mirror of
https://github.com/AFLplusplus/AFLplusplus.git
synced 2025-06-17 20:28:08 +00:00
lots of fixes
This commit is contained in:
@ -38,8 +38,10 @@ typedef struct my_mutator {
|
||||
|
||||
} my_mutator_t;
|
||||
|
||||
#define DEBUG \
|
||||
#undef DEBUGF
|
||||
#define DEBUGF \
|
||||
if (unlikely(debug)) fprintf
|
||||
#define IFDEBUG if (unlikely(debug))
|
||||
|
||||
static afl_state *afl_ptr;
|
||||
static int debug = AUTOTOKENS_DEBUG;
|
||||
@ -57,12 +59,12 @@ static unordered_map<u32, vector<u32> *> id_mapping;
|
||||
static unordered_map<string, u32> token_to_id;
|
||||
static unordered_map<u32, string> id_to_token;
|
||||
static string whitespace = AUTOTOKENS_WHITESPACE;
|
||||
static string output;
|
||||
static regex *regex_comment_custom;
|
||||
static regex regex_comment_star("/\\*([:print:]|\n)*?\\*/",
|
||||
regex::multiline | regex::optimize);
|
||||
static regex regex_word("[A-Za-z0-9_$.-]+", regex::optimize);
|
||||
static regex regex_whitespace(R"([ \t]+)", regex::optimize);
|
||||
static regex regex_string("\"[[:print:]]*?\"|'[[:print:]]*?'", regex::optimize);
|
||||
static vector<u32> *s; // the structure of the currently selected input
|
||||
|
||||
u32 good_whitespace_or_singleval() {
|
||||
@ -104,7 +106,7 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
MAX(AUTOTOKENS_CHANGE_MIN,
|
||||
MIN(m_size >> 3, HAVOC_CYCLES * afl_ptr->queue_cur->perf_score *
|
||||
afl_ptr->havoc_div / 256));
|
||||
// DEBUG(stderr, "structure size: %lu, rounds: %u \n", m.size(), rounds);
|
||||
// DEBUGF(stderr, "structure size: %lu, rounds: %u \n", m.size(), rounds);
|
||||
|
||||
#if AUTOTOKENS_SPLICE_DISABLE == 1
|
||||
#define AUTOTOKENS_MUT_MAX 12
|
||||
@ -112,7 +114,7 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
#define AUTOTOKENS_MUT_MAX 14
|
||||
#endif
|
||||
|
||||
u32 max_rand = AUTOTOKENS_MUT_MAX;
|
||||
u32 max_rand = AUTOTOKENS_MUT_MAX, new_item, pos;
|
||||
|
||||
for (i = 0; i < rounds; ++i) {
|
||||
|
||||
@ -122,8 +124,8 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
case 0 ... 7: /* fall through */
|
||||
{
|
||||
|
||||
u32 pos = rand_below(afl_ptr, m_size);
|
||||
u32 cur_item = m[pos], new_item;
|
||||
pos = rand_below(afl_ptr, m_size);
|
||||
u32 cur_item = m[pos];
|
||||
do {
|
||||
|
||||
new_item = rand_below(afl_ptr, current_id);
|
||||
@ -135,7 +137,7 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
((whitespace_ids < new_item && whitespace_ids >= cur_item) ||
|
||||
(whitespace_ids >= new_item && whitespace_ids < cur_item)))));
|
||||
|
||||
DEBUG(stderr, "MUT: %u -> %u\n", cur_item, new_item);
|
||||
DEBUGF(stderr, "MUT: %u -> %u\n", cur_item, new_item);
|
||||
m[pos] = new_item;
|
||||
break;
|
||||
|
||||
@ -144,7 +146,6 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
