tools/bugpoint/Miscompilation.cpp

   1 //===- Miscompilation.cpp - Debug program miscompilations -----------------===//
   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 optimizer and code generation miscompilation debugging
  11 // support.
  12 //
  13 //===----------------------------------------------------------------------===//
  14
  15 #include "BugDriver.h"
  16 #include "ListReducer.h"
  17 #include "llvm/Constants.h"
  18 #include "llvm/DerivedTypes.h"
  19 #include "llvm/Instructions.h"
  20 #include "llvm/Linker.h"
  21 #include "llvm/Module.h"
  22 #include "llvm/Pass.h"
  23 #include "llvm/Analysis/Verifier.h"
  24 #include "llvm/Support/Mangler.h"
  25 #include "llvm/Transforms/Utils/Cloning.h"
  26 #include "llvm/Support/CommandLine.h"
  27 #include "llvm/Support/FileUtilities.h"
  28 #include "llvm/Config/config.h"   // for HAVE_LINK_R
  29 using namespace llvm;
  30
  31 namespace llvm {
  32   extern cl::list<std::string> InputArgv;
  33 }
  34
  35 namespace {
  36   static llvm::cl::opt<bool>
  37     DisableLoopExtraction("disable-loop-extraction",
  38         cl::desc("Don't extract loops when searching for miscompilations"),
  39         cl::init(false));
  40   static llvm::cl::opt<bool>
  41     DisableBlockExtraction("disable-block-extraction",
  42         cl::desc("Don't extract blocks when searching for miscompilations"),
  43         cl::init(false));
  44
  45   class ReduceMiscompilingPasses : public ListReducer<const PassInfo*> {
  46     BugDriver &BD;
  47   public:
  48     ReduceMiscompilingPasses(BugDriver &bd) : BD(bd) {}
  49
  50     virtual TestResult doTest(std::vector<const PassInfo*> &Prefix,
  51                               std::vector<const PassInfo*> &Suffix);
  52   };
  53 }
  54
  55 /// TestResult - After passes have been split into a test group and a control
  56 /// group, see if they still break the program.
  57 ///
  58 ReduceMiscompilingPasses::TestResult
  59 ReduceMiscompilingPasses::doTest(std::vector<const PassInfo*> &Prefix,
  60                                  std::vector<const PassInfo*> &Suffix) {
  61   // First, run the program with just the Suffix passes.  If it is still broken
  62   // with JUST the kept passes, discard the prefix passes.
  63   outs() << "Checking to see if '" << getPassesString(Suffix)
  64          << "' compiles correctly: ";
  65
  66   std::string BitcodeResult;
  67   if (BD.runPasses(Suffix, BitcodeResult, false/*delete*/, true/*quiet*/)) {
  68     errs() << " Error running this sequence of passes"
  69            << " on the input program!\n";
  70     BD.setPassesToRun(Suffix);
  71     BD.EmitProgressBitcode("pass-error",  false);
  72     exit(BD.debugOptimizerCrash());
  73   }
  74
  75   // Check to see if the finished program matches the reference output...
  76   if (BD.diffProgram(BitcodeResult, "", true /*delete bitcode*/)) {
  77     outs() << " nope.\n";
  78     if (Suffix.empty()) {
  79       errs() << BD.getToolName() << ": I'm confused: the test fails when "
  80              << "no passes are run, nondeterministic program?\n";
  81       exit(1);
  82     }
  83     return KeepSuffix;         // Miscompilation detected!
  84   }
  85   outs() << " yup.\n";      // No miscompilation!
  86
  87   if (Prefix.empty()) return NoFailure;
  88
  89   // Next, see if the program is broken if we run the "prefix" passes first,
  90   // then separately run the "kept" passes.
  91   outs() << "Checking to see if '" << getPassesString(Prefix)
  92          << "' compiles correctly: ";
  93
  94   // If it is not broken with the kept passes, it's possible that the prefix
  95   // passes must be run before the kept passes to break it.  If the program
  96   // WORKS after the prefix passes, but then fails if running the prefix AND
  97   // kept passes, we can update our bitcode file to include the result of the
  98   // prefix passes, then discard the prefix passes.
  99   //
 100   if (BD.runPasses(Prefix, BitcodeResult, false/*delete*/, true/*quiet*/)) {
 101     errs() << " Error running this sequence of passes"
 102            << " on the input program!\n";
 103     BD.setPassesToRun(Prefix);
 104     BD.EmitProgressBitcode("pass-error",  false);
 105     exit(BD.debugOptimizerCrash());
 106   }
 107
 108   // If the prefix maintains the predicate by itself, only keep the prefix!
 109   if (BD.diffProgram(BitcodeResult)) {
 110     outs() << " nope.\n";
 111     sys::Path(BitcodeResult).eraseFromDisk();
 112     return KeepPrefix;
 113   }
 114   outs() << " yup.\n";      // No miscompilation!
 115
 116   // Ok, so now we know that the prefix passes work, try running the suffix
 117   // passes on the result of the prefix passes.
 118   //
 119   Module *PrefixOutput = ParseInputFile(BitcodeResult, BD.getContext());
 120   if (PrefixOutput == 0) {
 121     errs() << BD.getToolName() << ": Error reading bitcode file '"
 122            << BitcodeResult << "'!\n";
 123     exit(1);
 124   }
 125   sys::Path(BitcodeResult).eraseFromDisk();  // No longer need the file on disk
 126
 127   // Don't check if there are no passes in the suffix.
