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