afl-fuzz.c

/*
   american fuzzy lop - fuzzer code
   --------------------------------

   Written and maintained by Michal Zalewski <lcamtuf@google.com>

   Forkserver design by Jann Horn <jannhorn@googlemail.com>

   Copyright 2013, 2014, 2015 Google Inc. 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

   This is the real deal: the program takes an instrumented binary and
   attempts a variety of basic fuzzing tricks, paying close attention to
   how they affect the execution path.

 */

#define AFL_MAIN
#define MESSAGES_TO_STDOUT

#define _GNU_SOURCE
#define _FILE_OFFSET_BITS 64

#include "config.h"
#include "types.h"
#include "debug.h"
#include "alloc-inl.h"
#include "hash.h"

#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <errno.h>
#include <signal.h>
#include <dirent.h>
#include <ctype.h>
#include <fcntl.h>
#include <termios.h>
#include <dlfcn.h>

#include <sys/wait.h>
#include <sys/time.h>
#include <sys/shm.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <sys/resource.h>
#include <sys/mman.h>
#include <sys/ioctl.h>
#include <sys/file.h>

#if defined(__APPLE__) || defined(__FreeBSD__) || defined (__OpenBSD__)
#  include <sys/sysctl.h>
#endif /* __APPLE__ || __FreeBSD__ || __OpenBSD__ */

# define AFL_CTL_ASSOC_AREA (42)
# define AFL_CTL_DISASSOC_AREA (43)
# define AFL_CTL_GET_MMAP_OFFSET (44)

/* Lots of globals, but mostly for the status UI and other things where it
   really makes no sense to haul them around as function parameters. */

static u8 *in_dir,                    /* Input directory with test cases  */
          *out_file,                  /* File to fuzz, if any             */
          *out_dir,                   /* Working & output directory       */
          *sync_dir,                  /* Synchronization directory        */
          *sync_id,                   /* Fuzzer ID                        */
          *use_banner,                /* Display banner                   */
          *in_bitmap,                 /* Input bitmap                     */
          *doc_path,                  /* Path to documentation dir        */
          *target_path,               /* Path to target binary            */
          *orig_cmdline;              /* Original command line            */

static u32 exec_tmout = EXEC_TIMEOUT; /* Configurable exec timeout (ms)   */
static u64 mem_limit = MEM_LIMIT;     /* Memory cap for child (MB)        */

static u32 stats_update_freq = 1;     /* Stats update frequency (execs)   */

static u8  skip_deterministic,        /* Skip deterministic stages?       */
           force_deterministic,       /* Force deterministic stages?      */
           use_splicing,              /* Recombine input files?           */
           dumb_mode,                 /* Run in non-instrumented mode?    */
           score_changed,             /* Scoring for favorites changed?   */
           kill_signal,               /* Signal that killed the child     */
           resuming_fuzz,             /* Resuming an older fuzzing job?   */
           timeout_given,             /* Specific timeout given?          */
           not_on_tty,                /* stdout is not a tty              */
           term_too_small,            /* terminal dimensions too small    */
           uses_asan,                 /* Target uses ASAN?                */
           no_forkserver,             /* Disable forkserver?              */
           crash_mode,                /* Crash mode! Yeah!                */
           in_place_resume,           /* Attempt in-place resume?         */
           auto_changed,              /* Auto-generated tokens changed?   */
           no_cpu_meter_red,          /* Feng shui on the status screen   */
           no_var_check,              /* Don't detect variable behavior   */
           bitmap_changed = 1,        /* Time to update bitmap?           */
           qemu_mode,                 /* Running in QEMU mode?            */
           skip_requested,            /* Skip request, via SIGUSR1        */
           run_over10m;               /* Run time over 10 minutes?        */

static s32 out_fd,                    /* Persistent fd for out_file       */
           dev_urandom_fd = -1,       /* Persistent fd for /dev/urandom   */
           dev_null_fd = -1,          /* Persistent fd for /dev/null      */
           fsrv_ctl_fd,               /* Fork server control pipe (write) */
           fsrv_st_fd;                /* Fork server status pipe (read)   */

static s32 forksrv_pid,               /* PID of the fork server           */
           child_pid = -1,            /* PID of the fuzzed program        */
           out_dir_fd = -1;           /* FD of the lock file              */

static u8* trace_bits;                /* SHM with instrumentation bitmap  */

static u8  virgin_bits[MAP_SIZE],     /* Regions yet untouched by fuzzing */
           virgin_hang[MAP_SIZE],     /* Bits we haven't seen in hangs    */
           virgin_crash[MAP_SIZE];    /* Bits we haven't seen in crashes  */

/* static s32 shm_id;                    /\* ID of the SHM region             *\/ */
static int fd_afl_area;
static off_t mmap_offset;

static volatile u8 stop_soon,         /* Ctrl-C pressed?                  */
                   clear_screen = 1,  /* Window resized?                  */
                   child_timed_out;   /* Traced process timed out?        */

static u32 queued_paths,              /* Total number of queued testcases */
           queued_variable,           /* Testcases with variable behavior */
           queued_at_start,           /* Total number of initial inputs   */
           queued_discovered,         /* Items discovered during this run */
           queued_imported,           /* Items imported via -S            */
           queued_favored,            /* Paths deemed favorable           */
           queued_with_cov,           /* Paths with new coverage bytes    */
           pending_not_fuzzed,        /* Queued but not done yet          */
           pending_favored,           /* Pending favored paths            */
           cur_skipped_paths,         /* Abandoned inputs in cur cycle    */
           cur_depth,                 /* Current path depth               */
           max_depth,                 /* Max path depth                   */
           useless_at_start,          /* Number of useless starting paths */
           current_entry,             /* Current queue entry ID           */
           havoc_div = 1;             /* Cycle count divisor for havoc    */

static u64 total_crashes,             /* Total number of crashes          */
           unique_crashes,            /* Crashes with unique signatures   */
           total_hangs,               /* Total number of hangs            */
           unique_hangs,              /* Hangs with unique signatures     */
           total_execs,               /* Total execve() calls             */
           start_time,                /* Unix start time (ms)             */
           last_path_time,            /* Time for most recent path (ms)   */
           last_crash_time,           /* Time for most recent crash (ms)  */
           last_hang_time,            /* Time for most recent hang (ms)   */
           queue_cycle,               /* Queue round counter              */
           cycles_wo_finds,           /* Cycles without any new paths     */
           trim_execs,                /* Execs done to trim input files   */
           bytes_trim_in,             /* Bytes coming into the trimmer    */
           bytes_trim_out,            /* Bytes coming outa the trimmer    */
           blocks_eff_total,          /* Blocks subject to effector maps  */
           blocks_eff_select;         /* Blocks selected as fuzzable      */

static u32 subseq_hangs;              /* Number of hangs in a row         */

static u8 *stage_name = "init",       /* Name of the current fuzz stage   */
          *stage_short,               /* Short stage name                 */
          *syncing_party;             /* Currently syncing with...        */

static s32 stage_cur, stage_max;      /* Stage progression                */
static s32 splicing_with = -1;        /* Splicing with which test case?   */

static u32 syncing_case;              /* Syncing with case #...           */

static s32 stage_cur_byte,            /* Byte offset of current stage op  */
           stage_cur_val;             /* Value used for stage op          */

static u8  stage_val_type;            /* Value type (STAGE_VAL_*)         */

static u64 stage_finds[32],           /* Patterns found per fuzz stage    */
           stage_cycles[32];          /* Execs per fuzz stage             */

static u32 rand_cnt;                  /* Random number counter            */

static u64 total_cal_us,              /* Total calibration time (us)      */
           total_cal_cycles;          /* Total calibration cycles         */

static u64 total_bitmap_size,         /* Total bit count for all bitmaps  */
           total_bitmap_entries;      /* Number of bitmaps counted        */

static u32 cpu_core_count;            /* CPU core count                   */

static FILE* plot_file;               /* Gnuplot output file              */

struct queue_entry {

  u8* fname;                          /* File name for the test case      */
  u32 len;                            /* Input length                     */

  u8  cal_failed,                     /* Calibration failed?              */
      trim_done,                      /* Trimmed?                         */
      was_fuzzed,                     /* Had any fuzzing done yet?        */
      passed_det,                     /* Deterministic stages passed?     */
      has_new_cov,                    /* Triggers new coverage?           */
      var_behavior,                   /* Variable behavior?               */
      favored,                        /* Currently favored?               */
      fs_redundant;                   /* Marked as redundant in the fs?   */

  u32 bitmap_size,                    /* Number of bits set in bitmap     */
      exec_cksum;                     /* Checksum of the execution trace  */

  u64 exec_us,                        /* Execution time (us)              */
      handicap,                       /* Number of queue cycles behind    */
      depth;                          /* Path depth                       */

  u8* trace_mini;                     /* Trace bytes, if kept             */
  u32 tc_ref;                         /* Trace bytes ref count            */

  struct queue_entry *next,           /* Next element, if any             */
                     *next_100;       /* 100 elements ahead               */

};

static struct queue_entry *queue,     /* Fuzzing queue (linked list)      */
                          *queue_cur, /* Current offset within the queue  */
                          *queue_top, /* Top of the list                  */
                          *q_prev100; /* Previous 100 marker              */

static struct queue_entry*
  top_rated[MAP_SIZE];                /* Top entries for bitmap bytes     */

struct extra_data {
  u8* data;                           /* Dictionary token data            */
  u32 len;                            /* Dictionary token length          */
  u32 hit_cnt;                        /* Use count in the corpus          */
};

static struct extra_data* extras;     /* Extra tokens to fuzz with        */
static u32 extras_cnt;                /* Total number of tokens read      */

static struct extra_data* a_extras;   /* Automatically selected extras    */
static u32 a_extras_cnt;              /* Total number of tokens available */

static u8* (*post_handler)(u8* buf, u32* len);

/* Interesting values, as per config.h */

static s8  interesting_8[]  = { INTERESTING_8 };
static s16 interesting_16[] = { INTERESTING_8, INTERESTING_16 };
static s32 interesting_32[] = { INTERESTING_8, INTERESTING_16, INTERESTING_32 };

/* Fuzzing stages */

enum {
  /* 00 */ STAGE_FLIP1,
  /* 01 */ STAGE_FLIP2,
  /* 02 */ STAGE_FLIP4,
  /* 03 */ STAGE_FLIP8,
  /* 04 */ STAGE_FLIP16,
  /* 05 */ STAGE_FLIP32,
  /* 06 */ STAGE_ARITH8,
  /* 07 */ STAGE_ARITH16,
  /* 08 */ STAGE_ARITH32,
  /* 09 */ STAGE_INTEREST8,
  /* 10 */ STAGE_INTEREST16,
  /* 11 */ STAGE_INTEREST32,
  /* 12 */ STAGE_EXTRAS_UO,
  /* 13 */ STAGE_EXTRAS_UI,
  /* 14 */ STAGE_EXTRAS_AO,
  /* 15 */ STAGE_HAVOC,
  /* 16 */ STAGE_SPLICE
};

/* Stage value types */

enum {
  /* 00 */ STAGE_VAL_NONE,
  /* 01 */ STAGE_VAL_LE,
  /* 02 */ STAGE_VAL_BE
};

/* Execution status fault codes */

enum {
  /* 00 */ FAULT_NONE,
  /* 01 */ FAULT_HANG,
  /* 02 */ FAULT_CRASH,
  /* 03 */ FAULT_ERROR,
  /* 04 */ FAULT_NOINST,
  /* 05 */ FAULT_NOBITS
};


/* Get unix time in milliseconds */

static u64 get_cur_time(void) {

  struct timeval tv;
  struct timezone tz;

  gettimeofday(&tv, &tz);

  return (tv.tv_sec * 1000ULL) + (tv.tv_usec / 1000);

}


/* Get unix time in microseconds */

static u64 get_cur_time_us(void) {

  struct timeval tv;
  struct timezone tz;

  gettimeofday(&tv, &tz);

  return (tv.tv_sec * 1000000ULL) + tv.tv_usec;

}


/* Generate a random number (from 0 to limit - 1). This may
   have slight bias. */

static inline u32 UR(u32 limit) {

  if (!rand_cnt--) {

    u32 seed[2];

    ck_read(dev_urandom_fd, &seed, sizeof(seed), "/dev/urandom");

    srandom(seed[0]);
    rand_cnt = (RESEED_RNG / 2) + (seed[1] % RESEED_RNG);

  }

  return random() % limit;

}


#ifndef IGNORE_FINDS

/* Helper function to compare buffers; returns first and last differing offset. We
   use this to find reasonable locations for splicing two files. */

static void locate_diffs(u8* ptr1, u8* ptr2, u32 len, s32* first, s32* last) {

  s32 f_loc = -1;
  s32 l_loc = -1;
  u32 pos;

  for (pos = 0; pos < len; pos++) {

    if (*(ptr1++) != *(ptr2++)) {

      if (f_loc == -1) f_loc = pos;
      l_loc = pos;

    }

  }

  *first = f_loc;
  *last = l_loc;

  return;

}

#endif /* !IGNORE_FINDS */


/* Describe integer. Uses 12 cyclic static buffers for return values. The value
   returned should be five characters or less for all the integers we reasonably
   expect to see. */

static u8* DI(u64 val) {

  static u8 tmp[12][16];
  static u8 cur;

  cur = (cur + 1) % 12;

#define CHK_FORMAT(_divisor, _limit_mult, _fmt, _cast) do { \
    if (val < (_divisor) * (_limit_mult)) { \
      sprintf(tmp[cur], _fmt, ((_cast)val) / (_divisor)); \
      return tmp[cur]; \
    } \
  } while (0)

  /* 0-9999 */
  CHK_FORMAT(1, 10000, "%llu", u64);

  /* 10.0k - 99.9k */
  CHK_FORMAT(1000, 99.95, "%0.01fk", double);

  /* 100k - 999k */
  CHK_FORMAT(1000, 1000, "%lluk", u64);

  /* 1.00M - 9.99M */
  CHK_FORMAT(1000 * 1000, 9.995, "%0.02fM", double);

  /* 10.0M - 99.9M */
  CHK_FORMAT(1000 * 1000, 99.95, "%0.01fM", double);

  /* 100M - 999M */
  CHK_FORMAT(1000 * 1000, 1000, "%lluM", u64);

  /* 1.00G - 9.99G */
  CHK_FORMAT(1000LL * 1000 * 1000, 9.995, "%0.02fG", double);

  /* 10.0G - 99.9G */
  CHK_FORMAT(1000LL * 1000 * 1000, 99.95, "%0.01fG", double);

  /* 100G - 999G */
  CHK_FORMAT(1000LL * 1000 * 1000, 1000, "%lluG", u64);

  /* 1.00T - 9.99G */
  CHK_FORMAT(1000LL * 1000 * 1000 * 1000, 9.995, "%0.02fT", double);

  /* 10.0T - 99.9T */
  CHK_FORMAT(1000LL * 1000 * 1000 * 1000, 99.95, "%0.01fT", double);

  /* 100T+ */
  strcpy(tmp[cur], "infty");
  return tmp[cur];

}


/* Describe float. Similar to the above, except with a single 
   static buffer. */

static u8* DF(double val) {

  static u8 tmp[16];

  if (val < 99.995) {
    sprintf(tmp, "%0.02f", val);
    return tmp;
  }

  if (val < 999.95) {
    sprintf(tmp, "%0.01f", val);
    return tmp;
  }

  return DI((u64)val);

}


/* Describe integer as memory size. */

static u8* DMS(u64 val) {

  static u8 tmp[12][16];
  static u8 cur;

  cur = (cur + 1) % 12;

  /* 0-9999 */
  CHK_FORMAT(1, 10000, "%llu B", u64);

  /* 10.0k - 99.9k */
  CHK_FORMAT(1024, 99.95, "%0.01f kB", double);

  /* 100k - 999k */
  CHK_FORMAT(1024, 1000, "%llu kB", u64);

  /* 1.00M - 9.99M */
  CHK_FORMAT(1024 * 1024, 9.995, "%0.02f MB", double);

  /* 10.0M - 99.9M */
  CHK_FORMAT(1024 * 1024, 99.95, "%0.01f MB", double);

  /* 100M - 999M */
  CHK_FORMAT(1024 * 1024, 1000, "%llu MB", u64);

  /* 1.00G - 9.99G */
  CHK_FORMAT(1024LL * 1024 * 1024, 9.995, "%0.02f GB", double);

  /* 10.0G - 99.9G */
  CHK_FORMAT(1024LL * 1024 * 1024, 99.95, "%0.01f GB", double);

  /* 100G - 999G */
  CHK_FORMAT(1024LL * 1024 * 1024, 1000, "%llu GB", u64);

  /* 1.00T - 9.99G */
  CHK_FORMAT(1024LL * 1024 * 1024 * 1024, 9.995, "%0.02f TB", double);

  /* 10.0T - 99.9T */
  CHK_FORMAT(1024LL * 1024 * 1024 * 1024, 99.95, "%0.01f TB", double);

#undef CHK_FORMAT

  /* 100T+ */
  strcpy(tmp[cur], "infty");
  return tmp[cur];

}


/* Describe time delta. Returns one static buffer, 34 chars of less. */

static u8* DTD(u64 cur_ms, u64 event_ms) {

  static u8 tmp[64];
  u64 delta;
  s32 t_d, t_h, t_m, t_s;

  if (!event_ms) return "none seen yet";

  delta = cur_ms - event_ms;

  t_d = delta / 1000 / 60 / 60 / 24;
  t_h = (delta / 1000 / 60 / 60) % 24;
  t_m = (delta / 1000 / 60) % 60;
  t_s = (delta / 1000) % 60;

  sprintf(tmp, "%s days, %u hrs, %u min, %u sec", DI(t_d), t_h, t_m, t_s);
  return tmp;

}


/* Mark deterministic checks as done for a particular queue entry. We use the
   .state file to avoid repeating deterministic fuzzing when resuming aborted
   scans. */

static void mark_as_det_done(struct queue_entry* q) {

  u8* fn = strrchr(q->fname, '/');
  s32 fd;

  fn = alloc_printf("%s/queue/.state/deterministic_done/%s", out_dir, fn + 1);

  fd = open(fn, O_WRONLY | O_CREAT | O_EXCL, 0600);
  if (fd < 0) PFATAL("Unable to create '%s'", fn);
  close(fd);

  ck_free(fn);

  q->passed_det = 1;

}


/* Mark as variable. Create symlinks if possible to make it easier to examine
   the files. */

static void mark_as_variable(struct queue_entry* q) {

  u8 *fn = strrchr(q->fname, '/') + 1, *ldest;

  ldest = alloc_printf("../../%s", fn);
  fn = alloc_printf("%s/queue/.state/variable_behavior/%s", out_dir, fn);

  if (symlink(ldest, fn)) {

    s32 fd = open(fn, O_WRONLY | O_CREAT | O_EXCL, 0600);
    if (fd < 0) PFATAL("Unable to create '%s'", fn);
    close(fd);

  }

  ck_free(ldest);
  ck_free(fn);

  q->var_behavior = 1;

}


/* Mark / unmark as redundant (edge-only). This is not used for restoring state,
   but may be useful for post-processing datasets. */

static void mark_as_redundant(struct queue_entry* q, u8 state) {

  u8* fn;
  s32 fd;

  if (state == q->fs_redundant) return;

  q->fs_redundant = state;

  fn = strrchr(q->fname, '/');
  fn = alloc_printf("%s/queue/.state/redundant_edges/%s", out_dir, fn + 1);

  if (state) {

    fd = open(fn, O_WRONLY | O_CREAT | O_EXCL, 0600);
    if (fd < 0) PFATAL("Unable to create '%s'", fn);
    close(fd);

  } else {

    if (unlink(fn)) PFATAL("Unable to remove '%s'", fn);

  }

  ck_free(fn);

}


/* Append new test case to the queue. */

static void add_to_queue(u8* fname, u32 len, u8 passed_det) {

  struct queue_entry* q = ck_alloc(sizeof(struct queue_entry));

  q->fname        = fname;
  q->len          = len;
  q->depth        = cur_depth + 1;
  q->passed_det   = passed_det;

  if (q->depth > max_depth) max_depth = q->depth;

  if (queue_top) {

    queue_top->next = q;
    queue_top = q;

  } else q_prev100 = queue = queue_top = q;

  queued_paths++;
  pending_not_fuzzed++;

  if (!(queued_paths % 100)) {

    q_prev100->next_100 = q;
    q_prev100 = q;

  }

  last_path_time = get_cur_time();

}


/* Destroy the entire queue. */

static void destroy_queue(void) {

  struct queue_entry *q = queue, *n;

  while (q) {

    n = q->next;
    ck_free(q->fname);
    ck_free(q->trace_mini);
    ck_free(q);
    q = n;

  }

}


/* Write bitmap to file. The bitmap is useful mostly for the secret
   -B option, to focus a separate fuzzing session on a particular
   interesting input without rediscovering all the others. */

static void write_bitmap(void) {

  u8* fname;
  s32 fd;

  if (!bitmap_changed) return;
  bitmap_changed = 0;

  fname = alloc_printf("%s/fuzz_bitmap", out_dir);
  fd = open(fname, O_WRONLY | O_CREAT | O_TRUNC, 0600);

  if (fd < 0) PFATAL("Unable to open '%s'", fname);

  ck_write(fd, virgin_bits, MAP_SIZE, fname);

  close(fd);
  ck_free(fname);

}


/* Read bitmap from file. This is for the -B option again. */

static void read_bitmap(u8* fname) {

  s32 fd = open(fname, O_RDONLY);

  if (fd < 0) PFATAL("Unable to open '%s'", fname);

  ck_read(fd, virgin_bits, MAP_SIZE, fname);

  close(fd);

}


/* Check if the current execution path brings anything new to the table.
   Update virgin bits to reflect the finds. Returns 1 if the only change is
   the hit-count for a particular tuple; 2 if there are new tuples seen. 
   Updates the map, so subsequent calls will always return 0.