 128   if (Suffix.empty())
 129     return NoFailure;
 130
 131   outs() << "Checking to see if '" << getPassesString(Suffix)
 132             << "' passes compile correctly after the '"
 133             << getPassesString(Prefix) << "' passes: ";
 134
 135   Module *OriginalInput = BD.swapProgramIn(PrefixOutput);
 136   if (BD.runPasses(Suffix, BitcodeResult, false/*delete*/, true/*quiet*/)) {
 137     errs() << " Error running this sequence of passes"
 138            << " on the input program!\n";
 139     BD.setPassesToRun(Suffix);
 140     BD.EmitProgressBitcode("pass-error",  false);
 141     exit(BD.debugOptimizerCrash());
 142   }
 143
 144   // Run the result...
 145   if (BD.diffProgram(BitcodeResult, "", true/*delete bitcode*/)) {
 146     outs() << " nope.\n";
 147     delete OriginalInput;     // We pruned down the original input...
 148     return KeepSuffix;
 149   }
 150
 151   // Otherwise, we must not be running the bad pass anymore.
 152   outs() << " yup.\n";      // No miscompilation!
 153   delete BD.swapProgramIn(OriginalInput); // Restore orig program & free test
 154   return NoFailure;
 155 }
 156
 157 namespace {
 158   class ReduceMiscompilingFunctions : public ListReducer<Function*> {
 159     BugDriver &BD;
 160     bool (*TestFn)(BugDriver &, Module *, Module *);
 161   public:
 162     ReduceMiscompilingFunctions(BugDriver &bd,
 163                                 bool (*F)(BugDriver &, Module *, Module *))
 164       : BD(bd), TestFn(F) {}
 165
 166     virtual TestResult doTest(std::vector<Function*> &Prefix,
 167                               std::vector<Function*> &Suffix) {
 168       if (!Suffix.empty() && TestFuncs(Suffix))
 169         return KeepSuffix;
 170       if (!Prefix.empty() && TestFuncs(Prefix))
 171         return KeepPrefix;
 172       return NoFailure;
 173     }
 174
 175     bool TestFuncs(const std::vector<Function*> &Prefix);
 176   };
 177 }
 178
 179 /// TestMergedProgram - Given two modules, link them together and run the
 180 /// program, checking to see if the program matches the diff.  If the diff
 181 /// matches, return false, otherwise return true.  If the DeleteInputs argument
 182 /// is set to true then this function deletes both input modules before it
 183 /// returns.
 184 ///
 185 static bool TestMergedProgram(BugDriver &BD, Module *M1, Module *M2,
 186                               bool DeleteInputs) {
 187   // Link the two portions of the program back to together.
 188   std::string ErrorMsg;
 189   if (!DeleteInputs) {
 190     M1 = CloneModule(M1);
 191     M2 = CloneModule(M2);
 192   }
 193   if (Linker::LinkModules(M1, M2, &ErrorMsg)) {
 194     errs() << BD.getToolName() << ": Error linking modules together:"
 195            << ErrorMsg << '\n';
 196     exit(1);
 197   }
 198   delete M2;   // We are done with this module.
 199
 200   Module *OldProgram = BD.swapProgramIn(M1);
 201
 202   // Execute the program.  If it does not match the expected output, we must
 203   // return true.
 204   bool Broken = BD.diffProgram();
 205
 206   // Delete the linked module & restore the original
 207   BD.swapProgramIn(OldProgram);
 208   delete M1;
 209   return Broken;
 210 }
 211
 212 /// TestFuncs - split functions in a Module into two groups: those that are
 213 /// under consideration for miscompilation vs. those that are not, and test
 214 /// accordingly. Each group of functions becomes a separate Module.
 215 ///
 216 bool ReduceMiscompilingFunctions::TestFuncs(const std::vector<Function*>&Funcs){
 217   // Test to see if the function is misoptimized if we ONLY run it on the
 218   // functions listed in Funcs.
 219   outs() << "Checking to see if the program is misoptimized when "
 220          << (Funcs.size()==1 ? "this function is" : "these functions are")
 221          << " run through the pass"
 222          << (BD.getPassesToRun().size() == 1 ? "" : "es") << ":";
 223   PrintFunctionList(Funcs);
 224   outs() << '\n';
 225
 226   // Split the module into the two halves of the program we want.
 227   DenseMap<const Value*, Value*> ValueMap;
 228   Module *ToNotOptimize = CloneModule(BD.getProgram(), ValueMap);
 229   Module *ToOptimize = SplitFunctionsOutOfModule(ToNotOptimize, Funcs,
 230                                                  ValueMap);
 231
 232   // Run the predicate, note that the predicate will delete both input modules.
 233   return TestFn(BD, ToOptimize, ToNotOptimize);
 234 }
 235
 236 /// DisambiguateGlobalSymbols - Mangle symbols to guarantee uniqueness by
 237 /// modifying predominantly internal symbols rather than external ones.
 238 ///
 239 static void DisambiguateGlobalSymbols(Module *M) {
 240   // Try not to cause collisions by minimizing chances of renaming an
 241   // already-external symbol, so take in external globals and functions as-is.
 242   // The code should work correctly without disambiguation (assuming the same
 243   // mangler is used by the two code generators), but having symbols with the
 244   // same name causes warnings to be emitted by the code generator.
 245   Mangler Mang(*M);
 246   // Agree with the CBE on symbol naming
 247   Mang.markCharUnacceptable('.');
 248   for (Module::global_iterator I = M->global_begin(), E = M->global_end();
 249        I != E; ++I) {
 250     // Don't mangle asm names.
