ExtractFunction.cpp revision 9e9a0d5fc26878e51a58a8b57900fcbf952c2691 
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/System/Path.h"
33#include "llvm/System/Signals.h"
34#include <set>
35using namespace llvm;
36
37namespace llvm {
38  bool DisableSimplifyCFG = false;
39} // End llvm namespace
40
41namespace {
42  cl::opt<bool>
43  NoDCE ("disable-dce",
44         cl::desc("Do not use the -dce pass to reduce testcases"));
45  cl::opt<bool, true>
46  NoSCFG("disable-simplifycfg", cl::location(DisableSimplifyCFG),
47         cl::desc("Do not use the -simplifycfg pass to reduce testcases"));
48}
49
50/// deleteInstructionFromProgram - This method clones the current Program and
51/// deletes the specified instruction from the cloned module.  It then runs a
52/// series of cleanup passes (ADCE and SimplifyCFG) to eliminate any code which
53/// depends on the value.  The modified module is then returned.
54///
55Module *BugDriver::deleteInstructionFromProgram(const Instruction *I,
56                                                unsigned Simplification) const {
57  Module *Result = CloneModule(Program);
58
59  const BasicBlock *PBB = I->getParent();
60  const Function *PF = PBB->getParent();
61
62  Module::iterator RFI = Result->begin(); // Get iterator to corresponding fn
63  std::advance(RFI, std::distance(PF->getParent()->begin(),
64                                  Module::const_iterator(PF)));
65
66  Function::iterator RBI = RFI->begin();  // Get iterator to corresponding BB
67  std::advance(RBI, std::distance(PF->begin(), Function::const_iterator(PBB)));
68
69  BasicBlock::iterator RI = RBI->begin(); // Get iterator to corresponding inst
70  std::advance(RI, std::distance(PBB->begin(), BasicBlock::const_iterator(I)));
71  Instruction *TheInst = RI;              // Got the corresponding instruction!
72
73  // If this instruction produces a value, replace any users with null values
74  if (isa<StructType>(TheInst->getType()))
75    TheInst->replaceAllUsesWith(UndefValue::get(TheInst->getType()));
76  else if (TheInst->getType() != Type::VoidTy)
77    TheInst->replaceAllUsesWith(Context.getNullValue(TheInst->getType()));
78
79  // Remove the instruction from the program.
80  TheInst->getParent()->getInstList().erase(TheInst);
81
82
83  //writeProgramToFile("current.bc", Result);
84
85  // Spiff up the output a little bit.
86  PassManager Passes;
87  // Make sure that the appropriate target data is always used...
88  Passes.add(new TargetData(Result));
89
90  /// FIXME: If this used runPasses() like the methods below, we could get rid
91  /// of the -disable-* options!
92  if (Simplification > 1 && !NoDCE)
93    Passes.add(createDeadCodeEliminationPass());
94  if (Simplification && !DisableSimplifyCFG)
95    Passes.add(createCFGSimplificationPass());      // Delete dead control flow
96
97  Passes.add(createVerifierPass());
98  Passes.run(*Result);
99  return Result;
100}
101
102static const PassInfo *getPI(Pass *P) {
103  const PassInfo *PI = P->getPassInfo();
104  delete P;
105  return PI;
106}
107
108/// performFinalCleanups - This method clones the current Program and performs
109/// a series of cleanups intended to get rid of extra cruft on the module
110/// before handing it to the user.
111///
112Module *BugDriver::performFinalCleanups(Module *M, bool MayModifySemantics) {
113  // Make all functions external, so GlobalDCE doesn't delete them...
