1 //===- ExtractFunction.cpp - Extract a function from Program --------------===//
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 file implements several methods that are used to extract functions,
11 // loops, or portions of a module from the rest of the module.
12 //
13 //===----------------------------------------------------------------------===//
14 
15 #include "BugDriver.h"
16 #include "llvm/IR/Constants.h"
17 #include "llvm/IR/DataLayout.h"
18 #include "llvm/IR/DerivedTypes.h"
19 #include "llvm/IR/LLVMContext.h"
20 #include "llvm/IR/Module.h"
21 #include "llvm/IR/Verifier.h"
22 #include "llvm/Pass.h"
23 #include "llvm/PassManager.h"
24 #include "llvm/Support/CommandLine.h"
25 #include "llvm/Support/Debug.h"
26 #include "llvm/Support/FileUtilities.h"
27 #include "llvm/Support/Path.h"
28 #include "llvm/Support/Signals.h"
29 #include "llvm/Support/ToolOutputFile.h"
30 #include "llvm/Transforms/IPO.h"
31 #include "llvm/Transforms/Scalar.h"
32 #include "llvm/Transforms/Utils/Cloning.h"
33 #include "llvm/Transforms/Utils/CodeExtractor.h"
34 #include <set>
35 using namespace llvm;
36 using std::error_code;
37 
38 #define DEBUG_TYPE "bugpoint"
39 
40 namespace llvm {
41   bool DisableSimplifyCFG = false;
42   extern cl::opt<std::string> OutputPrefix;
43 } // End llvm namespace
44 
45 namespace {
46   cl::opt<bool>
47   NoDCE ("disable-dce",
48          cl::desc("Do not use the -dce pass to reduce testcases"));
49   cl::opt<bool, true>
50   NoSCFG("disable-simplifycfg", cl::location(DisableSimplifyCFG),
51          cl::desc("Do not use the -simplifycfg pass to reduce testcases"));
52 
53   Function* globalInitUsesExternalBA(GlobalVariable* GV) {
54     if (!GV->hasInitializer())
55       return nullptr;
56 
57     Constant *I = GV->getInitializer();
58 
59     // walk the values used by the initializer
60     // (and recurse into things like ConstantExpr)
61     std::vector<Constant*> Todo;
62     std::set<Constant*> Done;
63     Todo.push_back(I);
64 
65     while (!Todo.empty()) {
66       Constant* V = Todo.back();
67       Todo.pop_back();
68       Done.insert(V);
69 
70       if (BlockAddress *BA = dyn_cast<BlockAddress>(V)) {
71         Function *F = BA->getFunction();
72         if (F->isDeclaration())
73           return F;
74       }
75 
76       for (User::op_iterator i = V->op_begin(), e = V->op_end(); i != e; ++i) {
77         Constant *C = dyn_cast<Constant>(*i);
78         if (C && !isa<GlobalValue>(C) && !Done.count(C))
79           Todo.push_back(C);
80       }
81     }
82     return nullptr;
83   }
84 }  // end anonymous namespace
85 
86 /// deleteInstructionFromProgram - This method clones the current Program and
87 /// deletes the specified instruction from the cloned module.  It then runs a
88 /// series of cleanup passes (ADCE and SimplifyCFG) to eliminate any code which
89 /// depends on the value.  The modified module is then returned.
90 ///
91 Module *BugDriver::deleteInstructionFromProgram(const Instruction *I,
92                                                 unsigned Simplification) {
93   // FIXME, use vmap?
94   Module *Clone = CloneModule(Program);
95 
96   const BasicBlock *PBB = I->getParent();
97   const Function *PF = PBB->getParent();
98 
99   Module::iterator RFI = Clone->begin(); // Get iterator to corresponding fn
100   std::advance(RFI, std::distance(PF->getParent()->begin(),
101                                   Module::const_iterator(PF)));
102 
103   Function::iterator RBI = RFI->begin();  // Get iterator to corresponding BB
104   std::advance(RBI, std::distance(PF->begin(), Function::const_iterator(PBB)));
105 
106   BasicBlock::iterator RI = RBI->begin(); // Get iterator to corresponding inst
107   std::advance(RI, std::distance(PBB->begin(), BasicBlock::const_iterator(I)));
108   Instruction *TheInst = RI;              // Got the corresponding instruction!
109 
110   // If this instruction produces a value, replace any users with null values
111   if (!TheInst->getType()->isVoidTy())
112     TheInst->replaceAllUsesWith(Constant::getNullValue(TheInst->getType()));
113 
114   // Remove the instruction from the program.