 251     if (!I->hasName() || I->getName()[0] != 1)
 252       I->setName(Mang.getMangledName(I));
 253   }
 254   for (Module::iterator I = M->begin(), E = M->end(); I != E; ++I) {
 255     // Don't mangle asm names or intrinsics.
 256     if ((!I->hasName() || I->getName()[0] != 1) &&
 257         I->getIntrinsicID() == 0)
 258       I->setName(Mang.getMangledName(I));
 259   }
 260 }
 261
 262 /// ExtractLoops - Given a reduced list of functions that still exposed the bug,
 263 /// check to see if we can extract the loops in the region without obscuring the
 264 /// bug.  If so, it reduces the amount of code identified.
 265 ///
 266 static bool ExtractLoops(BugDriver &BD,
 267                          bool (*TestFn)(BugDriver &, Module *, Module *),
 268                          std::vector<Function*> &MiscompiledFunctions) {
 269   bool MadeChange = false;
 270   while (1) {
 271     if (BugpointIsInterrupted) return MadeChange;
 272
 273     DenseMap<const Value*, Value*> ValueMap;
 274     Module *ToNotOptimize = CloneModule(BD.getProgram(), ValueMap);
 275     Module *ToOptimize = SplitFunctionsOutOfModule(ToNotOptimize,
 276                                                    MiscompiledFunctions,
 277                                                    ValueMap);
 278     Module *ToOptimizeLoopExtracted = BD.ExtractLoop(ToOptimize);
 279     if (!ToOptimizeLoopExtracted) {
 280       // If the loop extractor crashed or if there were no extractible loops,
 281       // then this chapter of our odyssey is over with.
 282       delete ToNotOptimize;
 283       delete ToOptimize;
 284       return MadeChange;
 285     }
 286
 287     errs() << "Extracted a loop from the breaking portion of the program.\n";
 288
 289     // Bugpoint is intentionally not very trusting of LLVM transformations.  In
 290     // particular, we're not going to assume that the loop extractor works, so
 291     // we're going to test the newly loop extracted program to make sure nothing
 292     // has broken.  If something broke, then we'll inform the user and stop
 293     // extraction.
 294     AbstractInterpreter *AI = BD.switchToSafeInterpreter();
 295     if (TestMergedProgram(BD, ToOptimizeLoopExtracted, ToNotOptimize, false)) {
 296       BD.switchToInterpreter(AI);
 297
 298       // Merged program doesn't work anymore!
 299       errs() << "  *** ERROR: Loop extraction broke the program. :("
 300              << " Please report a bug!\n";
 301       errs() << "      Continuing on with un-loop-extracted version.\n";
 302
 303       BD.writeProgramToFile("bugpoint-loop-extract-fail-tno.bc", ToNotOptimize);
 304       BD.writeProgramToFile("bugpoint-loop-extract-fail-to.bc", ToOptimize);
 305       BD.writeProgramToFile("bugpoint-loop-extract-fail-to-le.bc",
 306                             ToOptimizeLoopExtracted);
 307
 308       errs() << "Please submit the bugpoint-loop-extract-fail-*.bc files.\n";
 309       delete ToOptimize;
 310       delete ToNotOptimize;
 311       delete ToOptimizeLoopExtracted;
 312       return MadeChange;
 313     }
 314     delete ToOptimize;
 315     BD.switchToInterpreter(AI);
 316
 317     outs() << "  Testing after loop extraction:\n";
 318     // Clone modules, the tester function will free them.
 319     Module *TOLEBackup = CloneModule(ToOptimizeLoopExtracted);
 320     Module *TNOBackup  = CloneModule(ToNotOptimize);
 321     if (!TestFn(BD, ToOptimizeLoopExtracted, ToNotOptimize)) {
 322       outs() << "*** Loop extraction masked the problem.  Undoing.\n";
 323       // If the program is not still broken, then loop extraction did something
 324       // that masked the error.  Stop loop extraction now.
 325       delete TOLEBackup;
 326       delete TNOBackup;
 327       return MadeChange;
 328     }
 329     ToOptimizeLoopExtracted = TOLEBackup;
 330     ToNotOptimize = TNOBackup;
 331
 332     outs() << "*** Loop extraction successful!\n";
 333
 334     std::vector<std::pair<std::string, const FunctionType*> > MisCompFunctions;
 335     for (Module::iterator I = ToOptimizeLoopExtracted->begin(),
 336            E = ToOptimizeLoopExtracted->end(); I != E; ++I)
 337       if (!I->isDeclaration())
 338         MisCompFunctions.push_back(std::make_pair(I->getName(),
 339                                                   I->getFunctionType()));
 340
 341     // Okay, great!  Now we know that we extracted a loop and that loop
 342     // extraction both didn't break the program, and didn't mask the problem.
 343     // Replace the current program with the loop extracted version, and try to
 344     // extract another loop.
 345     std::string ErrorMsg;
 346     if (Linker::LinkModules(ToNotOptimize, ToOptimizeLoopExtracted, &ErrorMsg)){
 347       errs() << BD.getToolName() << ": Error linking modules together:"
 348              << ErrorMsg << '\n';
 349       exit(1);
 350     }
 351     delete ToOptimizeLoopExtracted;
 352
 353     // All of the Function*'s in the MiscompiledFunctions list are in the old
 354     // module.  Update this list to include all of the functions in the
 355     // optimized and loop extracted module.
