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 #include "llvm/Transforms/IPO.h"
23 #include "llvm/ADT/SmallPtrSet.h"
24 #include "llvm/ADT/Statistic.h"
25 #include "llvm/Analysis/CallGraph.h"
26 #include "llvm/IR/Module.h"
27 #include "llvm/Pass.h"
28 #include "llvm/Support/CommandLine.h"
29 #include "llvm/Support/Debug.h"
30 #include "llvm/Support/raw_ostream.h"
31 #include "llvm/Transforms/Utils/GlobalStatus.h"
32 #include "llvm/Transforms/Utils/ModuleUtils.h"
33 #include <fstream>
34 #include <set>
35 using namespace llvm;
36
37 #define DEBUG_TYPE "internalize"
38
39 STATISTIC(NumAliases , "Number of aliases internalized");
40 STATISTIC(NumFunctions, "Number of functions internalized");
41 STATISTIC(NumGlobals , "Number of global vars internalized");
42
43 // APIFile - A file which contains a list of symbols that should not be marked
44 // external.
45 static cl::opt<std::string>
46 APIFile("internalize-public-api-file", cl::value_desc("filename"),
47 cl::desc("A file containing list of symbol names to preserve"));
48
49 // APIList - A list of symbols that should not be marked internal.
50 static cl::list<std::string>
51 APIList("internalize-public-api-list", cl::value_desc("list"),
52 cl::desc("A list of symbol names to preserve"),
53 cl::CommaSeparated);
54
55 namespace {
56 class InternalizePass : public ModulePass {
57 std::set<std::string> ExternalNames;
58 public:
59 static char ID; // Pass identification, replacement for typeid
60 explicit InternalizePass();
61 explicit InternalizePass(ArrayRef<const char *> ExportList);
62 void LoadFile(const char *Filename);
63 bool runOnModule(Module &M) override;
64
getAnalysisUsage(AnalysisUsage & AU) const65 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
InternalizePass()76 InternalizePass::InternalizePass() : 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 }
82
InternalizePass(ArrayRef<const char * > ExportList)83 InternalizePass::InternalizePass(ArrayRef<const char *> ExportList)
84 : ModulePass(ID) {
85 initializeInternalizePassPass(*PassRegistry::getPassRegistry());
86 for(ArrayRef<const char *>::const_iterator itr = ExportList.begin();
87 itr != ExportList.end(); itr++) {
88 ExternalNames.insert(*itr);
89 }
90 }
91
LoadFile(const char * Filename)92 void InternalizePass::LoadFile(const char *Filename) {
93 // Load the APIFile...
94 std::ifstream In(Filename);
95 if (!In.good()) {
96 errs() << "WARNING: Internalize couldn't load file '" << Filename
97 << "'! Continuing as if it's empty.\n";
98 return; // Just continue as if the file were empty
99 }
100 while (In) {
101 std::string Symbol;
102 In >> Symbol;
103 if (!Symbol.empty())
104 ExternalNames.insert(Symbol);
105 }
106 }
107
shouldInternalize(const GlobalValue & GV,const std::set<std::string> & ExternalNames)108 static bool shouldInternalize(const GlobalValue &GV,
109 const std::set<std::string> &ExternalNames) {
110 // Function must be defined here
111 if (GV.isDeclaration())
112 return false;
113
114 // Available externally is really just a "declaration with a body".
115 if (GV.hasAvailableExternallyLinkage())
116 return false;
117
118 // Assume that dllexported symbols are referenced elsewhere
119 if (GV.hasDLLExportStorageClass())
120 return false;
121
122 // Already has internal linkage
123 if (GV.hasLocalLinkage())
124 return false;
125
126 // Marked to keep external?
