reference, declarationdefinition
definition → references, declarations, derived classes, virtual overrides
reference to multiple definitions → definitions
unreferenced
    1
    2
    3
    4
    5
    6
    7
    8
    9
   10
   11
   12
   13
   14
   15
   16
   17
   18
   19
   20
   21
   22
   23
   24
   25
   26
   27
   28
   29
   30
   31
   32
   33
   34
   35
   36
   37
   38
   39
   40
   41
   42
   43
   44
   45
   46
   47
   48
   49
   50
   51
   52
   53
   54
   55
   56
   57
   58
   59
   60
   61
   62
   63
   64
   65
   66
   67
   68
   69
   70
   71
   72
   73
   74
   75
   76
   77
   78
   79
   80
   81
   82
   83
   84
   85
   86
   87
   88
   89
   90
   91
   92
   93
   94
   95
   96
   97
   98
   99
  100
  101
  102
  103
  104
  105
  106
  107
  108
  109
  110
  111
  112
  113
  114
  115
  116
  117
  118
  119
  120
  121
  122
  123
  124
  125
  126
  127
  128
  129
  130
  131
  132
  133
  134
  135
  136
  137
  138
  139
  140
  141
  142
  143
  144
  145
  146
  147
  148
  149
  150
  151
  152
  153
  154
  155
  156
  157
  158
  159
  160
  161
  162
  163
  164
  165
  166
  167
  168
  169
  170
  171
  172
  173
  174
  175
  176
  177
  178
  179
  180
  181
  182
  183
  184
  185
  186
  187
  188
  189
  190
  191
  192
  193
  194
  195
  196
  197
  198
  199
  200
  201
  202
  203
  204
  205
  206
  207
  208
  209
  210
  211
  212
  213
  214
  215
  216
  217
  218
  219
  220
  221
  222
  223
  224
  225
  226
  227
  228
  229
  230
  231
  232
  233
  234
  235
  236
  237
  238
  239
  240
  241
  242
  243
  244
  245
  246
  247
  248
  249
  250
  251
  252
  253
  254
  255
  256
  257
  258
  259
  260
  261
  262
  263
  264
  265
  266
  267
  268
  269
  270
  271
  272
  273
  274
  275
  276
  277
  278
  279
  280
  281
//===- FuzzerDataFlowTrace.cpp - DataFlowTrace                ---*- C++ -* ===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
// fuzzer::DataFlowTrace
//===----------------------------------------------------------------------===//

#include "FuzzerDataFlowTrace.h"

#include "FuzzerCommand.h"
#include "FuzzerIO.h"
#include "FuzzerRandom.h"
#include "FuzzerSHA1.h"
#include "FuzzerUtil.h"

#include <cstdlib>
#include <fstream>
#include <numeric>
#include <queue>
#include <sstream>
#include <string>
#include <unordered_map>
#include <unordered_set>
#include <vector>

namespace fuzzer {
static const char *kFunctionsTxt = "functions.txt";

bool BlockCoverage::AppendCoverage(const std::string &S) {
  std::stringstream SS(S);
  return AppendCoverage(SS);
}

// Coverage lines have this form:
// CN X Y Z T
// where N is the number of the function, T is the total number of instrumented
// BBs, and X,Y,Z, if present, are the indecies of covered BB.
// BB #0, which is the entry block, is not explicitly listed.
bool BlockCoverage::AppendCoverage(std::istream &IN) {
  std::string L;
  while (std::getline(IN, L, '\n')) {
    if (L.empty())
      continue;
    std::stringstream SS(L.c_str() + 1);
    size_t FunctionId  = 0;
    SS >> FunctionId;
    if (L[0] == 'F') {
      FunctionsWithDFT.insert(FunctionId);
      continue;
    }
    if (L[0] != 'C') continue;
    Vector<uint32_t> CoveredBlocks;
    while (true) {
      uint32_t BB = 0;
      SS >> BB;
      if (!SS) break;
      CoveredBlocks.push_back(BB);
    }
    if (CoveredBlocks.empty()) return false;
    uint32_t NumBlocks = CoveredBlocks.back();
    CoveredBlocks.pop_back();
    for (auto BB : CoveredBlocks)
      if (BB >= NumBlocks) return false;
    auto It = Functions.find(FunctionId);
    auto &Counters =
        It == Functions.end()
            ? Functions.insert({FunctionId, Vector<uint32_t>(NumBlocks)})
                  .first->second
            : It->second;

    if (Counters.size() != NumBlocks) return false;  // wrong number of blocks.