 356     MiscompiledFunctions.clear();
 357     for (unsigned i = 0, e = MisCompFunctions.size(); i != e; ++i) {
 358       Function *NewF = ToNotOptimize->getFunction(MisCompFunctions[i].first);
 359
 360       assert(NewF && "Function not found??");
 361       assert(NewF->getFunctionType() == MisCompFunctions[i].second &&
 362              "found wrong function type?");
 363       MiscompiledFunctions.push_back(NewF);
 364     }
 365
 366     BD.setNewProgram(ToNotOptimize);
 367     MadeChange = true;
 368   }
 369 }
 370
 371 namespace {
 372   class ReduceMiscompiledBlocks : public ListReducer<BasicBlock*> {
 373     BugDriver &BD;
 374     bool (*TestFn)(BugDriver &, Module *, Module *);
 375     std::vector<Function*> FunctionsBeingTested;
 376   public:
 377     ReduceMiscompiledBlocks(BugDriver &bd,
 378                             bool (*F)(BugDriver &, Module *, Module *),
 379                             const std::vector<Function*> &Fns)
 380       : BD(bd), TestFn(F), FunctionsBeingTested(Fns) {}
 381
 382     virtual TestResult doTest(std::vector<BasicBlock*> &Prefix,
 383                               std::vector<BasicBlock*> &Suffix) {
 384       if (!Suffix.empty() && TestFuncs(Suffix))
 385         return KeepSuffix;
 386       if (TestFuncs(Prefix))
 387         return KeepPrefix;
 388       return NoFailure;
 389     }
 390
 391     bool TestFuncs(const std::vector<BasicBlock*> &Prefix);
 392   };
 393 }
 394
 395 /// TestFuncs - Extract all blocks for the miscompiled functions except for the
 396 /// specified blocks.  If the problem still exists, return true.
 397 ///
 398 bool ReduceMiscompiledBlocks::TestFuncs(const std::vector<BasicBlock*> &BBs) {
 399   // Test to see if the function is misoptimized if we ONLY run it on the
 400   // functions listed in Funcs.
 401   outs() << "Checking to see if the program is misoptimized when all ";
 402   if (!BBs.empty()) {
 403     outs() << "but these " << BBs.size() << " blocks are extracted: ";
 404     for (unsigned i = 0, e = BBs.size() < 10 ? BBs.size() : 10; i != e; ++i)
 405       outs() << BBs[i]->getName() << " ";
 406     if (BBs.size() > 10) outs() << "...";
 407   } else {
 408     outs() << "blocks are extracted.";
 409   }
 410   outs() << '\n';
 411
 412   // Split the module into the two halves of the program we want.
 413   DenseMap<const Value*, Value*> ValueMap;
 414   Module *ToNotOptimize = CloneModule(BD.getProgram(), ValueMap);
 415   Module *ToOptimize = SplitFunctionsOutOfModule(ToNotOptimize,
 416                                                  FunctionsBeingTested,
 417                                                  ValueMap);
 418
 419   // Try the extraction.  If it doesn't work, then the block extractor crashed
 420   // or something, in which case bugpoint can't chase down this possibility.
 421   if (Module *New = BD.ExtractMappedBlocksFromModule(BBs, ToOptimize)) {
 422     delete ToOptimize;
 423     // Run the predicate, not that the predicate will delete both input modules.
 424     return TestFn(BD, New, ToNotOptimize);
 425   }
 426   delete ToOptimize;
 427   delete ToNotOptimize;
 428   return false;
 429 }
 430
 431
 432 /// ExtractBlocks - Given a reduced list of functions that still expose the bug,
 433 /// extract as many basic blocks from the region as possible without obscuring
 434 /// the bug.
 435 ///
 436 static bool ExtractBlocks(BugDriver &BD,
 437                           bool (*TestFn)(BugDriver &, Module *, Module *),
 438                           std::vector<Function*> &MiscompiledFunctions) {
 439   if (BugpointIsInterrupted) return false;
 440
 441   std::vector<BasicBlock*> Blocks;
 442   for (unsigned i = 0, e = MiscompiledFunctions.size(); i != e; ++i)
 443     for (Function::iterator I = MiscompiledFunctions[i]->begin(),
 444            E = MiscompiledFunctions[i]->end(); I != E; ++I)
 445       Blocks.push_back(I);
 446
 447   // Use the list reducer to identify blocks that can be extracted without
 448   // obscuring the bug.  The Blocks list will end up containing blocks that must
 449   // be retained from the original program.
 450   unsigned OldSize = Blocks.size();
 451
 452   // Check to see if all blocks are extractible first.
 453   if (ReduceMiscompiledBlocks(BD, TestFn,
 454                   MiscompiledFunctions).TestFuncs(std::vector<BasicBlock*>())) {
 455     Blocks.clear();
 456   } else {
 457     ReduceMiscompiledBlocks(BD, TestFn,MiscompiledFunctions).reduceList(Blocks);
 458     if (Blocks.size() == OldSize)
 459       return false;
 460   }
 461
 462   DenseMap<const Value*, Value*> ValueMap;
 463   Module *ProgClone = CloneModule(BD.getProgram(), ValueMap);
 464   Module *ToExtract = SplitFunctionsOutOfModule(ProgClone,
 465                                                 MiscompiledFunctions,
 466                                                 ValueMap);
 467   Module *Extracted = BD.ExtractMappedBlocksFromModule(Blocks, ToExtract);
 468   if (Extracted == 0) {
 469     // Weird, extraction should have worked.
 470     errs() << "Nondeterministic problem extracting blocks??\n";
 471     delete ProgClone;
 472     delete ToExtract;
 473     return false;
 474   }
 475
 476   // Otherwise, block extraction succeeded.  Link the two program fragments back
 477   // together.