   This function is called after every exec() on a fairly large buffer, so
   it needs to be fast. We do this in 32-bit and 64-bit flavors. */

#define FFL(_b) (0xffULL << ((_b) << 3))
#define FF(_b)  (0xff << ((_b) << 3))

static inline u8 has_new_bits(u8* virgin_map) {

#ifdef __x86_64__

  u64* current = (u64*)trace_bits;
  u64* virgin  = (u64*)virgin_map;

  u32  i = (MAP_SIZE >> 3);

#else

  u32* current = (u32*)trace_bits;
  u32* virgin  = (u32*)virgin_map;

  u32  i = (MAP_SIZE >> 2);

#endif /* ^__x86_64__ */

  u8   ret = 0;

  while (i--) {

#ifdef __x86_64__

    u64 cur = *current;
    u64 vir = *virgin;

#else

    u32 cur = *current;
    u32 vir = *virgin;

#endif /* ^__x86_64__ */

    /* Optimize for *current == ~*virgin, since this will almost always be the
       case. */

    if (cur & vir) {

      if (ret < 2) {

        /* This trace did not have any new bytes yet; see if there's any
           current[] byte that is non-zero when virgin[] is 0xff. */

#ifdef __x86_64__

        if (((cur & FFL(0)) && (vir & FFL(0)) == FFL(0)) ||
            ((cur & FFL(1)) && (vir & FFL(1)) == FFL(1)) ||
            ((cur & FFL(2)) && (vir & FFL(2)) == FFL(2)) ||
            ((cur & FFL(3)) && (vir & FFL(3)) == FFL(3)) ||
            ((cur & FFL(4)) && (vir & FFL(4)) == FFL(4)) ||
            ((cur & FFL(5)) && (vir & FFL(5)) == FFL(5)) ||
            ((cur & FFL(6)) && (vir & FFL(6)) == FFL(6)) ||
            ((cur & FFL(7)) && (vir & FFL(7)) == FFL(7))) ret = 2;
        else ret = 1;

#else

        if (((cur & FF(0)) && (vir & FF(0)) == FF(0)) ||
            ((cur & FF(1)) && (vir & FF(1)) == FF(1)) ||
            ((cur & FF(2)) && (vir & FF(2)) == FF(2)) ||
            ((cur & FF(3)) && (vir & FF(3)) == FF(3))) ret = 2;
        else ret = 1;

#endif /* ^__x86_64__ */

      }

      *virgin = vir & ~cur;

    }

    current++;
    virgin++;

  }

  if (ret && virgin_map == virgin_bits) bitmap_changed = 1;

  return ret;

}


/* Count the number of bits set in the provided bitmap. Used for the status
   screen several times every second, does not have to be fast. */

static u32 count_bits(u8* mem) {

  u32* ptr = (u32*)mem;
  u32  i   = (MAP_SIZE >> 2);
  u32  ret = 0;

  while (i--) {

    u32 v = *(ptr++);

    /* This gets called on the inverse, virgin bitmap; optimize for sparse
       data. */

    if (v == 0xffffffff) {
      ret += 32;
      continue;
    }

    v -= ((v >> 1) & 0x55555555);
    v = (v & 0x33333333) + ((v >> 2) & 0x33333333);
    ret += (((v + (v >> 4)) & 0xF0F0F0F) * 0x01010101) >> 24;

  }

  return ret;

}


/* Count the number of bytes set in the bitmap. Called fairly sporadically,
   mostly to update the status screen or calibrate and examine confirmed
   new paths. */

static u32 count_bytes(u8* mem) {

  u32* ptr = (u32*)mem;
  u32  i   = (MAP_SIZE >> 2);
  u32  ret = 0;

  while (i--) {

    u32 v = *(ptr++);

    if (!v) continue;
    if (v & FF(0)) ret++;
    if (v & FF(1)) ret++;
    if (v & FF(2)) ret++;
    if (v & FF(3)) ret++;

  }

  return ret;

}


/* Count the number of non-255 bytes set in the bitmap. Used strictly for the
   status screen, several calls per second or so. */

static u32 count_non_255_bytes(u8* mem) {

  u32* ptr = (u32*)mem;
  u32  i   = (MAP_SIZE >> 2);
  u32  ret = 0;

  while (i--) {

    u32 v = *(ptr++);

    /* This is called on the virgin bitmap, so optimize for the most likely
       case. */

    if (v == 0xffffffff) continue;
    if ((v & FF(0)) != FF(0)) ret++;
    if ((v & FF(1)) != FF(1)) ret++;
    if ((v & FF(2)) != FF(2)) ret++;
    if ((v & FF(3)) != FF(3)) ret++;

  }

  return ret;

}


/* Destructively simplify trace by eliminating hit count information
   and replacing it with 0x80 or 0x01 depending on whether the tuple
   is hit or not. Called on every new crash or hang, should be
   reasonably fast. */

#define AREP4(_sym)   (_sym), (_sym), (_sym), (_sym)
#define AREP8(_sym)   AREP4(_sym), AREP4(_sym)
#define AREP16(_sym)  AREP8(_sym), AREP8(_sym)
#define AREP32(_sym)  AREP16(_sym), AREP16(_sym)
#define AREP64(_sym)  AREP32(_sym), AREP32(_sym)
#define AREP128(_sym) AREP64(_sym), AREP64(_sym)

static u8 simplify_lookup[256] = { 
  /*    4 */ 1, 128, 128, 128,
  /*   +4 */ AREP4(128),
  /*   +8 */ AREP8(128),
  /*  +16 */ AREP16(128),
  /*  +32 */ AREP32(128),
  /*  +64 */ AREP64(128),
  /* +128 */ AREP128(128)
};

#ifdef __x86_64__

static void simplify_trace(u64* mem) {

  u32 i = MAP_SIZE >> 3;

  while (i--) {

    /* Optimize for sparse bitmaps. */

    if (*mem) {

      u8* mem8 = (u8*)mem;

      mem8[0] = simplify_lookup[mem8[0]];
      mem8[1] = simplify_lookup[mem8[1]];
      mem8[2] = simplify_lookup[mem8[2]];
      mem8[3] = simplify_lookup[mem8[3]];
      mem8[4] = simplify_lookup[mem8[4]];
      mem8[5] = simplify_lookup[mem8[5]];
      mem8[6] = simplify_lookup[mem8[6]];
      mem8[7] = simplify_lookup[mem8[7]];

    } else *mem = 0x0101010101010101ULL;

    mem++;

  }

}

#else

static void simplify_trace(u32* mem) {

  u32 i = MAP_SIZE >> 2;

  while (i--) {

    /* Optimize for sparse bitmaps. */

    if (*mem) {

      u8* mem8 = (u8*)mem;

      mem8[0] = simplify_lookup[mem8[0]];
      mem8[1] = simplify_lookup[mem8[1]];
      mem8[2] = simplify_lookup[mem8[2]];
      mem8[3] = simplify_lookup[mem8[3]];

    } else *mem = 0x01010101;

    mem++;
  }

}

#endif /* ^__x86_64__ */


/* Destructively classify execution counts in a trace. This is used as a
   preprocessing step for any newly acquired traces. Called on every exec,
   must be fast. */

static u8 count_class_lookup[256] = {

  /* 0 - 3:       4 */ 0, 1, 2, 4,
  /* 4 - 7:      +4 */ AREP4(8),
  /* 8 - 15:     +8 */ AREP8(16),
  /* 16 - 31:   +16 */ AREP16(32),
  /* 32 - 127:  +96 */ AREP64(64), AREP32(64),
  /* 128+:     +128 */ AREP128(128)

};

#ifdef __x86_64__

static inline void classify_counts(u64* mem) {

  u32 i = MAP_SIZE >> 3;

  while (i--) {

    /* Optimize for sparse bitmaps. */

    if (*mem) {
      u8* mem8 = (u8*)mem;

      mem8[0] = count_class_lookup[mem8[0]];
      mem8[1] = count_class_lookup[mem8[1]];
      mem8[2] = count_class_lookup[mem8[2]];
      mem8[3] = count_class_lookup[mem8[3]];
      mem8[4] = count_class_lookup[mem8[4]];
      mem8[5] = count_class_lookup[mem8[5]];
      mem8[6] = count_class_lookup[mem8[6]];
      mem8[7] = count_class_lookup[mem8[7]];
    }

    mem++;

  }

}

#else

static inline void classify_counts(u32* mem) {

  u32 i = MAP_SIZE >> 2;

  while (i--) {

    /* Optimize for sparse bitmaps. */

    if (*mem) {

      u8* mem8 = (u8*)mem;

      mem8[0] = count_class_lookup[mem8[0]];
      mem8[1] = count_class_lookup[mem8[1]];
      mem8[2] = count_class_lookup[mem8[2]];
      mem8[3] = count_class_lookup[mem8[3]];

    }

    mem++;

  }

}

#endif /* ^__x86_64__ */



/* Compact trace bytes into a smaller bitmap. We effectively just drop the
   count information here. This is called only sporadically, for some
   new paths. */

static void minimize_bits(u8* dst, u8* src) {

  u32 i = 0;

  while (i < MAP_SIZE) {

    if (*(src++)) dst[i >> 3] |= 1 << (i & 7);
    i++;

  }

}


/* When we bump into a new path, we call this to see if the path appears
   more "favorable" than any of the existing ones. The purpose of the
   "favorables" is to have a minimal set of paths that trigger all the bits
   seen in the bitmap so far, and focus on fuzzing them at the expense of
   the rest.

   The first step of the process is to maintain a list of top_rated[] entries
   for every byte in the bitmap. We win that slot if there is no previous
   contender, or if the contender has a more favorable speed x size factor. */

static void update_bitmap_score(struct queue_entry* q) {

  u32 i;
  u64 fav_factor = q->exec_us * q->len;

  /* For every byte set in trace_bits[], see if there is a previous winner,
     and how it compares to us. */

  for (i = 0; i < MAP_SIZE; i++)

    if (trace_bits[i]) {

       if (top_rated[i]) {

         /* Faster-executing or smaller test cases are favored. */

         if (fav_factor > top_rated[i]->exec_us * top_rated[i]->len) continue;

         /* Looks like we're going to win. Decrease ref count for the
            previous winner, discard its trace_bits[] if necessary. */

         if (!--top_rated[i]->tc_ref) {
           ck_free(top_rated[i]->trace_mini);
           top_rated[i]->trace_mini = 0;
         }

       }

       /* Insert ourselves as the new winner. */

       top_rated[i] = q;
       q->tc_ref++;

       if (!q->trace_mini) {
         q->trace_mini = ck_alloc(MAP_SIZE >> 3);
         minimize_bits(q->trace_mini, trace_bits);
       }

       score_changed = 1;

     }

}


/* The second part of the mechanism discussed above is a routine that
   goes over top_rated[] entries, and then sequentially grabs winners for
   previously-unseen bytes (temp_v) and marks them as favored, at least
   until the next run. The favored entries are given more air time during
   all fuzzing steps. */

static void cull_queue(void) {

  struct queue_entry* q;
  static u8 temp_v[MAP_SIZE >> 3];
  u32 i;

  if (dumb_mode || !score_changed) return;

  score_changed = 0;

  memset(temp_v, 255, MAP_SIZE >> 3);

  queued_favored  = 0;
  pending_favored = 0;

  q = queue;

  while (q) {
    q->favored = 0;
    q = q->next;
  }

  /* Let's see if anything in the bitmap isn't captured in temp_v.
     If yes, and if it has a top_rated[] contender, let's use it. */

  for (i = 0; i < MAP_SIZE; i++)
    if (top_rated[i] && (temp_v[i >> 3] & (1 << (i & 7)))) {

      u32 j = MAP_SIZE >> 3;

      /* Remove all bits belonging to the current entry from temp_v. */

      while (j--) 
        if (top_rated[i]->trace_mini[j])
          temp_v[j] &= ~top_rated[i]->trace_mini[j];

      top_rated[i]->favored = 1;
      queued_favored++;

      if (!top_rated[i]->was_fuzzed) pending_favored++;

    }

  q = queue;

  while (q) {
    mark_as_redundant(q, !q->favored);
    q = q->next;
  }

}


/* Configure shared memory and virgin_bits. This is called at startup. */

static void setup_shm(void) {

  if (!in_bitmap) memset(virgin_bits, 255, MAP_SIZE);

  memset(virgin_hang, 255, MAP_SIZE);
  memset(virgin_crash, 255, MAP_SIZE);

  fd_afl_area = open("/dev/afl", O_RDWR);
  if (fd_afl_area < 0) PFATAL("open(/dev/afl) failed");

  if (dup2(fd_afl_area, DEVAFL_FD) < 0) PFATAL("dup2(fd_afl_area, 42) failed");
  close(fd_afl_area);
  fd_afl_area = DEVAFL_FD;

  mmap_offset = ioctl(fd_afl_area, AFL_CTL_GET_MMAP_OFFSET, 0);

  trace_bits = mmap(NULL, MAP_SIZE, PROT_READ | PROT_WRITE, MAP_SHARED, fd_afl_area, mmap_offset);
  
  if (trace_bits == MAP_FAILED) PFATAL("mmap() failed");

}


/* Load postprocessor, if available. */

static void setup_post(void) {

  void* dh;
  u8* fn = getenv("AFL_POST_LIBRARY");
  u32 tlen = 6;

  if (!fn) return;

  ACTF("Loading postprocessor from '%s'...", fn);

  dh = dlopen(fn, RTLD_NOW);
  if (!dh) FATAL("%s", dlerror());

  post_handler = dlsym(dh, "afl_postprocess");
  if (!post_handler) FATAL("Symbol 'afl_postprocess' not found.");

  /* Do a quick test. It's better to segfault now than later =) */

  post_handler("hello", &tlen);

  OKF("Postprocessor installed successfully.");

}


/* Read all testcases from the input directory, then queue them for testing.
   Called at startup. */

static void read_testcases(void) {

  struct dirent **nl;
  s32 nl_cnt;
  u32 i;
  u8* fn;

  /* Auto-detect non-in-place resumption attempts. */

  fn = alloc_printf("%s/queue", in_dir);
  if (!access(fn, F_OK)) in_dir = fn; else ck_free(fn);

  ACTF("Scanning '%s'...", in_dir);

  /* We use scandir() + alphasort() rather than readdir() because otherwise,
     the ordering  of test cases would vary somewhat randomly and would be
     difficult to control. */

  nl_cnt = scandir(in_dir, &nl, NULL, alphasort);

  if (nl_cnt < 0) {

    if (errno == ENOENT || errno == ENOTDIR)

      SAYF("\n" cLRD "[-] " cRST
           "The input directory does not seem to be valid - try again. The fuzzer needs\n"
           "    one or more test case to start with - ideally, a small file under 1 kB\n"
           "    or so. The cases must be stored as regular files directly in the input\n"
           "    directory.\n");

    PFATAL("Unable to open '%s'", in_dir);

  }

  for (i = 0; i < nl_cnt; i++) {

    struct stat st;

    u8* fn = alloc_printf("%s/%s", in_dir, nl[i]->d_name);
    u8* dfn = alloc_printf("%s/.state/deterministic_done/%s", in_dir, nl[i]->d_name);

    u8  passed_det = 0;

    free(nl[i]); /* not tracked */
 
    if (lstat(fn, &st) || access(fn, R_OK))
      PFATAL("Unable to access '%s'", fn);

    /* This also takes care of . and .. */

    if (!S_ISREG(st.st_mode) || !st.st_size || strstr(fn, "/README.txt")) {

      ck_free(fn);
      ck_free(dfn);
      continue;

    }

    if (st.st_size > MAX_FILE) 
      FATAL("Test case '%s' is too big (%s, limit is %s)", fn,
            DMS(st.st_size), DMS(MAX_FILE));

    /* Check for metadata that indicates that deterministic fuzzing
       is complete for this entry. We don't want to repeat deterministic
       fuzzing when resuming aborted scans, because it would be pointless
       and probably very time-consuming. */

    if (!access(dfn, F_OK)) passed_det = 1;
    ck_free(dfn);

    add_to_queue(fn, st.st_size, passed_det);

  }

  free(nl); /* not tracked */

  if (!queued_paths) {

    SAYF("\n" cLRD "[-] " cRST
         "Looks like there are no valid test cases in the input directory! The fuzzer\n"
         "    needs one or more test case to start with - ideally, a small file under\n"
         "    1 kB or so. The cases must be stored as regular files directly in the\n"
         "    input directory.\n");

    FATAL("No usable test cases in '%s'", in_dir);

  }

  last_path_time = 0;
  queued_at_start = queued_paths;

}


/* Helper function for load_extras. */

static int compare_extras_len(const void* p1, const void* p2) {
  struct extra_data *e1 = (struct extra_data*)p1,
                    *e2 = (struct extra_data*)p2;

  return e1->len - e2->len;
}

static int compare_extras_use_d(const void* p1, const void* p2) {
  struct extra_data *e1 = (struct extra_data*)p1,
                    *e2 = (struct extra_data*)p2;

  return e2->hit_cnt - e1->hit_cnt;
}


/* Read extras from a file, sort by size. */

static void load_extras_file(u8* fname, u32* min_len, u32* max_len,
                             u32 dict_level) {

  FILE* f;
  u8  buf[MAX_LINE];
  u8  *lptr;
  u32 cur_line = 0;

  f = fopen(fname, "r");

  if (!f) PFATAL("Unable to open '%s'", fname);

  while ((lptr = fgets(buf, MAX_LINE, f))) {

    u8 *rptr, *wptr;
    u32 klen = 0;

    cur_line++;

    /* Trim on left and right. */

    while (isspace(*lptr)) lptr++;

    rptr = lptr + strlen(lptr) - 1;
    while (rptr >= lptr && isspace(*rptr)) rptr--;
    rptr++;
    *rptr = 0;

    /* Skip empty lines and comments. */

    if (!*lptr || *lptr == '#') continue;

    /* All other lines must end with '"', which we can consume. */

    rptr--;

    if (rptr < lptr || *rptr != '"')
      FATAL("Malformed name=\"value\" pair in line %u.", cur_line);

    *rptr = 0;

    /* Skip alphanumerics and dashes (label). */

    while (isalnum(*lptr) || *lptr == '_') lptr++;

    /* If @number follows, parse that. */

    if (*lptr == '@') {

      lptr++;
      if (atoi(lptr) > dict_level) continue;
      while (isdigit(*lptr)) lptr++;

    }

    /* Skip whitespace and = signs. */

    while (isspace(*lptr) || *lptr == '=') lptr++;

    /* Consume opening '"'. */

    if (*lptr != '"')
      FATAL("Malformed name=\"keyword\" pair in line %u.", cur_line);

    lptr++;

    if (!*lptr) FATAL("Empty keyword in line %u.", cur_line);

    /* Okay, let's allocate memory and copy data between "...", handling
       \xNN escaping, \\, and \". */

    extras = ck_realloc_block(extras, (extras_cnt + 1) *
               sizeof(struct extra_data));

    wptr = extras[extras_cnt].data = ck_alloc(rptr - lptr);

    while (*lptr) {

      char* hexdigits = "0123456789abcdef";

      switch (*lptr) {

        case 1 ... 31:
        case 128 ... 255:
          FATAL("Non-printable characters in line %u.", cur_line);

        case '\\':

          lptr++;

          if (*lptr == '\\' || *lptr == '"') {
            *(wptr++) = *(lptr++);
            klen++;
            break;
          }

          if (*lptr != 'x' || !isxdigit(lptr[1]) || !isxdigit(lptr[2]))
            FATAL("Invalid escaping (not \\xNN) in line %u.", cur_line);

          *(wptr++) =
            ((strchr(hexdigits, tolower(lptr[1])) - hexdigits) << 4) |
            (strchr(hexdigits, tolower(lptr[2])) - hexdigits);

          lptr += 3;
          klen++;

          break;

        default:

          *(wptr++) = *(lptr++);
          klen++;

      }

    }

    extras[extras_cnt].len = klen;

    if (extras[extras_cnt].len > MAX_DICT_FILE)
      FATAL("Keyword too big in line %u (%s, limit is %s)", cur_line,
            DMS(klen), DMS(MAX_DICT_FILE));

    if (*min_len > klen) *min_len = klen;
    if (*max_len < klen) *max_len = klen;

    extras_cnt++;

  }

  fclose(f);

}


/* Read extras from the extras directory and sort them by size. */

static void load_extras(u8* dir) {

  DIR* d;
  struct dirent* de;
  u32 min_len = MAX_DICT_FILE, max_len = 0, dict_level = 0;
  u8* x;

  /* If the name ends with @, extract level and continue. */

  if ((x = strchr(dir, '@'))) {

    *x = 0;
    dict_level = atoi(x + 1);

  }

  ACTF("Loading extra dictionary from '%s' (level %u)...", dir, dict_level);

  d = opendir(dir);

  if (!d) {

    if (errno == ENOTDIR) {
      load_extras_file(dir, &min_len, &max_len, dict_level);
      goto check_and_sort;
    }

    PFATAL("Unable to open '%s'", dir);

  }

  if (x) FATAL("Dictinary levels not supported for directories.");

  while ((de = readdir(d))) {

    struct stat st;
    u8* fn = alloc_printf("%s/%s", dir, de->d_name);
    s32 fd;

    if (lstat(fn, &st) || access(fn, R_OK))
      PFATAL("Unable to access '%s'", fn);

    /* This also takes care of . and .. */
    if (!S_ISREG(st.st_mode) || !st.st_size) {

      ck_free(fn);
      continue;

    }

    if (st.st_size > MAX_DICT_FILE)
      FATAL("Extra '%s' is too big (%s, limit is %s)", fn,
            DMS(st.st_size), DMS(MAX_DICT_FILE));

    if (min_len > st.st_size) min_len = st.st_size;
    if (max_len < st.st_size) max_len = st.st_size;

    extras = ck_realloc_block(extras, (extras_cnt + 1) *
               sizeof(struct extra_data));

    extras[extras_cnt].data = ck_alloc(st.st_size);
    extras[extras_cnt].len  = st.st_size;

    fd = open(fn, O_RDONLY);

    if (fd < 0) PFATAL("Unable to open '%s'", fn);

    ck_read(fd, extras[extras_cnt].data, st.st_size, fn);

    close(fd);
    ck_free(fn);

    extras_cnt++;

  }

  closedir(d);

check_and_sort:

  if (!extras_cnt) FATAL("No usable files in '%s'", dir);

  qsort(extras, extras_cnt, sizeof(struct extra_data), compare_extras_len);

  OKF("Loaded %u extra tokens, size range %s to %s.", extras_cnt,
      DMS(min_len), DMS(max_len));

  if (max_len > 32)
    WARNF("Some tokens are relatively large (%s) - consider trimming.",
          DMS(max_len));

  if (extras_cnt > MAX_DET_EXTRAS)
    WARNF("More than %u tokens - will use them probabilistically.",
          MAX_DET_EXTRAS);

}




/* Helper function for maybe_add_auto() */

static inline u8 memcmp_nocase(u8* m1, u8* m2, u32 len) {

  while (len--) if (tolower(*(m1++)) ^ tolower(*(m2++))) return 1;
  return 0;

}


/* Maybe add automatic extra. */

static void maybe_add_auto(u8* mem, u32 len) {

  u32 i;

  /* Allow users to specify that they don't want auto dictionaries. */

  if (!MAX_AUTO_EXTRAS || !USE_AUTO_EXTRAS) return;

  /* Skip runs of identical bytes. */

  for (i = 1; i < len; i++)
    if (mem[0] ^ mem[i]) break;

  if (i == len) return;

  /* Reject builtin interesting values. */

  if (len == 2) {

    i = sizeof(interesting_16) >> 1;

    while (i--) 
      if (*((u16*)mem) == interesting_16[i] ||
          *((u16*)mem) == SWAP16(interesting_16[i])) return;

  }

  if (len == 4) {

    i = sizeof(interesting_32) >> 2;

    while (i--) 
      if (*((u32*)mem) == interesting_32[i] ||
          *((u32*)mem) == SWAP32(interesting_32[i])) return;

  }

  /* Reject anything that matches existing extras. Do a case-insensitive
     match. We optimize by exploiting the fact that extras[] are sorted
     by size. */

  for (i = 0; i < extras_cnt; i++)
    if (extras[i].len >= len) break;

  for (; i < extras_cnt && extras[i].len == len; i++)
    if (!memcmp_nocase(extras[i].data, mem, len)) return;

  /* Last but not least, check a_extras[] for matches. There are no
     guarantees of a particular sort order. */

  auto_changed = 1;

  for (i = 0; i < a_extras_cnt; i++) {

    if (a_extras[i].len == len && !memcmp_nocase(a_extras[i].data, mem, len)) {

      a_extras[i].hit_cnt++;
      goto sort_a_extras;

    }

  }

  /* At this point, looks like we're dealing with a new entry. So, let's
     append it if we have room. Otherwise, let's randomly evict some other
     entry from the bottom half of the list. */

  if (a_extras_cnt < MAX_AUTO_EXTRAS) {

    a_extras = ck_realloc_block(a_extras, (a_extras_cnt + 1) *
                                sizeof(struct extra_data));

    a_extras[a_extras_cnt].data = ck_memdup(mem, len);
    a_extras[a_extras_cnt].len  = len;
    a_extras_cnt++;

  } else {

    i = MAX_AUTO_EXTRAS / 2 +
        UR((MAX_AUTO_EXTRAS + 1) / 2);

    ck_free(a_extras[i].data);

    a_extras[i].data    = ck_memdup(mem, len);
    a_extras[i].len     = len;
    a_extras[i].hit_cnt = 0;

  }

sort_a_extras:

  /* First, sort all auto extras by use count, descending order. */

  qsort(a_extras, a_extras_cnt, sizeof(struct extra_data),
        compare_extras_use_d);

  /* Then, sort the top USE_AUTO_EXTRAS entries by size. */

  qsort(a_extras, MIN(USE_AUTO_EXTRAS, a_extras_cnt),
        sizeof(struct extra_data), compare_extras_len);

}


/* Save automatically generated extras. */

static void save_auto(void) {

  u32 i;

  if (!auto_changed) return;
  auto_changed = 0;

  for (i = 0; i < MIN(USE_AUTO_EXTRAS, a_extras_cnt); i++) {

    u8* fn = alloc_printf("%s/queue/.state/auto_extras/auto_%06u", out_dir, i);
    s32 fd;

    fd = open(fn, O_WRONLY | O_CREAT | O_TRUNC, 0600);

    if (fd < 0) PFATAL("Unable to create '%s'", fn);

    ck_write(fd, a_extras[i].data, a_extras[i].len, fn);

    close(fd);
    ck_free(fn);

  }

}


/* Load automatically generated extras. */

static void load_auto(void) {

  u32 i;

  for (i = 0; i < USE_AUTO_EXTRAS; i++) {

    u8  tmp[MAX_AUTO_EXTRA + 1];
    u8* fn = alloc_printf("%s/.state/auto_extras/auto_%06u", in_dir, i);
    s32 fd, len;

    fd = open(fn, O_RDONLY, 0600);

    if (fd < 0) {

      if (errno != ENOENT) PFATAL("Unable to open '%s'", fn);
      ck_free(fn);
      break;

    }

    /* We read one byte more to cheaply detect tokens that are too
       long (and skip them). */

    len = read(fd, tmp, MAX_AUTO_EXTRA + 1);

    if (len < 0) PFATAL("Unable to read from '%s'", fn);

    if (len >= MIN_AUTO_EXTRA && len <= MAX_AUTO_EXTRA)
      maybe_add_auto(tmp, len);

    close(fd);
    ck_free(fn);

  }

  if (i) OKF("Loaded %u auto-discovered dictionary tokens.", i);
  else OKF("No auto-generated dictionary tokens to reuse.");

}


/* Destroy extras. */

static void destroy_extras(void) {

  u32 i;

  for (i = 0; i < extras_cnt; i++) 
    ck_free(extras[i].data);

  ck_free(extras);

  for (i = 0; i < a_extras_cnt; i++) 
    ck_free(a_extras[i].data);

  ck_free(a_extras);

}


/* Spin up fork server (instrumented mode only). The idea is explained here:

   http://lcamtuf.blogspot.com/2014/10/fuzzing-binaries-without-execve.html

   In essence, the instrumentation allows us to skip execve(), and just keep
   cloning a stopped child. So, we just execute once, and then send commands
   through a pipe. The other part of this logic is in afl-as.h. */

static void init_forkserver(char** argv) {

  static struct itimerval it;
  int st_pipe[2], ctl_pipe[2];
  int status;
  s32 rlen;

  ACTF("Spinning up the fork server...");

  if (pipe(st_pipe) || pipe(ctl_pipe)) PFATAL("pipe() failed");

  forksrv_pid = fork();

  if (forksrv_pid < 0) PFATAL("fork() failed");

  if (!forksrv_pid) {

    struct rlimit r;
    munmap(trace_bits, MAP_SIZE);
    /* Umpf. On OpenBSD, the default fd limit for root users is set to
       soft 128. Let's try to fix that... */

    if (!getrlimit(RLIMIT_NOFILE, &r) && r.rlim_cur < FORKSRV_FD + 2) {

      r.rlim_cur = FORKSRV_FD + 2;
      setrlimit(RLIMIT_NOFILE, &r); /* Ignore errors */

    }

    if (mem_limit) {

      r.rlim_max = r.rlim_cur = ((rlim_t)mem_limit) << 20;