114  for (Module::iterator I = M->begin(), E = M->end(); I != E; ++I)
115    I->setLinkage(GlobalValue::ExternalLinkage);
116
117  std::vector<const PassInfo*> CleanupPasses;
118  CleanupPasses.push_back(getPI(createGlobalDCEPass()));
119  CleanupPasses.push_back(getPI(createDeadTypeEliminationPass()));
120
121  if (MayModifySemantics)
122    CleanupPasses.push_back(getPI(createDeadArgHackingPass()));
123  else
124    CleanupPasses.push_back(getPI(createDeadArgEliminationPass()));
125
126  Module *New = runPassesOn(M, CleanupPasses);
127  if (New == 0) {
128    errs() << "Final cleanups failed.  Sorry. :(  Please report a bug!\n";
129    return M;
130  }
131  delete M;
132  return New;
133}
134
135
136/// ExtractLoop - Given a module, extract up to one loop from it into a new
137/// function.  This returns null if there are no extractable loops in the
138/// program or if the loop extractor crashes.
139Module *BugDriver::ExtractLoop(Module *M) {
140  std::vector<const PassInfo*> LoopExtractPasses;
141  LoopExtractPasses.push_back(getPI(createSingleLoopExtractorPass()));
142
143  Module *NewM = runPassesOn(M, LoopExtractPasses);
144  if (NewM == 0) {
145    Module *Old = swapProgramIn(M);
146    outs() << "*** Loop extraction failed: ";
147    EmitProgressBitcode("loopextraction", true);
148    outs() << "*** Sorry. :(  Please report a bug!\n";
149    swapProgramIn(Old);
150    return 0;
151  }
152
153  // Check to see if we created any new functions.  If not, no loops were
154  // extracted and we should return null.  Limit the number of loops we extract
155  // to avoid taking forever.
156  static unsigned NumExtracted = 32;
157  if (M->size() == NewM->size() || --NumExtracted == 0) {
158    delete NewM;
159    return 0;
160  } else {
161    assert(M->size() < NewM->size() && "Loop extract removed functions?");
162    Module::iterator MI = NewM->begin();
163    for (unsigned i = 0, e = M->size(); i != e; ++i)
164      ++MI;
165  }
166
167  return NewM;
168}
169
170
171// DeleteFunctionBody - "Remove" the function by deleting all of its basic
172// blocks, making it external.
173//
174void llvm::DeleteFunctionBody(Function *F) {
175  // delete the body of the function...
176  F->deleteBody();
177  assert(F->isDeclaration() && "This didn't make the function external!");
178}
179
180/// GetTorInit - Given a list of entries for static ctors/dtors, return them
181/// as a constant array.
182static Constant *GetTorInit(std::vector<std::pair<Function*, int> > &TorList) {
183  assert(!TorList.empty() && "Don't create empty tor list!");
184  std::vector<Constant*> ArrayElts;
185  for (unsigned i = 0, e = TorList.size(); i != e; ++i) {
186    std::vector<Constant*> Elts;
187    Elts.push_back(ConstantInt::get(Type::Int32Ty, TorList[i].second));
188    Elts.push_back(TorList[i].first);
189    ArrayElts.push_back(ConstantStruct::get(Elts));
190  }
191  return ConstantArray::get(ArrayType::get(ArrayElts[0]->getType(),
192                                           ArrayElts.size()),
193                            ArrayElts);
194}
195
196/// SplitStaticCtorDtor - A module was recently split into two parts, M1/M2, and
197/// M1 has all of the global variables.  If M2 contains any functions that are
198/// static ctors/dtors, we need to add an llvm.global_[cd]tors global to M2, and
199/// prune appropriate entries out of M1s list.
200static void SplitStaticCtorDtor(const char *GlobalName, Module *M1, Module *M2,
201                                DenseMap<const Value*, Value*> ValueMap) {
202  GlobalVariable *GV = M1->getNamedGlobal(GlobalName);
203  if (!GV || GV->isDeclaration() || GV->hasLocalLinkage() ||
204      !GV->use_empty()) return;
205
206  std::vector<std::pair<Function*, int> > M1Tors, M2Tors;
207  ConstantArray *InitList = dyn_cast<ConstantArray>(GV->getInitializer());
208  if (!InitList) return;
209
210  for (unsigned i = 0, e = InitList->getNumOperands(); i != e; ++i) {
211    if (ConstantStruct *CS = dyn_cast<ConstantStruct>(InitList->getOperand(i))){
212      if (CS->getNumOperands() != 2) return;  // Not array of 2-element structs.