    Counters[0]++;
    for (auto BB : CoveredBlocks)
      Counters[BB]++;
  }
  return true;
}

// Assign weights to each function.
// General principles:
//   * any uncovered function gets weight 0.
//   * a function with lots of uncovered blocks gets bigger weight.
//   * a function with a less frequently executed code gets bigger weight.
Vector<double> BlockCoverage::FunctionWeights(size_t NumFunctions) const {
  Vector<double> Res(NumFunctions);
  for (auto It : Functions) {
    auto FunctionID = It.first;
    auto Counters = It.second;
    assert(FunctionID < NumFunctions);
    auto &Weight = Res[FunctionID];
    // Give higher weight if the function has a DFT.
    Weight = FunctionsWithDFT.count(FunctionID) ? 1000. : 1;
    // Give higher weight to functions with less frequently seen basic blocks.
    Weight /= SmallestNonZeroCounter(Counters);
    // Give higher weight to functions with the most uncovered basic blocks.
    Weight *= NumberOfUncoveredBlocks(Counters) + 1;
  }
  return Res;
}

void DataFlowTrace::ReadCoverage(const std::string &DirPath) {
  Vector<SizedFile> Files;
  GetSizedFilesFromDir(DirPath, &Files);
  for (auto &SF : Files) {
    auto Name = Basename(SF.File);
    if (Name == kFunctionsTxt) continue;
    if (!CorporaHashes.count(Name)) continue;
    std::ifstream IF(SF.File);
    Coverage.AppendCoverage(IF);
  }
}

static void DFTStringAppendToVector(Vector<uint8_t> *DFT,
                                    const std::string &DFTString) {
  assert(DFT->size() == DFTString.size());
  for (size_t I = 0, Len = DFT->size(); I < Len; I++)
    (*DFT)[I] = DFTString[I] == '1';
}

// converts a string of '0' and '1' into a Vector<uint8_t>
static Vector<uint8_t> DFTStringToVector(const std::string &DFTString) {
  Vector<uint8_t> DFT(DFTString.size());
  DFTStringAppendToVector(&DFT, DFTString);
  return DFT;
}

static bool ParseError(const char *Err, const std::string &Line) {
  Printf("DataFlowTrace: parse error: %s: Line: %s\n", Err, Line.c_str());
  return false;
}

// TODO(metzman): replace std::string with std::string_view for
// better performance. Need to figure our how to use string_view on Windows.
static bool ParseDFTLine(const std::string &Line, size_t *FunctionNum,
                         std::string *DFTString) {
  if (!Line.empty() && Line[0] != 'F')
    return false; // Ignore coverage.
  size_t SpacePos = Line.find(' ');
  if (SpacePos == std::string::npos)
    return ParseError("no space in the trace line", Line);
  if (Line.empty() || Line[0] != 'F')
    return ParseError("the trace line doesn't start with 'F'", Line);
  *FunctionNum = std::atol(Line.c_str() + 1);
  const char *Beg = Line.c_str() + SpacePos + 1;
  const char *End = Line.c_str() + Line.size();
  assert(Beg < End);
  size_t Len = End - Beg;
  for (size_t I = 0; I < Len; I++) {
    if (Beg[I] != '0' && Beg[I] != '1')
      return ParseError("the trace should contain only 0 or 1", Line);
  }
  *DFTString = Beg;
  return true;
}

bool DataFlowTrace::Init(const std::string &DirPath, std::string *FocusFunction,
                         Vector<SizedFile> &CorporaFiles, Random &Rand) {
  if (DirPath.empty()) return false;
  Printf("INFO: DataFlowTrace: reading from '%s'\n", DirPath.c_str());
  Vector<SizedFile> Files;
  GetSizedFilesFromDir(DirPath, &Files);
  std::string L;
  size_t FocusFuncIdx = SIZE_MAX;
  Vector<std::string> FunctionNames;

  // Collect the hashes of the corpus files.
  for (auto &SF : CorporaFiles)
    CorporaHashes.insert(Hash(FileToVector(SF.File)));