127 if (ExternalNames.count(GV.getName()))
128 return false;
129
130 return true;
131 }
132
runOnModule(Module & M)133 bool InternalizePass::runOnModule(Module &M) {
134 CallGraphWrapperPass *CGPass = getAnalysisIfAvailable<CallGraphWrapperPass>();
135 CallGraph *CG = CGPass ? &CGPass->getCallGraph() : nullptr;
136 CallGraphNode *ExternalNode = CG ? CG->getExternalCallingNode() : nullptr;
137 bool Changed = false;
138
139 SmallPtrSet<GlobalValue *, 8> Used;
140 collectUsedGlobalVariables(M, Used, false);
141
142 // We must assume that globals in llvm.used have a reference that not even
143 // the linker can see, so we don't internalize them.
144 // For llvm.compiler.used the situation is a bit fuzzy. The assembler and
145 // linker can drop those symbols. If this pass is running as part of LTO,
146 // one might think that it could just drop llvm.compiler.used. The problem
147 // is that even in LTO llvm doesn't see every reference. For example,
148 // we don't see references from function local inline assembly. To be
149 // conservative, we internalize symbols in llvm.compiler.used, but we
150 // keep llvm.compiler.used so that the symbol is not deleted by llvm.
151 for (GlobalValue *V : Used) {
152 ExternalNames.insert(V->getName());
153 }
154
155 // Mark all functions not in the api as internal.
156 for (Module::iterator I = M.begin(), E = M.end(); I != E; ++I) {
157 if (!shouldInternalize(*I, ExternalNames))
158 continue;
159
160 I->setVisibility(GlobalValue::DefaultVisibility);
161 I->setLinkage(GlobalValue::InternalLinkage);
162
163 if (ExternalNode)
164 // Remove a callgraph edge from the external node to this function.
165 ExternalNode->removeOneAbstractEdgeTo((*CG)[I]);
166
167 Changed = true;
168 ++NumFunctions;
169 DEBUG(dbgs() << "Internalizing func " << I->getName() << "\n");
170 }
171
172 // Never internalize the llvm.used symbol. It is used to implement
173 // attribute((used)).
174 // FIXME: Shouldn't this just filter on llvm.metadata section??
175 ExternalNames.insert("llvm.used");
176 ExternalNames.insert("llvm.compiler.used");
177
178 // Never internalize anchors used by the machine module info, else the info
179 // won't find them. (see MachineModuleInfo.)
180 ExternalNames.insert("llvm.global_ctors");
181 ExternalNames.insert("llvm.global_dtors");
182 ExternalNames.insert("llvm.global.annotations");
183
184 // Never internalize symbols code-gen inserts.
185 // FIXME: We should probably add this (and the __stack_chk_guard) via some
186 // type of call-back in CodeGen.
187 ExternalNames.insert("__stack_chk_fail");
188 ExternalNames.insert("__stack_chk_guard");
189
190 // Mark all global variables with initializers that are not in the api as
191 // internal as well.
192 for (Module::global_iterator I = M.global_begin(), E = M.global_end();
193 I != E; ++I) {
194 if (!shouldInternalize(*I, ExternalNames))
195 continue;
196
197 I->setVisibility(GlobalValue::DefaultVisibility);
198 I->setLinkage(GlobalValue::InternalLinkage);
199 Changed = true;
200 ++NumGlobals;
201 DEBUG(dbgs() << "Internalized gvar " << I->getName() << "\n");
202 }
203
204 // Mark all aliases that are not in the api as internal as well.
205 for (Module::alias_iterator I = M.alias_begin(), E = M.alias_end();
206 I != E; ++I) {
207 if (!shouldInternalize(*I, ExternalNames))
208 continue;
209
210 I->setVisibility(GlobalValue::DefaultVisibility);
211 I->setLinkage(GlobalValue::InternalLinkage);
212 Changed = true;
213 ++NumAliases;
214 DEBUG(dbgs() << "Internalized alias " << I->getName() << "\n");
215 }
216
217 return Changed;
218 }
219
createInternalizePass()220 ModulePass *llvm::createInternalizePass() { return new InternalizePass(); }
221
createInternalizePass(ArrayRef<const char * > ExportList)222 ModulePass *llvm::createInternalizePass(ArrayRef<const char *> ExportList) {
223 return new InternalizePass(ExportList);
224 }
225