213
214      if (CS->getOperand(1)->isNullValue())
215        break;  // Found a null terminator, stop here.
216
217      ConstantInt *CI = dyn_cast<ConstantInt>(CS->getOperand(0));
218      int Priority = CI ? CI->getSExtValue() : 0;
219
220      Constant *FP = CS->getOperand(1);
221      if (ConstantExpr *CE = dyn_cast<ConstantExpr>(FP))
222        if (CE->isCast())
223          FP = CE->getOperand(0);
224      if (Function *F = dyn_cast<Function>(FP)) {
225        if (!F->isDeclaration())
226          M1Tors.push_back(std::make_pair(F, Priority));
227        else {
228          // Map to M2's version of the function.
229          F = cast<Function>(ValueMap[F]);
230          M2Tors.push_back(std::make_pair(F, Priority));
231        }
232      }
233    }
234  }
235
236  GV->eraseFromParent();
237  if (!M1Tors.empty()) {
238    Constant *M1Init = GetTorInit(M1Tors);
239    new GlobalVariable(*M1, M1Init->getType(), false,
240                       GlobalValue::AppendingLinkage,
241                       M1Init, GlobalName);
242  }
243
244  GV = M2->getNamedGlobal(GlobalName);
245  assert(GV && "Not a clone of M1?");
246  assert(GV->use_empty() && "llvm.ctors shouldn't have uses!");
247
248  GV->eraseFromParent();
249  if (!M2Tors.empty()) {
250    Constant *M2Init = GetTorInit(M2Tors);
251    new GlobalVariable(*M2, M2Init->getType(), false,
252                       GlobalValue::AppendingLinkage,
253                       M2Init, GlobalName);
254  }
255}
256
257
258/// SplitFunctionsOutOfModule - Given a module and a list of functions in the
259/// module, split the functions OUT of the specified module, and place them in
260/// the new module.
261Module *
262llvm::SplitFunctionsOutOfModule(Module *M,
263                                const std::vector<Function*> &F,
264                                DenseMap<const Value*, Value*> &ValueMap) {
265  // Make sure functions & globals are all external so that linkage
266  // between the two modules will work.
267  for (Module::iterator I = M->begin(), E = M->end(); I != E; ++I)
268    I->setLinkage(GlobalValue::ExternalLinkage);
269  for (Module::global_iterator I = M->global_begin(), E = M->global_end();
270       I != E; ++I) {
271    if (I->hasName() && I->getName()[0] == '\01')
272      I->setName(I->getName().substr(1));
273    I->setLinkage(GlobalValue::ExternalLinkage);
274  }
275
276  DenseMap<const Value*, Value*> NewValueMap;
277  Module *New = CloneModule(M, NewValueMap);
278
279  // Make sure global initializers exist only in the safe module (CBE->.so)
280  for (Module::global_iterator I = New->global_begin(), E = New->global_end();
281       I != E; ++I)
282    I->setInitializer(0);  // Delete the initializer to make it external
283
284  // Remove the Test functions from the Safe module
285  std::set<Function *> TestFunctions;
286  for (unsigned i = 0, e = F.size(); i != e; ++i) {
287    Function *TNOF = cast<Function>(ValueMap[F[i]]);
288    DEBUG(errs() << "Removing function ");
289    DEBUG(WriteAsOperand(errs(), TNOF, false));
290    DEBUG(errs() << "\n");
291    TestFunctions.insert(cast<Function>(NewValueMap[TNOF]));
292    DeleteFunctionBody(TNOF);       // Function is now external in this module!
293  }
294
295
296  // Remove the Safe functions from the Test module
297  for (Module::iterator I = New->begin(), E = New->end(); I != E; ++I)
298    if (!TestFunctions.count(I))
299      DeleteFunctionBody(I);
300
301
302  // Make sure that there is a global ctor/dtor array in both halves of the
303  // module if they both have static ctor/dtor functions.
