diff options
author | Dimitry Andric <dim@FreeBSD.org> | 2017-12-18 20:11:54 +0000 |
---|---|---|
committer | Dimitry Andric <dim@FreeBSD.org> | 2017-12-18 20:11:54 +0000 |
commit | cdf4f3055e964bb585f294cf77cb549ead82783f (patch) | |
tree | 7bceeca766b3fbe491245bc926a083f78c35d1de /lib/fuzzer | |
parent | 625108084a3ec7c19c7745004c5af0ed7aa417a9 (diff) | |
download | src-test2-cdf4f3055e964bb585f294cf77cb549ead82783f.tar.gz src-test2-cdf4f3055e964bb585f294cf77cb549ead82783f.zip |
Notes
Diffstat (limited to 'lib/fuzzer')
53 files changed, 9276 insertions, 0 deletions
diff --git a/lib/fuzzer/CMakeLists.txt b/lib/fuzzer/CMakeLists.txt new file mode 100644 index 000000000000..9769be52ae01 --- /dev/null +++ b/lib/fuzzer/CMakeLists.txt @@ -0,0 +1,80 @@ +set(LIBFUZZER_SOURCES + FuzzerClangCounters.cpp + FuzzerCrossOver.cpp + FuzzerDriver.cpp + FuzzerExtFunctionsDlsym.cpp + FuzzerExtFunctionsDlsymWin.cpp + FuzzerExtFunctionsWeak.cpp + FuzzerExtraCounters.cpp + FuzzerIO.cpp + FuzzerIOPosix.cpp + FuzzerIOWindows.cpp + FuzzerLoop.cpp + FuzzerMerge.cpp + FuzzerMutate.cpp + FuzzerSHA1.cpp + FuzzerShmemPosix.cpp + FuzzerShmemWindows.cpp + FuzzerTracePC.cpp + FuzzerUtil.cpp + FuzzerUtilDarwin.cpp + FuzzerUtilFuchsia.cpp + FuzzerUtilLinux.cpp + FuzzerUtilPosix.cpp + FuzzerUtilWindows.cpp + ) + +CHECK_CXX_SOURCE_COMPILES(" + static thread_local int blah; + int main() { + return 0; + } + " HAS_THREAD_LOCAL) + +set(LIBFUZZER_CFLAGS ${SANITIZER_COMMON_CFLAGS}) + +append_list_if(COMPILER_RT_HAS_OMIT_FRAME_POINTER_FLAG -fno-omit-frame-pointer LIBFUZZER_CFLAGS) + +if (CMAKE_CXX_FLAGS MATCHES "fsanitize-coverage") + list(APPEND LIBFUZZER_CFLAGS -fno-sanitize-coverage=trace-pc-guard,edge,trace-cmp,indirect-calls,8bit-counters) +endif() + +if(NOT HAS_THREAD_LOCAL) + list(APPEND LIBFUZZER_CFLAGS -Dthread_local=__thread) +endif() + +if(APPLE) + set(FUZZER_SUPPORTED_OS osx) +endif() + +add_compiler_rt_object_libraries(RTfuzzer + OS ${FUZZER_SUPPORTED_OS} + ARCHS ${FUZZER_SUPPORTED_ARCH} + SOURCES ${LIBFUZZER_SOURCES} + CFLAGS ${LIBFUZZER_CFLAGS}) + +add_compiler_rt_object_libraries(RTfuzzer_main + OS ${FUZZER_SUPPORTED_OS} + ARCHS ${FUZZER_SUPPORTED_ARCH} + SOURCES FuzzerMain.cpp + CFLAGS ${LIBFUZZER_CFLAGS}) + +add_compiler_rt_runtime(clang_rt.fuzzer + STATIC + OS ${FUZZER_SUPPORTED_OS} + ARCHS ${FUZZER_SUPPORTED_ARCH} + OBJECT_LIBS RTfuzzer RTfuzzer_main + CFLAGS ${LIBFUZZER_CFLAGS} + PARENT_TARGET fuzzer) + +add_compiler_rt_runtime(clang_rt.fuzzer_no_main + STATIC + OS ${FUZZER_SUPPORTED_OS} + ARCHS ${FUZZER_SUPPORTED_ARCH} + OBJECT_LIBS RTfuzzer + CFLAGS ${LIBFUZZER_CFLAGS} + PARENT_TARGET fuzzer) + +if(COMPILER_RT_INCLUDE_TESTS) + add_subdirectory(tests) +endif() diff --git a/lib/fuzzer/FuzzerClangCounters.cpp b/lib/fuzzer/FuzzerClangCounters.cpp new file mode 100644 index 000000000000..f69e922cf004 --- /dev/null +++ b/lib/fuzzer/FuzzerClangCounters.cpp @@ -0,0 +1,49 @@ +//===- FuzzerExtraCounters.cpp - Extra coverage counters ------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Coverage counters from Clang's SourceBasedCodeCoverage. +//===----------------------------------------------------------------------===// + +// Support for SourceBasedCodeCoverage is experimental: +// * Works only for the main binary, not DSOs yet. +// * Works only on Linux. +// * Does not implement print_pcs/print_coverage yet. +// * Is not fully evaluated for performance and sensitivity. +// We expect large performance drop due to 64-bit counters, +// and *maybe* better sensitivity due to more fine-grained counters. +// Preliminary comparison on a single benchmark (RE2) shows +// a bit worse sensitivity though. + +#include "FuzzerDefs.h" + +#if LIBFUZZER_LINUX +__attribute__((weak)) extern uint64_t __start___llvm_prf_cnts; +__attribute__((weak)) extern uint64_t __stop___llvm_prf_cnts; +namespace fuzzer { +uint64_t *ClangCountersBegin() { return &__start___llvm_prf_cnts; } +uint64_t *ClangCountersEnd() { return &__stop___llvm_prf_cnts; } +} // namespace fuzzer +#else +// TODO: Implement on Mac (if the data shows it's worth it). +//__attribute__((visibility("hidden"))) +//extern uint64_t CountersStart __asm("section$start$__DATA$__llvm_prf_cnts"); +//__attribute__((visibility("hidden"))) +//extern uint64_t CountersEnd __asm("section$end$__DATA$__llvm_prf_cnts"); +namespace fuzzer { +uint64_t *ClangCountersBegin() { return nullptr; } +uint64_t *ClangCountersEnd() { return nullptr; } +} // namespace fuzzer +#endif + +namespace fuzzer { +ATTRIBUTE_NO_SANITIZE_ALL +void ClearClangCounters() { // hand-written memset, don't asan-ify. + for (auto P = ClangCountersBegin(); P < ClangCountersEnd(); P++) + *P = 0; +} +} diff --git a/lib/fuzzer/FuzzerCommand.h b/lib/fuzzer/FuzzerCommand.h new file mode 100644 index 000000000000..c5500ed21fc1 --- /dev/null +++ b/lib/fuzzer/FuzzerCommand.h @@ -0,0 +1,180 @@ +//===- FuzzerCommand.h - Interface representing a process -------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// FuzzerCommand represents a command to run in a subprocess. It allows callers +// to manage command line arguments and output and error streams. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_COMMAND_H +#define LLVM_FUZZER_COMMAND_H + +#include "FuzzerDefs.h" +#include "FuzzerIO.h" + +#include <algorithm> +#include <sstream> +#include <string> +#include <vector> + +namespace fuzzer { + +class Command final { +public: + // This command line flag is used to indicate that the remaining command line + // is immutable, meaning this flag effectively marks the end of the mutable + // argument list. + static inline const char *ignoreRemainingArgs() { + static const char *kIgnoreRemaining = "-ignore_remaining_args=1"; + return kIgnoreRemaining; + } + + Command() : CombinedOutAndErr(false) {} + + explicit Command(const Vector<std::string> &ArgsToAdd) + : Args(ArgsToAdd), CombinedOutAndErr(false) {} + + explicit Command(const Command &Other) + : Args(Other.Args), CombinedOutAndErr(Other.CombinedOutAndErr), + OutputFile(Other.OutputFile) {} + + Command &operator=(const Command &Other) { + Args = Other.Args; + CombinedOutAndErr = Other.CombinedOutAndErr; + OutputFile = Other.OutputFile; + return *this; + } + + ~Command() {} + + // Returns true if the given Arg is present in Args. Only checks up to + // "-ignore_remaining_args=1". + bool hasArgument(const std::string &Arg) const { + auto i = endMutableArgs(); + return std::find(Args.begin(), i, Arg) != i; + } + + // Gets all of the current command line arguments, **including** those after + // "-ignore-remaining-args=1". + const Vector<std::string> &getArguments() const { return Args; } + + // Adds the given argument before "-ignore_remaining_args=1", or at the end + // if that flag isn't present. + void addArgument(const std::string &Arg) { + Args.insert(endMutableArgs(), Arg); + } + + // Adds all given arguments before "-ignore_remaining_args=1", or at the end + // if that flag isn't present. + void addArguments(const Vector<std::string> &ArgsToAdd) { + Args.insert(endMutableArgs(), ArgsToAdd.begin(), ArgsToAdd.end()); + } + + // Removes the given argument from the command argument list. Ignores any + // occurrences after "-ignore_remaining_args=1", if present. + void removeArgument(const std::string &Arg) { + auto i = endMutableArgs(); + Args.erase(std::remove(Args.begin(), i, Arg), i); + } + + // Like hasArgument, but checks for "-[Flag]=...". + bool hasFlag(const std::string &Flag) { + std::string Arg("-" + Flag + "="); + auto IsMatch = [&](const std::string &Other) { + return Arg.compare(0, std::string::npos, Other, 0, Arg.length()) == 0; + }; + return std::any_of(Args.begin(), endMutableArgs(), IsMatch); + } + + // Returns the value of the first instance of a given flag, or an empty string + // if the flag isn't present. Ignores any occurrences after + // "-ignore_remaining_args=1", if present. + std::string getFlagValue(const std::string &Flag) { + std::string Arg("-" + Flag + "="); + auto IsMatch = [&](const std::string &Other) { + return Arg.compare(0, std::string::npos, Other, 0, Arg.length()) == 0; + }; + auto i = endMutableArgs(); + auto j = std::find_if(Args.begin(), i, IsMatch); + std::string result; + if (j != i) { + result = j->substr(Arg.length()); + } + return result; + } + + // Like AddArgument, but adds "-[Flag]=[Value]". + void addFlag(const std::string &Flag, const std::string &Value) { + addArgument("-" + Flag + "=" + Value); + } + + // Like RemoveArgument, but removes "-[Flag]=...". + void removeFlag(const std::string &Flag) { + std::string Arg("-" + Flag + "="); + auto IsMatch = [&](const std::string &Other) { + return Arg.compare(0, std::string::npos, Other, 0, Arg.length()) == 0; + }; + auto i = endMutableArgs(); + Args.erase(std::remove_if(Args.begin(), i, IsMatch), i); + } + + // Returns whether the command's stdout is being written to an output file. + bool hasOutputFile() const { return !OutputFile.empty(); } + + // Returns the currently set output file. + const std::string &getOutputFile() const { return OutputFile; } + + // Configures the command to redirect its output to the name file. + void setOutputFile(const std::string &FileName) { OutputFile = FileName; } + + // Returns whether the command's stderr is redirected to stdout. + bool isOutAndErrCombined() const { return CombinedOutAndErr; } + + // Sets whether to redirect the command's stderr to its stdout. + void combineOutAndErr(bool combine = true) { CombinedOutAndErr = combine; } + + // Returns a string representation of the command. On many systems this will + // be the equivalent command line. + std::string toString() const { + std::stringstream SS; + for (auto arg : getArguments()) + SS << arg << " "; + if (hasOutputFile()) + SS << ">" << getOutputFile() << " "; + if (isOutAndErrCombined()) + SS << "2>&1 "; + std::string result = SS.str(); + if (!result.empty()) + result = result.substr(0, result.length() - 1); + return result; + } + +private: + Command(Command &&Other) = delete; + Command &operator=(Command &&Other) = delete; + + Vector<std::string>::iterator endMutableArgs() { + return std::find(Args.begin(), Args.end(), ignoreRemainingArgs()); + } + + Vector<std::string>::const_iterator endMutableArgs() const { + return std::find(Args.begin(), Args.end(), ignoreRemainingArgs()); + } + + // The command arguments. Args[0] is the command name. + Vector<std::string> Args; + + // True indicates stderr is redirected to stdout. + bool CombinedOutAndErr; + + // If not empty, stdout is redirected to the named file. + std::string OutputFile; +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_COMMAND_H diff --git a/lib/fuzzer/FuzzerCorpus.h b/lib/fuzzer/FuzzerCorpus.h new file mode 100644 index 000000000000..2da929835f45 --- /dev/null +++ b/lib/fuzzer/FuzzerCorpus.h @@ -0,0 +1,302 @@ +//===- FuzzerCorpus.h - Internal header for the Fuzzer ----------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// fuzzer::InputCorpus +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_CORPUS +#define LLVM_FUZZER_CORPUS + +#include "FuzzerDefs.h" +#include "FuzzerIO.h" +#include "FuzzerRandom.h" +#include "FuzzerSHA1.h" +#include "FuzzerTracePC.h" +#include <algorithm> +#include <numeric> +#include <random> +#include <unordered_set> + +namespace fuzzer { + +struct InputInfo { + Unit U; // The actual input data. + uint8_t Sha1[kSHA1NumBytes]; // Checksum. + // Number of features that this input has and no smaller input has. + size_t NumFeatures = 0; + size_t Tmp = 0; // Used by ValidateFeatureSet. + // Stats. + size_t NumExecutedMutations = 0; + size_t NumSuccessfullMutations = 0; + bool MayDeleteFile = false; + bool Reduced = false; + Vector<uint32_t> UniqFeatureSet; + float FeatureFrequencyScore = 1.0; +}; + +class InputCorpus { + static const size_t kFeatureSetSize = 1 << 21; + public: + InputCorpus(const std::string &OutputCorpus) : OutputCorpus(OutputCorpus) { + memset(InputSizesPerFeature, 0, sizeof(InputSizesPerFeature)); + memset(SmallestElementPerFeature, 0, sizeof(SmallestElementPerFeature)); + memset(FeatureFrequency, 0, sizeof(FeatureFrequency)); + } + ~InputCorpus() { + for (auto II : Inputs) + delete II; + } + size_t size() const { return Inputs.size(); } + size_t SizeInBytes() const { + size_t Res = 0; + for (auto II : Inputs) + Res += II->U.size(); + return Res; + } + size_t NumActiveUnits() const { + size_t Res = 0; + for (auto II : Inputs) + Res += !II->U.empty(); + return Res; + } + size_t MaxInputSize() const { + size_t Res = 0; + for (auto II : Inputs) + Res = std::max(Res, II->U.size()); + return Res; + } + bool empty() const { return Inputs.empty(); } + const Unit &operator[] (size_t Idx) const { return Inputs[Idx]->U; } + void AddToCorpus(const Unit &U, size_t NumFeatures, bool MayDeleteFile, + const Vector<uint32_t> &FeatureSet) { + assert(!U.empty()); + if (FeatureDebug) + Printf("ADD_TO_CORPUS %zd NF %zd\n", Inputs.size(), NumFeatures); + Inputs.push_back(new InputInfo()); + InputInfo &II = *Inputs.back(); + II.U = U; + II.NumFeatures = NumFeatures; + II.MayDeleteFile = MayDeleteFile; + II.UniqFeatureSet = FeatureSet; + std::sort(II.UniqFeatureSet.begin(), II.UniqFeatureSet.end()); + ComputeSHA1(U.data(), U.size(), II.Sha1); + Hashes.insert(Sha1ToString(II.Sha1)); + UpdateCorpusDistribution(); + PrintCorpus(); + // ValidateFeatureSet(); + } + + // Debug-only + void PrintUnit(const Unit &U) { + if (!FeatureDebug) return; + for (uint8_t C : U) { + if (C != 'F' && C != 'U' && C != 'Z') + C = '.'; + Printf("%c", C); + } + } + + // Debug-only + void PrintFeatureSet(const Vector<uint32_t> &FeatureSet) { + if (!FeatureDebug) return; + Printf("{"); + for (uint32_t Feature: FeatureSet) + Printf("%u,", Feature); + Printf("}"); + } + + // Debug-only + void PrintCorpus() { + if (!FeatureDebug) return; + Printf("======= CORPUS:\n"); + int i = 0; + for (auto II : Inputs) { + if (std::find(II->U.begin(), II->U.end(), 'F') != II->U.end()) { + Printf("[%2d] ", i); + Printf("%s sz=%zd ", Sha1ToString(II->Sha1).c_str(), II->U.size()); + PrintUnit(II->U); + Printf(" "); + PrintFeatureSet(II->UniqFeatureSet); + Printf("\n"); + } + i++; + } + } + + void Replace(InputInfo *II, const Unit &U) { + assert(II->U.size() > U.size()); + Hashes.erase(Sha1ToString(II->Sha1)); + DeleteFile(*II); + ComputeSHA1(U.data(), U.size(), II->Sha1); + Hashes.insert(Sha1ToString(II->Sha1)); + II->U = U; + II->Reduced = true; + UpdateCorpusDistribution(); + } + + bool HasUnit(const Unit &U) { return Hashes.count(Hash(U)); } + bool HasUnit(const std::string &H) { return Hashes.count(H); } + InputInfo &ChooseUnitToMutate(Random &Rand) { + InputInfo &II = *Inputs[ChooseUnitIdxToMutate(Rand)]; + assert(!II.U.empty()); + return II; + }; + + // Returns an index of random unit from the corpus to mutate. + size_t ChooseUnitIdxToMutate(Random &Rand) { + size_t Idx = static_cast<size_t>(CorpusDistribution(Rand)); + assert(Idx < Inputs.size()); + return Idx; + } + + void PrintStats() { + for (size_t i = 0; i < Inputs.size(); i++) { + const auto &II = *Inputs[i]; + Printf(" [%zd %s]\tsz: %zd\truns: %zd\tsucc: %zd\n", i, + Sha1ToString(II.Sha1).c_str(), II.U.size(), + II.NumExecutedMutations, II.NumSuccessfullMutations); + } + } + + void PrintFeatureSet() { + for (size_t i = 0; i < kFeatureSetSize; i++) { + if(size_t Sz = GetFeature(i)) + Printf("[%zd: id %zd sz%zd] ", i, SmallestElementPerFeature[i], Sz); + } + Printf("\n\t"); + for (size_t i = 0; i < Inputs.size(); i++) + if (size_t N = Inputs[i]->NumFeatures) + Printf(" %zd=>%zd ", i, N); + Printf("\n"); + } + + void DeleteFile(const InputInfo &II) { + if (!OutputCorpus.empty() && II.MayDeleteFile) + RemoveFile(DirPlusFile(OutputCorpus, Sha1ToString(II.Sha1))); + } + + void DeleteInput(size_t Idx) { + InputInfo &II = *Inputs[Idx]; + DeleteFile(II); + Unit().swap(II.U); + if (FeatureDebug) + Printf("EVICTED %zd\n", Idx); + } + + bool AddFeature(size_t Idx, uint32_t NewSize, bool Shrink) { + assert(NewSize); + Idx = Idx % kFeatureSetSize; + uint32_t OldSize = GetFeature(Idx); + if (OldSize == 0 || (Shrink && OldSize > NewSize)) { + if (OldSize > 0) { + size_t OldIdx = SmallestElementPerFeature[Idx]; + InputInfo &II = *Inputs[OldIdx]; + assert(II.NumFeatures > 0); + II.NumFeatures--; + if (II.NumFeatures == 0) + DeleteInput(OldIdx); + } else { + NumAddedFeatures++; + } + NumUpdatedFeatures++; + if (FeatureDebug) + Printf("ADD FEATURE %zd sz %d\n", Idx, NewSize); + SmallestElementPerFeature[Idx] = Inputs.size(); + InputSizesPerFeature[Idx] = NewSize; + return true; + } + return false; + } + + void UpdateFeatureFrequency(size_t Idx) { + FeatureFrequency[Idx % kFeatureSetSize]++; + } + float GetFeatureFrequency(size_t Idx) const { + return FeatureFrequency[Idx % kFeatureSetSize]; + } + void UpdateFeatureFrequencyScore(InputInfo *II) { + const float kMin = 0.01, kMax = 100.; + II->FeatureFrequencyScore = kMin; + for (auto Idx : II->UniqFeatureSet) + II->FeatureFrequencyScore += 1. / (GetFeatureFrequency(Idx) + 1.); + II->FeatureFrequencyScore = Min(II->FeatureFrequencyScore, kMax); + } + + size_t NumFeatures() const { return NumAddedFeatures; } + size_t NumFeatureUpdates() const { return NumUpdatedFeatures; } + +private: + + static const bool FeatureDebug = false; + + size_t GetFeature(size_t Idx) const { return InputSizesPerFeature[Idx]; } + + void ValidateFeatureSet() { + if (FeatureDebug) + PrintFeatureSet(); + for (size_t Idx = 0; Idx < kFeatureSetSize; Idx++) + if (GetFeature(Idx)) + Inputs[SmallestElementPerFeature[Idx]]->Tmp++; + for (auto II: Inputs) { + if (II->Tmp != II->NumFeatures) + Printf("ZZZ %zd %zd\n", II->Tmp, II->NumFeatures); + assert(II->Tmp == II->NumFeatures); + II->Tmp = 0; + } + } + + // Updates the probability distribution for the units in the corpus. + // Must be called whenever the corpus or unit weights are changed. + // + // Hypothesis: units added to the corpus last are more interesting. + // + // Hypothesis: inputs with infrequent features are more interesting. + void UpdateCorpusDistribution() { + size_t N = Inputs.size(); + assert(N); + Intervals.resize(N + 1); + Weights.resize(N); + std::iota(Intervals.begin(), Intervals.end(), 0); + for (size_t i = 0; i < N; i++) + Weights[i] = Inputs[i]->NumFeatures + ? (i + 1) * Inputs[i]->FeatureFrequencyScore + : 0.; + if (FeatureDebug) { + for (size_t i = 0; i < N; i++) + Printf("%zd ", Inputs[i]->NumFeatures); + Printf("NUM\n"); + for (size_t i = 0; i < N; i++) + Printf("%f ", Inputs[i]->FeatureFrequencyScore); + Printf("SCORE\n"); + for (size_t i = 0; i < N; i++) + Printf("%f ", Weights[i]); + Printf("Weights\n"); + } + CorpusDistribution = std::piecewise_constant_distribution<double>( + Intervals.begin(), Intervals.end(), Weights.begin()); + } + std::piecewise_constant_distribution<double> CorpusDistribution; + + Vector<double> Intervals; + Vector<double> Weights; + + std::unordered_set<std::string> Hashes; + Vector<InputInfo*> Inputs; + + size_t NumAddedFeatures = 0; + size_t NumUpdatedFeatures = 0; + uint32_t InputSizesPerFeature[kFeatureSetSize]; + uint32_t SmallestElementPerFeature[kFeatureSetSize]; + float FeatureFrequency[kFeatureSetSize]; + + std::string OutputCorpus; +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_CORPUS diff --git a/lib/fuzzer/FuzzerCrossOver.cpp b/lib/fuzzer/FuzzerCrossOver.cpp new file mode 100644 index 000000000000..8b0fd7d529a8 --- /dev/null +++ b/lib/fuzzer/FuzzerCrossOver.cpp @@ -0,0 +1,52 @@ +//===- FuzzerCrossOver.cpp - Cross over two test inputs -------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Cross over test inputs. +//===----------------------------------------------------------------------===// + +#include "FuzzerDefs.h" +#include "FuzzerMutate.h" +#include "FuzzerRandom.h" +#include <cstring> + +namespace fuzzer { + +// Cross Data1 and Data2, store the result (up to MaxOutSize bytes) in Out. +size_t MutationDispatcher::CrossOver(const uint8_t *Data1, size_t Size1, + const uint8_t *Data2, size_t Size2, + uint8_t *Out, size_t MaxOutSize) { + assert(Size1 || Size2); + MaxOutSize = Rand(MaxOutSize) + 1; + size_t OutPos = 0; + size_t Pos1 = 0; + size_t Pos2 = 0; + size_t *InPos = &Pos1; + size_t InSize = Size1; + const uint8_t *Data = Data1; + bool CurrentlyUsingFirstData = true; + while (OutPos < MaxOutSize && (Pos1 < Size1 || Pos2 < Size2)) { + // Merge a part of Data into Out. + size_t OutSizeLeft = MaxOutSize - OutPos; + if (*InPos < InSize) { + size_t InSizeLeft = InSize - *InPos; + size_t MaxExtraSize = std::min(OutSizeLeft, InSizeLeft); + size_t ExtraSize = Rand(MaxExtraSize) + 1; + memcpy(Out + OutPos, Data + *InPos, ExtraSize); + OutPos += ExtraSize; + (*InPos) += ExtraSize; + } + // Use the other input data on the next iteration. + InPos = CurrentlyUsingFirstData ? &Pos2 : &Pos1; + InSize = CurrentlyUsingFirstData ? Size2 : Size1; + Data = CurrentlyUsingFirstData ? Data2 : Data1; + CurrentlyUsingFirstData = !CurrentlyUsingFirstData; + } + return OutPos; +} + +} // namespace fuzzer diff --git a/lib/fuzzer/FuzzerDefs.h b/lib/fuzzer/FuzzerDefs.h new file mode 100644 index 000000000000..5942efc47a4a --- /dev/null +++ b/lib/fuzzer/FuzzerDefs.h @@ -0,0 +1,167 @@ +//===- FuzzerDefs.h - Internal header for the Fuzzer ------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Basic definitions. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_DEFS_H +#define LLVM_FUZZER_DEFS_H + +#include <cassert> +#include <cstddef> +#include <cstdint> +#include <cstring> +#include <string> +#include <vector> +#include <set> +#include <memory> + +// Platform detection. +#ifdef __linux__ +#define LIBFUZZER_APPLE 0 +#define LIBFUZZER_FUCHSIA 0 +#define LIBFUZZER_LINUX 1 +#define LIBFUZZER_NETBSD 0 +#define LIBFUZZER_WINDOWS 0 +#elif __APPLE__ +#define LIBFUZZER_APPLE 1 +#define LIBFUZZER_FUCHSIA 0 +#define LIBFUZZER_LINUX 0 +#define LIBFUZZER_NETBSD 0 +#define LIBFUZZER_WINDOWS 0 +#elif __NetBSD__ +#define LIBFUZZER_APPLE 0 +#define LIBFUZZER_FUCHSIA 0 +#define LIBFUZZER_LINUX 0 +#define LIBFUZZER_NETBSD 1 +#define LIBFUZZER_WINDOWS 0 +#elif _WIN32 +#define LIBFUZZER_APPLE 0 +#define LIBFUZZER_FUCHSIA 0 +#define LIBFUZZER_LINUX 0 +#define LIBFUZZER_NETBSD 0 +#define LIBFUZZER_WINDOWS 1 +#elif __Fuchsia__ +#define LIBFUZZER_APPLE 0 +#define LIBFUZZER_FUCHSIA 1 +#define LIBFUZZER_LINUX 0 +#define LIBFUZZER_NETBSD 0 +#define LIBFUZZER_WINDOWS 0 +#else +#error "Support for your platform has not been implemented" +#endif + +#ifndef __has_attribute +# define __has_attribute(x) 0 +#endif + +#define LIBFUZZER_POSIX (LIBFUZZER_APPLE || LIBFUZZER_LINUX || LIBFUZZER_NETBSD) + +#ifdef __x86_64 +# if __has_attribute(target) +# define ATTRIBUTE_TARGET_POPCNT __attribute__((target("popcnt"))) +# else +# define ATTRIBUTE_TARGET_POPCNT +# endif +#else +# define ATTRIBUTE_TARGET_POPCNT +#endif + + +#ifdef __clang__ // avoid gcc warning. +# if __has_attribute(no_sanitize) +# define ATTRIBUTE_NO_SANITIZE_MEMORY __attribute__((no_sanitize("memory"))) +# else +# define ATTRIBUTE_NO_SANITIZE_MEMORY +# endif +# define ALWAYS_INLINE __attribute__((always_inline)) +#else +# define ATTRIBUTE_NO_SANITIZE_MEMORY +# define ALWAYS_INLINE +#endif // __clang__ + +#define ATTRIBUTE_NO_SANITIZE_ADDRESS __attribute__((no_sanitize_address)) + +#if defined(__has_feature) +# if __has_feature(address_sanitizer) +# define ATTRIBUTE_NO_SANITIZE_ALL ATTRIBUTE_NO_SANITIZE_ADDRESS +# elif __has_feature(memory_sanitizer) +# define ATTRIBUTE_NO_SANITIZE_ALL ATTRIBUTE_NO_SANITIZE_MEMORY +# else +# define ATTRIBUTE_NO_SANITIZE_ALL +# endif +#else +# define ATTRIBUTE_NO_SANITIZE_ALL +#endif + +#if LIBFUZZER_WINDOWS +#define ATTRIBUTE_INTERFACE __declspec(dllexport) +#else +#define ATTRIBUTE_INTERFACE __attribute__((visibility("default"))) +#endif + +namespace fuzzer { + +template <class T> T Min(T a, T b) { return a < b ? a : b; } +template <class T> T Max(T a, T b) { return a > b ? a : b; } + +class Random; +class Dictionary; +class DictionaryEntry; +class MutationDispatcher; +struct FuzzingOptions; +class InputCorpus; +struct InputInfo; +struct ExternalFunctions; + +// Global interface to functions that may or may not be available. +extern ExternalFunctions *EF; + +// We are using a custom allocator to give a different symbol name to STL +// containers in order to avoid ODR violations. +template<typename T> + class fuzzer_allocator: public std::allocator<T> { + public: + template<class Other> + struct rebind { typedef fuzzer_allocator<Other> other; }; + }; + +template<typename T> +using Vector = std::vector<T, fuzzer_allocator<T>>; + +template<typename T> +using Set = std::set<T, std::less<T>, fuzzer_allocator<T>>; + +typedef Vector<uint8_t> Unit; +typedef Vector<Unit> UnitVector; +typedef int (*UserCallback)(const uint8_t *Data, size_t Size); + +int FuzzerDriver(int *argc, char ***argv, UserCallback Callback); + +struct ScopedDoingMyOwnMemOrStr { + ScopedDoingMyOwnMemOrStr() { DoingMyOwnMemOrStr++; } + ~ScopedDoingMyOwnMemOrStr() { DoingMyOwnMemOrStr--; } + static int DoingMyOwnMemOrStr; +}; + +inline uint8_t Bswap(uint8_t x) { return x; } +inline uint16_t Bswap(uint16_t x) { return __builtin_bswap16(x); } +inline uint32_t Bswap(uint32_t x) { return __builtin_bswap32(x); } +inline uint64_t Bswap(uint64_t x) { return __builtin_bswap64(x); } + +uint8_t *ExtraCountersBegin(); +uint8_t *ExtraCountersEnd(); +void ClearExtraCounters(); + +uint64_t *ClangCountersBegin(); +uint64_t *ClangCountersEnd(); +void ClearClangCounters(); + +} // namespace fuzzer + +#endif // LLVM_FUZZER_DEFS_H diff --git a/lib/fuzzer/FuzzerDictionary.h b/lib/fuzzer/FuzzerDictionary.h new file mode 100644 index 000000000000..daf7d003ea91 --- /dev/null +++ b/lib/fuzzer/FuzzerDictionary.h @@ -0,0 +1,127 @@ +//===- FuzzerDictionary.h - Internal header for the Fuzzer ------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// fuzzer::Dictionary +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_DICTIONARY_H +#define LLVM_FUZZER_DICTIONARY_H + +#include "FuzzerDefs.h" +#include "FuzzerIO.h" +#include "FuzzerUtil.h" +#include <algorithm> +#include <limits> + +namespace fuzzer { +// A simple POD sized array of bytes. +template <size_t kMaxSizeT> class FixedWord { +public: + static const size_t kMaxSize = kMaxSizeT; + FixedWord() {} + FixedWord(const uint8_t *B, uint8_t S) { Set(B, S); } + + void Set(const uint8_t *B, uint8_t S) { + assert(S <= kMaxSize); + memcpy(Data, B, S); + Size = S; + } + + bool operator==(const FixedWord<kMaxSize> &w) const { + ScopedDoingMyOwnMemOrStr scoped_doing_my_own_mem_os_str; + return Size == w.Size && 0 == memcmp(Data, w.Data, Size); + } + + bool operator<(const FixedWord<kMaxSize> &w) const { + ScopedDoingMyOwnMemOrStr scoped_doing_my_own_mem_os_str; + if (Size != w.Size) + return Size < w.Size; + return memcmp(Data, w.Data, Size) < 0; + } + + static size_t GetMaxSize() { return kMaxSize; } + const uint8_t *data() const { return Data; } + uint8_t size() const { return Size; } + +private: + uint8_t Size = 0; + uint8_t Data[kMaxSize]; +}; + +typedef FixedWord<64> Word; + +class DictionaryEntry { + public: + DictionaryEntry() {} + DictionaryEntry(Word W) : W(W) {} + DictionaryEntry(Word W, size_t PositionHint) : W(W), PositionHint(PositionHint) {} + const Word &GetW() const { return W; } + + bool HasPositionHint() const { return PositionHint != std::numeric_limits<size_t>::max(); } + size_t GetPositionHint() const { + assert(HasPositionHint()); + return PositionHint; + } + void IncUseCount() { UseCount++; } + void IncSuccessCount() { SuccessCount++; } + size_t GetUseCount() const { return UseCount; } + size_t GetSuccessCount() const {return SuccessCount; } + + void Print(const char *PrintAfter = "\n") { + PrintASCII(W.data(), W.size()); + if (HasPositionHint()) + Printf("@%zd", GetPositionHint()); + Printf("%s", PrintAfter); + } + +private: + Word W; + size_t PositionHint = std::numeric_limits<size_t>::max(); + size_t UseCount = 0; + size_t SuccessCount = 0; +}; + +class Dictionary { + public: + static const size_t kMaxDictSize = 1 << 14; + + bool ContainsWord(const Word &W) const { + return std::any_of(begin(), end(), [&](const DictionaryEntry &DE) { + return DE.GetW() == W; + }); + } + const DictionaryEntry *begin() const { return &DE[0]; } + const DictionaryEntry *end() const { return begin() + Size; } + DictionaryEntry & operator[] (size_t Idx) { + assert(Idx < Size); + return DE[Idx]; + } + void push_back(DictionaryEntry DE) { + if (Size < kMaxDictSize) + this->DE[Size++] = DE; + } + void clear() { Size = 0; } + bool empty() const { return Size == 0; } + size_t size() const { return Size; } + +private: + DictionaryEntry DE[kMaxDictSize]; + size_t Size = 0; +}; + +// Parses one dictionary entry. +// If successfull, write the enty to Unit and returns true, +// otherwise returns false. +bool ParseOneDictionaryEntry(const std::string &Str, Unit *U); +// Parses the dictionary file, fills Units, returns true iff all lines +// were parsed succesfully. +bool ParseDictionaryFile(const std::string &Text, Vector<Unit> *Units); + +} // namespace fuzzer + +#endif // LLVM_FUZZER_DICTIONARY_H diff --git a/lib/fuzzer/FuzzerDriver.cpp b/lib/fuzzer/FuzzerDriver.cpp new file mode 100644 index 000000000000..f6b642daeda7 --- /dev/null +++ b/lib/fuzzer/FuzzerDriver.cpp @@ -0,0 +1,767 @@ +//===- FuzzerDriver.cpp - FuzzerDriver function and flags -----------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// FuzzerDriver and flag parsing. +//===----------------------------------------------------------------------===// + +#include "FuzzerCommand.h" +#include "FuzzerCorpus.h" +#include "FuzzerIO.h" +#include "FuzzerInterface.h" +#include "FuzzerInternal.h" +#include "FuzzerMutate.h" +#include "FuzzerRandom.h" +#include "FuzzerShmem.h" +#include "FuzzerTracePC.h" +#include <algorithm> +#include <atomic> +#include <chrono> +#include <cstdlib> +#include <cstring> +#include <mutex> +#include <string> +#include <thread> + +// This function should be present in the libFuzzer so that the client +// binary can test for its existence. +extern "C" __attribute__((used)) void __libfuzzer_is_present() {} + +namespace fuzzer { + +// Program arguments. +struct FlagDescription { + const char *Name; + const char *Description; + int Default; + int *IntFlag; + const char **StrFlag; + unsigned int *UIntFlag; +}; + +struct { +#define FUZZER_DEPRECATED_FLAG(Name) +#define FUZZER_FLAG_INT(Name, Default, Description) int Name; +#define FUZZER_FLAG_UNSIGNED(Name, Default, Description) unsigned int Name; +#define FUZZER_FLAG_STRING(Name, Description) const char *Name; +#include "FuzzerFlags.def" +#undef FUZZER_DEPRECATED_FLAG +#undef FUZZER_FLAG_INT +#undef FUZZER_FLAG_UNSIGNED +#undef FUZZER_FLAG_STRING +} Flags; + +static const FlagDescription FlagDescriptions [] { +#define FUZZER_DEPRECATED_FLAG(Name) \ + {#Name, "Deprecated; don't use", 0, nullptr, nullptr, nullptr}, +#define FUZZER_FLAG_INT(Name, Default, Description) \ + {#Name, Description, Default, &Flags.Name, nullptr, nullptr}, +#define FUZZER_FLAG_UNSIGNED(Name, Default, Description) \ + {#Name, Description, static_cast<int>(Default), \ + nullptr, nullptr, &Flags.Name}, +#define FUZZER_FLAG_STRING(Name, Description) \ + {#Name, Description, 0, nullptr, &Flags.Name, nullptr}, +#include "FuzzerFlags.def" +#undef FUZZER_DEPRECATED_FLAG +#undef FUZZER_FLAG_INT +#undef FUZZER_FLAG_UNSIGNED +#undef FUZZER_FLAG_STRING +}; + +static const size_t kNumFlags = + sizeof(FlagDescriptions) / sizeof(FlagDescriptions[0]); + +static Vector<std::string> *Inputs; +static std::string *ProgName; + +static void PrintHelp() { + Printf("Usage:\n"); + auto Prog = ProgName->c_str(); + Printf("\nTo run fuzzing pass 0 or more directories.\n"); + Printf("%s [-flag1=val1 [-flag2=val2 ...] ] [dir1 [dir2 ...] ]\n", Prog); + + Printf("\nTo run individual tests without fuzzing pass 1 or more files:\n"); + Printf("%s [-flag1=val1 [-flag2=val2 ...] ] file1 [file2 ...]\n", Prog); + + Printf("\nFlags: (strictly in form -flag=value)\n"); + size_t MaxFlagLen = 0; + for (size_t F = 0; F < kNumFlags; F++) + MaxFlagLen = std::max(strlen(FlagDescriptions[F].Name), MaxFlagLen); + + for (size_t F = 0; F < kNumFlags; F++) { + const auto &D = FlagDescriptions[F]; + if (strstr(D.Description, "internal flag") == D.Description) continue; + Printf(" %s", D.Name); + for (size_t i = 0, n = MaxFlagLen - strlen(D.Name); i < n; i++) + Printf(" "); + Printf("\t"); + Printf("%d\t%s\n", D.Default, D.Description); + } + Printf("\nFlags starting with '--' will be ignored and " + "will be passed verbatim to subprocesses.\n"); +} + +static const char *FlagValue(const char *Param, const char *Name) { + size_t Len = strlen(Name); + if (Param[0] == '-' && strstr(Param + 1, Name) == Param + 1 && + Param[Len + 1] == '=') + return &Param[Len + 2]; + return nullptr; +} + +// Avoid calling stol as it triggers a bug in clang/glibc build. +static long MyStol(const char *Str) { + long Res = 0; + long Sign = 1; + if (*Str == '-') { + Str++; + Sign = -1; + } + for (size_t i = 0; Str[i]; i++) { + char Ch = Str[i]; + if (Ch < '0' || Ch > '9') + return Res; + Res = Res * 10 + (Ch - '0'); + } + return Res * Sign; +} + +static bool ParseOneFlag(const char *Param) { + if (Param[0] != '-') return false; + if (Param[1] == '-') { + static bool PrintedWarning = false; + if (!PrintedWarning) { + PrintedWarning = true; + Printf("INFO: libFuzzer ignores flags that start with '--'\n"); + } + for (size_t F = 0; F < kNumFlags; F++) + if (FlagValue(Param + 1, FlagDescriptions[F].Name)) + Printf("WARNING: did you mean '%s' (single dash)?\n", Param + 1); + return true; + } + for (size_t F = 0; F < kNumFlags; F++) { + const char *Name = FlagDescriptions[F].Name; + const char *Str = FlagValue(Param, Name); + if (Str) { + if (FlagDescriptions[F].IntFlag) { + int Val = MyStol(Str); + *FlagDescriptions[F].IntFlag = Val; + if (Flags.verbosity >= 2) + Printf("Flag: %s %d\n", Name, Val); + return true; + } else if (FlagDescriptions[F].UIntFlag) { + unsigned int Val = std::stoul(Str); + *FlagDescriptions[F].UIntFlag = Val; + if (Flags.verbosity >= 2) + Printf("Flag: %s %u\n", Name, Val); + return true; + } else if (FlagDescriptions[F].StrFlag) { + *FlagDescriptions[F].StrFlag = Str; + if (Flags.verbosity >= 2) + Printf("Flag: %s %s\n", Name, Str); + return true; + } else { // Deprecated flag. + Printf("Flag: %s: deprecated, don't use\n", Name); + return true; + } + } + } + Printf("\n\nWARNING: unrecognized flag '%s'; " + "use -help=1 to list all flags\n\n", Param); + return true; +} + +// We don't use any library to minimize dependencies. +static void ParseFlags(const Vector<std::string> &Args) { + for (size_t F = 0; F < kNumFlags; F++) { + if (FlagDescriptions[F].IntFlag) + *FlagDescriptions[F].IntFlag = FlagDescriptions[F].Default; + if (FlagDescriptions[F].UIntFlag) + *FlagDescriptions[F].UIntFlag = + static_cast<unsigned int>(FlagDescriptions[F].Default); + if (FlagDescriptions[F].StrFlag) + *FlagDescriptions[F].StrFlag = nullptr; + } + Inputs = new Vector<std::string>; + for (size_t A = 1; A < Args.size(); A++) { + if (ParseOneFlag(Args[A].c_str())) { + if (Flags.ignore_remaining_args) + break; + continue; + } + Inputs->push_back(Args[A]); + } +} + +static std::mutex Mu; + +static void PulseThread() { + while (true) { + SleepSeconds(600); + std::lock_guard<std::mutex> Lock(Mu); + Printf("pulse...\n"); + } +} + +static void WorkerThread(const Command &BaseCmd, std::atomic<unsigned> *Counter, + unsigned NumJobs, std::atomic<bool> *HasErrors) { + while (true) { + unsigned C = (*Counter)++; + if (C >= NumJobs) break; + std::string Log = "fuzz-" + std::to_string(C) + ".log"; + Command Cmd(BaseCmd); + Cmd.setOutputFile(Log); + Cmd.combineOutAndErr(); + if (Flags.verbosity) { + std::string CommandLine = Cmd.toString(); + Printf("%s\n", CommandLine.c_str()); + } + int ExitCode = ExecuteCommand(Cmd); + if (ExitCode != 0) + *HasErrors = true; + std::lock_guard<std::mutex> Lock(Mu); + Printf("================== Job %u exited with exit code %d ============\n", + C, ExitCode); + fuzzer::CopyFileToErr(Log); + } +} + +std::string CloneArgsWithoutX(const Vector<std::string> &Args, + const char *X1, const char *X2) { + std::string Cmd; + for (auto &S : Args) { + if (FlagValue(S.c_str(), X1) || FlagValue(S.c_str(), X2)) + continue; + Cmd += S + " "; + } + return Cmd; +} + +static int RunInMultipleProcesses(const Vector<std::string> &Args, + unsigned NumWorkers, unsigned NumJobs) { + std::atomic<unsigned> Counter(0); + std::atomic<bool> HasErrors(false); + Command Cmd(Args); + Cmd.removeFlag("jobs"); + Cmd.removeFlag("workers"); + Vector<std::thread> V; + std::thread Pulse(PulseThread); + Pulse.detach(); + for (unsigned i = 0; i < NumWorkers; i++) + V.push_back(std::thread(WorkerThread, std::ref(Cmd), &Counter, NumJobs, &HasErrors)); + for (auto &T : V) + T.join(); + return HasErrors ? 