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