#ifdef RLIMIT_AS

      setrlimit(RLIMIT_AS, &r); /* Ignore errors */

#else

      /* This takes care of OpenBSD, which doesn't have RLIMIT_AS, but
         according to reliable sources, RLIMIT_DATA covers anonymous
         maps - so we should be getting good protection against OOM bugs. */

      setrlimit(RLIMIT_DATA, &r); /* Ignore errors */

#endif /* ^RLIMIT_AS */


    }

    /* Dumping cores is slow and can lead to anomalies if SIGKILL is delivered
       before the dump is complete. */

    r.rlim_max = r.rlim_cur = 0;

    setrlimit(RLIMIT_CORE, &r); /* Ignore errors */

    /* Isolate the process and configure standard descriptors. If out_file is
       specified, stdin is /dev/null; otherwise, out_fd is cloned instead. */

    setsid();

    dup2(dev_null_fd, 1);
    dup2(dev_null_fd, 2);

    if (out_file) {

      dup2(dev_null_fd, 0);

    } else {

      dup2(out_fd, 0);
      close(out_fd);

    }

    /* Set up control and status pipes, close the unneeded original fds. */

    if (dup2(ctl_pipe[0], FORKSRV_FD) < 0) PFATAL("dup2() failed");
    if (dup2(st_pipe[1], FORKSRV_FD + 1) < 0) PFATAL("dup2() failed");

    close(ctl_pipe[0]);
    close(ctl_pipe[1]);
    close(st_pipe[0]);
    close(st_pipe[1]);

    close(out_dir_fd);
    close(dev_null_fd);
    close(dev_urandom_fd);
    close(fileno(plot_file));

    /* This should improve performance a bit, since it stops the linker from
       doing extra work post-fork(). */

    if (!getenv("LD_BIND_LAZY")) setenv("LD_BIND_NOW", "1", 0);

    /* Set sane defaults for ASAN if nothing else specified. */

    setenv("ASAN_OPTIONS", "abort_on_error=1:"
                           "detect_leaks=0:"
                           "allocator_may_return_null=1", 0);

    /* MSAN is tricky, because it doesn't support abort_on_error=1 at this
       point. So, we do this in a very hacky way. */

    setenv("MSAN_OPTIONS", "exit_code=" STRINGIFY(MSAN_ERROR) ":"
                           "msan_track_origins=0", 0);

    execv(target_path, argv);

    /* Use a distinctive bitmap signature to tell the parent about execv()
       falling through. */

    *(u32*)trace_bits = EXEC_FAIL_SIG;
    exit(0);

  }

  /* Close the unneeded endpoints. */

  close(ctl_pipe[0]);
  close(st_pipe[1]);

  fsrv_ctl_fd = ctl_pipe[1];
  fsrv_st_fd  = st_pipe[0];

  /* Wait for the fork server to come up, but don't wait too long. */

  it.it_value.tv_sec = ((exec_tmout * FORK_WAIT_MULT) / 1000);
  it.it_value.tv_usec = ((exec_tmout * FORK_WAIT_MULT) % 1000) * 1000;

  setitimer(ITIMER_REAL, &it, NULL);

  rlen = read(fsrv_st_fd, &status, 4);

  it.it_value.tv_sec = 0;
  it.it_value.tv_usec = 0;

  setitimer(ITIMER_REAL, &it, NULL);

  /* If we have a four-byte "hello" message from the server, we're all set.
     Otherwise, try to figure out what went wrong. */

  if (rlen == 4) {
    OKF("All right - fork server is up.");
    return;
  }

  if (child_timed_out)
    FATAL("Timeout while initializing fork server (adjusting -t may help)");

  if (waitpid(forksrv_pid, &status, 0) <= 0)
    PFATAL("waitpid() failed");

  if (WIFSIGNALED(status)) {

    if (mem_limit && mem_limit < 500 && uses_asan) {

      SAYF("\n" cLRD "[-] " cRST
           "Whoops, the target binary crashed suddenly, before receiving any input\n"
           "    from the fuzzer! Since it seems to be built with ASAN and you have a\n"
           "    restrictive memory limit configured, this is expected; please read\n"
           "    %s/notes_for_asan.txt for help.\n", doc_path);

    } else if (!mem_limit) {

      SAYF("\n" cLRD "[-] " cRST
           "Whoops, the target binary crashed suddenly, before receiving any input\n"
           "    from the fuzzer! There are several probable explanations:\n\n"

           "    - The binary is just buggy and explodes entirely on its own. If so, you\n"
           "      need to fix the underlying problem or find a better replacement.\n\n"

#ifdef __APPLE__

           "    - On MacOS X, the semantics of fork() syscalls are non-standard and may\n"
           "      break afl-fuzz performance optimizations when running platform-specific\n"
           "      targets. To fix this, set AFL_NO_FORKSRV=1 in the environment.\n\n"

#endif /* __APPLE__ */

           "    - Less likely, there is a horrible bug in the fuzzer. If other options\n"
           "      fail, poke <lcamtuf@coredump.cx> for troubleshooting tips.\n");

    } else {

      SAYF("\n" cLRD "[-] " cRST
           "Whoops, the target binary crashed suddenly, before receiving any input\n"
           "    from the fuzzer! There are several probable explanations:\n\n"

           "    - The current memory limit (%s) is too restrictive, causing the\n"
           "      target to hit an OOM condition in the dynamic linker. Try bumping up\n"
           "      the limit with the -m setting in the command line. A simple way confirm\n"
           "      this diagnosis would be:\n\n"

#ifdef RLIMIT_AS
           "      ( ulimit -Sv $[%llu << 10]; /path/to/fuzzed_app )\n\n"
#else
           "      ( ulimit -Sd $[%llu << 10]; /path/to/fuzzed_app )\n\n"
#endif /* ^RLIMIT_AS */

           "      Tip: you can use http://jwilk.net/software/recidivm to quickly\n"
           "      estimate the required amount of virtual memory for the binary.\n\n"

           "    - The binary is just buggy and explodes entirely on its own. If so, you\n"
           "      need to fix the underlying problem or find a better replacement.\n\n"

#ifdef __APPLE__

           "    - On MacOS X, the semantics of fork() syscalls are non-standard and may\n"
           "      break afl-fuzz performance optimizations when running platform-specific\n"
           "      targets. To fix this, set AFL_NO_FORKSRV=1 in the environment.\n\n"

#endif /* __APPLE__ */

           "    - Less likely, there is a horrible bug in the fuzzer. If other options\n"
           "      fail, poke <lcamtuf@coredump.cx> for troubleshooting tips.\n",
           DMS(mem_limit << 20), mem_limit - 1);

    }

    FATAL("Fork server crashed with signal %d", WTERMSIG(status));

  }

  if (*(u32*)trace_bits == EXEC_FAIL_SIG)
    FATAL("Unable to execute target application ('%s')", argv[0]);

  if (mem_limit && mem_limit < 500 && uses_asan) {

    SAYF("\n" cLRD "[-] " cRST
           "Hmm, looks like the target binary terminated before we could complete a\n"
           "    handshake with the injected code. Since it seems to be built with ASAN and\n"
           "    you have a restrictive memory limit configured, this is expected; please\n"
           "    read %s/notes_for_asan.txt for help.\n", doc_path);

  } else if (!mem_limit) {

    SAYF("\n" cLRD "[-] " cRST
         "Hmm, looks like the target binary terminated before we could complete a\n"
         "    handshake with the injected code. Perhaps there is a horrible bug in the\n"
         "    fuzzer. Poke <lcamtuf@coredump.cx> for troubleshooting tips.\n");

  } else {

    SAYF("\n" cLRD "[-] " cRST
         "Hmm, looks like the target binary terminated before we could complete a\n"
         "    handshake with the injected code. There are %s probable explanations:\n\n"

         "%s"
         "    - The current memory limit (%s) is too restrictive, causing an OOM\n"
         "      fault in the dynamic linker. This can be fixed with the -m option. A\n"
         "      simple way to confirm the diagnosis may be:\n\n"

#ifdef RLIMIT_AS
         "      ( ulimit -Sv $[%llu << 10]; /path/to/fuzzed_app )\n\n"
#else
         "      ( ulimit -Sd $[%llu << 10]; /path/to/fuzzed_app )\n\n"
#endif /* ^RLIMIT_AS */

         "      Tip: you can use http://jwilk.net/software/recidivm to quickly\n"
         "      estimate the required amount of virtual memory for the binary.\n\n"

         "    - Less likely, there is a horrible bug in the fuzzer. If other options\n"
         "      fail, poke <lcamtuf@coredump.cx> for troubleshooting tips.\n",
         getenv(DEFER_ENV_VAR) ? "three" : "two",
         getenv(DEFER_ENV_VAR) ?
         "    - You are using deferred forkserver, but __AFL_INIT() is never\n"
         "      reached before the program terminates.\n\n" : "",
         DMS(mem_limit << 20), mem_limit - 1);

  }

  FATAL("Fork server handshake failed");

}


static void afl_disassoc_area(void)
{
  ioctl(fd_afl_area, AFL_CTL_DISASSOC_AREA);
}

static void afl_assoc_area(void)
{
  ioctl(fd_afl_area, AFL_CTL_ASSOC_AREA);
}

static void afl_manual_fork(char** argv)
{
  /* If we're running in "dumb" mode, we can't rely on the fork server
     logic compiled into the target program, so we will just keep calling
     execve(). There is a bit of code duplication between here and 
     init_forkserver(), but c'est la vie. */
  child_pid = fork();

  if (child_pid < 0)
    PFATAL("fork() failed");

  if (!child_pid) {
    struct rlimit r;

    if (mem_limit) {
      r.rlim_max = r.rlim_cur = ((rlim_t)mem_limit) << 20;
#ifdef RLIMIT_AS
      setrlimit(RLIMIT_AS, &r); /* Ignore errors */
#else
      setrlimit(RLIMIT_DATA, &r); /* Ignore errors */
#endif
    }

    r.rlim_max = r.rlim_cur = 0;
    setrlimit(RLIMIT_CORE, &r); /* Ignore errors */

    /* Isolate the process and configure standard descriptors. If out_file is
       specified, stdin is /dev/null; otherwise, out_fd is cloned instead. */

    setsid();

    dup2(dev_null_fd, 1);
    dup2(dev_null_fd, 2);

    if (out_file) {
      dup2(dev_null_fd, 0);
    } else {
      dup2(out_fd, 0);
      close(out_fd);
    }

    /* On Linux, would be faster to use O_CLOEXEC. Maybe TODO. */
    close(dev_null_fd);
    close(out_dir_fd);
    close(dev_urandom_fd);
    close(fileno(plot_file));

    /* Set sane defaults for ASAN if nothing else specified. */
    setenv("ASAN_OPTIONS", "abort_on_error=1:"
	   "detect_leaks=0:"
	   "allocator_may_return_null=1", 0);
    setenv("MSAN_OPTIONS", "exit_code=" STRINGIFY(MSAN_ERROR) ":"
	   "msan_track_origins=0", 0);

    afl_assoc_area();

    execv(target_path, argv);

    /* Use a distinctive bitmap value to tell the parent about execv()
       falling through. */
    *(u32*)trace_bits = EXEC_FAIL_SIG;
    exit(0);
  }
}

static int afl_forkserver_fork(void)
{
  s32 res;
  u32 go = 0;

  /* In non-dumb mode, we have the fork server up and running, so simply
     tell it to have at it, and then read back PID. */

  if ((res = write(fsrv_ctl_fd, &go, 4)) != 4) {
    if (stop_soon)
      return -1;
    RPFATAL(res, "Unable to request new process from fork server (OOM?)");
  }

  if ((res = read(fsrv_st_fd, &child_pid, 4)) != 4) {
    if (stop_soon)
      return -1;
    RPFATAL(res, "Unable to request new process from fork server (OOM?)");
  }

  if (child_pid <= 0)
    FATAL("Fork server is misbehaving (OOM?)");
  return 0;
}

int afl_fork(char** argv)
{
  if (dumb_mode == 1 || no_forkserver) {
    afl_manual_fork(argv);
    return 0;
  } else {
    return afl_forkserver_fork();
  }
}

void afl_manual_waitchild(int* status)
{
   if (waitpid(child_pid, status, 0) <= 0)
     PFATAL("waitpid() failed");
}

void afl_forkserver_waitchild(int* status)
{
  s32 res;

  if ((res = read(fsrv_st_fd, status, 4)) != 4) {
    if (stop_soon)
      return;
    RPFATAL(res, "Unable to communicate with fork server");
  }
}

void afl_waitchild(int* status)
{
  if (dumb_mode == 1 || no_forkserver) {
    afl_manual_waitchild(status);
  } else {
    afl_forkserver_waitchild(status);
  }
}

static void afl_clear_child_timer(void)
{
  static struct itimerval it;

  setitimer(ITIMER_REAL, &it, NULL);
}

/* Configure timeout, as requested by user, then wait for child to terminate. */
static void afl_set_child_timer(void)
{
  static struct itimerval it;
  it.it_value.tv_sec = (exec_tmout / 1000);
  it.it_value.tv_usec = (exec_tmout % 1000) * 1000;

  setitimer(ITIMER_REAL, &it, NULL);
  /* The SIGALRM handler simply kills the child_pid and sets child_timed_out. */
}

static void afl_sync_fs(void)
{
  static int should_sync = 0;

  if ((++should_sync % 4096) == 0) {
    sync();
    should_sync = 0;
  }
}

static u8 afl_report_child_status(int status, u32 tb4)
{
  /* Report outcome to caller. */

  if (child_timed_out) return FAULT_HANG;

  if (WIFSIGNALED(status) && !stop_soon) {
    kill_signal = WTERMSIG(status);
    return FAULT_CRASH;
  }

  /* A somewhat nasty hack for MSAN, which doesn't support abort_on_error and
     must use a special exit code. */

  if (uses_asan && WEXITSTATUS(status) == MSAN_ERROR) {
    kill_signal = 0;
    return FAULT_CRASH;
  }

  if ((dumb_mode == 1 || no_forkserver) && tb4 == EXEC_FAIL_SIG)
    return FAULT_ERROR;

  return FAULT_NONE;
}

/* Execute target application, monitoring for timeouts. Return status
   information. The called program will update trace_bits[]. */

static u8 run_target(char** argv)
{
  int status = 0;

  child_timed_out = 0;

  afl_sync_fs();

  if (afl_fork(argv) == -1)
    return 0;

  afl_set_child_timer();
  afl_waitchild(&status);
  afl_clear_child_timer();

  child_pid = 0;
  total_execs++;

  afl_disassoc_area();

  /* Any subsequent operations on trace_bits must not be moved by the
     compiler below this point. Past this location, trace_bits[] behave
     very normally and do not have to be treated as volatile. */
  MEM_BARRIER();

#ifdef __x86_64__
  classify_counts((u64*)trace_bits);
#else
  classify_counts((u32*)trace_bits);
#endif /* ^__x86_64__ */

  return afl_report_child_status(status, *(u32*)trace_bits);
}


/* Write modified data to file for testing. If out_file is set, the old file
   is unlinked and a new one is created. Otherwise, out_fd is rewound and
   truncated. */

static void write_to_testcase(void* mem, u32 len) {

  s32 fd = out_fd;

  if (out_file) {

    unlink(out_file); /* Ignore errors. */

    fd = open(out_file, O_WRONLY | O_CREAT | O_EXCL, 0600);

    if (fd < 0) PFATAL("Unable to create '%s'", out_file);

  } else lseek(fd, 0, SEEK_SET);

  ck_write(fd, mem, len, out_file);

  if (!out_file) {

    if (ftruncate(fd, len)) PFATAL("ftruncate() failed");
    lseek(fd, 0, SEEK_SET);

  } else close(fd);

}


/* The same, but with an adjustable gap. Used for trimming. */

static void write_with_gap(void* mem, u32 len, u32 skip_at, u32 skip_len) {

  s32 fd = out_fd;
  u32 tail_len = len - skip_at - skip_len;

  if (out_file) {

    unlink(out_file); /* Ignore errors. */

    fd = open(out_file, O_WRONLY | O_CREAT | O_EXCL, 0600);

    if (fd < 0) PFATAL("Unable to create '%s'", out_file);

  } else lseek(fd, 0, SEEK_SET);

  if (skip_at) ck_write(fd, mem, skip_at, out_file);

  if (tail_len) ck_write(fd, mem + skip_at + skip_len, tail_len, out_file);

  if (!out_file) {

    if (ftruncate(fd, len - skip_len)) PFATAL("ftruncate() failed");
    lseek(fd, 0, SEEK_SET);

  } else close(fd);

}


static void show_stats(void);

/* Calibrate a new test case. This is done when processing the input directory
   to warn about flaky or otherwise problematic test cases early on; and when
   new paths are discovered to detect variable behavior and so on. */

static u8 calibrate_case(char** argv, struct queue_entry* q, u8* use_mem,
                         u32 handicap, u8 from_queue) {

  u8  fault = 0, new_bits = 0, var_detected = 0, first_run = (q->exec_cksum == 0);
  u64 start_us, stop_us;

  s32 old_sc = stage_cur, old_sm = stage_max, old_tmout = exec_tmout;
  u8* old_sn = stage_name;

  /* Be a bit more generous about timeouts when resuming sessions, or when
     trying to calibrate already-added finds. This helps avoid trouble due
     to intermittent latency. */

  if (!from_queue || resuming_fuzz)
    exec_tmout = MAX(exec_tmout + CAL_TMOUT_ADD,
                     exec_tmout * CAL_TMOUT_PERC / 100);

  q->cal_failed++;

  stage_name = "calibration";
  stage_max  = no_var_check ? CAL_CYCLES_NO_VAR : CAL_CYCLES;

  /* Make sure the forkserver is up before we do anything, and let's not
     count its spin-up time toward binary calibration. */

  if (dumb_mode != 1 && !no_forkserver && !forksrv_pid)
    init_forkserver(argv);

  start_us = get_cur_time_us();

  for (stage_cur = 0; stage_cur < stage_max; stage_cur++) {

    u32 cksum;

    if (!first_run && !(stage_cur % stats_update_freq)) show_stats();

    write_to_testcase(use_mem, q->len);

    fault = run_target(argv);

    /* stop_soon is set by the handler for Ctrl+C. When it's pressed,
       we want to bail out quickly. */

    if (stop_soon || fault != crash_mode) goto abort_calibration;

    if (!dumb_mode && !stage_cur && !count_bytes(trace_bits)) {
      fault = FAULT_NOINST;
      goto abort_calibration;
    }

    cksum = hash32(trace_bits, MAP_SIZE, HASH_CONST);

    if (q->exec_cksum != cksum) {

      u8 hnb = has_new_bits(virgin_bits);
      if (hnb > new_bits) new_bits = hnb;

      if (!no_var_check && q->exec_cksum) {

        var_detected = 1;
        stage_max    = CAL_CYCLES_LONG;

      } else q->exec_cksum = cksum;

    }

  }

  stop_us = get_cur_time_us();

  total_cal_us     += stop_us - start_us;
  total_cal_cycles += stage_max;

  /* OK, let's collect some stats about the performance of this test case.
     This is used for fuzzing air time calculations in calculate_score(). */

  q->exec_us     = (stop_us - start_us) / stage_max;
  q->bitmap_size = count_bytes(trace_bits);
  q->handicap    = handicap;
  q->cal_failed  = 0;

  total_bitmap_size += q->bitmap_size;
  total_bitmap_entries++;

  update_bitmap_score(q);

  /* If this case didn't result in new output from the instrumentation, tell
     parent. This is a non-critical problem, but something to warn the user
     about. */

  if (!dumb_mode && first_run && !fault && !new_bits) fault = FAULT_NOBITS;

abort_calibration:

  if (new_bits == 2 && !q->has_new_cov) {
    q->has_new_cov = 1;
    queued_with_cov++;
  }

  /* Mark variable paths. */

  if (var_detected && !q->var_behavior) {
    mark_as_variable(q);
    queued_variable++;
  }

  stage_name = old_sn;
  stage_cur  = old_sc;
  stage_max  = old_sm;
  exec_tmout = old_tmout;

  if (!first_run) show_stats();

  return fault;

}


/* Examine map coverage. Called once, for first test case. */

static void check_map_coverage(void) {

  u32 i;

  if (count_bytes(trace_bits) < 100) return;

  for (i = (1 << (MAP_SIZE_POW2 - 1)); i < MAP_SIZE; i++)
    if (trace_bits[i]) return;

  WARNF("Recompile binary with newer version of afl to improve coverage!");

}


/* Perform dry run of all test cases to confirm that the app is working as
   expected. This is done only for the initial inputs, and only once. */

static void perform_dry_run(char** argv) {

  struct queue_entry* q = queue;
  u32 cal_failures = 0;
  u8* skip_crashes = getenv("AFL_SKIP_CRASHES");

  while (q) {

    u8* use_mem;
    u8  res;
    s32 fd;

    u8* fn = strrchr(q->fname, '/') + 1;

    ACTF("Attempting dry run with '%s'...", fn);

    fd = open(q->fname, O_RDONLY);
    if (fd < 0) PFATAL("Unable to open '%s'", q->fname);

    use_mem = ck_alloc_nozero(q->len);

    if (read(fd, use_mem, q->len) != q->len)
      FATAL("Short read from '%s'", q->fname);

    close(fd);

    res = calibrate_case(argv, q, use_mem, 0, 1);
    ck_free(use_mem);

    if (stop_soon) return;

    if (res == crash_mode || res == FAULT_NOBITS)
      SAYF(cGRA "    len = %u, map size = %u, exec speed = %llu us\n" cRST, 
           q->len, q->bitmap_size, q->exec_us);

    switch (res) {

      case FAULT_NONE:

        if (q == queue) check_map_coverage();

        if (crash_mode) FATAL("Test case '%s' does *NOT* crash", fn);

        break;

      case FAULT_HANG:

        if (timeout_given) {

          /* The -t nn+ syntax in the command line sets timeout_given to '2' and
             instructs afl-fuzz to tolerate but skip queue entries that time
             out. */

          if (timeout_given > 1) {
            WARNF("Test case results in a hang (skipping)");
            q->cal_failed = CAL_CHANCES;
            cal_failures++;
            break;
          }

          SAYF("\n" cLRD "[-] " cRST
               "The program took more than %u ms to process one of the initial test cases.\n"
               "    Usually, the right thing to do is to relax the -t option - or to delete it\n"
               "    altogether and allow the fuzzer to auto-calibrate. That said, if you know\n"
               "    what you are doing and want to simply skip the unruly test cases, append\n"
               "    '+' at the end of the value passed to -t ('-t %u+').\n", exec_tmout,
               exec_tmout);

          FATAL("Test case '%s' results in a hang", fn);

        } else {

          SAYF("\n" cLRD "[-] " cRST
               "The program took more than %u ms to process one of the initial test cases.\n"
               "    This is bad news; raising the limit with the -t option is possible, but\n"
               "    will probably make the fuzzing process extremely slow.\n\n"

               "    If this test case is just a fluke, the other option is to just avoid it\n"
               "    altogether, and find one that is less of a CPU hog.\n", exec_tmout);

          FATAL("Test case '%s' results in a hang", fn);

        }

      case FAULT_CRASH:  

        if (crash_mode) break;

        if (skip_crashes) {
          WARNF("Test case results in a crash (skipping)");
          q->cal_failed = CAL_CHANCES;
          cal_failures++;
          break;
        }

        if (mem_limit) {

          SAYF("\n" cLRD "[-] " cRST
               "Oops, the program crashed with one of the test cases provided. There are\n"
               "    several possible explanations:\n\n"

               "    - The test case causes known crashes under normal working conditions. If\n"
               "      so, please remove it. The fuzzer should be seeded with interesting\n"
               "      inputs - but not ones that cause an outright crash.\n\n"

               "    - The current memory limit (%s) is too low for this program, causing\n"
               "      it to die due to OOM when parsing valid files. To fix this, try\n"
               "      bumping it up with the -m setting in the command line. If in doubt,\n"
               "      try something along the lines of:\n\n"

#ifdef RLIMIT_AS
               "      ( ulimit -Sv $[%llu << 10]; /path/to/binary [...] <testcase )\n\n"
#else
               "      ( ulimit -Sd $[%llu << 10]; /path/to/binary [...] <testcase )\n\n"
#endif /* ^RLIMIT_AS */

               "      Tip: you can use http://jwilk.net/software/recidivm to quickly\n"
               "      estimate the required amount of virtual memory for the binary. Also,\n"
               "      if you are using ASAN, see %s/notes_for_asan.txt.\n\n"

#ifdef __APPLE__
  
               "    - On MacOS X, the semantics of fork() syscalls are non-standard and may\n"
               "      break afl-fuzz performance optimizations when running platform-specific\n"
               "      binaries. To fix this, set AFL_NO_FORKSRV=1 in the environment.\n\n"

#endif /* __APPLE__ */

               "    - Least likely, there is a horrible bug in the fuzzer. If other options\n"
               "      fail, poke <lcamtuf@coredump.cx> for troubleshooting tips.\n",
               DMS(mem_limit << 20), mem_limit - 1, doc_path);

        } else {

          SAYF("\n" cLRD "[-] " cRST
               "Oops, the program crashed with one of the test cases provided. There are\n"
               "    several possible explanations:\n\n"

               "    - The test case causes known crashes under normal working conditions. If\n"
               "      so, please remove it. The fuzzer should be seeded with interesting\n"
               "      inputs - but not ones that cause an outright crash.\n\n"

#ifdef __APPLE__
  
               "    - On MacOS X, the semantics of fork() syscalls are non-standard and may\n"
               "      break afl-fuzz performance optimizations when running platform-specific\n"
               "      binaries. To fix this, set AFL_NO_FORKSRV=1 in the environment.\n\n"