1 : 0; +} + +static void RssThread(Fuzzer *F, size_t RssLimitMb) { + while (true) { + SleepSeconds(1); + size_t Peak = GetPeakRSSMb(); + if (Peak > RssLimitMb) + F->RssLimitCallback(); + } +} + +static void StartRssThread(Fuzzer *F, size_t RssLimitMb) { + if (!RssLimitMb) return; + std::thread T(RssThread, F, RssLimitMb); + T.detach(); +} + +int RunOneTest(Fuzzer *F, const char *InputFilePath, size_t MaxLen) { + Unit U = FileToVector(InputFilePath); + if (MaxLen && MaxLen < U.size()) + U.resize(MaxLen); + F->ExecuteCallback(U.data(), U.size()); + F->TryDetectingAMemoryLeak(U.data(), U.size(), true); + return 0; +} + +static bool AllInputsAreFiles() { + if (Inputs->empty()) return false; + for (auto &Path : *Inputs) + if (!IsFile(Path)) + return false; + return true; +} + +static std::string GetDedupTokenFromFile(const std::string &Path) { + auto S = FileToString(Path); + auto Beg = S.find("DEDUP_TOKEN:"); + if (Beg == std::string::npos) + return ""; + auto End = S.find('\n', Beg); + if (End == std::string::npos) + return ""; + return S.substr(Beg, End - Beg); +} + +int CleanseCrashInput(const Vector<std::string> &Args, + const FuzzingOptions &Options) { + if (Inputs->size() != 1 || !Flags.exact_artifact_path) { + Printf("ERROR: -cleanse_crash should be given one input file and" + " -exact_artifact_path\n"); + exit(1); + } + std::string InputFilePath = Inputs->at(0); + std::string OutputFilePath = Flags.exact_artifact_path; + Command Cmd(Args); + Cmd.removeFlag("cleanse_crash"); + + assert(Cmd.hasArgument(InputFilePath)); + Cmd.removeArgument(InputFilePath); + + auto LogFilePath = DirPlusFile( + TmpDir(), "libFuzzerTemp." + std::to_string(GetPid()) + ".txt"); + auto TmpFilePath = DirPlusFile( + TmpDir(), "libFuzzerTemp." + std::to_string(GetPid()) + ".repro"); + Cmd.addArgument(TmpFilePath); + Cmd.setOutputFile(LogFilePath); + Cmd.combineOutAndErr(); + + std::string CurrentFilePath = InputFilePath; + auto U = FileToVector(CurrentFilePath); + size_t Size = U.size(); + + const Vector<uint8_t> ReplacementBytes = {' ', 0xff}; + for (int NumAttempts = 0; NumAttempts < 5; NumAttempts++) { + bool Changed = false; + for (size_t Idx = 0; Idx < Size; Idx++) { + Printf("CLEANSE[%d]: Trying to replace byte %zd of %zd\n", NumAttempts, + Idx, Size); + uint8_t OriginalByte = U[Idx]; + if (ReplacementBytes.end() != std::find(ReplacementBytes.begin(), + ReplacementBytes.end(), + OriginalByte)) + continue; + for (auto NewByte : ReplacementBytes) { + U[Idx] = NewByte; + WriteToFile(U, TmpFilePath); + auto ExitCode = ExecuteCommand(Cmd); + RemoveFile(TmpFilePath); + if (!ExitCode) { + U[Idx] = OriginalByte; + } else { + Changed = true; + Printf("CLEANSE: Replaced byte %zd with 0x%x\n", Idx, NewByte); + WriteToFile(U, OutputFilePath); + break; + } + } + } + if (!Changed) break; + } + RemoveFile(LogFilePath); + return 0; +} + +int MinimizeCrashInput(const Vector<std::string> &Args, + const FuzzingOptions &Options) { + if (Inputs->size() != 1) { + Printf("ERROR: -minimize_crash should be given one input file\n"); + exit(1); + } + std::string InputFilePath = Inputs->at(0); + Command BaseCmd(Args); + BaseCmd.removeFlag("minimize_crash"); + BaseCmd.removeFlag("exact_artifact_path"); + assert(BaseCmd.hasArgument(InputFilePath)); + BaseCmd.removeArgument(InputFilePath); + if (Flags.runs <= 0 && Flags.max_total_time == 0) { + Printf("INFO: you need to specify -runs=N or " + "-max_total_time=N with -minimize_crash=1\n" + "INFO: defaulting to -max_total_time=600\n"); + BaseCmd.addFlag("max_total_time", "600"); + } + + auto LogFilePath = DirPlusFile( + TmpDir(), "libFuzzerTemp." + std::to_string(GetPid()) + ".txt"); + BaseCmd.setOutputFile(LogFilePath); + BaseCmd.combineOutAndErr(); + + std::string CurrentFilePath = InputFilePath; + while (true) { + Unit U = FileToVector(CurrentFilePath); + Printf("CRASH_MIN: minimizing crash input: '%s' (%zd bytes)\n", + CurrentFilePath.c_str(), U.size()); + + Command Cmd(BaseCmd); + Cmd.addArgument(CurrentFilePath); + + std::string CommandLine = Cmd.toString(); + Printf("CRASH_MIN: executing: %s\n", CommandLine.c_str()); + int ExitCode = ExecuteCommand(Cmd); + if (ExitCode == 0) { + Printf("ERROR: the input %s did not crash\n", CurrentFilePath.c_str()); + exit(1); + } + Printf("CRASH_MIN: '%s' (%zd bytes) caused a crash. Will try to minimize " + "it further\n", + CurrentFilePath.c_str(), U.size()); + auto DedupToken1 = GetDedupTokenFromFile(LogFilePath); + if (!DedupToken1.empty()) + Printf("CRASH_MIN: DedupToken1: %s\n", DedupToken1.c_str()); + + std::string ArtifactPath = + Flags.exact_artifact_path + ? Flags.exact_artifact_path + : Options.ArtifactPrefix + "minimized-from-" + Hash(U); + Cmd.addFlag("minimize_crash_internal_step", "1"); + Cmd.addFlag("exact_artifact_path", ArtifactPath); + CommandLine = Cmd.toString(); + Printf("CRASH_MIN: executing: %s\n", CommandLine.c_str()); + ExitCode = ExecuteCommand(Cmd); + CopyFileToErr(LogFilePath); + if (ExitCode == 0) { + if (Flags.exact_artifact_path) { + CurrentFilePath = Flags.exact_artifact_path; + WriteToFile(U, CurrentFilePath); + } + Printf("CRASH_MIN: failed to minimize beyond %s (%d bytes), exiting\n", + CurrentFilePath.c_str(), U.size()); + break; + } + auto DedupToken2 = GetDedupTokenFromFile(LogFilePath); + if (!DedupToken2.empty()) + Printf("CRASH_MIN: DedupToken2: %s\n", DedupToken2.c_str()); + + if (DedupToken1 != DedupToken2) { + if (Flags.exact_artifact_path) { + CurrentFilePath = Flags.exact_artifact_path; + WriteToFile(U, CurrentFilePath); + } + Printf("CRASH_MIN: mismatch in dedup tokens" + " (looks like a different bug). Won't minimize further\n"); + break; + } + + CurrentFilePath = ArtifactPath; + Printf("*********************************\n"); + } + RemoveFile(LogFilePath); + return 0; +} + +int MinimizeCrashInputInternalStep(Fuzzer *F, InputCorpus *Corpus) { + assert(Inputs->size() == 1); + std::string InputFilePath = Inputs->at(0); + Unit U = FileToVector(InputFilePath); + Printf("INFO: Starting MinimizeCrashInputInternalStep: %zd\n", U.size()); + if (U.size() < 2) { + Printf("INFO: The input is small enough, exiting\n"); + exit(0); + } + F->SetMaxInputLen(U.size()); + F->SetMaxMutationLen(U.size() - 1); + F->MinimizeCrashLoop(U); + Printf("INFO: Done MinimizeCrashInputInternalStep, no crashes found\n"); + exit(0); + return 0; +} + +int AnalyzeDictionary(Fuzzer *F, const Vector<Unit>& Dict, + UnitVector& Corpus) { + Printf("Started dictionary minimization (up to %d tests)\n", + Dict.size() * Corpus.size() * 2); + + // Scores and usage count for each dictionary unit. + Vector<int> Scores(Dict.size()); + Vector<int> Usages(Dict.size()); + + Vector<size_t> InitialFeatures; + Vector<size_t> ModifiedFeatures; + for (auto &C : Corpus) { + // Get coverage for the testcase without modifications. + F->ExecuteCallback(C.data(), C.size()); + InitialFeatures.clear(); + TPC.CollectFeatures([&](size_t Feature) { + InitialFeatures.push_back(Feature); + }); + + for (size_t i = 0; i < Dict.size(); ++i) { + Vector<uint8_t> Data = C; + auto StartPos = std::search(Data.begin(), Data.end(), + Dict[i].begin(), Dict[i].end()); + // Skip dictionary unit, if the testcase does not contain it. + if (StartPos == Data.end()) + continue; + + ++Usages[i]; + while (StartPos != Data.end()) { + // Replace all occurrences of dictionary unit in the testcase. + auto EndPos = StartPos + Dict[i].size(); + for (auto It = StartPos; It != EndPos; ++It) + *It ^= 0xFF; + + StartPos = std::search(EndPos, Data.end(), + Dict[i].begin(), Dict[i].end()); + } + + // Get coverage for testcase with masked occurrences of dictionary unit. + F->ExecuteCallback(Data.data(), Data.size()); + ModifiedFeatures.clear(); + TPC.CollectFeatures([&](size_t Feature) { + ModifiedFeatures.push_back(Feature); + }); + + if (InitialFeatures == ModifiedFeatures) + --Scores[i]; + else + Scores[i] += 2; + } + } + + Printf("###### Useless dictionary elements. ######\n"); + for (size_t i = 0; i < Dict.size(); ++i) { + // Dictionary units with positive score are treated as useful ones. + if (Scores[i] > 0) + continue; + + Printf("\""); + PrintASCII(Dict[i].data(), Dict[i].size(), "\""); + Printf(" # Score: %d, Used: %d\n", Scores[i], Usages[i]); + } + Printf("###### End of useless dictionary elements. ######\n"); + return 0; +} + +int FuzzerDriver(int *argc, char ***argv, UserCallback Callback) { + using namespace fuzzer; + assert(argc && argv && "Argument pointers cannot be nullptr"); + std::string Argv0((*argv)[0]); + EF = new ExternalFunctions(); + if (EF->LLVMFuzzerInitialize) + EF->LLVMFuzzerInitialize(argc, argv); + const Vector<std::string> Args(*argv, *argv + *argc); + assert(!Args.empty()); + ProgName = new std::string(Args[0]); + if (Argv0 != *ProgName) { + Printf("ERROR: argv[0] has been modified in LLVMFuzzerInitialize\n"); + exit(1); + } + ParseFlags(Args); + if (Flags.help) { + PrintHelp(); + return 0; + } + + if (Flags.close_fd_mask & 2) + DupAndCloseStderr(); + if (Flags.close_fd_mask & 1) + CloseStdout(); + + if (Flags.jobs > 0 && Flags.workers == 0) { + Flags.workers = std::min(NumberOfCpuCores() / 2, Flags.jobs); + if (Flags.workers > 1) + Printf("Running %u workers\n", Flags.workers); + } + + if (Flags.workers > 0 && Flags.jobs > 0) + return RunInMultipleProcesses(Args, Flags.workers, Flags.jobs); + + FuzzingOptions Options; + Options.Verbosity = Flags.verbosity; + Options.MaxLen = Flags.max_len; + Options.ExperimentalLenControl = Flags.experimental_len_control; + Options.UnitTimeoutSec = Flags.timeout; + Options.ErrorExitCode = Flags.error_exitcode; + Options.TimeoutExitCode = Flags.timeout_exitcode; + Options.MaxTotalTimeSec = Flags.max_total_time; + Options.DoCrossOver = Flags.cross_over; + Options.MutateDepth = Flags.mutate_depth; + Options.ReduceDepth = Flags.reduce_depth; + Options.UseCounters = Flags.use_counters; + Options.UseMemmem = Flags.use_memmem; + Options.UseCmp = Flags.use_cmp; + Options.UseValueProfile = Flags.use_value_profile; + Options.Shrink = Flags.shrink; + Options.ReduceInputs = Flags.reduce_inputs; + Options.ShuffleAtStartUp = Flags.shuffle; + Options.PreferSmall = Flags.prefer_small; + Options.ReloadIntervalSec = Flags.reload; + Options.OnlyASCII = Flags.only_ascii; + Options.DetectLeaks = Flags.detect_leaks; + Options.PurgeAllocatorIntervalSec = Flags.purge_allocator_interval; + Options.TraceMalloc = Flags.trace_malloc; + Options.RssLimitMb = Flags.rss_limit_mb; + Options.MallocLimitMb = Flags.malloc_limit_mb; + if (!Options.MallocLimitMb) + Options.MallocLimitMb = Options.RssLimitMb; + if (Flags.runs >= 0) + Options.MaxNumberOfRuns = Flags.runs; + if (!Inputs->empty() && !Flags.minimize_crash_internal_step) + Options.OutputCorpus = (*Inputs)[0]; + Options.ReportSlowUnits = Flags.report_slow_units; + if (Flags.artifact_prefix) + Options.ArtifactPrefix = Flags.artifact_prefix; + if (Flags.exact_artifact_path) + Options.ExactArtifactPath = Flags.exact_artifact_path; + Vector<Unit> Dictionary; + if (Flags.dict) + if (!ParseDictionaryFile(FileToString(Flags.dict), &Dictionary)) + return 1; + if (Flags.verbosity > 0 && !Dictionary.empty()) + Printf("Dictionary: %zd entries\n", Dictionary.size()); + bool DoPlainRun = AllInputsAreFiles(); + Options.SaveArtifacts = + !DoPlainRun || Flags.minimize_crash_internal_step; + Options.PrintNewCovPcs = Flags.print_pcs; + Options.PrintNewCovFuncs = Flags.print_funcs; + Options.PrintFinalStats = Flags.print_final_stats; + Options.PrintCorpusStats = Flags.print_corpus_stats; + Options.PrintCoverage = Flags.print_coverage; + Options.DumpCoverage = Flags.dump_coverage; + Options.UseClangCoverage = Flags.use_clang_coverage; + Options.UseFeatureFrequency = Flags.use_feature_frequency; + if (Flags.exit_on_src_pos) + Options.ExitOnSrcPos = Flags.exit_on_src_pos; + if (Flags.exit_on_item) + Options.ExitOnItem = Flags.exit_on_item; + + unsigned Seed = Flags.seed; + // Initialize Seed. + if (Seed == 0) + Seed = + std::chrono::system_clock::now().time_since_epoch().count() + GetPid(); + if (Flags.verbosity) + Printf("INFO: Seed: %u\n", Seed); + + Random Rand(Seed); + auto *MD = new MutationDispatcher(Rand, Options); + auto *Corpus = new InputCorpus(Options.OutputCorpus); + auto *F = new Fuzzer(Callback, *Corpus, *MD, Options); + + for (auto &U: Dictionary) + if (U.size() <= Word::GetMaxSize()) + MD->AddWordToManualDictionary(Word(U.data(), U.size())); + + StartRssThread(F, Flags.rss_limit_mb); + + Options.HandleAbrt = Flags.handle_abrt; + Options.HandleBus = Flags.handle_bus; + Options.HandleFpe = Flags.handle_fpe; + Options.HandleIll = Flags.handle_ill; + Options.HandleInt = Flags.handle_int; + Options.HandleSegv = Flags.handle_segv; + Options.HandleTerm = Flags.handle_term; + Options.HandleXfsz = Flags.handle_xfsz; + Options.HandleUsr1 = Flags.handle_usr1; + Options.HandleUsr2 = Flags.handle_usr2; + SetSignalHandler(Options); + + std::atexit(Fuzzer::StaticExitCallback); + + if (Flags.minimize_crash) + return MinimizeCrashInput(Args, Options); + + if (Flags.minimize_crash_internal_step) + return MinimizeCrashInputInternalStep(F, Corpus); + + if (Flags.cleanse_crash) + return CleanseCrashInput(Args, Options); + + if (auto Name = Flags.run_equivalence_server) { + SMR.Destroy(Name); + if (!SMR.Create(Name)) { + Printf("ERROR: can't create shared memory region\n"); + return 1; + } + Printf("INFO: EQUIVALENCE SERVER UP\n"); + while (true) { + SMR.WaitClient(); + size_t Size = SMR.ReadByteArraySize(); + SMR.WriteByteArray(nullptr, 0); + const Unit tmp(SMR.GetByteArray(), SMR.GetByteArray() + Size); + F->ExecuteCallback(tmp.data(), tmp.size()); + SMR.PostServer(); + } + return 0; + } + + if (auto Name = Flags.use_equivalence_server) { + if (!SMR.Open(Name)) { + Printf("ERROR: can't open shared memory region\n"); + return 1; + } + Printf("INFO: EQUIVALENCE CLIENT UP\n"); + } + + if (DoPlainRun) { + Options.SaveArtifacts = false; + int Runs = std::max(1, Flags.runs); + Printf("%s: Running %zd inputs %d time(s) each.\n", ProgName->c_str(), + Inputs->size(), Runs); + for (auto &Path : *Inputs) { + auto StartTime = system_clock::now(); + Printf("Running: %s\n", Path.c_str()); + for (int Iter = 0; Iter < Runs; Iter++) + RunOneTest(F, Path.c_str(), Options.MaxLen); + auto StopTime = system_clock::now(); + auto MS = duration_cast<milliseconds>(StopTime - StartTime).count(); + Printf("Executed %s in %zd ms\n", Path.c_str(), (long)MS); + } + Printf("***\n" + "*** NOTE: fuzzing was not performed, you have only\n" + "*** executed the target code on a fixed set of inputs.\n" + "***\n"); + F->PrintFinalStats(); + exit(0); + } + + if (Flags.merge) { + F->CrashResistantMerge(Args, *Inputs, + Flags.load_coverage_summary, + Flags.save_coverage_summary, + Flags.merge_control_file); + exit(0); + } + + if (Flags.merge_inner) { + const size_t kDefaultMaxMergeLen = 1 << 20; + if (Options.MaxLen == 0) + F->SetMaxInputLen(kDefaultMaxMergeLen); + assert(Flags.merge_control_file); + F->CrashResistantMergeInternalStep(Flags.merge_control_file); + exit(0); + } + + if (Flags.analyze_dict) { + size_t MaxLen = INT_MAX; // Large max length. + UnitVector InitialCorpus; + for (auto &Inp : *Inputs) { + Printf("Loading corpus dir: %s\n", Inp.c_str()); + ReadDirToVectorOfUnits(Inp.c_str(), &InitialCorpus, nullptr, + MaxLen, /*ExitOnError=*/false); + } + + if (Dictionary.empty() || Inputs->empty()) { + Printf("ERROR: can't analyze dict without dict and corpus provided\n"); + return 1; + } + if (AnalyzeDictionary(F, Dictionary, InitialCorpus)) { + Printf("Dictionary analysis failed\n"); + exit(1); + } + Printf("Dictionary analysis suceeded\n"); + exit(0); + } + + F->Loop(*Inputs); + + if (Flags.verbosity) + Printf("Done %zd runs in %zd second(s)\n", F->getTotalNumberOfRuns(), + F->secondsSinceProcessStartUp()); + F->PrintFinalStats(); + + exit(0); // Don't let F destroy itself. +} + +// Storage for global ExternalFunctions object. +ExternalFunctions *EF = nullptr; + +} // namespace fuzzer diff --git a/lib/fuzzer/FuzzerExtFunctions.def b/lib/fuzzer/FuzzerExtFunctions.def new file mode 100644 index 000000000000..25a655bfd71d --- /dev/null +++ b/lib/fuzzer/FuzzerExtFunctions.def @@ -0,0 +1,47 @@ +//===- FuzzerExtFunctions.def - External functions --------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// This defines the external function pointers that +// ``fuzzer::ExternalFunctions`` should contain and try to initialize. The +// EXT_FUNC macro must be defined at the point of inclusion. The signature of +// the macro is: +// +// EXT_FUNC(<name>, <return_type>, <function_signature>, <warn_if_missing>) +//===----------------------------------------------------------------------===// + +// Optional user functions +EXT_FUNC(LLVMFuzzerInitialize, int, (int *argc, char ***argv), false); +EXT_FUNC(LLVMFuzzerCustomMutator, size_t, + (uint8_t * Data, size_t Size, size_t MaxSize, unsigned int Seed), + false); +EXT_FUNC(LLVMFuzzerCustomCrossOver, size_t, + (const uint8_t * Data1, size_t Size1, + const uint8_t * Data2, size_t Size2, + uint8_t * Out, size_t MaxOutSize, unsigned int Seed), + false); + +// Sanitizer functions +EXT_FUNC(__lsan_enable, void, (), false); +EXT_FUNC(__lsan_disable, void, (), false); +EXT_FUNC(__lsan_do_recoverable_leak_check, int, (), false); +EXT_FUNC(__sanitizer_install_malloc_and_free_hooks, int, + (void (*malloc_hook)(const volatile void *, size_t), + void (*free_hook)(const volatile void *)), + false); +EXT_FUNC(__sanitizer_purge_allocator, void, (), false); +EXT_FUNC(__sanitizer_print_memory_profile, int, (size_t, size_t), false); +EXT_FUNC(__sanitizer_print_stack_trace, void, (), true); +EXT_FUNC(__sanitizer_symbolize_pc, void, + (void *, const char *fmt, char *out_buf, size_t out_buf_size), false); +EXT_FUNC(__sanitizer_get_module_and_offset_for_pc, int, + (void *pc, char *module_path, + size_t module_path_len,void **pc_offset), false); +EXT_FUNC(__sanitizer_set_death_callback, void, (void (*)(void)), true); +EXT_FUNC(__sanitizer_set_report_fd, void, (void*), false); +EXT_FUNC(__sanitizer_dump_coverage, void, (const uintptr_t *, uintptr_t), + false); diff --git a/lib/fuzzer/FuzzerExtFunctions.h b/lib/fuzzer/FuzzerExtFunctions.h new file mode 100644 index 000000000000..2672a385478d --- /dev/null +++ b/lib/fuzzer/FuzzerExtFunctions.h @@ -0,0 +1,35 @@ +//===- FuzzerExtFunctions.h - Interface to external functions ---*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Defines an interface to (possibly optional) functions. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_EXT_FUNCTIONS_H +#define LLVM_FUZZER_EXT_FUNCTIONS_H + +#include <stddef.h> +#include <stdint.h> + +namespace fuzzer { + +struct ExternalFunctions { + // Initialize function pointers. Functions that are not available will be set + // to nullptr. Do not call this constructor before ``main()`` has been + // entered. + ExternalFunctions(); + +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + RETURN_TYPE(*NAME) FUNC_SIG = nullptr + +#include "FuzzerExtFunctions.def" + +#undef EXT_FUNC +}; +} // namespace fuzzer + +#endif diff --git a/lib/fuzzer/FuzzerExtFunctionsDlsym.cpp b/lib/fuzzer/FuzzerExtFunctionsDlsym.cpp new file mode 100644 index 000000000000..06bddd5de38f --- /dev/null +++ b/lib/fuzzer/FuzzerExtFunctionsDlsym.cpp @@ -0,0 +1,52 @@ +//===- FuzzerExtFunctionsDlsym.cpp - Interface to external functions ------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Implementation for operating systems that support dlsym(). We only use it on +// Apple platforms for now. We don't use this approach on Linux because it +// requires that clients of LibFuzzer pass ``--export-dynamic`` to the linker. +// That is a complication we don't wish to expose to clients right now. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_APPLE + +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" +#include <dlfcn.h> + +using namespace fuzzer; + +template <typename T> +static T GetFnPtr(const char *FnName, bool WarnIfMissing) { + dlerror(); // Clear any previous errors. + void *Fn = dlsym(RTLD_DEFAULT, FnName); + if (Fn == nullptr) { + if (WarnIfMissing) { + const char *ErrorMsg = dlerror(); + Printf("WARNING: Failed to find function \"%s\".", FnName); + if (ErrorMsg) + Printf(" Reason %s.", ErrorMsg); + Printf("\n"); + } + } + return reinterpret_cast<T>(Fn); +} + +namespace fuzzer { + +ExternalFunctions::ExternalFunctions() { +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + this->NAME = GetFnPtr<decltype(ExternalFunctions::NAME)>(#NAME, WARN) + +#include "FuzzerExtFunctions.def" + +#undef EXT_FUNC +} + +} // namespace fuzzer + +#endif // LIBFUZZER_APPLE diff --git a/lib/fuzzer/FuzzerExtFunctionsDlsymWin.cpp b/lib/fuzzer/FuzzerExtFunctionsDlsymWin.cpp new file mode 100644 index 000000000000..321b3ec5d414 --- /dev/null +++ b/lib/fuzzer/FuzzerExtFunctionsDlsymWin.cpp @@ -0,0 +1,62 @@ +//===- FuzzerExtFunctionsDlsymWin.cpp - Interface to external functions ---===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Implementation using dynamic loading for Windows. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_WINDOWS + +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" +#include "Windows.h" + +// This must be included after Windows.h. +#include "Psapi.h" + +namespace fuzzer { + +ExternalFunctions::ExternalFunctions() { + HMODULE Modules[1024]; + DWORD BytesNeeded; + HANDLE CurrentProcess = GetCurrentProcess(); + + if (!EnumProcessModules(CurrentProcess, Modules, sizeof(Modules), + &BytesNeeded)) { + Printf("EnumProcessModules failed (error: %d).\n", GetLastError()); + exit(1); + } + + if (sizeof(Modules) < BytesNeeded) { + Printf("Error: the array is not big enough to hold all loaded modules.\n"); + exit(1); + } + + for (size_t i = 0; i < (BytesNeeded / sizeof(HMODULE)); i++) + { + FARPROC Fn; +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + if (this->NAME == nullptr) { \ + Fn = GetProcAddress(Modules[i], #NAME); \ + if (Fn == nullptr) \ + Fn = GetProcAddress(Modules[i], #NAME "__dll"); \ + this->NAME = (decltype(ExternalFunctions::NAME)) Fn; \ + } +#include "FuzzerExtFunctions.def" +#undef EXT_FUNC + } + +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + if (this->NAME == nullptr && WARN) \ + Printf("WARNING: Failed to find function \"%s\".\n", #NAME); +#include "FuzzerExtFunctions.def" +#undef EXT_FUNC +} + +} // namespace fuzzer + +#endif // LIBFUZZER_WINDOWS diff --git a/lib/fuzzer/FuzzerExtFunctionsWeak.cpp b/lib/fuzzer/FuzzerExtFunctionsWeak.cpp new file mode 100644 index 000000000000..5a90723986af --- /dev/null +++ b/lib/fuzzer/FuzzerExtFunctionsWeak.cpp @@ -0,0 +1,54 @@ +//===- FuzzerExtFunctionsWeak.cpp - Interface to external functions -------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Implementation for Linux. This relies on the linker's support for weak +// symbols. We don't use this approach on Apple platforms because it requires +// clients of LibFuzzer to pass ``-U _<symbol_name>`` to the linker to allow +// weak symbols to be undefined. That is a complication we don't want to expose +// to clients right now. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_LINUX || LIBFUZZER_NETBSD || LIBFUZZER_FUCHSIA + +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" + +extern "C" { +// Declare these symbols as weak to allow them to be optionally defined. +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + __attribute__((weak)) RETURN_TYPE NAME FUNC_SIG + +#include "FuzzerExtFunctions.def" + +#undef EXT_FUNC +} + +using namespace fuzzer; + +static void CheckFnPtr(void *FnPtr, const char *FnName, bool WarnIfMissing) { + if (FnPtr == nullptr && WarnIfMissing) { + Printf("WARNING: Failed to find function \"%s\".\n", FnName); + } +} + +namespace fuzzer { + +ExternalFunctions::ExternalFunctions() { +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + this->NAME = ::NAME; \ + CheckFnPtr(reinterpret_cast<void *>(reinterpret_cast<uintptr_t>(::NAME)), \ + #NAME, WARN); + +#include "FuzzerExtFunctions.def" + +#undef EXT_FUNC +} + +} // namespace fuzzer + +#endif // LIBFUZZER_LINUX || LIBFUZZER_NETBSD diff --git a/lib/fuzzer/FuzzerExtFunctionsWeakAlias.cpp b/lib/fuzzer/FuzzerExtFunctionsWeakAlias.cpp new file mode 100644 index 000000000000..e10f7b4dcac2 --- /dev/null +++ b/lib/fuzzer/FuzzerExtFunctionsWeakAlias.cpp @@ -0,0 +1,56 @@ +//===- FuzzerExtFunctionsWeakAlias.cpp - Interface to external functions --===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Implementation using weak aliases. Works for Windows. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_WINDOWS + +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" + +using namespace fuzzer; + +extern "C" { +// Declare these symbols as weak to allow them to be optionally defined. +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + RETURN_TYPE NAME##Def FUNC_SIG { \ + Printf("ERROR: Function \"%s\" not defined.\n", #NAME); \ + exit(1); \ + } \ + RETURN_TYPE NAME FUNC_SIG __attribute__((weak, alias(#NAME "Def"))); + +#include "FuzzerExtFunctions.def" + +#undef EXT_FUNC +} + +template <typename T> +static T *GetFnPtr(T *Fun, T *FunDef, const char *FnName, bool WarnIfMissing) { + if (Fun == FunDef) { + if (WarnIfMissing) + Printf("WARNING: Failed to find function \"%s\".\n", FnName); + return nullptr; + } + return Fun; +} + +namespace fuzzer { + +ExternalFunctions::ExternalFunctions() { +#define EXT_FUNC(NAME, RETURN_TYPE, FUNC_SIG, WARN) \ + this->NAME = GetFnPtr<decltype(::NAME)>(::NAME, ::NAME##Def, #NAME, WARN); + +#include "FuzzerExtFunctions.def" + +#undef EXT_FUNC +} + +} // namespace fuzzer + +#endif // LIBFUZZER_WINDOWS diff --git a/lib/fuzzer/FuzzerExtraCounters.cpp b/lib/fuzzer/FuzzerExtraCounters.cpp new file mode 100644 index 000000000000..0e7a7761bf80 --- /dev/null +++ b/lib/fuzzer/FuzzerExtraCounters.cpp @@ -0,0 +1,41 @@ +//===- FuzzerExtraCounters.cpp - Extra coverage counters ------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Extra coverage counters defined by user code. +//===----------------------------------------------------------------------===// + +#include "FuzzerDefs.h" + +#if LIBFUZZER_LINUX || LIBFUZZER_NETBSD +__attribute__((weak)) extern uint8_t __start___libfuzzer_extra_counters; +__attribute__((weak)) extern uint8_t __stop___libfuzzer_extra_counters; + +namespace fuzzer { +uint8_t *ExtraCountersBegin() { return &__start___libfuzzer_extra_counters; } +uint8_t *ExtraCountersEnd() { return &__stop___libfuzzer_extra_counters; } +ATTRIBUTE_NO_SANITIZE_ALL +void ClearExtraCounters() { // hand-written memset, don't asan-ify. + uintptr_t *Beg = reinterpret_cast<uintptr_t*>(ExtraCountersBegin()); + uintptr_t *End = reinterpret_cast<uintptr_t*>(ExtraCountersEnd()); + for (; Beg < End; Beg++) { + *Beg = 0; + __asm__ __volatile__("" : : : "memory"); + } +} + +} // namespace fuzzer + +#else +// TODO: implement for other platforms. +namespace fuzzer { +uint8_t *ExtraCountersBegin() { return nullptr; } +uint8_t *ExtraCountersEnd() { return nullptr; } +void ClearExtraCounters() {} +} // namespace fuzzer + +#endif diff --git a/lib/fuzzer/FuzzerFlags.def b/lib/fuzzer/FuzzerFlags.def new file mode 100644 index 000000000000..a32102a7da07 --- /dev/null +++ b/lib/fuzzer/FuzzerFlags.def @@ -0,0 +1,150 @@ +//===- FuzzerFlags.def - Run-time flags -------------------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Flags. FUZZER_FLAG_INT/FUZZER_FLAG_STRING macros should be defined at the +// point of inclusion. We are not using any flag parsing library for better +// portability and independence. +//===----------------------------------------------------------------------===// +FUZZER_FLAG_INT(verbosity, 1, "Verbosity level.") +FUZZER_FLAG_UNSIGNED(seed, 0, "Random seed. If 0, seed is generated.") +FUZZER_FLAG_INT(runs, -1, + "Number of individual test runs (-1 for infinite runs).") +FUZZER_FLAG_INT(max_len, 0, "Maximum length of the test input. " + "If 0, libFuzzer tries to guess a good value based on the corpus " + "and reports it. ") +FUZZER_FLAG_INT(experimental_len_control, 0, "experimental flag") +FUZZER_FLAG_INT(cross_over, 1, "If 1, cross over inputs.") +FUZZER_FLAG_INT(mutate_depth, 5, + "Apply this number of consecutive mutations to each input.") +FUZZER_FLAG_INT(reduce_depth, 0, "Experimental/internal. " + "Reduce depth if mutations lose unique features") +FUZZER_FLAG_INT(shuffle, 1, "Shuffle inputs at startup") +FUZZER_FLAG_INT(prefer_small, 1, + "If 1, always prefer smaller inputs during the corpus shuffle.") +FUZZER_FLAG_INT( + timeout, 1200, + "Timeout in seconds (if positive). " + "If one unit runs more than this number of seconds the process will abort.") +FUZZER_FLAG_INT(error_exitcode, 77, "When libFuzzer itself reports a bug " + "this exit code will be used.") +FUZZER_FLAG_INT(timeout_exitcode, 77, "When libFuzzer reports a timeout " + "this exit code will be used.") +FUZZER_FLAG_INT(max_total_time, 0, "If positive, indicates the maximal total " + "time in seconds to run the fuzzer.") +FUZZER_FLAG_INT(help, 0, "Print help.") +FUZZER_FLAG_INT(merge, 0, "If 1, the 2-nd, 3-rd, etc corpora will be " + "merged into the 1-st corpus. Only interesting units will be taken. " + "This flag can be used to minimize a corpus.") +FUZZER_FLAG_STRING(merge_inner, "internal flag") +FUZZER_FLAG_STRING(merge_control_file, + "Specify a control file used for the merge proccess. " + "If a merge process gets killed it tries to leave this file " + "in a state suitable for resuming the merge. " + "By default a temporary file will be used.") +FUZZER_FLAG_STRING(save_coverage_summary, "Experimental:" + " save coverage summary to a given file." + " Used with -merge=1") +FUZZER_FLAG_STRING(load_coverage_summary, "Experimental:" + " load coverage summary from a given file." + " Treat this coverage as belonging to the first corpus. " + " Used with -merge=1") +FUZZER_FLAG_INT(minimize_crash, 0, "If 1, minimizes the provided" + " crash input. Use with -runs=N or -max_total_time=N to limit " + "the number attempts." + " Use with -exact_artifact_path to specify the output." + " Combine with ASAN_OPTIONS=dedup_token_length=3 (or similar) to ensure that" + " the minimized input triggers the same crash." + ) +FUZZER_FLAG_INT(cleanse_crash, 0, "If 1, tries to cleanse the provided" + " crash input to make it contain fewer original bytes." + " Use with -exact_artifact_path to specify the output." + ) +FUZZER_FLAG_INT(minimize_crash_internal_step, 0, "internal flag") +FUZZER_FLAG_INT(use_counters, 1, "Use coverage counters") +FUZZER_FLAG_INT(use_memmem, 1, + "Use hints from intercepting memmem, strstr, etc") +FUZZER_FLAG_INT(use_value_profile, 0, + "Experimental. Use value profile to guide fuzzing.") +FUZZER_FLAG_INT(use_cmp, 1, "Use CMP traces to guide mutations") +FUZZER_FLAG_INT(shrink, 0, "Experimental. Try to shrink corpus inputs.") +FUZZER_FLAG_INT(reduce_inputs, 1, + "Try to reduce the size of inputs while preserving their full feature sets") +FUZZER_FLAG_UNSIGNED(jobs, 0, "Number of jobs to run. If jobs >= 1 we spawn" + " this number of jobs in separate worker processes" + " with stdout/stderr redirected to fuzz-JOB.log.") +FUZZER_FLAG_UNSIGNED(workers, 0, + "Number of simultaneous worker processes to run the jobs." + " If zero, \"min(jobs,NumberOfCpuCores()/2)\" is used.") +FUZZER_FLAG_INT(reload, 1, + "Reload the main corpus every <N> seconds to get new units" + " discovered by other processes. If 0, disabled") +FUZZER_FLAG_INT(report_slow_units, 10, + "Report slowest units if they run for more than this number of seconds.") +FUZZER_FLAG_INT(only_ascii, 0, + "If 1, generate only ASCII (isprint+isspace) inputs.") +FUZZER_FLAG_STRING(dict, "Experimental. Use the dictionary file.") +FUZZER_FLAG_STRING(artifact_prefix, "Write fuzzing artifacts (crash, " + "timeout, or slow inputs) as " + "$(artifact_prefix)file") +FUZZER_FLAG_STRING(exact_artifact_path, + "Write the single artifact on failure (crash, timeout) " + "as $(exact_artifact_path). This overrides -artifact_prefix " + "and will not use checksum in the file name. Do not " + "use the same path for several parallel processes.") +FUZZER_FLAG_INT(print_pcs, 0, "If 1, print out newly covered PCs.") +FUZZER_FLAG_INT(print_funcs, 2, "If >=1, print out at most this number of " + "newly covered functions.") +FUZZER_FLAG_INT(print_final_stats, 0, "If 1, print statistics at exit.") +FUZZER_FLAG_INT(print_corpus_stats, 0, + "If 1, print statistics on corpus elements at exit.") +FUZZER_FLAG_INT(print_coverage, 0, "If 1, print coverage information as text" + " at exit.") +FUZZER_FLAG_INT(dump_coverage, 0, "Deprecated." + " If 1, dump coverage information as a" + " .sancov file at exit.") +FUZZER_FLAG_INT(handle_segv, 1, "If 1, try to intercept SIGSEGV.") +FUZZER_FLAG_INT(handle_bus, 1, "If 1, try to intercept SIGBUS.") +FUZZER_FLAG_INT(handle_abrt, 1, "If 1, try to intercept SIGABRT.") +FUZZER_FLAG_INT(handle_ill, 1, "If 1, try to intercept SIGILL.") +FUZZER_FLAG_INT(handle_fpe, 1, "If 1, try to intercept SIGFPE.") +FUZZER_FLAG_INT(handle_int, 1, "If 1, try to intercept SIGINT.") +FUZZER_FLAG_INT(handle_term, 1, "If 1, try to intercept SIGTERM.") +FUZZER_FLAG_INT(handle_xfsz, 1, "If 1, try to intercept SIGXFSZ.") +FUZZER_FLAG_INT(handle_usr1, 1, "If 1, try to intercept SIGUSR1.") +FUZZER_FLAG_INT(handle_usr2, 1, "If 1, try to intercept SIGUSR2.") +FUZZER_FLAG_INT(close_fd_mask, 0, "If 1, close stdout at startup; " + "if 2, close stderr; if 3, close both. " + "Be careful, this will also close e.g. stderr of asan.") +FUZZER_FLAG_INT(detect_leaks, 1, "If 1, and if LeakSanitizer is enabled " + "try to detect memory leaks during fuzzing (i.e. not only at shut down).") +FUZZER_FLAG_INT(purge_allocator_interval, 1, "Purge allocator caches and " + "quarantines every <N> seconds. When rss_limit_mb is specified (>0), " + "purging starts when RSS exceeds 50% of rss_limit_mb. Pass " + "purge_allocator_interval=-1 to disable this functionality.") +FUZZER_FLAG_INT(trace_malloc, 0, "If >= 1 will print all mallocs/frees. " + "If >= 2 will also print stack traces.") +FUZZER_FLAG_INT(rss_limit_mb, 2048, "If non-zero, the fuzzer will exit upon" + "reaching this limit of RSS memory usage.") +FUZZER_FLAG_INT(malloc_limit_mb, 0, "If non-zero, the fuzzer will exit " + "if the target tries to allocate this number of Mb with one malloc call. " + "If zero (default) same limit as rss_limit_mb is applied.") +FUZZER_FLAG_STRING(exit_on_src_pos, "Exit if a newly found PC originates" + " from the given source location. Example: -exit_on_src_pos=foo.cc:123. " + "Used primarily for testing libFuzzer itself.") +FUZZER_FLAG_STRING(exit_on_item, "Exit if an item with a given sha1 sum" + " was added to the corpus. " + "Used primarily for testing libFuzzer itself.") +FUZZER_FLAG_INT(ignore_remaining_args, 0, "If 1, ignore all arguments passed " + "after this one. Useful for fuzzers that need to do their own " + "argument parsing.") + +FUZZER_FLAG_STRING(run_equivalence_server, "Experimental") +FUZZER_FLAG_STRING(use_equivalence_server, "Experimental") +FUZZER_FLAG_INT(analyze_dict, 0, "Experimental") +FUZZER_FLAG_INT(use_clang_coverage, 0, "Experimental") +FUZZER_FLAG_INT(use_feature_frequency, 0, "Experimental/internal") diff --git a/lib/fuzzer/FuzzerIO.cpp b/lib/fuzzer/FuzzerIO.cpp new file mode 100644 index 000000000000..dac5ec658f1c --- /dev/null +++ b/lib/fuzzer/FuzzerIO.cpp @@ -0,0 +1,129 @@ +//===- FuzzerIO.cpp - IO utils. -------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// IO functions. +//===----------------------------------------------------------------------===// + +#include "FuzzerIO.h" +#include "FuzzerDefs.h" +#include "FuzzerExtFunctions.h" +#include <algorithm> +#include <cstdarg> +#include <fstream> +#include <iterator> +#include <sys/stat.h> +#include <sys/types.h> + +namespace fuzzer { + +static FILE *OutputFile = stderr; + +long GetEpoch(const std::string &Path) { + struct stat St; + if (stat(Path.c_str(), &St)) + return 0; // Can't stat, be conservative. + return St.st_mtime; +} + +Unit FileToVector(const std::string &Path, size_t MaxSize, bool ExitOnError) { + std::ifstream T(Path); + if (ExitOnError && !T) { + Printf("No such directory: %s; exiting\n", Path.c_str()); + exit(1); + } + + T.seekg(0, T.end); + auto EndPos = T.tellg(); + if (EndPos < 0) return {}; + size_t FileLen = EndPos; + if (MaxSize) + FileLen = std::min(FileLen, MaxSize); + + T.seekg(0, T.beg); + Unit Res(FileLen); + T.read(reinterpret_cast<char *>(Res.data()), FileLen); + return Res; +} + +std::string FileToString(const std::string &Path) { + std::ifstream T(Path); + return std::string((std::istreambuf_iterator<char>(T)), + std::istreambuf_iterator<char>()); +} + +void CopyFileToErr(const std::string &Path) { + Printf("%s", FileToString(Path).c_str()); +} + +void WriteToFile(const Unit &U, const std::string &Path) { + // Use raw C interface because this function may be called from a sig handler. + FILE *Out = fopen(Path.c_str(), "w"); + if (!Out) return; + fwrite(U.data(), sizeof(U[0]), U.size(), Out); + fclose(Out); +} + +void ReadDirToVectorOfUnits(const char *Path, Vector<Unit> *V, + long *Epoch, size_t MaxSize, bool ExitOnError) { + long E = Epoch ? *Epoch : 0; + Vector<std::string> Files; + ListFilesInDirRecursive(Path, Epoch, &Files, /*TopDir*/true); + size_t NumLoaded = 0; + for (size_t i = 0; i < Files.size(); i++) { + auto &X = Files[i]; + if (Epoch && GetEpoch(X) < E) continue; + NumLoaded++; + if ((NumLoaded & (NumLoaded - 1)) == 0 && NumLoaded >= 1024) + Printf("Loaded %zd/%zd files from %s\n", NumLoaded, Files.size(), Path); + auto S = FileToVector(X, MaxSize, ExitOnError); + if (!S.empty()) + V->push_back(S); + } +} + + +void GetSizedFilesFromDir(const std::string &Dir, Vector<SizedFile> *V) { + Vector<std::string> Files; + ListFilesInDirRecursive(Dir, 0, &Files, /*TopDir*/true); + for (auto &File : Files) + if (size_t Size = FileSize(File)) + V->push_back({File, Size}); +} + +std::string DirPlusFile(const std::string &DirPath, + const std::string &FileName) { + return DirPath + GetSeparator() + FileName; +} + +void DupAndCloseStderr() { + int OutputFd = DuplicateFile(2); + if (OutputFd > 0) { + FILE *NewOutputFile = OpenFile(OutputFd, "w"); + if (NewOutputFile) { + OutputFile = NewOutputFile; + if (EF->__sanitizer_set_report_fd) + EF->__sanitizer_set_report_fd( + reinterpret_cast<void *>(GetHandleFromFd(OutputFd))); + DiscardOutput(2); + } + } +} + +void CloseStdout() { + DiscardOutput(1); +} + +void Printf(const char *Fmt, ...) { + va_list ap; + va_start(ap, Fmt); + vfprintf(OutputFile, Fmt, ap); + va_end(ap); + fflush(OutputFile); +} + +} // namespace fuzzer diff --git a/lib/fuzzer/FuzzerIO.h b/lib/fuzzer/FuzzerIO.h new file mode 100644 index 000000000000..ea9f0d5a6703 --- /dev/null +++ b/lib/fuzzer/FuzzerIO.h @@ -0,0 +1,85 @@ +//===- FuzzerIO.h - Internal header for IO utils ----------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// IO interface. