1 //===-- Internalize.cpp - Mark functions internal -------------------------===//
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 pass loops over all of the functions and variables in the input module.
11 // If the function or variable is not in the list of external names given to
12 // the pass it is marked as internal.
13 //
14 //===----------------------------------------------------------------------===//
15 
16 #define DEBUG_TYPE "internalize"
17 #include "llvm/Transforms/IPO.h"
18 #include "llvm/ADT/SmallPtrSet.h"
19 #include "llvm/ADT/Statistic.h"
20 #include "llvm/Analysis/CallGraph.h"
21 #include "llvm/IR/Module.h"
22 #include "llvm/Pass.h"
23 #include "llvm/Support/CommandLine.h"
24 #include "llvm/Support/Debug.h"
25 #include "llvm/Support/raw_ostream.h"
26 #include "llvm/Transforms/Utils/ModuleUtils.h"
27 #include <fstream>
28 #include <set>
29 using namespace llvm;
30 
31 STATISTIC(NumAliases  , "Number of aliases internalized");
32 STATISTIC(NumFunctions, "Number of functions internalized");
33 STATISTIC(NumGlobals  , "Number of global vars internalized");
34 
35 // APIFile - A file which contains a list of symbols that should not be marked
36 // external.
37 static cl::opt<std::string>
38 APIFile("internalize-public-api-file", cl::value_desc("filename"),
39         cl::desc("A file containing list of symbol names to preserve"));
40 
41 // APIList - A list of symbols that should not be marked internal.
42 static cl::list<std::string>
43 APIList("internalize-public-api-list", cl::value_desc("list"),
44         cl::desc("A list of symbol names to preserve"),
45         cl::CommaSeparated);
46 
47 static cl::list<std::string>
48 DSOList("internalize-dso-list", cl::value_desc("list"),
49         cl::desc("A list of symbol names need for a dso symbol table"),
50         cl::CommaSeparated);
51 
52 namespace {
53   class InternalizePass : public ModulePass {
54     std::set<std::string> ExternalNames;
55     std::set<std::string> DSONames;
56   public:
57     static char ID; // Pass identification, replacement for typeid
58     explicit InternalizePass();
59     explicit InternalizePass(ArrayRef<const char *> ExportList,
60                              ArrayRef<const char *> DSOList);
61     void LoadFile(const char *Filename);
62     virtual bool runOnModule(Module &M);
63 
64     virtual void getAnalysisUsage(AnalysisUsage &AU) const {
65       AU.setPreservesCFG();
66       AU.addPreserved<CallGraph>();
67     }
68   };
69 } // end anonymous namespace
70 
71 char InternalizePass::ID = 0;
72 INITIALIZE_PASS(InternalizePass, "internalize",
73                 "Internalize Global Symbols", false, false)
74 
75 InternalizePass::InternalizePass()
76   : ModulePass(ID) {
77   initializeInternalizePassPass(*PassRegistry::getPassRegistry());
78   if (!APIFile.empty())           // If a filename is specified, use it.
79     LoadFile(APIFile.c_str());
80   ExternalNames.insert(APIList.begin(), APIList.end());
81   DSONames.insert(DSOList.begin(), DSOList.end());
82 }
83 
84 InternalizePass::InternalizePass(ArrayRef<const char *> ExportList,
85                                  ArrayRef<const char *> DSOList)
86   : ModulePass(ID){
87   initializeInternalizePassPass(*PassRegistry::getPassRegistry());
88   for(ArrayRef<const char *>::const_iterator itr = ExportList.begin();
89         itr != ExportList.end(); itr++) {
90     ExternalNames.insert(*itr);
91   }
92   for(ArrayRef<const char *>::const_iterator itr = DSOList.begin();
93         itr != DSOList.end(); itr++) {
94     DSONames.insert(*itr);
95   }
96 }
97 
98 void InternalizePass::LoadFile(const char *Filename) {
99   // Load the APIFile...
100   std::ifstream In(Filename);
101   if (!In.good()) {
102     errs() << "WARNING: Internalize couldn't load file '" << Filename
103          << "'! Continuing as if it's empty.\n";
104     return; // Just continue as if the file were empty
105   }
106   while (In) {
107     std::string Symbol;
108     In >> Symbol;
109     if (!Symbol.empty())
110       ExternalNames.insert(Symbol);
111   }
112 }
113 
114 static bool shouldInternalize(const GlobalValue &GV,
115                               const std::set<std::string> &ExternalNames,
116                               const std::set<std::string> &DSONames) {
117   // Function must be defined here
118   if (GV.isDeclaration())
119     return false;
120 
121   // Available externally is really just a "declaration with a body".
122   if (GV.hasAvailableExternallyLinkage())
123     return false;
124 
125   // Already has internal linkage
126   if (GV.hasLocalLinkage())
127     return false;
128 
129   // Marked to keep external?