#endif /* __APPLE__ */

               "    - Least likely, there is a horrible bug in the fuzzer. If other options\n"
               "      fail, poke <lcamtuf@coredump.cx> for troubleshooting tips.\n");

        }

        FATAL("Test case '%s' results in a crash", fn);

      case FAULT_ERROR:

        FATAL("Unable to execute target application ('%s')", argv[0]);

      case FAULT_NOINST:

        FATAL("No instrumentation detected");

      case FAULT_NOBITS: 

        useless_at_start++;

        if (!in_bitmap)
          WARNF("No new instrumentation output, test case may be useless.");

        break;

    }

    if (q->var_behavior) WARNF("Instrumentation output varies across runs.");

    q = q->next;

  }

  if (cal_failures) {

    if (cal_failures == queued_paths)
      FATAL("All test cases time out%s, giving up!",
            skip_crashes ? " or crash" : "");

    WARNF("Skipped %u test cases (%0.02f%%) due to timeouts%s.", cal_failures,
          ((double)cal_failures) * 100 / queued_paths,
          skip_crashes ? " or crashes" : "");

    if (cal_failures * 5 > queued_paths)
      WARNF(cLRD "High percentage of rejected test cases, check settings!");

  }

  OKF("All test cases processed.");

}


/* Helper function: link() if possible, copy otherwise. */

static void link_or_copy(u8* old_path, u8* new_path) {

  s32 i = link(old_path, new_path);
  s32 sfd, dfd;
  u8* tmp;

  if (!i) return;

  sfd = open(old_path, O_RDONLY);
  if (sfd < 0) PFATAL("Unable to open '%s'", old_path);

  dfd = open(new_path, O_WRONLY | O_CREAT | O_EXCL, 0600);
  if (dfd < 0) PFATAL("Unable to create '%s'", new_path);

  tmp = ck_alloc(64 * 1024);

  while ((i = read(sfd, tmp, 64 * 1024)) > 0) 
    ck_write(dfd, tmp, i, new_path);

  if (i < 0) PFATAL("read() failed");

  ck_free(tmp);
  close(sfd);
  close(dfd);

}


static void nuke_resume_dir(void);

/* Create hard links for input test cases in the output directory, choosing
   good names and pivoting accordingly. */

static void pivot_inputs(void) {

  struct queue_entry* q = queue;
  u32 id = 0;

  ACTF("Creating hard links for all input files...");

  while (q) {

    u8  *nfn, *rsl = strrchr(q->fname, '/');
    u32 orig_id;

    if (!rsl) rsl = q->fname; else rsl++;

    /* If the original file name conforms to the syntax and the recorded
       ID matches the one we'd assign, just use the original file name.
       This is valuable for resuming fuzzing runs. */

#ifndef SIMPLE_FILES
#  define CASE_PREFIX "id:"
#else
#  define CASE_PREFIX "id_"
#endif /* ^!SIMPLE_FILES */

    if (!strncmp(rsl, CASE_PREFIX, 3) &&
        sscanf(rsl + 3, "%06u", &orig_id) == 1 && orig_id == id) {

      u8* src_str;
      u32 src_id;

      resuming_fuzz = 1;
      nfn = alloc_printf("%s/queue/%s", out_dir, rsl);

      /* Since we're at it, let's also try to find parent and figure out the
         appropriate depth for this entry. */

      src_str = strchr(rsl + 3, ':');

      if (src_str && sscanf(src_str + 1, "%06u", &src_id) == 1) {

        struct queue_entry* s = queue;
        while (src_id-- && s) s = s->next;
        if (s) q->depth = s->depth + 1;

        if (max_depth < q->depth) max_depth = q->depth;

      }

    } else {

      /* No dice - invent a new name, capturing the original one as a
         substring. */

#ifndef SIMPLE_FILES

      u8* use_name = strstr(rsl, ",orig:");

      if (use_name) use_name += 6; else use_name = rsl;
      nfn = alloc_printf("%s/queue/id:%06u,orig:%s", out_dir, id, use_name);

#else

      nfn = alloc_printf("%s/queue/id_%06u", out_dir, id);

#endif /* ^!SIMPLE_FILES */

    }

    /* Pivot to the new queue entry. */

    link_or_copy(q->fname, nfn);
    ck_free(q->fname);
    q->fname = nfn;

    /* Make sure that the passed_det value carries over, too. */

    if (q->passed_det) mark_as_det_done(q);

    q = q->next;
    id++;

  }

  if (in_place_resume) nuke_resume_dir();

}


#ifndef SIMPLE_FILES

/* Construct a file name for a new test case, capturing the operation
   that led to its discovery. Uses a static buffer. */

static u8* describe_op(u8 hnb) {

  static u8 ret[256];

  if (syncing_party) {

    sprintf(ret, "sync:%s,src:%06u", syncing_party, syncing_case);

  } else {

    sprintf(ret, "src:%06u", current_entry);

    if (splicing_with >= 0)
      sprintf(ret + strlen(ret), "+%06u", splicing_with);

    sprintf(ret + strlen(ret), ",op:%s", stage_short);

    if (stage_cur_byte >= 0) {

      sprintf(ret + strlen(ret), ",pos:%u", stage_cur_byte);

      if (stage_val_type != STAGE_VAL_NONE)
        sprintf(ret + strlen(ret), ",val:%s%+d", 
                (stage_val_type == STAGE_VAL_BE) ? "be:" : "",
                stage_cur_val);

    } else sprintf(ret + strlen(ret), ",rep:%u", stage_cur_val);

  }

  if (hnb == 2) strcat(ret, ",+cov");

  return ret;

}

#endif /* !SIMPLE_FILES */


/* Write a message accompanying the crash directory :-) */

static void write_crash_readme(void) {

  u8* fn = alloc_printf("%s/crashes/README.txt", out_dir);
  s32 fd;
  FILE* f;

  fd = open(fn, O_WRONLY | O_CREAT | O_EXCL, 0600);
  ck_free(fn);

  /* Do not die on errors here - that would be impolite. */

  if (fd < 0) return;

  f = fdopen(fd, "w");

  if (!f) {
    close(fd);
    return;
  }

  fprintf(f, "Command line used to find this crash:\n\n"

             "%s\n\n"

             "If you can't reproduce a bug outside of afl-fuzz, be sure to set the same\n"
             "memory limit. The limit used for this fuzzing session was %s.\n\n"

             "Need a tool to minimize test cases before investigating the crashes or sending\n"
             "them to a vendor? Check out the afl-tmin that comes with the fuzzer!\n\n"

             "Found any cool bugs in open-source tools using afl-fuzz? If yes, please drop\n"
             "me a mail at <lcamtuf@coredump.cx> once the issues are fixed - I'd love to\n"
             "add your finds to the gallery at:\n\n"

             "  http://lcamtuf.coredump.cx/afl/\n\n"

             "Thanks :-)\n",

             orig_cmdline, DMS(mem_limit << 20)); /* ignore errors */

  fclose(f);

}


/* Check if the result of an execve() during routine fuzzing is interesting,
   save or queue the input test case for further analysis if so. Returns 1 if
   entry is saved, 0 otherwise. */

static u8 save_if_interesting(char** argv, void* mem, u32 len, u8 fault) {

  u8  *fn = "";
  u8  hnb;
  s32 fd;
  u8  keeping = 0, res;

  if (fault == crash_mode) {

    /* Keep only if there are new bits in the map, add to queue for
       future fuzzing, etc. */

    if (!(hnb = has_new_bits(virgin_bits))) {
      if (crash_mode) total_crashes++;
      return 0;
    }    

#ifndef SIMPLE_FILES

    fn = alloc_printf("%s/queue/id:%06u,%s", out_dir, queued_paths,
                      describe_op(hnb));

#else

    fn = alloc_printf("%s/queue/id_%06u", out_dir, queued_paths);

#endif /* ^!SIMPLE_FILES */

    add_to_queue(fn, len, 0);

    if (hnb == 2) {
      queue_top->has_new_cov = 1;
      queued_with_cov++;
    }

    queue_top->exec_cksum = hash32(trace_bits, MAP_SIZE, HASH_CONST);

    /* Try to calibrate inline; this also calls update_bitmap_score() when
       successful. */

    res = calibrate_case(argv, queue_top, mem, queue_cycle - 1, 0);

    if (res == FAULT_ERROR)
      FATAL("Unable to execute target application");

    fd = open(fn, O_WRONLY | O_CREAT | O_EXCL, 0600);
    if (fd < 0) PFATAL("Unable to create '%s'", fn);
    ck_write(fd, mem, len, fn);
    close(fd);

    keeping = 1;

  }

  switch (fault) {

    case FAULT_HANG:

      /* Hangs are not very interesting, but we're still obliged to keep
         a handful of samples. We use the presence of new bits in the
         hang-specific bitmap as a signal of uniqueness. In "dumb" mode, we
         just keep everything. */

      total_hangs++;

      if (unique_hangs >= KEEP_UNIQUE_HANG) return keeping;

      if (!dumb_mode) {

#ifdef __x86_64__
        simplify_trace((u64*)trace_bits);
#else
        simplify_trace((u32*)trace_bits);
#endif /* ^__x86_64__ */

        if (!has_new_bits(virgin_hang)) return keeping;

      }

#ifndef SIMPLE_FILES

      fn = alloc_printf("%s/hangs/id:%06llu,%s", out_dir,
                        unique_hangs, describe_op(0));

#else

      fn = alloc_printf("%s/hangs/id_%06llu", out_dir,
                        unique_hangs);

#endif /* ^!SIMPLE_FILES */

      unique_hangs++;

      last_hang_time = get_cur_time();

      break;

    case FAULT_CRASH:

      /* This is handled in a manner roughly similar to hangs,
         except for slightly different limits. */

      total_crashes++;

      if (unique_crashes >= KEEP_UNIQUE_CRASH) return keeping;

      if (!dumb_mode) {

#ifdef __x86_64__
        simplify_trace((u64*)trace_bits);
#else
        simplify_trace((u32*)trace_bits);
#endif /* ^__x86_64__ */

        if (!has_new_bits(virgin_crash)) return keeping;

      }

      if (!unique_crashes) write_crash_readme();

#ifndef SIMPLE_FILES

      fn = alloc_printf("%s/crashes/id:%06llu,sig:%02u,%s", out_dir,
                        unique_crashes, kill_signal, describe_op(0));

#else

      fn = alloc_printf("%s/crashes/id_%06llu_%02u", out_dir, unique_crashes,
                        kill_signal);

#endif /* ^!SIMPLE_FILES */

      unique_crashes++;

      last_crash_time = get_cur_time();

      break;

    case FAULT_ERROR: FATAL("Unable to execute target application");

    default: return keeping;

  }

  /* If we're here, we apparently want to save the crash or hang
     test case, too. */

  fd = open(fn, O_WRONLY | O_CREAT | O_EXCL, 0600);
  if (fd < 0) PFATAL("Unable to create '%s'", fn);
  ck_write(fd, mem, len, fn);
  close(fd);

  ck_free(fn);

  return keeping;

}


/* When resuming, try to find the queue position to start from. This makes sense
   only when resuming, and when we can find the original fuzzer_stats. */

static u32 find_start_position(void) {

  static u8 tmp[4096]; /* Ought to be enough for anybody. */

  u8  *fn, *off;
  s32 fd, i;
  u32 ret;

  if (!resuming_fuzz) return 0;

  if (in_place_resume) fn = alloc_printf("%s/fuzzer_stats", out_dir);
  else fn = alloc_printf("%s/../fuzzer_stats", in_dir);

  fd = open(fn, O_RDONLY);
  ck_free(fn);

  if (fd < 0) return 0;

  i = read(fd, tmp, sizeof(tmp) - 1); (void)i; /* Ignore errors */
  close(fd);

  off = strstr(tmp, "cur_path       : ");
  if (!off) return 0;

  ret = atoi(off + 17);
  if (ret >= queued_paths) ret = 0;
  return ret;

}


/* The same, but for timeouts. The idea is that when resuming sessions without
   -t given, we don't want to keep auto-scaling the timeout over and over
   again to prevent it from growing due to random flukes. */

static void find_timeout(void) {

  static u8 tmp[4096]; /* Ought to be enough for anybody. */

  u8  *fn, *off;
  s32 fd, i;
  u32 ret;

  if (!resuming_fuzz) return;

  if (in_place_resume) fn = alloc_printf("%s/fuzzer_stats", out_dir);
  else fn = alloc_printf("%s/../fuzzer_stats", in_dir);

  fd = open(fn, O_RDONLY);
  ck_free(fn);

  if (fd < 0) return;

  i = read(fd, tmp, sizeof(tmp) - 1); (void)i; /* Ignore errors */
  close(fd);

  off = strstr(tmp, "exec_timeout   : ");
  if (!off) return;

  ret = atoi(off + 17);
  if (ret <= 4) return;

  exec_tmout = ret;
  timeout_given = 3;

}


/* Update stats file for unattended monitoring. */

static void write_stats_file(double bitmap_cvg, double eps) {

  static double last_bcvg, last_eps;

  u8* fn = alloc_printf("%s/fuzzer_stats", out_dir);
  s32 fd;
  FILE* f;

  fd = open(fn, O_WRONLY | O_CREAT | O_TRUNC, 0600);

  if (fd < 0) PFATAL("Unable to create '%s'", fn);

  ck_free(fn);

  f = fdopen(fd, "w");

  if (!f) PFATAL("fdopen() failed");

  /* Keep last values in case we're called from another context
     where exec/sec stats and such are not readily available. */

  if (!bitmap_cvg && !eps) {
    bitmap_cvg = last_bcvg;
    eps        = last_eps;
  } else {
    last_bcvg = bitmap_cvg;
    last_eps  = eps;
  }

  fprintf(f, "start_time     : %llu\n"
             "last_update    : %llu\n"
             "fuzzer_pid     : %u\n"
             "cycles_done    : %llu\n"
             "execs_done     : %llu\n"
             "execs_per_sec  : %0.02f\n"
             "paths_total    : %u\n"
             "paths_favored  : %u\n"
             "paths_found    : %u\n"
             "paths_imported : %u\n"
             "max_depth      : %u\n"
             "cur_path       : %u\n"
             "pending_favs   : %u\n"
             "pending_total  : %u\n"
             "variable_paths : %u\n"
             "bitmap_cvg     : %0.02f%%\n"
             "unique_crashes : %llu\n"
             "unique_hangs   : %llu\n"
             "last_path      : %llu\n"
             "last_crash     : %llu\n"
             "last_hang      : %llu\n"
             "exec_timeout   : %u\n"
             "afl_banner     : %s\n"
             "afl_version    : " VERSION "\n"
             "command_line   : %s\n",
             start_time / 1000, get_cur_time() / 1000, getpid(),
             queue_cycle ? (queue_cycle - 1) : 0, total_execs, eps,
             queued_paths, queued_favored, queued_discovered, queued_imported,
             max_depth, current_entry, pending_favored, pending_not_fuzzed,
             queued_variable, bitmap_cvg, unique_crashes, unique_hangs,
             last_path_time / 1000, last_crash_time / 1000,
             last_hang_time / 1000, exec_tmout, use_banner, orig_cmdline);
             /* ignore errors */

  fclose(f);

}


/* Update the plot file if there is a reason to. */

static void maybe_update_plot_file(double bitmap_cvg, double eps) {

  static u32 prev_qp, prev_pf, prev_pnf, prev_ce, prev_md;
  static u64 prev_qc, prev_uc, prev_uh;

  if (prev_qp == queued_paths && prev_pf == pending_favored && 
      prev_pnf == pending_not_fuzzed && prev_ce == current_entry &&
      prev_qc == queue_cycle && prev_uc == unique_crashes &&
      prev_uh == unique_hangs && prev_md == max_depth) return;

  prev_qp  = queued_paths;
  prev_pf  = pending_favored;
  prev_pnf = pending_not_fuzzed;
  prev_ce  = current_entry;
  prev_qc  = queue_cycle;
  prev_uc  = unique_crashes;
  prev_uh  = unique_hangs;
  prev_md  = max_depth;

  /* Fields in the file:

     unix_time, cycles_done, cur_path, paths_total, paths_not_fuzzed,
     favored_not_fuzzed, unique_crashes, unique_hangs, max_depth,
     execs_per_sec */

  fprintf(plot_file, 
          "%llu, %llu, %u, %u, %u, %u, %0.02f%%, %llu, %llu, %u, %0.02f\n",
          get_cur_time() / 1000, queue_cycle - 1, current_entry, queued_paths,
          pending_not_fuzzed, pending_favored, bitmap_cvg, unique_crashes,
          unique_hangs, max_depth, eps); /* ignore errors */

  fflush(plot_file);

}



/* A helper function for maybe_delete_out_dir(), deleting all prefixed
   files in a directory. */

static u8 delete_files(u8* path, u8* prefix) {

  DIR* d;
  struct dirent* d_ent;

  d = opendir(path);

  if (!d) return 0;

  while ((d_ent = readdir(d))) {

    if (d_ent->d_name[0] != '.' && (!prefix ||
        !strncmp(d_ent->d_name, prefix, strlen(prefix)))) {

      u8* fname = alloc_printf("%s/%s", path, d_ent->d_name);
      if (unlink(fname)) PFATAL("Unable to delete '%s'", fname);
      ck_free(fname);

    }

  }

  closedir(d);

  return !!rmdir(path);

}


/* Get the number of runnable processes, with some simple smoothing. */

static double get_runnable_processes(void) {

  static double res;

#if defined(__APPLE__) || defined(__FreeBSD__) || defined (__OpenBSD__)

  /* I don't see any portable sysctl or so that would quickly give us the
     number of runnable processes; the 1-minute load average can be a
     semi-decent approximation, though. */

  if (getloadavg(&res, 1) != 1) return 0;

#else

  /* On Linux, /proc/stat is probably the best way; load averages are
     computed in funny ways and sometimes don't reflect extremely short-lived
     processes well. */

  FILE* f = fopen("/proc/stat", "r");
  u8 tmp[1024];
  u32 val = 0;

  if (!f) return 0;

  while (fgets(tmp, sizeof(tmp), f)) {

    if (!strncmp(tmp, "procs_running ", 14) ||
        !strncmp(tmp, "procs_blocked ", 14)) val += atoi(tmp + 14);

  }
 
  fclose(f);

  if (!res) {

    res = val;

  } else {

    res = res * (1.0 - 1.0 / AVG_SMOOTHING) +
          ((double)val) * (1.0 / AVG_SMOOTHING);

  }

#endif /* ^(__APPLE__ || __FreeBSD__ || __OpenBSD__) */

  return res;

}


/* Delete the temporary directory used for in-place session resume. */

static void nuke_resume_dir(void) {

  u8* fn;

  fn = alloc_printf("%s/_resume/.state/deterministic_done", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/_resume/.state/auto_extras", out_dir);
  if (delete_files(fn, "auto_")) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/_resume/.state/redundant_edges", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/_resume/.state/variable_behavior", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/_resume/.state", out_dir);
  if (rmdir(fn) && errno != ENOENT) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/_resume", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  return;

dir_cleanup_failed:

  FATAL("_resume directory cleanup failed");

}


/* Delete fuzzer output directory if we recognize it as ours, if the fuzzer
   is not currently running, and if the last run time isn't too great. */

static void maybe_delete_out_dir(void) {

  FILE* f;
  u8 *fn = alloc_printf("%s/fuzzer_stats", out_dir);

  /* See if the output directory is locked. If yes, bail out. If not,
     create a lock that will persist for the lifetime of the process
     (this requires leaving the descriptor open).*/

  out_dir_fd = open(out_dir, O_RDONLY);
  if (out_dir_fd < 0) PFATAL("Unable to open '%s'", out_dir);

#ifndef __sun

  if (flock(out_dir_fd, LOCK_EX | LOCK_NB) && errno == EWOULDBLOCK) {

    SAYF("\n" cLRD "[-] " cRST
         "Looks like the job output directory is being actively used by another\n"
         "    instance of afl-fuzz. You will need to choose a different %s\n"
         "    or stop the other process first.\n",
         sync_id ? "fuzzer ID" : "output location");

    FATAL("Directory '%s' is in use", out_dir);

  }

#endif /* !__sun */

  f = fopen(fn, "r");

  if (f) {

    u64 start_time, last_update;

    if (fscanf(f, "start_time     : %llu\n"
                  "last_update    : %llu\n", &start_time, &last_update) != 2)
      FATAL("Malformed data in '%s'", fn);

    fclose(f);

    /* Let's see how much work is at stake. */

    if (!in_place_resume && last_update - start_time > OUTPUT_GRACE * 60) {

      SAYF("\n" cLRD "[-] " cRST
           "The job output directory already exists and contains the results of more\n"
           "    than %u minutes worth of fuzzing. To avoid data loss, afl-fuzz will *NOT*\n"
           "    automatically delete this data for you.\n\n"

           "    If you wish to start a new session, remove or rename the directory manually,\n"
           "    or specify a different output location for this job. To resume the old\n"
           "    session, put '-' as the input directory in the command line ('-i -') and\n"
           "    try again.\n", OUTPUT_GRACE);

       FATAL("At-risk data found in in '%s'", out_dir);

    }

  }

  ck_free(fn);

  /* The idea for in-place resume is pretty simple: we temporarily move the old
     queue/ to a new location that gets deleted once import to the new queue/
     is finished. If _resume/ already exists, the current queue/ may be
     incomplete due to an earlier abort, so we want to use the old _resume/
     dir instead, and we let rename() fail silently. */

  if (in_place_resume) {

    u8* orig_q = alloc_printf("%s/queue", out_dir);

    in_dir = alloc_printf("%s/_resume", out_dir);

    rename(orig_q, in_dir); /* Ignore errors */

    OKF("Output directory exists, will attempt session resume.");

    ck_free(orig_q);

  } else {

    OKF("Output directory exists but deemed OK to reuse.");

  }

  ACTF("Deleting old session data...");

  /* Okay, let's get the ball rolling! First, we need to get rid of the entries
     in <out_dir>/.synced/.../id:*, if any are present. */

  fn = alloc_printf("%s/.synced", out_dir);
  if (delete_files(fn, NULL)) goto dir_cleanup_failed;
  ck_free(fn);

  /* Next, we need to clean up <out_dir>/queue/.state/ subdirectories: */

  fn = alloc_printf("%s/queue/.state/deterministic_done", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/queue/.state/auto_extras", out_dir);
  if (delete_files(fn, "auto_")) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/queue/.state/redundant_edges", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/queue/.state/variable_behavior", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  /* Then, get rid of the .state subdirectory itself (should be empty by now)
     and everything matching <out_dir>/queue/id:*. */

  fn = alloc_printf("%s/queue/.state", out_dir);
  if (rmdir(fn) && errno != ENOENT) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/queue", out_dir);
  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  /* All right, let's do <out_dir>/crashes/id:* and <out_dir>/hangs/id:*. */

  if (!in_place_resume) {

    fn = alloc_printf("%s/crashes/README.txt", out_dir);
    unlink(fn); /* Ignore errors */
    ck_free(fn);

  }

  fn = alloc_printf("%s/crashes", out_dir);

  /* Make backup of the crashes directory if it's not empty and if we're
     doing in-place resume. */

  if (in_place_resume && rmdir(fn)) {

    time_t cur_t = time(0);
    struct tm* t = localtime(&cur_t);

#ifndef SIMPLE_FILES

    u8* nfn = alloc_printf("%s.%04u-%02u-%02u-%02u:%02u:%02u", fn,
                           t->tm_year + 1900, t->tm_mon + 1, t->tm_mday,
                           t->tm_hour, t->tm_min, t->tm_sec);

#else

    u8* nfn = alloc_printf("%s_%04u%02u%02u%02u%02u%02u", fn,
                           t->tm_year + 1900, t->tm_mon + 1, t->tm_mday,
                           t->tm_hour, t->tm_min, t->tm_sec);

#endif /* ^!SIMPLE_FILES */

    rename(fn, nfn); /* Ignore errors. */
    ck_free(nfn);

  }

  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/hangs", out_dir);

  /* Backup hangs, too. */

  if (in_place_resume && rmdir(fn)) {

    time_t cur_t = time(0);
    struct tm* t = localtime(&cur_t);

#ifndef SIMPLE_FILES

    u8* nfn = alloc_printf("%s.%04u-%02u-%02u-%02u:%02u:%02u", fn,
                           t->tm_year + 1900, t->tm_mon + 1, t->tm_mday,
                           t->tm_hour, t->tm_min, t->tm_sec);

#else

    u8* nfn = alloc_printf("%s_%04u%02u%02u%02u%02u%02u", fn,
                           t->tm_year + 1900, t->tm_mon + 1, t->tm_mday,
                           t->tm_hour, t->tm_min, t->tm_sec);

#endif /* ^!SIMPLE_FILES */

    rename(fn, nfn); /* Ignore errors. */
    ck_free(nfn);

  }

  if (delete_files(fn, CASE_PREFIX)) goto dir_cleanup_failed;
  ck_free(fn);

  /* And now, for some finishing touches. */

  fn = alloc_printf("%s/.cur_input", out_dir);
  if (unlink(fn) && errno != ENOENT) goto dir_cleanup_failed;
  ck_free(fn);

  fn = alloc_printf("%s/fuzz_bitmap", out_dir);
  if (unlink(fn) && errno != ENOENT) goto dir_cleanup_failed;
  ck_free(fn);

  if (!in_place_resume) {
    fn  = alloc_printf("%s/fuzzer_stats", out_dir);
    if (unlink(fn) && errno != ENOENT) goto dir_cleanup_failed;
    ck_free(fn);
  }

  fn = alloc_printf("%s/plot_data", out_dir);
  if (unlink(fn) && errno != ENOENT) goto dir_cleanup_failed;
  ck_free(fn);