304  SplitStaticCtorDtor("llvm.global_ctors", M, New, NewValueMap);
305  SplitStaticCtorDtor("llvm.global_dtors", M, New, NewValueMap);
306
307  return New;
308}
309
310//===----------------------------------------------------------------------===//
311// Basic Block Extraction Code
312//===----------------------------------------------------------------------===//
313
314/// ExtractMappedBlocksFromModule - Extract all but the specified basic blocks
315/// into their own functions.  The only detail is that M is actually a module
316/// cloned from the one the BBs are in, so some mapping needs to be performed.
317/// If this operation fails for some reason (ie the implementation is buggy),
318/// this function should return null, otherwise it returns a new Module.
319Module *BugDriver::ExtractMappedBlocksFromModule(const
320                                                 std::vector<BasicBlock*> &BBs,
321                                                 Module *M) {
322  char *ExtraArg = NULL;
323
324  sys::Path uniqueFilename("bugpoint-extractblocks");
325  std::string ErrMsg;
326  if (uniqueFilename.createTemporaryFileOnDisk(true, &ErrMsg)) {
327    outs() << "*** Basic Block extraction failed!\n";
328    errs() << "Error creating temporary file: " << ErrMsg << "\n";
329    M = swapProgramIn(M);
330    EmitProgressBitcode("basicblockextractfail", true);
331    swapProgramIn(M);
332    return 0;
333  }
334  sys::RemoveFileOnSignal(uniqueFilename);
335
336  std::string ErrorInfo;
337  raw_fd_ostream BlocksToNotExtractFile(uniqueFilename.c_str(),
338                                        /*Binary=*/false, /*Force=*/true,
339                                        ErrorInfo);
340  if (!ErrorInfo.empty()) {
341    outs() << "*** Basic Block extraction failed!\n";
342    errs() << "Error writing list of blocks to not extract: " << ErrorInfo
343           << "\n";
344    M = swapProgramIn(M);
345    EmitProgressBitcode("basicblockextractfail", true);
346    swapProgramIn(M);
347    return 0;
348  }
349  for (std::vector<BasicBlock*>::const_iterator I = BBs.begin(), E = BBs.end();
350       I != E; ++I) {
351    BasicBlock *BB = *I;
352    // If the BB doesn't have a name, give it one so we have something to key
353    // off of.
354    if (!BB->hasName()) BB->setName("tmpbb");
355    BlocksToNotExtractFile << BB->getParent()->getNameStr() << " "
356                           << BB->getName() << "\n";
357  }
358  BlocksToNotExtractFile.close();
359
360  const char *uniqueFN = uniqueFilename.c_str();
361  ExtraArg = (char*)malloc(23 + strlen(uniqueFN));
362  strcat(strcpy(ExtraArg, "--extract-blocks-file="), uniqueFN);
363
364  std::vector<const PassInfo*> PI;
365  std::vector<BasicBlock *> EmptyBBs; // This parameter is ignored.
366  PI.push_back(getPI(createBlockExtractorPass(EmptyBBs)));
367  Module *Ret = runPassesOn(M, PI, false, 1, &ExtraArg);
368
369  if (uniqueFilename.exists())
370    uniqueFilename.eraseFromDisk(); // Free disk space
371  free(ExtraArg);
372
373  if (Ret == 0) {
374    outs() << "*** Basic Block extraction failed, please report a bug!\n";
375    M = swapProgramIn(M);
376    EmitProgressBitcode("basicblockextractfail", true);
377    swapProgramIn(M);
378  }
379  return Ret;
380}
381