lib/Transforms/IPO/ConstantMerge.cpp

   1 //===- ConstantMerge.cpp - Merge duplicate global constants ---------------===//
   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 interface to a pass that merges duplicate global
  10 // constants together into a single constant that is shared.  This is useful
  11 // because some passes (ie TraceValues) insert a lot of string constants into
  12 // the program, regardless of whether or not an existing string is available.
  13 //
  14 // Algorithm: ConstantMerge is designed to build up a map of available constants
  15 // and eliminate duplicates when it is initialized.
  16 //
  17 //===----------------------------------------------------------------------===//
  18
  19 #include "llvm/Transforms/IPO/ConstantMerge.h"
  20 #include "llvm/ADT/DenseMap.h"
  21 #include "llvm/ADT/SmallPtrSet.h"
  22 #include "llvm/ADT/SmallVector.h"
  23 #include "llvm/ADT/Statistic.h"
  24 #include "llvm/IR/Constants.h"
  25 #include "llvm/IR/DataLayout.h"
  26 #include "llvm/IR/DerivedTypes.h"
  27 #include "llvm/IR/GlobalValue.h"
  28 #include "llvm/IR/GlobalVariable.h"
  29 #include "llvm/IR/LLVMContext.h"
  30 #include "llvm/IR/Module.h"
  31 #include "llvm/Pass.h"
  32 #include "llvm/Support/Casting.h"
  33 #include "llvm/Transforms/IPO.h"
  34 #include <algorithm>
  35 #include <cassert>
  36 #include <utility>
  37
  38 using namespace llvm;
  39
  40 #define DEBUG_TYPE "constmerge"
  41
  42 STATISTIC(NumIdenticalMerged, "Number of identical global constants merged");
  43
  44 /// Find values that are marked as llvm.used.
  45 static void FindUsedValues(GlobalVariable *LLVMUsed,
  46                            SmallPtrSetImpl<const GlobalValue*> &UsedValues) {
  47   if (!LLVMUsed) return;
  48   ConstantArray *Inits = cast<ConstantArray>(LLVMUsed->getInitializer());
  49
  50   for (unsigned i = 0, e = Inits->getNumOperands(); i != e; ++i) {
  51     Value *Operand = Inits->getOperand(i)->stripPointerCastsNoFollowAliases();
  52     GlobalValue *GV = cast<GlobalValue>(Operand);
  53     UsedValues.insert(GV);
  54   }
  55 }
  56
  57 // True if A is better than B.
  58 static bool IsBetterCanonical(const GlobalVariable &A,
  59                               const GlobalVariable &B) {
  60   if (!A.hasLocalLinkage() && B.hasLocalLinkage())
  61     return true;
  62
  63   if (A.hasLocalLinkage() && !B.hasLocalLinkage())
  64     return false;
  65
  66   return A.hasGlobalUnnamedAddr();
  67 }
  68
  69 static bool hasMetadataOtherThanDebugLoc(const GlobalVariable *GV) {
  70   SmallVector<std::pair<unsigned, MDNode *>, 4> MDs;
  71   GV->getAllMetadata(MDs);
  72   for (const auto &V : MDs)
  73     if (V.first != LLVMContext::MD_dbg)
  74       return true;
  75   return false;
  76 }
  77
  78 static void copyDebugLocMetadata(const GlobalVariable *From,
  79                                  GlobalVariable *To) {
  80   SmallVector<DIGlobalVariableExpression *, 1> MDs;
  81   From->getDebugInfo(MDs);
  82   for (auto MD : MDs)
  83     To->addDebugInfo(MD);
  84 }
  85
  86 static unsigned getAlignment(GlobalVariable *GV) {
  87   unsigned Align = GV->getAlignment();
  88   if (Align)
  89     return Align;
  90   return GV->getParent()->getDataLayout().getPreferredAlignment(GV);
  91 }
  92
  93 static bool
  94 isUnmergeableGlobal(GlobalVariable *GV,
  95                     const SmallPtrSetImpl<const GlobalValue *> &UsedGlobals) {
  96   // Only process constants with initializers in the default address space.