/* INSERT (m_size +1 so we insert also after last place) */
|
||||
case 8 ... 9: {
|
||||
|
||||
u32 new_item;
|
||||
do {
|
||||
|
||||
new_item = rand_below(afl_ptr, current_id);
|
||||
@ -154,7 +155,7 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
u32 pos = rand_below(afl_ptr, m_size + 1);
|
||||
m.insert(m.begin() + pos, new_item);
|
||||
++m_size;
|
||||
DEBUG(stderr, "INS: %u at %u\n", new_item, pos);
|
||||
DEBUGF(stderr, "INS: %u at %u\n", new_item, pos);
|
||||
|
||||
if (likely(!alternative_tokenize)) {
|
||||
|
||||
@ -212,7 +213,8 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
m.insert(m.begin() + dst_off, src->begin() + src_off,
|
||||
src->begin() + src_off + n);
|
||||
m_size += n;
|
||||
DEBUG(stderr, "SPLICE-INS: %u at %u\n", n, dst_off);
|
||||
DEBUGF(stderr, "SPLICE-INS: %u at %u\n", n, dst_off);
|
||||
|
||||
break;
|
||||
|
||||
}
|
||||
@ -231,13 +233,36 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
copy(src->begin() + src_off, src->begin() + src_off + n,
|
||||
m.begin() + dst_off);
|
||||
|
||||
DEBUG(stderr, "SPLICE-MUT: %u at %u\n", n, dst_off);
|
||||
DEBUGF(stderr, "SPLICE-MUT: %u at %u\n", n, dst_off);
|
||||
break;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
if (likely(!alternative_tokenize)) {
|
||||
|
||||
// do we need a whitespace/token at the beginning?
|
||||
if (dst_off && id_to_token[m[dst_off - 1]].size() > 1 &&
|
||||
id_to_token[m[dst_off]].size() > 1) {
|
||||
|
||||
m.insert(m.begin() + dst_off, good_whitespace_or_singleval());
|
||||
++m_size;
|
||||
|
||||
}
|
||||
|
||||
// do we need a whitespace/token at the end?
|
||||
if (dst_off + n < m_size &&
|
||||
id_to_token[m[dst_off + n - 1]].size() > 1 &&
|
||||
id_to_token[m[dst_off + n]].size() > 1) {
|
||||
|
||||
m.insert(m.begin() + dst_off + n, good_whitespace_or_singleval());
|
||||
++m_size;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
break;
|
||||
|
||||
}
|
||||
@ -249,11 +274,32 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
|
||||
if (m_size > 8) {
|
||||
|
||||
m.erase(m.begin() + rand_below(afl_ptr, m_size));
|
||||
--m_size;
|
||||
do {
|
||||
|
||||
pos = rand_below(afl_ptr, m_size);
|
||||
|
||||
} while (unlikely(pos < whitespace_ids));
|
||||
|
||||
// if what we delete will result in a missing whitespace/token,
|
||||
// instead of deleting we switch the item to a whitespace or token.
|
||||
if (likely(!alternative_tokenize) && pos && pos < m_size &&
|
||||
id_to_token[m[pos - 1]].size() > 1 &&
|
||||
id_to_token[m[pos + 1]].size() > 1) {
|
||||
|
||||
m[pos] = good_whitespace_or_singleval();
|
||||
|
||||
} else {
|
||||
|
||||
m.erase(m.begin() + pos);
|
||||
--m_size;
|
||||
|
||||
}
|
||||
|
||||
} else {
|
||||
|
||||
// if the data is already too small do not try to make it smaller
|
||||
// again this run.