115   TheInst->getParent()->getInstList().erase(TheInst);
116 
117   // Spiff up the output a little bit.
118   std::vector<std::string> Passes;
119 
120   /// Can we get rid of the -disable-* options?
121   if (Simplification > 1 && !NoDCE)
122     Passes.push_back("dce");
123   if (Simplification && !DisableSimplifyCFG)
124     Passes.push_back("simplifycfg");      // Delete dead control flow
125 
126   Passes.push_back("verify");
127   Module *New = runPassesOn(Clone, Passes);
128   delete Clone;
129   if (!New) {
130     errs() << "Instruction removal failed.  Sorry. :(  Please report a bug!\n";
131     exit(1);
132   }
133   return New;
134 }
135 
136 /// performFinalCleanups - This method clones the current Program and performs
137 /// a series of cleanups intended to get rid of extra cruft on the module
138 /// before handing it to the user.
139 ///
140 Module *BugDriver::performFinalCleanups(Module *M, bool MayModifySemantics) {
141   // Make all functions external, so GlobalDCE doesn't delete them...
142   for (Module::iterator I = M->begin(), E = M->end(); I != E; ++I)
143     I->setLinkage(GlobalValue::ExternalLinkage);
144 
145   std::vector<std::string> CleanupPasses;
146   CleanupPasses.push_back("globaldce");
147 
148   if (MayModifySemantics)
149     CleanupPasses.push_back("deadarghaX0r");
150   else
151     CleanupPasses.push_back("deadargelim");
152 
153   Module *New = runPassesOn(M, CleanupPasses);
154   if (!New) {
155     errs() << "Final cleanups failed.  Sorry. :(  Please report a bug!\n";
156     return M;
157   }
158   delete M;
159   return New;
160 }
161 
162 
163 /// ExtractLoop - Given a module, extract up to one loop from it into a new
164 /// function.  This returns null if there are no extractable loops in the
165 /// program or if the loop extractor crashes.
166 Module *BugDriver::ExtractLoop(Module *M) {
167   std::vector<std::string> LoopExtractPasses;
168   LoopExtractPasses.push_back("loop-extract-single");
169 
170   Module *NewM = runPassesOn(M, LoopExtractPasses);
171   if (!NewM) {
172     outs() << "*** Loop extraction failed: ";
173     EmitProgressBitcode(M, "loopextraction", true);
174     outs() << "*** Sorry. :(  Please report a bug!\n";
175     return nullptr;
176   }
177 
178   // Check to see if we created any new functions.  If not, no loops were
179   // extracted and we should return null.  Limit the number of loops we extract
180   // to avoid taking forever.
181   static unsigned NumExtracted = 32;
182   if (M->size() == NewM->size() || --NumExtracted == 0) {
183     delete NewM;
184     return nullptr;
185   } else {
186     assert(M->size() < NewM->size() && "Loop extract removed functions?");
187     Module::iterator MI = NewM->begin();
188     for (unsigned i = 0, e = M->size(); i != e; ++i)
189       ++MI;
190   }
191 
192   return NewM;
193 }
194 
195 
196 // DeleteFunctionBody - "Remove" the function by deleting all of its basic
197 // blocks, making it external.
198 //
199 void llvm::DeleteFunctionBody(Function *F) {
200   // delete the body of the function...
201   F->deleteBody();
202   assert(F->isDeclaration() && "This didn't make the function external!");
203 }
204 
205 /// GetTorInit - Given a list of entries for static ctors/dtors, return them
206 /// as a constant array.
207 static Constant *GetTorInit(std::vector<std::pair<Function*, int> > &TorList) {
208   assert(!TorList.empty() && "Don't create empty tor list!");
209   std::vector<Constant*> ArrayElts;
210   Type *Int32Ty = Type::getInt32Ty(TorList[0].first->getContext());
211 
212   StructType *STy =
213     StructType::get(Int32Ty, TorList[0].first->getType(), NULL);
214   for (unsigned i = 0, e = TorList.size(); i != e; ++i) {
215     Constant *Elts[] = {
216       ConstantInt::get(Int32Ty, TorList[i].second),
217       TorList[i].first
218     };
219     ArrayElts.push_back(ConstantStruct::get(STy, Elts));
220   }
221   return ConstantArray::get(ArrayType::get(ArrayElts[0]->getType(),
222                                            ArrayElts.size()),
223                             ArrayElts);
224 }
225 
226 /// SplitStaticCtorDtor - A module was recently split into two parts, M1/M2, and
227 /// M1 has all of the global variables.  If M2 contains any functions that are
228 /// static ctors/dtors, we need to add an llvm.global_[cd]tors global to M2, and
229 /// prune appropriate entries out of M1s list.