  OKF("Output dir cleanup successful.");

  /* Wow... is that all? If yes, celebrate! */

  return;

dir_cleanup_failed:

  SAYF("\n" cLRD "[-] " cRST
       "Whoops, the fuzzer tried to reuse your output directory, but bumped into\n"
       "    some files that shouldn't be there or that couldn't be removed - so it\n"
       "    decided to abort! This happened while processing this path:\n\n"

       "    %s\n\n"
       "    Please examine and manually delete the files, or specify a different\n"
       "    output location for the tool.\n", fn);

  FATAL("Output directory cleanup failed");

}


static void check_term_size(void);


/* A spiffy retro stats screen! This is called every stats_update_freq
   execve() calls, plus in several other circumstances. */

static void show_stats(void) {

  static u64 last_stats_ms, last_plot_ms, last_ms, last_execs;
  static double avg_exec;
  double t_byte_ratio;

  u64 cur_ms;
  u32 t_bytes, t_bits;

  u32 banner_len, banner_pad;
  u8  tmp[256];

  cur_ms = get_cur_time();

  /* If not enough time has passed since last UI update, bail out. */

  if (cur_ms - last_ms < 1000 / UI_TARGET_HZ) return;

  /* Check if we're past the 10 minute mark. */

  if (cur_ms - start_time > 10 * 60 * 1000) run_over10m = 1;

  /* Calculate smoothed exec speed stats. */

  if (!last_execs) {
  
    avg_exec = ((double)total_execs) * 1000 / (cur_ms - start_time);

  } else {

    double cur_avg = ((double)(total_execs - last_execs)) * 1000 /
                     (cur_ms - last_ms);

    /* If there is a dramatic (5x+) jump in speed, reset the indicator
       more quickly. */

    if (cur_avg * 5 < avg_exec || cur_avg / 5 > avg_exec)
      avg_exec = cur_avg;

    avg_exec = avg_exec * (1.0 - 1.0 / AVG_SMOOTHING) +
               cur_avg * (1.0 / AVG_SMOOTHING);

  }

  last_ms = cur_ms;
  last_execs = total_execs;

  /* Tell the callers when to contact us (as measured in execs). */

  stats_update_freq = avg_exec / (UI_TARGET_HZ * 10);
  if (!stats_update_freq) stats_update_freq = 1;

  /* Do some bitmap stats. */

  t_bytes = count_non_255_bytes(virgin_bits);
  t_byte_ratio = ((double)t_bytes * 100) / MAP_SIZE;

  /* Roughly every minute, update fuzzer stats and save auto tokens. */

  if (cur_ms - last_stats_ms > STATS_UPDATE_SEC * 1000) {

    last_stats_ms = cur_ms;
    write_stats_file(t_byte_ratio, avg_exec);
    save_auto();
    write_bitmap();

  }

  /* Every now and then, write plot data. */

  if (cur_ms - last_plot_ms > PLOT_UPDATE_SEC * 1000) {

    last_plot_ms = cur_ms;
    maybe_update_plot_file(t_byte_ratio, avg_exec);
 
  }

  /* Honor AFL_EXIT_WHEN_DONE. */

  if (!dumb_mode && cycles_wo_finds > 20 && !pending_not_fuzzed &&
      getenv("AFL_EXIT_WHEN_DONE")) stop_soon = 1;

  /* If we're not on TTY, bail out. */

  if (not_on_tty) return;

  /* Compute some mildly useful bitmap stats. */

  t_bits = (MAP_SIZE << 3) - count_bits(virgin_bits);

  /* Now, for the visuals... */

  if (clear_screen) {

    SAYF(TERM_CLEAR CURSOR_HIDE);
    clear_screen = 0;

    check_term_size();

  }

  SAYF(TERM_HOME);

  if (term_too_small) {

    SAYF(cBRI "Your terminal is too small to display the UI.\n"
         "Please resize terminal window to at least 80x25.\n" cNOR);

    return;

  }

  /* Let's start by drawing a centered banner. */

  banner_len = (crash_mode ? 24 : 22) + strlen(VERSION) + strlen(use_banner);
  banner_pad = (80 - banner_len) / 2;
  memset(tmp, ' ', banner_pad);

  sprintf(tmp + banner_pad, "%s " cLCY VERSION cLGN
          " (%s)",  crash_mode ? cPIN "peruvian were-rabbit" : 
          cYEL "american fuzzy lop", use_banner);

  SAYF("\n%s\n\n", tmp);

  /* "Handy" shortcuts for drawing boxes... */

#define bSTG    bSTART cGRA
#define bH2     bH bH
#define bH5     bH2 bH2 bH
#define bH10    bH5 bH5
#define bH20    bH10 bH10
#define bH30    bH20 bH10
#define SP5     "     "
#define SP10    SP5 SP5
#define SP20    SP10 SP10

  /* Lord, forgive me this. */

  SAYF(SET_G1 bSTG bLT bH bSTOP cCYA " process timing " bSTG bH30 bH5 bH2 bHB
       bH bSTOP cCYA " overall results " bSTG bH5 bRT "\n");

  if (dumb_mode) {

    strcpy(tmp, cNOR);

  } else {

    /* First queue cycle: don't stop now! */
    if (queue_cycle == 1) strcpy(tmp, cMGN); else

    /* Subsequent cycles, but we're still making finds. */
    if (cycles_wo_finds < 3) strcpy(tmp, cYEL); else

    /* No finds for a long time and no test cases to try. */
    if (cycles_wo_finds > 20 && !pending_not_fuzzed) strcpy(tmp, cLGN);

    /* Default: cautiously OK to stop? */
    else strcpy(tmp, cLBL);

  }

  SAYF(bV bSTOP "        run time : " cNOR "%-34s " bSTG bV bSTOP
       "  cycles done : %s%-5s  " bSTG bV "\n",
       DTD(cur_ms, start_time), tmp, DI(queue_cycle - 1));

  /* We want to warn people about not seeing new paths after a full cycle,
     except when resuming fuzzing or running in non-instrumented mode. */

  if (!dumb_mode && (last_path_time || resuming_fuzz || queue_cycle == 1 ||
      in_bitmap || crash_mode)) {

    SAYF(bV bSTOP "   last new path : " cNOR "%-34s ",
         DTD(cur_ms, last_path_time));

  } else {

    if (dumb_mode)

      SAYF(bV bSTOP "   last new path : " cPIN "n/a" cNOR 
           " (non-instrumented mode)        ");

     else

      SAYF(bV bSTOP "   last new path : " cNOR "none yet " cLRD
           "(odd, check syntax!)      ");

  }

  SAYF(bSTG bV bSTOP "  total paths : " cNOR "%-5s  " bSTG bV "\n",
       DI(queued_paths));

  /* Highlight crashes in red if found, denote going over the KEEP_UNIQUE_CRASH
     limit with a '+' appended to the count. */

  sprintf(tmp, "%s%s", DI(unique_crashes),
          (unique_crashes >= KEEP_UNIQUE_CRASH) ? "+" : "");

  SAYF(bV bSTOP " last uniq crash : " cNOR "%-34s " bSTG bV bSTOP
       " uniq crashes : %s%-6s " bSTG bV "\n",
       DTD(cur_ms, last_crash_time), unique_crashes ? cLRD : cNOR,
       tmp);

  sprintf(tmp, "%s%s", DI(unique_hangs),
         (unique_hangs >= KEEP_UNIQUE_HANG) ? "+" : "");

  SAYF(bV bSTOP "  last uniq hang : " cNOR "%-34s " bSTG bV bSTOP 
       "   uniq hangs : " cNOR "%-6s " bSTG bV "\n",
       DTD(cur_ms, last_hang_time), tmp);

  SAYF(bVR bH bSTOP cCYA " cycle progress " bSTG bH20 bHB bH bSTOP cCYA
       " map coverage " bSTG bH bHT bH20 bH2 bH bVL "\n");

  /* This gets funny because we want to print several variable-length variables
     together, but then cram them into a fixed-width field - so we need to
     put them in a temporary buffer first. */

  sprintf(tmp, "%s%s (%0.02f%%)", DI(current_entry),
          queue_cur->favored ? "" : "*",
          ((double)current_entry * 100) / queued_paths);

  SAYF(bV bSTOP "  now processing : " cNOR "%-17s " bSTG bV bSTOP, tmp);


  sprintf(tmp, "%s (%0.02f%%)", DI(t_bytes), t_byte_ratio);

  SAYF("    map density : %s%-21s " bSTG bV "\n", t_byte_ratio > 70 ? cLRD : 
       ((t_bytes < 200 && !dumb_mode) ? cPIN : cNOR), tmp);

  sprintf(tmp, "%s (%0.02f%%)", DI(cur_skipped_paths),
          ((double)cur_skipped_paths * 100) / queued_paths);

  SAYF(bV bSTOP " paths timed out : " cNOR "%-17s " bSTG bV, tmp);

  sprintf(tmp, "%0.02f bits/tuple",
          t_bytes ? (((double)t_bits) / t_bytes) : 0);

  SAYF(bSTOP " count coverage : " cNOR "%-21s " bSTG bV "\n", tmp);

  SAYF(bVR bH bSTOP cCYA " stage progress " bSTG bH20 bX bH bSTOP cCYA
       " findings in depth " bSTG bH20 bVL "\n");

  sprintf(tmp, "%s (%0.02f%%)", DI(queued_favored),
          ((double)queued_favored) * 100 / queued_paths);

  /* Yeah... it's still going on... halp? */

  SAYF(bV bSTOP "  now trying : " cNOR "%-21s " bSTG bV bSTOP 
       " favored paths : " cNOR "%-22s " bSTG bV "\n", stage_name, tmp);

  if (!stage_max) {

    sprintf(tmp, "%s/-", DI(stage_cur));

  } else {

    sprintf(tmp, "%s/%s (%0.02f%%)", DI(stage_cur), DI(stage_max),
            ((double)stage_cur) * 100 / stage_max);

  }

  SAYF(bV bSTOP " stage execs : " cNOR "%-21s " bSTG bV bSTOP, tmp);

  sprintf(tmp, "%s (%0.02f%%)", DI(queued_with_cov),
          ((double)queued_with_cov) * 100 / queued_paths);

  SAYF("  new edges on : " cNOR "%-22s " bSTG bV "\n", tmp);

  sprintf(tmp, "%s (%s%s unique)", DI(total_crashes), DI(unique_crashes),
          (unique_crashes >= KEEP_UNIQUE_CRASH) ? "+" : "");

  if (crash_mode) {

    SAYF(bV bSTOP " total execs : " cNOR "%-21s " bSTG bV bSTOP
         "   new crashes : %s%-22s " bSTG bV "\n", DI(total_execs),
         unique_crashes ? cLRD : cNOR, tmp);

  } else {

    SAYF(bV bSTOP " total execs : " cNOR "%-21s " bSTG bV bSTOP
         " total crashes : %s%-22s " bSTG bV "\n", DI(total_execs),
         unique_crashes ? cLRD : cNOR, tmp);

  }

  /* Show a warning about slow execution. */

  if (avg_exec < 100) {

    sprintf(tmp, "%s/sec (%s)", DF(avg_exec), avg_exec < 20 ?
            "zzzz..." : "slow!");

    SAYF(bV bSTOP "  exec speed : " cLRD "%-21s ", tmp);

  } else {

    sprintf(tmp, "%s/sec", DF(avg_exec));
    SAYF(bV bSTOP "  exec speed : " cNOR "%-21s ", tmp);

  }

  sprintf(tmp, "%s (%s%s unique)", DI(total_hangs), DI(unique_hangs),
          (unique_hangs >= KEEP_UNIQUE_HANG) ? "+" : "");

  SAYF (bSTG bV bSTOP "   total hangs : " cNOR "%-22s " bSTG bV "\n", tmp);

  /* Aaaalmost there... hold on! */

  SAYF(bVR bH cCYA bSTOP " fuzzing strategy yields " bSTG bH10 bH bHT bH10
       bH5 bHB bH bSTOP cCYA " path geometry " bSTG bH5 bH2 bH bVL "\n");

  if (skip_deterministic) {

    strcpy(tmp, "n/a, n/a, n/a");

  } else {

    sprintf(tmp, "%s/%s, %s/%s, %s/%s",
            DI(stage_finds[STAGE_FLIP1]), DI(stage_cycles[STAGE_FLIP1]),
            DI(stage_finds[STAGE_FLIP2]), DI(stage_cycles[STAGE_FLIP2]),
            DI(stage_finds[STAGE_FLIP4]), DI(stage_cycles[STAGE_FLIP4]));

  }

  SAYF(bV bSTOP "   bit flips : " cNOR "%-37s " bSTG bV bSTOP "    levels : "
       cNOR "%-10s " bSTG bV "\n", tmp, DI(max_depth));

  if (!skip_deterministic)
    sprintf(tmp, "%s/%s, %s/%s, %s/%s",
            DI(stage_finds[STAGE_FLIP8]), DI(stage_cycles[STAGE_FLIP8]),
            DI(stage_finds[STAGE_FLIP16]), DI(stage_cycles[STAGE_FLIP16]),
            DI(stage_finds[STAGE_FLIP32]), DI(stage_cycles[STAGE_FLIP32]));

  SAYF(bV bSTOP "  byte flips : " cNOR "%-37s " bSTG bV bSTOP "   pending : "
       cNOR "%-10s " bSTG bV "\n", tmp, DI(pending_not_fuzzed));

  if (!skip_deterministic)
    sprintf(tmp, "%s/%s, %s/%s, %s/%s",
            DI(stage_finds[STAGE_ARITH8]), DI(stage_cycles[STAGE_ARITH8]),
            DI(stage_finds[STAGE_ARITH16]), DI(stage_cycles[STAGE_ARITH16]),
            DI(stage_finds[STAGE_ARITH32]), DI(stage_cycles[STAGE_ARITH32]));

  SAYF(bV bSTOP " arithmetics : " cNOR "%-37s " bSTG bV bSTOP "  pend fav : "
       cNOR "%-10s " bSTG bV "\n", tmp, DI(pending_favored));

  if (!skip_deterministic)
    sprintf(tmp, "%s/%s, %s/%s, %s/%s",
            DI(stage_finds[STAGE_INTEREST8]), DI(stage_cycles[STAGE_INTEREST8]),
            DI(stage_finds[STAGE_INTEREST16]), DI(stage_cycles[STAGE_INTEREST16]),
            DI(stage_finds[STAGE_INTEREST32]), DI(stage_cycles[STAGE_INTEREST32]));

  SAYF(bV bSTOP "  known ints : " cNOR "%-37s " bSTG bV bSTOP " own finds : "
       cNOR "%-10s " bSTG bV "\n", tmp, DI(queued_discovered));

  if (!skip_deterministic)
    sprintf(tmp, "%s/%s, %s/%s, %s/%s",
            DI(stage_finds[STAGE_EXTRAS_UO]), DI(stage_cycles[STAGE_EXTRAS_UO]),
            DI(stage_finds[STAGE_EXTRAS_UI]), DI(stage_cycles[STAGE_EXTRAS_UI]),
            DI(stage_finds[STAGE_EXTRAS_AO]), DI(stage_cycles[STAGE_EXTRAS_AO]));

  SAYF(bV bSTOP "  dictionary : " cNOR "%-37s " bSTG bV bSTOP
       "  imported : " cNOR "%-10s " bSTG bV "\n", tmp,
       sync_id ? DI(queued_imported) : (u8*)"n/a");

  sprintf(tmp, "%s/%s, %s/%s",
          DI(stage_finds[STAGE_HAVOC]), DI(stage_cycles[STAGE_HAVOC]),
          DI(stage_finds[STAGE_SPLICE]), DI(stage_cycles[STAGE_SPLICE]));

  SAYF(bV bSTOP "       havoc : " cNOR "%-37s " bSTG bV bSTOP 
       "  variable : %s%-10s " bSTG bV "\n", tmp, queued_variable ? cLRD : cNOR,
      no_var_check ? (u8*)"n/a" : DI(queued_variable));

  if (!bytes_trim_out) {

    sprintf(tmp, "n/a, ");

  } else {

    sprintf(tmp, "%0.02f%%/%s, ",
            ((double)(bytes_trim_in - bytes_trim_out)) * 100 / bytes_trim_in,
            DI(trim_execs));

  }

  if (!blocks_eff_total) {

    u8 tmp2[128];

    sprintf(tmp2, "n/a");
    strcat(tmp, tmp2);

  } else {

    u8 tmp2[128];

    sprintf(tmp2, "%0.02f%%",
            ((double)(blocks_eff_total - blocks_eff_select)) * 100 /
            blocks_eff_total);

    strcat(tmp, tmp2);

  }

  SAYF(bV bSTOP "        trim : " cNOR "%-37s " bSTG bVR bH20 bH2 bH2 bRB "\n"
       bLB bH30 bH20 bH2 bH bRB bSTOP cRST RESET_G1, tmp);

  /* Provide some CPU utilization stats. */

  if (cpu_core_count) {

    double cur_runnable = get_runnable_processes();
    u32 cur_utilization = cur_runnable * 100 / cpu_core_count;

    u8* cpu_color = cCYA;

    /* If we could still run one or more processes, use green. */

    if (cpu_core_count > 1 && cur_runnable + 1 <= cpu_core_count)
      cpu_color = cLGN;

    /* If we're clearly oversubscribed, use red. */

    if (!no_cpu_meter_red && cur_utilization >= 150) cpu_color = cLRD;

    SAYF(SP10 cGRA "   [cpu:%s%3u%%" cGRA "]\r" cRST,
         cpu_color, cur_utilization < 999 ? cur_utilization : 999);

  } else SAYF("\r");

  /* Hallelujah! */

  fflush(0);

}


/* Display quick statistics at the end of processing the input directory,
   plus a bunch of warnings. Some calibration stuff also ended up here,
   along with several hardcoded constants. Maybe clean up eventually. */

static void show_init_stats(void) {

  struct queue_entry* q = queue;
  u32 min_bits = 0, max_bits = 0;
  u64 min_us = 0, max_us = 0;
  u64 avg_us = 0;
  u32 max_len = 0;

  if (total_cal_cycles) avg_us = total_cal_us / total_cal_cycles;

  while (q) {

    if (!min_us || q->exec_us < min_us) min_us = q->exec_us;
    if (q->exec_us > max_us) max_us = q->exec_us;

    if (!min_bits || q->bitmap_size < min_bits) min_bits = q->bitmap_size;
    if (q->bitmap_size > max_bits) max_bits = q->bitmap_size;

    if (q->len > max_len) max_len = q->len;

    q = q->next;

  }

  SAYF("\n");

  if (avg_us > (qemu_mode ? 50000 : 10000)) 
    WARNF(cLRD "The target binary is pretty slow! See %s/perf_tips.txt.",
          doc_path);

  /* Let's keep things moving with slow binaries. */

  if (avg_us > 50000) havoc_div = 10;     /* 0-19 execs/sec   */
  else if (avg_us > 20000) havoc_div = 5; /* 20-49 execs/sec  */
  else if (avg_us > 10000) havoc_div = 2; /* 50-100 execs/sec */

  if (!resuming_fuzz) {

    if (max_len > 50 * 1024)
      WARNF(cLRD "Some test cases are huge (%s) - see %s/perf_tips.txt!",
            DMS(max_len), doc_path);
    else if (max_len > 10 * 1024)
      WARNF("Some test cases are big (%s) - see %s/perf_tips.txt.",
            DMS(max_len), doc_path);

    if (useless_at_start && !in_bitmap)
      WARNF(cLRD "Some test cases look useless. Consider using a smaller set.");

    if (queued_paths > 100)
      WARNF(cLRD "You probably have far too many input files! Consider trimming down.");
    else if (queued_paths > 20)
      WARNF("You have lots of input files; try starting small.");

  }

  OKF("Here are some useful stats:\n\n"

      cGRA "    Test case count : " cNOR "%u favored, %u variable, %u total\n"
      cGRA "       Bitmap range : " cNOR "%u to %u bits (average: %0.02f bits)\n"
      cGRA "        Exec timing : " cNOR "%s to %s us (average: %s us)\n",
      queued_favored, queued_variable, queued_paths, min_bits, max_bits, 
      ((double)total_bitmap_size) / (total_bitmap_entries ? total_bitmap_entries : 1),
      DI(min_us), DI(max_us), DI(avg_us));

  if (!timeout_given) {

    /* Figure out the appropriate timeout. The basic idea is: 5x average or
       1x max, rounded up to EXEC_TM_ROUND ms and capped at 1 second.

       If the program is slow, the multiplier is lowered to 2x or 3x, because
       random scheduler jitter is less likely to have any impact, and because
       our patience is wearing thin =) */

    if (avg_us > 50000) exec_tmout = avg_us * 2 / 1000;
    else if (avg_us > 10000) exec_tmout = avg_us * 3 / 1000;
    else exec_tmout = avg_us * 5 / 1000;

    exec_tmout = MAX(exec_tmout, max_us / 1000);
    exec_tmout = (exec_tmout + EXEC_TM_ROUND) / EXEC_TM_ROUND * EXEC_TM_ROUND;

    if (exec_tmout > EXEC_TIMEOUT) exec_tmout = EXEC_TIMEOUT;

    ACTF("No -t option specified, so I'll use exec timeout of %u ms.", 
         exec_tmout);

    timeout_given = 1;

  } else if (timeout_given == 3) {

    ACTF("Applying timeout settings from resumed session (%u ms).", exec_tmout);

  }

  OKF("All set and ready to roll!");

}


/* Find first power of two greater or equal to val. */

static u32 next_p2(u32 val) {

  u32 ret = 1;
  while (val > ret) ret <<= 1;
  return ret;

} 


/* Trim all new test cases to save cycles when doing deterministic checks. The
   trimmer uses power-of-two increments somewhere between 1/16 and 1/1024 of
   file size, to keep the stage short and sweet. */

static u8 trim_case(char** argv, struct queue_entry* q, u8* in_buf) {

  static u8 tmp[64];
  static u8 clean_trace[MAP_SIZE];

  u8  needs_write = 0, fault = 0;
  u32 trim_exec = 0;
  u32 remove_len;
  u32 len_p2;

  /* Although the trimmer will be less useful when variable behavior is
     detected, it will still work to some extent, so we don't check for
     this. */

  if (q->len < 5) return 0;

  stage_name = tmp;
  bytes_trim_in += q->len;

  /* Select initial chunk len, starting with large steps. */

  len_p2 = next_p2(q->len);

  remove_len = MAX(len_p2 / TRIM_START_STEPS, TRIM_MIN_BYTES);

  /* Continue until the number of steps gets too high or the stepover
     gets too small. */

  while (remove_len >= MAX(len_p2 / TRIM_END_STEPS, TRIM_MIN_BYTES)) {

    u32 remove_pos = remove_len;

    sprintf(tmp, "trim %s/%s", DI(remove_len), DI(remove_len));

    stage_cur = 0;
    stage_max = q->len / remove_len;

    while (remove_pos < q->len) {

      u32 trim_avail = MIN(remove_len, q->len - remove_pos);
      u32 cksum;

      write_with_gap(in_buf, q->len, remove_pos, trim_avail);

      fault = run_target(argv);
      trim_execs++;

      if (stop_soon || fault == FAULT_ERROR) goto abort_trimming;

      /* Note that we don't keep track of crashes or hangs here; maybe TODO? */

      cksum = hash32(trace_bits, MAP_SIZE, HASH_CONST);

      /* If the deletion had no impact on the trace, make it permanent. This
         isn't perfect for variable-path inputs, but we're just making a
         best-effort pass, so it's not a big deal if we end up with false
         negatives every now and then. */

      if (cksum == q->exec_cksum) {

        u32 move_tail = q->len - remove_pos - trim_avail;

        q->len -= trim_avail;
        len_p2  = next_p2(q->len);

        memmove(in_buf + remove_pos, in_buf + remove_pos + trim_avail, 
                move_tail);

        /* Let's save a clean trace, which will be needed by
           update_bitmap_score once we're done with the trimming stuff. */

        if (!needs_write) {

          needs_write = 1;
          memcpy(clean_trace, trace_bits, MAP_SIZE);

        }

      } else remove_pos += remove_len;

      /* Since this can be slow, update the screen every now and then. */

      if (!(trim_exec++ % stats_update_freq)) show_stats();
      stage_cur++;

    }

    remove_len >>= 1;

  }

  /* If we have made changes to in_buf, we also need to update the on-disk
     version of the test case. */

  if (needs_write) {

    s32 fd;

    unlink(q->fname); /* ignore errors */

    fd = open(q->fname, O_WRONLY | O_CREAT | O_EXCL, 0600);

    if (fd < 0) PFATAL("Unable to create '%s'", q->fname);

    ck_write(fd, in_buf, q->len, q->fname);
    close(fd);

    memcpy(trace_bits, clean_trace, MAP_SIZE);
    update_bitmap_score(q);

  }



abort_trimming:

  bytes_trim_out += q->len;
  return fault;

}


/* Write a modified test case, run program, process results. Handle
   error conditions, returning 1 if it's time to bail out. This is
   a helper function for fuzz_one(). */

static u8 common_fuzz_stuff(char** argv, u8* out_buf, u32 len) {

  u8 fault;

  if (post_handler) {

    out_buf = post_handler(out_buf, &len);
    if (!out_buf || !len) return 0;

  }

  write_to_testcase(out_buf, len);

  fault = run_target(argv);

  if (stop_soon) return 1;

  if (fault == FAULT_HANG) {

    if (subseq_hangs++ > HANG_LIMIT) {
      cur_skipped_paths++;
      return 1;
    }

  } else subseq_hangs = 0;