|
||||
|
||||
max_rand = AUTOTOKENS_MUT_MAX - 2;
|
||||
|
||||
}
|
||||
@ -262,14 +308,12 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
|
||||
}
|
||||
|
||||
// TODO: add full line insert splice, replace splace, delete
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
string output;
|
||||
u32 m_size_1 = m_size - 1;
|
||||
output = "";
|
||||
|
||||
for (i = 0; i < m_size; ++i) {
|
||||
|
||||
@ -282,31 +326,108 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
|
||||
|
||||
}
|
||||
|
||||
u32 mutated_size = output.size();
|
||||
u8 *mutated_out = (u8 *)afl_realloc((void **)out_buf, mutated_size);
|
||||
u32 mutated_size = (u32)output.size();
|
||||
u8 *mutated_out = (u8 *)output.data();
|
||||
|
||||
if (unlikely(!mutated_out)) {
|
||||
if (unlikely(mutated_size > max_size)) { mutated_size = max_size; }
|
||||
|
||||
*out_buf = NULL;
|
||||
return 0;
|
||||
IFDEBUG {
|
||||
|
||||
}
|
||||
|
||||
if (unlikely(debug)) {
|
||||
|
||||
DEBUG(stderr, "MUTATED to %u bytes:\n", mutated_size);
|
||||
DEBUGF(stderr, "MUTATED to %u bytes:\n", mutated_size);
|
||||
fwrite(output.data(), 1, mutated_size, stderr);
|
||||
DEBUG(stderr, "\n---\n");
|
||||
DEBUGF(stderr, "\n---\n");
|
||||
|
||||
}
|
||||
|
||||
memcpy(mutated_out, output.data(), mutated_size);
|
||||
*out_buf = mutated_out;
|
||||
++fuzz_count;
|
||||
return mutated_size;
|
||||
|
||||
}
|
||||
|
||||
/* I get f*cking stack overflow using C++ regex with a regex of
|
||||
"\"[[:print:]]*?\"" if this matches a long string even with regex::optimize
|
||||
enabled :-( */
|
||||
u8 my_search_string(string::const_iterator cur, string::const_iterator ende,
|
||||
string::const_iterator *match_begin,
|
||||
string::const_iterator *match_end) {
|
||||
|
||||
string::const_iterator start = cur, found_begin;
|
||||
u8 quote_type = 0;
|
||||
|
||||
while (cur < ende) {
|
||||
|
||||
switch (*cur) {
|
||||
|
||||
case '"': {
|
||||
|
||||
if (cur == start || *(cur - 1) != '\\') {
|
||||
|
||||
if (!quote_type) {
|
||||
|
||||
found_begin = cur;
|
||||
quote_type = 1;
|
||||
|
||||
} else if (quote_type == 1) {
|
||||
|
||||
*match_begin = found_begin;
|
||||
*match_end = cur + 1;
|
||||
return 1;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
break;
|
||||
|
||||
}
|
||||
|
||||
case '\'': {
|
||||
|
||||
if (cur == start || *(cur - 1) != '\\') {
|
||||
|
||||
if (!quote_type) {
|
||||
|
||||
found_begin = cur;
|
||||
quote_type = 2;
|
||||
|
||||
} else if (quote_type == 2) {
|
||||
|
||||
*match_begin = found_begin;
|
||||
*match_end = cur + 1;
|
||||
return 1;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
break;
|
||||
|
||||
}
|
||||
|
||||
case '\n':
|
||||
case '\r':
|
||||
case 0: {
|
||||
|
||||
quote_type = 0;
|
||||
break;
|
||||
|
||||
}
|
||||
|
||||
default:
|
||||
if (unlikely(quote_type && !isprint(*cur))) { quote_type = 0; }
|
||||
break;
|
||||
|
||||
}
|
||||
|
||||
++cur;
|
||||
|
||||
}
|
||||
|
||||
return 0;
|
||||
|
||||
}
|
||||
|
||||
/* We are not using afl_custom_queue_new_entry() because not every corpus entry
|
||||
will be necessarily fuzzed. so we use afl_custom_queue_get() instead */
|
||||
|
||||
@ -321,7 +442,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
(only_fav && !afl_ptr->queue_cur->favored)) {
|
||||
|
||||
s = NULL;
|
||||
DEBUG(stderr, "cmplog not ascii or only_fav and not favorite\n");
|
||||
DEBUGF(stderr, "cmplog not ascii or only_fav and not favorite\n");
