| /* |
| american fuzzy lop++ - redqueen implementation on top of cmplog |
| --------------------------------------------------------------- |
| |
| Originally written by Michal Zalewski |
| |
| Forkserver design by Jann Horn <jannhorn@googlemail.com> |
| |
| Now maintained by by Marc Heuse <mh@mh-sec.de>, |
| Heiko Eißfeldt <heiko.eissfeldt@hexco.de> and |
| Andrea Fioraldi <andreafioraldi@gmail.com> |
| |
| Copyright 2016, 2017 Google Inc. All rights reserved. |
| Copyright 2019-2020 AFLplusplus Project. All rights reserved. |
| |
| Licensed under the Apache License, Version 2.0 (the "License"); |
| you may not use this file except in compliance with the License. |
| You may obtain a copy of the License at: |
| |
| http://www.apache.org/licenses/LICENSE-2.0 |
| |
| Shared code to handle the shared memory. This is used by the fuzzer |
| as well the other components like afl-tmin, afl-showmap, etc... |
| |
| */ |
| |
| #include "afl-fuzz.h" |
| #include "cmplog.h" |
| |
| static char** its_argv; |
| |
| ///// Colorization |
| |
| struct range { |
| |
| u32 start; |
| u32 end; |
| struct range* next; |
| |
| }; |
| |
| struct range* add_range(struct range* ranges, u32 start, u32 end) { |
| |
| struct range* r = ck_alloc_nozero(sizeof(struct range)); |
| r->start = start; |
| r->end = end; |
| r->next = ranges; |
| return r; |
| |
| } |
| |
| struct range* pop_biggest_range(struct range** ranges) { |
| |
| struct range* r = *ranges; |
| struct range* prev = NULL; |
| struct range* rmax = NULL; |
| struct range* prev_rmax = NULL; |
| u32 max_size = 0; |
| |
| while (r) { |
| |
| u32 s = r->end - r->start; |
| if (s >= max_size) { |
| |
| max_size = s; |
| prev_rmax = prev; |
| rmax = r; |
| |
| } |
| |
| prev = r; |
| r = r->next; |
| |
| } |
| |
| if (rmax) { |
| |
| if (prev_rmax) |
| prev_rmax->next = rmax->next; |
| else |
| *ranges = rmax->next; |
| |
| } |
| |
| return rmax; |
| |
| } |
| |
| u8 get_exec_checksum(u8* buf, u32 len, u32* cksum) { |
| |
| if (unlikely(common_fuzz_stuff(its_argv, buf, len))) return 1; |
| |
| *cksum = hash32(trace_bits, MAP_SIZE, HASH_CONST); |
| return 0; |
| |
| } |
| |
| static void rand_replace(u8* buf, u32 len) { |
| |
| u32 i; |
| for (i = 0; i < len; ++i) |
| buf[i] = UR(256); |
| |
| } |
| |
| u8 colorization(u8* buf, u32 len, u32 exec_cksum) { |
| |
| struct range* ranges = add_range(NULL, 0, len); |
| u8* backup = ck_alloc_nozero(len); |
| |
| u8 needs_write = 0; |
| |
| u64 orig_hit_cnt, new_hit_cnt; |
| orig_hit_cnt = queued_paths + unique_crashes; |
| |
| stage_name = "colorization"; |
| stage_short = "colorization"; |
| stage_max = 1000; |
| |
| struct range* rng; |
| stage_cur = stage_max; |
| while ((rng = pop_biggest_range(&ranges)) != NULL && stage_cur) { |
| |
| u32 s = rng->end - rng->start; |
| if (s == 0) goto empty_range; |
| |
| memcpy(backup, buf + rng->start, s); |
| rand_replace(buf + rng->start, s); |
| |
| u32 cksum; |
| if (unlikely(get_exec_checksum(buf, len, &cksum))) goto checksum_fail; |
| |
| if (cksum != exec_cksum) { |
| |
| ranges = add_range(ranges, rng->start, rng->start + s / 2); |
| ranges = add_range(ranges, rng->start + s / 2 + 1, rng->end); |
| memcpy(buf + rng->start, backup, s); |
| |
| } else |
| |
| needs_write = 1; |
| |
| empty_range: |
| ck_free(rng); |
| --stage_cur; |
| |
| } |
| |
| new_hit_cnt = queued_paths + unique_crashes; |
| stage_finds[STAGE_COLORIZATION] += new_hit_cnt - orig_hit_cnt; |
