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