|
||||
return 0;
|
||||
|
||||
}
|
||||
@ -356,7 +477,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
}
|
||||
|
||||
++extras_cnt;
|
||||
DEBUG(stderr, "Added from dictionary: \"%s\"\n", ptr);
|
||||
DEBUGF(stderr, "Added from dictionary: \"%s\"\n", ptr);
|
||||
|
||||
}
|
||||
|
||||
@ -385,7 +506,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
}
|
||||
|
||||
++a_extras_cnt;
|
||||
DEBUG(stderr, "Added from auto dictionary: \"%s\"\n", ptr);
|
||||
DEBUGF(stderr, "Added from auto dictionary: \"%s\"\n", ptr);
|
||||
|
||||
}
|
||||
|
||||
@ -415,7 +536,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
fclose(fp);
|
||||
file_mapping[fn] = structure; // NULL ptr so we don't read the file again
|
||||
s = NULL;
|
||||
DEBUG(stderr, "Too short (%lu) %s\n", len, filename);
|
||||
DEBUGF(stderr, "Too short (%lu) %s\n", len, filename);
|
||||
return 0;
|
||||
|
||||
}
|
||||
@ -443,14 +564,14 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
file_mapping[fn] = NULL;
|
||||
s = NULL;
|
||||
DEBUG(stderr, "Not text (%lu) %s\n", len, filename);
|
||||
DEBUGF(stderr, "Not text (%lu) %s\n", len, filename);
|
||||
return 0;
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
// DEBUG(stderr, "Read %lu bytes for %s\nBefore comment trim:\n%s\n",
|
||||
// DEBUGF(stderr, "Read %lu bytes for %s\nBefore comment trim:\n%s\n",
|
||||
// input.size(), filename, input.c_str());
|
||||
|
||||
if (regex_comment_custom) {
|
||||
@ -463,14 +584,14 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
DEBUG(stderr, "After replace %lu bytes for %s\n%s\n", input.size(),
|
||||
DEBUGF(stderr, "After replace %lu bytes for %s\n%s\n", input.size(),
|
||||
filename, input.c_str());
|
||||
|
||||
u32 spaces = count(input.begin(), input.end(), ' ');
|
||||
u32 tabs = count(input.begin(), input.end(), '\t');
|
||||
u32 linefeeds = count(input.begin(), input.end(), '\n');
|
||||
bool ends_with_linefeed = input[input.length() - 1] == '\n';
|
||||
DEBUG(stderr, "spaces=%u tabs=%u linefeeds=%u ends=%u\n", spaces, tabs,
|
||||
DEBUGF(stderr, "spaces=%u tabs=%u linefeeds=%u ends=%u\n", spaces, tabs,
|
||||
linefeeds, ends_with_linefeed);
|
||||
all_spaces += spaces;
|
||||
all_tabs += tabs;
|
||||
@ -479,26 +600,29 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
// now extract all tokens
|
||||
vector<string> tokens;
|
||||
smatch match;
|
||||
string::const_iterator cur = input.begin(), ende = input.end(), found, prev;
|
||||
string::const_iterator cur = input.begin(), ende = input.end(), found, prev,
|
||||
match_begin, match_end;
|
||||
|
||||
DEBUG(stderr, "START!\n");
|
||||
DEBUGF(stderr, "START!\n");
|
||||
|
||||
if (likely(!alternative_tokenize)) {
|
||||
|
||||
while (regex_search(cur, ende, match, regex_string,
|
||||
regex_constants::match_any |
|
||||
regex_constants::match_not_null |
|
||||
regex_constants::match_continuous)) {
|
||||
while (my_search_string(cur, ende, &match_begin, &match_end)) {
|
||||
|
||||
prev = cur;
|
||||
found = match[0].first;
|
||||
cur = match[0].second;
|
||||
DEBUG(stderr,
|
||||
found = match_begin;
|
||||
cur = match_end;
|
||||
|
||||
IFDEBUG {
|
||||
|
||||
string foo(match_begin, match_end);
|
||||
DEBUGF(stderr,
|
||||
"string %s found at start %lu offset %lu continue at %lu\n",
|
||||
match[0].str().c_str(), prev - input.begin(), match.position(),
|
||||
foo.c_str(), prev - input.begin(), found - prev,