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_IO_H +#define LLVM_FUZZER_IO_H + +#include "FuzzerDefs.h" + +namespace fuzzer { + +long GetEpoch(const std::string &Path); + +Unit FileToVector(const std::string &Path, size_t MaxSize = 0, + bool ExitOnError = true); + +std::string FileToString(const std::string &Path); + +void CopyFileToErr(const std::string &Path); + +void WriteToFile(const Unit &U, const std::string &Path); + +void ReadDirToVectorOfUnits(const char *Path, Vector<Unit> *V, + long *Epoch, size_t MaxSize, bool ExitOnError); + +// Returns "Dir/FileName" or equivalent for the current OS. +std::string DirPlusFile(const std::string &DirPath, + const std::string &FileName); + +// Returns the name of the dir, similar to the 'dirname' utility. +std::string DirName(const std::string &FileName); + +// Returns path to a TmpDir. +std::string TmpDir(); + +bool IsInterestingCoverageFile(const std::string &FileName); + +void DupAndCloseStderr(); + +void CloseStdout(); + +void Printf(const char *Fmt, ...); + +// Print using raw syscalls, useful when printing at early init stages. +void RawPrint(const char *Str); + +// Platform specific functions: +bool IsFile(const std::string &Path); +size_t FileSize(const std::string &Path); + +void ListFilesInDirRecursive(const std::string &Dir, long *Epoch, + Vector<std::string> *V, bool TopDir); + +struct SizedFile { + std::string File; + size_t Size; + bool operator<(const SizedFile &B) const { return Size < B.Size; } +}; + +void GetSizedFilesFromDir(const std::string &Dir, Vector<SizedFile> *V); + +char GetSeparator(); + +FILE* OpenFile(int Fd, const char *Mode); + +int CloseFile(int Fd); + +int DuplicateFile(int Fd); + +void RemoveFile(const std::string &Path); + +void DiscardOutput(int Fd); + +intptr_t GetHandleFromFd(int fd); + +} // namespace fuzzer + +#endif // LLVM_FUZZER_IO_H diff --git a/lib/fuzzer/FuzzerIOPosix.cpp b/lib/fuzzer/FuzzerIOPosix.cpp new file mode 100644 index 000000000000..2257751c662d --- /dev/null +++ b/lib/fuzzer/FuzzerIOPosix.cpp @@ -0,0 +1,140 @@ +//===- FuzzerIOPosix.cpp - IO utils for Posix. ----------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// IO functions implementation using Posix API. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_POSIX || LIBFUZZER_FUCHSIA + +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" +#include <cstdarg> +#include <cstdio> +#include <dirent.h> +#include <fstream> +#include <iterator> +#include <libgen.h> +#include <sys/stat.h> +#include <sys/types.h> +#include <unistd.h> + +namespace fuzzer { + +bool IsFile(const std::string &Path) { + struct stat St; + if (stat(Path.c_str(), &St)) + return false; + return S_ISREG(St.st_mode); +} + +static bool IsDirectory(const std::string &Path) { + struct stat St; + if (stat(Path.c_str(), &St)) + return false; + return S_ISDIR(St.st_mode); +} + +size_t FileSize(const std::string &Path) { + struct stat St; + if (stat(Path.c_str(), &St)) + return 0; + return St.st_size; +} + +void ListFilesInDirRecursive(const std::string &Dir, long *Epoch, + Vector<std::string> *V, bool TopDir) { + auto E = GetEpoch(Dir); + if (Epoch) + if (E && *Epoch >= E) return; + + DIR *D = opendir(Dir.c_str()); + if (!D) { + Printf("No such directory: %s; exiting\n", Dir.c_str()); + exit(1); + } + while (auto E = readdir(D)) { + std::string Path = DirPlusFile(Dir, E->d_name); + if (E->d_type == DT_REG || E->d_type == DT_LNK || + (E->d_type == DT_UNKNOWN && IsFile(Path))) + V->push_back(Path); + else if ((E->d_type == DT_DIR || + (E->d_type == DT_UNKNOWN && IsDirectory(Path))) && + *E->d_name != '.') + ListFilesInDirRecursive(Path, Epoch, V, false); + } + closedir(D); + if (Epoch && TopDir) + *Epoch = E; +} + +char GetSeparator() { + return '/'; +} + +FILE* OpenFile(int Fd, const char* Mode) { + return fdopen(Fd, Mode); +} + +int CloseFile(int fd) { + return close(fd); +} + +int DuplicateFile(int Fd) { + return dup(Fd); +} + +void RemoveFile(const std::string &Path) { + unlink(Path.c_str()); +} + +void DiscardOutput(int Fd) { + FILE* Temp = fopen("/dev/null", "w"); + if (!Temp) + return; + dup2(fileno(Temp), Fd); + fclose(Temp); +} + +intptr_t GetHandleFromFd(int fd) { + return static_cast<intptr_t>(fd); +} + +std::string DirName(const std::string &FileName) { + char *Tmp = new char[FileName.size() + 1]; + memcpy(Tmp, FileName.c_str(), FileName.size() + 1); + std::string Res = dirname(Tmp); + delete [] Tmp; + return Res; +} + +std::string TmpDir() { + if (auto Env = getenv("TMPDIR")) + return Env; + return "/tmp"; +} + +bool IsInterestingCoverageFile(const std::string &FileName) { + if (FileName.find("compiler-rt/lib/") != std::string::npos) + return false; // sanitizer internal. + if (FileName.find("/usr/lib/") != std::string::npos) + return false; + if (FileName.find("/usr/include/") != std::string::npos) + return false; + if (FileName == "<null>") + return false; + return true; +} + + +void RawPrint(const char *Str) { + write(2, Str, strlen(Str)); +} + +} // namespace fuzzer + +#endif // LIBFUZZER_POSIX diff --git a/lib/fuzzer/FuzzerIOWindows.cpp b/lib/fuzzer/FuzzerIOWindows.cpp new file mode 100644 index 000000000000..74853646b213 --- /dev/null +++ b/lib/fuzzer/FuzzerIOWindows.cpp @@ -0,0 +1,323 @@ +//===- FuzzerIOWindows.cpp - IO utils for Windows. ------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// IO functions implementation for Windows. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_WINDOWS + +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" +#include <cstdarg> +#include <cstdio> +#include <fstream> +#include <io.h> +#include <iterator> +#include <sys/stat.h> +#include <sys/types.h> +#include <windows.h> + +namespace fuzzer { + +static bool IsFile(const std::string &Path, const DWORD &FileAttributes) { + + if (FileAttributes & FILE_ATTRIBUTE_NORMAL) + return true; + + if (FileAttributes & FILE_ATTRIBUTE_DIRECTORY) + return false; + + HANDLE FileHandle( + CreateFileA(Path.c_str(), 0, FILE_SHARE_READ, NULL, OPEN_EXISTING, + FILE_FLAG_BACKUP_SEMANTICS, 0)); + + if (FileHandle == INVALID_HANDLE_VALUE) { + Printf("CreateFileA() failed for \"%s\" (Error code: %lu).\n", Path.c_str(), + GetLastError()); + return false; + } + + DWORD FileType = GetFileType(FileHandle); + + if (FileType == FILE_TYPE_UNKNOWN) { + Printf("GetFileType() failed for \"%s\" (Error code: %lu).\n", Path.c_str(), + GetLastError()); + CloseHandle(FileHandle); + return false; + } + + if (FileType != FILE_TYPE_DISK) { + CloseHandle(FileHandle); + return false; + } + + CloseHandle(FileHandle); + return true; +} + +bool IsFile(const std::string &Path) { + DWORD Att = GetFileAttributesA(Path.c_str()); + + if (Att == INVALID_FILE_ATTRIBUTES) { + Printf("GetFileAttributesA() failed for \"%s\" (Error code: %lu).\n", + Path.c_str(), GetLastError()); + return false; + } + + return IsFile(Path, Att); +} + +void ListFilesInDirRecursive(const std::string &Dir, long *Epoch, + Vector<std::string> *V, bool TopDir) { + auto E = GetEpoch(Dir); + if (Epoch) + if (E && *Epoch >= E) return; + + std::string Path(Dir); + assert(!Path.empty()); + if (Path.back() != '\\') + Path.push_back('\\'); + Path.push_back('*'); + + // Get the first directory entry. + WIN32_FIND_DATAA FindInfo; + HANDLE FindHandle(FindFirstFileA(Path.c_str(), &FindInfo)); + if (FindHandle == INVALID_HANDLE_VALUE) + { + if (GetLastError() == ERROR_FILE_NOT_FOUND) + return; + Printf("No such directory: %s; exiting\n", Dir.c_str()); + exit(1); + } + + do { + std::string FileName = DirPlusFile(Dir, FindInfo.cFileName); + + if (FindInfo.dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY) { + size_t FilenameLen = strlen(FindInfo.cFileName); + if ((FilenameLen == 1 && FindInfo.cFileName[0] == '.') || + (FilenameLen == 2 && FindInfo.cFileName[0] == '.' && + FindInfo.cFileName[1] == '.')) + continue; + + ListFilesInDirRecursive(FileName, Epoch, V, false); + } + else if (IsFile(FileName, FindInfo.dwFileAttributes)) + V->push_back(FileName); + } while (FindNextFileA(FindHandle, &FindInfo)); + + DWORD LastError = GetLastError(); + if (LastError != ERROR_NO_MORE_FILES) + Printf("FindNextFileA failed (Error code: %lu).\n", LastError); + + FindClose(FindHandle); + + if (Epoch && TopDir) + *Epoch = E; +} + +char GetSeparator() { + return '\\'; +} + +FILE* OpenFile(int Fd, const char* Mode) { + return _fdopen(Fd, Mode); +} + +int CloseFile(int Fd) { + return _close(Fd); +} + +int DuplicateFile(int Fd) { + return _dup(Fd); +} + +void RemoveFile(const std::string &Path) { + _unlink(Path.c_str()); +} + +void DiscardOutput(int Fd) { + FILE* Temp = fopen("nul", "w"); + if (!Temp) + return; + _dup2(_fileno(Temp), Fd); + fclose(Temp); +} + +intptr_t GetHandleFromFd(int fd) { + return _get_osfhandle(fd); +} + +static bool IsSeparator(char C) { + return C == '\\' || C == '/'; +} + +// Parse disk designators, like "C:\". If Relative == true, also accepts: "C:". +// Returns number of characters considered if successful. +static size_t ParseDrive(const std::string &FileName, const size_t Offset, + bool Relative = true) { + if (Offset + 1 >= FileName.size() || FileName[Offset + 1] != ':') + return 0; + if (Offset + 2 >= FileName.size() || !IsSeparator(FileName[Offset + 2])) { + if (!Relative) // Accept relative path? + return 0; + else + return 2; + } + return 3; +} + +// Parse a file name, like: SomeFile.txt +// Returns number of characters considered if successful. +static size_t ParseFileName(const std::string &FileName, const size_t Offset) { + size_t Pos = Offset; + const size_t End = FileName.size(); + for(; Pos < End && !IsSeparator(FileName[Pos]); ++Pos) + ; + return Pos - Offset; +} + +// Parse a directory ending in separator, like: `SomeDir\` +// Returns number of characters considered if successful. +static size_t ParseDir(const std::string &FileName, const size_t Offset) { + size_t Pos = Offset; + const size_t End = FileName.size(); + if (Pos >= End || IsSeparator(FileName[Pos])) + return 0; + for(; Pos < End && !IsSeparator(FileName[Pos]); ++Pos) + ; + if (Pos >= End) + return 0; + ++Pos; // Include separator. + return Pos - Offset; +} + +// Parse a servername and share, like: `SomeServer\SomeShare\` +// Returns number of characters considered if successful. +static size_t ParseServerAndShare(const std::string &FileName, + const size_t Offset) { + size_t Pos = Offset, Res; + if (!(Res = ParseDir(FileName, Pos))) + return 0; + Pos += Res; + if (!(Res = ParseDir(FileName, Pos))) + return 0; + Pos += Res; + return Pos - Offset; +} + +// Parse the given Ref string from the position Offset, to exactly match the given +// string Patt. +// Returns number of characters considered if successful. +static size_t ParseCustomString(const std::string &Ref, size_t Offset, + const char *Patt) { + size_t Len = strlen(Patt); + if (Offset + Len > Ref.size()) + return 0; + return Ref.compare(Offset, Len, Patt) == 0 ? Len : 0; +} + +// Parse a location, like: +// \\?\UNC\Server\Share\ \\?\C:\ \\Server\Share\ \ C:\ C: +// Returns number of characters considered if successful. +static size_t ParseLocation(const std::string &FileName) { + size_t Pos = 0, Res; + + if ((Res = ParseCustomString(FileName, Pos, R"(\\?\)"))) { + Pos += Res; + if ((Res = ParseCustomString(FileName, Pos, R"(UNC\)"))) { + Pos += Res; + if ((Res = ParseServerAndShare(FileName, Pos))) + return Pos + Res; + return 0; + } + if ((Res = ParseDrive(FileName, Pos, false))) + return Pos + Res; + return 0; + } + + if (Pos < FileName.size() && IsSeparator(FileName[Pos])) { + ++Pos; + if (Pos < FileName.size() && IsSeparator(FileName[Pos])) { + ++Pos; + if ((Res = ParseServerAndShare(FileName, Pos))) + return Pos + Res; + return 0; + } + return Pos; + } + + if ((Res = ParseDrive(FileName, Pos))) + return Pos + Res; + + return Pos; +} + +std::string DirName(const std::string &FileName) { + size_t LocationLen = ParseLocation(FileName); + size_t DirLen = 0, Res; + while ((Res = ParseDir(FileName, LocationLen + DirLen))) + DirLen += Res; + size_t FileLen = ParseFileName(FileName, LocationLen + DirLen); + + if (LocationLen + DirLen + FileLen != FileName.size()) { + Printf("DirName() failed for \"%s\", invalid path.\n", FileName.c_str()); + exit(1); + } + + if (DirLen) { + --DirLen; // Remove trailing separator. + if (!FileLen) { // Path ended in separator. + assert(DirLen); + // Remove file name from Dir. + while (DirLen && !IsSeparator(FileName[LocationLen + DirLen - 1])) + --DirLen; + if (DirLen) // Remove trailing separator. + --DirLen; + } + } + + if (!LocationLen) { // Relative path. + if (!DirLen) + return "."; + return std::string(".\\").append(FileName, 0, DirLen); + } + + return FileName.substr(0, LocationLen + DirLen); +} + +std::string TmpDir() { + std::string Tmp; + Tmp.resize(MAX_PATH + 1); + DWORD Size = GetTempPathA(Tmp.size(), &Tmp[0]); + if (Size == 0) { + Printf("Couldn't get Tmp path.\n"); + exit(1); + } + Tmp.resize(Size); + return Tmp; +} + +bool IsInterestingCoverageFile(const std::string &FileName) { + if (FileName.find("Program Files") != std::string::npos) + return false; + if (FileName.find("compiler-rt\\lib\\") != std::string::npos) + return false; // sanitizer internal. + if (FileName == "<null>") + return false; + return true; +} + +void RawPrint(const char *Str) { + // Not tested, may or may not work. Fix if needed. + Printf("%s", Str); +} + +} // namespace fuzzer + +#endif // LIBFUZZER_WINDOWS diff --git a/lib/fuzzer/FuzzerInterface.h b/lib/fuzzer/FuzzerInterface.h new file mode 100644 index 000000000000..c2c0a39843c0 --- /dev/null +++ b/lib/fuzzer/FuzzerInterface.h @@ -0,0 +1,67 @@ +//===- FuzzerInterface.h - Interface header for the Fuzzer ------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Define the interface between libFuzzer and the library being tested. +//===----------------------------------------------------------------------===// + +// NOTE: the libFuzzer interface is thin and in the majority of cases +// you should not include this file into your target. In 95% of cases +// all you need is to define the following function in your file: +// extern "C" int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size); + +// WARNING: keep the interface in C. + +#ifndef LLVM_FUZZER_INTERFACE_H +#define LLVM_FUZZER_INTERFACE_H + +#include <stddef.h> +#include <stdint.h> + +#ifdef __cplusplus +extern "C" { +#endif // __cplusplus + +// Mandatory user-provided target function. +// Executes the code under test with [Data, Data+Size) as the input. +// libFuzzer will invoke this function *many* times with different inputs. +// Must return 0. +int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size); + +// Optional user-provided initialization function. +// If provided, this function will be called by libFuzzer once at startup. +// It may read and modify argc/argv. +// Must return 0. +int LLVMFuzzerInitialize(int *argc, char ***argv); + +// Optional user-provided custom mutator. +// Mutates raw data in [Data, Data+Size) inplace. +// Returns the new size, which is not greater than MaxSize. +// Given the same Seed produces the same mutation. +size_t LLVMFuzzerCustomMutator(uint8_t *Data, size_t Size, size_t MaxSize, + unsigned int Seed); + +// Optional user-provided custom cross-over function. +// Combines pieces of Data1 & Data2 together into Out. +// Returns the new size, which is not greater than MaxOutSize. +// Should produce the same mutation given the same Seed. +size_t LLVMFuzzerCustomCrossOver(const uint8_t *Data1, size_t Size1, + const uint8_t *Data2, size_t Size2, + uint8_t *Out, size_t MaxOutSize, + unsigned int Seed); + +// Experimental, may go away in future. +// libFuzzer-provided function to be used inside LLVMFuzzerCustomMutator. +// Mutates raw data in [Data, Data+Size) inplace. +// Returns the new size, which is not greater than MaxSize. +size_t LLVMFuzzerMutate(uint8_t *Data, size_t Size, size_t MaxSize); + +#ifdef __cplusplus +} // extern "C" +#endif // __cplusplus + +#endif // LLVM_FUZZER_INTERFACE_H diff --git a/lib/fuzzer/FuzzerInternal.h b/lib/fuzzer/FuzzerInternal.h new file mode 100644 index 000000000000..2b2638f1f8f2 --- /dev/null +++ b/lib/fuzzer/FuzzerInternal.h @@ -0,0 +1,155 @@ +//===- FuzzerInternal.h - Internal header for the Fuzzer --------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Define the main class fuzzer::Fuzzer and most functions. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_INTERNAL_H +#define LLVM_FUZZER_INTERNAL_H + +#include "FuzzerDefs.h" +#include "FuzzerExtFunctions.h" +#include "FuzzerInterface.h" +#include "FuzzerOptions.h" +#include "FuzzerSHA1.h" +#include "FuzzerValueBitMap.h" +#include <algorithm> +#include <atomic> +#include <chrono> +#include <climits> +#include <cstdlib> +#include <string.h> + +namespace fuzzer { + +using namespace std::chrono; + +class Fuzzer { +public: + + Fuzzer(UserCallback CB, InputCorpus &Corpus, MutationDispatcher &MD, + FuzzingOptions Options); + ~Fuzzer(); + void Loop(const Vector<std::string> &CorpusDirs); + void ReadAndExecuteSeedCorpora(const Vector<std::string> &CorpusDirs); + void MinimizeCrashLoop(const Unit &U); + void RereadOutputCorpus(size_t MaxSize); + + size_t secondsSinceProcessStartUp() { + return duration_cast<seconds>(system_clock::now() - ProcessStartTime) + .count(); + } + + bool TimedOut() { + return Options.MaxTotalTimeSec > 0 && + secondsSinceProcessStartUp() > + static_cast<size_t>(Options.MaxTotalTimeSec); + } + + size_t execPerSec() { + size_t Seconds = secondsSinceProcessStartUp(); + return Seconds ? TotalNumberOfRuns / Seconds : 0; + } + + size_t getTotalNumberOfRuns() { return TotalNumberOfRuns; } + + static void StaticAlarmCallback(); + static void StaticCrashSignalCallback(); + static void StaticExitCallback(); + static void StaticInterruptCallback(); + static void StaticFileSizeExceedCallback(); + static void StaticGracefulExitCallback(); + + void ExecuteCallback(const uint8_t *Data, size_t Size); + bool RunOne(const uint8_t *Data, size_t Size, bool MayDeleteFile = false, + InputInfo *II = nullptr, bool *FoundUniqFeatures = nullptr); + + // Merge Corpora[1:] into Corpora[0]. + void Merge(const Vector<std::string> &Corpora); + void CrashResistantMerge(const Vector<std::string> &Args, + const Vector<std::string> &Corpora, + const char *CoverageSummaryInputPathOrNull, + const char *CoverageSummaryOutputPathOrNull, + const char *MergeControlFilePathOrNull); + void CrashResistantMergeInternalStep(const std::string &ControlFilePath); + MutationDispatcher &GetMD() { return MD; } + void PrintFinalStats(); + void SetMaxInputLen(size_t MaxInputLen); + void SetMaxMutationLen(size_t MaxMutationLen); + void RssLimitCallback(); + + bool InFuzzingThread() const { return IsMyThread; } + size_t GetCurrentUnitInFuzzingThead(const uint8_t **Data) const; + void TryDetectingAMemoryLeak(const uint8_t *Data, size_t Size, + bool DuringInitialCorpusExecution); + + void HandleMalloc(size_t Size); + void AnnounceOutput(const uint8_t *Data, size_t Size); + +private: + void AlarmCallback(); + void CrashCallback(); + void ExitCallback(); + void MaybeExitGracefully(); + void CrashOnOverwrittenData(); + void InterruptCallback(); + void MutateAndTestOne(); + void PurgeAllocator(); + void ReportNewCoverage(InputInfo *II, const Unit &U); + void PrintPulseAndReportSlowInput(const uint8_t *Data, size_t Size); + void WriteToOutputCorpus(const Unit &U); + void WriteUnitToFileWithPrefix(const Unit &U, const char *Prefix); + void PrintStats(const char *Where, const char *End = "\n", size_t Units = 0); + void PrintStatusForNewUnit(const Unit &U, const char *Text); + void CheckExitOnSrcPosOrItem(); + + static void StaticDeathCallback(); + void DumpCurrentUnit(const char *Prefix); + void DeathCallback(); + + void AllocateCurrentUnitData(); + uint8_t *CurrentUnitData = nullptr; + std::atomic<size_t> CurrentUnitSize; + uint8_t BaseSha1[kSHA1NumBytes]; // Checksum of the base unit. + bool RunningCB = false; + + bool GracefulExitRequested = false; + + size_t TotalNumberOfRuns = 0; + size_t NumberOfNewUnitsAdded = 0; + + size_t LastCorpusUpdateRun = 0; + + bool HasMoreMallocsThanFrees = false; + size_t NumberOfLeakDetectionAttempts = 0; + + system_clock::time_point LastAllocatorPurgeAttemptTime = system_clock::now(); + + UserCallback CB; + InputCorpus &Corpus; + MutationDispatcher &MD; + FuzzingOptions Options; + + system_clock::time_point ProcessStartTime = system_clock::now(); + system_clock::time_point UnitStartTime, UnitStopTime; + long TimeOfLongestUnitInSeconds = 0; + long EpochOfLastReadOfOutputCorpus = 0; + + size_t MaxInputLen = 0; + size_t MaxMutationLen = 0; + size_t TmpMaxMutationLen = 0; + + Vector<uint32_t> UniqFeatureSetTmp; + + // Need to know our own thread. + static thread_local bool IsMyThread; +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_INTERNAL_H diff --git a/lib/fuzzer/FuzzerLoop.cpp b/lib/fuzzer/FuzzerLoop.cpp new file mode 100644 index 000000000000..5b451ca122d7 --- /dev/null +++ b/lib/fuzzer/FuzzerLoop.cpp @@ -0,0 +1,839 @@ +//===- FuzzerLoop.cpp - Fuzzer's main loop --------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Fuzzer's main loop. +//===----------------------------------------------------------------------===// + +#include "FuzzerCorpus.h" +#include "FuzzerIO.h" +#include "FuzzerInternal.h" +#include "FuzzerMutate.h" +#include "FuzzerRandom.h" +#include "FuzzerShmem.h" +#include "FuzzerTracePC.h" +#include <algorithm> +#include <cstring> +#include <memory> +#include <mutex> +#include <set> + +#if defined(__has_include) +#if __has_include(<sanitizer / lsan_interface.h>) +#include <sanitizer/lsan_interface.h> +#endif +#endif + +#define NO_SANITIZE_MEMORY +#if defined(__has_feature) +#if __has_feature(memory_sanitizer) +#undef NO_SANITIZE_MEMORY +#define NO_SANITIZE_MEMORY __attribute__((no_sanitize_memory)) +#endif +#endif + +namespace fuzzer { +static const size_t kMaxUnitSizeToPrint = 256; + +thread_local bool Fuzzer::IsMyThread; + +SharedMemoryRegion SMR; + +// Only one Fuzzer per process. +static Fuzzer *F; + +// Leak detection is expensive, so we first check if there were more mallocs +// than frees (using the sanitizer malloc hooks) and only then try to call lsan. +struct MallocFreeTracer { + void Start(int TraceLevel) { + this->TraceLevel = TraceLevel; + if (TraceLevel) + Printf("MallocFreeTracer: START\n"); + Mallocs = 0; + Frees = 0; + } + // Returns true if there were more mallocs than frees. + bool Stop() { + if (TraceLevel) + Printf("MallocFreeTracer: STOP %zd %zd (%s)\n", Mallocs.load(), + Frees.load(), Mallocs == Frees ? "same" : "DIFFERENT"); + bool Result = Mallocs > Frees; + Mallocs = 0; + Frees = 0; + TraceLevel = 0; + return Result; + } + std::atomic<size_t> Mallocs; + std::atomic<size_t> Frees; + int TraceLevel = 0; + + std::recursive_mutex TraceMutex; + bool TraceDisabled = false; +}; + +static MallocFreeTracer AllocTracer; + +// Locks printing and avoids nested hooks triggered from mallocs/frees in +// sanitizer. +class TraceLock { +public: + TraceLock() : Lock(AllocTracer.TraceMutex) { + AllocTracer.TraceDisabled = !AllocTracer.TraceDisabled; + } + ~TraceLock() { AllocTracer.TraceDisabled = !AllocTracer.TraceDisabled; } + + bool IsDisabled() const { + // This is already inverted value. + return !AllocTracer.TraceDisabled; + } + +private: + std::lock_guard<std::recursive_mutex> Lock; +}; + +ATTRIBUTE_NO_SANITIZE_MEMORY +void MallocHook(const volatile void *ptr, size_t size) { + size_t N = AllocTracer.Mallocs++; + F->HandleMalloc(size); + if (int TraceLevel = AllocTracer.TraceLevel) { + TraceLock Lock; + if (Lock.IsDisabled()) + return; + Printf("MALLOC[%zd] %p %zd\n", N, ptr, size); + if (TraceLevel >= 2 && EF) + EF->__sanitizer_print_stack_trace(); + } +} + +ATTRIBUTE_NO_SANITIZE_MEMORY +void FreeHook(const volatile void *ptr) { + size_t N = AllocTracer.Frees++; + if (int TraceLevel = AllocTracer.TraceLevel) { + TraceLock Lock; + if (Lock.IsDisabled()) + return; + Printf("FREE[%zd] %p\n", N, ptr); + if (TraceLevel >= 2 && EF) + EF->__sanitizer_print_stack_trace(); + } +} + +// Crash on a single malloc that exceeds the rss limit. +void Fuzzer::HandleMalloc(size_t Size) { + if (!Options.MallocLimitMb || (Size >> 20) < (size_t)Options.MallocLimitMb) + return; + Printf("==%d== ERROR: libFuzzer: out-of-memory (malloc(%zd))\n", GetPid(), + Size); + Printf(" To change the out-of-memory limit use -rss_limit_mb=<N>\n\n"); + if (EF->__sanitizer_print_stack_trace) + EF->__sanitizer_print_stack_trace(); + DumpCurrentUnit("oom-"); + Printf("SUMMARY: libFuzzer: out-of-memory\n"); + PrintFinalStats(); + _Exit(Options.ErrorExitCode); // Stop right now. +} + +Fuzzer::Fuzzer(UserCallback CB, InputCorpus &Corpus, MutationDispatcher &MD, + FuzzingOptions Options) + : CB(CB), Corpus(Corpus), MD(MD), Options(Options) { + if (EF->__sanitizer_set_death_callback) + EF->__sanitizer_set_death_callback(StaticDeathCallback); + assert(!F); + F = this; + TPC.ResetMaps(); + IsMyThread = true; + if (Options.DetectLeaks && EF->__sanitizer_install_malloc_and_free_hooks) + EF->__sanitizer_install_malloc_and_free_hooks(MallocHook, FreeHook); + TPC.SetUseCounters(Options.UseCounters); + TPC.SetUseValueProfile(Options.UseValueProfile); + TPC.SetUseClangCoverage(Options.UseClangCoverage); + + if (Options.Verbosity) + TPC.PrintModuleInfo(); + if (!Options.OutputCorpus.empty() && Options.ReloadIntervalSec) + EpochOfLastReadOfOutputCorpus = GetEpoch(Options.OutputCorpus); + MaxInputLen = MaxMutationLen = Options.MaxLen; + TmpMaxMutationLen = Max(size_t(4), Corpus.MaxInputSize()); + AllocateCurrentUnitData(); + CurrentUnitSize = 0; + memset(BaseSha1, 0, sizeof(BaseSha1)); +} + +Fuzzer::~Fuzzer() {} + +void Fuzzer::AllocateCurrentUnitData() { + if (CurrentUnitData || MaxInputLen == 0) + return; + CurrentUnitData = new uint8_t[MaxInputLen]; +} + +void Fuzzer::StaticDeathCallback() { + assert(F); + F->DeathCallback(); +} + +void Fuzzer::DumpCurrentUnit(const char *Prefix) { + if (!CurrentUnitData) + return; // Happens when running individual inputs. + MD.PrintMutationSequence(); + Printf("; base unit: %s\n", Sha1ToString(BaseSha1).c_str()); + size_t UnitSize = CurrentUnitSize; + if (UnitSize <= kMaxUnitSizeToPrint) { + PrintHexArray(CurrentUnitData, UnitSize, "\n"); + PrintASCII(CurrentUnitData, UnitSize, "\n"); + } + WriteUnitToFileWithPrefix({CurrentUnitData, CurrentUnitData + UnitSize}, + Prefix); +} + +NO_SANITIZE_MEMORY +void Fuzzer::DeathCallback() { + DumpCurrentUnit("crash-"); + PrintFinalStats(); +} + +void Fuzzer::StaticAlarmCallback() { + assert(F); + F->AlarmCallback(); +} + +void Fuzzer::StaticCrashSignalCallback() { + assert(F); + F->CrashCallback(); +} + +void Fuzzer::StaticExitCallback() { + assert(F); + F->ExitCallback(); +} + +void Fuzzer::StaticInterruptCallback() { + assert(F); + F->InterruptCallback(); +} + +void Fuzzer::StaticGracefulExitCallback() { + assert(F); + F->GracefulExitRequested = true; + Printf("INFO: signal received, trying to exit gracefully\n"); +} + +void Fuzzer::StaticFileSizeExceedCallback() { + Printf("==%lu== ERROR: libFuzzer: file size exceeded\n", GetPid()); + exit(1); +} + +void Fuzzer::CrashCallback() { + Printf("==%lu== ERROR: libFuzzer: deadly signal\n", GetPid()); + if (EF->__sanitizer_print_stack_trace) + EF->__sanitizer_print_stack_trace(); + Printf("NOTE: libFuzzer has rudimentary signal handlers.\n" + " Combine libFuzzer with AddressSanitizer or similar for better " + "crash reports.\n"); + Printf("SUMMARY: libFuzzer: deadly signal\n"); + DumpCurrentUnit("crash-"); + PrintFinalStats(); + _Exit(Options.ErrorExitCode); // Stop right now. +} + +void Fuzzer::ExitCallback() { + if (!RunningCB) + return; // This exit did not come from the user callback + Printf("==%lu== ERROR: libFuzzer: fuzz target exited\n", GetPid()); + if (EF->__sanitizer_print_stack_trace) + EF->__sanitizer_print_stack_trace(); + Printf("SUMMARY: libFuzzer: fuzz target exited\n"); + DumpCurrentUnit("crash-"); + PrintFinalStats(); + _Exit(Options.ErrorExitCode); +} + +void Fuzzer::MaybeExitGracefully() { + if (!GracefulExitRequested) return; + Printf("==%lu== INFO: libFuzzer: exiting as requested\n", GetPid()); + PrintFinalStats(); + _Exit(0); +} + +void Fuzzer::InterruptCallback() { + Printf("==%lu== libFuzzer: run interrupted; exiting\n", GetPid()); + PrintFinalStats(); + _Exit(0); // Stop right now, don't perform any at-exit actions. +} + +NO_SANITIZE_MEMORY +void Fuzzer::AlarmCallback() { + assert(Options.UnitTimeoutSec > 0); + // In Windows Alarm callback is executed by a different thread. +#if !LIBFUZZER_WINDOWS + if (!InFuzzingThread()) + return; +#endif + if (!RunningCB) + return; // We have not started running units yet. + size_t Seconds = + duration_cast<seconds>(system_clock::now() - UnitStartTime).count(); + if (Seconds == 0) + return; + if (Options.Verbosity >= 2) + Printf("AlarmCallback %zd\n", Seconds); + if (Seconds >= (size_t)Options.UnitTimeoutSec) { + Printf("ALARM: working on the last Unit for %zd seconds\n", Seconds); + Printf(" and the timeout value is %d (use -timeout=N to change)\n", + Options.UnitTimeoutSec); + DumpCurrentUnit("timeout-"); + Printf("==%lu== ERROR: libFuzzer: timeout after %d seconds\n", GetPid(), + Seconds); + if (EF->__sanitizer_print_stack_trace) + EF->__sanitizer_print_stack_trace(); + Printf("SUMMARY: libFuzzer: timeout\n"); + PrintFinalStats(); + _Exit(Options.TimeoutExitCode); // Stop right now. + } +} + +void Fuzzer::RssLimitCallback() { + Printf( + "==%lu== ERROR: libFuzzer: out-of-memory (used: %zdMb; limit: %zdMb)\n", + GetPid(), GetPeakRSSMb(), Options.RssLimitMb); + Printf(" To change the out-of-memory limit use -rss_limit_mb=<N>\n\n"); + if (EF->__sanitizer_print_memory_profile) + EF->__sanitizer_print_memory_profile(95, 8); + DumpCurrentUnit("oom-"); + Printf("SUMMARY: libFuzzer: out-of-memory\n"); + PrintFinalStats(); + _Exit(Options.ErrorExitCode); // Stop right now. +} + +void Fuzzer::PrintStats(const char *Where, const char *End, size_t Units) { + size_t ExecPerSec = execPerSec(); + if (!Options.Verbosity) + return; + Printf("#%zd\t%s", TotalNumberOfRuns, Where); + if (size_t N = TPC.GetTotalPCCoverage()) + Printf(" cov: %zd", N); + if (size_t N = Corpus.NumFeatures()) + Printf(" ft: %zd", N); + if (!Corpus.empty()) { + Printf(" corp: %zd", Corpus.NumActiveUnits()); + if (size_t N = Corpus.SizeInBytes()) { + if (N < (1 << 14)) + Printf("/%zdb", N); + else if (N < (1 << 24)) + Printf("/%zdKb", N >> 10); + else + Printf("/%zdMb", N >> 20); + } + } + if (Units) + Printf(" units: %zd", Units); + + Printf(" exec/s: %zd", ExecPerSec); + Printf(" rss: %zdMb", GetPeakRSSMb()); + Printf("%s", End); +} + +void Fuzzer::PrintFinalStats() { + if (Options.PrintCoverage) + TPC.PrintCoverage(); + if (Options.DumpCoverage) + TPC.DumpCoverage(); + if (Options.PrintCorpusStats) + Corpus.PrintStats(); + if (!Options.PrintFinalStats) + return; + size_t ExecPerSec = execPerSec(); + Printf("stat::number_of_executed_units: %zd\n", TotalNumberOfRuns); + Printf("stat::average_exec_per_sec: %zd\n", ExecPerSec); + Printf("stat::new_units_added: %zd\n", NumberOfNewUnitsAdded); + Printf("stat::slowest_unit_time_sec: %zd\n", TimeOfLongestUnitInSeconds); + Printf("stat::peak_rss_mb: %zd\n", GetPeakRSSMb()); +} + +void Fuzzer::SetMaxInputLen(size_t MaxInputLen) { + assert(this->MaxInputLen == 0); // Can only reset MaxInputLen from 0 to non-0. + assert(MaxInputLen); + this->MaxInputLen = MaxInputLen; + this->MaxMutationLen = MaxInputLen; + AllocateCurrentUnitData(); + Printf("INFO: -max_len is not provided; " + "libFuzzer will not generate inputs larger than %zd bytes\n", + MaxInputLen); +} + +void Fuzzer::SetMaxMutationLen(size_t MaxMutationLen) { + assert(MaxMutationLen && MaxMutationLen <= MaxInputLen); + this->MaxMutationLen = MaxMutationLen; +} + +void Fuzzer::CheckExitOnSrcPosOrItem() { + if (!Options.ExitOnSrcPos.empty()) { + static auto *PCsSet = new Set<uintptr_t>; + auto HandlePC = [&](uintptr_t PC) { + if (!PCsSet->insert(PC).second) + return; + std::string Descr = DescribePC("%F %L", PC + 1); + if (Descr.find(Options.ExitOnSrcPos) != std::string::npos) { + Printf("INFO: found line matching '%s', exiting.\n", + Options.ExitOnSrcPos.c_str()); + _Exit(0); + } + }; + TPC.ForEachObservedPC(HandlePC); + } + if (!Options.ExitOnItem.empty()) { + if (Corpus.HasUnit(Options.ExitOnItem)) { + Printf("INFO: found item with checksum '%s', exiting.\n", + Options.ExitOnItem.c_str()); + _Exit(0); + } + } +} + +void Fuzzer::RereadOutputCorpus(size_t MaxSize) { + if (Options.OutputCorpus.empty() || !Options.ReloadIntervalSec) + return; + Vector<Unit> AdditionalCorpus; + ReadDirToVectorOfUnits(Options.OutputCorpus.c_str(), &AdditionalCorpus, + &EpochOfLastReadOfOutputCorpus, MaxSize, + /*ExitOnError*/ false); + if (Options.Verbosity >= 2) + Printf("Reload: read %zd new units.