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