|
||||
cur - input.begin());
|
||||
|
||||
}
|
||||
|
||||
if (prev < found) { // there are items between search start and find
|
||||
while (prev < found) {
|
||||
|
||||
@ -512,7 +636,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
}
|
||||
|
||||
tokens.push_back(std::string(start, prev));
|
||||
DEBUG(stderr, "WHITESPACE %ld \"%s\"\n", prev - start,
|
||||
DEBUGF(stderr, "WHITESPACE %ld \"%s\"\n", prev - start,
|
||||
tokens[tokens.size() - 1].c_str());
|
||||
|
||||
} else if (isalnum(*prev) || *prev == '$' || *prev == '_') {
|
||||
@ -525,14 +649,14 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
tokens.push_back(std::string(start, prev));
|
||||
DEBUG(stderr, "IDENTIFIER %ld \"%s\"\n", prev - start,
|
||||
tokens.push_back(string(start, prev));
|
||||
DEBUGF(stderr, "IDENTIFIER %ld \"%s\"\n", prev - start,
|
||||
tokens[tokens.size() - 1].c_str());
|
||||
|
||||
} else {
|
||||
|
||||
tokens.push_back(std::string(prev, prev + 1));
|
||||
DEBUG(stderr, "OTHER \"%c\"\n", *prev);
|
||||
tokens.push_back(string(prev, prev + 1));
|
||||
DEBUGF(stderr, "OTHER \"%c\"\n", *prev);
|
||||
++prev;
|
||||
|
||||
}
|
||||
@ -541,11 +665,12 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
if (match[0].length() > 0) { tokens.push_back(match[0]); }
|
||||
tokens.push_back(string(match_begin, match_end));
|
||||
DEBUGF(stderr, "TOK: %s\n", tokens[tokens.size() - 1].c_str());
|
||||
|
||||
}
|
||||
|
||||
DEBUG(stderr, "AFTER all strings\n");
|
||||
DEBUGF(stderr, "AFTER all strings\n");
|
||||
|
||||
if (cur < ende) {
|
||||
|
||||
@ -561,7 +686,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
}
|
||||
|
||||
tokens.push_back(std::string(start, cur));
|
||||
DEBUG(stderr, "WHITESPACE %ld \"%s\"\n", cur - start,
|
||||
DEBUGF(stderr, "WHITESPACE %ld \"%s\"\n", cur - start,
|
||||
tokens[tokens.size() - 1].c_str());
|
||||
|
||||
} else if (isalnum(*cur) || *cur == '$' || *cur == '_') {
|
||||
@ -575,13 +700,13 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
}
|
||||
|
||||
tokens.push_back(std::string(start, cur));
|
||||
DEBUG(stderr, "IDENTIFIER %ld \"%s\"\n", cur - start,
|
||||
DEBUGF(stderr, "IDENTIFIER %ld \"%s\"\n", cur - start,
|
||||
tokens[tokens.size() - 1].c_str());
|
||||
|
||||
} else {
|
||||
|
||||
tokens.push_back(std::string(cur, cur + 1));
|
||||
DEBUG(stderr, "OTHER \"%c\"\n", *cur);
|
||||
DEBUGF(stderr, "OTHER \"%c\"\n", *cur);
|
||||
++cur;
|
||||
|
||||
}
|
||||
@ -593,19 +718,21 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
} else {
|
||||
|
||||
// alternative tokenize
|
||||
|
||||
while (regex_search(cur, ende, match, regex_string,
|
||||
regex_constants::match_any |
|
||||
regex_constants::match_not_null |
|
||||
regex_constants::match_continuous)) {
|
||||
while (my_search_string(cur, ende, &match_begin, &match_end)) {
|
||||
|
||||
prev = cur;
|
||||
found = match[0].first;
|
||||
cur = match[0].second;
|
||||
DEBUG(stderr,
|
||||
found = match_begin;
|
||||
cur = match_end;
|
||||
IFDEBUG {
|
||||
|
||||
string foo(match_begin, match_end);
|
||||
DEBUGF(stderr,
|
||||
"string %s found at start %lu offset %lu continue at %lu\n",
|
||||
match[0].str().c_str(), prev - input.begin(), match.position(),
|
||||
foo.c_str(), prev - input.begin(), found - prev,
|
||||
cur - input.begin());
|
||||
|
||||
}
|
||||
|
||||
if (prev < found) { // there are items between search start and find
|
||||
