// -*- C++ -*- //===------------------------- fuzzing.cpp -------------------------------===// // // The LLVM Compiler Infrastructure // // This file is dual licensed under the MIT and the University of Illinois Open // Source Licenses. See LICENSE.TXT for details. // //===----------------------------------------------------------------------===// // A set of routines to use when fuzzing the algorithms in libc++ // Each one tests a single algorithm. // // They all have the form of: // int `algorithm`(const uint8_t *data, size_t size); // // They perform the operation, and then check to see if the results are correct. // If so, they return zero, and non-zero otherwise. // // For example, sort calls std::sort, then checks two things: // (1) The resulting vector is sorted // (2) The resulting vector contains the same elements as the original data. #include "fuzzing.h" #include <vector> #include <algorithm> #include <functional> #include <regex> #include <cassert> #include <iostream> // If we had C++14, we could use the four iterator version of is_permutation and equal namespace fuzzing { // This is a struct we can use to test the stable_XXX algorithms. // perform the operation on the key, then check the order of the payload. struct stable_test { uint8_t key; size_t payload; stable_test(uint8_t k) : key(k), payload(0) {} stable_test(uint8_t k, size_t p) : key(k), payload(p) {} }; void swap(stable_test &lhs, stable_test &rhs) { using std::swap; swap(lhs.key, rhs.key); swap(lhs.payload, rhs.payload); } struct key_less { bool operator () (const stable_test &lhs, const stable_test &rhs) const { return lhs.key < rhs.key; } }; struct payload_less { bool operator () (const stable_test &lhs, const stable_test &rhs) const { return lhs.payload < rhs.payload; } }; struct total_less { bool operator () (const stable_test &lhs, const stable_test &rhs) const { return lhs.key == rhs.key ? lhs.payload < rhs.payload : lhs.key < rhs.key; } }; bool operator==(const stable_test &lhs, const stable_test &rhs) { return lhs.key == rhs.key && lhs.payload == rhs.payload; } template<typename T> struct is_even { bool operator () (const T &t) const { return t % 2 == 0; } }; template<> struct is_even<stable_test> { bool operator () (const stable_test &t) const { return t.key % 2 == 0; } }; typedef std::vector<uint8_t> Vec; typedef std::vector<stable_test> StableVec; typedef StableVec::const_iterator SVIter; // Cheap version of is_permutation // Builds a set of buckets for each of the key values. // Sums all the payloads. // Not 100% perfect, but _way_ faster bool is_permutation(SVIter first1, SVIter last1, SVIter first2) { size_t xBuckets[256] = {0}; size_t xPayloads[256] = {0}; size_t yBuckets[256] = {0}; size_t yPayloads[256] = {0}; for (; first1 != last1; ++first1, ++first2) { xBuckets [first1->key]++; xPayloads[first1->key] += first1->payload; yBuckets [first2->key]++; yPayloads[first2->key] += first2->payload; } for (size_t i = 0; i < 256; ++i) { if (xBuckets[i] != yBuckets[i]) return false; if (xPayloads[i] != yPayloads[i]) return false; } return true; } template <typename Iter1, typename Iter2> bool is_permutation(Iter1 first1, Iter1 last1, Iter2 first2) { static_assert((std::is_same<typename std::iterator_traits<Iter1>::value_type, uint8_t>::value), ""); static_assert((std::is_same<typename std::iterator_traits<Iter2>::value_type, uint8_t>::value), ""); size_t xBuckets[256] = {0}; size_t yBuckets[256] = {0}; for (; first1 != last1; ++first1, ++first2) { xBuckets [*first1]++; yBuckets [*first2]++; } for (size_t i = 0; i < 256; ++i) if (xBuckets[i] != yBuckets[i]) return false; return true; } // == sort == int sort(const uint8_t *data, size_t size) { Vec working(data, data + size); std::sort(working.begin(), working.end()); if (!std::is_sorted(working.begin(), working.end())) return 1; if (!fuzzing::is_permutation(data, data + size, working.cbegin())) return 99; return 0; } // == stable_sort == int stable_sort(const uint8_t *data, size_t size) { StableVec input; for (size_t i = 0; i < size; ++i) input.push_back(stable_test(data[i], i)); StableVec working = input; std::stable_sort(working.begin(), working.end(), key_less()); if (!std::is_sorted(working.begin(), working.end(), key_less())) return 1; auto iter = working.begin(); while (iter != working.end()) { auto range = std::equal_range(iter, working.end(), *iter, key_less()); if (!std::is_sorted(range.first, range.second, total_less())) return 2; iter = range.second; } if (!fuzzing::is_permutation(input.cbegin(), input.cend(), working.cbegin())) return 99; return 0; } // == partition == int partition(const uint8_t *data, size_t size) { Vec working(data, data + size); auto iter = std::partition(working.begin(), working.end(), is_even<uint8_t>()); if (!std::all_of (working.begin(), iter, is_even<uint8_t>())) return 1; if (!std::none_of(iter, working.end(), is_even<uint8_t>())) return 2; if (!fuzzing::is_permutation(data, data + size, working.cbegin())) return 99; return 0; } // == partition_copy == int partition_copy(const uint8_t *data, size_t size) { Vec v1, v2; auto iter = std::partition_copy(data, data + size, std::back_inserter<Vec>(v1), std::back_inserter<Vec>(v2), is_even<uint8_t>()); // The two vectors should add up to the original size if (v1.size() + v2.size() != size) return 1; // All of the even values should be in the first vector, and none in the second if (!std::all_of (v1.begin(), v1.end(), is_even<uint8_t>())) return 2; if (!std::none_of(v2.begin(), v2.end(), is_even<uint8_t>())) return 3; // Every value in both vectors has to be in the original // Make a copy of the input, and sort it Vec v0{data, data + size}; std::sort(v0.begin(), v0.end()); // Sort each vector and ensure that all of the elements appear in the original input std::sort(v1.begin(), v1.end()); if (!std::includes(v0.begin(), v0.end(), v1.begin(), v1.end())) return 4; std::sort(v2.begin(), v2.end()); if (!std::includes(v0.begin(), v0.end(), v2.begin(), v2.end())) return 5; // This, while simple, is really slow - 20 seconds on a 500K element input. // for (auto v: v1) // if (std::find(data, data + size, v) == data + size) return 4; // // for (auto v: v2) // if (std::find(data, data + size, v) == data + size) return 5; return 0; } // == stable_partition == int stable_partition (const uint8_t *data, size_t size) { StableVec input; for (size_t i = 0; i < size; ++i) input.push_back(stable_test(data[i], i)); StableVec working = input; auto iter = std::stable_partition(working.begin(), working.end(), is_even<stable_test>()); if (!std::all_of (working.begin(), iter, is_even<stable_test>())) return 1; if (!std::none_of(iter, working.end(), is_even<stable_test>())) return 2; if (!std::is_sorted(working.begin(), iter, payload_less())) return 3; if (!std::is_sorted(iter, working.end(), payload_less())) return 4; if (!fuzzing::is_permutation(input.cbegin(), input.cend(), working.cbegin())) return 99; return 0; } // == nth_element == // use the first element as a position into the data int nth_element (const uint8_t *data, size_t size) { if (size <= 1) return 0; const size_t partition_point = data[0] % size; Vec working(data + 1, data + size); const auto partition_iter = working.begin() + partition_point; std::nth_element(working.begin(), partition_iter, working.end()); // nth may be the end iterator, in this case nth_element has no effect. if (partition_iter == working.end()) { if (!std::equal(data + 1, data + size, working.begin())) return 98; } else { const uint8_t nth = *partition_iter; if (!std::all_of(working.begin(), partition_iter, [=](uint8_t v) { return v <= nth; })) return 1; if (!std::all_of(partition_iter, working.end(), [=](uint8_t v) { return v >= nth; })) return 2; if (!fuzzing::is_permutation(data + 1, data + size, working.cbegin())) return 99; } return 0; } // == partial_sort == // use the first element as a position into the data int partial_sort (const uint8_t *data, size_t size) { if (size <= 1) return 0; const size_t sort_point = data[0] % size; Vec working(data + 1, data + size); const auto sort_iter = working.begin() + sort_point; std::partial_sort(working.begin(), sort_iter, working.end()); if (sort_iter != working.end()) { const uint8_t nth = *std::min_element(sort_iter, working.end()); if (!std::all_of(working.begin(), sort_iter, [=](uint8_t v) { return v <= nth; })) return 1; if (!std::all_of(sort_iter, working.end(), [=](uint8_t v) { return v >= nth; })) return 2; } if (!std::is_sorted(working.begin(), sort_iter)) return 3; if (!fuzzing::is_permutation(data + 1, data + size, working.cbegin())) return 99; return 0; } // == partial_sort_copy == // use the first element as a count int partial_sort_copy (const uint8_t *data, size_t size) { if (size <= 1) return 0; const size_t num_results = data[0] % size; Vec results(num_results); (void) std::partial_sort_copy(data + 1, data + size, results.begin(), results.end()); // The results have to be sorted if (!std::is_sorted(results.begin(), results.end())) return 1; // All the values in results have to be in the original data for (auto v: results) if (std::find(data + 1, data + size, v) == data + size) return 2; // The things in results have to be the smallest N in the original data Vec sorted(data + 1, data + size); std::sort(sorted.begin(), sorted.end()); if (!std::equal(results.begin(), results.end(), sorted.begin())) return 3; return 0; } // The second sequence has been "uniqued" template <typename Iter1, typename Iter2> static bool compare_unique(Iter1 first1, Iter1 last1, Iter2 first2, Iter2 last2) { assert(first1 != last1 && first2 != last2); if (*first1 != *first2) return false; uint8_t last_value = *first1; ++first1; ++first2; while(first1 != last1 && first2 != last2) { // Skip over dups in the first sequence while (*first1 == last_value) if (++first1 == last1) return false; if (*first1 != *first2) return false; last_value = *first1; ++first1; ++first2; } // Still stuff left in the 'uniqued' sequence - oops if (first1 == last1 && first2 != last2) return false; // Still stuff left in the original sequence - better be all the same while (first1 != last1) { if (*first1 != last_value) return false; ++first1; } return true; } // == unique == int unique (const uint8_t *data, size_t size) { Vec working(data, data + size); std::sort(working.begin(), working.end()); Vec results = working; Vec::iterator new_end = std::unique(results.begin(), results.end()); Vec::iterator it; // scratch iterator // Check the size of the unique'd sequence. // it should only be zero if the input sequence was empty. if (results.begin() == new_end) return working.size() == 0 ? 0 : 1; // 'results' is sorted if (!std::is_sorted(results.begin(), new_end)) return 2; // All the elements in 'results' must be different it = results.begin(); uint8_t prev_value = *it++; for (; it != new_end; ++it) { if (*it == prev_value) return 3; prev_value = *it; } // Every element in 'results' must be in 'working' for (it = results.begin(); it != new_end; ++it) if (std::find(working.begin(), working.end(), *it) == working.end()) return 4; // Every element in 'working' must be in 'results' for (auto v : working) if (std::find(results.begin(), new_end, v) == new_end) return 5; return 0; } // == unique_copy == int unique_copy (const uint8_t *data, size_t size) { Vec working(data, data + size); std::sort(working.begin(), working.end()); Vec results; (void) std::unique_copy(working.begin(), working.end(), std::back_inserter<Vec>(results)); Vec::iterator it; // scratch iterator // Check the size of the unique'd sequence. // it should only be zero if the input sequence was empty. if (results.size() == 0) return working.size() == 0 ? 0 : 1; // 'results' is sorted if (!std::is_sorted(results.begin(), results.end())) return 2; // All the elements in 'results' must be different it = results.begin(); uint8_t prev_value = *it++; for (; it != results.end(); ++it) { if (*it == prev_value) return 3; prev_value = *it; } // Every element in 'results' must be in 'working' for (auto v : results) if (std::find(working.begin(), working.end(), v) == working.end()) return 4; // Every element in 'working' must be in 'results' for (auto v : working) if (std::find(results.begin(), results.end(), v) == results.end()) return 5; return 0; } // -- regex fuzzers static int regex_helper(const uint8_t *data, size_t size, std::regex::flag_type flag) { if (size > 0) { try { std::string s((const char *)data, size); std::regex re(s, flag); return std::regex_match(s, re) ? 