\n", AdditionalCorpus.size()); + bool Reloaded = false; + for (auto &U : AdditionalCorpus) { + if (U.size() > MaxSize) + U.resize(MaxSize); + if (!Corpus.HasUnit(U)) { + if (RunOne(U.data(), U.size())) { + CheckExitOnSrcPosOrItem(); + Reloaded = true; + } + } + } + if (Reloaded) + PrintStats("RELOAD"); +} + +void Fuzzer::PrintPulseAndReportSlowInput(const uint8_t *Data, size_t Size) { + auto TimeOfUnit = + duration_cast<seconds>(UnitStopTime - UnitStartTime).count(); + if (!(TotalNumberOfRuns & (TotalNumberOfRuns - 1)) && + secondsSinceProcessStartUp() >= 2) + PrintStats("pulse "); + if (TimeOfUnit > TimeOfLongestUnitInSeconds * 1.1 && + TimeOfUnit >= Options.ReportSlowUnits) { + TimeOfLongestUnitInSeconds = TimeOfUnit; + Printf("Slowest unit: %zd s:\n", TimeOfLongestUnitInSeconds); + WriteUnitToFileWithPrefix({Data, Data + Size}, "slow-unit-"); + } +} + +bool Fuzzer::RunOne(const uint8_t *Data, size_t Size, bool MayDeleteFile, + InputInfo *II, bool *FoundUniqFeatures) { + if (!Size) + return false; + + ExecuteCallback(Data, Size); + + UniqFeatureSetTmp.clear(); + size_t FoundUniqFeaturesOfII = 0; + size_t NumUpdatesBefore = Corpus.NumFeatureUpdates(); + TPC.CollectFeatures([&](size_t Feature) { + if (Options.UseFeatureFrequency) + Corpus.UpdateFeatureFrequency(Feature); + if (Corpus.AddFeature(Feature, Size, Options.Shrink)) + UniqFeatureSetTmp.push_back(Feature); + if (Options.ReduceInputs && II) + if (std::binary_search(II->UniqFeatureSet.begin(), + II->UniqFeatureSet.end(), Feature)) + FoundUniqFeaturesOfII++; + }); + if (FoundUniqFeatures) + *FoundUniqFeatures = FoundUniqFeaturesOfII; + PrintPulseAndReportSlowInput(Data, Size); + size_t NumNewFeatures = Corpus.NumFeatureUpdates() - NumUpdatesBefore; + if (NumNewFeatures) { + TPC.UpdateObservedPCs(); + Corpus.AddToCorpus({Data, Data + Size}, NumNewFeatures, MayDeleteFile, + UniqFeatureSetTmp); + return true; + } + if (II && FoundUniqFeaturesOfII && + FoundUniqFeaturesOfII == II->UniqFeatureSet.size() && + II->U.size() > Size) { + Corpus.Replace(II, {Data, Data + Size}); + return true; + } + return false; +} + +size_t Fuzzer::GetCurrentUnitInFuzzingThead(const uint8_t **Data) const { + assert(InFuzzingThread()); + *Data = CurrentUnitData; + return CurrentUnitSize; +} + +void Fuzzer::CrashOnOverwrittenData() { + Printf("==%d== ERROR: libFuzzer: fuzz target overwrites it's const input\n", + GetPid()); + DumpCurrentUnit("crash-"); + Printf("SUMMARY: libFuzzer: out-of-memory\n"); + _Exit(Options.ErrorExitCode); // Stop right now. +} + +// Compare two arrays, but not all bytes if the arrays are large. +static bool LooseMemeq(const uint8_t *A, const uint8_t *B, size_t Size) { + const size_t Limit = 64; + if (Size <= 64) + return !memcmp(A, B, Size); + // Compare first and last Limit/2 bytes. + return !memcmp(A, B, Limit / 2) && + !memcmp(A + Size - Limit / 2, B + Size - Limit / 2, Limit / 2); +} + +void Fuzzer::ExecuteCallback(const uint8_t *Data, size_t Size) { + TPC.RecordInitialStack(); + TotalNumberOfRuns++; + assert(InFuzzingThread()); + if (SMR.IsClient()) + SMR.WriteByteArray(Data, Size); + // We copy the contents of Unit into a separate heap buffer + // so that we reliably find buffer overflows in it. + uint8_t *DataCopy = new uint8_t[Size]; + memcpy(DataCopy, Data, Size); + if (CurrentUnitData && CurrentUnitData != Data) + memcpy(CurrentUnitData, Data, Size); + CurrentUnitSize = Size; + AllocTracer.Start(Options.TraceMalloc); + UnitStartTime = system_clock::now(); + TPC.ResetMaps(); + RunningCB = true; + int Res = CB(DataCopy, Size); + RunningCB = false; + UnitStopTime = system_clock::now(); + (void)Res; + assert(Res == 0); + HasMoreMallocsThanFrees = AllocTracer.Stop(); + if (!LooseMemeq(DataCopy, Data, Size)) + CrashOnOverwrittenData(); + CurrentUnitSize = 0; + delete[] DataCopy; +} + +void Fuzzer::WriteToOutputCorpus(const Unit &U) { + if (Options.OnlyASCII) + assert(IsASCII(U)); + if (Options.OutputCorpus.empty()) + return; + std::string Path = DirPlusFile(Options.OutputCorpus, Hash(U)); + WriteToFile(U, Path); + if (Options.Verbosity >= 2) + Printf("Written %zd bytes to %s\n", U.size(), Path.c_str()); +} + +void Fuzzer::WriteUnitToFileWithPrefix(const Unit &U, const char *Prefix) { + if (!Options.SaveArtifacts) + return; + std::string Path = Options.ArtifactPrefix + Prefix + Hash(U); + if (!Options.ExactArtifactPath.empty()) + Path = Options.ExactArtifactPath; // Overrides ArtifactPrefix. + WriteToFile(U, Path); + Printf("artifact_prefix='%s'; Test unit written to %s\n", + Options.ArtifactPrefix.c_str(), Path.c_str()); + if (U.size() <= kMaxUnitSizeToPrint) + Printf("Base64: %s\n", Base64(U).c_str()); +} + +void Fuzzer::PrintStatusForNewUnit(const Unit &U, const char *Text) { + if (!Options.PrintNEW) + return; + PrintStats(Text, ""); + if (Options.Verbosity) { + Printf(" L: %zd/%zd ", U.size(), Corpus.MaxInputSize()); + MD.PrintMutationSequence(); + Printf("\n"); + } +} + +void Fuzzer::ReportNewCoverage(InputInfo *II, const Unit &U) { + II->NumSuccessfullMutations++; + MD.RecordSuccessfulMutationSequence(); + PrintStatusForNewUnit(U, II->Reduced ? "REDUCE" : "NEW "); + WriteToOutputCorpus(U); + NumberOfNewUnitsAdded++; + CheckExitOnSrcPosOrItem(); // Check only after the unit is saved to corpus. + LastCorpusUpdateRun = TotalNumberOfRuns; +} + +// Tries detecting a memory leak on the particular input that we have just +// executed before calling this function. +void Fuzzer::TryDetectingAMemoryLeak(const uint8_t *Data, size_t Size, + bool DuringInitialCorpusExecution) { + if (!HasMoreMallocsThanFrees) + return; // mallocs==frees, a leak is unlikely. + if (!Options.DetectLeaks) + return; + if (!DuringInitialCorpusExecution && + TotalNumberOfRuns >= Options.MaxNumberOfRuns) + return; + if (!&(EF->__lsan_enable) || !&(EF->__lsan_disable) || + !(EF->__lsan_do_recoverable_leak_check)) + return; // No lsan. + // Run the target once again, but with lsan disabled so that if there is + // a real leak we do not report it twice. + EF->__lsan_disable(); + ExecuteCallback(Data, Size); + EF->__lsan_enable(); + if (!HasMoreMallocsThanFrees) + return; // a leak is unlikely. + if (NumberOfLeakDetectionAttempts++ > 1000) { + Options.DetectLeaks = false; + Printf("INFO: libFuzzer disabled leak detection after every mutation.\n" + " Most likely the target function accumulates allocated\n" + " memory in a global state w/o actually leaking it.\n" + " You may try running this binary with -trace_malloc=[12]" + " to get a trace of mallocs and frees.\n" + " If LeakSanitizer is enabled in this process it will still\n" + " run on the process shutdown.\n"); + return; + } + // Now perform the actual lsan pass. This is expensive and we must ensure + // we don't call it too often. + if (EF->__lsan_do_recoverable_leak_check()) { // Leak is found, report it. + if (DuringInitialCorpusExecution) + Printf("\nINFO: a leak has been found in the initial corpus.\n\n"); + Printf("INFO: to ignore leaks on libFuzzer side use -detect_leaks=0.\n\n"); + CurrentUnitSize = Size; + DumpCurrentUnit("leak-"); + PrintFinalStats(); + _Exit(Options.ErrorExitCode); // not exit() to disable lsan further on. + } +} + +void Fuzzer::MutateAndTestOne() { + MD.StartMutationSequence(); + + auto &II = Corpus.ChooseUnitToMutate(MD.GetRand()); + if (Options.UseFeatureFrequency) + Corpus.UpdateFeatureFrequencyScore(&II); + const auto &U = II.U; + memcpy(BaseSha1, II.Sha1, sizeof(BaseSha1)); + assert(CurrentUnitData); + size_t Size = U.size(); + assert(Size <= MaxInputLen && "Oversized Unit"); + memcpy(CurrentUnitData, U.data(), Size); + + assert(MaxMutationLen > 0); + + size_t CurrentMaxMutationLen = + Min(MaxMutationLen, Max(U.size(), TmpMaxMutationLen)); + assert(CurrentMaxMutationLen > 0); + + for (int i = 0; i < Options.MutateDepth; i++) { + if (TotalNumberOfRuns >= Options.MaxNumberOfRuns) + break; + MaybeExitGracefully(); + size_t NewSize = 0; + NewSize = MD.Mutate(CurrentUnitData, Size, CurrentMaxMutationLen); + assert(NewSize > 0 && "Mutator returned empty unit"); + assert(NewSize <= CurrentMaxMutationLen && "Mutator return oversized unit"); + Size = NewSize; + II.NumExecutedMutations++; + + bool FoundUniqFeatures = false; + bool NewCov = RunOne(CurrentUnitData, Size, /*MayDeleteFile=*/true, &II, + &FoundUniqFeatures); + TryDetectingAMemoryLeak(CurrentUnitData, Size, + /*DuringInitialCorpusExecution*/ false); + if (NewCov) { + ReportNewCoverage(&II, {CurrentUnitData, CurrentUnitData + Size}); + break; // We will mutate this input more in the next rounds. + } + if (Options.ReduceDepth && !FoundUniqFeatures) + break; + } +} + +void Fuzzer::PurgeAllocator() { + if (Options.PurgeAllocatorIntervalSec < 0 || !EF->__sanitizer_purge_allocator) + return; + if (duration_cast<seconds>(system_clock::now() - + LastAllocatorPurgeAttemptTime) + .count() < Options.PurgeAllocatorIntervalSec) + return; + + if (Options.RssLimitMb <= 0 || + GetPeakRSSMb() > static_cast<size_t>(Options.RssLimitMb) / 2) + EF->__sanitizer_purge_allocator(); + + LastAllocatorPurgeAttemptTime = system_clock::now(); +} + +void Fuzzer::ReadAndExecuteSeedCorpora(const Vector<std::string> &CorpusDirs) { + const size_t kMaxSaneLen = 1 << 20; + const size_t kMinDefaultLen = 4096; + Vector<SizedFile> SizedFiles; + size_t MaxSize = 0; + size_t MinSize = -1; + size_t TotalSize = 0; + size_t LastNumFiles = 0; + for (auto &Dir : CorpusDirs) { + GetSizedFilesFromDir(Dir, &SizedFiles); + Printf("INFO: % 8zd files found in %s\n", SizedFiles.size() - LastNumFiles, + Dir.c_str()); + LastNumFiles = SizedFiles.size(); + } + for (auto &File : SizedFiles) { + MaxSize = Max(File.Size, MaxSize); + MinSize = Min(File.Size, MinSize); + TotalSize += File.Size; + } + if (Options.MaxLen == 0) + SetMaxInputLen(std::min(std::max(kMinDefaultLen, MaxSize), kMaxSaneLen)); + assert(MaxInputLen > 0); + + // Test the callback with empty input and never try it again. + uint8_t dummy = 0; + ExecuteCallback(&dummy, 0); + + if (SizedFiles.empty()) { + Printf("INFO: A corpus is not provided, starting from an empty corpus\n"); + Unit U({'\n'}); // Valid ASCII input. + RunOne(U.data(), U.size()); + } else { + Printf("INFO: seed corpus: files: %zd min: %zdb max: %zdb total: %zdb" + " rss: %zdMb\n", + SizedFiles.size(), MinSize, MaxSize, TotalSize, GetPeakRSSMb()); + if (Options.ShuffleAtStartUp) + std::shuffle(SizedFiles.begin(), SizedFiles.end(), MD.GetRand()); + + if (Options.PreferSmall) { + std::stable_sort(SizedFiles.begin(), SizedFiles.end()); + assert(SizedFiles.front().Size <= SizedFiles.back().Size); + } + + // Load and execute inputs one by one. + for (auto &SF : SizedFiles) { + auto U = FileToVector(SF.File, MaxInputLen, /*ExitOnError=*/false); + assert(U.size() <= MaxInputLen); + RunOne(U.data(), U.size()); + CheckExitOnSrcPosOrItem(); + TryDetectingAMemoryLeak(U.data(), U.size(), + /*DuringInitialCorpusExecution*/ true); + } + } + + PrintStats("INITED"); + if (Corpus.empty()) { + Printf("ERROR: no interesting inputs were found. " + "Is the code instrumented for coverage? Exiting.\n"); + exit(1); + } +} + +void Fuzzer::Loop(const Vector<std::string> &CorpusDirs) { + ReadAndExecuteSeedCorpora(CorpusDirs); + TPC.SetPrintNewPCs(Options.PrintNewCovPcs); + TPC.SetPrintNewFuncs(Options.PrintNewCovFuncs); + system_clock::time_point LastCorpusReload = system_clock::now(); + if (Options.DoCrossOver) + MD.SetCorpus(&Corpus); + while (true) { + auto Now = system_clock::now(); + if (duration_cast<seconds>(Now - LastCorpusReload).count() >= + Options.ReloadIntervalSec) { + RereadOutputCorpus(MaxInputLen); + LastCorpusReload = system_clock::now(); + } + if (TotalNumberOfRuns >= Options.MaxNumberOfRuns) + break; + if (TimedOut()) + break; + + // Update TmpMaxMutationLen + if (Options.ExperimentalLenControl) { + if (TmpMaxMutationLen < MaxMutationLen && + TotalNumberOfRuns - LastCorpusUpdateRun > + Options.ExperimentalLenControl * Log(TmpMaxMutationLen)) { + TmpMaxMutationLen = + Min(MaxMutationLen, TmpMaxMutationLen + Log(TmpMaxMutationLen)); + if (TmpMaxMutationLen <= MaxMutationLen) + Printf("#%zd\tTEMP_MAX_LEN: %zd (%zd %zd)\n", TotalNumberOfRuns, + TmpMaxMutationLen, Options.ExperimentalLenControl, + LastCorpusUpdateRun); + LastCorpusUpdateRun = TotalNumberOfRuns; + } + } else { + TmpMaxMutationLen = MaxMutationLen; + } + + // Perform several mutations and runs. + MutateAndTestOne(); + + PurgeAllocator(); + } + + PrintStats("DONE ", "\n"); + MD.PrintRecommendedDictionary(); +} + +void Fuzzer::MinimizeCrashLoop(const Unit &U) { + if (U.size() <= 1) + return; + while (!TimedOut() && TotalNumberOfRuns < Options.MaxNumberOfRuns) { + MD.StartMutationSequence(); + memcpy(CurrentUnitData, U.data(), U.size()); + for (int i = 0; i < Options.MutateDepth; i++) { + size_t NewSize = MD.Mutate(CurrentUnitData, U.size(), MaxMutationLen); + assert(NewSize > 0 && NewSize <= MaxMutationLen); + ExecuteCallback(CurrentUnitData, NewSize); + PrintPulseAndReportSlowInput(CurrentUnitData, NewSize); + TryDetectingAMemoryLeak(CurrentUnitData, NewSize, + /*DuringInitialCorpusExecution*/ false); + } + } +} + +void Fuzzer::AnnounceOutput(const uint8_t *Data, size_t Size) { + if (SMR.IsServer()) { + SMR.WriteByteArray(Data, Size); + } else if (SMR.IsClient()) { + SMR.PostClient(); + SMR.WaitServer(); + size_t OtherSize = SMR.ReadByteArraySize(); + uint8_t *OtherData = SMR.GetByteArray(); + if (Size != OtherSize || memcmp(Data, OtherData, Size) != 0) { + size_t i = 0; + for (i = 0; i < Min(Size, OtherSize); i++) + if (Data[i] != OtherData[i]) + break; + Printf("==%lu== ERROR: libFuzzer: equivalence-mismatch. Sizes: %zd %zd; " + "offset %zd\n", + GetPid(), Size, OtherSize, i); + DumpCurrentUnit("mismatch-"); + Printf("SUMMARY: libFuzzer: equivalence-mismatch\n"); + PrintFinalStats(); + _Exit(Options.ErrorExitCode); + } + } +} + +} // namespace fuzzer + +extern "C" { + +size_t LLVMFuzzerMutate(uint8_t *Data, size_t Size, size_t MaxSize) { + assert(fuzzer::F); + return fuzzer::F->GetMD().DefaultMutate(Data, Size, MaxSize); +} + +// Experimental +void LLVMFuzzerAnnounceOutput(const uint8_t *Data, size_t Size) { + assert(fuzzer::F); + fuzzer::F->AnnounceOutput(Data, Size); +} +} // extern "C" diff --git a/lib/fuzzer/FuzzerMain.cpp b/lib/fuzzer/FuzzerMain.cpp new file mode 100644 index 000000000000..af8657200be2 --- /dev/null +++ b/lib/fuzzer/FuzzerMain.cpp @@ -0,0 +1,21 @@ +//===- FuzzerMain.cpp - main() function and flags -------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// main() and flags. +//===----------------------------------------------------------------------===// + +#include "FuzzerDefs.h" + +extern "C" { +// This function should be defined by the user. +int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size); +} // extern "C" + +int main(int argc, char **argv) { + return fuzzer::FuzzerDriver(&argc, &argv, LLVMFuzzerTestOneInput); +} diff --git a/lib/fuzzer/FuzzerMerge.cpp b/lib/fuzzer/FuzzerMerge.cpp new file mode 100644 index 000000000000..5f3052a39c16 --- /dev/null +++ b/lib/fuzzer/FuzzerMerge.cpp @@ -0,0 +1,390 @@ +//===- FuzzerMerge.cpp - merging corpora ----------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Merging corpora. +//===----------------------------------------------------------------------===// + +#include "FuzzerCommand.h" +#include "FuzzerMerge.h" +#include "FuzzerIO.h" +#include "FuzzerInternal.h" +#include "FuzzerTracePC.h" +#include "FuzzerUtil.h" + +#include <fstream> +#include <iterator> +#include <set> +#include <sstream> + +namespace fuzzer { + +bool Merger::Parse(const std::string &Str, bool ParseCoverage) { + std::istringstream SS(Str); + return Parse(SS, ParseCoverage); +} + +void Merger::ParseOrExit(std::istream &IS, bool ParseCoverage) { + if (!Parse(IS, ParseCoverage)) { + Printf("MERGE: failed to parse the control file (unexpected error)\n"); + exit(1); + } +} + +// The control file example: +// +// 3 # The number of inputs +// 1 # The number of inputs in the first corpus, <= the previous number +// file0 +// file1 +// file2 # One file name per line. +// STARTED 0 123 # FileID, file size +// DONE 0 1 4 6 8 # FileID COV1 COV2 ... +// STARTED 1 456 # If DONE is missing, the input crashed while processing. +// STARTED 2 567 +// DONE 2 8 9 +bool Merger::Parse(std::istream &IS, bool ParseCoverage) { + LastFailure.clear(); + std::string Line; + + // Parse NumFiles. + if (!std::getline(IS, Line, '\n')) return false; + std::istringstream L1(Line); + size_t NumFiles = 0; + L1 >> NumFiles; + if (NumFiles == 0 || NumFiles > 10000000) return false; + + // Parse NumFilesInFirstCorpus. + if (!std::getline(IS, Line, '\n')) return false; + std::istringstream L2(Line); + NumFilesInFirstCorpus = NumFiles + 1; + L2 >> NumFilesInFirstCorpus; + if (NumFilesInFirstCorpus > NumFiles) return false; + + // Parse file names. + Files.resize(NumFiles); + for (size_t i = 0; i < NumFiles; i++) + if (!std::getline(IS, Files[i].Name, '\n')) + return false; + + // Parse STARTED and DONE lines. + size_t ExpectedStartMarker = 0; + const size_t kInvalidStartMarker = -1; + size_t LastSeenStartMarker = kInvalidStartMarker; + Vector<uint32_t> TmpFeatures; + while (std::getline(IS, Line, '\n')) { + std::istringstream ISS1(Line); + std::string Marker; + size_t N; + ISS1 >> Marker; + ISS1 >> N; + if (Marker == "STARTED") { + // STARTED FILE_ID FILE_SIZE + if (ExpectedStartMarker != N) + return false; + ISS1 >> Files[ExpectedStartMarker].Size; + LastSeenStartMarker = ExpectedStartMarker; + assert(ExpectedStartMarker < Files.size()); + ExpectedStartMarker++; + } else if (Marker == "DONE") { + // DONE FILE_ID COV1 COV2 COV3 ... + size_t CurrentFileIdx = N; + if (CurrentFileIdx != LastSeenStartMarker) + return false; + LastSeenStartMarker = kInvalidStartMarker; + if (ParseCoverage) { + TmpFeatures.clear(); // use a vector from outer scope to avoid resizes. + while (ISS1 >> std::hex >> N) + TmpFeatures.push_back(N); + std::sort(TmpFeatures.begin(), TmpFeatures.end()); + Files[CurrentFileIdx].Features = TmpFeatures; + } + } else { + return false; + } + } + if (LastSeenStartMarker != kInvalidStartMarker) + LastFailure = Files[LastSeenStartMarker].Name; + + FirstNotProcessedFile = ExpectedStartMarker; + return true; +} + +size_t Merger::ApproximateMemoryConsumption() const { + size_t Res = 0; + for (const auto &F: Files) + Res += sizeof(F) + F.Features.size() * sizeof(F.Features[0]); + return Res; +} + +// Decides which files need to be merged (add thost to NewFiles). +// Returns the number of new features added. +size_t Merger::Merge(const Set<uint32_t> &InitialFeatures, + Vector<std::string> *NewFiles) { + NewFiles->clear(); + assert(NumFilesInFirstCorpus <= Files.size()); + Set<uint32_t> AllFeatures(InitialFeatures); + + // What features are in the initial corpus? + for (size_t i = 0; i < NumFilesInFirstCorpus; i++) { + auto &Cur = Files[i].Features; + AllFeatures.insert(Cur.begin(), Cur.end()); + } + size_t InitialNumFeatures = AllFeatures.size(); + + // Remove all features that we already know from all other inputs. + for (size_t i = NumFilesInFirstCorpus; i < Files.size(); i++) { + auto &Cur = Files[i].Features; + Vector<uint32_t> Tmp; + std::set_difference(Cur.begin(), Cur.end(), AllFeatures.begin(), + AllFeatures.end(), std::inserter(Tmp, Tmp.begin())); + Cur.swap(Tmp); + } + + // Sort. Give preference to + // * smaller files + // * files with more features. + std::sort(Files.begin() + NumFilesInFirstCorpus, Files.end(), + [&](const MergeFileInfo &a, const MergeFileInfo &b) -> bool { + if (a.Size != b.Size) + return a.Size < b.Size; + return a.Features.size() > b.Features.size(); + }); + + // One greedy pass: add the file's features to AllFeatures. + // If new features were added, add this file to NewFiles. + for (size_t i = NumFilesInFirstCorpus; i < Files.size(); i++) { + auto &Cur = Files[i].Features; + // Printf("%s -> sz %zd ft %zd\n", Files[i].Name.c_str(), + // Files[i].Size, Cur.size()); + size_t OldSize = AllFeatures.size(); + AllFeatures.insert(Cur.begin(), Cur.end()); + if (AllFeatures.size() > OldSize) + NewFiles->push_back(Files[i].Name); + } + return AllFeatures.size() - InitialNumFeatures; +} + +void Merger::PrintSummary(std::ostream &OS) { + for (auto &File : Files) { + OS << std::hex; + OS << File.Name << " size: " << File.Size << " features: "; + for (auto Feature : File.Features) + OS << " " << Feature; + OS << "\n"; + } +} + +Set<uint32_t> Merger::AllFeatures() const { + Set<uint32_t> S; + for (auto &File : Files) + S.insert(File.Features.begin(), File.Features.end()); + return S; +} + +Set<uint32_t> Merger::ParseSummary(std::istream &IS) { + std::string Line, Tmp; + Set<uint32_t> Res; + while (std::getline(IS, Line, '\n')) { + size_t N; + std::istringstream ISS1(Line); + ISS1 >> Tmp; // Name + ISS1 >> Tmp; // size: + assert(Tmp == "size:" && "Corrupt summary file"); + ISS1 >> std::hex; + ISS1 >> N; // File Size + ISS1 >> Tmp; // features: + assert(Tmp == "features:" && "Corrupt summary file"); + while (ISS1 >> std::hex >> N) + Res.insert(N); + } + return Res; +} + +// Inner process. May crash if the target crashes. +void Fuzzer::CrashResistantMergeInternalStep(const std::string &CFPath) { + Printf("MERGE-INNER: using the control file '%s'\n", CFPath.c_str()); + Merger M; + std::ifstream IF(CFPath); + M.ParseOrExit(IF, false); + IF.close(); + if (!M.LastFailure.empty()) + Printf("MERGE-INNER: '%s' caused a failure at the previous merge step\n", + M.LastFailure.c_str()); + + Printf("MERGE-INNER: %zd total files;" + " %zd processed earlier; will process %zd files now\n", + M.Files.size(), M.FirstNotProcessedFile, + M.Files.size() - M.FirstNotProcessedFile); + + std::ofstream OF(CFPath, std::ofstream::out | std::ofstream::app); + Set<size_t> AllFeatures; + for (size_t i = M.FirstNotProcessedFile; i < M.Files.size(); i++) { + MaybeExitGracefully(); + auto U = FileToVector(M.Files[i].Name); + if (U.size() > MaxInputLen) { + U.resize(MaxInputLen); + U.shrink_to_fit(); + } + std::ostringstream StartedLine; + // Write the pre-run marker. + OF << "STARTED " << std::dec << i << " " << U.size() << "\n"; + OF.flush(); // Flush is important since Command::Execute may crash. + // Run. + TPC.ResetMaps(); + ExecuteCallback(U.data(), U.size()); + // Collect coverage. We are iterating over the files in this order: + // * First, files in the initial corpus ordered by size, smallest first. + // * Then, all other files, smallest first. + // So it makes no sense to record all features for all files, instead we + // only record features that were not seen before. + Set<size_t> UniqFeatures; + TPC.CollectFeatures([&](size_t Feature) { + if (AllFeatures.insert(Feature).second) + UniqFeatures.insert(Feature); + }); + // Show stats. + if (!(TotalNumberOfRuns & (TotalNumberOfRuns - 1))) + PrintStats("pulse "); + // Write the post-run marker and the coverage. + OF << "DONE " << i; + for (size_t F : UniqFeatures) + OF << " " << std::hex << F; + OF << "\n"; + OF.flush(); + } +} + +static void WriteNewControlFile(const std::string &CFPath, + const Vector<SizedFile> &AllFiles, + size_t NumFilesInFirstCorpus) { + RemoveFile(CFPath); + std::ofstream ControlFile(CFPath); + ControlFile << AllFiles.size() << "\n"; + ControlFile << NumFilesInFirstCorpus << "\n"; + for (auto &SF: AllFiles) + ControlFile << SF.File << "\n"; + if (!ControlFile) { + Printf("MERGE-OUTER: failed to write to the control file: %s\n", + CFPath.c_str()); + exit(1); + } +} + +// Outer process. Does not call the target code and thus sohuld not fail. +void Fuzzer::CrashResistantMerge(const Vector<std::string> &Args, + const Vector<std::string> &Corpora, + const char *CoverageSummaryInputPathOrNull, + const char *CoverageSummaryOutputPathOrNull, + const char *MergeControlFilePathOrNull) { + if (Corpora.size() <= 1) { + Printf("Merge requires two or more corpus dirs\n"); + return; + } + auto CFPath = + MergeControlFilePathOrNull + ? MergeControlFilePathOrNull + : DirPlusFile(TmpDir(), + "libFuzzerTemp." + std::to_string(GetPid()) + ".txt"); + + size_t NumAttempts = 0; + if (MergeControlFilePathOrNull && FileSize(MergeControlFilePathOrNull)) { + Printf("MERGE-OUTER: non-empty control file provided: '%s'\n", + MergeControlFilePathOrNull); + Merger M; + std::ifstream IF(MergeControlFilePathOrNull); + if (M.Parse(IF, /*ParseCoverage=*/false)) { + Printf("MERGE-OUTER: control file ok, %zd files total," + " first not processed file %zd\n", + M.Files.size(), M.FirstNotProcessedFile); + if (!M.LastFailure.empty()) + Printf("MERGE-OUTER: '%s' will be skipped as unlucky " + "(merge has stumbled on it the last time)\n", + M.LastFailure.c_str()); + if (M.FirstNotProcessedFile >= M.Files.size()) { + Printf("MERGE-OUTER: nothing to do, merge has been completed before\n"); + exit(0); + } + + NumAttempts = M.Files.size() - M.FirstNotProcessedFile; + } else { + Printf("MERGE-OUTER: bad control file, will overwrite it\n"); + } + } + + if (!NumAttempts) { + // The supplied control file is empty or bad, create a fresh one. + Vector<SizedFile> AllFiles; + GetSizedFilesFromDir(Corpora[0], &AllFiles); + size_t NumFilesInFirstCorpus = AllFiles.size(); + std::sort(AllFiles.begin(), AllFiles.end()); + for (size_t i = 1; i < Corpora.size(); i++) + GetSizedFilesFromDir(Corpora[i], &AllFiles); + std::sort(AllFiles.begin() + NumFilesInFirstCorpus, AllFiles.end()); + Printf("MERGE-OUTER: %zd files, %zd in the initial corpus\n", + AllFiles.size(), NumFilesInFirstCorpus); + WriteNewControlFile(CFPath, AllFiles, NumFilesInFirstCorpus); + NumAttempts = AllFiles.size(); + } + + // Execute the inner process until it passes. + // Every inner process should execute at least one input. + Command BaseCmd(Args); + BaseCmd.removeFlag("merge"); + bool Success = false; + for (size_t Attempt = 1; Attempt <= NumAttempts; Attempt++) { + MaybeExitGracefully(); + Printf("MERGE-OUTER: attempt %zd\n", Attempt); + Command Cmd(BaseCmd); + Cmd.addFlag("merge_control_file", CFPath); + Cmd.addFlag("merge_inner", "1"); + auto ExitCode = ExecuteCommand(Cmd); + if (!ExitCode) { + Printf("MERGE-OUTER: succesfull in %zd attempt(s)\n", Attempt); + Success = true; + break; + } + } + if (!Success) { + Printf("MERGE-OUTER: zero succesfull attempts, exiting\n"); + exit(1); + } + // Read the control file and do the merge. + Merger M; + std::ifstream IF(CFPath); + IF.seekg(0, IF.end); + Printf("MERGE-OUTER: the control file has %zd bytes\n", (size_t)IF.tellg()); + IF.seekg(0, IF.beg); + M.ParseOrExit(IF, true); + IF.close(); + Printf("MERGE-OUTER: consumed %zdMb (%zdMb rss) to parse the control file\n", + M.ApproximateMemoryConsumption() >> 20, GetPeakRSSMb()); + if (CoverageSummaryOutputPathOrNull) { + Printf("MERGE-OUTER: writing coverage summary for %zd files to %s\n", + M.Files.size(), CoverageSummaryOutputPathOrNull); + std::ofstream SummaryOut(CoverageSummaryOutputPathOrNull); + M.PrintSummary(SummaryOut); + } + Vector<std::string> NewFiles; + Set<uint32_t> InitialFeatures; + if (CoverageSummaryInputPathOrNull) { + std::ifstream SummaryIn(CoverageSummaryInputPathOrNull); + InitialFeatures = M.ParseSummary(SummaryIn); + Printf("MERGE-OUTER: coverage summary loaded from %s, %zd features found\n", + CoverageSummaryInputPathOrNull, InitialFeatures.size()); + } + size_t NumNewFeatures = M.Merge(InitialFeatures, &NewFiles); + Printf("MERGE-OUTER: %zd new files with %zd new features added\n", + NewFiles.size(), NumNewFeatures); + for (auto &F: NewFiles) + WriteToOutputCorpus(FileToVector(F, MaxInputLen)); + // We are done, delete the control file if it was a temporary one. + if (!MergeControlFilePathOrNull) + RemoveFile(CFPath); +} + +} // namespace fuzzer diff --git a/lib/fuzzer/FuzzerMerge.h b/lib/fuzzer/FuzzerMerge.h new file mode 100644 index 000000000000..e54885a1ebae --- /dev/null +++ b/lib/fuzzer/FuzzerMerge.h @@ -0,0 +1,80 @@ +//===- FuzzerMerge.h - merging corpa ----------------------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Merging Corpora. +// +// The task: +// Take the existing corpus (possibly empty) and merge new inputs into +// it so that only inputs with new coverage ('features') are added. +// The process should tolerate the crashes, OOMs, leaks, etc. +// +// Algorithm: +// The outter process collects the set of files and writes their names +// into a temporary "control" file, then repeatedly launches the inner +// process until all inputs are processed. +// The outer process does not actually execute the target code. +// +// The inner process reads the control file and sees a) list of all the inputs +// and b) the last processed input. Then it starts processing the inputs one +// by one. Before processing every input it writes one line to control file: +// STARTED INPUT_ID INPUT_SIZE +// After processing an input it write another line: +// DONE INPUT_ID Feature1 Feature2 Feature3 ... +// If a crash happens while processing an input the last line in the control +// file will be "STARTED INPUT_ID" and so the next process will know +// where to resume. +// +// Once all inputs are processed by the innner process(es) the outer process +// reads the control files and does the merge based entirely on the contents +// of control file. +// It uses a single pass greedy algorithm choosing first the smallest inputs +// within the same size the inputs that have more new features. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_MERGE_H +#define LLVM_FUZZER_MERGE_H + +#include "FuzzerDefs.h" + +#include <istream> +#include <ostream> +#include <set> +#include <vector> + +namespace fuzzer { + +struct MergeFileInfo { + std::string Name; + size_t Size = 0; + Vector<uint32_t> Features; +}; + +struct Merger { + Vector<MergeFileInfo> Files; + size_t NumFilesInFirstCorpus = 0; + size_t FirstNotProcessedFile = 0; + std::string LastFailure; + + bool Parse(std::istream &IS, bool ParseCoverage); + bool Parse(const std::string &Str, bool ParseCoverage); + void ParseOrExit(std::istream &IS, bool ParseCoverage); + void PrintSummary(std::ostream &OS); + Set<uint32_t> ParseSummary(std::istream &IS); + size_t Merge(const Set<uint32_t> &InitialFeatures, + Vector<std::string> *NewFiles); + size_t Merge(Vector<std::string> *NewFiles) { + return Merge(Set<uint32_t>{}, NewFiles); + } + size_t ApproximateMemoryConsumption() const; + Set<uint32_t> AllFeatures() const; +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_MERGE_H diff --git a/lib/fuzzer/FuzzerMutate.cpp b/lib/fuzzer/FuzzerMutate.cpp new file mode 100644 index 000000000000..9ee5299f1b60 --- /dev/null +++ b/lib/fuzzer/FuzzerMutate.cpp @@ -0,0 +1,533 @@ +//===- FuzzerMutate.cpp - Mutate a test input -----------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Mutate a test input. +//===----------------------------------------------------------------------===// + +#include "FuzzerMutate.h" +#include "FuzzerCorpus.h" +#include "FuzzerDefs.h" +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" +#include "FuzzerOptions.h" + +namespace fuzzer { + +const size_t Dictionary::kMaxDictSize; + +static void PrintASCII(const Word &W, const char *PrintAfter) { + PrintASCII(W.data(), W.size(), PrintAfter); +} + +MutationDispatcher::MutationDispatcher(Random &Rand, + const FuzzingOptions &Options) + : Rand(Rand), Options(Options) { + DefaultMutators.insert( + DefaultMutators.begin(), + { + {&MutationDispatcher::Mutate_EraseBytes, "EraseBytes"}, + {&MutationDispatcher::Mutate_InsertByte, "InsertByte"}, + {&MutationDispatcher::Mutate_InsertRepeatedBytes, + "InsertRepeatedBytes"}, + {&MutationDispatcher::Mutate_ChangeByte, "ChangeByte"}, + {&MutationDispatcher::Mutate_ChangeBit, "ChangeBit"}, + {&MutationDispatcher::Mutate_ShuffleBytes, "ShuffleBytes"}, + {&MutationDispatcher::Mutate_ChangeASCIIInteger, "ChangeASCIIInt"}, + {&MutationDispatcher::Mutate_ChangeBinaryInteger, "ChangeBinInt"}, + {&MutationDispatcher::Mutate_CopyPart, "CopyPart"}, + {&MutationDispatcher::Mutate_CrossOver, "CrossOver"}, + {&MutationDispatcher::Mutate_AddWordFromManualDictionary, + "ManualDict"}, + {&MutationDispatcher::Mutate_AddWordFromPersistentAutoDictionary, + "PersAutoDict"}, + }); + if(Options.UseCmp) + DefaultMutators.push_back( + {&MutationDispatcher::Mutate_AddWordFromTORC, "CMP"}); + + if (EF->LLVMFuzzerCustomMutator) + Mutators.push_back({&MutationDispatcher::Mutate_Custom, "Custom"}); + else + Mutators = DefaultMutators; + + if (EF->LLVMFuzzerCustomCrossOver) + Mutators.push_back( + {&MutationDispatcher::Mutate_CustomCrossOver, "CustomCrossOver"}); +} + +static char RandCh(Random &Rand) { + if (Rand.RandBool()) return Rand(256); + const char *Special = "!*'();:@&=+$,/?%#[]012Az-`~.\xff\x00"; + return Special[Rand(sizeof(Special) - 1)]; +} + +size_t MutationDispatcher::Mutate_Custom(uint8_t *Data, size_t Size, + size_t MaxSize) { + return EF->LLVMFuzzerCustomMutator(Data, Size, MaxSize, Rand.Rand()); +} + +size_t MutationDispatcher::Mutate_CustomCrossOver(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (!Corpus || Corpus->size() < 2 || Size == 0) + return 0; + size_t Idx = Rand(Corpus->size()); + const Unit &Other = (*Corpus)[Idx]; + if (Other.empty()) + return 0; + CustomCrossOverInPlaceHere.resize(MaxSize); + auto &U = CustomCrossOverInPlaceHere; + size_t NewSize = EF->LLVMFuzzerCustomCrossOver( + Data, Size, Other.data(), Other.size(), U.data(), U.size(), Rand.Rand()); + if (!NewSize) + return 0; + assert(NewSize <= MaxSize && "CustomCrossOver returned overisized unit"); + memcpy(Data, U.data(), NewSize); + return NewSize; +} + +size_t MutationDispatcher::Mutate_ShuffleBytes(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size > MaxSize || Size == 0) return 0; + size_t ShuffleAmount = + Rand(std::min(Size, (size_t)8)) + 1; // [1,8] and <= Size. + size_t ShuffleStart = Rand(Size - ShuffleAmount); + assert(ShuffleStart + ShuffleAmount <= Size); + std::shuffle(Data + ShuffleStart, Data + ShuffleStart + ShuffleAmount, Rand); + return Size; +} + +size_t MutationDispatcher::Mutate_EraseBytes(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size <= 1) return 0; + size_t N = Rand(Size / 2) + 1; + assert(N < Size); + size_t Idx = Rand(Size - N + 1); + // Erase Data[Idx:Idx+N]. + memmove(Data + Idx, Data + Idx + N, Size - Idx - N); + // Printf("Erase: %zd %zd => %zd; Idx %zd\n", N, Size, Size - N, Idx); + return Size - N; +} + +size_t MutationDispatcher::Mutate_InsertByte(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size >= MaxSize) return 0; + size_t Idx = Rand(Size + 1); + // Insert new value at Data[Idx]. + memmove(Data + Idx + 1, Data + Idx, Size - Idx); + Data[Idx] = RandCh(Rand); + return Size + 1; +} + +size_t MutationDispatcher::Mutate_InsertRepeatedBytes(uint8_t *Data, + size_t Size, + size_t MaxSize) { + const size_t kMinBytesToInsert = 3; + if (Size + kMinBytesToInsert >= MaxSize) return 0; + size_t MaxBytesToInsert = std::min(MaxSize - Size, (size_t)128); + size_t N = Rand(MaxBytesToInsert - kMinBytesToInsert + 1) + kMinBytesToInsert; + assert(Size + N <= MaxSize && N); + size_t Idx = Rand(Size + 1); + // Insert new values at Data[Idx]. + memmove(Data + Idx + N, Data + Idx, Size - Idx); + // Give preference to 0x00 and 0xff. + uint8_t Byte = Rand.RandBool() ? Rand(256) : (Rand.RandBool() ? 