| stage_cycles[STAGE_COLORIZATION] += stage_max - stage_cur; |
| ck_free(backup); |
| |
| while (ranges) { |
| |
| rng = ranges; |
| ranges = ranges->next; |
| ck_free(rng); |
| |
| } |
| |
| // save the input with the high entropy |
| |
| if (needs_write) { |
| |
| s32 fd; |
| |
| if (no_unlink) { |
| |
| fd = open(queue_cur->fname, O_WRONLY | O_CREAT | O_TRUNC, 0600); |
| |
| } else { |
| |
| unlink(queue_cur->fname); /* ignore errors */ |
| fd = open(queue_cur->fname, O_WRONLY | O_CREAT | O_EXCL, 0600); |
| |
| } |
| |
| if (fd < 0) PFATAL("Unable to create '%s'", queue_cur->fname); |
| |
| ck_write(fd, buf, len, queue_cur->fname); |
| queue_cur->len = len; // no-op, just to be 100% safe |
| |
| close(fd); |
| |
| } |
| |
| return 0; |
| |
| checksum_fail: |
| ck_free(backup); |
| |
| while (ranges) { |
| |
| rng = ranges; |
| ranges = ranges->next; |
| ck_free(rng); |
| |
| } |
| |
| return 1; |
| |
| } |
| |
| ///// Input to State replacement |
| |
| u8 its_fuzz(u32 idx, u32 size, u8* buf, u32 len, u8* status) { |
| |
| u64 orig_hit_cnt, new_hit_cnt; |
| |
| orig_hit_cnt = queued_paths + unique_crashes; |
| |
| if (unlikely(common_fuzz_stuff(its_argv, buf, len))) return 1; |
| |
| new_hit_cnt = queued_paths + unique_crashes; |
| |
| if (unlikely(new_hit_cnt != orig_hit_cnt)) { |
| |
| *status = 1; |
| |
| } else { |
| |
| if (size >= MIN_AUTO_EXTRA && size <= MAX_AUTO_EXTRA) |
| maybe_add_auto(&buf[idx], size); |
| *status = 2; |
| |
| } |
| |
| return 0; |
| |
| } |
| |
| u8 cmp_extend_encoding(struct cmp_header* h, u64 pattern, u64 repl, u32 idx, |
| u8* orig_buf, u8* buf, u32 len, u8 do_reverse, |
| u8* status) { |
| |
| u64* buf_64 = (u64*)&buf[idx]; |
| u32* buf_32 = (u32*)&buf[idx]; |
| u16* buf_16 = (u16*)&buf[idx]; |
| // u8* buf_8 = &buf[idx]; |
| u64* o_buf_64 = (u64*)&orig_buf[idx]; |
| u32* o_buf_32 = (u32*)&orig_buf[idx]; |
| u16* o_buf_16 = (u16*)&orig_buf[idx]; |
| // u8* o_buf_8 = &orig_buf[idx]; |
| |
| u32 its_len = len - idx; |
| *status = 0; |
| |
| if (SHAPE_BYTES(h->shape) == 8) { |
| |
| if (its_len >= 8 && *buf_64 == pattern && *o_buf_64 == pattern) { |
| |
| *buf_64 = repl; |
| if (unlikely(its_fuzz(idx, 8, buf, len, status))) return 1; |
| *buf_64 = pattern; |
| |
| } |
| |
| // reverse encoding |
| if (do_reverse) |
| if (unlikely(cmp_extend_encoding(h, SWAP64(pattern), SWAP64(repl), idx, |
| orig_buf, buf, len, 0, status))) |
| return 1; |
| |
| } |
| |
| if (SHAPE_BYTES(h->shape) == 4 || *status == 2) { |
| |
| if (its_len >= 4 && *buf_32 == (u32)pattern && *o_buf_32 == (u32)pattern) { |
| |
| *buf_32 = (u32)repl; |
| if (unlikely(its_fuzz(idx, 4, buf, len, status))) return 1; |
| *buf_32 = pattern; |
| |
| } |
| |
| // reverse encoding |
| if (do_reverse) |
| if (unlikely(cmp_extend_encoding(h, SWAP32(pattern), SWAP32(repl), idx, |
| orig_buf, buf, len, 0, status))) |
| return 1; |
| |
| } |
| |
| if (SHAPE_BYTES(h->shape) == 2 || *status == 2) { |
| |
| if (its_len >= 2 && *buf_16 == (u16)pattern && *o_buf_16 == (u16)pattern) { |
| |
| *buf_16 = (u16)repl; |
| if (unlikely(its_fuzz(idx, 2, buf, len, status))) return 1; |
| *buf_16 = (u16)pattern; |
| |
| } |
| |
| // reverse encoding |
| if (do_reverse) |
| if (unlikely(cmp_extend_encoding(h, SWAP16(pattern), SWAP16(repl), idx, |
| orig_buf, buf, len, 0, status))) |
| return 1; |
| |
| } |
| |
| /*if (SHAPE_BYTES(h->shape) == 1 || *status == 2) { |
| |
| if (its_len >= 2 && *buf_8 == (u8)pattern && *o_buf_8 == (u8)pattern) { |