 478   delete ToExtract;
 479
 480   std::vector<std::pair<std::string, const FunctionType*> > MisCompFunctions;
 481   for (Module::iterator I = Extracted->begin(), E = Extracted->end();
 482        I != E; ++I)
 483     if (!I->isDeclaration())
 484       MisCompFunctions.push_back(std::make_pair(I->getName(),
 485                                                 I->getFunctionType()));
 486
 487   std::string ErrorMsg;
 488   if (Linker::LinkModules(ProgClone, Extracted, &ErrorMsg)) {
 489     errs() << BD.getToolName() << ": Error linking modules together:"
 490            << ErrorMsg << '\n';
 491     exit(1);
 492   }
 493   delete Extracted;
 494
 495   // Set the new program and delete the old one.
 496   BD.setNewProgram(ProgClone);
 497
 498   // Update the list of miscompiled functions.
 499   MiscompiledFunctions.clear();
 500
 501   for (unsigned i = 0, e = MisCompFunctions.size(); i != e; ++i) {
 502     Function *NewF = ProgClone->getFunction(MisCompFunctions[i].first);
 503     assert(NewF && "Function not found??");
 504     assert(NewF->getFunctionType() == MisCompFunctions[i].second &&
 505            "Function has wrong type??");
 506     MiscompiledFunctions.push_back(NewF);
 507   }
 508
 509   return true;
 510 }
 511
 512
 513 /// DebugAMiscompilation - This is a generic driver to narrow down
 514 /// miscompilations, either in an optimization or a code generator.
 515 ///
 516 static std::vector<Function*>
 517 DebugAMiscompilation(BugDriver &BD,
 518                      bool (*TestFn)(BugDriver &, Module *, Module *)) {
 519   // Okay, now that we have reduced the list of passes which are causing the
 520   // failure, see if we can pin down which functions are being
 521   // miscompiled... first build a list of all of the non-external functions in
 522   // the program.
 523   std::vector<Function*> MiscompiledFunctions;
 524   Module *Prog = BD.getProgram();
 525   for (Module::iterator I = Prog->begin(), E = Prog->end(); I != E; ++I)
 526     if (!I->isDeclaration())
 527       MiscompiledFunctions.push_back(I);
 528
 529   // Do the reduction...
 530   if (!BugpointIsInterrupted)
 531     ReduceMiscompilingFunctions(BD, TestFn).reduceList(MiscompiledFunctions);
 532
 533   outs() << "\n*** The following function"
 534          << (MiscompiledFunctions.size() == 1 ? " is" : "s are")
 535          << " being miscompiled: ";
 536   PrintFunctionList(MiscompiledFunctions);
 537   outs() << '\n';
 538
 539   // See if we can rip any loops out of the miscompiled functions and still
 540   // trigger the problem.
 541
 542   if (!BugpointIsInterrupted && !DisableLoopExtraction &&
 543       ExtractLoops(BD, TestFn, MiscompiledFunctions)) {
 544     // Okay, we extracted some loops and the problem still appears.  See if we
 545     // can eliminate some of the created functions from being candidates.
 546
 547     // Loop extraction can introduce functions with the same name (foo_code).
 548     // Make sure to disambiguate the symbols so that when the program is split
 549     // apart that we can link it back together again.
 550     DisambiguateGlobalSymbols(BD.getProgram());
 551
 552     // Do the reduction...
 553     if (!BugpointIsInterrupted)
 554       ReduceMiscompilingFunctions(BD, TestFn).reduceList(MiscompiledFunctions);
 555
 556     outs() << "\n*** The following function"
 557            << (MiscompiledFunctions.size() == 1 ? " is" : "s are")
 558            << " being miscompiled: ";
 559     PrintFunctionList(MiscompiledFunctions);
 560     outs() << '\n';
 561   }
 562
 563   if (!BugpointIsInterrupted && !DisableBlockExtraction &&
 564       ExtractBlocks(BD, TestFn, MiscompiledFunctions)) {
 565     // Okay, we extracted some blocks and the problem still appears.  See if we
 566     // can eliminate some of the created functions from being candidates.
 567
 568     // Block extraction can introduce functions with the same name (foo_code).
 569     // Make sure to disambiguate the symbols so that when the program is split
 570     // apart that we can link it back together again.
 571     DisambiguateGlobalSymbols(BD.getProgram());
 572
 573     // Do the reduction...
 574     ReduceMiscompilingFunctions(BD, TestFn).reduceList(MiscompiledFunctions);
 575
 576     outs() << "\n*** The following function"
 577            << (MiscompiledFunctions.size() == 1 ? " is" : "s are")
 578            << " being miscompiled: ";
 579     PrintFunctionList(MiscompiledFunctions);
 580     outs() << '\n';
 581   }
 582
 583   return MiscompiledFunctions;
 584 }
 585
 586 /// TestOptimizer - This is the predicate function used to check to see if the
 587 /// "Test" portion of the program is misoptimized.  If so, return true.  In any
 588 /// case, both module arguments are deleted.
 589 ///
 590 static bool TestOptimizer(BugDriver &BD, Module *Test, Module *Safe) {
 591   // Run the optimization passes on ToOptimize, producing a transformed version
 592   // of the functions being tested.