230 static void SplitStaticCtorDtor(const char *GlobalName, Module *M1, Module *M2,
231                                 ValueToValueMapTy &VMap) {
232   GlobalVariable *GV = M1->getNamedGlobal(GlobalName);
233   if (!GV || GV->isDeclaration() || GV->hasLocalLinkage() ||
234       !GV->use_empty()) return;
235 
236   std::vector<std::pair<Function*, int> > M1Tors, M2Tors;
237   ConstantArray *InitList = dyn_cast<ConstantArray>(GV->getInitializer());
238   if (!InitList) return;
239 
240   for (unsigned i = 0, e = InitList->getNumOperands(); i != e; ++i) {
241     if (ConstantStruct *CS = dyn_cast<ConstantStruct>(InitList->getOperand(i))){
242       if (CS->getNumOperands() != 2) return;  // Not array of 2-element structs.
243 
244       if (CS->getOperand(1)->isNullValue())
245         break;  // Found a null terminator, stop here.
246 
247       ConstantInt *CI = dyn_cast<ConstantInt>(CS->getOperand(0));
248       int Priority = CI ? CI->getSExtValue() : 0;
249 
250       Constant *FP = CS->getOperand(1);
251       if (ConstantExpr *CE = dyn_cast<ConstantExpr>(FP))
252         if (CE->isCast())
253           FP = CE->getOperand(0);
254       if (Function *F = dyn_cast<Function>(FP)) {
255         if (!F->isDeclaration())
256           M1Tors.push_back(std::make_pair(F, Priority));
257         else {
258           // Map to M2's version of the function.
259           F = cast<Function>(VMap[F]);
260           M2Tors.push_back(std::make_pair(F, Priority));
261         }
262       }
263     }
264   }
265 
266   GV->eraseFromParent();
267   if (!M1Tors.empty()) {
268     Constant *M1Init = GetTorInit(M1Tors);
269     new GlobalVariable(*M1, M1Init->getType(), false,
270                        GlobalValue::AppendingLinkage,
271                        M1Init, GlobalName);
272   }
273 
274   GV = M2->getNamedGlobal(GlobalName);
275   assert(GV && "Not a clone of M1?");
276   assert(GV->use_empty() && "llvm.ctors shouldn't have uses!");
277 
278   GV->eraseFromParent();
279   if (!M2Tors.empty()) {
280     Constant *M2Init = GetTorInit(M2Tors);
281     new GlobalVariable(*M2, M2Init->getType(), false,
282                        GlobalValue::AppendingLinkage,
283                        M2Init, GlobalName);
284   }
285 }
286 
287 
288 /// SplitFunctionsOutOfModule - Given a module and a list of functions in the
289 /// module, split the functions OUT of the specified module, and place them in
290 /// the new module.
291 Module *
292 llvm::SplitFunctionsOutOfModule(Module *M,
293                                 const std::vector<Function*> &F,
294                                 ValueToValueMapTy &VMap) {
295   // Make sure functions & globals are all external so that linkage
296   // between the two modules will work.
297   for (Module::iterator I = M->begin(), E = M->end(); I != E; ++I)
298     I->setLinkage(GlobalValue::ExternalLinkage);
299   for (Module::global_iterator I = M->global_begin(), E = M->global_end();
300        I != E; ++I) {
301     if (I->hasName() && I->getName()[0] == '\01')
302       I->setName(I->getName().substr(1));
303     I->setLinkage(GlobalValue::ExternalLinkage);
304   }
305 
306   ValueToValueMapTy NewVMap;
307   Module *New = CloneModule(M, NewVMap);
308 
309   // Remove the Test functions from the Safe module
310   std::set<Function *> TestFunctions;