| |
| *buf_8 = (u8)repl; |
| if (unlikely(its_fuzz(idx, 1, buf, len, status))) |
| return 1; |
| *buf_16 = (u16)pattern; |
| |
| } |
| |
| }*/ |
| |
| return 0; |
| |
| } |
| |
| void try_to_add_to_dict(u64 v, u8 shape) { |
| |
| u8* b = (u8*)&v; |
| |
| u32 k; |
| u8 cons_ff = 0, cons_0 = 0; |
| for (k = 0; k < shape; ++k) { |
| |
| if (b[k] == 0) |
| ++cons_0; |
| else if (b[k] == 0xff) |
| ++cons_0; |
| else |
| cons_0 = cons_ff = 0; |
| |
| if (cons_0 > 1 || cons_ff > 1) return; |
| |
| } |
| |
| maybe_add_auto((u8*)&v, shape); |
| |
| u64 rev; |
| switch (shape) { |
| |
| case 1: break; |
| case 2: |
| rev = SWAP16((u16)v); |
| maybe_add_auto((u8*)&rev, shape); |
| break; |
| case 4: |
| rev = SWAP32((u32)v); |
| maybe_add_auto((u8*)&rev, shape); |
| break; |
| case 8: |
| rev = SWAP64(v); |
| maybe_add_auto((u8*)&rev, shape); |
| break; |
| |
| } |
| |
| } |
| |
| u8 cmp_fuzz(u32 key, u8* orig_buf, u8* buf, u32 len) { |
| |
| struct cmp_header* h = &cmp_map->headers[key]; |
| u32 i, j, idx; |
| |
| u32 loggeds = h->hits; |
| if (h->hits > CMP_MAP_H) loggeds = CMP_MAP_H; |
| |
| u8 status; |
| // opt not in the paper |
| u32 fails = 0; |
| |
| for (i = 0; i < loggeds; ++i) { |
| |
| struct cmp_operands* o = &cmp_map->log[key][i]; |
| |
| // opt not in the paper |
| for (j = 0; j < i; ++j) |
| if (cmp_map->log[key][j].v0 == o->v0 && cmp_map->log[key][i].v1 == o->v1) |
| goto cmp_fuzz_next_iter; |
| |
| for (idx = 0; idx < len && fails < 8; ++idx) { |
| |
| if (unlikely(cmp_extend_encoding(h, o->v0, o->v1, idx, orig_buf, buf, len, |
| 1, &status))) |
| return 1; |
| if (status == 2) |
| ++fails; |
| else if (status == 1) |
| break; |
| |
| if (unlikely(cmp_extend_encoding(h, o->v1, o->v0, idx, orig_buf, buf, len, |
| 1, &status))) |
| return 1; |
| if (status == 2) |
| ++fails; |
| else if (status == 1) |
| break; |
| |
| } |
| |
| // If failed, add to dictionary |
| if (fails == 8) { |
| |
| try_to_add_to_dict(o->v0, SHAPE_BYTES(h->shape)); |
| try_to_add_to_dict(o->v1, SHAPE_BYTES(h->shape)); |
| |
| } |
| |
| cmp_fuzz_next_iter: |
| stage_cur++; |
| |
| } |
| |
| return 0; |
| |
| } |
| |
| ///// Input to State stage |
| |
| // queue_cur->exec_cksum |
| u8 input_to_state_stage(char** argv, u8* orig_buf, u8* buf, u32 len, |
| u32 exec_cksum) { |
| |
| its_argv = argv; |
| |
| if (unlikely(colorization(buf, len, exec_cksum))) return 1; |
| |
| // do it manually, forkserver clear only trace_bits |
| memset(cmp_map->headers, 0, sizeof(cmp_map->headers)); |
| |
| if (unlikely(common_fuzz_cmplog_stuff(argv, buf, len))) return 1; |
| |
| u64 orig_hit_cnt, new_hit_cnt; |
| u64 orig_execs = total_execs; |
| orig_hit_cnt = queued_paths + unique_crashes; |
| |
| stage_name = "input-to-state"; |
| stage_short = "its"; |
| stage_max = 0; |
| stage_cur = 0; |
| |
| u32 k; |
| for (k = 0; k < CMP_MAP_W; ++k) { |
| |
| if (!cmp_map->headers[k].hits) continue; |
| if (cmp_map->headers[k].hits > CMP_MAP_H) |
| stage_max += CMP_MAP_H; |
| else |
| stage_max += cmp_map->headers[k].hits; |
| |
| } |
| |
| for (k = 0; k < CMP_MAP_W; ++k) { |
| |
| if (!cmp_map->headers[k].hits) continue; |
| cmp_fuzz(k, orig_buf, buf, len); |
| |
| } |
| |
| memcpy(orig_buf, buf, len); |
| |
| new_hit_cnt = queued_paths + unique_crashes; |
| stage_finds[STAGE_ITS] += new_hit_cnt - orig_hit_cnt; |
| stage_cycles[STAGE_ITS] += total_execs - orig_execs; |
| |
| return 0; |
| |
| } |
| |