0 : 255); + for (size_t i = 0; i < N; i++) + Data[Idx + i] = Byte; + return Size + N; +} + +size_t MutationDispatcher::Mutate_ChangeByte(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size > MaxSize) return 0; + size_t Idx = Rand(Size); + Data[Idx] = RandCh(Rand); + return Size; +} + +size_t MutationDispatcher::Mutate_ChangeBit(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size > MaxSize) return 0; + size_t Idx = Rand(Size); + Data[Idx] ^= 1 << Rand(8); + return Size; +} + +size_t MutationDispatcher::Mutate_AddWordFromManualDictionary(uint8_t *Data, + size_t Size, + size_t MaxSize) { + return AddWordFromDictionary(ManualDictionary, Data, Size, MaxSize); +} + +size_t MutationDispatcher::ApplyDictionaryEntry(uint8_t *Data, size_t Size, + size_t MaxSize, + DictionaryEntry &DE) { + const Word &W = DE.GetW(); + bool UsePositionHint = DE.HasPositionHint() && + DE.GetPositionHint() + W.size() < Size && + Rand.RandBool(); + if (Rand.RandBool()) { // Insert W. + if (Size + W.size() > MaxSize) return 0; + size_t Idx = UsePositionHint ? DE.GetPositionHint() : Rand(Size + 1); + memmove(Data + Idx + W.size(), Data + Idx, Size - Idx); + memcpy(Data + Idx, W.data(), W.size()); + Size += W.size(); + } else { // Overwrite some bytes with W. + if (W.size() > Size) return 0; + size_t Idx = UsePositionHint ? DE.GetPositionHint() : Rand(Size - W.size()); + memcpy(Data + Idx, W.data(), W.size()); + } + return Size; +} + +// Somewhere in the past we have observed a comparison instructions +// with arguments Arg1 Arg2. This function tries to guess a dictionary +// entry that will satisfy that comparison. +// It first tries to find one of the arguments (possibly swapped) in the +// input and if it succeeds it creates a DE with a position hint. +// Otherwise it creates a DE with one of the arguments w/o a position hint. +DictionaryEntry MutationDispatcher::MakeDictionaryEntryFromCMP( + const void *Arg1, const void *Arg2, + const void *Arg1Mutation, const void *Arg2Mutation, + size_t ArgSize, const uint8_t *Data, + size_t Size) { + ScopedDoingMyOwnMemOrStr scoped_doing_my_own_mem_os_str; + bool HandleFirst = Rand.RandBool(); + const void *ExistingBytes, *DesiredBytes; + Word W; + const uint8_t *End = Data + Size; + for (int Arg = 0; Arg < 2; Arg++) { + ExistingBytes = HandleFirst ? Arg1 : Arg2; + DesiredBytes = HandleFirst ? Arg2Mutation : Arg1Mutation; + HandleFirst = !HandleFirst; + W.Set(reinterpret_cast<const uint8_t*>(DesiredBytes), ArgSize); + const size_t kMaxNumPositions = 8; + size_t Positions[kMaxNumPositions]; + size_t NumPositions = 0; + for (const uint8_t *Cur = Data; + Cur < End && NumPositions < kMaxNumPositions; Cur++) { + Cur = + (const uint8_t *)SearchMemory(Cur, End - Cur, ExistingBytes, ArgSize); + if (!Cur) break; + Positions[NumPositions++] = Cur - Data; + } + if (!NumPositions) continue; + return DictionaryEntry(W, Positions[Rand(NumPositions)]); + } + DictionaryEntry DE(W); + return DE; +} + + +template <class T> +DictionaryEntry MutationDispatcher::MakeDictionaryEntryFromCMP( + T Arg1, T Arg2, const uint8_t *Data, size_t Size) { + if (Rand.RandBool()) Arg1 = Bswap(Arg1); + if (Rand.RandBool()) Arg2 = Bswap(Arg2); + T Arg1Mutation = Arg1 + Rand(-1, 1); + T Arg2Mutation = Arg2 + Rand(-1, 1); + return MakeDictionaryEntryFromCMP(&Arg1, &Arg2, &Arg1Mutation, &Arg2Mutation, + sizeof(Arg1), Data, Size); +} + +DictionaryEntry MutationDispatcher::MakeDictionaryEntryFromCMP( + const Word &Arg1, const Word &Arg2, const uint8_t *Data, size_t Size) { + return MakeDictionaryEntryFromCMP(Arg1.data(), Arg2.data(), Arg1.data(), + Arg2.data(), Arg1.size(), Data, Size); +} + +size_t MutationDispatcher::Mutate_AddWordFromTORC( + uint8_t *Data, size_t Size, size_t MaxSize) { + Word W; + DictionaryEntry DE; + switch (Rand(4)) { + case 0: { + auto X = TPC.TORC8.Get(Rand.Rand()); + DE = MakeDictionaryEntryFromCMP(X.A, X.B, Data, Size); + } break; + case 1: { + auto X = TPC.TORC4.Get(Rand.Rand()); + if ((X.A >> 16) == 0 && (X.B >> 16) == 0 && Rand.RandBool()) + DE = MakeDictionaryEntryFromCMP((uint16_t)X.A, (uint16_t)X.B, Data, Size); + else + DE = MakeDictionaryEntryFromCMP(X.A, X.B, Data, Size); + } break; + case 2: { + auto X = TPC.TORCW.Get(Rand.Rand()); + DE = MakeDictionaryEntryFromCMP(X.A, X.B, Data, Size); + } break; + case 3: if (Options.UseMemmem) { + auto X = TPC.MMT.Get(Rand.Rand()); + DE = DictionaryEntry(X); + } break; + default: + assert(0); + } + if (!DE.GetW().size()) return 0; + Size = ApplyDictionaryEntry(Data, Size, MaxSize, DE); + if (!Size) return 0; + DictionaryEntry &DERef = + CmpDictionaryEntriesDeque[CmpDictionaryEntriesDequeIdx++ % + kCmpDictionaryEntriesDequeSize]; + DERef = DE; + CurrentDictionaryEntrySequence.push_back(&DERef); + return Size; +} + +size_t MutationDispatcher::Mutate_AddWordFromPersistentAutoDictionary( + uint8_t *Data, size_t Size, size_t MaxSize) { + return AddWordFromDictionary(PersistentAutoDictionary, Data, Size, MaxSize); +} + +size_t MutationDispatcher::AddWordFromDictionary(Dictionary &D, uint8_t *Data, + size_t Size, size_t MaxSize) { + if (Size > MaxSize) return 0; + if (D.empty()) return 0; + DictionaryEntry &DE = D[Rand(D.size())]; + Size = ApplyDictionaryEntry(Data, Size, MaxSize, DE); + if (!Size) return 0; + DE.IncUseCount(); + CurrentDictionaryEntrySequence.push_back(&DE); + return Size; +} + +// Overwrites part of To[0,ToSize) with a part of From[0,FromSize). +// Returns ToSize. +size_t MutationDispatcher::CopyPartOf(const uint8_t *From, size_t FromSize, + uint8_t *To, size_t ToSize) { + // Copy From[FromBeg, FromBeg + CopySize) into To[ToBeg, ToBeg + CopySize). + size_t ToBeg = Rand(ToSize); + size_t CopySize = Rand(ToSize - ToBeg) + 1; + assert(ToBeg + CopySize <= ToSize); + CopySize = std::min(CopySize, FromSize); + size_t FromBeg = Rand(FromSize - CopySize + 1); + assert(FromBeg + CopySize <= FromSize); + memmove(To + ToBeg, From + FromBeg, CopySize); + return ToSize; +} + +// Inserts part of From[0,ToSize) into To. +// Returns new size of To on success or 0 on failure. +size_t MutationDispatcher::InsertPartOf(const uint8_t *From, size_t FromSize, + uint8_t *To, size_t ToSize, + size_t MaxToSize) { + if (ToSize >= MaxToSize) return 0; + size_t AvailableSpace = MaxToSize - ToSize; + size_t MaxCopySize = std::min(AvailableSpace, FromSize); + size_t CopySize = Rand(MaxCopySize) + 1; + size_t FromBeg = Rand(FromSize - CopySize + 1); + assert(FromBeg + CopySize <= FromSize); + size_t ToInsertPos = Rand(ToSize + 1); + assert(ToInsertPos + CopySize <= MaxToSize); + size_t TailSize = ToSize - ToInsertPos; + if (To == From) { + MutateInPlaceHere.resize(MaxToSize); + memcpy(MutateInPlaceHere.data(), From + FromBeg, CopySize); + memmove(To + ToInsertPos + CopySize, To + ToInsertPos, TailSize); + memmove(To + ToInsertPos, MutateInPlaceHere.data(), CopySize); + } else { + memmove(To + ToInsertPos + CopySize, To + ToInsertPos, TailSize); + memmove(To + ToInsertPos, From + FromBeg, CopySize); + } + return ToSize + CopySize; +} + +size_t MutationDispatcher::Mutate_CopyPart(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size > MaxSize || Size == 0) return 0; + if (Rand.RandBool()) + return CopyPartOf(Data, Size, Data, Size); + else + return InsertPartOf(Data, Size, Data, Size, MaxSize); +} + +size_t MutationDispatcher::Mutate_ChangeASCIIInteger(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size > MaxSize) return 0; + size_t B = Rand(Size); + while (B < Size && !isdigit(Data[B])) B++; + if (B == Size) return 0; + size_t E = B; + while (E < Size && isdigit(Data[E])) E++; + assert(B < E); + // now we have digits in [B, E). + // strtol and friends don't accept non-zero-teminated data, parse it manually. + uint64_t Val = Data[B] - '0'; + for (size_t i = B + 1; i < E; i++) + Val = Val * 10 + Data[i] - '0'; + + // Mutate the integer value. + switch(Rand(5)) { + case 0: Val++; break; + case 1: Val--; break; + case 2: Val /= 2; break; + case 3: Val *= 2; break; + case 4: Val = Rand(Val * Val); break; + default: assert(0); + } + // Just replace the bytes with the new ones, don't bother moving bytes. + for (size_t i = B; i < E; i++) { + size_t Idx = E + B - i - 1; + assert(Idx >= B && Idx < E); + Data[Idx] = (Val % 10) + '0'; + Val /= 10; + } + return Size; +} + +template<class T> +size_t ChangeBinaryInteger(uint8_t *Data, size_t Size, Random &Rand) { + if (Size < sizeof(T)) return 0; + size_t Off = Rand(Size - sizeof(T) + 1); + assert(Off + sizeof(T) <= Size); + T Val; + if (Off < 64 && !Rand(4)) { + Val = Size; + if (Rand.RandBool()) + Val = Bswap(Val); + } else { + memcpy(&Val, Data + Off, sizeof(Val)); + T Add = Rand(21); + Add -= 10; + if (Rand.RandBool()) + Val = Bswap(T(Bswap(Val) + Add)); // Add assuming different endiannes. + else + Val = Val + Add; // Add assuming current endiannes. + if (Add == 0 || Rand.RandBool()) // Maybe negate. + Val = -Val; + } + memcpy(Data + Off, &Val, sizeof(Val)); + return Size; +} + +size_t MutationDispatcher::Mutate_ChangeBinaryInteger(uint8_t *Data, + size_t Size, + size_t MaxSize) { + if (Size > MaxSize) return 0; + switch (Rand(4)) { + case 3: return ChangeBinaryInteger<uint64_t>(Data, Size, Rand); + case 2: return ChangeBinaryInteger<uint32_t>(Data, Size, Rand); + case 1: return ChangeBinaryInteger<uint16_t>(Data, Size, Rand); + case 0: return ChangeBinaryInteger<uint8_t>(Data, Size, Rand); + default: assert(0); + } + return 0; +} + +size_t MutationDispatcher::Mutate_CrossOver(uint8_t *Data, size_t Size, + size_t MaxSize) { + if (Size > MaxSize) return 0; + if (!Corpus || Corpus->size() < 2 || Size == 0) return 0; + size_t Idx = Rand(Corpus->size()); + const Unit &O = (*Corpus)[Idx]; + if (O.empty()) return 0; + MutateInPlaceHere.resize(MaxSize); + auto &U = MutateInPlaceHere; + size_t NewSize = 0; + switch(Rand(3)) { + case 0: + NewSize = CrossOver(Data, Size, O.data(), O.size(), U.data(), U.size()); + break; + case 1: + NewSize = InsertPartOf(O.data(), O.size(), U.data(), U.size(), MaxSize); + if (!NewSize) + NewSize = CopyPartOf(O.data(), O.size(), U.data(), U.size()); + break; + case 2: + NewSize = CopyPartOf(O.data(), O.size(), U.data(), U.size()); + break; + default: assert(0); + } + assert(NewSize > 0 && "CrossOver returned empty unit"); + assert(NewSize <= MaxSize && "CrossOver returned overisized unit"); + memcpy(Data, U.data(), NewSize); + return NewSize; +} + +void MutationDispatcher::StartMutationSequence() { + CurrentMutatorSequence.clear(); + CurrentDictionaryEntrySequence.clear(); +} + +// Copy successful dictionary entries to PersistentAutoDictionary. +void MutationDispatcher::RecordSuccessfulMutationSequence() { + for (auto DE : CurrentDictionaryEntrySequence) { + // PersistentAutoDictionary.AddWithSuccessCountOne(DE); + DE->IncSuccessCount(); + assert(DE->GetW().size()); + // Linear search is fine here as this happens seldom. + if (!PersistentAutoDictionary.ContainsWord(DE->GetW())) + PersistentAutoDictionary.push_back({DE->GetW(), 1}); + } +} + +void MutationDispatcher::PrintRecommendedDictionary() { + Vector<DictionaryEntry> V; + for (auto &DE : PersistentAutoDictionary) + if (!ManualDictionary.ContainsWord(DE.GetW())) + V.push_back(DE); + if (V.empty()) return; + Printf("###### Recommended dictionary. ######\n"); + for (auto &DE: V) { + assert(DE.GetW().size()); + Printf("\""); + PrintASCII(DE.GetW(), "\""); + Printf(" # Uses: %zd\n", DE.GetUseCount()); + } + Printf("###### End of recommended dictionary. ######\n"); +} + +void MutationDispatcher::PrintMutationSequence() { + Printf("MS: %zd ", CurrentMutatorSequence.size()); + for (auto M : CurrentMutatorSequence) + Printf("%s-", M.Name); + if (!CurrentDictionaryEntrySequence.empty()) { + Printf(" DE: "); + for (auto DE : CurrentDictionaryEntrySequence) { + Printf("\""); + PrintASCII(DE->GetW(), "\"-"); + } + } +} + +size_t MutationDispatcher::Mutate(uint8_t *Data, size_t Size, size_t MaxSize) { + return MutateImpl(Data, Size, MaxSize, Mutators); +} + +size_t MutationDispatcher::DefaultMutate(uint8_t *Data, size_t Size, + size_t MaxSize) { + return MutateImpl(Data, Size, MaxSize, DefaultMutators); +} + +// Mutates Data in place, returns new size. +size_t MutationDispatcher::MutateImpl(uint8_t *Data, size_t Size, + size_t MaxSize, + Vector<Mutator> &Mutators) { + assert(MaxSize > 0); + // Some mutations may fail (e.g. can't insert more bytes if Size == MaxSize), + // in which case they will return 0. + // Try several times before returning un-mutated data. + for (int Iter = 0; Iter < 100; Iter++) { + auto M = Mutators[Rand(Mutators.size())]; + size_t NewSize = (this->*(M.Fn))(Data, Size, MaxSize); + if (NewSize && NewSize <= MaxSize) { + if (Options.OnlyASCII) + ToASCII(Data, NewSize); + CurrentMutatorSequence.push_back(M); + return NewSize; + } + } + *Data = ' '; + return 1; // Fallback, should not happen frequently. +} + +void MutationDispatcher::AddWordToManualDictionary(const Word &W) { + ManualDictionary.push_back( + {W, std::numeric_limits<size_t>::max()}); +} + +} // namespace fuzzer diff --git a/lib/fuzzer/FuzzerMutate.h b/lib/fuzzer/FuzzerMutate.h new file mode 100644 index 000000000000..4aa58af9902d --- /dev/null +++ b/lib/fuzzer/FuzzerMutate.h @@ -0,0 +1,150 @@ +//===- FuzzerMutate.h - Internal header for the Fuzzer ----------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// fuzzer::MutationDispatcher +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_MUTATE_H +#define LLVM_FUZZER_MUTATE_H + +#include "FuzzerDefs.h" +#include "FuzzerDictionary.h" +#include "FuzzerOptions.h" +#include "FuzzerRandom.h" + +namespace fuzzer { + +class MutationDispatcher { +public: + MutationDispatcher(Random &Rand, const FuzzingOptions &Options); + ~MutationDispatcher() {} + /// Indicate that we are about to start a new sequence of mutations. + void StartMutationSequence(); + /// Print the current sequence of mutations. + void PrintMutationSequence(); + /// Indicate that the current sequence of mutations was successfull. + void RecordSuccessfulMutationSequence(); + /// Mutates data by invoking user-provided mutator. + size_t Mutate_Custom(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by invoking user-provided crossover. + size_t Mutate_CustomCrossOver(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by shuffling bytes. + size_t Mutate_ShuffleBytes(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by erasing bytes. + size_t Mutate_EraseBytes(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by inserting a byte. + size_t Mutate_InsertByte(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by inserting several repeated bytes. + size_t Mutate_InsertRepeatedBytes(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by chanding one byte. + size_t Mutate_ChangeByte(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by chanding one bit. + size_t Mutate_ChangeBit(uint8_t *Data, size_t Size, size_t MaxSize); + /// Mutates data by copying/inserting a part of data into a different place. + size_t Mutate_CopyPart(uint8_t *Data, size_t Size, size_t MaxSize); + + /// Mutates data by adding a word from the manual dictionary. + size_t Mutate_AddWordFromManualDictionary(uint8_t *Data, size_t Size, + size_t MaxSize); + + /// Mutates data by adding a word from the TORC. + size_t Mutate_AddWordFromTORC(uint8_t *Data, size_t Size, size_t MaxSize); + + /// Mutates data by adding a word from the persistent automatic dictionary. + size_t Mutate_AddWordFromPersistentAutoDictionary(uint8_t *Data, size_t Size, + size_t MaxSize); + + /// Tries to find an ASCII integer in Data, changes it to another ASCII int. + size_t Mutate_ChangeASCIIInteger(uint8_t *Data, size_t Size, size_t MaxSize); + /// Change a 1-, 2-, 4-, or 8-byte integer in interesting ways. + size_t Mutate_ChangeBinaryInteger(uint8_t *Data, size_t Size, size_t MaxSize); + + /// CrossOver Data with some other element of the corpus. + size_t Mutate_CrossOver(uint8_t *Data, size_t Size, size_t MaxSize); + + /// Applies one of the configured mutations. + /// Returns the new size of data which could be up to MaxSize. + size_t Mutate(uint8_t *Data, size_t Size, size_t MaxSize); + /// Applies one of the default mutations. Provided as a service + /// to mutation authors. + size_t DefaultMutate(uint8_t *Data, size_t Size, size_t MaxSize); + + /// Creates a cross-over of two pieces of Data, returns its size. + size_t CrossOver(const uint8_t *Data1, size_t Size1, const uint8_t *Data2, + size_t Size2, uint8_t *Out, size_t MaxOutSize); + + void AddWordToManualDictionary(const Word &W); + + void PrintRecommendedDictionary(); + + void SetCorpus(const InputCorpus *Corpus) { this->Corpus = Corpus; } + + Random &GetRand() { return Rand; } + +private: + + struct Mutator { + size_t (MutationDispatcher::*Fn)(uint8_t *Data, size_t Size, size_t Max); + const char *Name; + }; + + size_t AddWordFromDictionary(Dictionary &D, uint8_t *Data, size_t Size, + size_t MaxSize); + size_t MutateImpl(uint8_t *Data, size_t Size, size_t MaxSize, + Vector<Mutator> &Mutators); + + size_t InsertPartOf(const uint8_t *From, size_t FromSize, uint8_t *To, + size_t ToSize, size_t MaxToSize); + size_t CopyPartOf(const uint8_t *From, size_t FromSize, uint8_t *To, + size_t ToSize); + size_t ApplyDictionaryEntry(uint8_t *Data, size_t Size, size_t MaxSize, + DictionaryEntry &DE); + + template <class T> + DictionaryEntry MakeDictionaryEntryFromCMP(T Arg1, T Arg2, + const uint8_t *Data, size_t Size); + DictionaryEntry MakeDictionaryEntryFromCMP(const Word &Arg1, const Word &Arg2, + const uint8_t *Data, size_t Size); + DictionaryEntry MakeDictionaryEntryFromCMP(const void *Arg1, const void *Arg2, + const void *Arg1Mutation, + const void *Arg2Mutation, + size_t ArgSize, + const uint8_t *Data, size_t Size); + + Random &Rand; + const FuzzingOptions Options; + + // Dictionary provided by the user via -dict=DICT_FILE. + Dictionary ManualDictionary; + // Temporary dictionary modified by the fuzzer itself, + // recreated periodically. + Dictionary TempAutoDictionary; + // Persistent dictionary modified by the fuzzer, consists of + // entries that led to successfull discoveries in the past mutations. + Dictionary PersistentAutoDictionary; + + Vector<Mutator> CurrentMutatorSequence; + Vector<DictionaryEntry *> CurrentDictionaryEntrySequence; + + static const size_t kCmpDictionaryEntriesDequeSize = 16; + DictionaryEntry CmpDictionaryEntriesDeque[kCmpDictionaryEntriesDequeSize]; + size_t CmpDictionaryEntriesDequeIdx = 0; + + const InputCorpus *Corpus = nullptr; + Vector<uint8_t> MutateInPlaceHere; + // CustomCrossOver needs its own buffer as a custom implementation may call + // LLVMFuzzerMutate, which in turn may resize MutateInPlaceHere. + Vector<uint8_t> CustomCrossOverInPlaceHere; + + Vector<Mutator> Mutators; + Vector<Mutator> DefaultMutators; +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_MUTATE_H diff --git a/lib/fuzzer/FuzzerOptions.h b/lib/fuzzer/FuzzerOptions.h new file mode 100644 index 000000000000..15a378020b85 --- /dev/null +++ b/lib/fuzzer/FuzzerOptions.h @@ -0,0 +1,75 @@ +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// fuzzer::FuzzingOptions +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_OPTIONS_H +#define LLVM_FUZZER_OPTIONS_H + +#include "FuzzerDefs.h" + +namespace fuzzer { + +struct FuzzingOptions { + int Verbosity = 1; + size_t MaxLen = 0; + size_t ExperimentalLenControl = 0; + int UnitTimeoutSec = 300; + int TimeoutExitCode = 77; + int ErrorExitCode = 77; + int MaxTotalTimeSec = 0; + int RssLimitMb = 0; + int MallocLimitMb = 0; + bool DoCrossOver = true; + int MutateDepth = 5; + bool ReduceDepth = false; + bool UseCounters = false; + bool UseMemmem = true; + bool UseCmp = false; + bool UseValueProfile = false; + bool Shrink = false; + bool ReduceInputs = false; + int ReloadIntervalSec = 1; + bool ShuffleAtStartUp = true; + bool PreferSmall = true; + size_t MaxNumberOfRuns = -1L; + int ReportSlowUnits = 10; + bool OnlyASCII = false; + std::string OutputCorpus; + std::string ArtifactPrefix = "./"; + std::string ExactArtifactPath; + std::string ExitOnSrcPos; + std::string ExitOnItem; + bool SaveArtifacts = true; + bool PrintNEW = true; // Print a status line when new units are found; + bool PrintNewCovPcs = false; + int PrintNewCovFuncs = 0; + bool PrintFinalStats = false; + bool PrintCorpusStats = false; + bool PrintCoverage = false; + bool DumpCoverage = false; + bool UseClangCoverage = false; + bool DetectLeaks = true; + int PurgeAllocatorIntervalSec = 1; + int UseFeatureFrequency = false; + int TraceMalloc = 0; + bool HandleAbrt = false; + bool HandleBus = false; + bool HandleFpe = false; + bool HandleIll = false; + bool HandleInt = false; + bool HandleSegv = false; + bool HandleTerm = false; + bool HandleXfsz = false; + bool HandleUsr1 = false; + bool HandleUsr2 = false; +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_OPTIONS_H diff --git a/lib/fuzzer/FuzzerRandom.h b/lib/fuzzer/FuzzerRandom.h new file mode 100644 index 000000000000..8a1aa3ef5fdc --- /dev/null +++ b/lib/fuzzer/FuzzerRandom.h @@ -0,0 +1,34 @@ +//===- FuzzerRandom.h - Internal header for the Fuzzer ----------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// fuzzer::Random +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_RANDOM_H +#define LLVM_FUZZER_RANDOM_H + +#include <random> + +namespace fuzzer { +class Random : public std::mt19937 { + public: + Random(unsigned int seed) : std::mt19937(seed) {} + result_type operator()() { return this->std::mt19937::operator()(); } + size_t Rand() { return this->operator()(); } + size_t RandBool() { return Rand() % 2; } + size_t operator()(size_t n) { return n ? Rand() % n : 0; } + intptr_t operator()(intptr_t From, intptr_t To) { + assert(From < To); + intptr_t RangeSize = To - From + 1; + return operator()(RangeSize) + From; + } +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_RANDOM_H diff --git a/lib/fuzzer/FuzzerSHA1.cpp b/lib/fuzzer/FuzzerSHA1.cpp new file mode 100644 index 000000000000..d2f8e811bbf8 --- /dev/null +++ b/lib/fuzzer/FuzzerSHA1.cpp @@ -0,0 +1,222 @@ +//===- FuzzerSHA1.h - Private copy of the SHA1 implementation ---*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// This code is taken from public domain +// (http://oauth.googlecode.com/svn/code/c/liboauth/src/sha1.c) +// and modified by adding anonymous namespace, adding an interface +// function fuzzer::ComputeSHA1() and removing unnecessary code. +// +// lib/Fuzzer can not use SHA1 implementation from openssl because +// openssl may not be available and because we may be fuzzing openssl itself. +// For the same reason we do not want to depend on SHA1 from LLVM tree. +//===----------------------------------------------------------------------===// + +#include "FuzzerSHA1.h" +#include "FuzzerDefs.h" + +/* This code is public-domain - it is based on libcrypt + * placed in the public domain by Wei Dai and other contributors. + */ + +#include <iomanip> +#include <sstream> +#include <stdint.h> +#include <string.h> + +namespace { // Added for LibFuzzer + +#ifdef __BIG_ENDIAN__ +# define SHA_BIG_ENDIAN +#elif defined __LITTLE_ENDIAN__ +/* override */ +#elif defined __BYTE_ORDER +# if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ +# define SHA_BIG_ENDIAN +# endif +#else // ! defined __LITTLE_ENDIAN__ +# include <endian.h> // machine/endian.h +# if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ +# define SHA_BIG_ENDIAN +# endif +#endif + + +/* header */ + +#define HASH_LENGTH 20 +#define BLOCK_LENGTH 64 + +typedef struct sha1nfo { + uint32_t buffer[BLOCK_LENGTH/4]; + uint32_t state[HASH_LENGTH/4]; + uint32_t byteCount; + uint8_t bufferOffset; + uint8_t keyBuffer[BLOCK_LENGTH]; + uint8_t innerHash[HASH_LENGTH]; +} sha1nfo; + +/* public API - prototypes - TODO: doxygen*/ + +/** + */ +void sha1_init(sha1nfo *s); +/** + */ +void sha1_writebyte(sha1nfo *s, uint8_t data); +/** + */ +void sha1_write(sha1nfo *s, const char *data, size_t len); +/** + */ +uint8_t* sha1_result(sha1nfo *s); + + +/* code */ +#define SHA1_K0 0x5a827999 +#define SHA1_K20 0x6ed9eba1 +#define SHA1_K40 0x8f1bbcdc +#define SHA1_K60 0xca62c1d6 + +void sha1_init(sha1nfo *s) { + s->state[0] = 0x67452301; + s->state[1] = 0xefcdab89; + s->state[2] = 0x98badcfe; + s->state[3] = 0x10325476; + s->state[4] = 0xc3d2e1f0; + s->byteCount = 0; + s->bufferOffset = 0; +} + +uint32_t sha1_rol32(uint32_t number, uint8_t bits) { + return ((number << bits) | (number >> (32-bits))); +} + +void sha1_hashBlock(sha1nfo *s) { + uint8_t i; + uint32_t a,b,c,d,e,t; + + a=s->state[0]; + b=s->state[1]; + c=s->state[2]; + d=s->state[3]; + e=s->state[4]; + for (i=0; i<80; i++) { + if (i>=16) { + t = s->buffer[(i+13)&15] ^ s->buffer[(i+8)&15] ^ s->buffer[(i+2)&15] ^ s->buffer[i&15]; + s->buffer[i&15] = sha1_rol32(t,1); + } + if (i<20) { + t = (d ^ (b & (c ^ d))) + SHA1_K0; + } else if (i<40) { + t = (b ^ c ^ d) + SHA1_K20; + } else if (i<60) { + t = ((b & c) | (d & (b | c))) + SHA1_K40; + } else { + t = (b ^ c ^ d) + SHA1_K60; + } + t+=sha1_rol32(a,5) + e + s->buffer[i&15]; + e=d; + d=c; + c=sha1_rol32(b,30); + b=a; + a=t; + } + s->state[0] += a; + s->state[1] += b; + s->state[2] += c; + s->state[3] += d; + s->state[4] += e; +} + +void sha1_addUncounted(sha1nfo *s, uint8_t data) { + uint8_t * const b = (uint8_t*) s->buffer; +#ifdef SHA_BIG_ENDIAN + b[s->bufferOffset] = data; +#else + b[s->bufferOffset ^ 3] = data; +#endif + s->bufferOffset++; + if (s->bufferOffset == BLOCK_LENGTH) { + sha1_hashBlock(s); + s->bufferOffset = 0; + } +} + +void sha1_writebyte(sha1nfo *s, uint8_t data) { + ++s->byteCount; + sha1_addUncounted(s, data); +} + +void sha1_write(sha1nfo *s, const char *data, size_t len) { + for (;len--;) sha1_writebyte(s, (uint8_t) *data++); +} + +void sha1_pad(sha1nfo *s) { + // Implement SHA-1 padding (fips180-2 §5.1.1) + + // Pad with 0x80 followed by 0x00 until the end of the block + sha1_addUncounted(s, 0x80); + while (s->bufferOffset != 56) sha1_addUncounted(s, 0x00); + + // Append length in the last 8 bytes + sha1_addUncounted(s, 0); // We're only using 32 bit lengths + sha1_addUncounted(s, 0); // But SHA-1 supports 64 bit lengths + sha1_addUncounted(s, 0); // So zero pad the top bits + sha1_addUncounted(s, s->byteCount >> 29); // Shifting to multiply by 8 + sha1_addUncounted(s, s->byteCount >> 21); // as SHA-1 supports bitstreams as well as + sha1_addUncounted(s, s->byteCount >> 13); // byte. + sha1_addUncounted(s, s->byteCount >> 5); + sha1_addUncounted(s, s->byteCount << 3); +} + +uint8_t* sha1_result(sha1nfo *s) { + // Pad to complete the last block + sha1_pad(s); + +#ifndef SHA_BIG_ENDIAN + // Swap byte order back + int i; + for (i=0; i<5; i++) { + s->state[i]= + (((s->state[i])<<24)& 0xff000000) + | (((s->state[i])<<8) & 0x00ff0000) + | (((s->state[i])>>8) & 0x0000ff00) + | (((s->state[i])>>24)& 0x000000ff); + } +#endif + + // Return pointer to hash (20 characters) + return (uint8_t*) s->state; +} + +} // namespace; Added for LibFuzzer + +namespace fuzzer { + +// The rest is added for LibFuzzer +void ComputeSHA1(const uint8_t *Data, size_t Len, uint8_t *Out) { + sha1nfo s; + sha1_init(&s); + sha1_write(&s, (const char*)Data, Len); + memcpy(Out, sha1_result(&s), HASH_LENGTH); +} + +std::string Sha1ToString(const uint8_t Sha1[kSHA1NumBytes]) { + std::stringstream SS; + for (int i = 0; i < kSHA1NumBytes; i++) + SS << std::hex << std::setfill('0') << std::setw(2) << (unsigned)Sha1[i]; + return SS.str(); +} + +std::string Hash(const Unit &U) { + uint8_t Hash[kSHA1NumBytes]; + ComputeSHA1(U.data(), U.size(), Hash); + return Sha1ToString(Hash); +} + +} diff --git a/lib/fuzzer/FuzzerSHA1.h b/lib/fuzzer/FuzzerSHA1.h new file mode 100644 index 000000000000..3b5e6e807f42 --- /dev/null +++ b/lib/fuzzer/FuzzerSHA1.h @@ -0,0 +1,33 @@ +//===- FuzzerSHA1.h - Internal header for the SHA1 utils --------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// SHA1 utils. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_SHA1_H +#define LLVM_FUZZER_SHA1_H + +#include "FuzzerDefs.h" +#include <cstddef> +#include <stdint.h> + +namespace fuzzer { + +// Private copy of SHA1 implementation. +static const int kSHA1NumBytes = 20; + +// Computes SHA1 hash of 'Len' bytes in 'Data', writes kSHA1NumBytes to 'Out'. +void ComputeSHA1(const uint8_t *Data, size_t Len, uint8_t *Out); + +std::string Sha1ToString(const uint8_t Sha1[kSHA1NumBytes]); + +std::string Hash(const Unit &U); + +} // namespace fuzzer + +#endif // LLVM_FUZZER_SHA1_H diff --git a/lib/fuzzer/FuzzerShmem.h b/lib/fuzzer/FuzzerShmem.h new file mode 100644 index 000000000000..53568e0acb69 --- /dev/null +++ b/lib/fuzzer/FuzzerShmem.h @@ -0,0 +1,69 @@ +//===- FuzzerShmem.h - shared memory interface ------------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// SharedMemoryRegion +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_SHMEM_H +#define LLVM_FUZZER_SHMEM_H + +#include <algorithm> +#include <cstring> +#include <string> + +#include "FuzzerDefs.h" + +namespace fuzzer { + +class SharedMemoryRegion { + public: + bool Create(const char *Name); + bool Open(const char *Name); + bool Destroy(const char *Name); + uint8_t *GetData() { return Data; } + void PostServer() {Post(0);} + void WaitServer() {Wait(0);} + void PostClient() {Post(1);} + void WaitClient() {Wait(1);} + + size_t WriteByteArray(const uint8_t *Bytes, size_t N) { + assert(N <= kShmemSize - sizeof(N)); + memcpy(GetData(), &N, sizeof(N)); + memcpy(GetData() + sizeof(N), Bytes, N); + assert(N == ReadByteArraySize()); + return N; + } + size_t ReadByteArraySize() { + size_t Res; + memcpy(&Res, GetData(), sizeof(Res)); + return Res; + } + uint8_t *GetByteArray() { return GetData() + sizeof(size_t); } + + bool IsServer() const { return Data && IAmServer; } + bool IsClient() const { return Data && !IAmServer; } + +private: + + static const size_t kShmemSize = 1 << 22; + bool IAmServer; + std::string Path(const char *Name); + std::string SemName(const char *Name, int Idx); + void Post(int Idx); + void Wait(int Idx); + + bool Map(int fd); + uint8_t *Data = nullptr; + void *Semaphore[2]; +}; + +extern SharedMemoryRegion SMR; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_SHMEM_H diff --git a/lib/fuzzer/FuzzerShmemFuchsia.cpp b/lib/fuzzer/FuzzerShmemFuchsia.cpp new file mode 100644 index 000000000000..e9ce50c2ac8a --- /dev/null +++ b/lib/fuzzer/FuzzerShmemFuchsia.cpp @@ -0,0 +1,38 @@ +//===- FuzzerShmemPosix.cpp - Posix shared memory ---------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// SharedMemoryRegion. For Fuchsia, this is just stubs as equivalence servers +// are not currently supported. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" + +#if LIBFUZZER_FUCHSIA + +#include "FuzzerShmem.h" + +namespace fuzzer { + +bool SharedMemoryRegion::Create(const char *Name) { + return false; +} + +bool SharedMemoryRegion::Open(const char *Name) { + return false; +} + +bool SharedMemoryRegion::Destroy(const char *Name) { + return false; +} + +void SharedMemoryRegion::Post(int Idx) {} + +void SharedMemoryRegion::Wait(int Idx) {} + +} // namespace fuzzer + +#endif // LIBFUZZER_FUCHSIA diff --git a/lib/fuzzer/FuzzerShmemPosix.cpp b/lib/fuzzer/FuzzerShmemPosix.cpp new file mode 100644 index 000000000000..50cdcfb509dc --- /dev/null +++ b/lib/fuzzer/FuzzerShmemPosix.cpp @@ -0,0 +1,103 @@ +//===- FuzzerShmemPosix.cpp - Posix shared memory ---------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// SharedMemoryRegion +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_POSIX + +#include "FuzzerIO.h" +#include "FuzzerShmem.h" + +#include <errno.h> +#include <fcntl.h> +#include <semaphore.h> +#include <stdio.h> +#include <stdlib.h> +#include <sys/mman.h> +#include <sys/stat.h> +#include <sys/types.h> +#include <unistd.h> + +namespace fuzzer { + +std::string SharedMemoryRegion::Path(const char *Name) { + return DirPlusFile(TmpDir(), Name); +} + +std::string SharedMemoryRegion::SemName(const char *Name, int Idx) { + std::string Res(Name); + return Res + (char)('0' + Idx); +} + +bool SharedMemoryRegion::Map(int fd) { + Data = + (uint8_t *)mmap(0, kShmemSize, PROT_WRITE | PROT_READ, MAP_SHARED, fd, 0); + if (Data == (uint8_t*)-1) + return false; + return true; +} + +bool SharedMemoryRegion::Create(const char *Name) { + int fd = open(Path(Name).c_str(), O_CREAT | O_RDWR, 0777); + if (fd < 0) return false; + if (ftruncate(fd, kShmemSize) < 0) return false; + if (!Map(fd)) + return false; + for (int i = 0; i < 2; i++) { + sem_unlink(SemName(Name, i).c_str()); + Semaphore[i] = sem_open(SemName(Name, i).c_str(), O_CREAT, 0644, 0); + if (Semaphore[i] == (void *)-1) + return false; + } + IAmServer = true; + return true; +} + +bool SharedMemoryRegion::Open(const char *Name) { + int fd = open(Path(Name).c_str(), O_RDWR); + if (fd < 0) return false; + struct stat stat_res; + if (0 != fstat(fd, &stat_res)) + return false; + assert(stat_res.st_size == kShmemSize); + if (!Map(fd)) + return false; + for (int i = 0; i < 2; i++) { + Semaphore[i] = sem_open(SemName(Name, i).c_str(), 0); + if (Semaphore[i] == (void *)-1) + return false; + } + IAmServer = false; + return true; +} + +bool SharedMemoryRegion::Destroy(const char *Name) { + return 0 == unlink(Path(Name).c_str()); +} + +void SharedMemoryRegion::Post(int Idx) { + assert(Idx == 0 || Idx == 1); + sem_post((sem_t*)Semaphore[Idx]); +} + +void SharedMemoryRegion::Wait(int Idx) { + assert(Idx == 0 || Idx == 1); + for (int i = 0; i < 10 && sem_wait((sem_t*)Semaphore[Idx]); i++) { + // sem_wait may fail if interrupted by a signal. + sleep(i); + if (i) + Printf("%s: sem_wait[%d] failed %s\n", i < 9 ? "WARNING" : "ERROR", i, + strerror(errno)); + if (i == 9) abort(); + } +} + +} // namespace fuzzer + +#endif // LIBFUZZER_POSIX diff --git a/lib/fuzzer/FuzzerShmemWindows.cpp b/lib/fuzzer/FuzzerShmemWindows.cpp new file mode 100644 index 000000000000..d330ebf4fd07 --- /dev/null +++ b/lib/fuzzer/FuzzerShmemWindows.cpp @@ -0,0 +1,64 @@ +//===- FuzzerShmemWindows.cpp - Posix shared memory -------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// SharedMemoryRegion +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_WINDOWS + +#include "FuzzerIO.h" +#include "FuzzerShmem.h" + +#include <fcntl.h> +#include <stdio.h> +#include <sys/stat.h> +#include <sys/types.h> + +namespace fuzzer { + +std::string SharedMemoryRegion::Path(const char *Name) { + return DirPlusFile(TmpDir(), Name); +} + +std::string SharedMemoryRegion::SemName(const char *Name, int Idx) { + std::string Res(Name); + return Res + (char)('0' + Idx); +} + +bool SharedMemoryRegion::Map(int fd) { + assert(0 && "UNIMPLEMENTED"); + return false; +} + +bool SharedMemoryRegion::Create(const char *Name) { + assert(0 && "UNIMPLEMENTED"); + return false; +} + +bool SharedMemoryRegion::Open(const char *Name) { + assert(0 && "UNIMPLEMENTED"); + return false; +} + +bool SharedMemoryRegion::Destroy(const char *Name) { + assert(0 && "UNIMPLEMENTED"); + return false; +} + +void SharedMemoryRegion::Post(int Idx) { + assert(0 && "UNIMPLEMENTED"); +} + +void SharedMemoryRegion::Wait(int Idx) { + Semaphore[1] = nullptr; + assert(0 && "UNIMPLEMENTED"); +} + +} // namespace fuzzer + +#endif // LIBFUZZER_WINDOWS diff --git a/lib/fuzzer/FuzzerTracePC.cpp b/lib/fuzzer/FuzzerTracePC.cpp new file mode 100644 index 000000000000..5e9f9f2f6dcc --- /dev/null +++ b/lib/fuzzer/FuzzerTracePC.cpp @@ -0,0 +1,602 @@ +//===- FuzzerTracePC.cpp - PC tracing--------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Trace PCs. +// This module implements __sanitizer_cov_trace_pc_guard[_init], +// the callback required for -fsanitize-coverage=trace-pc-guard instrumentation. +// +//===----------------------------------------------------------------------===// + +#include "FuzzerTracePC.h" +#include "FuzzerCorpus.h" +#include "FuzzerDefs.h" +#include "FuzzerDictionary.h" +#include "FuzzerExtFunctions.h" +#include "FuzzerIO.h" +#include "FuzzerUtil.h" +#include "FuzzerValueBitMap.h" +#include <set> + +// The coverage counters and PCs. +// These are declared as global variables named "__sancov_*" to simplify +// experiments with inlined instrumentation. +alignas(64) ATTRIBUTE_INTERFACE +uint8_t __sancov_trace_pc_guard_8bit_counters[fuzzer::TracePC::kNumPCs]; + +ATTRIBUTE_INTERFACE +uintptr_t __sancov_trace_pc_pcs[fuzzer::TracePC::kNumPCs]; + +// Used by -fsanitize-coverage=stack-depth to track stack depth +ATTRIBUTE_INTERFACE __attribute__((tls_model("initial-exec"))) +thread_local uintptr_t __sancov_lowest_stack; + +namespace fuzzer { + +TracePC TPC; + +int ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr; + +uint8_t *TracePC::Counters() const { + return __sancov_trace_pc_guard_8bit_counters; +} + +uintptr_t *TracePC::PCs() const { + return __sancov_trace_pc_pcs; +} + +size_t TracePC::GetTotalPCCoverage() { + if (ObservedPCs.size()) + return ObservedPCs.size(); + size_t Res = 0; + for (size_t i = 1, N = GetNumPCs(); i < N; i++) + if (PCs()[i]) + Res++; + return Res; +} + + +void TracePC::HandleInline8bitCountersInit(uint8_t *Start, uint8_t *Stop) { + if (Start == Stop) return; + if (NumModulesWithInline8bitCounters && + ModuleCounters[NumModulesWithInline8bitCounters-1].Start == Start) return; + assert(NumModulesWithInline8bitCounters < + sizeof(ModuleCounters) / sizeof(ModuleCounters[0])); + ModuleCounters[NumModulesWithInline8bitCounters++] = {Start, Stop}; + NumInline8bitCounters += Stop - Start; +} + +void TracePC::HandlePCsInit(const uintptr_t *Start, const uintptr_t *Stop) { + const PCTableEntry *B = reinterpret_cast<const PCTableEntry *>(Start); + const PCTableEntry *E = reinterpret_cast<const PCTableEntry *>(Stop); + if (NumPCTables && ModulePCTable[NumPCTables - 1].Start == B) return; + assert(NumPCTables < sizeof(ModulePCTable) / sizeof(ModulePCTable[0])); + ModulePCTable[NumPCTables++] = {B, E}; + NumPCsInPCTables += E - B; +} + +void TracePC::HandleInit(uint32_t *Start, uint32_t *Stop) { + if (Start == Stop || *Start) return; + assert(NumModules < sizeof(Modules) / sizeof(Modules[0])); + for (uint32_t *P = Start; P < Stop; P++) { + NumGuards++; + if (NumGuards == kNumPCs) { + RawPrint( + "WARNING: The binary has too many instrumented PCs.