 593   outs() << "  Optimizing functions being tested: ";
 594   Module *Optimized = BD.runPassesOn(Test, BD.getPassesToRun(),
 595                                      /*AutoDebugCrashes*/true);
 596   outs() << "done.\n";
 597   delete Test;
 598
 599   outs() << "  Checking to see if the merged program executes correctly: ";
 600   bool Broken = TestMergedProgram(BD, Optimized, Safe, true);
 601   outs() << (Broken ? " nope.\n" : " yup.\n");
 602   return Broken;
 603 }
 604
 605
 606 /// debugMiscompilation - This method is used when the passes selected are not
 607 /// crashing, but the generated output is semantically different from the
 608 /// input.
 609 ///
 610 bool BugDriver::debugMiscompilation() {
 611   // Make sure something was miscompiled...
 612   if (!BugpointIsInterrupted)
 613     if (!ReduceMiscompilingPasses(*this).reduceList(PassesToRun)) {
 614       errs() << "*** Optimized program matches reference output!  No problem"
 615              << " detected...\nbugpoint can't help you with your problem!\n";
 616       return false;
 617     }
 618
 619   outs() << "\n*** Found miscompiling pass"
 620          << (getPassesToRun().size() == 1 ? "" : "es") << ": "
 621          << getPassesString(getPassesToRun()) << '\n';
 622   EmitProgressBitcode("passinput");
 623
 624   std::vector<Function*> MiscompiledFunctions =
 625     DebugAMiscompilation(*this, TestOptimizer);
 626
 627   // Output a bunch of bitcode files for the user...
 628   outs() << "Outputting reduced bitcode files which expose the problem:\n";
 629   DenseMap<const Value*, Value*> ValueMap;
 630   Module *ToNotOptimize = CloneModule(getProgram(), ValueMap);
 631   Module *ToOptimize = SplitFunctionsOutOfModule(ToNotOptimize,
 632                                                  MiscompiledFunctions,
 633                                                  ValueMap);
 634
 635   outs() << "  Non-optimized portion: ";
 636   ToNotOptimize = swapProgramIn(ToNotOptimize);
 637   EmitProgressBitcode("tonotoptimize", true);
 638   setNewProgram(ToNotOptimize);   // Delete hacked module.
 639
 640   outs() << "  Portion that is input to optimizer: ";
 641   ToOptimize = swapProgramIn(ToOptimize);
 642   EmitProgressBitcode("tooptimize");
 643   setNewProgram(ToOptimize);      // Delete hacked module.
 644
 645   return false;
 646 }
 647
 648 /// CleanupAndPrepareModules - Get the specified modules ready for code
 649 /// generator testing.
 650 ///
 651 static void CleanupAndPrepareModules(BugDriver &BD, Module *&Test,
 652                                      Module *Safe) {
 653   // Clean up the modules, removing extra cruft that we don't need anymore...
 654   Test = BD.performFinalCleanups(Test);
 655
 656   // If we are executing the JIT, we have several nasty issues to take care of.
 657   if (!BD.isExecutingJIT()) return;
 658
 659   // First, if the main function is in the Safe module, we must add a stub to
 660   // the Test module to call into it.  Thus, we create a new function `main'
 661   // which just calls the old one.
 662   if (Function *oldMain = Safe->getFunction("main"))
 663     if (!oldMain->isDeclaration()) {
 664       // Rename it
 665       oldMain->setName("llvm_bugpoint_old_main");
 666       // Create a NEW `main' function with same type in the test module.
 667       Function *newMain = Function::Create(oldMain->getFunctionType(),
 668                                            GlobalValue::ExternalLinkage,
 669                                            "main", Test);
 670       // Create an `oldmain' prototype in the test module, which will
 671       // corresponds to the real main function in the same module.
 672       Function *oldMainProto = Function::Create(oldMain->getFunctionType(),
 673                                                 GlobalValue::ExternalLinkage,
 674                                                 oldMain->getName(), Test);
 675       // Set up and remember the argument list for the main function.
 676       std::vector<Value*> args;
 677       for (Function::arg_iterator
 678              I = newMain->arg_begin(), E = newMain->arg_end(),
 679              OI = oldMain->arg_begin(); I != E; ++I, ++OI) {
 680         I->setName(OI->getName());    // Copy argument names from oldMain
 681         args.push_back(I);
 682       }
 683
 684       // Call the old main function and return its result
 685       BasicBlock *BB = BasicBlock::Create("entry", newMain);
 686       CallInst *call = CallInst::Create(oldMainProto, args.begin(), args.end(),
 687                                         "", BB);
 688
 689       // If the type of old function wasn't void, return value of call
 690       ReturnInst::Create(call, BB);
 691     }
 692
 693   // The second nasty issue we must deal with in the JIT is that the Safe
 694   // module cannot directly reference any functions defined in the test
 695   // module.  Instead, we use a JIT API call to dynamically resolve the
 696   // symbol.
 697
 698   // Add the resolver to the Safe module.
 699   // Prototype: void *getPointerToNamedFunction(const char* Name)
 700   Constant *resolverFunc =
 701     Safe->getOrInsertFunction("getPointerToNamedFunction",
 702                         PointerType::getUnqual(Type::Int8Ty),
 703                         PointerType::getUnqual(Type::Int8Ty), (Type *)0);
 704
 705   // Use the function we just added to get addresses of functions we need.
 706   for (Module::iterator F = Safe->begin(), E = Safe->end(); F != E; ++F) {
 707     if (F->isDeclaration() && !F->use_empty() && &*F != resolverFunc &&
 708         !F->isIntrinsic() /* ignore intrinsics */) {
 709       Function *TestFn = Test->getFunction(F->getName());
 710
 711       // Don't forward functions which are external in the test module too.