311   for (unsigned i = 0, e = F.size(); i != e; ++i) {
312     Function *TNOF = cast<Function>(VMap[F[i]]);
313     DEBUG(errs() << "Removing function ");
314     DEBUG(TNOF->printAsOperand(errs(), false));
315     DEBUG(errs() << "\n");
316     TestFunctions.insert(cast<Function>(NewVMap[TNOF]));
317     DeleteFunctionBody(TNOF);       // Function is now external in this module!
318   }
319 
320 
321   // Remove the Safe functions from the Test module
322   for (Module::iterator I = New->begin(), E = New->end(); I != E; ++I)
323     if (!TestFunctions.count(I))
324       DeleteFunctionBody(I);
325 
326 
327   // Try to split the global initializers evenly
328   for (Module::global_iterator I = M->global_begin(), E = M->global_end();
329        I != E; ++I) {
330     GlobalVariable *GV = cast<GlobalVariable>(NewVMap[I]);
331     if (Function *TestFn = globalInitUsesExternalBA(I)) {
332       if (Function *SafeFn = globalInitUsesExternalBA(GV)) {
333         errs() << "*** Error: when reducing functions, encountered "
334                   "the global '";
335         GV->printAsOperand(errs(), false);
336         errs() << "' with an initializer that references blockaddresses "
337                   "from safe function '" << SafeFn->getName()
338                << "' and from test function '" << TestFn->getName() << "'.\n";
339         exit(1);
340       }
341       I->setInitializer(nullptr);  // Delete the initializer to make it external
342     } else {
343       // If we keep it in the safe module, then delete it in the test module
344       GV->setInitializer(nullptr);
345     }
346   }
347 
348   // Make sure that there is a global ctor/dtor array in both halves of the
349   // module if they both have static ctor/dtor functions.
350   SplitStaticCtorDtor("llvm.global_ctors", M, New, NewVMap);
351   SplitStaticCtorDtor("llvm.global_dtors", M, New, NewVMap);
352 
353   return New;
354 }
355 
356 //===----------------------------------------------------------------------===//
357 // Basic Block Extraction Code
358 //===----------------------------------------------------------------------===//
359 
360 /// ExtractMappedBlocksFromModule - Extract all but the specified basic blocks
361 /// into their own functions.  The only detail is that M is actually a module
362 /// cloned from the one the BBs are in, so some mapping needs to be performed.
363 /// If this operation fails for some reason (ie the implementation is buggy),
364 /// this function should return null, otherwise it returns a new Module.
365 Module *BugDriver::ExtractMappedBlocksFromModule(const
366                                                  std::vector<BasicBlock*> &BBs,
367                                                  Module *M) {
368   SmallString<128> Filename;
369   int FD;
370   error_code EC = sys::fs::createUniqueFile(
371       OutputPrefix + "-extractblocks%%%%%%%", FD, Filename);
372   if (EC) {
373     outs() << "*** Basic Block extraction failed!\n";
374     errs() << "Error creating temporary file: " << EC.message() << "\n";
375     EmitProgressBitcode(M, "basicblockextractfail", true);
376     return nullptr;
377   }
378   sys::RemoveFileOnSignal(Filename);
379 
380   tool_output_file BlocksToNotExtractFile(Filename.c_str(), FD);
381   for (std::vector<BasicBlock*>::const_iterator I = BBs.begin(), E = BBs.end();
382        I != E; ++I) {
383     BasicBlock *BB = *I;
384     // If the BB doesn't have a name, give it one so we have something to key
385     // off of.
386     if (!BB->hasName()) BB->setName("tmpbb");
387     BlocksToNotExtractFile.os() << BB->getParent()->getName() << " "
388                                 << BB->getName() << "\n";
389   }
390   BlocksToNotExtractFile.os().close();
391   if (BlocksToNotExtractFile.os().has_error()) {
392     errs() << "Error writing list of blocks to not extract\n";
393     EmitProgressBitcode(M, "basicblockextractfail", true);
394     BlocksToNotExtractFile.os().clear_error();
395     return nullptr;
396   }
397   BlocksToNotExtractFile.keep();
398 
399   std::string uniqueFN = "--extract-blocks-file=";
400   uniqueFN += Filename.str();
401   const char *ExtraArg = uniqueFN.c_str();
402 
403   std::vector<std::string> PI;
404   PI.push_back("extract-blocks");
405   Module *Ret = runPassesOn(M, PI, false, 1, &ExtraArg);
406 
407   sys::fs::remove(Filename.c_str());
408 
409   if (!Ret) {
410     outs() << "*** Basic Block extraction failed, please report a bug!\n";
411     EmitProgressBitcode(M, "basicblockextractfail", true);
412   }
413   return Ret;
414 }
415