\n" + " You may want to reduce the size of the binary\n" + " for more efficient fuzzing and precise coverage data\n"); + } + *P = NumGuards % kNumPCs; + } + Modules[NumModules].Start = Start; + Modules[NumModules].Stop = Stop; + NumModules++; +} + +void TracePC::PrintModuleInfo() { + if (NumGuards) { + Printf("INFO: Loaded %zd modules (%zd guards): ", NumModules, NumGuards); + for (size_t i = 0; i < NumModules; i++) + Printf("%zd [%p, %p), ", Modules[i].Stop - Modules[i].Start, + Modules[i].Start, Modules[i].Stop); + Printf("\n"); + } + if (NumModulesWithInline8bitCounters) { + Printf("INFO: Loaded %zd modules (%zd inline 8-bit counters): ", + NumModulesWithInline8bitCounters, NumInline8bitCounters); + for (size_t i = 0; i < NumModulesWithInline8bitCounters; i++) + Printf("%zd [%p, %p), ", ModuleCounters[i].Stop - ModuleCounters[i].Start, + ModuleCounters[i].Start, ModuleCounters[i].Stop); + Printf("\n"); + } + if (NumPCTables) { + Printf("INFO: Loaded %zd PC tables (%zd PCs): ", NumPCTables, + NumPCsInPCTables); + for (size_t i = 0; i < NumPCTables; i++) { + Printf("%zd [%p,%p), ", ModulePCTable[i].Stop - ModulePCTable[i].Start, + ModulePCTable[i].Start, ModulePCTable[i].Stop); + } + Printf("\n"); + + if ((NumGuards && NumGuards != NumPCsInPCTables) || + (NumInline8bitCounters && NumInline8bitCounters != NumPCsInPCTables)) { + Printf("ERROR: The size of coverage PC tables does not match the\n" + "number of instrumented PCs. This might be a compiler bug,\n" + "please contact the libFuzzer developers.\n" + "Also check https://bugs.llvm.org/show_bug.cgi?id=34636\n" + "for possible workarounds (tl;dr: don't use the old GNU ld)\n"); + _Exit(1); + } + } + if (size_t NumClangCounters = ClangCountersEnd() - ClangCountersBegin()) + Printf("INFO: %zd Clang Coverage Counters\n", NumClangCounters); +} + +ATTRIBUTE_NO_SANITIZE_ALL +void TracePC::HandleCallerCallee(uintptr_t Caller, uintptr_t Callee) { + const uintptr_t kBits = 12; + const uintptr_t kMask = (1 << kBits) - 1; + uintptr_t Idx = (Caller & kMask) | ((Callee & kMask) << kBits); + ValueProfileMap.AddValueModPrime(Idx); +} + +void TracePC::UpdateObservedPCs() { + Vector<uintptr_t> CoveredFuncs; + auto ObservePC = [&](uintptr_t PC) { + if (ObservedPCs.insert(PC).second && DoPrintNewPCs) + PrintPC("\tNEW_PC: %p %F %L\n", "\tNEW_PC: %p\n", PC + 1); + }; + + auto Observe = [&](const PCTableEntry &TE) { + if (TE.PCFlags & 1) + if (ObservedFuncs.insert(TE.PC).second && NumPrintNewFuncs) + CoveredFuncs.push_back(TE.PC); + ObservePC(TE.PC); + }; + + if (NumPCsInPCTables) { + if (NumInline8bitCounters == NumPCsInPCTables) { + for (size_t i = 0; i < NumModulesWithInline8bitCounters; i++) { + uint8_t *Beg = ModuleCounters[i].Start; + size_t Size = ModuleCounters[i].Stop - Beg; + assert(Size == + (size_t)(ModulePCTable[i].Stop - ModulePCTable[i].Start)); + for (size_t j = 0; j < Size; j++) + if (Beg[j]) + Observe(ModulePCTable[i].Start[j]); + } + } else if (NumGuards == NumPCsInPCTables) { + size_t GuardIdx = 1; + for (size_t i = 0; i < NumModules; i++) { + uint32_t *Beg = Modules[i].Start; + size_t Size = Modules[i].Stop - Beg; + assert(Size == + (size_t)(ModulePCTable[i].Stop - ModulePCTable[i].Start)); + for (size_t j = 0; j < Size; j++, GuardIdx++) + if (Counters()[GuardIdx]) + Observe(ModulePCTable[i].Start[j]); + } + } + } + if (size_t NumClangCounters = + ClangCountersEnd() - ClangCountersBegin()) { + auto P = ClangCountersBegin(); + for (size_t Idx = 0; Idx < NumClangCounters; Idx++) + if (P[Idx]) + ObservePC((uintptr_t)Idx); + } + + for (size_t i = 0, N = Min(CoveredFuncs.size(), NumPrintNewFuncs); i < N; i++) { + Printf("\tNEW_FUNC[%zd/%zd]: ", i, CoveredFuncs.size()); + PrintPC("%p %F %L\n", "%p\n", CoveredFuncs[i] + 1); + } +} + +inline ALWAYS_INLINE uintptr_t GetPreviousInstructionPc(uintptr_t PC) { + // TODO: this implementation is x86 only. + // see sanitizer_common GetPreviousInstructionPc for full implementation. + return PC - 1; +} + +inline ALWAYS_INLINE uintptr_t GetNextInstructionPc(uintptr_t PC) { + // TODO: this implementation is x86 only. + // see sanitizer_common GetPreviousInstructionPc for full implementation. + return PC + 1; +} + +static std::string GetModuleName(uintptr_t PC) { + char ModulePathRaw[4096] = ""; // What's PATH_MAX in portable C++? + void *OffsetRaw = nullptr; + if (!EF->__sanitizer_get_module_and_offset_for_pc( + reinterpret_cast<void *>(PC), ModulePathRaw, + sizeof(ModulePathRaw), &OffsetRaw)) + return ""; + return ModulePathRaw; +} + +void TracePC::PrintCoverage() { + if (!EF->__sanitizer_symbolize_pc || + !EF->__sanitizer_get_module_and_offset_for_pc) { + Printf("INFO: __sanitizer_symbolize_pc or " + "__sanitizer_get_module_and_offset_for_pc is not available," + " not printing coverage\n"); + return; + } + Printf("COVERAGE:\n"); + std::string LastFunctionName = ""; + std::string LastFileStr = ""; + Set<size_t> UncoveredLines; + Set<size_t> CoveredLines; + + auto FunctionEndCallback = [&](const std::string &CurrentFunc, + const std::string &CurrentFile) { + if (LastFunctionName != CurrentFunc) { + if (CoveredLines.empty() && !UncoveredLines.empty()) { + Printf("UNCOVERED_FUNC: %s\n", LastFunctionName.c_str()); + } else { + for (auto Line : UncoveredLines) { + if (!CoveredLines.count(Line)) + Printf("UNCOVERED_LINE: %s %s:%zd\n", LastFunctionName.c_str(), + LastFileStr.c_str(), Line); + } + } + + UncoveredLines.clear(); + CoveredLines.clear(); + LastFunctionName = CurrentFunc; + LastFileStr = CurrentFile; + } + }; + + for (size_t i = 0; i < NumPCTables; i++) { + auto &M = ModulePCTable[i]; + assert(M.Start < M.Stop); + auto ModuleName = GetModuleName(M.Start->PC); + for (auto Ptr = M.Start; Ptr < M.Stop; Ptr++) { + auto PC = Ptr->PC; + auto VisualizePC = GetNextInstructionPc(PC); + bool IsObserved = ObservedPCs.count(PC); + std::string FileStr = DescribePC("%s", VisualizePC); + if (!IsInterestingCoverageFile(FileStr)) continue; + std::string FunctionStr = DescribePC("%F", VisualizePC); + FunctionEndCallback(FunctionStr, FileStr); + std::string LineStr = DescribePC("%l", VisualizePC); + size_t Line = std::stoul(LineStr); + if (IsObserved && CoveredLines.insert(Line).second) + Printf("COVERED: %s %s:%zd\n", FunctionStr.c_str(), FileStr.c_str(), + Line); + else + UncoveredLines.insert(Line); + } + } + FunctionEndCallback("", ""); +} + +void TracePC::DumpCoverage() { + if (EF->__sanitizer_dump_coverage) { + Vector<uintptr_t> PCsCopy(GetNumPCs()); + for (size_t i = 0; i < GetNumPCs(); i++) + PCsCopy[i] = PCs()[i] ? GetPreviousInstructionPc(PCs()[i]) : 0; + EF->__sanitizer_dump_coverage(PCsCopy.data(), PCsCopy.size()); + } +} + +// Value profile. +// We keep track of various values that affect control flow. +// These values are inserted into a bit-set-based hash map. +// Every new bit in the map is treated as a new coverage. +// +// For memcmp/strcmp/etc the interesting value is the length of the common +// prefix of the parameters. +// For cmp instructions the interesting value is a XOR of the parameters. +// The interesting value is mixed up with the PC and is then added to the map. + +ATTRIBUTE_NO_SANITIZE_ALL +void TracePC::AddValueForMemcmp(void *caller_pc, const void *s1, const void *s2, + size_t n, bool StopAtZero) { + if (!n) return; + size_t Len = std::min(n, Word::GetMaxSize()); + const uint8_t *A1 = reinterpret_cast<const uint8_t *>(s1); + const uint8_t *A2 = reinterpret_cast<const uint8_t *>(s2); + uint8_t B1[Word::kMaxSize]; + uint8_t B2[Word::kMaxSize]; + // Copy the data into locals in this non-msan-instrumented function + // to avoid msan complaining further. + size_t Hash = 0; // Compute some simple hash of both strings. + for (size_t i = 0; i < Len; i++) { + B1[i] = A1[i]; + B2[i] = A2[i]; + size_t T = B1[i]; + Hash ^= (T << 8) | B2[i]; + } + size_t I = 0; + for (; I < Len; I++) + if (B1[I] != B2[I] || (StopAtZero && B1[I] == 0)) + break; + size_t PC = reinterpret_cast<size_t>(caller_pc); + size_t Idx = (PC & 4095) | (I << 12); + ValueProfileMap.AddValue(Idx); + TORCW.Insert(Idx ^ Hash, Word(B1, Len), Word(B2, Len)); +} + +template <class T> +ATTRIBUTE_TARGET_POPCNT ALWAYS_INLINE +ATTRIBUTE_NO_SANITIZE_ALL +void TracePC::HandleCmp(uintptr_t PC, T Arg1, T Arg2) { + uint64_t ArgXor = Arg1 ^ Arg2; + uint64_t ArgDistance = __builtin_popcountll(ArgXor) + 1; // [1,65] + uintptr_t Idx = ((PC & 4095) + 1) * ArgDistance; + if (sizeof(T) == 4) + TORC4.Insert(ArgXor, Arg1, Arg2); + else if (sizeof(T) == 8) + TORC8.Insert(ArgXor, Arg1, Arg2); + ValueProfileMap.AddValue(Idx); +} + +static size_t InternalStrnlen(const char *S, size_t MaxLen) { + size_t Len = 0; + for (; Len < MaxLen && S[Len]; Len++) {} + return Len; +} + +// Finds min of (strlen(S1), strlen(S2)). +// Needed bacause one of these strings may actually be non-zero terminated. +static size_t InternalStrnlen2(const char *S1, const char *S2) { + size_t Len = 0; + for (; S1[Len] && S2[Len]; Len++) {} + return Len; +} + +void TracePC::ClearInlineCounters() { + for (size_t i = 0; i < NumModulesWithInline8bitCounters; i++) { + uint8_t *Beg = ModuleCounters[i].Start; + size_t Size = ModuleCounters[i].Stop - Beg; + memset(Beg, 0, Size); + } +} + +ATTRIBUTE_NO_SANITIZE_ALL +void TracePC::RecordInitialStack() { + int stack; + __sancov_lowest_stack = InitialStack = reinterpret_cast<uintptr_t>(&stack); +} + +uintptr_t TracePC::GetMaxStackOffset() const { + return InitialStack - __sancov_lowest_stack; // Stack grows down +} + +} // namespace fuzzer + +extern "C" { +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +void __sanitizer_cov_trace_pc_guard(uint32_t *Guard) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + uint32_t Idx = *Guard; + __sancov_trace_pc_pcs[Idx] = PC; + __sancov_trace_pc_guard_8bit_counters[Idx]++; +} + +// Best-effort support for -fsanitize-coverage=trace-pc, which is available +// in both Clang and GCC. +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +void __sanitizer_cov_trace_pc() { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + uintptr_t Idx = PC & (((uintptr_t)1 << fuzzer::TracePC::kTracePcBits) - 1); + __sancov_trace_pc_pcs[Idx] = PC; + __sancov_trace_pc_guard_8bit_counters[Idx]++; +} + +ATTRIBUTE_INTERFACE +void __sanitizer_cov_trace_pc_guard_init(uint32_t *Start, uint32_t *Stop) { + fuzzer::TPC.HandleInit(Start, Stop); +} + +ATTRIBUTE_INTERFACE +void __sanitizer_cov_8bit_counters_init(uint8_t *Start, uint8_t *Stop) { + fuzzer::TPC.HandleInline8bitCountersInit(Start, Stop); +} + +ATTRIBUTE_INTERFACE +void __sanitizer_cov_pcs_init(const uintptr_t *pcs_beg, + const uintptr_t *pcs_end) { + fuzzer::TPC.HandlePCsInit(pcs_beg, pcs_end); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +void __sanitizer_cov_trace_pc_indir(uintptr_t Callee) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCallerCallee(PC, Callee); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_cmp8(uint64_t Arg1, uint64_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +// Now the __sanitizer_cov_trace_const_cmp[1248] callbacks just mimic +// the behaviour of __sanitizer_cov_trace_cmp[1248] ones. This, however, +// should be changed later to make full use of instrumentation. +void __sanitizer_cov_trace_const_cmp8(uint64_t Arg1, uint64_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_cmp4(uint32_t Arg1, uint32_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_const_cmp4(uint32_t Arg1, uint32_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_cmp2(uint16_t Arg1, uint16_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_const_cmp2(uint16_t Arg1, uint16_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_cmp1(uint8_t Arg1, uint8_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_const_cmp1(uint8_t Arg1, uint8_t Arg2) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Arg1, Arg2); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_switch(uint64_t Val, uint64_t *Cases) { + uint64_t N = Cases[0]; + uint64_t ValSizeInBits = Cases[1]; + uint64_t *Vals = Cases + 2; + // Skip the most common and the most boring case. + if (Vals[N - 1] < 256 && Val < 256) + return; + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + size_t i; + uint64_t Token = 0; + for (i = 0; i < N; i++) { + Token = Val ^ Vals[i]; + if (Val < Vals[i]) + break; + } + + if (ValSizeInBits == 16) + fuzzer::TPC.HandleCmp(PC + i, static_cast<uint16_t>(Token), (uint16_t)(0)); + else if (ValSizeInBits == 32) + fuzzer::TPC.HandleCmp(PC + i, static_cast<uint32_t>(Token), (uint32_t)(0)); + else + fuzzer::TPC.HandleCmp(PC + i, Token, (uint64_t)(0)); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_div4(uint32_t Val) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Val, (uint32_t)0); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_div8(uint64_t Val) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Val, (uint64_t)0); +} + +ATTRIBUTE_INTERFACE +ATTRIBUTE_NO_SANITIZE_ALL +ATTRIBUTE_TARGET_POPCNT +void __sanitizer_cov_trace_gep(uintptr_t Idx) { + uintptr_t PC = reinterpret_cast<uintptr_t>(__builtin_return_address(0)); + fuzzer::TPC.HandleCmp(PC, Idx, (uintptr_t)0); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_memcmp(void *caller_pc, const void *s1, + const void *s2, size_t n, int result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + if (result == 0) return; // No reason to mutate. + if (n <= 1) return; // Not interesting. + fuzzer::TPC.AddValueForMemcmp(caller_pc, s1, s2, n, /*StopAtZero*/false); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_strncmp(void *caller_pc, const char *s1, + const char *s2, size_t n, int result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + if (result == 0) return; // No reason to mutate. + size_t Len1 = fuzzer::InternalStrnlen(s1, n); + size_t Len2 = fuzzer::InternalStrnlen(s2, n); + n = std::min(n, Len1); + n = std::min(n, Len2); + if (n <= 1) return; // Not interesting. + fuzzer::TPC.AddValueForMemcmp(caller_pc, s1, s2, n, /*StopAtZero*/true); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_strcmp(void *caller_pc, const char *s1, + const char *s2, int result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + if (result == 0) return; // No reason to mutate. + size_t N = fuzzer::InternalStrnlen2(s1, s2); + if (N <= 1) return; // Not interesting. + fuzzer::TPC.AddValueForMemcmp(caller_pc, s1, s2, N, /*StopAtZero*/true); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_strncasecmp(void *called_pc, const char *s1, + const char *s2, size_t n, int result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + return __sanitizer_weak_hook_strncmp(called_pc, s1, s2, n, result); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_strcasecmp(void *called_pc, const char *s1, + const char *s2, int result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + return __sanitizer_weak_hook_strcmp(called_pc, s1, s2, result); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_strstr(void *called_pc, const char *s1, + const char *s2, char *result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + fuzzer::TPC.MMT.Add(reinterpret_cast<const uint8_t *>(s2), strlen(s2)); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_strcasestr(void *called_pc, const char *s1, + const char *s2, char *result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + fuzzer::TPC.MMT.Add(reinterpret_cast<const uint8_t *>(s2), strlen(s2)); +} + +ATTRIBUTE_INTERFACE ATTRIBUTE_NO_SANITIZE_MEMORY +void __sanitizer_weak_hook_memmem(void *called_pc, const void *s1, size_t len1, + const void *s2, size_t len2, void *result) { + if (fuzzer::ScopedDoingMyOwnMemOrStr::DoingMyOwnMemOrStr) return; + fuzzer::TPC.MMT.Add(reinterpret_cast<const uint8_t *>(s2), len2); +} +} // extern "C" diff --git a/lib/fuzzer/FuzzerTracePC.h b/lib/fuzzer/FuzzerTracePC.h new file mode 100644 index 000000000000..dc65cd72091a --- /dev/null +++ b/lib/fuzzer/FuzzerTracePC.h @@ -0,0 +1,296 @@ +//===- FuzzerTracePC.h - Internal header for the Fuzzer ---------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// fuzzer::TracePC +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_TRACE_PC +#define LLVM_FUZZER_TRACE_PC + +#include "FuzzerDefs.h" +#include "FuzzerDictionary.h" +#include "FuzzerValueBitMap.h" + +#include <set> + +namespace fuzzer { + +// TableOfRecentCompares (TORC) remembers the most recently performed +// comparisons of type T. +// We record the arguments of CMP instructions in this table unconditionally +// because it seems cheaper this way than to compute some expensive +// conditions inside __sanitizer_cov_trace_cmp*. +// After the unit has been executed we may decide to use the contents of +// this table to populate a Dictionary. +template<class T, size_t kSizeT> +struct TableOfRecentCompares { + static const size_t kSize = kSizeT; + struct Pair { + T A, B; + }; + ATTRIBUTE_NO_SANITIZE_ALL + void Insert(size_t Idx, const T &Arg1, const T &Arg2) { + Idx = Idx % kSize; + Table[Idx].A = Arg1; + Table[Idx].B = Arg2; + } + + Pair Get(size_t I) { return Table[I % kSize]; } + + Pair Table[kSize]; +}; + +template <size_t kSizeT> +struct MemMemTable { + static const size_t kSize = kSizeT; + Word MemMemWords[kSize]; + Word EmptyWord; + + void Add(const uint8_t *Data, size_t Size) { + if (Size <= 2) return; + Size = std::min(Size, Word::GetMaxSize()); + size_t Idx = SimpleFastHash(Data, Size) % kSize; + MemMemWords[Idx].Set(Data, Size); + } + const Word &Get(size_t Idx) { + for (size_t i = 0; i < kSize; i++) { + const Word &W = MemMemWords[(Idx + i) % kSize]; + if (W.size()) return W; + } + EmptyWord.Set(nullptr, 0); + return EmptyWord; + } +}; + +class TracePC { + public: + static const size_t kNumPCs = 1 << 21; + // How many bits of PC are used from __sanitizer_cov_trace_pc. + static const size_t kTracePcBits = 18; + + void HandleInit(uint32_t *Start, uint32_t *Stop); + void HandleInline8bitCountersInit(uint8_t *Start, uint8_t *Stop); + void HandlePCsInit(const uintptr_t *Start, const uintptr_t *Stop); + void HandleCallerCallee(uintptr_t Caller, uintptr_t Callee); + template <class T> void HandleCmp(uintptr_t PC, T Arg1, T Arg2); + size_t GetTotalPCCoverage(); + void SetUseCounters(bool UC) { UseCounters = UC; } + void SetUseClangCoverage(bool UCC) { UseClangCoverage = UCC; } + void SetUseValueProfile(bool VP) { UseValueProfile = VP; } + void SetPrintNewPCs(bool P) { DoPrintNewPCs = P; } + void SetPrintNewFuncs(size_t P) { NumPrintNewFuncs = P; } + void UpdateObservedPCs(); + template <class Callback> void CollectFeatures(Callback CB) const; + + void ResetMaps() { + ValueProfileMap.Reset(); + if (NumModules) + memset(Counters(), 0, GetNumPCs()); + ClearExtraCounters(); + ClearInlineCounters(); + if (UseClangCoverage) + ClearClangCounters(); + } + + void ClearInlineCounters(); + + void UpdateFeatureSet(size_t CurrentElementIdx, size_t CurrentElementSize); + void PrintFeatureSet(); + + void PrintModuleInfo(); + + void PrintCoverage(); + void DumpCoverage(); + + void AddValueForMemcmp(void *caller_pc, const void *s1, const void *s2, + size_t n, bool StopAtZero); + + TableOfRecentCompares<uint32_t, 32> TORC4; + TableOfRecentCompares<uint64_t, 32> TORC8; + TableOfRecentCompares<Word, 32> TORCW; + MemMemTable<1024> MMT; + + size_t GetNumPCs() const { + return NumGuards == 0 ? (1 << kTracePcBits) : Min(kNumPCs, NumGuards + 1); + } + uintptr_t GetPC(size_t Idx) { + assert(Idx < GetNumPCs()); + return PCs()[Idx]; + } + + void RecordInitialStack(); + uintptr_t GetMaxStackOffset() const; + + template<class CallBack> + void ForEachObservedPC(CallBack CB) { + for (auto PC : ObservedPCs) + CB(PC); + } + +private: + bool UseCounters = false; + bool UseValueProfile = false; + bool UseClangCoverage = false; + bool DoPrintNewPCs = false; + size_t NumPrintNewFuncs = 0; + + struct Module { + uint32_t *Start, *Stop; + }; + + Module Modules[4096]; + size_t NumModules; // linker-initialized. + size_t NumGuards; // linker-initialized. + + struct { uint8_t *Start, *Stop; } ModuleCounters[4096]; + size_t NumModulesWithInline8bitCounters; // linker-initialized. + size_t NumInline8bitCounters; + + struct PCTableEntry { + uintptr_t PC, PCFlags; + }; + + struct { const PCTableEntry *Start, *Stop; } ModulePCTable[4096]; + size_t NumPCTables; + size_t NumPCsInPCTables; + + uint8_t *Counters() const; + uintptr_t *PCs() const; + + Set<uintptr_t> ObservedPCs; + Set<uintptr_t> ObservedFuncs; + + ValueBitMap ValueProfileMap; + uintptr_t InitialStack; +}; + +template <class Callback> +// void Callback(size_t FirstFeature, size_t Idx, uint8_t Value); +ATTRIBUTE_NO_SANITIZE_ALL +void ForEachNonZeroByte(const uint8_t *Begin, const uint8_t *End, + size_t FirstFeature, Callback Handle8bitCounter) { + typedef uintptr_t LargeType; + const size_t Step = sizeof(LargeType) / sizeof(uint8_t); + const size_t StepMask = Step - 1; + auto P = Begin; + // Iterate by 1 byte until either the alignment boundary or the end. + for (; reinterpret_cast<uintptr_t>(P) & StepMask && P < End; P++) + if (uint8_t V = *P) + Handle8bitCounter(FirstFeature, P - Begin, V); + + // Iterate by Step bytes at a time. + for (; P < End; P += Step) + if (LargeType Bundle = *reinterpret_cast<const LargeType *>(P)) + for (size_t I = 0; I < Step; I++, Bundle >>= 8) + if (uint8_t V = Bundle & 0xff) + Handle8bitCounter(FirstFeature, P - Begin + I, V); + + // Iterate by 1 byte until the end. + for (; P < End; P++) + if (uint8_t V = *P) + Handle8bitCounter(FirstFeature, P - Begin, V); +} + +// Given a non-zero Counter returns a number in the range [0,7]. +template<class T> +unsigned CounterToFeature(T Counter) { + // Returns a feature number by placing Counters into buckets as illustrated + // below. + // + // Counter bucket: [1] [2] [3] [4-7] [8-15] [16-31] [32-127] [128+] + // Feature number: 0 1 2 3 4 5 6 7 + // + // This is a heuristic taken from AFL (see + // http://lcamtuf.coredump.cx/afl/technical_details.txt). + // + // This implementation may change in the future so clients should + // not rely on it. + assert(Counter); + unsigned Bit = 0; + /**/ if (Counter >= 128) Bit = 7; + else if (Counter >= 32) Bit = 6; + else if (Counter >= 16) Bit = 5; + else if (Counter >= 8) Bit = 4; + else if (Counter >= 4) Bit = 3; + else if (Counter >= 3) Bit = 2; + else if (Counter >= 2) Bit = 1; + return Bit; +} + +template <class Callback> // void Callback(size_t Feature) +ATTRIBUTE_NO_SANITIZE_ADDRESS +__attribute__((noinline)) +void TracePC::CollectFeatures(Callback HandleFeature) const { + uint8_t *Counters = this->Counters(); + size_t N = GetNumPCs(); + auto Handle8bitCounter = [&](size_t FirstFeature, + size_t Idx, uint8_t Counter) { + if (UseCounters) + HandleFeature(FirstFeature + Idx * 8 + CounterToFeature(Counter)); + else + HandleFeature(FirstFeature + Idx); + }; + + size_t FirstFeature = 0; + + if (!NumInline8bitCounters) { + ForEachNonZeroByte(Counters, Counters + N, FirstFeature, Handle8bitCounter); + FirstFeature += N * 8; + } + + if (NumInline8bitCounters) { + for (size_t i = 0; i < NumModulesWithInline8bitCounters; i++) { + ForEachNonZeroByte(ModuleCounters[i].Start, ModuleCounters[i].Stop, + FirstFeature, Handle8bitCounter); + FirstFeature += 8 * (ModuleCounters[i].Stop - ModuleCounters[i].Start); + } + } + + if (size_t NumClangCounters = ClangCountersEnd() - ClangCountersBegin()) { + auto P = ClangCountersBegin(); + for (size_t Idx = 0; Idx < NumClangCounters; Idx++) + if (auto Cnt = P[Idx]) { + if (UseCounters) + HandleFeature(FirstFeature + Idx * 8 + CounterToFeature(Cnt)); + else + HandleFeature(FirstFeature + Idx); + } + FirstFeature += NumClangCounters; + } + + ForEachNonZeroByte(ExtraCountersBegin(), ExtraCountersEnd(), FirstFeature, + Handle8bitCounter); + FirstFeature += (ExtraCountersEnd() - ExtraCountersBegin()) * 8; + + if (UseValueProfile) { + ValueProfileMap.ForEach([&](size_t Idx) { + HandleFeature(FirstFeature + Idx); + }); + FirstFeature += ValueProfileMap.SizeInBits(); + } + + // Step function, grows similar to 8 * Log_2(A). + auto StackDepthStepFunction = [](uint32_t A) -> uint32_t { + uint32_t Log2 = Log(A); + if (Log2 < 3) return A; + Log2 -= 3; + return (Log2 + 1) * 8 + ((A >> Log2) & 7); + }; + assert(StackDepthStepFunction(1024) == 64); + assert(StackDepthStepFunction(1024 * 4) == 80); + assert(StackDepthStepFunction(1024 * 1024) == 144); + + if (auto MaxStackOffset = GetMaxStackOffset()) + HandleFeature(FirstFeature + StackDepthStepFunction(MaxStackOffset / 8)); +} + +extern TracePC TPC; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_TRACE_PC diff --git a/lib/fuzzer/FuzzerUtil.cpp b/lib/fuzzer/FuzzerUtil.cpp new file mode 100644 index 000000000000..96b37d34815d --- /dev/null +++ b/lib/fuzzer/FuzzerUtil.cpp @@ -0,0 +1,215 @@ +//===- FuzzerUtil.cpp - Misc utils ----------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Misc utils. +//===----------------------------------------------------------------------===// + +#include "FuzzerUtil.h" +#include "FuzzerIO.h" +#include "FuzzerInternal.h" +#include <cassert> +#include <chrono> +#include <cstring> +#include <errno.h> +#include <signal.h> +#include <sstream> +#include <stdio.h> +#include <sys/types.h> +#include <thread> + +namespace fuzzer { + +void PrintHexArray(const uint8_t *Data, size_t Size, + const char *PrintAfter) { + for (size_t i = 0; i < Size; i++) + Printf("0x%x,", (unsigned)Data[i]); + Printf("%s", PrintAfter); +} + +void Print(const Unit &v, const char *PrintAfter) { + PrintHexArray(v.data(), v.size(), PrintAfter); +} + +void PrintASCIIByte(uint8_t Byte) { + if (Byte == '\\') + Printf("\\\\"); + else if (Byte == '"') + Printf("\\\""); + else if (Byte >= 32 && Byte < 127) + Printf("%c", Byte); + else + Printf("\\x%02x", Byte); +} + +void PrintASCII(const uint8_t *Data, size_t Size, const char *PrintAfter) { + for (size_t i = 0; i < Size; i++) + PrintASCIIByte(Data[i]); + Printf("%s", PrintAfter); +} + +void PrintASCII(const Unit &U, const char *PrintAfter) { + PrintASCII(U.data(), U.size(), PrintAfter); +} + +bool ToASCII(uint8_t *Data, size_t Size) { + bool Changed = false; + for (size_t i = 0; i < Size; i++) { + uint8_t &X = Data[i]; + auto NewX = X; + NewX &= 127; + if (!isspace(NewX) && !isprint(NewX)) + NewX = ' '; + Changed |= NewX != X; + X = NewX; + } + return Changed; +} + +bool IsASCII(const Unit &U) { return IsASCII(U.data(), U.size()); } + +bool IsASCII(const uint8_t *Data, size_t Size) { + for (size_t i = 0; i < Size; i++) + if (!(isprint(Data[i]) || isspace(Data[i]))) return false; + return true; +} + +bool ParseOneDictionaryEntry(const std::string &Str, Unit *U) { + U->clear(); + if (Str.empty()) return false; + size_t L = 0, R = Str.size() - 1; // We are parsing the range [L,R]. + // Skip spaces from both sides. + while (L < R && isspace(Str[L])) L++; + while (R > L && isspace(Str[R])) R--; + if (R - L < 2) return false; + // Check the closing " + if (Str[R] != '"') return false; + R--; + // Find the opening " + while (L < R && Str[L] != '"') L++; + if (L >= R) return false; + assert(Str[L] == '\"'); + L++; + assert(L <= R); + for (size_t Pos = L; Pos <= R; Pos++) { + uint8_t V = (uint8_t)Str[Pos]; + if (!isprint(V) && !isspace(V)) return false; + if (V =='\\') { + // Handle '\\' + if (Pos + 1 <= R && (Str[Pos + 1] == '\\' || Str[Pos + 1] == '"')) { + U->push_back(Str[Pos + 1]); + Pos++; + continue; + } + // Handle '\xAB' + if (Pos + 3 <= R && Str[Pos + 1] == 'x' + && isxdigit(Str[Pos + 2]) && isxdigit(Str[Pos + 3])) { + char Hex[] = "0xAA"; + Hex[2] = Str[Pos + 2]; + Hex[3] = Str[Pos + 3]; + U->push_back(strtol(Hex, nullptr, 16)); + Pos += 3; + continue; + } + return false; // Invalid escape. + } else { + // Any other character. + U->push_back(V); + } + } + return true; +} + +bool ParseDictionaryFile(const std::string &Text, Vector<Unit> *Units) { + if (Text.empty()) { + Printf("ParseDictionaryFile: file does not exist or is empty\n"); + return false; + } + std::istringstream ISS(Text); + Units->clear(); + Unit U; + int LineNo = 0; + std::string S; + while (std::getline(ISS, S, '\n')) { + LineNo++; + size_t Pos = 0; + while (Pos < S.size() && isspace(S[Pos])) Pos++; // Skip spaces. + if (Pos == S.size()) continue; // Empty line. + if (S[Pos] == '#') continue; // Comment line. + if (ParseOneDictionaryEntry(S, &U)) { + Units->push_back(U); + } else { + Printf("ParseDictionaryFile: error in line %d\n\t\t%s\n", LineNo, + S.c_str()); + return false; + } + } + return true; +} + +std::string Base64(const Unit &U) { + static const char Table[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ" + "abcdefghijklmnopqrstuvwxyz" + "0123456789+/"; + std::string Res; + size_t i; + for (i = 0; i + 2 < U.size(); i += 3) { + uint32_t x = (U[i] << 16) + (U[i + 1] << 8) + U[i + 2]; + Res += Table[(x >> 18) & 63]; + Res += Table[(x >> 12) & 63]; + Res += Table[(x >> 6) & 63]; + Res += Table[x & 63]; + } + if (i + 1 == U.size()) { + uint32_t x = (U[i] << 16); + Res += Table[(x >> 18) & 63]; + Res += Table[(x >> 12) & 63]; + Res += "=="; + } else if (i + 2 == U.size()) { + uint32_t x = (U[i] << 16) + (U[i + 1] << 8); + Res += Table[(x >> 18) & 63]; + Res += Table[(x >> 12) & 63]; + Res += Table[(x >> 6) & 63]; + Res += "="; + } + return Res; +} + +std::string DescribePC(const char *SymbolizedFMT, uintptr_t PC) { + if (!EF->__sanitizer_symbolize_pc) return "<can not symbolize>"; + char PcDescr[1024] = {}; + EF->__sanitizer_symbolize_pc(reinterpret_cast<void*>(PC), + SymbolizedFMT, PcDescr, sizeof(PcDescr)); + PcDescr[sizeof(PcDescr) - 1] = 0; // Just in case. + return PcDescr; +} + +void PrintPC(const char *SymbolizedFMT, const char *FallbackFMT, uintptr_t PC) { + if (EF->__sanitizer_symbolize_pc) + Printf("%s", DescribePC(SymbolizedFMT, PC).c_str()); + else + Printf(FallbackFMT, PC); +} + +unsigned NumberOfCpuCores() { + unsigned N = std::thread::hardware_concurrency(); + if (!N) { + Printf("WARNING: std::thread::hardware_concurrency not well defined for " + "your platform. Assuming CPU count of 1.\n"); + N = 1; + } + return N; +} + +size_t SimpleFastHash(const uint8_t *Data, size_t Size) { + size_t Res = 0; + for (size_t i = 0; i < Size; i++) + Res = Res * 11 + Data[i]; + return Res; +} + +} // namespace fuzzer diff --git a/lib/fuzzer/FuzzerUtil.h b/lib/fuzzer/FuzzerUtil.h new file mode 100644 index 000000000000..f2ed028ce78e --- /dev/null +++ b/lib/fuzzer/FuzzerUtil.h @@ -0,0 +1,87 @@ +//===- FuzzerUtil.h - Internal header for the Fuzzer Utils ------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Util functions. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_UTIL_H +#define LLVM_FUZZER_UTIL_H + +#include "FuzzerDefs.h" +#include "FuzzerCommand.h" + +namespace fuzzer { + +void PrintHexArray(const Unit &U, const char *PrintAfter = ""); + +void PrintHexArray(const uint8_t *Data, size_t Size, + const char *PrintAfter = ""); + +void PrintASCII(const uint8_t *Data, size_t Size, const char *PrintAfter = ""); + +void PrintASCII(const Unit &U, const char *PrintAfter = ""); + +// Changes U to contain only ASCII (isprint+isspace) characters. +// Returns true iff U has been changed. +bool ToASCII(uint8_t *Data, size_t Size); + +bool IsASCII(const Unit &U); + +bool IsASCII(const uint8_t *Data, size_t Size); + +std::string Base64(const Unit &U); + +void PrintPC(const char *SymbolizedFMT, const char *FallbackFMT, uintptr_t PC); + +std::string DescribePC(const char *SymbolizedFMT, uintptr_t PC); + +unsigned NumberOfCpuCores(); + +// Platform specific functions. +void SetSignalHandler(const FuzzingOptions& Options); + +void SleepSeconds(int Seconds); + +unsigned long GetPid(); + +size_t GetPeakRSSMb(); + +int ExecuteCommand(const Command &Cmd); + +FILE *OpenProcessPipe(const char *Command, const char *Mode); + +const void *SearchMemory(const void *haystack, size_t haystacklen, + const void *needle, size_t needlelen); + +std::string CloneArgsWithoutX(const Vector<std::string> &Args, + const char *X1, const char *X2); + +inline std::string CloneArgsWithoutX(const Vector<std::string> &Args, + const char *X) { + return CloneArgsWithoutX(Args, X, X); +} + +inline std::pair<std::string, std::string> SplitBefore(std::string X, + std::string S) { + auto Pos = S.find(X); + if (Pos == std::string::npos) + return std::make_pair(S, ""); + return std::make_pair(S.substr(0, Pos), S.substr(Pos)); +} + +std::string DisassembleCmd(const std::string &FileName); + +std::string SearchRegexCmd(const std::string &Regex); + +size_t SimpleFastHash(const uint8_t *Data, size_t Size); + +inline uint32_t Log(uint32_t X) { return 32 - __builtin_clz(X) - 1; } + +} // namespace fuzzer + +#endif // LLVM_FUZZER_UTIL_H diff --git a/lib/fuzzer/FuzzerUtilDarwin.cpp b/lib/fuzzer/FuzzerUtilDarwin.cpp new file mode 100644 index 000000000000..4bfbc11a50cf --- /dev/null +++ b/lib/fuzzer/FuzzerUtilDarwin.cpp @@ -0,0 +1,162 @@ +//===- FuzzerUtilDarwin.cpp - Misc utils ----------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Misc utils for Darwin. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_APPLE +#include "FuzzerCommand.h" +#include "FuzzerIO.h" +#include <mutex> +#include <signal.h> +#include <spawn.h> +#include <stdlib.h> +#include <string.h> +#include <sys/wait.h> + +// There is no header for this on macOS so declare here +extern "C" char **environ; + +namespace fuzzer { + +static std::mutex SignalMutex; +// Global variables used to keep track of how signal handling should be +// restored. They should **not** be accessed without holding `SignalMutex`. +static int ActiveThreadCount = 0; +static struct sigaction OldSigIntAction; +static struct sigaction OldSigQuitAction; +static sigset_t OldBlockedSignalsSet; + +// This is a reimplementation of Libc's `system()`. On Darwin the Libc +// implementation contains a mutex which prevents it from being used +// concurrently. This implementation **can** be used concurrently. It sets the +// signal handlers when the first thread enters and restores them when the last +// thread finishes execution of the function and ensures this is not racey by +// using a mutex. +int ExecuteCommand(const Command &Cmd) { + std::string CmdLine = Cmd.toString(); + posix_spawnattr_t SpawnAttributes; + if (posix_spawnattr_init(&SpawnAttributes)) + return -1; + // Block and ignore signals of the current process when the first thread + // enters. + { + std::lock_guard<std::mutex> Lock(SignalMutex); + if (ActiveThreadCount == 0) { + static struct sigaction IgnoreSignalAction; + sigset_t BlockedSignalsSet; + memset(&IgnoreSignalAction, 0, sizeof(IgnoreSignalAction)); + IgnoreSignalAction.sa_handler = SIG_IGN; + + if (sigaction(SIGINT, &IgnoreSignalAction, &OldSigIntAction) == -1) { + Printf("Failed to ignore SIGINT\n"); + (void)posix_spawnattr_destroy(&SpawnAttributes); + return -1; + } + if (sigaction(SIGQUIT, &IgnoreSignalAction, &OldSigQuitAction) == -1) { + Printf("Failed to ignore SIGQUIT\n"); + // Try our best to restore the signal handlers. + (void)sigaction(SIGINT, &OldSigIntAction, NULL); + (void)posix_spawnattr_destroy(&SpawnAttributes); + return -1; + } + + (void)sigemptyset(&BlockedSignalsSet); + (void)sigaddset(&BlockedSignalsSet, SIGCHLD); + if (sigprocmask(SIG_BLOCK, &BlockedSignalsSet, &OldBlockedSignalsSet) == + -1) { + Printf("Failed to block SIGCHLD\n"); + // Try our best to restore the signal handlers. + (void)sigaction(SIGQUIT, &OldSigQuitAction, NULL); + (void)sigaction(SIGINT, &OldSigIntAction, NULL); + (void)posix_spawnattr_destroy(&SpawnAttributes); + return -1; + } + } + ++ActiveThreadCount; + } + + // NOTE: Do not introduce any new `return` statements past this + // point. It is important that `ActiveThreadCount` always be decremented + // when leaving this function. + + // Make sure the child process uses the default handlers for the + // following signals rather than inheriting what the parent has. + sigset_t DefaultSigSet; + (void)sigemptyset(&DefaultSigSet); + (void)sigaddset(&DefaultSigSet, SIGQUIT); + (void)sigaddset(&DefaultSigSet, SIGINT); + (void)posix_spawnattr_setsigdefault(&SpawnAttributes, &DefaultSigSet); + // Make sure the child process doesn't block SIGCHLD + (void)posix_spawnattr_setsigmask(&SpawnAttributes, &OldBlockedSignalsSet); + short SpawnFlags = POSIX_SPAWN_SETSIGDEF | POSIX_SPAWN_SETSIGMASK; + (void)posix_spawnattr_setflags(&SpawnAttributes, SpawnFlags); + + pid_t Pid; + char **Environ = environ; // Read from global + const char *CommandCStr = CmdLine.c_str(); + char *const Argv[] = { + strdup("sh"), + strdup("-c"), + strdup(CommandCStr), + NULL + }; + int ErrorCode = 0, ProcessStatus = 0; + // FIXME: We probably shouldn't hardcode the shell path. + ErrorCode = posix_spawn(&Pid, "/bin/sh", NULL, &SpawnAttributes, + Argv, Environ); + (void)posix_spawnattr_destroy(&SpawnAttributes); + if (!ErrorCode) { + pid_t SavedPid = Pid; + do { + // Repeat until call completes uninterrupted. + Pid = waitpid(SavedPid, &ProcessStatus, /*options=*/0); + } while (Pid == -1 && errno == EINTR); + if (Pid == -1) { + // Fail for some other reason. + ProcessStatus = -1; + } + } else if (ErrorCode == ENOMEM || ErrorCode == EAGAIN) { + // Fork failure. + ProcessStatus = -1; + } else { + // Shell execution failure. + ProcessStatus = W_EXITCODE(127, 0); + } + for (unsigned i = 0, n = sizeof(Argv) / sizeof(Argv[0]); i < n; ++i) + free(Argv[i]); + + // Restore the signal handlers of the current process when the last thread + // using this function finishes. + { + std::lock_guard<std::mutex> Lock(SignalMutex); + --ActiveThreadCount; + if (ActiveThreadCount == 0) { + bool FailedRestore = false; + if (sigaction(SIGINT, &OldSigIntAction, NULL) == -1) { + Printf("Failed to restore SIGINT handling\n"); + FailedRestore = true; + } + if (sigaction(SIGQUIT, &OldSigQuitAction, NULL) == -1) { + Printf("Failed to restore SIGQUIT handling\n"); + FailedRestore = true; + } + if (sigprocmask(SIG_BLOCK, &OldBlockedSignalsSet, NULL) == -1) { + Printf("Failed to unblock SIGCHLD\n"); + FailedRestore = true; + } + if (FailedRestore) + ProcessStatus = -1; + } + } + return ProcessStatus; +} + +} // namespace fuzzer + +#endif // LIBFUZZER_APPLE diff --git a/lib/fuzzer/FuzzerUtilFuchsia.cpp b/lib/fuzzer/FuzzerUtilFuchsia.cpp new file mode 100644 index 000000000000..a5fdc37fb5a6 --- /dev/null +++ b/lib/fuzzer/FuzzerUtilFuchsia.cpp @@ -0,0 +1,228 @@ +//===- FuzzerUtilFuchsia.cpp - Misc utils for Fuchsia. --------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Misc utils implementation using Fuchsia/Zircon APIs. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" + +#if LIBFUZZER_FUCHSIA + +#include "FuzzerInternal.h" +#include "FuzzerUtil.h" +#include <cerrno> +#include <cinttypes> +#include <cstdint> +#include <fbl/unique_fd.h> +#include <fcntl.h> +#include <launchpad/launchpad.h> +#include <string> +#include <thread> +#include <zircon/errors.h> +#include <zircon/status.h> +#include <zircon/syscalls.h> +#include <zircon/syscalls/port.h> +#include <zircon/types.h> +#include <zx/object.h> +#include <zx/port.h> +#include <zx/process.h> +#include <zx/time.h> + +namespace fuzzer { + +namespace { + +// A magic value for the Zircon exception port, chosen to spell 'FUZZING' +// when interpreted as a byte sequence on little-endian platforms. +const uint64_t kFuzzingCrash = 0x474e495a5a5546; + +void AlarmHandler(int Seconds) { + while (true) { + SleepSeconds(Seconds); + Fuzzer::StaticAlarmCallback(); + } +} + +void InterruptHandler() { + // Ctrl-C sends ETX in Zircon. + while (getchar() != 0x03); + Fuzzer::StaticInterruptCallback(); +} + +void CrashHandler(zx::port *Port) { + std::unique_ptr<zx::port> ExceptionPort(Port); + zx_port_packet_t Packet; + ExceptionPort->wait(ZX_TIME_INFINITE, &Packet, 0); + // Unbind as soon as possible so we don't receive exceptions from this thread. + if (zx_task_bind_exception_port(ZX_HANDLE_INVALID, ZX_HANDLE_INVALID, + kFuzzingCrash, 0) != ZX_OK) { + // Shouldn't happen; if it does the safest option is to just exit. + Printf("libFuzzer: unable to unbind exception port; aborting!\n"); + exit(1); + } + if (Packet.key != kFuzzingCrash) { + Printf("libFuzzer: invalid crash key: %" PRIx64 "; aborting!\n", + Packet.key); + exit(1); + } + // CrashCallback should not return from this call + Fuzzer::StaticCrashSignalCallback(); +} + +} // namespace + +// Platform specific functions. +void SetSignalHandler(const FuzzingOptions &Options) { + zx_status_t rc; + + // Set up alarm handler if needed. + if (Options.UnitTimeoutSec > 0) { + std::thread T(AlarmHandler, Options.UnitTimeoutSec / 2 + 1); + T.detach(); + } + + // Set up interrupt handler if needed. + if (Options.HandleInt || Options.HandleTerm) { + std::thread T(InterruptHandler); + T.detach(); + } + + // Early exit if no crash handler needed. + if (!Options.HandleSegv && !Options.HandleBus && !Options.HandleIll && + !Options.HandleFpe && !Options.HandleAbrt) + return; + + // Create an exception port + zx::port *ExceptionPort = new zx::port(); + if ((rc = zx::port::create(0, ExceptionPort)) != ZX_OK) { + Printf("libFuzzer: zx_port_create failed: %s\n", zx_status_get_string(rc)); + exit(1); + } + + // Bind the port to receive exceptions from our process + if ((rc = zx_task_bind_exception_port(zx_process_self(), ExceptionPort->get(), + kFuzzingCrash, 0)) != ZX_OK) { + Printf("libFuzzer: unable to bind exception port: %s\n", + zx_status_get_string(rc)); + exit(1); + } + + // Set up the crash handler. + std::thread T(CrashHandler, ExceptionPort); + T.detach(); +} + +void SleepSeconds(int Seconds) { + zx::nanosleep(zx::deadline_after(ZX_SEC(Seconds))); +} + +unsigned long GetPid() { + zx_status_t rc; + zx_info_handle_basic_t Info; + if ((rc = zx_object_get_info(zx_process_self(), ZX_INFO_HANDLE_BASIC, &Info, + sizeof(Info), NULL, NULL)) != ZX_OK) { + Printf("libFuzzer: unable to get info about self: %s\n", + zx_status_get_string(rc)); + exit(1); + } + return Info.koid; +} + +size_t GetPeakRSSMb() { + zx_status_t rc; + zx_info_task_stats_t Info; + if ((rc = zx_object_get_info(zx_process_self(), ZX_INFO_TASK_STATS, &Info, + sizeof(Info), NULL, NULL)) != ZX_OK) { + Printf("libFuzzer: unable to get info about self: %s\n", + zx_status_get_string(rc)); + exit(1); + } + return (Info.mem_private_bytes + Info.mem_shared_bytes) >> 20; +} + +int ExecuteCommand(const Command &Cmd) { + zx_status_t rc; + + // Convert arguments to C array + auto Args = Cmd.getArguments(); + size_t Argc = Args.size(); + assert(Argc != 0); + std::unique_ptr<const char *[]> Argv(new const char *[Argc]); + for (size_t i = 0; i < Argc; ++i) + Argv[i] = Args[i].c_str(); + + // Create the basic launchpad. Clone everything except stdio. + launchpad_t *lp; + launchpad_create(ZX_HANDLE_INVALID, Argv[0], &lp); + launchpad_load_from_file(lp, Argv[0]); + launchpad_set_args(lp, Argc, Argv.get()); + launchpad_clone(lp, LP_CLONE_ALL & (~LP_CLONE_FDIO_STDIO)); + + // Determine stdout + int FdOut = STDOUT_FILENO; + fbl::unique_fd OutputFile; + if (Cmd.hasOutputFile()) { + auto Filename = Cmd.getOutputFile(); + OutputFile.reset(open(Filename.c_str(), O_WRONLY | O_CREAT | O_TRUNC, 0)); + if (!OutputFile) { + Printf("libFuzzer: failed to open %s: %s\n", Filename.c_str(), + strerror(errno)); + return ZX_ERR_IO; + } + FdOut = OutputFile.get(); + } + + // Determine stderr + int FdErr = STDERR_FILENO; + if (Cmd.isOutAndErrCombined()) + FdErr = FdOut; + + // Clone the file descriptors into the new process + if ((rc = launchpad_clone_fd(lp, STDIN_FILENO, STDIN_FILENO)) != ZX_OK || + (rc = launchpad_clone_fd(lp, FdOut, STDOUT_FILENO)) != ZX_OK || + (rc = launchpad_clone_fd(lp, FdErr, STDERR_FILENO)) != ZX_OK) { + Printf("libFuzzer: failed to clone FDIO: %s\n", zx_status_get_string(rc)); + return rc; + } + + // Start the process + zx_handle_t ProcessHandle = ZX_HANDLE_INVALID; + const char *ErrorMsg = nullptr; + if ((rc = launchpad_go(lp, &ProcessHandle, &ErrorMsg)) != ZX_OK) { + Printf("libFuzzer: failed to launch '%s': %s, %s\n", Argv[0], ErrorMsg, + zx_status_get_string(rc)); + return rc; + } + zx::process Process(ProcessHandle); + + // Now join the process and return the exit status. + if ((rc = Process.wait_one(ZX_PROCESS_TERMINATED, ZX_TIME_INFINITE, + nullptr)) != ZX_OK) { + Printf("libFuzzer: failed to join '%s': %s\n", Argv[0], + zx_status_get_string(rc)); + return rc; + } + + zx_info_process_t Info; + if ((rc = Process.get_info(ZX_INFO_PROCESS, &Info, sizeof(Info), nullptr, + nullptr)) != ZX_OK) { + Printf("libFuzzer: unable to get return code from '%s': %s\n", Argv[0], + zx_status_get_string(rc)); + return rc; + } + + return Info.return_code; +} + +const void *SearchMemory(const void *Data, size_t DataLen, const void *Patt, + size_t PattLen) { + return memmem(Data, DataLen, Patt, PattLen); +} + +} // namespace fuzzer + +#endif // LIBFUZZER_FUCHSIA diff --git a/lib/fuzzer/FuzzerUtilLinux.cpp b/lib/fuzzer/FuzzerUtilLinux.cpp new file mode 100644 index 000000000000..c7cf2c0a778b --- /dev/null +++ b/lib/fuzzer/FuzzerUtilLinux.cpp @@ -0,0 +1,26 @@ +//===- FuzzerUtilLinux.cpp - Misc utils for Linux. ------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Misc utils for Linux. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_LINUX || LIBFUZZER_NETBSD +#include "FuzzerCommand.h" + +#include <stdlib.h> + +namespace fuzzer { + +int ExecuteCommand(const Command &Cmd) { + std::string CmdLine = Cmd.toString(); + return system(CmdLine.c_str()); +} + +} // namespace fuzzer + +#endif // LIBFUZZER_LINUX || LIBFUZZER_NETBSD diff --git a/lib/fuzzer/FuzzerUtilPosix.cpp b/lib/fuzzer/FuzzerUtilPosix.cpp new file mode 100644 index 000000000000..934b7aa98ff1 --- /dev/null +++ b/lib/fuzzer/FuzzerUtilPosix.cpp @@ -0,0 +1,151 @@ +//===- FuzzerUtilPosix.cpp - Misc utils for Posix. ------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Misc utils implementation using Posix API. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_POSIX +#include "FuzzerIO.h" +#include "FuzzerInternal.h" +#include <cassert> +#include <chrono> +#include <cstring> +#include <errno.h> +#include <iomanip> +#include <signal.h> +#include <stdio.h> +#include <sys/resource.h> +#include <sys/syscall.h> +#include <sys/time.h> +#include <sys/types.h> +#include <thread> +#include <unistd.h> + +namespace fuzzer { + +static void AlarmHandler(int, siginfo_t *, void *) { + Fuzzer::StaticAlarmCallback(); +} + +static void CrashHandler(int, siginfo_t *, void *) { + Fuzzer::StaticCrashSignalCallback(); +} + +static void InterruptHandler(int, siginfo_t *, void *) { + Fuzzer::StaticInterruptCallback(); +} + +static void GracefulExitHandler(int, siginfo_t *, void *) { + Fuzzer::StaticGracefulExitCallback(); +} + +static void FileSizeExceedHandler(int, siginfo_t *, void *) { + Fuzzer::StaticFileSizeExceedCallback(); +} + +static void SetSigaction(int signum, + void (*callback)(int, siginfo_t *, void *)) { + struct sigaction sigact = {}; + if (sigaction(signum, nullptr, &sigact)) { + Printf("libFuzzer: sigaction failed with %d\n", errno); + exit(1); + } + if (sigact.sa_flags & SA_SIGINFO) { + if (sigact.sa_sigaction) + return; + } else { + if (sigact.sa_handler != SIG_DFL && sigact.sa_handler != SIG_IGN && + sigact.sa_handler != SIG_ERR) + return; + } + + sigact = {}; + sigact.sa_sigaction = callback; + if (sigaction(signum, &sigact, 0)) { + Printf("libFuzzer: sigaction failed with %d\n", errno); + exit(1); + } +} + +void SetTimer(int Seconds) { + struct itimerval T { + {Seconds, 0}, { Seconds, 0 } + }; + if (setitimer(ITIMER_REAL, &T, nullptr)) { + Printf("libFuzzer: setitimer failed with %d\n", errno); + exit(1); + } + SetSigaction(SIGALRM, AlarmHandler); +} + +void SetSignalHandler(const FuzzingOptions& Options) { + if (Options.UnitTimeoutSec > 0) + SetTimer(Options.UnitTimeoutSec / 2 + 1); + if (Options.HandleInt) + SetSigaction(SIGINT, InterruptHandler); + if (Options.HandleTerm) + SetSigaction(SIGTERM, InterruptHandler); + if (Options.HandleSegv) + SetSigaction(SIGSEGV, CrashHandler); + if (Options.HandleBus) + SetSigaction(SIGBUS, CrashHandler); + if (Options.HandleAbrt) + SetSigaction(SIGABRT, CrashHandler); + if (Options.HandleIll) + SetSigaction(SIGILL, CrashHandler); + if (Options.HandleFpe) + SetSigaction(SIGFPE, CrashHandler); + if (Options.HandleXfsz) + SetSigaction(SIGXFSZ, FileSizeExceedHandler); + if (Options.HandleUsr1) + SetSigaction(SIGUSR1, GracefulExitHandler); + if (Options.HandleUsr2) + SetSigaction(SIGUSR2, GracefulExitHandler); +} + +void SleepSeconds(int Seconds) { + sleep(Seconds); // Use C API to avoid coverage from instrumented libc++. +} + +unsigned long GetPid() { return (unsigned long)getpid(); } + +size_t GetPeakRSSMb() { + struct rusage usage; + if (getrusage(RUSAGE_SELF, &usage)) + return 0; + if (LIBFUZZER_LINUX) { + // ru_maxrss is in KiB + return usage.ru_maxrss >> 10; + } else if (LIBFUZZER_APPLE) { + // ru_maxrss is in bytes + return usage.ru_maxrss >> 20; + } + assert(0 && "GetPeakRSSMb() is not implemented for your platform"); + return 0; +} + +FILE *OpenProcessPipe(const char *Command, const char *Mode) { + return popen(Command, Mode); +} + +const void *SearchMemory(const void *Data, size_t DataLen, const void *Patt, + size_t PattLen) { + return memmem(Data, DataLen, Patt, PattLen); +} + +std::string DisassembleCmd(const std::string &FileName) { + return "objdump -d " + FileName; +} + +std::string SearchRegexCmd(const std::string &Regex) { + return "grep '" + Regex + "'"; +} + +} // namespace fuzzer + +#endif // LIBFUZZER_POSIX diff --git a/lib/fuzzer/FuzzerUtilWindows.cpp b/lib/fuzzer/FuzzerUtilWindows.cpp new file mode 100644 index 000000000000..8227e778ea0a --- /dev/null +++ b/lib/fuzzer/FuzzerUtilWindows.cpp @@ -0,0 +1,194 @@ +//===- FuzzerUtilWindows.cpp - Misc utils for Windows. --------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// Misc utils implementation for Windows. +//===----------------------------------------------------------------------===// +#include "FuzzerDefs.h" +#if LIBFUZZER_WINDOWS +#include "FuzzerCommand.h" +#include "FuzzerIO.h" +#include "FuzzerInternal.h" +#include <cassert> +#include <chrono> +#include <cstring> +#include <errno.h> +#include <iomanip> +#include <signal.h> +#include <stdio.h> +#include <sys/types.h> +#include <windows.h> + +// This must be included after windows.h. +#include <Psapi.h> + +namespace fuzzer { + +static const FuzzingOptions* HandlerOpt = nullptr; + +static LONG CALLBACK ExceptionHandler(PEXCEPTION_POINTERS ExceptionInfo) { + switch (ExceptionInfo->ExceptionRecord->ExceptionCode) { + case EXCEPTION_ACCESS_VIOLATION: + case EXCEPTION_ARRAY_BOUNDS_EXCEEDED: + case EXCEPTION_STACK_OVERFLOW: + if (HandlerOpt->HandleSegv) + Fuzzer::StaticCrashSignalCallback(); + break; + case EXCEPTION_DATATYPE_MISALIGNMENT: + case EXCEPTION_IN_PAGE_ERROR: + if (HandlerOpt->HandleBus) + Fuzzer::StaticCrashSignalCallback(); + break; + case EXCEPTION_ILLEGAL_INSTRUCTION: + case EXCEPTION_PRIV_INSTRUCTION: + if (HandlerOpt->HandleIll) + Fuzzer::StaticCrashSignalCallback(); + break; + case EXCEPTION_FLT_DENORMAL_OPERAND: + case EXCEPTION_FLT_DIVIDE_BY_ZERO: + case EXCEPTION_FLT_INEXACT_RESULT: + case EXCEPTION_FLT_INVALID_OPERATION: + case EXCEPTION_FLT_OVERFLOW: + case EXCEPTION_FLT_STACK_CHECK: + case EXCEPTION_FLT_UNDERFLOW: + case EXCEPTION_INT_DIVIDE_BY_ZERO: + case EXCEPTION_INT_OVERFLOW: + if (HandlerOpt->HandleFpe) + Fuzzer::StaticCrashSignalCallback(); + break; + // TODO: handle (Options.HandleXfsz) + } + return EXCEPTION_CONTINUE_SEARCH; +} + +BOOL WINAPI CtrlHandler(DWORD dwCtrlType) { + switch (dwCtrlType) { + case CTRL_C_EVENT: + if (HandlerOpt->HandleInt) + Fuzzer::StaticInterruptCallback(); + return TRUE; + case CTRL_BREAK_EVENT: + if (HandlerOpt->HandleTerm) + Fuzzer::StaticInterruptCallback(); + return TRUE; + } + return FALSE; +} + +void CALLBACK AlarmHandler(PVOID, BOOLEAN) { + Fuzzer::StaticAlarmCallback(); +} + +class TimerQ { + HANDLE TimerQueue; + public: + TimerQ() : TimerQueue(NULL) {}; + ~TimerQ() { + if (TimerQueue) + DeleteTimerQueueEx(TimerQueue, NULL); + }; + void SetTimer(int Seconds) { + if (!TimerQueue) { + TimerQueue = CreateTimerQueue(); + if (!TimerQueue) { + Printf("libFuzzer: CreateTimerQueue failed.\n"); + exit(1); + } + } + HANDLE Timer; + if (!CreateTimerQueueTimer(&Timer, TimerQueue, AlarmHandler, NULL, + Seconds*1000, Seconds*1000, 0)) { + Printf("libFuzzer: CreateTimerQueueTimer failed.\n"); + exit(1); + } + }; +}; + +static TimerQ Timer; + +static void CrashHandler(int) { Fuzzer::StaticCrashSignalCallback(); } + +void SetSignalHandler(const FuzzingOptions& Options) { + HandlerOpt = &Options; + + if (Options.UnitTimeoutSec > 0) + Timer.SetTimer(Options.UnitTimeoutSec / 2 + 1); + + if (Options.HandleInt || Options.HandleTerm) + if (!SetConsoleCtrlHandler(CtrlHandler, TRUE)) { + DWORD LastError = GetLastError(); + Printf("libFuzzer: SetConsoleCtrlHandler failed (Error code: %lu).\n", + LastError); + exit(1); + } + + if (Options.HandleSegv || Options.HandleBus || Options.HandleIll || + Options.HandleFpe) + SetUnhandledExceptionFilter(ExceptionHandler); + + if (Options.HandleAbrt) + if (SIG_ERR == signal(SIGABRT, CrashHandler)) { + Printf("libFuzzer: signal failed with %d\n", errno); + exit(1); + } +} + +void SleepSeconds(int Seconds) { Sleep(Seconds * 1000); } + +unsigned long GetPid() { return GetCurrentProcessId(); } + +size_t GetPeakRSSMb() { + PROCESS_MEMORY_COUNTERS info; + if (!GetProcessMemoryInfo(GetCurrentProcess(), &info, sizeof(info))) + return 0; + return info.PeakWorkingSetSize >> 20; +} + +FILE *OpenProcessPipe(const char *Command, const char *Mode) { + return _popen(Command, Mode); +} + +int ExecuteCommand(const Command &Cmd) { + std::string CmdLine = Cmd.toString(); + return system(CmdLine.c_str()); +} + +const void *SearchMemory(const void *Data, size_t DataLen, const void *Patt, + size_t PattLen) { + // TODO: make this implementation more efficient. + const char *Cdata = (const char *)Data; + const char *Cpatt = (const char *)Patt; + + if (!Data || !Patt || DataLen == 0 || PattLen == 0 || DataLen < PattLen) + return NULL; + + if (PattLen == 1) + return memchr(Data, *Cpatt, DataLen); + + const char *End = Cdata + DataLen - PattLen + 1; + + for (const char *It = Cdata; It < End; ++It) + if (It[0] == Cpatt[0] && memcmp(It, Cpatt, PattLen) == 0) + return It; + + return NULL; +} + +std::string DisassembleCmd(const std::string &FileName) { + if (ExecuteCommand("dumpbin /summary > nul") == 0) + return "dumpbin /disasm " + FileName; + Printf("libFuzzer: couldn't find tool to disassemble (dumpbin)\n"); + exit(1); +} + +std::string SearchRegexCmd(const std::string &Regex) { + return "findstr /r \"" + Regex + "\""; +} + +} // namespace fuzzer + +#endif // LIBFUZZER_WINDOWS diff --git a/lib/fuzzer/FuzzerValueBitMap.h b/lib/fuzzer/FuzzerValueBitMap.h new file mode 100644 index 000000000000..13d7cbd95dd7 --- /dev/null +++ b/lib/fuzzer/FuzzerValueBitMap.h @@ -0,0 +1,73 @@ +//===- FuzzerValueBitMap.h - INTERNAL - Bit map -----------------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// ValueBitMap. +//===----------------------------------------------------------------------===// + +#ifndef LLVM_FUZZER_VALUE_BIT_MAP_H +#define LLVM_FUZZER_VALUE_BIT_MAP_H + +#include "FuzzerDefs.h" + +namespace fuzzer { + +// A bit map containing kMapSizeInWords bits. +struct ValueBitMap { + static const size_t kMapSizeInBits = 1 << 16; + static const size_t kMapPrimeMod = 65371; // Largest Prime < kMapSizeInBits; + static const size_t kBitsInWord = (sizeof(uintptr_t) * 8); + static const size_t kMapSizeInWords = kMapSizeInBits / kBitsInWord; + public: + + // Clears all bits. + void Reset() { memset(Map, 0, sizeof(Map)); } + + // Computes a hash function of Value and sets the corresponding bit. + // Returns true if the bit was changed from 0 to 1. + ATTRIBUTE_NO_SANITIZE_ALL + inline bool AddValue(uintptr_t Value) { + uintptr_t Idx = Value % kMapSizeInBits; + uintptr_t WordIdx = Idx / kBitsInWord; + uintptr_t BitIdx = Idx % kBitsInWord; + uintptr_t Old = Map[WordIdx]; + uintptr_t New = Old | (1UL << BitIdx); + Map[WordIdx] = New; + return New != Old; + } + + ATTRIBUTE_NO_SANITIZE_ALL + inline bool AddValueModPrime(uintptr_t Value) { + return AddValue(Value % kMapPrimeMod); + } + + inline bool Get(uintptr_t Idx) { + assert(Idx < kMapSizeInBits); + uintptr_t WordIdx = Idx / kBitsInWord; + uintptr_t BitIdx = Idx % kBitsInWord; + return Map[WordIdx] & (1UL << BitIdx); + } + + size_t SizeInBits() const { return kMapSizeInBits; } + + template <class Callback> + ATTRIBUTE_NO_SANITIZE_ALL + void ForEach(Callback CB) const { + for (size_t i = 0; i < kMapSizeInWords; i++) + if (uintptr_t M = Map[i]) + for (size_t j = 0; j < sizeof(M) * 8; j++) + if (M & ((uintptr_t)1 << j)) + CB(i * sizeof(M) * 8 + j); + } + + private: + uintptr_t Map[kMapSizeInWords] __attribute__((aligned(512))); +}; + +} // namespace fuzzer + +#endif // LLVM_FUZZER_VALUE_BIT_MAP_H diff --git a/lib/fuzzer/README.txt b/lib/fuzzer/README.txt new file mode 100644 index 000000000000..3eee01c77672 --- /dev/null +++ b/lib/fuzzer/README.txt @@ -0,0 +1 @@ +See http://llvm.org/docs/LibFuzzer.html diff --git a/lib/fuzzer/afl/afl_driver.cpp b/lib/fuzzer/afl/afl_driver.cpp new file mode 100644 index 000000000000..bbe5be795ed4 --- /dev/null +++ b/lib/fuzzer/afl/afl_driver.cpp @@ -0,0 +1,347 @@ +//===- afl_driver.cpp - a glue between AFL and libFuzzer --------*- C++ -* ===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +//===----------------------------------------------------------------------===// + +/* This file allows to fuzz libFuzzer-style target functions + (LLVMFuzzerTestOneInput) with AFL using AFL's persistent (in-process) mode. + +Usage: +################################################################################ +cat << EOF > test_fuzzer.cc +#include <stddef.h> +#include <stdint.h> +extern "C" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) { + if (size > 0 && data[0] == 'H') + if (size > 1 && data[1] == 'I') + if (size > 2 && data[2] == '!') + __builtin_trap(); + return 0; +} +EOF +# Build your target with -fsanitize-coverage=trace-pc-guard using fresh clang. +clang -g -fsanitize-coverage=trace-pc-guard test_fuzzer.cc -c +# Build afl-llvm-rt.o.c from the AFL distribution. +clang -c -w $AFL_HOME/llvm_mode/afl-llvm-rt.o.c +# Build this file, link it with afl-llvm-rt.o.o and the target code. +clang++ afl_driver.cpp test_fuzzer.o afl-llvm-rt.o.o +# Run AFL: +rm -rf IN OUT; mkdir IN OUT; echo z > IN/z; +$AFL_HOME/afl-fuzz -i IN -o OUT ./a.out +################################################################################ +Environment Variables: +There are a few environment variables that can be set to use features that +afl-fuzz doesn't have. + +AFL_DRIVER_STDERR_DUPLICATE_FILENAME: Setting this *appends* stderr to the file +specified. If the file does not exist, it is created. This is useful for getting +stack traces (when using ASAN for example) or original error messages on hard to +reproduce bugs. + +AFL_DRIVER_EXTRA_STATS_FILENAME: Setting this causes afl_driver to write extra +statistics to the file specified. Currently these are peak_rss_mb +(the peak amount of virtual memory used in MB) and slowest_unit_time_secs. If +the file does not exist it is created. If the file does exist then +afl_driver assumes it was restarted by afl-fuzz and will try to read old +statistics from the file. If that fails then the process will quit. + +*/ +#include <assert.h> +#include <errno.h> +#include <signal.h> +#include <stdint.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/resource.h> +#include <sys/time.h> +#include <unistd.h> + +#include <fstream> +#include <iostream> +#include <vector> + +// Platform detection. Copied from FuzzerInternal.h +#ifdef __linux__ +#define LIBFUZZER_LINUX 1 +#define LIBFUZZER_APPLE 0 +#define LIBFUZZER_NETBSD 0 +#elif __APPLE__ +#define LIBFUZZER_LINUX 0 +#define LIBFUZZER_APPLE 1 +#define LIBFUZZER_NETBSD 0 +#elif __NetBSD__ +#define LIBFUZZER_LINUX 0 +#define LIBFUZZER_APPLE 0 +#define LIBFUZZER_NETBSD 1 +#else +#error "Support for your platform has not been implemented" +#endif + +// Used to avoid repeating error checking boilerplate. If cond is false, a +// fatal error has occured in the program. In this event print error_message +// to stderr and abort(). Otherwise do nothing. Note that setting +// AFL_DRIVER_STDERR_DUPLICATE_FILENAME may cause error_message to be appended +// to the file as well, if the error occurs after the duplication is performed. +#define CHECK_ERROR(cond, error_message) \ + if (!(cond)) { \ + fprintf(stderr, "%s\n", (error_message)); \ + abort(); \ + } + +// libFuzzer interface is thin, so we don't include any libFuzzer headers. +extern "C" { +int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size); +__attribute__((weak)) int LLVMFuzzerInitialize(int *argc, char ***argv); +} + +// Notify AFL about persistent mode. +static volatile char AFL_PERSISTENT[] = "##SIG_AFL_PERSISTENT##"; +extern "C" int __afl_persistent_loop(unsigned int); +static volatile char suppress_warning2 = AFL_PERSISTENT[0]; + +// Notify AFL about deferred forkserver. +static volatile char AFL_DEFER_FORKSVR[] = "##SIG_AFL_DEFER_FORKSRV##"; +extern "C" void __afl_manual_init(); +static volatile char suppress_warning1 = AFL_DEFER_FORKSVR[0]; + +// Input buffer. +static const size_t kMaxAflInputSize = 1 << 20; +static uint8_t AflInputBuf[kMaxAflInputSize]; + +// Variables we need for writing to the extra stats file. +static FILE *extra_stats_file = NULL; +static uint32_t previous_peak_rss = 0; +static time_t slowest_unit_time_secs = 0; +static const int kNumExtraStats = 2; +static const char *kExtraStatsFormatString = "peak_rss_mb : %u\n" + "slowest_unit_time_sec : %u\n"; + +// Copied from FuzzerUtil.cpp. +size_t GetPeakRSSMb() { + struct rusage usage; + if (getrusage(RUSAGE_SELF, &usage)) + return 0; + if (LIBFUZZER_LINUX || LIBFUZZER_NETBSD) { + // ru_maxrss is in KiB + return usage.ru_maxrss >> 10; + } else if (LIBFUZZER_APPLE) { + // ru_maxrss is in bytes + return usage.ru_maxrss >> 20; + } + assert(0 && "GetPeakRSSMb() is not implemented for your platform"); + return 0; +} + +// Based on SetSigaction in FuzzerUtil.cpp +static void SetSigaction(int signum, + void (*callback)(int, siginfo_t *, void *)) { + struct sigaction sigact; + memset(&sigact, 0, sizeof(sigact)); + sigact.sa_sigaction = callback; + if (sigaction(signum, &sigact, 0)) { + fprintf(stderr, "libFuzzer: sigaction failed with %d\n", errno); + exit(1); + } +} + +// Write extra stats to the file specified by the user. If none is specified +// this function will never be called. +static void write_extra_stats() { + uint32_t peak_rss = GetPeakRSSMb(); + + if (peak_rss < previous_peak_rss) + peak_rss = previous_peak_rss; + + int chars_printed = fprintf(extra_stats_file, kExtraStatsFormatString, + peak_rss, slowest_unit_time_secs); + + CHECK_ERROR(chars_printed != 0, "Failed to write extra_stats_file"); + + CHECK_ERROR(fclose(extra_stats_file) == 0, + "Failed to close extra_stats_file"); +} + +// Call write_extra_stats before we exit. +static void crash_handler(int, siginfo_t *, void *) { + // Make sure we don't try calling write_extra_stats again if we crashed while + // trying to call it. + static bool first_crash = true; + CHECK_ERROR(first_crash, + "Crashed in crash signal handler. This is a bug in the fuzzer."); + + first_crash = false; + write_extra_stats(); +} + +// If the user has specified an extra_stats_file through the environment +// variable AFL_DRIVER_EXTRA_STATS_FILENAME, then perform necessary set up +// to write stats to it on exit. If no file is specified, do nothing. Otherwise +// install signal and exit handlers to write to the file when the process exits. +// Then if the file doesn't exist create it and set extra stats to 0. But if it +// does exist then read the initial values of the extra stats from the file +// and check that the file is writable. +static void maybe_initialize_extra_stats() { + // If AFL_DRIVER_EXTRA_STATS_FILENAME isn't set then we have nothing to do. + char *extra_stats_filename = getenv("AFL_DRIVER_EXTRA_STATS_FILENAME"); + if (!extra_stats_filename) + return; + + // Open the file and find the previous peak_rss_mb value. + // This is necessary because the fuzzing process is restarted after N + // iterations are completed. So we may need to get this value from a previous + // process to be accurate. + extra_stats_file = fopen(extra_stats_filename, "r"); + + // If extra_stats_file already exists: read old stats from it. + if (extra_stats_file) { + int matches = fscanf(extra_stats_file, kExtraStatsFormatString, + &previous_peak_rss, &slowest_unit_time_secs); + + // Make sure we have read a real extra stats file and that we have used it + // to set slowest_unit_time_secs and previous_peak_rss. + CHECK_ERROR(matches == kNumExtraStats, "Extra stats file is corrupt"); + + CHECK_ERROR(fclose(extra_stats_file) == 0, "Failed to close file"); + + // Now open the file for writing. + extra_stats_file = fopen(extra_stats_filename, "w"); + CHECK_ERROR(extra_stats_file, + "Failed to open extra stats file for writing"); + } else { + // Looks like this is the first time in a fuzzing job this is being called. + extra_stats_file = fopen(extra_stats_filename, "w+"); + CHECK_ERROR(extra_stats_file, "failed to create extra stats file"); + } + + // Make sure that crash_handler gets called on any kind of fatal error. + int crash_signals[] = {SIGSEGV, SIGBUS, SIGABRT, SIGILL, SIGFPE, SIGINT, + SIGTERM}; + + const size_t num_signals = sizeof(crash_signals) / sizeof(crash_signals[0]); + + for (size_t idx = 0; idx < num_signals; idx++) + SetSigaction(crash_signals[idx], crash_handler); + + // Make sure it gets called on other kinds of exits. + atexit(write_extra_stats); +} + +// If the user asks us to duplicate stderr, then do it. +static void maybe_duplicate_stderr() { + char* stderr_duplicate_filename = + getenv("AFL_DRIVER_STDERR_DUPLICATE_FILENAME"); + + if (!stderr_duplicate_filename) + return; + + FILE* stderr_duplicate_stream = + freopen(stderr_duplicate_filename, "a+", stderr); + + if (!stderr_duplicate_stream) { + fprintf( + stderr, + "Failed to duplicate stderr to AFL_DRIVER_STDERR_DUPLICATE_FILENAME"); + abort(); + } +} + +// Define LLVMFuzzerMutate to avoid link failures for targets that use it +// with libFuzzer's LLVMFuzzerCustomMutator. +extern "C" size_t LLVMFuzzerMutate(uint8_t *Data, size_t Size, size_t MaxSize) { + assert(false && "LLVMFuzzerMutate should not be called from afl_driver"); + return 0; +} + +// Execute any files provided as parameters. +int ExecuteFilesOnyByOne(int argc, char **argv) { + for (int i = 1; i < argc; i++) { + std::ifstream in(argv[i]); + in.seekg(0, in.end); + size_t length = in.tellg(); + in.seekg (0, in.beg); + std::cout << "Reading " << length << " bytes from " << argv[i] << std::endl; + // Allocate exactly length bytes so that we reliably catch buffer overflows. + std::vector<char> bytes(length); + in.read(bytes.data(), bytes.size()); + assert(in); + LLVMFuzzerTestOneInput(reinterpret_cast<const uint8_t *>(bytes.data()), + bytes.size()); + std::cout << "Execution successfull" << std::endl; + } + return 0; +} + +int main(int argc, char **argv) { + fprintf(stderr, + "======================= INFO =========================\n" + "This binary is built for AFL-fuzz.\n" + "To run the target function on individual input(s) execute this:\n" + " %s < INPUT_FILE\n" + "or\n" + " %s INPUT_FILE1 [INPUT_FILE2 ... ]\n" + "To fuzz with afl-fuzz execute this:\n" + " afl-fuzz [afl-flags] %s [-N]\n" + "afl-fuzz will run N iterations before " + "re-spawning the process (default: 1000)\n" + "======================================================\n", + argv[0], argv[0], argv[0]); + if (LLVMFuzzerInitialize) + LLVMFuzzerInitialize(&argc, &argv); + // Do any other expensive one-time initialization here. + + maybe_duplicate_stderr(); + maybe_initialize_extra_stats(); + + __afl_manual_init(); + + int N = 1000; + if (argc == 2 && argv[1][0] == '-') + N = atoi(argv[1] + 1); + else if(argc == 2 && (N = atoi(argv[1])) > 0) + fprintf(stderr, "WARNING: using the deprecated call style `%s %d`\n", + argv[0], N); + else if (argc > 1) + return ExecuteFilesOnyByOne(argc, argv); + + assert(N > 0); + + // Call LLVMFuzzerTestOneInput here so that coverage caused by initialization + // on the first execution of LLVMFuzzerTestOneInput is ignored. + uint8_t dummy_input[1] = {0}; + LLVMFuzzerTestOneInput(dummy_input, 1); + + time_t unit_time_secs; + int num_runs = 0; + while (__afl_persistent_loop(N)) { + ssize_t n_read = read(0, AflInputBuf, kMaxAflInputSize); + if (n_read > 0) { + // Copy AflInputBuf into a separate buffer to let asan find buffer + // overflows. Don't use unique_ptr/etc to avoid extra dependencies. + uint8_t *copy = new uint8_t[n_read]; + memcpy(copy, AflInputBuf, n_read); + + struct timeval unit_start_time; + CHECK_ERROR(gettimeofday(&unit_start_time, NULL) == 0, + "Calling gettimeofday failed"); + + num_runs++; + LLVMFuzzerTestOneInput(copy, n_read); + + struct timeval unit_stop_time; + CHECK_ERROR(gettimeofday(&unit_stop_time, NULL) == 0, + "Calling gettimeofday failed"); + + // Update slowest_unit_time_secs if we see a new max. + unit_time_secs = unit_stop_time.tv_sec - unit_start_time.tv_sec; + if (slowest_unit_time_secs < unit_time_secs) + slowest_unit_time_secs = unit_time_secs; + + delete[] copy; + } + } + fprintf(stderr, "%s: successfully executed %d input(s)\n", argv[0], num_runs); +} diff --git a/lib/fuzzer/build.sh b/lib/fuzzer/build.sh new file mode 100755 index 000000000000..4556af5daf7d --- /dev/null +++ b/lib/fuzzer/build.sh @@ -0,0 +1,11 @@ +#!/bin/bash +LIBFUZZER_SRC_DIR=$(dirname $0) +CXX="${CXX:-clang}" +for f in $LIBFUZZER_SRC_DIR/*.cpp; do + $CXX -g -O2 -fno-omit-frame-pointer -std=c++11 $f -c & +done +wait +rm -f libFuzzer.a +ar ru libFuzzer.a Fuzzer*.o +rm -f Fuzzer*.o + diff --git a/lib/fuzzer/scripts/unbalanced_allocs.py b/lib/fuzzer/scripts/unbalanced_allocs.py new file mode 100755 index 000000000000..a4ce187679d7 --- /dev/null +++ b/lib/fuzzer/scripts/unbalanced_allocs.py @@ -0,0 +1,93 @@ +#!