1 //===-- Internalize.cpp - Mark functions internal -------------------------===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 // This pass loops over all of the functions and variables in the input module.
10 // If the function or variable does not need to be preserved according to the
11 // client supplied callback, it is marked as internal.
12 //
13 // This transformation would not be legal in a regular compilation, but it gets
14 // extra information from the linker about what is safe.
15 //
16 // For example: Internalizing a function with external linkage. Only if we are
17 // told it is only used from within this module, it is safe to do it.
18 //
19 //===----------------------------------------------------------------------===//
20 
21 #include "llvm/Transforms/IPO/Internalize.h"
22 #include "llvm/ADT/SmallPtrSet.h"
23 #include "llvm/ADT/Statistic.h"
24 #include "llvm/ADT/StringSet.h"
25 #include "llvm/Analysis/CallGraph.h"
26 #include "llvm/IR/Module.h"
27 #include "llvm/InitializePasses.h"
28 #include "llvm/Pass.h"
29 #include "llvm/Support/CommandLine.h"
30 #include "llvm/Support/Debug.h"
31 #include "llvm/Support/LineIterator.h"
32 #include "llvm/Support/MemoryBuffer.h"
33 #include "llvm/Support/raw_ostream.h"
34 #include "llvm/Transforms/IPO.h"
35 #include "llvm/Transforms/Utils/GlobalStatus.h"
36 using namespace llvm;
37 
38 #define DEBUG_TYPE "internalize"
39 
40 STATISTIC(NumAliases, "Number of aliases internalized");
41 STATISTIC(NumFunctions, "Number of functions internalized");
42 STATISTIC(NumGlobals, "Number of global vars internalized");
43 
44 // APIFile - A file which contains a list of symbols that should not be marked
45 // external.
46 static cl::opt<std::string>
47     APIFile("internalize-public-api-file", cl::value_desc("filename"),
48             cl::desc("A file containing list of symbol names to preserve"));
49 
50 // APIList - A list of symbols that should not be marked internal.
51 static cl::list<std::string>
52     APIList("internalize-public-api-list", cl::value_desc("list"),
53             cl::desc("A list of symbol names to preserve"), cl::CommaSeparated);
54 
55 namespace {
56 // Helper to load an API list to preserve from file and expose it as a functor
57 // for internalization.
58 class PreserveAPIList {
59 public:
60   PreserveAPIList() {
61     if (!APIFile.empty())
62       LoadFile(APIFile);
63     ExternalNames.insert(APIList.begin(), APIList.end());
64   }
65 
66   bool operator()(const GlobalValue &GV) {
67     return ExternalNames.count(GV.getName());
68   }
69 
70 private:
71   // Contains the set of symbols loaded from file
72   StringSet<> ExternalNames;
73 
74   void LoadFile(StringRef Filename) {
75     // Load the APIFile...
76     ErrorOr<std::unique_ptr<MemoryBuffer>> Buf =
77         MemoryBuffer::getFile(Filename);
78     if (!Buf) {
79       errs() << "WARNING: Internalize couldn't load file '" << Filename
80              << "'! Continuing as if it's empty.\n";
81       return; // Just continue as if the file were empty
82     }
83     for (line_iterator I(*Buf->get(), true), E; I != E; ++I)
84       ExternalNames.insert(*I);
85   }
86 };
87 } // end anonymous namespace
88 
89 bool InternalizePass::shouldPreserveGV(const GlobalValue &GV) {
90   // Function must be defined here
91   if (GV.isDeclaration())
92     return true;
93 
94   // Available externally is really just a "declaration with a body".
95   if (GV.hasAvailableExternallyLinkage())
96     return true;
97 
98   // Assume that dllexported symbols are referenced elsewhere
99   if (GV.hasDLLExportStorageClass())
100     return true;
101 
102   // Already local, has nothing to do.
103   if (GV.hasLocalLinkage())
104     return false;
105 
106   // Check some special cases
107   if (AlwaysPreserved.count(GV.getName()))
108     return true;
109 
110   return MustPreserveGV(GV);
111 }
112 
113 bool InternalizePass::maybeInternalize(
114     GlobalValue &GV, const DenseSet<const Comdat *> &ExternalComdats) {
115   if (Comdat *C = GV.getComdat()) {
116     if (ExternalComdats.count(C))
117       return false;
118 
119     // If a comdat is not externally visible we can drop it.
120     if (auto GO = dyn_cast<GlobalObject>(&GV))
121       GO->setComdat(nullptr);
122 
123     if (GV.hasLocalLinkage())
124       return false;
125   } else {
126     if (GV.hasLocalLinkage())
127       return false;
128 
129     if (shouldPreserveGV(GV))
130       return false;
131   }
132 
133   GV.setVisibility(GlobalValue::DefaultVisibility);
134   GV.setLinkage(GlobalValue::InternalLinkage);
135   return true;
136 }
137 
138 // If GV is part of a comdat and is externally visible, keep track of its
139 // comdat so that we don't internalize any of its members.
140 void InternalizePass::checkComdatVisibility(
141     GlobalValue &GV, DenseSet<const Comdat *> &ExternalComdats) {
142   Comdat *C = GV.getComdat();
143   if (!C)
144     return;
145 
146   if (shouldPreserveGV(GV))
147     ExternalComdats.insert(C);
148 }
149 
150 bool InternalizePass::internalizeModule(Module &M, CallGraph *CG) {
151   bool Changed = false;
152   CallGraphNode *ExternalNode = CG ? CG->getExternalCallingNode() : nullptr;
153 
154   SmallPtrSet<GlobalValue *, 8> Used;
155   collectUsedGlobalVariables(M, Used, false);
156 
157   // Collect comdat visiblity information for the module.