  97   return !GV->isConstant() || !GV->hasDefinitiveInitializer() ||
  98          GV->getType()->getAddressSpace() != 0 || GV->hasSection() ||
  99          // Don't touch values marked with attribute(used).
 100          UsedGlobals.count(GV);
 101 }
 102
 103 enum class CanMerge { No, Yes };
 104 static CanMerge makeMergeable(GlobalVariable *Old, GlobalVariable *New) {
 105   if (!Old->hasGlobalUnnamedAddr() && !New->hasGlobalUnnamedAddr())
 106     return CanMerge::No;
 107   if (hasMetadataOtherThanDebugLoc(Old))
 108     return CanMerge::No;
 109   assert(!hasMetadataOtherThanDebugLoc(New));
 110   if (!Old->hasGlobalUnnamedAddr())
 111     New->setUnnamedAddr(GlobalValue::UnnamedAddr::None);
 112   return CanMerge::Yes;
 113 }
 114
 115 static void replace(Module &M, GlobalVariable *Old, GlobalVariable *New) {
 116   Constant *NewConstant = New;
 117
 118   LLVM_DEBUG(dbgs() << "Replacing global: @" << Old->getName() << " -> @"
 119                     << New->getName() << "\n");
 120
 121   // Bump the alignment if necessary.
 122   if (Old->getAlignment() || New->getAlignment())
 123     New->setAlignment(std::max(getAlignment(Old), getAlignment(New)));
 124
 125   copyDebugLocMetadata(Old, New);
 126   Old->replaceAllUsesWith(NewConstant);
 127
 128   // Delete the global value from the module.
 129   assert(Old->hasLocalLinkage() &&
 130          "Refusing to delete an externally visible global variable.");
 131   Old->eraseFromParent();
 132 }
 133
 134 static bool mergeConstants(Module &M) {
 135   // Find all the globals that are marked "used".  These cannot be merged.
 136   SmallPtrSet<const GlobalValue*, 8> UsedGlobals;
 137   FindUsedValues(M.getGlobalVariable("llvm.used"), UsedGlobals);
 138   FindUsedValues(M.getGlobalVariable("llvm.compiler.used"), UsedGlobals);
 139
 140   // Map unique constants to globals.
 141   DenseMap<Constant *, GlobalVariable *> CMap;
 142
 143   SmallVector<std::pair<GlobalVariable *, GlobalVariable *>, 32>
 144       SameContentReplacements;
 145
 146   size_t ChangesMade = 0;
 147   size_t OldChangesMade = 0;
 148
 149   // Iterate constant merging while we are still making progress.  Merging two
 150   // constants together may allow us to merge other constants together if the
 151   // second level constants have initializers which point to the globals that
 152   // were just merged.
 153   while (true) {
 154     // Find the canonical constants others will be merged with.
 155     for (Module::global_iterator GVI = M.global_begin(), E = M.global_end();
 156          GVI != E; ) {
 157       GlobalVariable *GV = &*GVI++;
 158
 159       // If this GV is dead, remove it.
 160       GV->removeDeadConstantUsers();
 161       if (GV->use_empty() && GV->hasLocalLinkage()) {
 162         GV->eraseFromParent();
 163         ++ChangesMade;
 164         continue;
 165       }
 166
 167       if (isUnmergeableGlobal(GV, UsedGlobals))
 168         continue;
 169
 170       // This transformation is legal for weak ODR globals in the sense it
 171       // doesn't change semantics, but we really don't want to perform it
 172       // anyway; it's likely to pessimize code generation, and some tools
 173       // (like the Darwin linker in cases involving CFString) don't expect it.
 174       if (GV->isWeakForLinker())
 175         continue;
 176
 177       // Don't touch globals with metadata other then !dbg.
 178       if (hasMetadataOtherThanDebugLoc(GV))
 179         continue;
 180
 181       Constant *Init = GV->getInitializer();
 182
 183       // Check to see if the initializer is already known.
