1 //===- ConstantMerge.cpp - Merge duplicate global constants ---------------===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 //
10 // This file defines the interface to a pass that merges duplicate global
11 // constants together into a single constant that is shared.  This is useful
12 // because some passes (ie TraceValues) insert a lot of string constants into
13 // the program, regardless of whether or not an existing string is available.
14 //
15 // Algorithm: ConstantMerge is designed to build up a map of available constants
16 // and eliminate duplicates when it is initialized.
17 //
18 //===----------------------------------------------------------------------===//
19 
20 #define DEBUG_TYPE "constmerge"
21 #include "llvm/Transforms/IPO.h"
22 #include "llvm/ADT/DenseMap.h"
23 #include "llvm/ADT/PointerIntPair.h"
24 #include "llvm/ADT/SmallPtrSet.h"
25 #include "llvm/ADT/Statistic.h"
26 #include "llvm/IR/Constants.h"
27 #include "llvm/IR/DataLayout.h"
28 #include "llvm/IR/DerivedTypes.h"
29 #include "llvm/IR/Module.h"
30 #include "llvm/IR/Operator.h"
31 #include "llvm/Pass.h"
32 using namespace llvm;
33 
34 STATISTIC(NumMerged, "Number of global constants merged");
35 
36 namespace {
37   struct ConstantMerge : public ModulePass {
38     static char ID; // Pass identification, replacement for typeid
39     ConstantMerge() : ModulePass(ID) {
40       initializeConstantMergePass(*PassRegistry::getPassRegistry());
41     }
42 
43     // For this pass, process all of the globals in the module, eliminating
44     // duplicate constants.
45     bool runOnModule(Module &M) override;
46 
47     // Return true iff we can determine the alignment of this global variable.
48     bool hasKnownAlignment(GlobalVariable *GV) const;
49 
50     // Return the alignment of the global, including converting the default
51     // alignment to a concrete value.
52     unsigned getAlignment(GlobalVariable *GV) const;
53 
54     const DataLayout *DL;
55   };
56 }
57 
58 char ConstantMerge::ID = 0;
59 INITIALIZE_PASS(ConstantMerge, "constmerge",
60                 "Merge Duplicate Global Constants", false, false)
61 
62 ModulePass *llvm::createConstantMergePass() { return new ConstantMerge(); }
63 
64 
65 
66 /// Find values that are marked as llvm.used.
67 static void FindUsedValues(GlobalVariable *LLVMUsed,
68                            SmallPtrSet<const GlobalValue*, 8> &UsedValues) {
69   if (LLVMUsed == 0) return;
70   ConstantArray *Inits = cast<ConstantArray>(LLVMUsed->getInitializer());
71 
72   for (unsigned i = 0, e = Inits->getNumOperands(); i != e; ++i) {
73     Value *Operand = Inits->getOperand(i)->stripPointerCastsNoFollowAliases();
74     GlobalValue *GV = cast<GlobalValue>(Operand);
75     UsedValues.insert(GV);
76   }
77 }
78 
79 // True if A is better than B.
80 static bool IsBetterCanonical(const GlobalVariable &A,
81                               const GlobalVariable &B) {
82   if (!A.hasLocalLinkage() && B.hasLocalLinkage())
83     return true;
84 
85   if (A.hasLocalLinkage() && !B.hasLocalLinkage())
86     return false;
87 
88   return A.hasUnnamedAddr();
89 }
90 
91 bool ConstantMerge::hasKnownAlignment(GlobalVariable *GV) const {
92   return DL || GV->getAlignment() != 0;
93 }
94 
95 unsigned ConstantMerge::getAlignment(GlobalVariable *GV) const {
96   unsigned Align = GV->getAlignment();
97   if (Align)
98     return Align;
99   if (DL)
100     return DL->getPreferredAlignment(GV);
101   return 0;
102 }
103 
104 bool ConstantMerge::runOnModule(Module &M) {
105   DataLayoutPass *DLP = getAnalysisIfAvailable<DataLayoutPass>();
106   DL = DLP ? &DLP->getDataLayout() : 0;
107 
108   // Find all the globals that are marked "used".  These cannot be merged.
109   SmallPtrSet<const GlobalValue*, 8> UsedGlobals;
110   FindUsedValues(M.getGlobalVariable("llvm.used"), UsedGlobals);
111   FindUsedValues(M.getGlobalVariable("llvm.compiler.used"), UsedGlobals);
112 
113   // Map unique <constants, has-unknown-alignment> pairs to globals.  We don't
114   // want to merge globals of unknown alignment with those of explicit
115   // alignment.  If we have DataLayout, we always know the alignment.
116   DenseMap<PointerIntPair<Constant*, 1, bool>, GlobalVariable*> CMap;
117 
118   // Replacements - This vector contains a list of replacements to perform.