  // Read functions.txt
  std::ifstream IF(DirPlusFile(DirPath, kFunctionsTxt));
  size_t NumFunctions = 0;
  while (std::getline(IF, L, '\n')) {
    FunctionNames.push_back(L);
    NumFunctions++;
    if (*FocusFunction == L)
      FocusFuncIdx = NumFunctions - 1;
  }
  if (!NumFunctions)
    return false;

  if (*FocusFunction == "auto") {
    // AUTOFOCUS works like this:
    // * reads the coverage data from the DFT files.
    // * assigns weights to functions based on coverage.
    // * chooses a random function according to the weights.
    ReadCoverage(DirPath);
    auto Weights = Coverage.FunctionWeights(NumFunctions);
    Vector<double> Intervals(NumFunctions + 1);
    std::iota(Intervals.begin(), Intervals.end(), 0);
    auto Distribution = std::piecewise_constant_distribution<double>(
        Intervals.begin(), Intervals.end(), Weights.begin());
    FocusFuncIdx = static_cast<size_t>(Distribution(Rand));
    *FocusFunction = FunctionNames[FocusFuncIdx];
    assert(FocusFuncIdx < NumFunctions);
    Printf("INFO: AUTOFOCUS: %zd %s\n", FocusFuncIdx,
           FunctionNames[FocusFuncIdx].c_str());
    for (size_t i = 0; i < NumFunctions; i++) {
      if (!Weights[i]) continue;
      Printf("  [%zd] W %g\tBB-tot %u\tBB-cov %u\tEntryFreq %u:\t%s\n", i,
             Weights[i], Coverage.GetNumberOfBlocks(i),
             Coverage.GetNumberOfCoveredBlocks(i), Coverage.GetCounter(i, 0),
             FunctionNames[i].c_str());
    }
  }

  if (!NumFunctions || FocusFuncIdx == SIZE_MAX || Files.size() <= 1)
    return false;

  // Read traces.
  size_t NumTraceFiles = 0;
  size_t NumTracesWithFocusFunction = 0;
  for (auto &SF : Files) {
    auto Name = Basename(SF.File);
    if (Name == kFunctionsTxt) continue;
    if (!CorporaHashes.count(Name)) continue;  // not in the corpus.
    NumTraceFiles++;
    // Printf("=== %s\n", Name.c_str());
    std::ifstream IF(SF.File);
    while (std::getline(IF, L, '\n')) {
      size_t FunctionNum = 0;
      std::string DFTString;
      if (ParseDFTLine(L, &FunctionNum, &DFTString) &&
          FunctionNum == FocusFuncIdx) {
        NumTracesWithFocusFunction++;

        if (FunctionNum >= NumFunctions)
          return ParseError("N is greater than the number of functions", L);
        Traces[Name] = DFTStringToVector(DFTString);
        // Print just a few small traces.
        if (NumTracesWithFocusFunction <= 3 && DFTString.size() <= 16)
          Printf("%s => |%s|\n", Name.c_str(), std::string(DFTString).c_str());
        break; // No need to parse the following lines.
      }
    }
  }
  Printf("INFO: DataFlowTrace: %zd trace files, %zd functions, "
         "%zd traces with focus function\n",
         NumTraceFiles, NumFunctions, NumTracesWithFocusFunction);
  return NumTraceFiles > 0;
}

int CollectDataFlow(const std::string &DFTBinary, const std::string &DirPath,
                    const Vector<SizedFile> &CorporaFiles) {
  Printf("INFO: collecting data flow: bin: %s dir: %s files: %zd\n",
         DFTBinary.c_str(), DirPath.c_str(), CorporaFiles.size());
  static char DFSanEnv[] = "DFSAN_OPTIONS=fast16labels=1:warn_unimplemented=0";
  putenv(DFSanEnv);
  MkDir(DirPath);
  for (auto &F : CorporaFiles) {
    // For every input F we need to collect the data flow and the coverage.
    // Data flow collection may fail if we request too many DFSan tags at once.
    // So, we start from requesting all tags in range [0,Size) and if that fails
    // we then request tags in [0,Size/2) and [Size/2, Size), and so on.
    // Function number => DFT.
    auto OutPath = DirPlusFile(DirPath, Hash(FileToVector(F.File)));
    std::unordered_map<size_t, Vector<uint8_t>> DFTMap;
    std::unordered_set<std::string> Cov;
    Command Cmd;
    Cmd.addArgument(DFTBinary);
    Cmd.addArgument(F.File);
    Cmd.addArgument(OutPath);
    Printf("CMD: %s\n", Cmd.toString().c_str());
    ExecuteCommand(Cmd);
  }
  // Write functions.txt if it's currently empty or doesn't exist.
  auto FunctionsTxtPath = DirPlusFile(DirPath, kFunctionsTxt);
  if (FileToString(FunctionsTxtPath).empty()) {
    Command Cmd;
    Cmd.addArgument(DFTBinary);
    Cmd.setOutputFile(FunctionsTxtPath);
    ExecuteCommand(Cmd);
  }
  return 0;
}

}  // namespace fuzzer