 184       GlobalVariable *&Slot = CMap[Init];
 185
 186       // If this is the first constant we find or if the old one is local,
 187       // replace with the current one. If the current is externally visible
 188       // it cannot be replace, but can be the canonical constant we merge with.
 189       bool FirstConstantFound = !Slot;
 190       if (FirstConstantFound || IsBetterCanonical(*GV, *Slot)) {
 191         Slot = GV;
 192         LLVM_DEBUG(dbgs() << "Cmap[" << *Init << "] = " << GV->getName()
 193                           << (FirstConstantFound ? "\n" : " (updated)\n"));
 194       }
 195     }
 196
 197     // Identify all globals that can be merged together, filling in the
 198     // SameContentReplacements vector. We cannot do the replacement in this pass
 199     // because doing so may cause initializers of other globals to be rewritten,
 200     // invalidating the Constant* pointers in CMap.
 201     for (Module::global_iterator GVI = M.global_begin(), E = M.global_end();
 202          GVI != E; ) {
 203       GlobalVariable *GV = &*GVI++;
 204
 205       if (isUnmergeableGlobal(GV, UsedGlobals))
 206         continue;
 207
 208       // We can only replace constant with local linkage.
 209       if (!GV->hasLocalLinkage())
 210         continue;
 211
 212       Constant *Init = GV->getInitializer();
 213
 214       // Check to see if the initializer is already known.
 215       auto Found = CMap.find(Init);
 216       if (Found == CMap.end())
 217         continue;
 218
 219       GlobalVariable *Slot = Found->second;
 220       if (Slot == GV)
 221         continue;
 222
 223       if (makeMergeable(GV, Slot) == CanMerge::No)
 224         continue;
 225
 226       // Make all uses of the duplicate constant use the canonical version.
 227       LLVM_DEBUG(dbgs() << "Will replace: @" << GV->getName() << " -> @"
 228                         << Slot->getName() << "\n");
 229       SameContentReplacements.push_back(std::make_pair(GV, Slot));
 230     }
 231
 232     // Now that we have figured out which replacements must be made, do them all
 233     // now.  This avoid invalidating the pointers in CMap, which are unneeded
 234     // now.
 235     for (unsigned i = 0, e = SameContentReplacements.size(); i != e; ++i) {
 236       GlobalVariable *Old = SameContentReplacements[i].first;
 237       GlobalVariable *New = SameContentReplacements[i].second;
 238       replace(M, Old, New);
 239       ++ChangesMade;
 240       ++NumIdenticalMerged;
 241     }
 242
 243     if (ChangesMade == OldChangesMade)
 244       break;
 245     OldChangesMade = ChangesMade;
 246
 247     SameContentReplacements.clear();
 248     CMap.clear();
 249   }
 250
 251   return ChangesMade;
 252 }
 253
 254 PreservedAnalyses ConstantMergePass::run(Module &M, ModuleAnalysisManager &) {
 255   if (!mergeConstants(M))
 256     return PreservedAnalyses::all();
 257   return PreservedAnalyses::none();
 258 }
 259
 260 namespace {
 261
 262 struct ConstantMergeLegacyPass : public ModulePass {
 263   static char ID; // Pass identification, replacement for typeid
 264
 265   ConstantMergeLegacyPass() : ModulePass(ID) {
 266     initializeConstantMergeLegacyPassPass(*PassRegistry::getPassRegistry());
 267   }
 268
 269   // For this pass, process all of the globals in the module, eliminating
 270   // duplicate constants.
 271   bool runOnModule(Module &M) override {
 272     if (skipModule(M))
 273       return false;
 274     return mergeConstants(M);
 275   }
 276 };
 277
 278 } // end anonymous namespace
 279
 280 char ConstantMergeLegacyPass::ID = 0;
 281
 282 INITIALIZE_PASS(ConstantMergeLegacyPass, "constmerge",
 283                 "Merge Duplicate Global Constants", false, false)
 284
 285 ModulePass *llvm::createConstantMergePass() {
 286   return new ConstantMergeLegacyPass();
 287 }