sregex_token_iterator it{prev, found, regex_whitespace, -1};
|
||||
vector<std::string> tokenized{it, {}};
|
||||
@ -619,9 +746,9 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
tokenized.end());
|
||||
tokens.reserve(tokens.size() + tokenized.size() * 2 + 1);
|
||||
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
DEBUG(stderr, "tokens: %lu input size: %lu\n", tokenized.size(),
|
||||
DEBUGF(stderr, "tokens1: %lu input size: %lu\n", tokenized.size(),
|
||||
input.size());
|
||||
for (auto x : tokenized) {
|
||||
|
||||
@ -636,10 +763,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
string::const_iterator c = token.begin(), e = token.end(), f, p;
|
||||
smatch m;
|
||||
|
||||
while (regex_search(c, e, m, regex_word,
|
||||
regex_constants::match_any |
|
||||
regex_constants::match_not_null |
|
||||
regex_constants::match_continuous)) {
|
||||
while (regex_search(c, e, m, regex_word)) {
|
||||
|
||||
p = c;
|
||||
f = m[0].first;
|
||||
@ -649,10 +773,10 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
// there are items between search start and find
|
||||
while (p < f) {
|
||||
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
string foo(p, p + 1);
|
||||
DEBUG(stderr, "before string: \"%s\"\n", foo.c_str());
|
||||
DEBUGF(stderr, "before string: \"%s\"\n", foo.c_str());
|
||||
|
||||
}
|
||||
|
||||
@ -661,18 +785,19 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
/*
|
||||
IFDEBUG {
|
||||
|
||||
string foo(p, f);
|
||||
DEBUG(stderr, "before string: \"%s\"\n",
|
||||
foo.c_str()); tokens.push_back(std::string(p, f));
|
||||
*/
|
||||
DEBUGF(stderr, "before string: \"%s\"\n", foo.c_str());
|
||||
tokens.push_back(std::string(p, f));
|
||||
|
||||
}
|
||||
|
||||
DEBUG(
|
||||
stderr,
|
||||
"SUBstring \"%s\" found at start %lu offset %lu continue at "
|
||||
"%lu\n",
|
||||
}
|
||||
|
||||
DEBUGF(stderr,
|
||||
"SUBstring \"%s\" found at start %lu offset %lu continue "
|
||||
"at %lu\n",
|
||||
m[0].str().c_str(), p - input.begin(), m.position(),
|
||||
c - token.begin());
|
||||
tokens.push_back(m[0].str());
|
||||
@ -683,10 +808,10 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
while (c < e) {
|
||||
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
string foo(c, c + 1);
|
||||
DEBUG(stderr, "after string: \"%s\"\n", foo.c_str());
|
||||
DEBUGF(stderr, "after string: \"%s\"\n", foo.c_str());
|
||||
|
||||
}
|
||||
|
||||
@ -695,17 +820,14 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
/*
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
string foo(c, e);
|
||||
DEBUG(stderr, "after string: \"%s\"\n",
|
||||
foo.c_str());
|
||||
DEBUGF(stderr, "after string: \"%s\"\n", foo.c_str());
|
||||
|
||||
}
|
||||
|
||||
tokens.push_back(std::string(c, e));
|
||||
*/
|
||||
|
||||
}
|
||||
|
||||
@ -713,7 +835,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
if (match[0].length() > 0) { tokens.push_back(match[0]); }
|
||||
tokens.push_back(string(match_begin, match_end));
|
||||
|
||||
}
|
||||
|
||||
@ -727,9 +849,9 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
tokenized.end());
|
||||
tokens.reserve(tokens.size() + tokenized.size() * 2 + 1);
|
||||
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
DEBUG(stderr, "tokens: %lu input size: %lu\n", tokenized.size(),
|
||||
DEBUGF(stderr, "tokens2: %lu input size: %lu\n", tokenized.size(),
|
||||
input.size());
|
||||
for (auto x : tokenized) {
|
||||
