1 //===- SplitModule.cpp - Split a module into partitions -------------------===// 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 // This file defines the function llvm::SplitModule, which splits a module 10 // into multiple linkable partitions. It can be used to implement parallel code 11 // generation for link-time optimization. 12 // 13 //===----------------------------------------------------------------------===// 14 15 #include "llvm/Transforms/Utils/SplitModule.h" 16 #include "llvm/ADT/DenseMap.h" 17 #include "llvm/ADT/EquivalenceClasses.h" 18 #include "llvm/ADT/SmallPtrSet.h" 19 #include "llvm/ADT/SmallVector.h" 20 #include "llvm/ADT/StringRef.h" 21 #include "llvm/IR/Comdat.h" 22 #include "llvm/IR/Constant.h" 23 #include "llvm/IR/Constants.h" 24 #include "llvm/IR/Function.h" 25 #include "llvm/IR/GlobalAlias.h" 26 #include "llvm/IR/GlobalObject.h" 27 #include "llvm/IR/GlobalIndirectSymbol.h" 28 #include "llvm/IR/GlobalValue.h" 29 #include "llvm/IR/GlobalVariable.h" 30 #include "llvm/IR/Instruction.h" 31 #include "llvm/IR/Module.h" 32 #include "llvm/IR/User.h" 33 #include "llvm/IR/Value.h" 34 #include "llvm/Support/Casting.h" 35 #include "llvm/Support/Debug.h" 36 #include "llvm/Support/ErrorHandling.h" 37 #include "llvm/Support/MD5.h" 38 #include "llvm/Support/raw_ostream.h" 39 #include "llvm/Transforms/Utils/Cloning.h" 40 #include "llvm/Transforms/Utils/ValueMapper.h" 41 #include <algorithm> 42 #include <cassert> 43 #include <iterator> 44 #include <memory> 45 #include <queue> 46 #include <utility> 47 #include <vector> 48 49 using namespace llvm; 50 51 #define DEBUG_TYPE "split-module" 52 53 namespace { 54 55 using ClusterMapType = EquivalenceClasses<const GlobalValue *>; 56 using ComdatMembersType = DenseMap<const Comdat *, const GlobalValue *>; 57 using ClusterIDMapType = DenseMap<const GlobalValue *, unsigned>; 58 59 } // end anonymous namespace 60 61 static void addNonConstUser(ClusterMapType &GVtoClusterMap, 62 const GlobalValue *GV, const User *U) { 63 assert((!isa<Constant>(U) || isa<GlobalValue>(U)) && "Bad user"); 64 65 if (const Instruction *I = dyn_cast<Instruction>(U)) { 66 const GlobalValue *F = I->getParent()->getParent(); 67 GVtoClusterMap.unionSets(GV, F); 68 } else if (isa<GlobalIndirectSymbol>(U) || isa<Function>(U) || 69 isa<GlobalVariable>(U)) { 70 GVtoClusterMap.unionSets(GV, cast<GlobalValue>(U)); 71 } else { 72 llvm_unreachable("Underimplemented use case"); 73 } 74 } 75 76 // Adds all GlobalValue users of V to the same cluster as GV. 77 static void addAllGlobalValueUsers(ClusterMapType &GVtoClusterMap, 78 const GlobalValue *GV, const Value *V) { 79 for (auto *U : V->users()) { 80 SmallVector<const User *, 4> Worklist; 81 Worklist.push_back(U); 82 while (!Worklist.empty()) { 83 const User *UU = Worklist.pop_back_val(); 84 // For each constant that is not a GV (a pure const) recurse. 85 if (isa<Constant>(UU) && !isa<GlobalValue>(UU)) { 86 Worklist.append(UU->user_begin(), UU->user_end()); 87 continue; 88 } 89 addNonConstUser(GVtoClusterMap, GV, UU); 90 } 91 } 92 } 93 94 // Find partitions for module in the way that no locals need to be 95 // globalized. 