 712       if (TestFn && !TestFn->isDeclaration()) {
 713         // 1. Add a string constant with its name to the global file
 714         Constant *InitArray = ConstantArray::get(F->getName());
 715         GlobalVariable *funcName =
 716           new GlobalVariable(*Safe, InitArray->getType(), true /*isConstant*/,
 717                              GlobalValue::InternalLinkage, InitArray,
 718                              F->getName() + "_name");
 719
 720         // 2. Use `GetElementPtr *funcName, 0, 0' to convert the string to an
 721         // sbyte* so it matches the signature of the resolver function.
 722
 723         // GetElementPtr *funcName, ulong 0, ulong 0
 724         std::vector<Constant*> GEPargs(2, Constant::getNullValue(Type::Int32Ty));
 725         Value *GEP =
 726                 ConstantExpr::getGetElementPtr(funcName, &GEPargs[0], 2);
 727         std::vector<Value*> ResolverArgs;
 728         ResolverArgs.push_back(GEP);
 729
 730         // Rewrite uses of F in global initializers, etc. to uses of a wrapper
 731         // function that dynamically resolves the calls to F via our JIT API
 732         if (!F->use_empty()) {
 733           // Create a new global to hold the cached function pointer.
 734           Constant *NullPtr = ConstantPointerNull::get(F->getType());
 735           GlobalVariable *Cache =
 736             new GlobalVariable(*F->getParent(), F->getType(),
 737                                false, GlobalValue::InternalLinkage,
 738                                NullPtr,F->getName()+".fpcache");
 739
 740           // Construct a new stub function that will re-route calls to F
 741           const FunctionType *FuncTy = F->getFunctionType();
 742           Function *FuncWrapper = Function::Create(FuncTy,
 743                                                    GlobalValue::InternalLinkage,
 744                                                    F->getName() + "_wrapper",
 745                                                    F->getParent());
 746           BasicBlock *EntryBB  = BasicBlock::Create("entry", FuncWrapper);
 747           BasicBlock *DoCallBB = BasicBlock::Create("usecache", FuncWrapper);
 748           BasicBlock *LookupBB = BasicBlock::Create("lookupfp", FuncWrapper);
 749
 750           // Check to see if we already looked up the value.
 751           Value *CachedVal = new LoadInst(Cache, "fpcache", EntryBB);
 752           Value *IsNull = new ICmpInst(*EntryBB, ICmpInst::ICMP_EQ, CachedVal,
 753                                        NullPtr, "isNull");
 754           BranchInst::Create(LookupBB, DoCallBB, IsNull, EntryBB);
 755
 756           // Resolve the call to function F via the JIT API:
 757           //
 758           // call resolver(GetElementPtr...)
 759           CallInst *Resolver =
 760             CallInst::Create(resolverFunc, ResolverArgs.begin(),
 761                              ResolverArgs.end(), "resolver", LookupBB);
 762
 763           // Cast the result from the resolver to correctly-typed function.
 764           CastInst *CastedResolver =
 765             new BitCastInst(Resolver,
 766                             PointerType::getUnqual(F->getFunctionType()),
 767                             "resolverCast", LookupBB);
 768
 769           // Save the value in our cache.
 770           new StoreInst(CastedResolver, Cache, LookupBB);
 771           BranchInst::Create(DoCallBB, LookupBB);
 772
 773           PHINode *FuncPtr = PHINode::Create(NullPtr->getType(),
 774                                              "fp", DoCallBB);
 775           FuncPtr->addIncoming(CastedResolver, LookupBB);
 776           FuncPtr->addIncoming(CachedVal, EntryBB);
 777
 778           // Save the argument list.
 779           std::vector<Value*> Args;
 780           for (Function::arg_iterator i = FuncWrapper->arg_begin(),
 781                  e = FuncWrapper->arg_end(); i != e; ++i)
 782             Args.push_back(i);
 783
 784           // Pass on the arguments to the real function, return its result
 785           if (F->getReturnType() == Type::VoidTy) {
 786             CallInst::Create(FuncPtr, Args.begin(), Args.end(), "", DoCallBB);
 787             ReturnInst::Create(DoCallBB);
 788           } else {
 789             CallInst *Call = CallInst::Create(FuncPtr, Args.begin(), Args.end(),
 790                                               "retval", DoCallBB);
 791             ReturnInst::Create(Call, DoCallBB);
 792           }
 793
 794           // Use the wrapper function instead of the old function
 795           F->replaceAllUsesWith(FuncWrapper);
 796         }
 797       }
 798     }
 799   }
 800
 801   if (verifyModule(*Test) || verifyModule(*Safe)) {
 802     errs() << "Bugpoint has a bug, which corrupted a module!!\n";
 803     abort();
 804   }
 805 }
 806
 807
 808
 809 /// TestCodeGenerator - This is the predicate function used to check to see if
 810 /// the "Test" portion of the program is miscompiled by the code generator under
 811 /// test.  If so, return true.  In any case, both module arguments are deleted.
