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