96 // Try to balance pack those partitions into N files since this roughly equals 97 // thread balancing for the backend codegen step. 98 static void findPartitions(Module &M, ClusterIDMapType &ClusterIDMap, 99 unsigned N) { 100 // At this point module should have the proper mix of globals and locals. 101 // As we attempt to partition this module, we must not change any 102 // locals to globals. 103 LLVM_DEBUG(dbgs() << "Partition module with (" << M.size() << ")functions\n"); 104 ClusterMapType GVtoClusterMap; 105 ComdatMembersType ComdatMembers; 106 107 auto recordGVSet = [&GVtoClusterMap, &ComdatMembers](GlobalValue &GV) { 108 if (GV.isDeclaration()) 109 return; 110 111 if (!GV.hasName()) 112 GV.setName("__llvmsplit_unnamed"); 113 114 // Comdat groups must not be partitioned. For comdat groups that contain 115 // locals, record all their members here so we can keep them together. 116 // Comdat groups that only contain external globals are already handled by 117 // the MD5-based partitioning. 118 if (const Comdat *C = GV.getComdat()) { 119 auto &Member = ComdatMembers[C]; 120 if (Member) 121 GVtoClusterMap.unionSets(Member, &GV); 122 else 123 Member = &GV; 124 } 125 126 // For aliases we should not separate them from their aliasees regardless 127 // of linkage. 128 if (auto *GIS = dyn_cast<GlobalIndirectSymbol>(&GV)) { 129 if (const GlobalObject *Base = GIS->getBaseObject()) 130 GVtoClusterMap.unionSets(&GV, Base); 131 } 132 133 if (const Function *F = dyn_cast<Function>(&GV)) { 134 for (const BasicBlock &BB : *F) { 135 BlockAddress *BA = BlockAddress::lookup(&BB); 136 if (!BA || !BA->isConstantUsed()) 137 continue; 138 addAllGlobalValueUsers(GVtoClusterMap, F, BA); 139 } 140 } 141 142 if (GV.hasLocalLinkage()) 143 addAllGlobalValueUsers(GVtoClusterMap, &GV, &GV); 144 }; 145 146 llvm::for_each(M.functions(), recordGVSet); 147 llvm::for_each(M.globals(), recordGVSet); 148 llvm::for_each(M.aliases(), recordGVSet); 149 150 // Assigned all GVs to merged clusters while balancing number of objects in 151 // each. 152 auto CompareClusters = [](const std::pair<unsigned, unsigned> &a, 153 const std::pair<unsigned, unsigned> &b) { 154 if (a.second || b.second) 155 return a.second > b.second; 156 else 157 return a.first > b.first; 158 }; 159 160 std::priority_queue<std::pair<unsigned, unsigned>, 161 std::vector<std::pair<unsigned, unsigned>>, 162 decltype(CompareClusters)> 163 BalancinQueue(CompareClusters); 164 // Pre-populate priority queue with N slot blanks. 165 for (unsigned i = 0; i < N; ++i) 166 BalancinQueue.push(std::make_pair(i, 0)); 167 168 using SortType = std::pair<unsigned, ClusterMapType::iterator>; 169 170 SmallVector<SortType, 64> Sets; 171 SmallPtrSet<const GlobalValue *, 32> Visited; 172 173 // To guarantee determinism, we have to sort SCC according to size. 174 // When size is the same, use leader's name. 175 for (ClusterMapType::iterator I = GVtoClusterMap.begin(), 176 E = GVtoClusterMap.end(); I != E; ++I) 177 if (I->isLeader()) 178 Sets.push_back( 179 std::make_pair(std::distance(GVtoClusterMap.member_begin(I), 180 GVtoClusterMap.member_end()), I)); 181 182 llvm::sort(Sets, [](const SortType &a, const SortType &b) { 183 if (a.first == b.first) 184 return a.second->getData()->getName() > b.second->getData()->getName(); 185 else 186 return a.first > b.first; 187 }); 188 189 for (auto &I : Sets) { 190 unsigned CurrentClusterID = BalancinQueue.top().first; 191 unsigned CurrentClusterSize = BalancinQueue.top().second; 192 BalancinQueue.pop(); 193 194 LLVM_DEBUG(dbgs() << "Root[" << CurrentClusterID << "] cluster_size(" 195 << I.first << ") ----> " << I.second->getData()->getName() 196 << "\n"); 197 198 for (ClusterMapType::member_iterator MI = 199 GVtoClusterMap.findLeader(I.second); 200 MI != GVtoClusterMap.member_end(); ++MI) { 201 if (!Visited.insert(*MI).second) 202 continue; 203 LLVM_DEBUG(dbgs() << "----> " << (*MI)->getName() 204 << ((*MI)->hasLocalLinkage() ? " l " : " e ") << "\n"); 205 Visited.insert(*MI); 206 ClusterIDMap[*MI] = CurrentClusterID; 207 CurrentClusterSize++; 208 } 209 // Add this set size to the number of entries in this cluster. 210 BalancinQueue.push(std::make_pair(CurrentClusterID, CurrentClusterSize)); 211 } 212 } 213 214 static void externalize(GlobalValue *GV) { 215 if (GV->hasLocalLinkage()) { 216 GV->setLinkage(GlobalValue::ExternalLinkage); 217 GV->setVisibility(GlobalValue::HiddenVisibility); 218 } 219 220 // Unnamed entities must be named consistently between modules. setName will 221 // give a distinct name to each such entity. 222 if (!GV->hasName()) 223 GV->setName("__llvmsplit_unnamed"); 224 } 225 226 // Returns whether GV should be in partition (0-based) I of N. 227 static bool isInPartition(const GlobalValue *GV, unsigned I, unsigned N) { 228 if (auto *GIS = dyn_cast<GlobalIndirectSymbol>(GV)) 229 if (const GlobalObject *Base = GIS->getBaseObject()) 230 GV = Base; 231 232 StringRef Name; 233 if (const Comdat *C = GV->getComdat()) 234 Name = C->getName(); 235 else 236 Name = GV->getName(); 237 238 // Partition by MD5 hash. We only need a few bits for evenness as the number 239 // of partitions will generally be in the 1-2 figure range; the low 16 bits 240 // are enough. 241 MD5 H; 242 MD5::MD5Result R; 243 H.update(Name); 244 H.final(R); 245 return (R[0] | (R[1] << 8)) % N == I; 246 } 247 248 void llvm::SplitModule( 249 Module &M, unsigned N, 250 function_ref<void(std::unique_ptr<Module> MPart)> ModuleCallback, 251 bool PreserveLocals) { 252 if (!PreserveLocals) { 253 for (Function &F : M) 254 externalize(&F); 255 for (GlobalVariable &GV : M.globals()) 256 externalize(&GV); 257 for (GlobalAlias &GA : M.aliases()) 258 externalize(&GA); 259 for (GlobalIFunc &GIF : M.ifuncs()) 260 externalize(&GIF); 261 } 262 263 // This performs splitting without a need for externalization, which might not 264 // always be possible. 265 ClusterIDMapType ClusterIDMap; 266 findPartitions(M, ClusterIDMap, N); 267 268 // FIXME: We should be able to reuse M as the last partition instead of 269 // cloning it. Note that the callers at the moment expect the module to 270 // be preserved, so will need some adjustments as well. 271 for (unsigned I = 0; I < N; ++I) { 272 ValueToValueMapTy VMap; 273 std::unique_ptr<Module> MPart( 274 CloneModule(M, VMap, [&](const GlobalValue *GV) { 275 if (ClusterIDMap.count(GV)) 276 return (ClusterIDMap[GV] == I); 277 else 278 return isInPartition(GV, I, N); 279 })); 280 if (I != 0) 281 MPart->setModuleInlineAsm(""); 282 ModuleCallback(std::move(MPart)); 283 } 284 } 285