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