362 
363   if (!ExtractedTSM) {
364     ES.reportError(ExtractedTSM.takeError());
365     R->failMaterialization();
366     return;
367   }
368 
369   if (auto Err = R->replace(std::make_unique<PartitioningIRMaterializationUnit>(
370           ES, *getManglingOptions(), std::move(TSM), *this))) {
371     ES.reportError(std::move(Err));
372     R->failMaterialization();
373     return;
374   }
375   BaseLayer.emit(std::move(R), std::move(*ExtractedTSM));
376 }
377 
378 } // end namespace orc
379 } // end namespace llvm
380