1 : 0; } catch (std::regex_error &ex) {} } return 0; } int regex_ECMAScript (const uint8_t *data, size_t size) { (void) regex_helper(data, size, std::regex_constants::ECMAScript); return 0; } int regex_POSIX (const uint8_t *data, size_t size) { (void) regex_helper(data, size, std::regex_constants::basic); return 0; } int regex_extended (const uint8_t *data, size_t size) { (void) regex_helper(data, size, std::regex_constants::extended); return 0; } int regex_awk (const uint8_t *data, size_t size) { (void) regex_helper(data, size, std::regex_constants::awk); return 0; } int regex_grep (const uint8_t *data, size_t size) { (void) regex_helper(data, size, std::regex_constants::grep); return 0; } int regex_egrep (const uint8_t *data, size_t size) { (void) regex_helper(data, size, std::regex_constants::egrep); return 0; } // -- heap fuzzers int make_heap (const uint8_t *data, size_t size) { Vec working(data, data + size); std::make_heap(working.begin(), working.end()); if (!std::is_heap(working.begin(), working.end())) return 1; if (!fuzzing::is_permutation(data, data + size, working.cbegin())) return 99; return 0; } int push_heap (const uint8_t *data, size_t size) { if (size < 2) return 0; // Make a heap from the first half of the data Vec working(data, data + size); auto iter = working.begin() + (size / 2); std::make_heap(working.begin(), iter); if (!std::is_heap(working.begin(), iter)) return 1; // Now push the rest onto the heap, one at a time ++iter; for (; iter != working.end(); ++iter) { std::push_heap(working.begin(), iter); if (!std::is_heap(working.begin(), iter)) return 2; } if (!fuzzing::is_permutation(data, data + size, working.cbegin())) return 99; return 0; } int pop_heap (const uint8_t *data, size_t size) { if (size < 2) return 0; Vec working(data, data + size); std::make_heap(working.begin(), working.end()); // Pop things off, one at a time auto iter = --working.end(); while (iter != working.begin()) { std::pop_heap(working.begin(), iter); if (!std::is_heap(working.begin(), --iter)) return 2; } return 0; } // -- search fuzzers int search (const uint8_t *data, size_t size) { if (size < 2) return 0; const size_t pat_size = data[0] * (size - 1) / std::numeric_limits<uint8_t>::max(); assert(pat_size <= size - 1); const uint8_t *pat_begin = data + 1; const uint8_t *pat_end = pat_begin + pat_size; const uint8_t *data_end = data + size; assert(pat_end <= data_end); // std::cerr << "data[0] = " << size_t(data[0]) << " "; // std::cerr << "Pattern size = " << pat_size << "; corpus is " << size - 1 << std::endl; auto it = std::search(pat_end, data_end, pat_begin, pat_end); if (it != data_end) // not found if (!std::equal(pat_begin, pat_end, it)) return 1; return 0; } template <typename S> static int search_helper (const uint8_t *data, size_t size) { if (size < 2) return 0; const size_t pat_size = data[0] * (size - 1) / std::numeric_limits<uint8_t>::max(); const uint8_t *pat_begin = data + 1; const uint8_t *pat_end = pat_begin + pat_size; const uint8_t *data_end = data + size; auto it = std::search(pat_end, data_end, S(pat_begin, pat_end)); if (it != data_end) // not found if (!std::equal(pat_begin, pat_end, it)) return 1; return 0; } // These are still in std::experimental // int search_boyer_moore (const uint8_t *data, size_t size) // { // return search_helper<std::boyer_moore_searcher<const uint8_t *>>(data, size); // } // // int search_boyer_moore_horspool (const uint8_t *data, size_t size) // { // return search_helper<std::boyer_moore_horspool_searcher<const uint8_t *>>(data, size); // } // -- set operation fuzzers template <typename S> static void set_helper (const uint8_t *data, size_t size, Vec &v1, Vec &v2) { assert(size > 1); const size_t pat_size = data[0] * (size - 1) / std::numeric_limits<uint8_t>::max(); const uint8_t *pat_begin = data + 1; const uint8_t *pat_end = pat_begin + pat_size; const uint8_t *data_end = data + size; v1.assign(pat_begin, pat_end); v2.assign(pat_end, data_end); std::sort(v1.begin(), v1.end()); std::sort(v2.begin(), v2.end()); } } // namespace fuzzing