 812 ///
 813 static bool TestCodeGenerator(BugDriver &BD, Module *Test, Module *Safe) {
 814   CleanupAndPrepareModules(BD, Test, Safe);
 815
 816   sys::Path TestModuleBC("bugpoint.test.bc");
 817   std::string ErrMsg;
 818   if (TestModuleBC.makeUnique(true, &ErrMsg)) {
 819     errs() << BD.getToolName() << "Error making unique filename: "
 820            << ErrMsg << "\n";
 821     exit(1);
 822   }
 823   if (BD.writeProgramToFile(TestModuleBC.toString(), Test)) {
 824     errs() << "Error writing bitcode to `" << TestModuleBC << "'\nExiting.";
 825     exit(1);
 826   }
 827   delete Test;
 828
 829   // Make the shared library
 830   sys::Path SafeModuleBC("bugpoint.safe.bc");
 831   if (SafeModuleBC.makeUnique(true, &ErrMsg)) {
 832     errs() << BD.getToolName() << "Error making unique filename: "
 833            << ErrMsg << "\n";
 834     exit(1);
 835   }
 836
 837   if (BD.writeProgramToFile(SafeModuleBC.toString(), Safe)) {
 838     errs() << "Error writing bitcode to `" << SafeModuleBC << "'\nExiting.";
 839     exit(1);
 840   }
 841   std::string SharedObject = BD.compileSharedObject(SafeModuleBC.toString());
 842   delete Safe;
 843
 844   // Run the code generator on the `Test' code, loading the shared library.
 845   // The function returns whether or not the new output differs from reference.
 846   int Result = BD.diffProgram(TestModuleBC.toString(), SharedObject, false);
 847
 848   if (Result)
 849     errs() << ": still failing!\n";
 850   else
 851     errs() << ": didn't fail.\n";
 852   TestModuleBC.eraseFromDisk();
 853   SafeModuleBC.eraseFromDisk();
 854   sys::Path(SharedObject).eraseFromDisk();
 855
 856   return Result;
 857 }
 858
 859
 860 /// debugCodeGenerator - debug errors in LLC, LLI, or CBE.
 861 ///
 862 bool BugDriver::debugCodeGenerator() {
 863   if ((void*)SafeInterpreter == (void*)Interpreter) {
 864     std::string Result = executeProgramSafely("bugpoint.safe.out");
 865     outs() << "\n*** The \"safe\" i.e. 'known good' backend cannot match "
 866            << "the reference diff.  This may be due to a\n    front-end "
 867            << "bug or a bug in the original program, but this can also "
 868            << "happen if bugpoint isn't running the program with the "
 869            << "right flags or input.\n    I left the result of executing "
 870            << "the program with the \"safe\" backend in this file for "
 871            << "you: '"
 872            << Result << "'.\n";
 873     return true;
 874   }
 875
 876   DisambiguateGlobalSymbols(Program);
 877
 878   std::vector<Function*> Funcs = DebugAMiscompilation(*this, TestCodeGenerator);
 879
 880   // Split the module into the two halves of the program we want.
 881   DenseMap<const Value*, Value*> ValueMap;
 882   Module *ToNotCodeGen = CloneModule(getProgram(), ValueMap);
 883   Module *ToCodeGen = SplitFunctionsOutOfModule(ToNotCodeGen, Funcs, ValueMap);
 884
 885   // Condition the modules
 886   CleanupAndPrepareModules(*this, ToCodeGen, ToNotCodeGen);
 887
 888   sys::Path TestModuleBC("bugpoint.test.bc");
 889   std::string ErrMsg;
 890   if (TestModuleBC.makeUnique(true, &ErrMsg)) {
 891     errs() << getToolName() << "Error making unique filename: "
 892            << ErrMsg << "\n";
 893     exit(1);
 894   }
 895
 896   if (writeProgramToFile(TestModuleBC.toString(), ToCodeGen)) {
 897     errs() << "Error writing bitcode to `" << TestModuleBC << "'\nExiting.";
 898     exit(1);
 899   }
 900   delete ToCodeGen;
 901
 902   // Make the shared library
 903   sys::Path SafeModuleBC("bugpoint.safe.bc");
 904   if (SafeModuleBC.makeUnique(true, &ErrMsg)) {
 905     errs() << getToolName() << "Error making unique filename: "
 906            << ErrMsg << "\n";
 907     exit(1);
 908   }
 909
 910   if (writeProgramToFile(SafeModuleBC.toString(), ToNotCodeGen)) {
 911     errs() << "Error writing bitcode to `" << SafeModuleBC << "'\nExiting.";
 912     exit(1);
 913   }
 914   std::string SharedObject = compileSharedObject(SafeModuleBC.toString());
 915   delete ToNotCodeGen;
 916
 917   outs() << "You can reproduce the problem with the command line: \n";
 918   if (isExecutingJIT()) {
 919     outs() << "  lli -load " << SharedObject << " " << TestModuleBC;
 920   } else {
 921     outs() << "  llc -f " << TestModuleBC << " -o " << TestModuleBC<< ".s\n";
 922     outs() << "  gcc " << SharedObject << " " << TestModuleBC
 923               << ".s -o " << TestModuleBC << ".exe";
 924 #if defined (HAVE_LINK_R)
 925     outs() << " -Wl,-R.";
 926 #endif
 927     outs() << "\n";
 928     outs() << "  " << TestModuleBC << ".exe";
 929   }
 930   for (unsigned i=0, e = InputArgv.size(); i != e; ++i)
 931     outs() << " " << InputArgv[i];
 932   outs() << '\n';
 933   outs() << "The shared object was created with:\n  llc -march=c "
 934          << SafeModuleBC << " -o temporary.c\n"
 935          << "  gcc -xc temporary.c -O2 -o " << SharedObject
 936 #if defined(sparc) || defined(__sparc__) || defined(__sparcv9)
 937          << " -G"            // Compile a shared library, `-G' for Sparc
 938 #else
 939          << " -fPIC -shared"       // `-shared' for Linux/X86, maybe others
 940 #endif
 941          << " -fno-strict-aliasing\n";
 942
 943   return false;
 944 }