  /* Users can hit us with SIGUSR1 to request the current input
     to be abandoned. */

  if (skip_requested) {

     skip_requested = 0;
     cur_skipped_paths++;
     return 1;

  }

  /* This handles FAULT_ERROR for us: */

  queued_discovered += save_if_interesting(argv, out_buf, len, fault);

  if (!(stage_cur % stats_update_freq) || stage_cur + 1 == stage_max)
    show_stats();

  return 0;

}


/* Helper to choose random block len for block operations in fuzz_one().
   Doesn't return zero, provided that max_len is > 0. */

static u32 choose_block_len(u32 limit) {

  u32 min_value, max_value;
  u32 rlim = MIN(queue_cycle, 3);

  if (!run_over10m) rlim = 1;

  switch (UR(rlim)) {

    case 0:  min_value = 1;
             max_value = HAVOC_BLK_SMALL;
             break;

    case 1:  min_value = HAVOC_BLK_SMALL;
             max_value = HAVOC_BLK_MEDIUM;
             break;

    default: min_value = HAVOC_BLK_MEDIUM;
             max_value = HAVOC_BLK_LARGE;


  }

  if (min_value >= limit) min_value = 1;

  return min_value + UR(MIN(max_value, limit) - min_value + 1);

}


/* Calculate case desirability score to adjust the length of havoc fuzzing.
   A helper function for fuzz_one(). Maybe some of these constants should
   go into config.h. */

static u32 calculate_score(struct queue_entry* q) {

  u32 avg_exec_us = total_cal_us / total_cal_cycles;
  u32 avg_bitmap_size = total_bitmap_size / total_bitmap_entries;
  u32 perf_score = 100;

  /* Adjust score based on execution speed of this path, compared to the
     global average. Multiplier ranges from 0.1x to 3x. Fast inputs are
     less expensive to fuzz, so we're giving them more air time. */

  if (q->exec_us * 0.1 > avg_exec_us) perf_score = 10;
  else if (q->exec_us * 0.25 > avg_exec_us) perf_score = 25;
  else if (q->exec_us * 0.5 > avg_exec_us) perf_score = 50;
  else if (q->exec_us * 0.75 > avg_exec_us) perf_score = 75;
  else if (q->exec_us * 4 < avg_exec_us) perf_score = 300;
  else if (q->exec_us * 3 < avg_exec_us) perf_score = 200;
  else if (q->exec_us * 2 < avg_exec_us) perf_score = 150;

  /* Adjust score based on bitmap size. The working theory is that better
     coverage translates to better targets. Multiplier from 0.25x to 3x. */

  if (q->bitmap_size * 0.3 > avg_bitmap_size) perf_score *= 3;
  else if (q->bitmap_size * 0.5 > avg_bitmap_size) perf_score *= 2;
  else if (q->bitmap_size * 0.75 > avg_bitmap_size) perf_score *= 1.5;
  else if (q->bitmap_size * 3 < avg_bitmap_size) perf_score *= 0.25;
  else if (q->bitmap_size * 2 < avg_bitmap_size) perf_score *= 0.5;
  else if (q->bitmap_size * 1.5 < avg_bitmap_size) perf_score *= 0.75;

  /* Adjust score based on handicap. Handicap is proportional to how late
     in the game we learned about this path. Latecomers are allowed to run
     for a bit longer until they catch up with the rest. */

  if (q->handicap >= 4) {

    perf_score *= 4;
    q->handicap -= 4;

  } else if (q->handicap) {

    perf_score *= 2;
    q->handicap--;

  }

  /* Final adjustment based on input depth, under the assumption that fuzzing
     deeper test cases is more likely to reveal stuff that can't be
     discovered with traditional fuzzers. */

  switch (q->depth) {

    case 0 ... 3:   break;
    case 4 ... 7:   perf_score *= 2; break;
    case 8 ... 13:  perf_score *= 4; break;
    case 14 ... 25: perf_score *= 6; break;
    default:        perf_score *= 8;

  }

  /* Make sure that we don't go over limit. */

  if (perf_score > HAVOC_MAX_MULT * 100) perf_score = HAVOC_MAX_MULT * 100;

  return perf_score;

}


/* Helper function to see if a particular change (xor_val = old ^ new) could
   be a product of deterministic bit flips with the lengths and stepovers
   attempted by afl-fuzz. This is used to avoid dupes in some of the
   deterministic fuzzing operations that follow bit flips. We also
   return 1 if xor_val is zero, which implies that the old and attempted new
   values are identical and the exec would be a waste of time. */

static u8 could_be_bitflip(u32 xor_val) {

  u32 sh = 0;

  if (!xor_val) return 1;

  /* Shift left until first bit set. */

  while (!(xor_val & 1)) { sh++; xor_val >>= 1; }

  /* 1-, 2-, and 4-bit patterns are OK anywhere. */

  if (xor_val == 1 || xor_val == 3 || xor_val == 15) return 1;

  /* 8-, 16-, and 32-bit patterns are OK only if shift factor is
     divisible by 8, since that's the stepover for these ops. */

  if (sh & 7) return 0;

  if (xor_val == 0xff || xor_val == 0xffff || xor_val == 0xffffffff)
    return 1;

  return 0;

}


/* Helper function to see if a particular value is reachable through
   arithmetic operations. Used for similar purposes. */

static u8 could_be_arith(u32 old_val, u32 new_val, u8 blen) {

  u32 i, ov = 0, nv = 0, diffs = 0;

  if (old_val == new_val) return 1;

  /* See if one-byte adjustments to any byte could produce this result. */

  for (i = 0; i < blen; i++) {

    u8 a = old_val >> (8 * i),
       b = new_val >> (8 * i);

    if (a != b) { diffs++; ov = a; nv = b; }

  }

  /* If only one byte differs and the values are within range, return 1. */

  if (diffs == 1) {

    if ((u8)(ov - nv) <= ARITH_MAX ||
        (u8)(nv - ov) <= ARITH_MAX) return 1;

  }

  if (blen == 1) return 0;

  /* See if two-byte adjustments to any byte would produce this result. */

  diffs = 0;

  for (i = 0; i < blen / 2; i++) {

    u16 a = old_val >> (16 * i),
        b = new_val >> (16 * i);

    if (a != b) { diffs++; ov = a; nv = b; }

  }

  /* If only one word differs and the values are within range, return 1. */

  if (diffs == 1) {

    if ((u16)(ov - nv) <= ARITH_MAX ||
        (u16)(nv - ov) <= ARITH_MAX) return 1;

    ov = SWAP16(ov); nv = SWAP16(nv);

    if ((u16)(ov - nv) <= ARITH_MAX ||
        (u16)(nv - ov) <= ARITH_MAX) return 1;

  }

  /* Finally, let's do the same thing for dwords. */

  if (blen == 4) {

    if ((u32)(old_val - new_val) <= ARITH_MAX ||
        (u32)(new_val - old_val) <= ARITH_MAX) return 1;

    new_val = SWAP32(new_val);
    old_val = SWAP32(old_val);

    if ((u32)(old_val - new_val) <= ARITH_MAX ||
        (u32)(new_val - old_val) <= ARITH_MAX) return 1;

  }

  return 0;

}


/* Last but not least, a similar helper to see if insertion of an 
   interesting integer is redundant given the insertions done for
   shorter blen. The last param (check_le) is set if the caller
   already executed LE insertion for current blen and wants to see
   if BE variant passed in new_val is unique. */

static u8 could_be_interest(u32 old_val, u32 new_val, u8 blen, u8 check_le) {

  u32 i, j;

  if (old_val == new_val) return 1;

  /* See if one-byte insertions from interesting_8 over old_val could
     produce new_val. */

  for (i = 0; i < blen; i++) {

    for (j = 0; j < sizeof(interesting_8); j++) {

      u32 tval = (old_val & ~(0xff << (i * 8))) |
                 (((u8)interesting_8[j]) << (i * 8));

      if (new_val == tval) return 1;

    }

  }

  /* Bail out unless we're also asked to examine two-byte LE insertions
     as a preparation for BE attempts. */

  if (blen == 2 && !check_le) return 0;

  /* See if two-byte insertions over old_val could give us new_val. */

  for (i = 0; i < blen - 1; i++) {

    for (j = 0; j < sizeof(interesting_16) / 2; j++) {

      u32 tval = (old_val & ~(0xffff << (i * 8))) |
                 (((u16)interesting_16[j]) << (i * 8));

      if (new_val == tval) return 1;

      /* Continue here only if blen > 2. */

      if (blen > 2) {

        tval = (old_val & ~(0xffff << (i * 8))) |
               (SWAP16(interesting_16[j]) << (i * 8));

        if (new_val == tval) return 1;

      }

    }

  }

  if (blen == 4 && check_le) {

    /* See if four-byte insertions could produce the same result
       (LE only). */

    for (j = 0; j < sizeof(interesting_32) / 4; j++)
      if (new_val == (u32)interesting_32[j]) return 1;

  }

  return 0;

}


/* Take the current entry from the queue, fuzz it for a while. This
   function is a tad too long... returns 0 if fuzzed successfully, 1 if
   skipped or bailed out. */

static u8 fuzz_one(char** argv) {

  s32 len, fd, temp_len, i, j;
  u8  *in_buf, *out_buf, *orig_in, *ex_tmp, *eff_map = 0;
  u64 havoc_queued,  orig_hit_cnt, new_hit_cnt;
  u32 splice_cycle = 0, perf_score = 100, orig_perf, prev_cksum, eff_cnt = 1;

  u8  ret_val = 1;

  u8  a_collect[MAX_AUTO_EXTRA];
  u32 a_len = 0;

#ifdef IGNORE_FINDS

  /* In IGNORE_FINDS mode, skip any entries that weren't in the
     initial data set. */

  if (queue_cur->depth > 1) return 1;

#else

  if (pending_favored) {

    /* If we have any favored, non-fuzzed new arrivals in the queue,
       possibly skip to them at the expense of already-fuzzed or non-favored
       cases. */

    if ((queue_cur->was_fuzzed || !queue_cur->favored) &&
        UR(100) < SKIP_TO_NEW_PROB) return 1;

  } else if (!dumb_mode && !queue_cur->favored && queued_paths > 10) {

    /* Otherwise, still possibly skip non-favored cases, albeit less often.
       The odds of skipping stuff are higher for already-fuzzed inputs and
       lower for never-fuzzed entries. */

    if (queue_cycle > 1 && !queue_cur->was_fuzzed) {

      if (UR(100) < SKIP_NFAV_NEW_PROB) return 1;

    } else {

      if (UR(100) < SKIP_NFAV_OLD_PROB) return 1;

    }

  }

#endif /* ^IGNORE_FINDS */

  if (not_on_tty)
    ACTF("Fuzzing test case #%u (%u total)...", current_entry, queued_paths);

  /* Map the test case into memory. */

  fd = open(queue_cur->fname, O_RDONLY);

  if (fd < 0) PFATAL("Unable to open '%s'", queue_cur->fname);

  len = queue_cur->len;

  orig_in = in_buf = mmap(0, len, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);

  if (orig_in == MAP_FAILED) PFATAL("Unable to mmap '%s'", queue_cur->fname);

  close(fd);

  /* We could mmap() out_buf as MAP_PRIVATE, but we end up clobbering every
     single byte anyway, so it wouldn't give us any performance or memory usage
     benefits. */

  out_buf = ck_alloc_nozero(len);

  subseq_hangs = 0;

  cur_depth = queue_cur->depth;

  /*******************************************
   * CALIBRATION (only if failed earlier on) *
   *******************************************/

  if (queue_cur->cal_failed) {

    u8 res = FAULT_HANG;

    if (queue_cur->cal_failed < CAL_CHANCES) {

      res = calibrate_case(argv, queue_cur, in_buf, queue_cycle - 1, 0);

      if (res == FAULT_ERROR)
        FATAL("Unable to execute target application");

    }

    if (stop_soon || res != crash_mode) {
      cur_skipped_paths++;
      goto abandon_entry;
    }

  }

  /************
   * TRIMMING *
   ************/

  if (!dumb_mode && !queue_cur->trim_done) {

    u8 res = trim_case(argv, queue_cur, in_buf);

    if (res == FAULT_ERROR)
      FATAL("Unable to execute target application");

    if (stop_soon) {
      cur_skipped_paths++;
      goto abandon_entry;
    }

    /* Don't retry trimming, even if it failed. */

    queue_cur->trim_done = 1;

    if (len != queue_cur->len) len = queue_cur->len;

  }

  memcpy(out_buf, in_buf, len);

  /*********************
   * PERFORMANCE SCORE *
   *********************/

  orig_perf = perf_score = calculate_score(queue_cur);

  /* Skip right away if -d is given, if we have done deterministic fuzzing on
     this entry ourselves (was_fuzzed), or if it has gone through deterministic
     testing in earlier, resumed runs (passed_det). */

  if (skip_deterministic || queue_cur->was_fuzzed || queue_cur->passed_det)
    goto havoc_stage;

  /*********************************************
   * SIMPLE BITFLIP (+dictionary construction) *
   *********************************************/

#define FLIP_BIT(_ar, _b) do { \
    u8* _arf = (u8*)(_ar); \
    u32 _bf = (_b); \
    _arf[(_bf) >> 3] ^= (128 >> ((_bf) & 7)); \
  } while (0)

  /* Single walking bit. */

  stage_short = "flip1";
  stage_max   = len << 3;
  stage_name  = "bitflip 1/1";

  stage_val_type = STAGE_VAL_NONE;

  orig_hit_cnt = queued_paths + unique_crashes;

  prev_cksum = queue_cur->exec_cksum;

  for (stage_cur = 0; stage_cur < stage_max; stage_cur++) {

    stage_cur_byte = stage_cur >> 3;

    FLIP_BIT(out_buf, stage_cur);

    if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;

    FLIP_BIT(out_buf, stage_cur);

    /* While flipping the least significant bit in every byte, pull of an extra
       trick to detect possible syntax tokens. In essence, the idea is that if
       you have a binary blob like this:

       xxxxxxxxIHDRxxxxxxxx

       ...and changing the leading and trailing bytes causes variable or no
       changes in program flow, but touching any character in the "IHDR" string
       always produces the same, distinctive path, it's highly likely that
       "IHDR" is an atomically-checked magic value of special significance to
       the fuzzed format.

       We do this here, rather than as a separate stage, because it's a nice
       way to keep the operation approximately "free" (i.e., no extra execs).
       
       Empirically, performing the check when flipping the least significant bit
       is advantageous, compared to doing it at the time of more disruptive
       changes, where the program flow may be affected in more violent ways.

       The caveat is that we won't generate dictionaries in the -d mode or -S
       mode - but that's probably a fair trade-off.

       This won't work particularly well with paths that exhibit variable
       behavior, but fails gracefully, so we'll carry out the checks anyway.

      */

    if (!dumb_mode && (stage_cur & 7) == 7) {

      u32 cksum = hash32(trace_bits, MAP_SIZE, HASH_CONST);

      if (stage_cur == stage_max - 1 && cksum == prev_cksum) {

        /* If at end of file and we are still collecting a string, grab the
           final character and force output. */

        if (a_len < MAX_AUTO_EXTRA) a_collect[a_len] = out_buf[stage_cur >> 3];
        a_len++;

        if (a_len >= MIN_AUTO_EXTRA && a_len <= MAX_AUTO_EXTRA)
          maybe_add_auto(a_collect, a_len);

      } else if (cksum != prev_cksum) {

        /* Otherwise, if the checksum has changed, see if we have something
           worthwhile queued up, and collect that if the answer is yes. */

        if (a_len >= MIN_AUTO_EXTRA && a_len <= MAX_AUTO_EXTRA)
          maybe_add_auto(a_collect, a_len);

        a_len = 0;
        prev_cksum = cksum;

      }

      /* Continue collecting string, but only if the bit flip actually made
         any difference - we don't want no-op tokens. */

      if (cksum != queue_cur->exec_cksum) {

        if (a_len < MAX_AUTO_EXTRA) a_collect[a_len] = out_buf[stage_cur >> 3];        
        a_len++;

      }

    }

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_FLIP1]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_FLIP1] += stage_max;

  if (queue_cur->passed_det) goto havoc_stage;

  /* Two walking bits. */

  stage_name  = "bitflip 2/1";
  stage_short = "flip2";
  stage_max   = (len << 3) - 1;

  orig_hit_cnt = new_hit_cnt;

  for (stage_cur = 0; stage_cur < stage_max; stage_cur++) {

    stage_cur_byte = stage_cur >> 3;

    FLIP_BIT(out_buf, stage_cur);
    FLIP_BIT(out_buf, stage_cur + 1);

    if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;

    FLIP_BIT(out_buf, stage_cur);
    FLIP_BIT(out_buf, stage_cur + 1);

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_FLIP2]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_FLIP2] += stage_max;

  /* Four walking bits. */

  stage_name  = "bitflip 4/1";
  stage_short = "flip4";
  stage_max   = (len << 3) - 3;

  orig_hit_cnt = new_hit_cnt;

  for (stage_cur = 0; stage_cur < stage_max; stage_cur++) {

    stage_cur_byte = stage_cur >> 3;

    FLIP_BIT(out_buf, stage_cur);
    FLIP_BIT(out_buf, stage_cur + 1);
    FLIP_BIT(out_buf, stage_cur + 2);
    FLIP_BIT(out_buf, stage_cur + 3);

    if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;

    FLIP_BIT(out_buf, stage_cur);
    FLIP_BIT(out_buf, stage_cur + 1);
    FLIP_BIT(out_buf, stage_cur + 2);
    FLIP_BIT(out_buf, stage_cur + 3);

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_FLIP4]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_FLIP4] += stage_max;

  /* Effector map setup. These macros calculate:

     EFF_APOS      - position of a particular file offset in the map.
     EFF_ALEN      - length of an map with a particular number of bytes.
     EFF_SPAN_ALEN - map span for a sequence of bytes.

   */

#define EFF_APOS(_p)          ((_p) >> EFF_MAP_SCALE2)
#define EFF_REM(_x)           ((_x) & ((1 << EFF_MAP_SCALE2) - 1))
#define EFF_ALEN(_l)          (EFF_APOS(_l) + !!EFF_REM(_l))
#define EFF_SPAN_ALEN(_p, _l) (EFF_APOS((_p) + (_l) - 1) - EFF_APOS(_p) + 1)

  /* Initialize effector map for the next step (see comments below). Always
     flag first and last byte as doing something. */

  eff_map    = ck_alloc(EFF_ALEN(len));
  eff_map[0] = 1;

  if (EFF_APOS(len - 1) != 0) {
    eff_map[EFF_APOS(len - 1)] = 1;
    eff_cnt++;
  }

  /* Walking byte. */

  stage_name  = "bitflip 8/8";
  stage_short = "flip8";
  stage_max   = len;

  orig_hit_cnt = new_hit_cnt;

  for (stage_cur = 0; stage_cur < stage_max; stage_cur++) {

    stage_cur_byte = stage_cur;

    out_buf[stage_cur] ^= 0xFF;

    if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;

    /* We also use this stage to pull off a simple trick: we identify
       bytes that seem to have no effect on the current execution path
       even when fully flipped - and we skip them during more expensive
       deterministic stages, such as arithmetics or known ints. */

    if (!eff_map[EFF_APOS(stage_cur)]) {

      u32 cksum;

      /* If in dumb mode or if the file is very short, just flag everything
         without wasting time on checksums. */

      if (!dumb_mode && len >= EFF_MIN_LEN)
        cksum = hash32(trace_bits, MAP_SIZE, HASH_CONST);
      else
        cksum = ~queue_cur->exec_cksum;

      if (cksum != queue_cur->exec_cksum) {
        eff_map[EFF_APOS(stage_cur)] = 1;
        eff_cnt++;
      }

    }

    out_buf[stage_cur] ^= 0xFF;

  }

  /* If the effector map is more than EFF_MAX_PERC dense, just flag the
     whole thing as worth fuzzing, since we wouldn't be saving much time
     anyway. */

  if (eff_cnt != EFF_ALEN(len) &&
      eff_cnt * 100 / EFF_ALEN(len) > EFF_MAX_PERC) {

    memset(eff_map, 1, EFF_ALEN(len));

    blocks_eff_select += EFF_ALEN(len);

  } else {

    blocks_eff_select += eff_cnt;

  }

  blocks_eff_total += EFF_ALEN(len);

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_FLIP8]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_FLIP8] += stage_max;

  /* Two walking bytes. */

  if (len < 2) goto skip_bitflip;

  stage_name  = "bitflip 16/8";
  stage_short = "flip16";
  stage_cur   = 0;
  stage_max   = len - 1;

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len - 1; i++) {

    /* Let's consult the effector map... */

    if (!eff_map[EFF_APOS(i)] && !eff_map[EFF_APOS(i + 1)]) {
      stage_max--;
      continue;
    }

    stage_cur_byte = i;

    *(u16*)(out_buf + i) ^= 0xFFFF;

    if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
    stage_cur++;

    *(u16*)(out_buf + i) ^= 0xFFFF;


  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_FLIP16]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_FLIP16] += stage_max;

  if (len < 4) goto skip_bitflip;

  /* Four walking bytes. */

  stage_name  = "bitflip 32/8";
  stage_short = "flip32";
  stage_cur   = 0;
  stage_max   = len - 3;

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len - 3; i++) {

    /* Let's consult the effector map... */
    if (!eff_map[EFF_APOS(i)] && !eff_map[EFF_APOS(i + 1)] &&
        !eff_map[EFF_APOS(i + 2)] && !eff_map[EFF_APOS(i + 3)]) {
      stage_max--;
      continue;
    }

    stage_cur_byte = i;

    *(u32*)(out_buf + i) ^= 0xFFFFFFFF;

    if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
    stage_cur++;

    *(u32*)(out_buf + i) ^= 0xFFFFFFFF;

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_FLIP32]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_FLIP32] += stage_max;

skip_bitflip:

  /**********************
   * ARITHMETIC INC/DEC *
   **********************/

  /* 8-bit arithmetics. */

  stage_name  = "arith 8/8";
  stage_short = "arith8";
  stage_cur   = 0;
  stage_max   = 2 * len * ARITH_MAX;

  stage_val_type = STAGE_VAL_LE;

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len; i++) {

    u8 orig = out_buf[i];

    /* Let's consult the effector map... */

    if (!eff_map[EFF_APOS(i)]) {
      stage_max -= 2 * ARITH_MAX;
      continue;
    }

    stage_cur_byte = i;

    for (j = 1; j <= ARITH_MAX; j++) {

      u8 r = orig ^ (orig + j);

      /* Do arithmetic operations only if the result couldn't be a product
         of a bitflip. */

      if (!could_be_bitflip(r)) {

        stage_cur_val = j;
        out_buf[i] = orig + j;

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      r =  orig ^ (orig - j);

      if (!could_be_bitflip(r)) {

        stage_cur_val = -j;
        out_buf[i] = orig - j;

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      out_buf[i] = orig;

    }

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_ARITH8]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_ARITH8] += stage_max;

  /* 16-bit arithmetics, both endians. */

  if (len < 2) goto skip_arith;

  stage_name  = "arith 16/8";
  stage_short = "arith16";
  stage_cur   = 0;
  stage_max   = 4 * (len - 1) * ARITH_MAX;

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len - 1; i++) {

    u16 orig = *(u16*)(out_buf + i);

    /* Let's consult the effector map... */

    if (!eff_map[EFF_APOS(i)] && !eff_map[EFF_APOS(i + 1)]) {
      stage_max -= 4 * ARITH_MAX;
      continue;
    }

    stage_cur_byte = i;

    for (j = 1; j <= ARITH_MAX; j++) {

      u16 r1 = orig ^ (orig + j),
          r2 = orig ^ (orig - j),
          r3 = orig ^ SWAP16(SWAP16(orig) + j),
          r4 = orig ^ SWAP16(SWAP16(orig) - j);

      /* Try little endian addition and subtraction first. Do it only
         if the operation would affect more than one byte (hence the 
         & 0xff overflow checks) and if it couldn't be a product of
         a bitflip. */

      stage_val_type = STAGE_VAL_LE; 

      if ((orig & 0xff) + j > 0xff && !could_be_bitflip(r1)) {

        stage_cur_val = j;
        *(u16*)(out_buf + i) = orig + j;

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;
 
      } else stage_max--;

      if ((orig & 0xff) < j && !could_be_bitflip(r2)) {

        stage_cur_val = -j;
        *(u16*)(out_buf + i) = orig - j;

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      /* Big endian comes next. Same deal. */

      stage_val_type = STAGE_VAL_BE;


      if ((orig >> 8) + j > 0xff && !could_be_bitflip(r3)) {

        stage_cur_val = j;
        *(u16*)(out_buf + i) = SWAP16(SWAP16(orig) + j);

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      if ((orig >> 8) < j && !could_be_bitflip(r4)) {

        stage_cur_val = -j;
        *(u16*)(out_buf + i) = SWAP16(SWAP16(orig) - j);

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      *(u16*)(out_buf + i) = orig;

    }

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_ARITH16]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_ARITH16] += stage_max;

  /* 32-bit arithmetics, both endians. */

  if (len < 4) goto skip_arith;

  stage_name  = "arith 32/8";
  stage_short = "arith32";
  stage_cur   = 0;
  stage_max   = 4 * (len - 3) * ARITH_MAX;

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len - 3; i++) {

    u32 orig = *(u32*)(out_buf + i);

    /* Let's consult the effector map... */

    if (!eff_map[EFF_APOS(i)] && !eff_map[EFF_APOS(i + 1)] &&
        !eff_map[EFF_APOS(i + 2)] && !eff_map[EFF_APOS(i + 3)]) {
      stage_max -= 4 * ARITH_MAX;
      continue;
    }

    stage_cur_byte = i;

    for (j = 1; j <= ARITH_MAX; j++) {

      u32 r1 = orig ^ (orig + j),
          r2 = orig ^ (orig - j),
          r3 = orig ^ SWAP32(SWAP32(orig) + j),
          r4 = orig ^ SWAP32(SWAP32(orig) - j);