|
||||
@ -744,10 +866,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
string::const_iterator c = token.begin(), e = token.end(), f, p;
|
||||
smatch m;
|
||||
|
||||
while (regex_search(c, e, m, regex_word,
|
||||
regex_constants::match_any |
|
||||
regex_constants::match_not_null |
|
||||
regex_constants::match_continuous)) {
|
||||
while (regex_search(c, e, m, regex_word)) {
|
||||
|
||||
p = c;
|
||||
f = m[0].first;
|
||||
@ -757,10 +876,10 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
// there are items between search start and find
|
||||
while (p < f) {
|
||||
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
string foo(p, p + 1);
|
||||
DEBUG(stderr, "before string: \"%s\"\n", foo.c_str());
|
||||
DEBUGF(stderr, "before string: \"%s\"\n", foo.c_str());
|
||||
|
||||
}
|
||||
|
||||
@ -769,23 +888,20 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
/*
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
string foo(p, f);
|
||||
DEBUG(stderr, "before string: \"%s\"\n",
|
||||
foo.c_str());
|
||||
DEBUGF(stderr, "before string: \"%s\"\n", foo.c_str());
|
||||
|
||||
}
|
||||
|
||||
tokens.push_back(std::string(p, f));
|
||||
*/
|
||||
|
||||
}
|
||||
|
||||
DEBUG(stderr,
|
||||
"SUB2string \"%s\" found at start %lu offset %lu continue at "
|
||||
"%lu\n",
|
||||
DEBUGF(stderr,
|
||||
"SUB2string \"%s\" found at start %lu offset %lu continue "
|
||||
"at %lu\n",
|
||||
m[0].str().c_str(), p - input.begin(), m.position(),
|
||||
c - token.begin());
|
||||
tokens.push_back(m[0].str());
|
||||
@ -796,10 +912,10 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
while (c < e) {
|
||||
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
string foo(c, c + 1);
|
||||
DEBUG(stderr, "after string: \"%s\"\n", foo.c_str());
|
||||
DEBUGF(stderr, "after string: \"%s\"\n", foo.c_str());
|
||||
|
||||
}
|
||||
|
||||
@ -808,16 +924,14 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
/*
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
string foo(c, e);
|
||||
DEBUG(stderr, "after string: \"%s\"\n", foo.c_str());
|
||||
DEBUGF(stderr, "after string: \"%s\"\n", foo.c_str());
|
||||
|
||||
}
|
||||
|
||||
tokens.push_back(std::string(c, e));
|
||||
*/
|
||||
|
||||
}
|
||||
|
||||
@ -827,22 +941,22 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
}
|
||||
|
||||
if (unlikely(debug)) {
|
||||
IFDEBUG {
|
||||
|
||||
DEBUG(stderr, "DUMPING TOKENS:\n");
|
||||
DEBUGF(stderr, "DUMPING TOKENS:\n");
|
||||
u32 size_1 = tokens.size() - 1;
|
||||
for (u32 i = 0; i < tokens.size(); ++i) {
|
||||
|
||||
DEBUG(stderr, "%s", tokens[i].c_str());
|
||||
DEBUGF(stderr, "%s", tokens[i].c_str());
|
||||
if (unlikely(alternative_tokenize && i < size_1)) {
|
||||
|
||||
DEBUG(stderr, "%s", whitespace.c_str());
|
||||
DEBUGF(stderr, "%s", whitespace.c_str());
|
||||
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
DEBUG(stderr, "---------------------------\n");
|
||||
DEBUGF(stderr, "---------------------------\n");
|
||||
|
||||
}
|
||||
|
||||
@ -850,7 +964,7 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
|
||||
file_mapping[fn] = NULL;
|
||||
s = NULL;
|
||||
DEBUG(stderr, "too few tokens\n");
|
||||
DEBUGF(stderr, "too few tokens\n");
|
||||
return 0;
|
||||
|
||||
}
|
||||
@ -886,21 +1000,23 @@ extern "C" unsigned char afl_custom_queue_get(void *data,
|
||||
all_structure_items += structure->size();
|
||||
|
||||
// we are done!