158   DenseSet<const Comdat *> ExternalComdats;
159   if (!M.getComdatSymbolTable().empty()) {
160     for (Function &F : M)
161       checkComdatVisibility(F, ExternalComdats);
162     for (GlobalVariable &GV : M.globals())
163       checkComdatVisibility(GV, ExternalComdats);
164     for (GlobalAlias &GA : M.aliases())
165       checkComdatVisibility(GA, ExternalComdats);
166   }
167 
168   // We must assume that globals in llvm.used have a reference that not even
169   // the linker can see, so we don't internalize them.
170   // For llvm.compiler.used the situation is a bit fuzzy. The assembler and
171   // linker can drop those symbols. If this pass is running as part of LTO,
172   // one might think that it could just drop llvm.compiler.used. The problem
173   // is that even in LTO llvm doesn't see every reference. For example,
174   // we don't see references from function local inline assembly. To be
175   // conservative, we internalize symbols in llvm.compiler.used, but we
176   // keep llvm.compiler.used so that the symbol is not deleted by llvm.
177   for (GlobalValue *V : Used) {
178     AlwaysPreserved.insert(V->getName());
179   }
180 
181   // Mark all functions not in the api as internal.
182   for (Function &I : M) {
183     if (!maybeInternalize(I, ExternalComdats))
184       continue;
185     Changed = true;
186 
187     if (ExternalNode)
188       // Remove a callgraph edge from the external node to this function.
189       ExternalNode->removeOneAbstractEdgeTo((*CG)[&I]);
190 
191     ++NumFunctions;
192     LLVM_DEBUG(dbgs() << "Internalizing func " << I.getName() << "\n");
193   }
194 
195   // Never internalize the llvm.used symbol.  It is used to implement
196   // attribute((used)).
197   // FIXME: Shouldn't this just filter on llvm.metadata section??
198   AlwaysPreserved.insert("llvm.used");
199   AlwaysPreserved.insert("llvm.compiler.used");
200 
201   // Never internalize anchors used by the machine module info, else the info
202   // won't find them.  (see MachineModuleInfo.)
203   AlwaysPreserved.insert("llvm.global_ctors");
204   AlwaysPreserved.insert("llvm.global_dtors");
205   AlwaysPreserved.insert("llvm.global.annotations");
206 
207   // Never internalize symbols code-gen inserts.
208   // FIXME: We should probably add this (and the __stack_chk_guard) via some
209   // type of call-back in CodeGen.
210   AlwaysPreserved.insert("__stack_chk_fail");
211   AlwaysPreserved.insert("__stack_chk_guard");
212 
213   // Mark all global variables with initializers that are not in the api as
214   // internal as well.
215   for (auto &GV : M.globals()) {
216     if (!maybeInternalize(GV, ExternalComdats))
217       continue;
218     Changed = true;
219 
220     ++NumGlobals;
221     LLVM_DEBUG(dbgs() << "Internalized gvar " << GV.getName() << "\n");
222   }
223 
224   // Mark all aliases that are not in the api as internal as well.
225   for (auto &GA : M.aliases()) {
226     if (!maybeInternalize(GA, ExternalComdats))
227       continue;
228     Changed = true;
229 
230     ++NumAliases;
231     LLVM_DEBUG(dbgs() << "Internalized alias " << GA.getName() << "\n");
232   }
233 
234   return Changed;
235 }
236 
237 InternalizePass::InternalizePass() : MustPreserveGV(PreserveAPIList()) {}
238 
239 PreservedAnalyses InternalizePass::run(Module &M, ModuleAnalysisManager &AM) {
240   if (!internalizeModule(M, AM.getCachedResult<CallGraphAnalysis>(M)))
241     return PreservedAnalyses::all();
242 
243   PreservedAnalyses PA;
244   PA.preserve<CallGraphAnalysis>();
245   return PA;
246 }
247 
248 namespace {
249 class InternalizeLegacyPass : public ModulePass {
250   // Client supplied callback to control wheter a symbol must be preserved.
251   std::function<bool(const GlobalValue &)> MustPreserveGV;
252 
253 public:
254   static char ID; // Pass identification, replacement for typeid
255 
256   InternalizeLegacyPass() : ModulePass(ID), MustPreserveGV(PreserveAPIList()) {}
257 
258   InternalizeLegacyPass(std::function<bool(const GlobalValue &)> MustPreserveGV)
259       : ModulePass(ID), MustPreserveGV(std::move(MustPreserveGV)) {
260     initializeInternalizeLegacyPassPass(*PassRegistry::getPassRegistry());
261   }
262 
263   bool runOnModule(Module &M) override {
264     if (skipModule(M))
265       return false;
266 
267     CallGraphWrapperPass *CGPass =
268         getAnalysisIfAvailable<CallGraphWrapperPass>();
269     CallGraph *CG = CGPass ? &CGPass->getCallGraph() : nullptr;
270     return internalizeModule(M, MustPreserveGV, CG);
271   }
272 
273   void getAnalysisUsage(AnalysisUsage &AU) const override {
274     AU.setPreservesCFG();
275     AU.addPreserved<CallGraphWrapperPass>();
276   }
277 };
278 }
279 
280 char InternalizeLegacyPass::ID = 0;
281 INITIALIZE_PASS(InternalizeLegacyPass, "internalize",
282                 "Internalize Global Symbols", false, false)
283 
284 ModulePass *llvm::createInternalizePass() {
285   return new InternalizeLegacyPass();
286 }
287 
288 ModulePass *llvm::createInternalizePass(
289     std::function<bool(const GlobalValue &)> MustPreserveGV) {
290   return new InternalizeLegacyPass(std::move(MustPreserveGV));
291 }
292