tools/bugpoint/ExtractFunction.cpp

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