/usr/bin/env python +#===- lib/fuzzer/scripts/unbalanced_allocs.py ------------------------------===# +# +# The LLVM Compiler Infrastructure +# +# This file is distributed under the University of Illinois Open Source +# License. See LICENSE.TXT for details. +# +#===------------------------------------------------------------------------===# +# +# Post-process -trace_malloc=2 output and printout only allocations and frees +# unbalanced inside of fuzzer runs. +# Usage: +# my_fuzzer -trace_malloc=2 -runs=10 2>&1 | unbalanced_allocs.py -skip=5 +# +#===------------------------------------------------------------------------===# + +import argparse +import sys + +_skip = 0 + +def PrintStack(line, stack): + global _skip + if _skip > 0: + return + print 'Unbalanced ' + line.rstrip(); + for l in stack: + print l.rstrip() + +def ProcessStack(line, f): + stack = [] + while line and line.startswith(' #'): + stack += [line] + line = f.readline() + return line, stack + +def ProcessFree(line, f, allocs): + if not line.startswith('FREE['): + return f.readline() + + addr = int(line.split()[1], 16) + next_line, stack = ProcessStack(f.readline(), f) + if addr in allocs: + del allocs[addr] + else: + PrintStack(line, stack) + return next_line + +def ProcessMalloc(line, f, allocs): + if not line.startswith('MALLOC['): + return ProcessFree(line, f, allocs) + + addr = int(line.split()[1], 16) + assert not addr in allocs + + next_line, stack = ProcessStack(f.readline(), f) + allocs[addr] = (line, stack) + return next_line + +def ProcessRun(line, f): + if not line.startswith('MallocFreeTracer: START'): + return ProcessMalloc(line, f, {}) + + allocs = {} + print line.rstrip() + line = f.readline() + while line: + if line.startswith('MallocFreeTracer: STOP'): + global _skip + _skip = _skip - 1 + for _, (l, s) in allocs.iteritems(): + PrintStack(l, s) + print line.rstrip() + return f.readline() + line = ProcessMalloc(line, f, allocs) + return line + +def ProcessFile(f): + line = f.readline() + while line: + line = ProcessRun(line, f); + +def main(argv): + parser = argparse.ArgumentParser() + parser.add_argument('--skip', default=0, help='number of runs to ignore') + args = parser.parse_args() + global _skip + _skip = int(args.skip) + 1 + ProcessFile(sys.stdin) + +if __name__ == '__main__': + main(sys.argv) diff --git a/lib/fuzzer/standalone/StandaloneFuzzTargetMain.c b/lib/fuzzer/standalone/StandaloneFuzzTargetMain.c new file mode 100644 index 000000000000..0d76ea49e796 --- /dev/null +++ b/lib/fuzzer/standalone/StandaloneFuzzTargetMain.c @@ -0,0 +1,41 @@ +/*===- StandaloneFuzzTargetMain.c - standalone main() for fuzz targets. ---===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// This main() function can be linked to a fuzz target (i.e. a library +// that exports LLVMFuzzerTestOneInput() and possibly LLVMFuzzerInitialize()) +// instead of libFuzzer. This main() function will not perform any fuzzing +// but will simply feed all input files one by one to the fuzz target. +// +// Use this file to provide reproducers for bugs when linking against libFuzzer +// or other fuzzing engine is undesirable. +//===----------------------------------------------------------------------===*/ +#include <assert.h> +#include <stdio.h> +#include <stdlib.h> + +extern int LLVMFuzzerTestOneInput(const unsigned char *data, size_t size); +__attribute__((weak)) extern int LLVMFuzzerInitialize(int *argc, char ***argv); +int main(int argc, char **argv) { + fprintf(stderr, "StandaloneFuzzTargetMain: running %d inputs\n", argc - 1); + if (LLVMFuzzerInitialize) + LLVMFuzzerInitialize(&argc, &argv); + for (int i = 1; i < argc; i++) { + fprintf(stderr, "Running: %s\n", argv[i]); + FILE *f = fopen(argv[i], "r"); + assert(f); + fseek(f, 0, SEEK_END); + size_t len = ftell(f); + fseek(f, 0, SEEK_SET); + unsigned char *buf = (unsigned char*)malloc(len); + size_t n_read = fread(buf, 1, len, f); + assert(n_read == len); + LLVMFuzzerTestOneInput(buf, len); + free(buf); + fprintf(stderr, "Done: %s: (%zd bytes)\n", argv[i], n_read); + } +} diff --git a/lib/fuzzer/tests/CMakeLists.txt b/lib/fuzzer/tests/CMakeLists.txt new file mode 100644 index 000000000000..dac8773597e8 --- /dev/null +++ b/lib/fuzzer/tests/CMakeLists.txt @@ -0,0 +1,46 @@ +set(LIBFUZZER_UNITTEST_CFLAGS + ${COMPILER_RT_UNITTEST_CFLAGS} + ${COMPILER_RT_GTEST_CFLAGS} + -I${COMPILER_RT_SOURCE_DIR}/lib/fuzzer + -fno-rtti + -Werror + -O2) + +add_custom_target(FuzzerUnitTests) +set_target_properties(FuzzerUnitTests PROPERTIES FOLDER "Compiler-RT Tests") + +set(LIBFUZZER_UNITTEST_LINK_FLAGS ${COMPILER_RT_UNITTEST_LINK_FLAGS}) +list(APPEND LIBFUZZER_UNITTEST_LINK_FLAGS --driver-mode=g++) + +if(APPLE) + list(APPEND LIBFUZZER_UNITTEST_LINK_FLAGS -lc++) +else() + list(APPEND LIBFUZZER_UNITTEST_LINK_FLAGS -lstdc++ -lpthread) +endif() + +foreach(arch ${FUZZER_SUPPORTED_ARCH}) + set(LIBFUZZER_TEST_RUNTIME RTFuzzerTest.${arch}) + if(APPLE) + set(LIBFUZZER_TEST_RUNTIME_OBJECTS + $<TARGET_OBJECTS:RTfuzzer.osx>) + else() + set(LIBFUZZER_TEST_RUNTIME_OBJECTS + $<TARGET_OBJECTS:RTfuzzer.${arch}>) + endif() + add_library(${LIBFUZZER_TEST_RUNTIME} STATIC + ${LIBFUZZER_TEST_RUNTIME_OBJECTS}) + set_target_properties(${LIBFUZZER_TEST_RUNTIME} PROPERTIES + ARCHIVE_OUTPUT_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR} + FOLDER "Compiler-RT Runtime tests") + + set(FuzzerTestObjects) + generate_compiler_rt_tests(FuzzerTestObjects + FuzzerUnitTests "Fuzzer-${arch}-Test" ${arch} + SOURCES FuzzerUnittest.cpp ${COMPILER_RT_GTEST_SOURCE} + RUNTIME ${LIBFUZZER_TEST_RUNTIME} + DEPS gtest + CFLAGS ${LIBFUZZER_UNITTEST_CFLAGS} + LINK_FLAGS ${LIBFUZZER_UNITTEST_LINK_FLAGS}) + set_target_properties(FuzzerUnitTests PROPERTIES + RUNTIME_OUTPUT_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}) +endforeach() diff --git a/lib/fuzzer/tests/FuzzerUnittest.cpp b/lib/fuzzer/tests/FuzzerUnittest.cpp new file mode 100644 index 000000000000..97ec3b4bb6af --- /dev/null +++ b/lib/fuzzer/tests/FuzzerUnittest.cpp @@ -0,0 +1,932 @@ +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. + +// Avoid ODR violations (LibFuzzer is built without ASan and this test is built +// with ASan) involving C++ standard library types when using libcxx. +#define _LIBCPP_HAS_NO_ASAN + +// Do not attempt to use LLVM ostream from gtest. +#define GTEST_NO_LLVM_RAW_OSTREAM 1 + +#include "FuzzerCorpus.h" +#include "FuzzerDictionary.h" +#include "FuzzerInternal.h" +#include "FuzzerMerge.h" +#include "FuzzerMutate.h" +#include "FuzzerRandom.h" +#include "FuzzerTracePC.h" +#include "gtest/gtest.h" +#include <memory> +#include <set> +#include <sstream> + +using namespace fuzzer; + +// For now, have LLVMFuzzerTestOneInput just to make it link. +// Later we may want to make unittests that actually call LLVMFuzzerTestOneInput. +extern "C" int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size) { + abort(); +} + +TEST(Fuzzer, CrossOver) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + Unit A({0, 1, 2}), B({5, 6, 7}); + Unit C; + Unit Expected[] = { + { 0 }, + { 0, 1 }, + { 0, 5 }, + { 0, 1, 2 }, + { 0, 1, 5 }, + { 0, 5, 1 }, + { 0, 5, 6 }, + { 0, 1, 2, 5 }, + { 0, 1, 5, 2 }, + { 0, 1, 5, 6 }, + { 0, 5, 1, 2 }, + { 0, 5, 1, 6 }, + { 0, 5, 6, 1 }, + { 0, 5, 6, 7 }, + { 0, 1, 2, 5, 6 }, + { 0, 1, 5, 2, 6 }, + { 0, 1, 5, 6, 2 }, + { 0, 1, 5, 6, 7 }, + { 0, 5, 1, 2, 6 }, + { 0, 5, 1, 6, 2 }, + { 0, 5, 1, 6, 7 }, + { 0, 5, 6, 1, 2 }, + { 0, 5, 6, 1, 7 }, + { 0, 5, 6, 7, 1 }, + { 0, 1, 2, 5, 6, 7 }, + { 0, 1, 5, 2, 6, 7 }, + { 0, 1, 5, 6, 2, 7 }, + { 0, 1, 5, 6, 7, 2 }, + { 0, 5, 1, 2, 6, 7 }, + { 0, 5, 1, 6, 2, 7 }, + { 0, 5, 1, 6, 7, 2 }, + { 0, 5, 6, 1, 2, 7 }, + { 0, 5, 6, 1, 7, 2 }, + { 0, 5, 6, 7, 1, 2 } + }; + for (size_t Len = 1; Len < 8; Len++) { + Set<Unit> FoundUnits, ExpectedUnitsWitThisLength; + for (int Iter = 0; Iter < 3000; Iter++) { + C.resize(Len); + size_t NewSize = MD->CrossOver(A.data(), A.size(), B.data(), B.size(), + C.data(), C.size()); + C.resize(NewSize); + FoundUnits.insert(C); + } + for (const Unit &U : Expected) + if (U.size() <= Len) + ExpectedUnitsWitThisLength.insert(U); + EXPECT_EQ(ExpectedUnitsWitThisLength, FoundUnits); + } +} + +TEST(Fuzzer, Hash) { + uint8_t A[] = {'a', 'b', 'c'}; + fuzzer::Unit U(A, A + sizeof(A)); + EXPECT_EQ("a9993e364706816aba3e25717850c26c9cd0d89d", fuzzer::Hash(U)); + U.push_back('d'); + EXPECT_EQ("81fe8bfe87576c3ecb22426f8e57847382917acf", fuzzer::Hash(U)); +} + +typedef size_t (MutationDispatcher::*Mutator)(uint8_t *Data, size_t Size, + size_t MaxSize); + +void TestEraseBytes(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + uint8_t REM0[8] = {0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t REM1[8] = {0x00, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t REM2[8] = {0x00, 0x11, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t REM3[8] = {0x00, 0x11, 0x22, 0x44, 0x55, 0x66, 0x77}; + uint8_t REM4[8] = {0x00, 0x11, 0x22, 0x33, 0x55, 0x66, 0x77}; + uint8_t REM5[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x66, 0x77}; + uint8_t REM6[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x77}; + uint8_t REM7[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66}; + + uint8_t REM8[6] = {0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t REM9[6] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55}; + uint8_t REM10[6] = {0x00, 0x11, 0x22, 0x55, 0x66, 0x77}; + + uint8_t REM11[5] = {0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t REM12[5] = {0x00, 0x11, 0x22, 0x33, 0x44}; + uint8_t REM13[5] = {0x00, 0x44, 0x55, 0x66, 0x77}; + + + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + int FoundMask = 0; + for (int i = 0; i < NumIter; i++) { + uint8_t T[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + size_t NewSize = (*MD.*M)(T, sizeof(T), sizeof(T)); + if (NewSize == 7 && !memcmp(REM0, T, 7)) FoundMask |= 1 << 0; + if (NewSize == 7 && !memcmp(REM1, T, 7)) FoundMask |= 1 << 1; + if (NewSize == 7 && !memcmp(REM2, T, 7)) FoundMask |= 1 << 2; + if (NewSize == 7 && !memcmp(REM3, T, 7)) FoundMask |= 1 << 3; + if (NewSize == 7 && !memcmp(REM4, T, 7)) FoundMask |= 1 << 4; + if (NewSize == 7 && !memcmp(REM5, T, 7)) FoundMask |= 1 << 5; + if (NewSize == 7 && !memcmp(REM6, T, 7)) FoundMask |= 1 << 6; + if (NewSize == 7 && !memcmp(REM7, T, 7)) FoundMask |= 1 << 7; + + if (NewSize == 6 && !memcmp(REM8, T, 6)) FoundMask |= 1 << 8; + if (NewSize == 6 && !memcmp(REM9, T, 6)) FoundMask |= 1 << 9; + if (NewSize == 6 && !memcmp(REM10, T, 6)) FoundMask |= 1 << 10; + + if (NewSize == 5 && !memcmp(REM11, T, 5)) FoundMask |= 1 << 11; + if (NewSize == 5 && !memcmp(REM12, T, 5)) FoundMask |= 1 << 12; + if (NewSize == 5 && !memcmp(REM13, T, 5)) FoundMask |= 1 << 13; + } + EXPECT_EQ(FoundMask, (1 << 14) - 1); +} + +TEST(FuzzerMutate, EraseBytes1) { + TestEraseBytes(&MutationDispatcher::Mutate_EraseBytes, 200); +} +TEST(FuzzerMutate, EraseBytes2) { + TestEraseBytes(&MutationDispatcher::Mutate, 2000); +} + +void TestInsertByte(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + int FoundMask = 0; + uint8_t INS0[8] = {0xF1, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66}; + uint8_t INS1[8] = {0x00, 0xF2, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66}; + uint8_t INS2[8] = {0x00, 0x11, 0xF3, 0x22, 0x33, 0x44, 0x55, 0x66}; + uint8_t INS3[8] = {0x00, 0x11, 0x22, 0xF4, 0x33, 0x44, 0x55, 0x66}; + uint8_t INS4[8] = {0x00, 0x11, 0x22, 0x33, 0xF5, 0x44, 0x55, 0x66}; + uint8_t INS5[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0xF6, 0x55, 0x66}; + uint8_t INS6[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0xF7, 0x66}; + uint8_t INS7[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0xF8}; + for (int i = 0; i < NumIter; i++) { + uint8_t T[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66}; + size_t NewSize = (*MD.*M)(T, 7, 8); + if (NewSize == 8 && !memcmp(INS0, T, 8)) FoundMask |= 1 << 0; + if (NewSize == 8 && !memcmp(INS1, T, 8)) FoundMask |= 1 << 1; + if (NewSize == 8 && !memcmp(INS2, T, 8)) FoundMask |= 1 << 2; + if (NewSize == 8 && !memcmp(INS3, T, 8)) FoundMask |= 1 << 3; + if (NewSize == 8 && !memcmp(INS4, T, 8)) FoundMask |= 1 << 4; + if (NewSize == 8 && !memcmp(INS5, T, 8)) FoundMask |= 1 << 5; + if (NewSize == 8 && !memcmp(INS6, T, 8)) FoundMask |= 1 << 6; + if (NewSize == 8 && !memcmp(INS7, T, 8)) FoundMask |= 1 << 7; + } + EXPECT_EQ(FoundMask, 255); +} + +TEST(FuzzerMutate, InsertByte1) { + TestInsertByte(&MutationDispatcher::Mutate_InsertByte, 1 << 15); +} +TEST(FuzzerMutate, InsertByte2) { + TestInsertByte(&MutationDispatcher::Mutate, 1 << 17); +} + +void TestInsertRepeatedBytes(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + int FoundMask = 0; + uint8_t INS0[7] = {0x00, 0x11, 0x22, 0x33, 'a', 'a', 'a'}; + uint8_t INS1[7] = {0x00, 0x11, 0x22, 'a', 'a', 'a', 0x33}; + uint8_t INS2[7] = {0x00, 0x11, 'a', 'a', 'a', 0x22, 0x33}; + uint8_t INS3[7] = {0x00, 'a', 'a', 'a', 0x11, 0x22, 0x33}; + uint8_t INS4[7] = {'a', 'a', 'a', 0x00, 0x11, 0x22, 0x33}; + + uint8_t INS5[8] = {0x00, 0x11, 0x22, 0x33, 'b', 'b', 'b', 'b'}; + uint8_t INS6[8] = {0x00, 0x11, 0x22, 'b', 'b', 'b', 'b', 0x33}; + uint8_t INS7[8] = {0x00, 0x11, 'b', 'b', 'b', 'b', 0x22, 0x33}; + uint8_t INS8[8] = {0x00, 'b', 'b', 'b', 'b', 0x11, 0x22, 0x33}; + uint8_t INS9[8] = {'b', 'b', 'b', 'b', 0x00, 0x11, 0x22, 0x33}; + + for (int i = 0; i < NumIter; i++) { + uint8_t T[8] = {0x00, 0x11, 0x22, 0x33}; + size_t NewSize = (*MD.*M)(T, 4, 8); + if (NewSize == 7 && !memcmp(INS0, T, 7)) FoundMask |= 1 << 0; + if (NewSize == 7 && !memcmp(INS1, T, 7)) FoundMask |= 1 << 1; + if (NewSize == 7 && !memcmp(INS2, T, 7)) FoundMask |= 1 << 2; + if (NewSize == 7 && !memcmp(INS3, T, 7)) FoundMask |= 1 << 3; + if (NewSize == 7 && !memcmp(INS4, T, 7)) FoundMask |= 1 << 4; + + if (NewSize == 8 && !memcmp(INS5, T, 8)) FoundMask |= 1 << 5; + if (NewSize == 8 && !memcmp(INS6, T, 8)) FoundMask |= 1 << 6; + if (NewSize == 8 && !memcmp(INS7, T, 8)) FoundMask |= 1 << 7; + if (NewSize == 8 && !memcmp(INS8, T, 8)) FoundMask |= 1 << 8; + if (NewSize == 8 && !memcmp(INS9, T, 8)) FoundMask |= 1 << 9; + + } + EXPECT_EQ(FoundMask, (1 << 10) - 1); +} + +TEST(FuzzerMutate, InsertRepeatedBytes1) { + TestInsertRepeatedBytes(&MutationDispatcher::Mutate_InsertRepeatedBytes, 10000); +} +TEST(FuzzerMutate, InsertRepeatedBytes2) { + TestInsertRepeatedBytes(&MutationDispatcher::Mutate, 300000); +} + +void TestChangeByte(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + int FoundMask = 0; + uint8_t CH0[8] = {0xF0, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH1[8] = {0x00, 0xF1, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH2[8] = {0x00, 0x11, 0xF2, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH3[8] = {0x00, 0x11, 0x22, 0xF3, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH4[8] = {0x00, 0x11, 0x22, 0x33, 0xF4, 0x55, 0x66, 0x77}; + uint8_t CH5[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0xF5, 0x66, 0x77}; + uint8_t CH6[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0xF5, 0x77}; + uint8_t CH7[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0xF7}; + for (int i = 0; i < NumIter; i++) { + uint8_t T[9] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + size_t NewSize = (*MD.*M)(T, 8, 9); + if (NewSize == 8 && !memcmp(CH0, T, 8)) FoundMask |= 1 << 0; + if (NewSize == 8 && !memcmp(CH1, T, 8)) FoundMask |= 1 << 1; + if (NewSize == 8 && !memcmp(CH2, T, 8)) FoundMask |= 1 << 2; + if (NewSize == 8 && !memcmp(CH3, T, 8)) FoundMask |= 1 << 3; + if (NewSize == 8 && !memcmp(CH4, T, 8)) FoundMask |= 1 << 4; + if (NewSize == 8 && !memcmp(CH5, T, 8)) FoundMask |= 1 << 5; + if (NewSize == 8 && !memcmp(CH6, T, 8)) FoundMask |= 1 << 6; + if (NewSize == 8 && !memcmp(CH7, T, 8)) FoundMask |= 1 << 7; + } + EXPECT_EQ(FoundMask, 255); +} + +TEST(FuzzerMutate, ChangeByte1) { + TestChangeByte(&MutationDispatcher::Mutate_ChangeByte, 1 << 15); +} +TEST(FuzzerMutate, ChangeByte2) { + TestChangeByte(&MutationDispatcher::Mutate, 1 << 17); +} + +void TestChangeBit(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + int FoundMask = 0; + uint8_t CH0[8] = {0x01, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH1[8] = {0x00, 0x13, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH2[8] = {0x00, 0x11, 0x02, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH3[8] = {0x00, 0x11, 0x22, 0x37, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH4[8] = {0x00, 0x11, 0x22, 0x33, 0x54, 0x55, 0x66, 0x77}; + uint8_t CH5[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x54, 0x66, 0x77}; + uint8_t CH6[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x76, 0x77}; + uint8_t CH7[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0xF7}; + for (int i = 0; i < NumIter; i++) { + uint8_t T[9] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + size_t NewSize = (*MD.*M)(T, 8, 9); + if (NewSize == 8 && !memcmp(CH0, T, 8)) FoundMask |= 1 << 0; + if (NewSize == 8 && !memcmp(CH1, T, 8)) FoundMask |= 1 << 1; + if (NewSize == 8 && !memcmp(CH2, T, 8)) FoundMask |= 1 << 2; + if (NewSize == 8 && !memcmp(CH3, T, 8)) FoundMask |= 1 << 3; + if (NewSize == 8 && !memcmp(CH4, T, 8)) FoundMask |= 1 << 4; + if (NewSize == 8 && !memcmp(CH5, T, 8)) FoundMask |= 1 << 5; + if (NewSize == 8 && !memcmp(CH6, T, 8)) FoundMask |= 1 << 6; + if (NewSize == 8 && !memcmp(CH7, T, 8)) FoundMask |= 1 << 7; + } + EXPECT_EQ(FoundMask, 255); +} + +TEST(FuzzerMutate, ChangeBit1) { + TestChangeBit(&MutationDispatcher::Mutate_ChangeBit, 1 << 16); +} +TEST(FuzzerMutate, ChangeBit2) { + TestChangeBit(&MutationDispatcher::Mutate, 1 << 18); +} + +void TestShuffleBytes(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + int FoundMask = 0; + uint8_t CH0[7] = {0x00, 0x22, 0x11, 0x33, 0x44, 0x55, 0x66}; + uint8_t CH1[7] = {0x11, 0x00, 0x33, 0x22, 0x44, 0x55, 0x66}; + uint8_t CH2[7] = {0x00, 0x33, 0x11, 0x22, 0x44, 0x55, 0x66}; + uint8_t CH3[7] = {0x00, 0x11, 0x22, 0x44, 0x55, 0x66, 0x33}; + uint8_t CH4[7] = {0x00, 0x11, 0x22, 0x33, 0x55, 0x44, 0x66}; + for (int i = 0; i < NumIter; i++) { + uint8_t T[7] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66}; + size_t NewSize = (*MD.*M)(T, 7, 7); + if (NewSize == 7 && !memcmp(CH0, T, 7)) FoundMask |= 1 << 0; + if (NewSize == 7 && !memcmp(CH1, T, 7)) FoundMask |= 1 << 1; + if (NewSize == 7 && !memcmp(CH2, T, 7)) FoundMask |= 1 << 2; + if (NewSize == 7 && !memcmp(CH3, T, 7)) FoundMask |= 1 << 3; + if (NewSize == 7 && !memcmp(CH4, T, 7)) FoundMask |= 1 << 4; + } + EXPECT_EQ(FoundMask, 31); +} + +TEST(FuzzerMutate, ShuffleBytes1) { + TestShuffleBytes(&MutationDispatcher::Mutate_ShuffleBytes, 1 << 16); +} +TEST(FuzzerMutate, ShuffleBytes2) { + TestShuffleBytes(&MutationDispatcher::Mutate, 1 << 20); +} + +void TestCopyPart(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + int FoundMask = 0; + uint8_t CH0[7] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x00, 0x11}; + uint8_t CH1[7] = {0x55, 0x66, 0x22, 0x33, 0x44, 0x55, 0x66}; + uint8_t CH2[7] = {0x00, 0x55, 0x66, 0x33, 0x44, 0x55, 0x66}; + uint8_t CH3[7] = {0x00, 0x11, 0x22, 0x00, 0x11, 0x22, 0x66}; + uint8_t CH4[7] = {0x00, 0x11, 0x11, 0x22, 0x33, 0x55, 0x66}; + + for (int i = 0; i < NumIter; i++) { + uint8_t T[7] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66}; + size_t NewSize = (*MD.*M)(T, 7, 7); + if (NewSize == 7 && !memcmp(CH0, T, 7)) FoundMask |= 1 << 0; + if (NewSize == 7 && !memcmp(CH1, T, 7)) FoundMask |= 1 << 1; + if (NewSize == 7 && !memcmp(CH2, T, 7)) FoundMask |= 1 << 2; + if (NewSize == 7 && !memcmp(CH3, T, 7)) FoundMask |= 1 << 3; + if (NewSize == 7 && !memcmp(CH4, T, 7)) FoundMask |= 1 << 4; + } + + uint8_t CH5[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x00, 0x11, 0x22}; + uint8_t CH6[8] = {0x22, 0x33, 0x44, 0x00, 0x11, 0x22, 0x33, 0x44}; + uint8_t CH7[8] = {0x00, 0x11, 0x22, 0x00, 0x11, 0x22, 0x33, 0x44}; + uint8_t CH8[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x22, 0x33, 0x44}; + uint8_t CH9[8] = {0x00, 0x11, 0x22, 0x22, 0x33, 0x44, 0x33, 0x44}; + + for (int i = 0; i < NumIter; i++) { + uint8_t T[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + size_t NewSize = (*MD.*M)(T, 5, 8); + if (NewSize == 8 && !memcmp(CH5, T, 8)) FoundMask |= 1 << 5; + if (NewSize == 8 && !memcmp(CH6, T, 8)) FoundMask |= 1 << 6; + if (NewSize == 8 && !memcmp(CH7, T, 8)) FoundMask |= 1 << 7; + if (NewSize == 8 && !memcmp(CH8, T, 8)) FoundMask |= 1 << 8; + if (NewSize == 8 && !memcmp(CH9, T, 8)) FoundMask |= 1 << 9; + } + + EXPECT_EQ(FoundMask, 1023); +} + +TEST(FuzzerMutate, CopyPart1) { + TestCopyPart(&MutationDispatcher::Mutate_CopyPart, 1 << 10); +} +TEST(FuzzerMutate, CopyPart2) { + TestCopyPart(&MutationDispatcher::Mutate, 1 << 13); +} + +void TestAddWordFromDictionary(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + uint8_t Word1[4] = {0xAA, 0xBB, 0xCC, 0xDD}; + uint8_t Word2[3] = {0xFF, 0xEE, 0xEF}; + MD->AddWordToManualDictionary(Word(Word1, sizeof(Word1))); + MD->AddWordToManualDictionary(Word(Word2, sizeof(Word2))); + int FoundMask = 0; + uint8_t CH0[7] = {0x00, 0x11, 0x22, 0xAA, 0xBB, 0xCC, 0xDD}; + uint8_t CH1[7] = {0x00, 0x11, 0xAA, 0xBB, 0xCC, 0xDD, 0x22}; + uint8_t CH2[7] = {0x00, 0xAA, 0xBB, 0xCC, 0xDD, 0x11, 0x22}; + uint8_t CH3[7] = {0xAA, 0xBB, 0xCC, 0xDD, 0x00, 0x11, 0x22}; + uint8_t CH4[6] = {0x00, 0x11, 0x22, 0xFF, 0xEE, 0xEF}; + uint8_t CH5[6] = {0x00, 0x11, 0xFF, 0xEE, 0xEF, 0x22}; + uint8_t CH6[6] = {0x00, 0xFF, 0xEE, 0xEF, 0x11, 0x22}; + uint8_t CH7[6] = {0xFF, 0xEE, 0xEF, 0x00, 0x11, 0x22}; + for (int i = 0; i < NumIter; i++) { + uint8_t T[7] = {0x00, 0x11, 0x22}; + size_t NewSize = (*MD.*M)(T, 3, 7); + if (NewSize == 7 && !memcmp(CH0, T, 7)) FoundMask |= 1 << 0; + if (NewSize == 7 && !memcmp(CH1, T, 7)) FoundMask |= 1 << 1; + if (NewSize == 7 && !memcmp(CH2, T, 7)) FoundMask |= 1 << 2; + if (NewSize == 7 && !memcmp(CH3, T, 7)) FoundMask |= 1 << 3; + if (NewSize == 6 && !memcmp(CH4, T, 6)) FoundMask |= 1 << 4; + if (NewSize == 6 && !memcmp(CH5, T, 6)) FoundMask |= 1 << 5; + if (NewSize == 6 && !memcmp(CH6, T, 6)) FoundMask |= 1 << 6; + if (NewSize == 6 && !memcmp(CH7, T, 6)) FoundMask |= 1 << 7; + } + EXPECT_EQ(FoundMask, 255); +} + +TEST(FuzzerMutate, AddWordFromDictionary1) { + TestAddWordFromDictionary( + &MutationDispatcher::Mutate_AddWordFromManualDictionary, 1 << 15); +} + +TEST(FuzzerMutate, AddWordFromDictionary2) { + TestAddWordFromDictionary(&MutationDispatcher::Mutate, 1 << 15); +} + +void TestChangeASCIIInteger(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + + uint8_t CH0[8] = {'1', '2', '3', '4', '5', '6', '7', '7'}; + uint8_t CH1[8] = {'1', '2', '3', '4', '5', '6', '7', '9'}; + uint8_t CH2[8] = {'2', '4', '6', '9', '1', '3', '5', '6'}; + uint8_t CH3[8] = {'0', '6', '1', '7', '2', '8', '3', '9'}; + int FoundMask = 0; + for (int i = 0; i < NumIter; i++) { + uint8_t T[8] = {'1', '2', '3', '4', '5', '6', '7', '8'}; + size_t NewSize = (*MD.*M)(T, 8, 8); + /**/ if (NewSize == 8 && !memcmp(CH0, T, 8)) FoundMask |= 1 << 0; + else if (NewSize == 8 && !memcmp(CH1, T, 8)) FoundMask |= 1 << 1; + else if (NewSize == 8 && !memcmp(CH2, T, 8)) FoundMask |= 1 << 2; + else if (NewSize == 8 && !memcmp(CH3, T, 8)) FoundMask |= 1 << 3; + else if (NewSize == 8) FoundMask |= 1 << 4; + } + EXPECT_EQ(FoundMask, 31); +} + +TEST(FuzzerMutate, ChangeASCIIInteger1) { + TestChangeASCIIInteger(&MutationDispatcher::Mutate_ChangeASCIIInteger, + 1 << 15); +} + +TEST(FuzzerMutate, ChangeASCIIInteger2) { + TestChangeASCIIInteger(&MutationDispatcher::Mutate, 1 << 15); +} + +void TestChangeBinaryInteger(Mutator M, int NumIter) { + std::unique_ptr<ExternalFunctions> t(new ExternalFunctions()); + fuzzer::EF = t.get(); + Random Rand(0); + std::unique_ptr<MutationDispatcher> MD(new MutationDispatcher(Rand, {})); + + uint8_t CH0[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x79}; + uint8_t CH1[8] = {0x00, 0x11, 0x22, 0x31, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH2[8] = {0xff, 0x10, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH3[8] = {0x00, 0x11, 0x2a, 0x33, 0x44, 0x55, 0x66, 0x77}; + uint8_t CH4[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x4f, 0x66, 0x77}; + uint8_t CH5[8] = {0xff, 0xee, 0xdd, 0xcc, 0xbb, 0xaa, 0x99, 0x88}; + uint8_t CH6[8] = {0x00, 0x11, 0x22, 0x00, 0x00, 0x00, 0x08, 0x77}; // Size + uint8_t CH7[8] = {0x00, 0x08, 0x00, 0x33, 0x44, 0x55, 0x66, 0x77}; // Sw(Size) + + int FoundMask = 0; + for (int i = 0; i < NumIter; i++) { + uint8_t T[8] = {0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77}; + size_t NewSize = (*MD.*M)(T, 8, 8); + /**/ if (NewSize == 8 && !memcmp(CH0, T, 8)) FoundMask |= 1 << 0; + else if (NewSize == 8 && !memcmp(CH1, T, 8)) FoundMask |= 1 << 1; + else if (NewSize == 8 && !memcmp(CH2, T, 8)) FoundMask |= 1 << 2; + else if (NewSize == 8 && !memcmp(CH3, T, 8)) FoundMask |= 1 << 3; + else if (NewSize == 8 && !memcmp(CH4, T, 8)) FoundMask |= 1 << 4; + else if (NewSize == 8 && !memcmp(CH5, T, 8)) FoundMask |= 1 << 5; + else if (NewSize == 8 && !memcmp(CH6, T, 8)) FoundMask |= 1 << 6; + else if (NewSize == 8 && !memcmp(CH7, T, 8)) FoundMask |= 1 << 7; + } + EXPECT_EQ(FoundMask, 255); +} + +TEST(FuzzerMutate, ChangeBinaryInteger1) { + TestChangeBinaryInteger(&MutationDispatcher::Mutate_ChangeBinaryInteger, + 1 << 12); +} + +TEST(FuzzerMutate, ChangeBinaryInteger2) { + TestChangeBinaryInteger(&MutationDispatcher::Mutate, 1 << 15); +} + + +TEST(FuzzerDictionary, ParseOneDictionaryEntry) { + Unit U; + EXPECT_FALSE(ParseOneDictionaryEntry("", &U)); + EXPECT_FALSE(ParseOneDictionaryEntry(" ", &U)); + EXPECT_FALSE(ParseOneDictionaryEntry("\t ", &U)); + EXPECT_FALSE(ParseOneDictionaryEntry(" \" ", &U)); + EXPECT_FALSE(ParseOneDictionaryEntry(" zz\" ", &U)); + EXPECT_FALSE(ParseOneDictionaryEntry(" \"zz ", &U)); + EXPECT_FALSE(ParseOneDictionaryEntry(" \"\" ", &U)); + EXPECT_TRUE(ParseOneDictionaryEntry("\"a\"", &U)); + EXPECT_EQ(U, Unit({'a'})); + EXPECT_TRUE(ParseOneDictionaryEntry("\"abc\"", &U)); + EXPECT_EQ(U, Unit({'a', 'b', 'c'})); + EXPECT_TRUE(ParseOneDictionaryEntry("abc=\"abc\"", &U)); + EXPECT_EQ(U, Unit({'a', 'b', 'c'})); + EXPECT_FALSE(ParseOneDictionaryEntry("\"\\\"", &U)); + EXPECT_TRUE(ParseOneDictionaryEntry("\"\\\\\"", &U)); + EXPECT_EQ(U, Unit({'\\'})); + EXPECT_TRUE(ParseOneDictionaryEntry("\"\\xAB\"", &U)); + EXPECT_EQ(U, Unit({0xAB})); + EXPECT_TRUE(ParseOneDictionaryEntry("\"\\xABz\\xDE\"", &U)); + EXPECT_EQ(U, Unit({0xAB, 'z', 0xDE})); + EXPECT_TRUE(ParseOneDictionaryEntry("\"#\"", &U)); + EXPECT_EQ(U, Unit({'#'})); + EXPECT_TRUE(ParseOneDictionaryEntry("\"\\\"\"", &U)); + EXPECT_EQ(U, Unit({'"'})); +} + +TEST(FuzzerDictionary, ParseDictionaryFile) { + Vector<Unit> Units; + EXPECT_FALSE(ParseDictionaryFile("zzz\n", &Units)); + EXPECT_FALSE(ParseDictionaryFile("", &Units)); + EXPECT_TRUE(ParseDictionaryFile("\n", &Units)); + EXPECT_EQ(Units.size(), 0U); + EXPECT_TRUE(ParseDictionaryFile("#zzzz a b c d\n", &Units)); + EXPECT_EQ(Units.size(), 0U); + EXPECT_TRUE(ParseDictionaryFile(" #zzzz\n", &Units)); + EXPECT_EQ(Units.size(), 0U); + EXPECT_TRUE(ParseDictionaryFile(" #zzzz\n", &Units)); + EXPECT_EQ(Units.size(), 0U); + EXPECT_TRUE(ParseDictionaryFile(" #zzzz\naaa=\"aa\"", &Units)); + EXPECT_EQ(Units, Vector<Unit>({Unit({'a', 'a'})})); + EXPECT_TRUE( + ParseDictionaryFile(" #zzzz\naaa=\"aa\"\n\nabc=\"abc\"", &Units)); + EXPECT_EQ(Units, + Vector<Unit>({Unit({'a', 'a'}), Unit({'a', 'b', 'c'})})); +} + +TEST(FuzzerUtil, Base64) { + EXPECT_EQ("", Base64({})); + EXPECT_EQ("YQ==", Base64({'a'})); + EXPECT_EQ("eA==", Base64({'x'})); + EXPECT_EQ("YWI=", Base64({'a', 'b'})); + EXPECT_EQ("eHk=", Base64({'x', 'y'})); + EXPECT_EQ("YWJj", Base64({'a', 'b', 'c'})); + EXPECT_EQ("eHl6", Base64({'x', 'y', 'z'})); + EXPECT_EQ("YWJjeA==", Base64({'a', 'b', 'c', 'x'})); + EXPECT_EQ("YWJjeHk=", Base64({'a', 'b', 'c', 'x', 'y'})); + EXPECT_EQ("YWJjeHl6", Base64({'a', 'b', 'c', 'x', 'y', 'z'})); +} + +TEST(Corpus, Distribution) { + Random Rand(0); + std::unique_ptr<InputCorpus> C(new InputCorpus("")); + size_t N = 10; + size_t TriesPerUnit = 1<<16; + for (size_t i = 0; i < N; i++) + C->AddToCorpus(Unit{ static_cast<uint8_t>(i) }, 1, false, {}); + + Vector<size_t> Hist(N); + for (size_t i = 0; i < N * TriesPerUnit; i++) { + Hist[C->ChooseUnitIdxToMutate(Rand)]++; + } + for (size_t i = 0; i < N; i++) { + // A weak sanity check that every unit gets invoked. + EXPECT_GT(Hist[i], TriesPerUnit / N / 3); + } +} + +TEST(Merge, Bad) { + const char *kInvalidInputs[] = { + "", + "x", + "3\nx", + "2\n3", + "2\n2", + "2\n2\nA\n", + "2\n2\nA\nB\nC\n", + "0\n0\n", + "1\n1\nA\nDONE 0", + "1\n1\nA\nSTARTED 1", + }; + Merger M; + for (auto S : kInvalidInputs) { + // fprintf(stderr, "TESTING:\n%s\n", S); + EXPECT_FALSE(M.Parse(S, false)); + } +} + +void EQ(const Vector<uint32_t> &A, const Vector<uint32_t> &B) { + EXPECT_EQ(A, B); +} + +void EQ(const Vector<std::string> &A, const Vector<std::string> &B) { + Set<std::string> a(A.begin(), A.end()); + Set<std::string> b(B.begin(), B.end()); + EXPECT_EQ(a, b); +} + +static void Merge(const std::string &Input, + const Vector<std::string> Result, + size_t NumNewFeatures) { + Merger M; + Vector<std::string> NewFiles; + EXPECT_TRUE(M.Parse(Input, true)); + std::stringstream SS; + M.PrintSummary(SS); + EXPECT_EQ(NumNewFeatures, M.Merge(&NewFiles)); + EXPECT_EQ(M.AllFeatures(), M.ParseSummary(SS)); + EQ(NewFiles, Result); +} + +TEST(Merge, Good) { + Merger M; + + EXPECT_TRUE(M.Parse("1\n0\nAA\n", false)); + EXPECT_EQ(M.Files.size(), 1U); + EXPECT_EQ(M.NumFilesInFirstCorpus, 0U); + EXPECT_EQ(M.Files[0].Name, "AA"); + EXPECT_TRUE(M.LastFailure.empty()); + EXPECT_EQ(M.FirstNotProcessedFile, 0U); + + EXPECT_TRUE(M.Parse("2\n1\nAA\nBB\nSTARTED 0 42\n", false)); + EXPECT_EQ(M.Files.size(), 2U); + EXPECT_EQ(M.NumFilesInFirstCorpus, 1U); + EXPECT_EQ(M.Files[0].Name, "AA"); + EXPECT_EQ(M.Files[1].Name, "BB"); + EXPECT_EQ(M.LastFailure, "AA"); + EXPECT_EQ(M.FirstNotProcessedFile, 1U); + + EXPECT_TRUE(M.Parse("3\n1\nAA\nBB\nC\n" + "STARTED 0 1000\n" + "DONE 0 1 2 3\n" + "STARTED 1 1001\n" + "DONE 1 4 5 6 \n" + "STARTED 2 1002\n" + "", true)); + EXPECT_EQ(M.Files.size(), 3U); + EXPECT_EQ(M.NumFilesInFirstCorpus, 1U); + EXPECT_EQ(M.Files[0].Name, "AA"); + EXPECT_EQ(M.Files[0].Size, 1000U); + EXPECT_EQ(M.Files[1].Name, "BB"); + EXPECT_EQ(M.Files[1].Size, 1001U); + EXPECT_EQ(M.Files[2].Name, "C"); + EXPECT_EQ(M.Files[2].Size, 1002U); + EXPECT_EQ(M.LastFailure, "C"); + EXPECT_EQ(M.FirstNotProcessedFile, 3U); + EQ(M.Files[0].Features, {1, 2, 3}); + EQ(M.Files[1].Features, {4, 5, 6}); + + + Vector<std::string> NewFiles; + + EXPECT_TRUE(M.Parse("3\n2\nAA\nBB\nC\n" + "STARTED 0 1000\nDONE 0 1 2 3\n" + "STARTED 1 1001\nDONE 1 4 5 6 \n" + "STARTED 2 1002\nDONE 2 6 1 3 \n" + "", true)); + EXPECT_EQ(M.Files.size(), 3U); + EXPECT_EQ(M.NumFilesInFirstCorpus, 2U); + EXPECT_TRUE(M.LastFailure.empty()); + EXPECT_EQ(M.FirstNotProcessedFile, 3U); + EQ(M.Files[0].Features, {1, 2, 3}); + EQ(M.Files[1].Features, {4, 5, 6}); + EQ(M.Files[2].Features, {1, 3, 6}); + EXPECT_EQ(0U, M.Merge(&NewFiles)); + EQ(NewFiles, {}); + + EXPECT_TRUE(M.Parse("3\n1\nA\nB\nC\n" + "STARTED 0 1000\nDONE 0 1 2 3\n" + "STARTED 1 1001\nDONE 1 4 5 6 \n" + "STARTED 2 1002\nDONE 2 6 1 3\n" + "", true)); + EQ(M.Files[0].Features, {1, 2, 3}); + EQ(M.Files[1].Features, {4, 5, 6}); + EQ(M.Files[2].Features, {1, 3, 6}); + EXPECT_EQ(3U, M.Merge(&NewFiles)); + EQ(NewFiles, {"B"}); + + // Same as the above, but with InitialFeatures. + EXPECT_TRUE(M.Parse("2\n0\nB\nC\n" + "STARTED 0 1001\nDONE 0 4 5 6 \n" + "STARTED 1 1002\nDONE 1 6 1 3\n" + "", true)); + EQ(M.Files[0].Features, {4, 5, 6}); + EQ(M.Files[1].Features, {1, 3, 6}); + Set<uint32_t> InitialFeatures; + InitialFeatures.insert(1); + InitialFeatures.insert(2); + InitialFeatures.insert(3); + EXPECT_EQ(3U, M.Merge(InitialFeatures, &NewFiles)); + EQ(NewFiles, {"B"}); +} + +TEST(Merge, Merge) { + + Merge("3\n1\nA\nB\nC\n" + "STARTED 0 1000\nDONE 0 1 2 3\n" + "STARTED 1 1001\nDONE 1 4 5 6 \n" + "STARTED 2 1002\nDONE 2 6 1 3 \n", + {"B"}, 3); + + Merge("3\n0\nA\nB\nC\n" + "STARTED 0 2000\nDONE 0 1 2 3\n" + "STARTED 1 1001\nDONE 1 4 5 6 \n" + "STARTED 2 1002\nDONE 2 6 1 3 \n", + {"A", "B", "C"}, 6); + + Merge("4\n0\nA\nB\nC\nD\n" + "STARTED 0 2000\nDONE 0 1 2 3\n" + "STARTED 1 1101\nDONE 1 4 5 6 \n" + "STARTED 2 1102\nDONE 2 6 1 3 100 \n" + "STARTED 3 1000\nDONE 3 1 \n", + {"A", "B", "C", "D"}, 7); + + Merge("4\n1\nA\nB\nC\nD\n" + "STARTED 0 2000\nDONE 0 4 5 6 7 8\n" + "STARTED 1 1100\nDONE 1 1 2 3 \n" + "STARTED 2 1100\nDONE 2 2 3 \n" + "STARTED 3 1000\nDONE 3 1 \n", + {"B", "D"}, 3); +} + +TEST(Fuzzer, ForEachNonZeroByte) { + const size_t N = 64; + alignas(64) uint8_t Ar[N + 8] = { + 0, 0, 0, 0, 0, 0, 0, 0, + 1, 2, 0, 0, 0, 0, 0, 0, + 0, 0, 3, 0, 4, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 5, 0, 6, 0, 0, + 0, 0, 0, 0, 0, 0, 7, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 8, + 9, 9, 9, 9, 9, 9, 9, 9, + }; + typedef Vector<std::pair<size_t, uint8_t> > Vec; + Vec Res, Expected; + auto CB = [&](size_t FirstFeature, size_t Idx, uint8_t V) { + Res.push_back({FirstFeature + Idx, V}); + }; + ForEachNonZeroByte(Ar, Ar + N, 100, CB); + Expected = {{108, 1}, {109, 2}, {118, 3}, {120, 4}, + {135, 5}, {137, 6}, {146, 7}, {163, 8}}; + EXPECT_EQ(Res, Expected); + + Res.clear(); + ForEachNonZeroByte(Ar + 9, Ar + N, 109, CB); + Expected = { {109, 2}, {118, 3}, {120, 4}, + {135, 5}, {137, 6}, {146, 7}, {163, 8}}; + EXPECT_EQ(Res, Expected); + + Res.clear(); + ForEachNonZeroByte(Ar + 9, Ar + N - 9, 109, CB); + Expected = { {109, 2}, {118, 3}, {120, 4}, + {135, 5}, {137, 6}, {146, 7}}; + EXPECT_EQ(Res, Expected); +} + +// FuzzerCommand unit tests. The arguments in the two helper methods below must +// match. +static void makeCommandArgs(Vector<std::string> *ArgsToAdd) { + assert(ArgsToAdd); + ArgsToAdd->clear(); + ArgsToAdd->push_back("foo"); + ArgsToAdd->push_back("-bar=baz"); + ArgsToAdd->push_back("qux"); + ArgsToAdd->push_back(Command::ignoreRemainingArgs()); + ArgsToAdd->push_back("quux"); + ArgsToAdd->push_back("-grault=garply"); +} + +static std::string makeCmdLine(const char *separator, const char *suffix) { + std::string CmdLine("foo -bar=baz qux "); + if (strlen(separator) != 0) { + CmdLine += separator; + CmdLine += " "; + } + CmdLine += Command::ignoreRemainingArgs(); + CmdLine += " quux -grault=garply"; + if (strlen(suffix) != 0) { + CmdLine += " "; + CmdLine += suffix; + } + return CmdLine; +} + +TEST(FuzzerCommand, Create) { + std::string CmdLine; + + // Default constructor + Command DefaultCmd; + + CmdLine = DefaultCmd.toString(); + EXPECT_EQ(CmdLine, ""); + + // Explicit constructor + Vector<std::string> ArgsToAdd; + makeCommandArgs(&ArgsToAdd); + Command InitializedCmd(ArgsToAdd); + + CmdLine = InitializedCmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "")); + + // Compare each argument + auto InitializedArgs = InitializedCmd.getArguments(); + auto i = ArgsToAdd.begin(); + auto j = InitializedArgs.begin(); + while (i != ArgsToAdd.end() && j != InitializedArgs.end()) { + EXPECT_EQ(*i++, *j++); + } + EXPECT_EQ(i, ArgsToAdd.end()); + EXPECT_EQ(j, InitializedArgs.end()); + + // Copy constructor + Command CopiedCmd(InitializedCmd); + + CmdLine = CopiedCmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "")); + + // Assignment operator + Command AssignedCmd; + AssignedCmd = CopiedCmd; + + CmdLine = AssignedCmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "")); +} + +TEST(FuzzerCommand, ModifyArguments) { + Vector<std::string> ArgsToAdd; + makeCommandArgs(&ArgsToAdd); + Command Cmd; + std::string CmdLine; + + Cmd.addArguments(ArgsToAdd); + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "")); + + Cmd.addArgument("waldo"); + EXPECT_TRUE(Cmd.hasArgument("waldo")); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("waldo", "")); + + Cmd.removeArgument("waldo"); + EXPECT_FALSE(Cmd.hasArgument("waldo")); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "")); +} + +TEST(FuzzerCommand, ModifyFlags) { + Vector<std::string> ArgsToAdd; + makeCommandArgs(&ArgsToAdd); + Command Cmd(ArgsToAdd); + std::string Value, CmdLine; + ASSERT_FALSE(Cmd.hasFlag("fred")); + + Value = Cmd.getFlagValue("fred"); + EXPECT_EQ(Value, ""); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "")); + + Cmd.addFlag("fred", "plugh"); + EXPECT_TRUE(Cmd.hasFlag("fred")); + + Value = Cmd.getFlagValue("fred"); + EXPECT_EQ(Value, "plugh"); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("-fred=plugh", "")); + + Cmd.removeFlag("fred"); + EXPECT_FALSE(Cmd.hasFlag("fred")); + + Value = Cmd.getFlagValue("fred"); + EXPECT_EQ(Value, ""); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "")); +} + +TEST(FuzzerCommand, SetOutput) { + Vector<std::string> ArgsToAdd; + makeCommandArgs(&ArgsToAdd); + Command Cmd(ArgsToAdd); + std::string CmdLine; + ASSERT_FALSE(Cmd.hasOutputFile()); + ASSERT_FALSE(Cmd.isOutAndErrCombined()); + + Cmd.combineOutAndErr(true); + EXPECT_TRUE(Cmd.isOutAndErrCombined()); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", "2>&1")); + + Cmd.combineOutAndErr(false); + EXPECT_FALSE(Cmd.isOutAndErrCombined()); + + Cmd.setOutputFile("xyzzy"); + EXPECT_TRUE(Cmd.hasOutputFile()); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", ">xyzzy")); + + Cmd.setOutputFile("thud"); + EXPECT_TRUE(Cmd.hasOutputFile()); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", ">thud")); + + Cmd.combineOutAndErr(); + EXPECT_TRUE(Cmd.isOutAndErrCombined()); + + CmdLine = Cmd.toString(); + EXPECT_EQ(CmdLine, makeCmdLine("", ">thud 2>&1")); +} + +int main(int argc, char **argv) { + testing::InitGoogleTest(&argc, argv); + return RUN_ALL_TESTS(); +} |