|
||||
DEBUG(stderr, "DONE! We have %lu tokens in the structure\n",
|
||||
DEBUGF(stderr, "DONE! We have %lu tokens in the structure\n",
|
||||
structure->size());
|
||||
|
||||
} else {
|
||||
}
|
||||
|
||||
else {
|
||||
|
||||
if (entry->second == NULL) {
|
||||
|
||||
DEBUG(stderr, "Skipping %s\n", filename);
|
||||
DEBUGF(stderr, "Skipping %s\n", filename);
|
||||
s = NULL;
|
||||
return 0;
|
||||
|
||||
}
|
||||
|
||||
s = entry->second;
|
||||
DEBUG(stderr, "OK %s\n", filename);
|
||||
DEBUGF(stderr, "OK %s\n", filename);
|
||||
|
||||
}
|
||||
|
||||
|
@ -150,6 +150,7 @@ def deinit(): # optional for Python
|
||||
sense to use it. You would only skip this if `post_process` is used to fix
|
||||
checksums etc. so if you are using it, e.g., as a post processing library.
|
||||
Note that a length > 0 *must* be returned!
|
||||
The returned output buffer is under **your** memory management!
|
||||
|
||||
- `describe` (optional):
|
||||
|
||||
|
@ -844,15 +844,16 @@ struct custom_mutator {
|
||||
/**
|
||||
* Perform custom mutations on a given input
|
||||
*
|
||||
* (Optional for now. Required in the future)
|
||||
* (Optional)
|
||||
*
|
||||
* @param data pointer returned in afl_custom_init by this custom mutator
|
||||
* Getting an add_buf can be skipped by using afl_custom_splice_optout().
|
||||
*
|
||||
* @param[in] data Pointer returned in afl_custom_init by this custom mutator
|
||||
* @param[in] buf Pointer to the input data to be mutated and the mutated
|
||||
* output
|
||||
* @param[in] buf_size Size of the input/output data
|
||||
* @param[out] out_buf the new buffer. We may reuse *buf if large enough.
|
||||
* *out_buf = NULL is treated as FATAL.
|
||||
* @param[in] add_buf Buffer containing the additional test case
|
||||
* @param[out] out_buf The new buffer, under your memory mgmt.
|
||||
* @param[in] add_buf Buffer containing an additional test case (splicing)
|
||||
* @param[in] add_buf_size Size of the additional test case
|
||||
* @param[in] max_size Maximum size of the mutated output. The mutation must
|
||||
* not produce data larger than max_size.
|
||||
|
@ -564,8 +564,7 @@ u8 fuzz_one_original(afl_state_t *afl) {
|
||||
if (afl->cmplog_lvl == 3 ||
|
||||
(afl->cmplog_lvl == 2 && afl->queue_cur->tc_ref) ||
|
||||
afl->queue_cur->favored ||
|
||||
!(afl->fsrv.total_execs % afl->queued_items) ||
|
||||
get_cur_time() - afl->last_find_time > 300000) { // 300 seconds
|
||||
get_cur_time() - afl->last_find_time > 600000) { // 600 seconds
|
||||
|
||||
if (input_to_state_stage(afl, in_buf, out_buf, len)) {
|
||||
|
||||
|
Reference in New Issue
Block a user