      /* Little endian first. Same deal as with 16-bit: we only want to
         try if the operation would have effect on more than two bytes. */

      stage_val_type = STAGE_VAL_LE; 

      if ((orig & 0xffff) + j > 0xffff && !could_be_bitflip(r1)) {

        stage_cur_val = j;
        *(u32*)(out_buf + i) = orig + j;

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      if ((orig & 0xffff) < j && !could_be_bitflip(r2)) {

        stage_cur_val = -j;
        *(u32*)(out_buf + i) = orig - j;

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      /* Big endian next. */

      stage_val_type = STAGE_VAL_BE;

      if ((SWAP32(orig) & 0xffff) + j > 0xffff && !could_be_bitflip(r3)) {

        stage_cur_val = j;
        *(u32*)(out_buf + i) = SWAP32(SWAP32(orig) + j);

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      if ((SWAP32(orig) & 0xffff) < j && !could_be_bitflip(r4)) {

        stage_cur_val = -j;
        *(u32*)(out_buf + i) = SWAP32(SWAP32(orig) - j);

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      *(u32*)(out_buf + i) = orig;

    }

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_ARITH32]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_ARITH32] += stage_max;

skip_arith:

  /**********************
   * INTERESTING VALUES *
   **********************/

  stage_name  = "interest 8/8";
  stage_short = "int8";
  stage_cur   = 0;
  stage_max   = len * sizeof(interesting_8);

  stage_val_type = STAGE_VAL_LE;

  orig_hit_cnt = new_hit_cnt;

  /* Setting 8-bit integers. */

  for (i = 0; i < len; i++) {

    u8 orig = out_buf[i];

    /* Let's consult the effector map... */

    if (!eff_map[EFF_APOS(i)]) {
      stage_max -= sizeof(interesting_8);
      continue;
    }

    stage_cur_byte = i;

    for (j = 0; j < sizeof(interesting_8); j++) {

      /* Skip if the value could be a product of bitflips or arithmetics. */

      if (could_be_bitflip(orig ^ (u8)interesting_8[j]) ||
          could_be_arith(orig, (u8)interesting_8[j], 1)) {
        stage_max--;
        continue;
      }

      stage_cur_val = interesting_8[j];
      out_buf[i] = interesting_8[j];

      if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;

      out_buf[i] = orig;
      stage_cur++;

    }

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_INTEREST8]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_INTEREST8] += stage_max;

  /* Setting 16-bit integers, both endians. */

  if (len < 2) goto skip_interest;

  stage_name  = "interest 16/8";
  stage_short = "int16";
  stage_cur   = 0;
  stage_max   = 2 * (len - 1) * (sizeof(interesting_16) >> 1);

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len - 1; i++) {

    u16 orig = *(u16*)(out_buf + i);

    /* Let's consult the effector map... */

    if (!eff_map[EFF_APOS(i)] && !eff_map[EFF_APOS(i + 1)]) {
      stage_max -= sizeof(interesting_16);
      continue;
    }

    stage_cur_byte = i;

    for (j = 0; j < sizeof(interesting_16) / 2; j++) {

      stage_cur_val = interesting_16[j];

      /* Skip if this could be a product of a bitflip, arithmetics,
         or single-byte interesting value insertion. */

      if (!could_be_bitflip(orig ^ (u16)interesting_16[j]) &&
          !could_be_arith(orig, (u16)interesting_16[j], 2) &&
          !could_be_interest(orig, (u16)interesting_16[j], 2, 0)) {

        stage_val_type = STAGE_VAL_LE;

        *(u16*)(out_buf + i) = interesting_16[j];

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      if ((u16)interesting_16[j] != SWAP16(interesting_16[j]) &&
          !could_be_bitflip(orig ^ SWAP16(interesting_16[j])) &&
          !could_be_arith(orig, SWAP16(interesting_16[j]), 2) &&
          !could_be_interest(orig, SWAP16(interesting_16[j]), 2, 1)) {

        stage_val_type = STAGE_VAL_BE;

        *(u16*)(out_buf + i) = SWAP16(interesting_16[j]);
        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

    }

    *(u16*)(out_buf + i) = orig;

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_INTEREST16]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_INTEREST16] += stage_max;

  if (len < 4) goto skip_interest;

  /* Setting 32-bit integers, both endians. */

  stage_name  = "interest 32/8";
  stage_short = "int32";
  stage_cur   = 0;
  stage_max   = 2 * (len - 3) * (sizeof(interesting_32) >> 2);

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len - 3; i++) {

    u32 orig = *(u32*)(out_buf + i);

    /* Let's consult the effector map... */

    if (!eff_map[EFF_APOS(i)] && !eff_map[EFF_APOS(i + 1)] &&
        !eff_map[EFF_APOS(i + 2)] && !eff_map[EFF_APOS(i + 3)]) {
      stage_max -= sizeof(interesting_32) >> 1;
      continue;
    }

    stage_cur_byte = i;

    for (j = 0; j < sizeof(interesting_32) / 4; j++) {

      stage_cur_val = interesting_32[j];

      /* Skip if this could be a product of a bitflip, arithmetics,
         or word interesting value insertion. */

      if (!could_be_bitflip(orig ^ (u32)interesting_32[j]) &&
          !could_be_arith(orig, interesting_32[j], 4) &&
          !could_be_interest(orig, interesting_32[j], 4, 0)) {

        stage_val_type = STAGE_VAL_LE;

        *(u32*)(out_buf + i) = interesting_32[j];

        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

      if ((u32)interesting_32[j] != SWAP32(interesting_32[j]) &&
          !could_be_bitflip(orig ^ SWAP32(interesting_32[j])) &&
          !could_be_arith(orig, SWAP32(interesting_32[j]), 4) &&
          !could_be_interest(orig, SWAP32(interesting_32[j]), 4, 1)) {

        stage_val_type = STAGE_VAL_BE;

        *(u32*)(out_buf + i) = SWAP32(interesting_32[j]);
        if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;
        stage_cur++;

      } else stage_max--;

    }

    *(u32*)(out_buf + i) = orig;

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_INTEREST32]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_INTEREST32] += stage_max;

skip_interest:

  /********************
   * DICTIONARY STUFF *
   ********************/

  if (!extras_cnt) goto skip_user_extras;

  /* Overwrite with user-supplied extras. */

  stage_name  = "user extras (over)";
  stage_short = "ext_UO";
  stage_cur   = 0;
  stage_max   = extras_cnt * len;

  stage_val_type = STAGE_VAL_NONE;

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len; i++) {

    u32 last_len = 0;

    stage_cur_byte = i;

    /* Extras are sorted by size, from smallest to largest. This means
       that we don't have to worry about restoring the buffer in
       between writes at a particular offset determined by the outer
       loop. */

    for (j = 0; j < extras_cnt; j++) {

      /* Skip extras probabilistically if extras_cnt > MAX_DET_EXTRAS. Also
         skip them if there's no room to insert the payload, if the token
         is redundant, or if its entire span has no bytes set in the effector
         map. */

      if ((extras_cnt > MAX_DET_EXTRAS && UR(extras_cnt) >= MAX_DET_EXTRAS) ||
          extras[j].len > len - i ||
          !memcmp(extras[j].data, out_buf + i, extras[j].len) ||
          !memchr(eff_map + EFF_APOS(i), 1, EFF_SPAN_ALEN(i, extras[j].len))) {

        stage_max--;
        continue;

      }

      last_len = extras[j].len;
      memcpy(out_buf + i, extras[j].data, last_len);

      if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;

      stage_cur++;

    }

    /* Restore all the clobbered memory. */
    memcpy(out_buf + i, in_buf + i, last_len);

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_EXTRAS_UO]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_EXTRAS_UO] += stage_max;

  /* Insertion of user-supplied extras. */

  stage_name  = "user extras (insert)";
  stage_short = "ext_UI";
  stage_cur   = 0;
  stage_max   = extras_cnt * len;

  orig_hit_cnt = new_hit_cnt;

  ex_tmp = ck_alloc(len + MAX_DICT_FILE);

  for (i = 0; i < len; i++) {

    stage_cur_byte = i;

    for (j = 0; j < extras_cnt; j++) {

      /* Insert token */
      memcpy(ex_tmp + i, extras[j].data, extras[j].len);

      /* Copy tail */
      memcpy(ex_tmp + i + extras[j].len, out_buf + i, len - i);

      if (common_fuzz_stuff(argv, ex_tmp, len + extras[j].len)) {
        ck_free(ex_tmp);
        goto abandon_entry;
      }

      stage_cur++;

    }

    /* Copy head */
    ex_tmp[i] = out_buf[i];

  }

  ck_free(ex_tmp);

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_EXTRAS_UI]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_EXTRAS_UI] += stage_max;

skip_user_extras:

  if (!a_extras_cnt) goto skip_extras;

  stage_name  = "auto extras (over)";
  stage_short = "ext_AO";
  stage_cur   = 0;
  stage_max   = MIN(a_extras_cnt, USE_AUTO_EXTRAS) * len;

  stage_val_type = STAGE_VAL_NONE;

  orig_hit_cnt = new_hit_cnt;

  for (i = 0; i < len; i++) {

    u32 last_len = 0;

    stage_cur_byte = i;

    for (j = 0; j < MIN(a_extras_cnt, USE_AUTO_EXTRAS); j++) {

      /* See the comment in the earlier code; extras are sorted by size. */

      if (a_extras[j].len > len - i ||
          !memcmp(a_extras[j].data, out_buf + i, a_extras[j].len) ||
          !memchr(eff_map + EFF_APOS(i), 1, EFF_SPAN_ALEN(i, a_extras[j].len))) {

        stage_max--;
        continue;

      }

      last_len = a_extras[j].len;
      memcpy(out_buf + i, a_extras[j].data, last_len);

      if (common_fuzz_stuff(argv, out_buf, len)) goto abandon_entry;

      stage_cur++;

    }

    /* Restore all the clobbered memory. */
    memcpy(out_buf + i, in_buf + i, last_len);

  }

  new_hit_cnt = queued_paths + unique_crashes;

  stage_finds[STAGE_EXTRAS_AO]  += new_hit_cnt - orig_hit_cnt;
  stage_cycles[STAGE_EXTRAS_AO] += stage_max;

skip_extras:

  /* If we made this to here without jumping to havoc_stage or abandon_entry,
     we're properly done with deterministic steps and can mark it as such
     in the .state/ directory. */

  if (!queue_cur->passed_det) mark_as_det_done(queue_cur);

  /****************
   * RANDOM HAVOC *
   ****************/

havoc_stage:

  stage_cur_byte = -1;

  /* The havoc stage mutation code is also invoked when splicing files; if the
     splice_cycle variable is set, generate different descriptions and such. */

  if (!splice_cycle) {

    stage_name  = "havoc";
    stage_short = "havoc";
    stage_max   = HAVOC_CYCLES * perf_score / havoc_div / 100;

  } else {

    static u8 tmp[32];

    perf_score = orig_perf;

    sprintf(tmp, "splice %u", splice_cycle);
    stage_name  = tmp;
    stage_short = "splice";
    stage_max   = SPLICE_HAVOC * perf_score / havoc_div / 100;

  }

  if (stage_max < HAVOC_MIN) stage_max = HAVOC_MIN;

  temp_len = len;

  orig_hit_cnt = queued_paths + unique_crashes;

  havoc_queued = queued_paths;

  /* We essentially just do several thousand runs (depending on perf_score)
     where we take the input file and make random stacked tweaks. */

  for (stage_cur = 0; stage_cur < stage_max; stage_cur++) {

    u32 use_stacking = 1 << (1 + UR(HAVOC_STACK_POW2));

    stage_cur_val = use_stacking;
 
    for (i = 0; i < use_stacking; i++) {

      switch (UR(15 + ((extras_cnt + a_extras_cnt) ? 2 : 0))) {

        case 0:

          /* Flip a single bit somewhere. Spooky! */

          FLIP_BIT(out_buf, UR(temp_len << 3));
          break;

        case 1: 

          /* Set byte to interesting value. */

          out_buf[UR(temp_len)] = interesting_8[UR(sizeof(interesting_8))];
          break;

        case 2:

          /* Set word to interesting value, randomly choosing endian. */

          if (temp_len < 2) break;

          if (UR(2)) {

            *(u16*)(out_buf + UR(temp_len - 1)) =
              interesting_16[UR(sizeof(interesting_16) >> 1)];

          } else {

            *(u16*)(out_buf + UR(temp_len - 1)) = SWAP16(
              interesting_16[UR(sizeof(interesting_16) >> 1)]);

          }

          break;

        case 3:

          /* Set dword to interesting value, randomly choosing endian. */

          if (temp_len < 4) break;

          if (UR(2)) {
  
            *(u32*)(out_buf + UR(temp_len - 3)) =
              interesting_32[UR(sizeof(interesting_32) >> 2)];

          } else {

            *(u32*)(out_buf + UR(temp_len - 3)) = SWAP32(
              interesting_32[UR(sizeof(interesting_32) >> 2)]);

          }

          break;

        case 4:

          /* Randomly subtract from byte. */

          out_buf[UR(temp_len)] -= 1 + UR(ARITH_MAX);
          break;

        case 5:

          /* Randomly add to byte. */

          out_buf[UR(temp_len)] += 1 + UR(ARITH_MAX);
          break;

        case 6:

          /* Randomly subtract from word, random endian. */

          if (temp_len < 2) break;

          if (UR(2)) {

            u32 pos = UR(temp_len - 1);

            *(u16*)(out_buf + pos) -= 1 + UR(ARITH_MAX);

          } else {

            u32 pos = UR(temp_len - 1);
            u16 num = 1 + UR(ARITH_MAX);

            *(u16*)(out_buf + pos) =
              SWAP16(SWAP16(*(u16*)(out_buf + pos)) - num);

          }

          break;

        case 7:

          /* Randomly add to word, random endian. */

          if (temp_len < 2) break;

          if (UR(2)) {

            u32 pos = UR(temp_len - 1);

            *(u16*)(out_buf + pos) += 1 + UR(ARITH_MAX);

          } else {

            u32 pos = UR(temp_len - 1);
            u16 num = 1 + UR(ARITH_MAX);

            *(u16*)(out_buf + pos) =
              SWAP16(SWAP16(*(u16*)(out_buf + pos)) + num);

          }

          break;

        case 8:

          /* Randomly subtract from dword, random endian. */

          if (temp_len < 4) break;

          if (UR(2)) {

            u32 pos = UR(temp_len - 3);

            *(u32*)(out_buf + pos) -= 1 + UR(ARITH_MAX);

          } else {

            u32 pos = UR(temp_len - 3);
            u32 num = 1 + UR(ARITH_MAX);

            *(u32*)(out_buf + pos) =
              SWAP32(SWAP32(*(u32*)(out_buf + pos)) - num);

          }

          break;

        case 9:

          /* Randomly add to dword, random endian. */

          if (temp_len < 4) break;

          if (UR(2)) {

            u32 pos = UR(temp_len - 3);

            *(u32*)(out_buf + pos) += 1 + UR(ARITH_MAX);

          } else {

            u32 pos = UR(temp_len - 3);
            u32 num = 1 + UR(ARITH_MAX);

            *(u32*)(out_buf + pos) =
              SWAP32(SWAP32(*(u32*)(out_buf + pos)) + num);

          }

          break;

        case 10:

          /* Just set a random byte to a random value. Because,
             why not. We use XOR with 1-255 to eliminate the
             possibility of a no-op. */

          out_buf[UR(temp_len)] ^= 1 + UR(255);
          break;

        case 11 ... 12: {

            /* Delete bytes. We're making this a bit more likely
               than insertion (the next option) in hopes of keeping
               files reasonably small. */

            u32 del_from, del_len;

            if (temp_len < 2) break;

            /* Don't delete too much. */

            del_len = choose_block_len(temp_len - 1);

            del_from = UR(temp_len - del_len + 1);

            memmove(out_buf + del_from, out_buf + del_from + del_len,
                    temp_len - del_from - del_len);

            temp_len -= del_len;

            break;

          }

        case 13:

          if (temp_len + HAVOC_BLK_LARGE < MAX_FILE) {

            /* Clone bytes (75%) or insert a block of constant bytes (25%). */

            u32 clone_from, clone_to, clone_len;
            u8* new_buf;

            clone_len  = choose_block_len(temp_len);

            clone_from = UR(temp_len - clone_len + 1);
            clone_to   = UR(temp_len);

            new_buf = ck_alloc_nozero(temp_len + clone_len);

            /* Head */

            memcpy(new_buf, out_buf, clone_to);

            /* Inserted part */

            if (UR(4))
              memcpy(new_buf + clone_to, out_buf + clone_from, clone_len);
            else
              memset(new_buf + clone_to, UR(256), clone_len);

            /* Tail */
            memcpy(new_buf + clone_to + clone_len, out_buf + clone_to,
                   temp_len - clone_to);

            ck_free(out_buf);
            out_buf = new_buf;
            temp_len += clone_len;

          }

          break;

        case 14: {

            /* Overwrite bytes with a randomly selected chunk (75%) or fixed
               bytes (25%). */

            u32 copy_from, copy_to, copy_len;

            if (temp_len < 2) break;

            copy_len  = choose_block_len(temp_len - 1);

            copy_from = UR(temp_len - copy_len + 1);
            copy_to   = UR(temp_len - copy_len + 1);

            if (UR(4)) {

              if (copy_from != copy_to)
                memmove(out_buf + copy_to, out_buf + copy_from, copy_len);

            } else memset(out_buf + copy_to, UR(256), copy_len);

            break;

          }

        /* Values 16 and 17 can be selected only if there are any extras
           present in the dictionaries. */

        case 16: {

            /* Overwrite bytes with an extra. */

            if (!extras_cnt || (a_extras_cnt && UR(2))) {

              /* No user-specified extras or odds in our favor. Let's use an
                 auto-detected one. */

              u32 use_extra = UR(a_extras_cnt);
              u32 extra_len = a_extras[use_extra].len;
              u32 insert_at;

              if (extra_len > temp_len) break;

              insert_at = UR(temp_len - extra_len + 1);
              memcpy(out_buf + insert_at, a_extras[use_extra].data, extra_len);

            } else {

              /* No auto extras or odds in our favor. Use the dictionary. */

              u32 use_extra = UR(extras_cnt);
              u32 extra_len = extras[use_extra].len;
              u32 insert_at;

              if (extra_len > temp_len) break;

              insert_at = UR(temp_len - extra_len + 1);
              memcpy(out_buf + insert_at, extras[use_extra].data, extra_len);

            }

            break;

          }

        case 17: {

            u32 use_extra, extra_len, insert_at = UR(temp_len);
            u8* new_buf;

            /* Insert an extra. Do the same dice-rolling stuff as for the
               previous case. */

            if (!extras_cnt || (a_extras_cnt && UR(2))) {

              use_extra = UR(a_extras_cnt);
              extra_len = a_extras[use_extra].len;

              if (temp_len + extra_len >= MAX_FILE) break;

              new_buf = ck_alloc_nozero(temp_len + extra_len);

              /* Head */
              memcpy(new_buf, out_buf, insert_at);

              /* Inserted part */
              memcpy(new_buf + insert_at, a_extras[use_extra].data, extra_len);

            } else {

              use_extra = UR(extras_cnt);
              extra_len = extras[use_extra].len;

              if (temp_len + extra_len >= MAX_FILE) break;

              new_buf = ck_alloc_nozero(temp_len + extra_len);

              /* Head */
              memcpy(new_buf, out_buf, insert_at);

              /* Inserted part */
              memcpy(new_buf + insert_at, extras[use_extra].data, extra_len);

            }

            /* Tail */
            memcpy(new_buf + insert_at + extra_len, out_buf + insert_at,
                   temp_len - insert_at);

            ck_free(out_buf);
            out_buf   = new_buf;
            temp_len += extra_len;

            break;

          }

      }

    }

    if (common_fuzz_stuff(argv, out_buf, temp_len))
      goto abandon_entry;

    /* out_buf might have been mangled a bit, so let's restore it to its
       original size and shape. */

    if (temp_len < len) out_buf = ck_realloc(out_buf, len);
    temp_len = len;
    memcpy(out_buf, in_buf, len);

    /* If we're finding new stuff, let's run for a bit longer, limits
       permitting. */

    if (queued_paths != havoc_queued) {

      if (perf_score <= HAVOC_MAX_MULT * 100) {
        stage_max  *= 2;
        perf_score *= 2;
      }

      havoc_queued = queued_paths;

    }

  }

  new_hit_cnt = queued_paths + unique_crashes;

  if (!splice_cycle) {
    stage_finds[STAGE_HAVOC]  += new_hit_cnt - orig_hit_cnt;
    stage_cycles[STAGE_HAVOC] += stage_max;
  } else {
    stage_finds[STAGE_SPLICE]  += new_hit_cnt - orig_hit_cnt;
    stage_cycles[STAGE_SPLICE] += stage_max;
  }

#ifndef IGNORE_FINDS

  /************
   * SPLICING *
   ************/

  /* This is a last-resort strategy triggered by a full round with no findings.
     It takes the current input file, randomly selects another input, and
     splices them together at some offset, then relies on the havoc
     code to mutate that blob. */

retry_splicing:

  if (use_splicing && splice_cycle++ < SPLICE_CYCLES &&
      queued_paths > 1 && queue_cur->len > 1) {

    struct queue_entry* target;
    u32 tid, split_at;
    u8* new_buf;
    s32 f_diff, l_diff;

    /* First of all, if we've modified in_buf for havoc, let's clean that
       up... */

    if (in_buf != orig_in) {
      ck_free(in_buf);
      in_buf = orig_in;
      len = queue_cur->len;
    }

    /* Pick a random queue entry and seek to it. Don't splice with yourself. */

    do { tid = UR(queued_paths); } while (tid == current_entry);

    splicing_with = tid;
    target = queue;

    while (tid >= 100) { target = target->next_100; tid -= 100; }
    while (tid--) target = target->next;

    /* Make sure that the target has a reasonable length. */

    while (target && (target->len < 2 || target == queue_cur)) {
      target = target->next;
      splicing_with++;
    }

    if (!target) goto retry_splicing;

    /* Read the testcase into a new buffer. */

    fd = open(target->fname, O_RDONLY);

    if (fd < 0) PFATAL("Unable to open '%s'", target->fname);

    new_buf = ck_alloc_nozero(target->len);

    ck_read(fd, new_buf, target->len, target->fname);

    close(fd);

    /* Find a suitable splicing location, somewhere between the first and
       the last differing byte. Bail out if the difference is just a single
       byte or so. */

    locate_diffs(in_buf, new_buf, MIN(len, target->len), &f_diff, &l_diff);

    if (f_diff < 0 || l_diff < 2 || f_diff == l_diff) {
      ck_free(new_buf);
      goto retry_splicing;
    }

    /* Split somewhere between the first and last differing byte. */

    split_at = f_diff + UR(l_diff - f_diff);

    /* Do the thing. */

    len = target->len;
    memcpy(new_buf, in_buf, split_at);
    in_buf = new_buf;

    ck_free(out_buf);
    out_buf = ck_alloc_nozero(len);
    memcpy(out_buf, in_buf, len);

    goto havoc_stage;

  }

#endif /* !IGNORE_FINDS */

  ret_val = 0;

abandon_entry:

  splicing_with = -1;

  /* Update pending_not_fuzzed count if we made it through the calibration
     cycle and have not seen this entry before. */

  if (!stop_soon && !queue_cur->cal_failed && !queue_cur->was_fuzzed) {
    queue_cur->was_fuzzed = 1;
    pending_not_fuzzed--;
    if (queue_cur->favored) pending_favored--;
  }

  munmap(orig_in, queue_cur->len);

  if (in_buf != orig_in) ck_free(in_buf);
  ck_free(out_buf);
  ck_free(eff_map);

  return ret_val;

#undef FLIP_BIT

}


/* Grab interesting test cases from other fuzzers. */

static void sync_fuzzers(char** argv) {

  DIR* sd;
  struct dirent* sd_ent;
  u32 sync_cnt = 0;

  sd = opendir(sync_dir);
  if (!sd) PFATAL("Unable to open '%s'", sync_dir);

  stage_max = stage_cur = 0;
  cur_depth = 0;

  /* Look at the entries created for every other fuzzer in the sync directory. */

  while ((sd_ent = readdir(sd))) {

    static u8 stage_tmp[128];

    DIR* qd;
    struct dirent* qd_ent;
    u8 *qd_path, *qd_synced_path;
    u32 min_accept = 0, next_min_accept;

    s32 id_fd;

    /* Skip dot files and our own output directory. */

    if (sd_ent->d_name[0] == '.' || !strcmp(sync_id, sd_ent->d_name)) continue;

    /* Skip anything that doesn't have a queue/ subdirectory. */

    qd_path = alloc_printf("%s/%s/queue", sync_dir, sd_ent->d_name);

    if (!(qd = opendir(qd_path))) {
      ck_free(qd_path);
      continue;
    }

    /* Retrieve the ID of the last seen test case. */

    qd_synced_path = alloc_printf("%s/.synced/%s", out_dir, sd_ent->d_name);

    id_fd = open(qd_synced_path, O_RDWR | O_CREAT, 0600);

    if (id_fd < 0) PFATAL("Unable to create '%s'", qd_synced_path);

    if (read(id_fd, &min_accept, sizeof(u32)) > 0) 
      lseek(id_fd, 0, SEEK_SET);

    next_min_accept = min_accept;