119   SmallVector<std::pair<GlobalVariable*, GlobalVariable*>, 32> Replacements;
120 
121   bool MadeChange = false;
122 
123   // Iterate constant merging while we are still making progress.  Merging two
124   // constants together may allow us to merge other constants together if the
125   // second level constants have initializers which point to the globals that
126   // were just merged.
127   while (1) {
128 
129     // First: Find the canonical constants others will be merged with.
130     for (Module::global_iterator GVI = M.global_begin(), E = M.global_end();
131          GVI != E; ) {
132       GlobalVariable *GV = GVI++;
133 
134       // If this GV is dead, remove it.
135       GV->removeDeadConstantUsers();
136       if (GV->use_empty() && GV->hasLocalLinkage()) {
137         GV->eraseFromParent();
138         continue;
139       }
140 
141       // Only process constants with initializers in the default address space.
142       if (!GV->isConstant() || !GV->hasDefinitiveInitializer() ||
143           GV->getType()->getAddressSpace() != 0 || GV->hasSection() ||
144           // Don't touch values marked with attribute(used).
145           UsedGlobals.count(GV))
146         continue;
147 
148       // This transformation is legal for weak ODR globals in the sense it
149       // doesn't change semantics, but we really don't want to perform it
150       // anyway; it's likely to pessimize code generation, and some tools
151       // (like the Darwin linker in cases involving CFString) don't expect it.
152       if (GV->isWeakForLinker())
153         continue;
154 
155       Constant *Init = GV->getInitializer();
156 
157       // Check to see if the initializer is already known.
158       PointerIntPair<Constant*, 1, bool> Pair(Init, hasKnownAlignment(GV));
159       GlobalVariable *&Slot = CMap[Pair];
160 
161       // If this is the first constant we find or if the old one is local,
162       // replace with the current one. If the current is externally visible
163       // it cannot be replace, but can be the canonical constant we merge with.
164       if (Slot == 0 || IsBetterCanonical(*GV, *Slot))
165         Slot = GV;
166     }
167 
168     // Second: identify all globals that can be merged together, filling in
169     // the Replacements vector.  We cannot do the replacement in this pass
170     // because doing so may cause initializers of other globals to be rewritten,
171     // invalidating the Constant* pointers in CMap.
172     for (Module::global_iterator GVI = M.global_begin(), E = M.global_end();
173          GVI != E; ) {
174       GlobalVariable *GV = GVI++;
175 
176       // Only process constants with initializers in the default address space.
177       if (!GV->isConstant() || !GV->hasDefinitiveInitializer() ||
178           GV->getType()->getAddressSpace() != 0 || GV->hasSection() ||
179           // Don't touch values marked with attribute(used).
180           UsedGlobals.count(GV))
181         continue;
182 
183       // We can only replace constant with local linkage.
184       if (!GV->hasLocalLinkage())
185         continue;
186 
187       Constant *Init = GV->getInitializer();
188 
189       // Check to see if the initializer is already known.
190       PointerIntPair<Constant*, 1, bool> Pair(Init, hasKnownAlignment(GV));
191       GlobalVariable *Slot = CMap[Pair];
192 
193       if (!Slot || Slot == GV)
194         continue;
195 
196       if (!Slot->hasUnnamedAddr() && !GV->hasUnnamedAddr())
197         continue;
198 
199       if (!GV->hasUnnamedAddr())
200         Slot->setUnnamedAddr(false);
201 
202       // Make all uses of the duplicate constant use the canonical version.
203       Replacements.push_back(std::make_pair(GV, Slot));
204     }
205 
206     if (Replacements.empty())
207       return MadeChange;
208     CMap.clear();
209 
210     // Now that we have figured out which replacements must be made, do them all
211     // now.  This avoid invalidating the pointers in CMap, which are unneeded
212     // now.
213     for (unsigned i = 0, e = Replacements.size(); i != e; ++i) {
214       // Bump the alignment if necessary.
215       if (Replacements[i].first->getAlignment() ||
216           Replacements[i].second->getAlignment()) {
217         Replacements[i].second->setAlignment(
218             std::max(getAlignment(Replacements[i].first),
219                      getAlignment(Replacements[i].second)));
220       }
221 
222       // Eliminate any uses of the dead global.
223       Replacements[i].first->replaceAllUsesWith(Replacements[i].second);
224 
225       // Delete the global value from the module.
226       assert(Replacements[i].first->hasLocalLinkage() &&
227              "Refusing to delete an externally visible global variable.");
228       Replacements[i].first->eraseFromParent();
229     }
230 
231     NumMerged += Replacements.size();
232     Replacements.clear();
233   }
234 }
235