130   if (ExternalNames.count(GV.getName()))
131     return false;
132 
133   // Not needed for the symbol table?
134   if (!DSONames.count(GV.getName()))
135     return true;
136 
137   // Not a linkonce. Someone can depend on it being on the symbol table.
138   if (!GV.hasLinkOnceLinkage())
139     return false;
140 
141   // The address is not important, we can hide it.
142   if (GV.hasUnnamedAddr())
143     return true;
144 
145   // FIXME: Check if the address is used.
146   return false;
147 }
148 
149 bool InternalizePass::runOnModule(Module &M) {
150   CallGraph *CG = getAnalysisIfAvailable<CallGraph>();
151   CallGraphNode *ExternalNode = CG ? CG->getExternalCallingNode() : 0;
152   bool Changed = false;
153 
154   SmallPtrSet<GlobalValue *, 8> Used;
155   collectUsedGlobalVariables(M, Used, false);
156 
157   // We must assume that globals in llvm.used have a reference that not even
158   // the linker can see, so we don't internalize them.
159   // For llvm.compiler.used the situation is a bit fuzzy. The assembler and
160   // linker can drop those symbols. If this pass is running as part of LTO,
161   // one might think that it could just drop llvm.compiler.used. The problem
162   // is that even in LTO llvm doesn't see every reference. For example,
163   // we don't see references from function local inline assembly. To be
164   // conservative, we internalize symbols in llvm.compiler.used, but we
165   // keep llvm.compiler.used so that the symbol is not deleted by llvm.
166   for (SmallPtrSet<GlobalValue *, 8>::iterator I = Used.begin(), E = Used.end();
167        I != E; ++I) {
168     GlobalValue *V = *I;
169     ExternalNames.insert(V->getName());
170   }
171 
172   // Mark all functions not in the api as internal.
173   // FIXME: maybe use private linkage?
174   for (Module::iterator I = M.begin(), E = M.end(); I != E; ++I) {
175     if (!shouldInternalize(*I, ExternalNames, DSONames))
176       continue;
177 
178     I->setLinkage(GlobalValue::InternalLinkage);
179 
180     if (ExternalNode)
181       // Remove a callgraph edge from the external node to this function.
182       ExternalNode->removeOneAbstractEdgeTo((*CG)[I]);
183 
184     Changed = true;
185     ++NumFunctions;
186     DEBUG(dbgs() << "Internalizing func " << I->getName() << "\n");
187   }
188 
189   // Never internalize the llvm.used symbol.  It is used to implement
190   // attribute((used)).
191   // FIXME: Shouldn't this just filter on llvm.metadata section??
192   ExternalNames.insert("llvm.used");
193   ExternalNames.insert("llvm.compiler.used");
194 
195   // Never internalize anchors used by the machine module info, else the info
196   // won't find them.  (see MachineModuleInfo.)
197   ExternalNames.insert("llvm.global_ctors");
198   ExternalNames.insert("llvm.global_dtors");
199   ExternalNames.insert("llvm.global.annotations");
200 
201   // Never internalize symbols code-gen inserts.
202   // FIXME: We should probably add this (and the __stack_chk_guard) via some
203   // type of call-back in CodeGen.
204   ExternalNames.insert("__stack_chk_fail");
205   ExternalNames.insert("__stack_chk_guard");
206 
207   // Mark all global variables with initializers that are not in the api as
208   // internal as well.
209   // FIXME: maybe use private linkage?
210   for (Module::global_iterator I = M.global_begin(), E = M.global_end();
211        I != E; ++I) {
212     if (!shouldInternalize(*I, ExternalNames, DSONames))
213       continue;
214 
215     I->setLinkage(GlobalValue::InternalLinkage);
216     Changed = true;
217     ++NumGlobals;
218     DEBUG(dbgs() << "Internalized gvar " << I->getName() << "\n");
219   }
220 
221   // Mark all aliases that are not in the api as internal as well.
222   for (Module::alias_iterator I = M.alias_begin(), E = M.alias_end();
223        I != E; ++I) {
224     if (!shouldInternalize(*I, ExternalNames, DSONames))
225       continue;
226 
227     I->setLinkage(GlobalValue::InternalLinkage);
228     Changed = true;
229     ++NumAliases;
230     DEBUG(dbgs() << "Internalized alias " << I->getName() << "\n");
231   }
232 
233   return Changed;
234 }
235 
236 ModulePass *llvm::createInternalizePass() {
237   return new InternalizePass();
238 }
239 
240 ModulePass *llvm::createInternalizePass(ArrayRef<const char *> ExportList,
241                                         ArrayRef<const char *> DSOList) {
242   return new InternalizePass(ExportList, DSOList);
243 }
244