    /* Show stats */    

    sprintf(stage_tmp, "sync %u", ++sync_cnt);
    stage_name = stage_tmp;
    stage_cur  = 0;
    stage_max  = 0;

    /* For every file queued by this fuzzer, parse ID and see if we have looked at
       it before; exec a test case if not. */

    while ((qd_ent = readdir(qd))) {

      u8* path;
      s32 fd;
      struct stat st;

      if (qd_ent->d_name[0] == '.' ||
          sscanf(qd_ent->d_name, CASE_PREFIX "%06u", &syncing_case) != 1 || 
          syncing_case < min_accept) continue;

      /* OK, sounds like a new one. Let's give it a try. */

      if (syncing_case >= next_min_accept)
        next_min_accept = syncing_case + 1;

      path = alloc_printf("%s/%s", qd_path, qd_ent->d_name);

      fd = open(path, O_RDONLY);
      if (fd < 0) PFATAL("Unable to open '%s'", path);

      if (fstat(fd, &st)) PFATAL("fstat() failed");

      /* Ignore zero-sized or oversized files. */

      if (st.st_size && st.st_size <= MAX_FILE) {

        u8  fault;
        u8* mem = mmap(0, st.st_size, PROT_READ, MAP_PRIVATE, fd, 0);

        if (mem == MAP_FAILED) PFATAL("Unable to mmap '%s'", path);

        /* See what happens. We rely on save_if_interesting() to catch major
           errors and save the test case. */

        write_to_testcase(mem, st.st_size);

        fault = run_target(argv);

        if (stop_soon) return;

        syncing_party = sd_ent->d_name;
        queued_imported += save_if_interesting(argv, mem, st.st_size, fault);
        syncing_party = 0;

        munmap(mem, st.st_size);

        if (!(stage_cur++ % stats_update_freq)) show_stats();

      }

      ck_free(path);
      close(fd);

    }

    ck_write(id_fd, &next_min_accept, sizeof(u32), qd_synced_path);

    close(id_fd);
    closedir(qd);
    ck_free(qd_path);
    ck_free(qd_synced_path);
    
  }  

  closedir(sd);

}


/* Handle stop signal (Ctrl-C, etc). */

static void handle_stop_sig(int sig) {

  stop_soon = 1; 

  if (child_pid > 0) kill(child_pid, SIGKILL);
  if (forksrv_pid > 0) kill(forksrv_pid, SIGKILL);

}


/* Handle skip request (SIGUSR1). */

static void handle_skipreq(int sig) {

  skip_requested = 1;

}

/* Handle timeout (SIGALRM). */

static void handle_timeout(int sig) {

  if (child_pid > 0) {

    child_timed_out = 1; 
    kill(child_pid, SIGKILL);

  } else if (child_pid == -1 && forksrv_pid > 0) {

    child_timed_out = 1; 
    kill(forksrv_pid, SIGKILL);

  }

}


/* Do a PATH search and find target binary to see that it exists and
   isn't a shell script - a common and painful mistake. We also check for
   a valid ELF header and for evidence of AFL instrumentation. */

static void check_binary(u8* fname) {

  u8* env_path = 0;
  struct stat st;

  s32 fd;
  u8* f_data;
  u32 f_len = 0;

  ACTF("Validating target binary...");

  if (strchr(fname, '/') || !(env_path = getenv("PATH"))) {

    target_path = ck_strdup(fname);
    if (stat(target_path, &st) || !S_ISREG(st.st_mode) ||
        !(st.st_mode & 0111) || (f_len = st.st_size) < 4)
      FATAL("Program '%s' not found or not executable", fname);

  } else {

    while (env_path) {

      u8 *cur_elem, *delim = strchr(env_path, ':');

      if (delim) {

        cur_elem = ck_alloc(delim - env_path + 1);
        memcpy(cur_elem, env_path, delim - env_path);
        delim++;

      } else cur_elem = ck_strdup(env_path);

      env_path = delim;

      if (cur_elem[0])
        target_path = alloc_printf("%s/%s", cur_elem, fname);
      else
        target_path = ck_strdup(fname);

      ck_free(cur_elem);

      if (!stat(target_path, &st) && S_ISREG(st.st_mode) &&
          (st.st_mode & 0111) && (f_len = st.st_size) >= 4) break;

      ck_free(target_path);
      target_path = 0;

    }

    if (!target_path) FATAL("Program '%s' not found or not executable", fname);

  }

  if (getenv("AFL_SKIP_BIN_CHECK")) return;

  /* Check for blatant user errors. */

  if ((!strncmp(target_path, "/tmp/", 5) && !strchr(target_path + 5, '/')) ||
      (!strncmp(target_path, "/var/tmp/", 9) && !strchr(target_path + 9, '/')))
     FATAL("Please don't keep binaries in /tmp or /var/tmp");

  fd = open(target_path, O_RDONLY);

  if (fd < 0) PFATAL("Unable to open '%s'", target_path);

  f_data = mmap(0, f_len, PROT_READ, MAP_PRIVATE, fd, 0);

  if (f_data == MAP_FAILED) PFATAL("Unable to mmap file '%s'", target_path);

  close(fd);

  if (f_data[0] == '#' && f_data[1] == '!') {

    SAYF("\n" cLRD "[-] " cRST
         "Oops, the target binary looks like a shell script. Some build systems will\n"
         "    sometimes generate shell stubs for dynamically linked programs; try static\n"
         "    library mode (./configure --disable-shared) if that's the case.\n\n"

         "    Another possible cause is that you are actually trying to use a shell\n" 
         "    wrapper around the fuzzed component. Invoking shell can slow down the\n" 
         "    fuzzing process by a factor of 20x or more; it's best to write the wrapper\n"
         "    in a compiled language instead.\n");

    FATAL("Program '%s' is a shell script", target_path);

  }

#ifndef __APPLE__

  if (f_data[0] != 0x7f || memcmp(f_data + 1, "ELF", 3))
    FATAL("Program '%s' is not an ELF binary", target_path);

#else

  if (f_data[0] != 0xCF || f_data[1] != 0xFA || f_data[2] != 0xED)
    FATAL("Program '%s' is not a 64-bit Mach-O binary", target_path);

#endif /* ^!__APPLE__ */

  if (!qemu_mode && !dumb_mode &&
      !memmem(f_data, f_len, SHM_ENV_VAR, strlen(SHM_ENV_VAR) + 1)) {

    SAYF("\n" cLRD "[-] " cRST
         "Looks like the target binary is not instrumented! The fuzzer depends on\n"
         "    compile-time instrumentation to isolate interesting test cases while\n"
         "    mutating the input data. For more information, and for tips on how to\n"
         "    instrument binaries, please see %s/README.\n\n"

         "    When source code is not available, you may be able to leverage QEMU\n"
         "    mode support. Consult the README for tips on how to enable this.\n"

         "    (It is also possible to use afl-fuzz as a traditional, \"dumb\" fuzzer.\n"
         "    For that, you can use the -n option - but expect much worse results.)\n",
         doc_path);

    FATAL("No instrumentation detected");

  }

  if (qemu_mode &&
      memmem(f_data, f_len, SHM_ENV_VAR, strlen(SHM_ENV_VAR) + 1)) {

    SAYF("\n" cLRD "[-] " cRST
         "This program appears to be instrumented with afl-gcc, but is being run in\n"
         "    QEMU mode (-Q). This is probably not what you want - this setup will be\n"
         "    slow and offer no practical benefits.\n");

    FATAL("Instrumentation found in -Q mode");

  }

  if (memmem(f_data, f_len, "libasan.so", 10) ||
      memmem(f_data, f_len, "__msan_init", 11)) uses_asan = 1;

  /* Detect persistent & deferred init signatures in the binary. */

  if (memmem(f_data, f_len, PERSIST_SIG, strlen(PERSIST_SIG) + 1)) {

    OKF(cPIN "Persistent mode binary detected.");
    setenv(PERSIST_ENV_VAR, "1", 1);
    no_var_check = 1;

  }

  if (memmem(f_data, f_len, DEFER_SIG, strlen(DEFER_SIG) + 1)) {

    OKF(cPIN "Deferred forkserver binary detected.");
    setenv(DEFER_ENV_VAR, "1", 1);

  }

  if (munmap(f_data, f_len)) PFATAL("unmap() failed");

}


/* Trim and possibly create a banner for the run. */

static void fix_up_banner(u8* name) {

  if (!use_banner) {

    if (sync_id) {

      use_banner = sync_id;

    } else {

      u8* trim = strrchr(name, '/');
      if (!trim) use_banner = name; else use_banner = trim + 1;

    }

  }

  if (strlen(use_banner) > 40) {

    u8* tmp = ck_alloc(44);
    sprintf(tmp, "%.40s...", use_banner);
    use_banner = tmp;

  }

}


/* Check if we're on TTY. */

static void check_if_tty(void) {

  struct winsize ws;

  if (ioctl(1, TIOCGWINSZ, &ws)) {

    if (errno == ENOTTY) {
      OKF("Looks like we're not running on a tty, so I'll be a bit less verbose.");
      not_on_tty = 1;
    }

    return;
  }

}


/* Check terminal dimensions after resize. */

static void check_term_size(void) {

  struct winsize ws;

  term_too_small = 0;

  if (ioctl(1, TIOCGWINSZ, &ws)) return;

  if (ws.ws_row < 25 || ws.ws_col < 80) term_too_small = 1;

}



/* Display usage hints. */

static void usage(u8* argv0) {

  SAYF("\n%s [ options ] -- /path/to/fuzzed_app [ ... ]\n\n"

       "Required parameters:\n\n"

       "  -i dir        - input directory with test cases\n"
       "  -o dir        - output directory for fuzzer findings\n\n"

       "Execution control settings:\n\n"

       "  -f file       - location read by the fuzzed program (stdin)\n"
       "  -t msec       - timeout for each run (auto-scaled, 50-%u ms)\n"
       "  -m megs       - memory limit for child process (%u MB)\n"
       "  -Q            - use binary-only instrumentation (QEMU mode)\n\n"     
 
       "Fuzzing behavior settings:\n\n"

       "  -d            - quick & dirty mode (skips deterministic steps)\n"
       "  -n            - fuzz without instrumentation (dumb mode)\n"
       "  -x dir        - optional fuzzer dictionary (see README)\n\n"

       "Other stuff:\n\n"

       "  -T text       - text banner to show on the screen\n"
       "  -M / -S id    - distributed mode (see parallel_fuzzing.txt)\n"
       "  -C            - crash exploration mode (the peruvian rabbit thing)\n\n"

       "For additional tips, please consult %s/README.\n\n",

       argv0, EXEC_TIMEOUT, MEM_LIMIT, doc_path);

  exit(1);

}


/* Prepare output directories and fds. */

static void setup_dirs_fds(void) {

  u8* tmp;
  s32 fd;

  ACTF("Setting up output directories...");

  if (sync_id && mkdir(sync_dir, 0700) && errno != EEXIST)
      PFATAL("Unable to create '%s'", sync_dir);

  if (mkdir(out_dir, 0700)) {

    if (errno != EEXIST) PFATAL("Unable to create '%s'", out_dir);

    maybe_delete_out_dir();

  } else {

    if (in_place_resume)
      FATAL("Resume attempted but old output directory not found");

    out_dir_fd = open(out_dir, O_RDONLY);

#ifndef __sun

    if (out_dir_fd < 0 || flock(out_dir_fd, LOCK_EX | LOCK_NB))
      PFATAL("Unable to flock() output directory.");

#endif /* !__sun */

  }

  /* Queue directory for any starting & discovered paths. */

  tmp = alloc_printf("%s/queue", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* Top-level directory for queue metadata used for session
     resume and related tasks. */

  tmp = alloc_printf("%s/queue/.state/", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* Directory for flagging queue entries that went through
     deterministic fuzzing in the past. */

  tmp = alloc_printf("%s/queue/.state/deterministic_done/", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* Directory with the auto-selected dictionary entries. */

  tmp = alloc_printf("%s/queue/.state/auto_extras/", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* The set of paths currently deemed redundant. */

  tmp = alloc_printf("%s/queue/.state/redundant_edges/", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* The set of paths showing variable behavior. */

  tmp = alloc_printf("%s/queue/.state/variable_behavior/", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* Sync directory for keeping track of cooperating fuzzers. */

  if (sync_id) {

    tmp = alloc_printf("%s/.synced/", out_dir);
    if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
    ck_free(tmp);

  }

  /* All recorded crashes. */

  tmp = alloc_printf("%s/crashes", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* All recorded hangs. */

  tmp = alloc_printf("%s/hangs", out_dir);
  if (mkdir(tmp, 0700)) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  /* Generally useful file descriptors. */

  dev_null_fd = open("/dev/null", O_RDWR);
  if (dev_null_fd < 0) PFATAL("Unable to open /dev/null");

  dev_urandom_fd = open("/dev/urandom", O_RDONLY);
  if (dev_urandom_fd < 0) PFATAL("Unable to open /dev/urandom");

  /* Gnuplot output file. */

  tmp = alloc_printf("%s/plot_data", out_dir);
  fd = open(tmp, O_WRONLY | O_CREAT | O_EXCL, 0600);
  if (fd < 0) PFATAL("Unable to create '%s'", tmp);
  ck_free(tmp);

  plot_file = fdopen(fd, "w");
  if (!plot_file) PFATAL("fdopen() failed");

  fprintf(plot_file, "# unix_time, cycles_done, cur_path, paths_total, "
                     "pending_total, pending_favs, map_size, unique_crashes, "
                     "unique_hangs, max_depth, execs_per_sec\n");
                     /* ignore errors */

}


/* Setup the output file for fuzzed data, if not using -f. */

static void setup_stdio_file(void) {

  u8* fn = alloc_printf("%s/.cur_input", out_dir);

  unlink(fn); /* Ignore errors */

  out_fd = open(fn, O_RDWR | O_CREAT | O_EXCL, 0600);

  if (out_fd < 0) PFATAL("Unable to create '%s'", fn);

  ck_free(fn);

}


/* Make sure that core dumps don't go to a program. */

static void check_crash_handling(void) {

#ifdef __APPLE__

  /* Yuck! There appears to be no simple C API to query for the state of 
     loaded daemons on MacOS X, and I'm a bit hesitant to do something
     more sophisticated, such as disabling crash reporting via Mach ports,
     until I get a box to test the code. So, for now, we check for crash
     reporting the awful way. */
  
  if (system("launchctl list 2>/dev/null | grep -q '\\.ReportCrash$'")) return;

  SAYF("\n" cLRD "[-] " cRST
       "Whoops, your system is configured to forward crash notifications to an\n"
       "    external crash reporting utility. This will cause issues due to the\n"
       "    extended delay between the fuzzed binary malfunctioning and this fact\n"
       "    being relayed to the fuzzer via the standard waitpid() API.\n\n"
       "    To avoid having crashes misinterpreted as hangs, please run the\n" 
       "    following commands:\n\n"

       "    SL=/System/Library; PL=com.apple.ReportCrash\n"
       "    launchctl unload -w ${SL}/LaunchAgents/${PL}.plist\n"
       "    sudo launchctl unload -w ${SL}/LaunchDaemons/${PL}.Root.plist\n");

  if (!getenv("AFL_I_DONT_CARE_ABOUT_MISSING_CRASHES"))
    FATAL("Crash reporter detected");

#else

  /* This is Linux specific, but I don't think there's anything equivalent on
     *BSD, so we can just let it slide for now. */

  s32 fd = open("/proc/sys/kernel/core_pattern", O_RDONLY);
  u8  fchar;

  if (fd < 0) return;

  ACTF("Checking core_pattern...");

  if (read(fd, &fchar, 1) == 1 && fchar == '|') {

    SAYF("\n" cLRD "[-] " cRST
         "Hmm, your system is configured to send core dump notifications to an\n"
         "    external utility. This will cause issues due to an extended delay\n"
         "    between the fuzzed binary malfunctioning and this information being\n"
         "    eventually relayed to the fuzzer via the standard waitpid() API.\n\n"

         "    To avoid having crashes misinterpreted as hangs, please log in as root\n" 
         "    and temporarily modify /proc/sys/kernel/core_pattern, like so:\n\n"

         "    echo core >/proc/sys/kernel/core_pattern\n");

    if (!getenv("AFL_I_DONT_CARE_ABOUT_MISSING_CRASHES"))
      FATAL("Pipe at the beginning of 'core_pattern'");

  }
 
  close(fd);

#endif /* ^__APPLE__ */

}


/* Check CPU governor. */

static void check_cpu_governor(void) {

  FILE* f;
  u8 tmp[128];
  u64 min = 0, max = 0;

  if (getenv("AFL_SKIP_CPUFREQ")) return;

  f = fopen("/sys/devices/system/cpu/cpu0/cpufreq/scaling_governor", "r");
  if (!f) return;

  ACTF("Checking CPU scaling governor...");

  if (!fgets(tmp, 128, f)) PFATAL("fgets() failed");

  fclose(f);

  if (!strncmp(tmp, "perf", 4)) return;

  f = fopen("/sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq", "r");

  if (f) {
    if (fscanf(f, "%llu", &min) != 1) min = 0;
    fclose(f);
  }

  f = fopen("/sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq", "r");

  if (f) {
    if (fscanf(f, "%llu", &max) != 1) max = 0;
    fclose(f);
  }

  if (min == max) return;

  SAYF("\n" cLRD "[-] " cRST
       "Whoops, your system uses on-demand CPU frequency scaling, adjusted\n"
       "    between %llu and %llu MHz. Unfortunately, the scaling algorithm in the\n"
       "    kernel is imperfect and can miss the short-lived processes spawned by\n"
       "    afl-fuzz. To keep things moving, run these commands as root:\n\n"

       "    cd /sys/devices/system/cpu\n"
       "    echo performance | tee cpu*/cpufreq/scaling_governor\n\n"

       "    You can later go back to the original state by replacing 'performance' with\n"
       "    'ondemand'. If you don't want to change the settings, set AFL_SKIP_CPUFREQ\n"
       "    to make afl-fuzz skip this check - but expect some performance drop.\n",
       min / 1024, max / 1024);

  FATAL("Suboptimal CPU scaling governor");

}


/* Count the number of logical CPU cores. */

static void get_core_count(void) {

  u32 cur_runnable = 0;

#if defined(__APPLE__) || defined(__FreeBSD__) || defined (__OpenBSD__)

  size_t s = sizeof(cpu_core_count);

  /* On *BSD systems, we can just use a sysctl to get the number of CPUs. */

#ifdef __APPLE__

  if (sysctlbyname("hw.logicalcpu", &cpu_core_count, &s, NULL, 0) < 0)
    return;

#else

  int s_name[2] = { CTL_HW, HW_NCPU };

  if (sysctl(s_name, 2, &cpu_core_count, &s, NULL, 0) < 0) return;

#endif /* ^__APPLE__ */

#else

  /* On Linux, a simple way is to look at /proc/stat, especially since we'd
     be parsing it anyway for other reasons later on. */

  FILE* f = fopen("/proc/stat", "r");
  u8 tmp[1024];

  if (!f) return;

  while (fgets(tmp, sizeof(tmp), f))
    if (!strncmp(tmp, "cpu", 3) && isdigit(tmp[3])) cpu_core_count++;

  fclose(f);
  
#endif /* ^(__APPLE__ || __FreeBSD__ || __OpenBSD__) */

  if (cpu_core_count) {

    cur_runnable = (u32)get_runnable_processes();

#if defined(__APPLE__) || defined(__FreeBSD__) || defined (__OpenBSD__)

    /* Add ourselves, since the 1-minute average doesn't include that yet. */

    cur_runnable++;

#endif /* __APPLE__ || __FreeBSD__ || __OpenBSD__ */

    OKF("You have %u CPU cores and %u runnable tasks (utilization: %0.0f%%).",
        cpu_core_count, cur_runnable, cur_runnable * 100.0 / cpu_core_count);

    if (cpu_core_count > 1) {

      if (cur_runnable > cpu_core_count * 1.5) {

        WARNF("System under apparent load, performance may be spotty.");

      } else if (cur_runnable + 1 <= cpu_core_count) {

        OKF("Try parallel jobs - see %s/parallel_fuzzing.txt.", doc_path);
  
      }

    }

  } else WARNF("Unable to figure out the number of CPU cores.");

}


/* Validate and fix up out_dir and sync_dir when using -S. */

static void fix_up_sync(void) {

  u8* x = sync_id;

  if (dumb_mode)
    FATAL("-S / -M and -n are mutually exclusive");

  if (skip_deterministic) {

    if (force_deterministic)
      FATAL("use -S instead of -M -d");
    else
      FATAL("-S already implies -d");

  }

  while (*x) {

    if (!isalnum(*x) && *x != '_' && *x != '-')
      FATAL("Non-alphanumeric fuzzer ID specified via -S or -M");

    x++;

  }

  if (strlen(sync_id) > 32) FATAL("Fuzzer ID too long");

  x = alloc_printf("%s/%s", out_dir, sync_id);

  sync_dir = out_dir;
  out_dir  = x;

  if (!force_deterministic) {
    skip_deterministic = 1;
    use_splicing = 1;
  }

}


/* Handle screen resize (SIGWINCH). */

static void handle_resize(int sig) {
  clear_screen = 1;
}


/* Check ASAN options. */

static void check_asan_opts(void) {
  u8* x = getenv("ASAN_OPTIONS");

  if (x && !strstr(x, "abort_on_error=1"))
    FATAL("Custom ASAN_OPTIONS set without abort_on_error=1 - please fix!");

  x = getenv("MSAN_OPTIONS");

  if (x && !strstr(x, "exit_code=" STRINGIFY(MSAN_ERROR)))
    FATAL("Custom MSAN_OPTIONS set without exit_code="
          STRINGIFY(MSAN_ERROR) " - please fix!");

} 


/* Detect @@ in args. */

static void detect_file_args(char** argv) {

  u32 i = 0;
  u8* cwd = getcwd(NULL, 0);

  if (!cwd) PFATAL("getcwd() failed");

  while (argv[i]) {

    u8* aa_loc = strstr(argv[i], "@@");

    if (aa_loc) {

      u8 *aa_subst, *n_arg;

      /* If we don't have a file name chosen yet, use a safe default. */

      if (!out_file)
        out_file = alloc_printf("%s/.cur_input", out_dir);

      /* Be sure that we're always using fully-qualified paths. */

      if (out_file[0] == '/') aa_subst = out_file;
      else aa_subst = alloc_printf("%s/%s", cwd, out_file);

      /* Construct a replacement argv value. */

      *aa_loc = 0;
      n_arg = alloc_printf("%s%s%s", argv[i], aa_subst, aa_loc + 2);
      argv[i] = n_arg;
      *aa_loc = '@';

      if (out_file[0] != '/') ck_free(aa_subst);

    }

    i++;

  }

  free(cwd); /* not tracked */

}


/* Set up signal handlers. More complicated that needs to be, because libc on
   Solaris doesn't resume interrupted reads(), sets SA_RESETHAND when you call
   siginterrupt(), and does other stupid things. */

static void setup_signal_handlers(void) {

  struct sigaction sa;

  sa.sa_handler   = NULL;
  sa.sa_flags     = SA_RESTART;
  sa.sa_sigaction = NULL;

  sigemptyset(&sa.sa_mask);

  /* Various ways of saying "stop". */

  sa.sa_handler = handle_stop_sig;
  sigaction(SIGHUP, &sa, NULL);
  sigaction(SIGINT, &sa, NULL);
  sigaction(SIGTERM, &sa, NULL);

  /* Exec timeout notifications. */

  sa.sa_handler = handle_timeout;
  sigaction(SIGALRM, &sa, NULL);

  /* Window resize */

  sa.sa_handler = handle_resize;
  sigaction(SIGWINCH, &sa, NULL);

  /* SIGUSR1: skip entry */

  sa.sa_handler = handle_skipreq;
  sigaction(SIGUSR1, &sa, NULL);

  /* Things we don't care about. */

  sa.sa_handler = SIG_IGN;
  sigaction(SIGTSTP, &sa, NULL);
  sigaction(SIGPIPE, &sa, NULL);

}


/* Rewrite argv for QEMU. */

static char** get_qemu_argv(u8* own_loc, char** argv, int argc) {

  char** new_argv = ck_alloc(sizeof(char*) * (argc + 4));
  u8 *tmp, *cp, *rsl, *own_copy;

  memcpy(new_argv + 3, argv + 1, sizeof(char*) * argc);

  new_argv[2] = target_path;
  new_argv[1] = "--";

  /* Now we need to actually find the QEMU binary to put in argv[0]. */

  tmp = getenv("AFL_PATH");

  if (tmp) {

    cp = alloc_printf("%s/afl-qemu-trace", tmp);

    if (access(cp, X_OK))
      FATAL("Unable to find '%s'", tmp);

    target_path = new_argv[0] = cp;
    return new_argv;

  }

  own_copy = ck_strdup(own_loc);
  rsl = strrchr(own_copy, '/');

  if (rsl) {

    *rsl = 0;

    cp = alloc_printf("%s/afl-qemu-trace", own_copy);
    ck_free(own_copy);

    if (!access(cp, X_OK)) {

      target_path = new_argv[0] = cp;
      return new_argv;

    }

  } else ck_free(own_copy);

  if (!access(BIN_PATH "/afl-qemu-trace", X_OK)) {

    target_path = new_argv[0] = ck_strdup(BIN_PATH "/afl-qemu-trace");
    return new_argv;

  }

  SAYF("\n" cLRD "[-] " cRST
       "Oops, unable to find the 'afl-qemu-trace' binary. The binary must be built\n"
       "    separately by following the instructions in qemu_mode/README.qemu. If you\n"
       "    already have the binary installed, you may need to specify AFL_PATH in the\n"
       "    environment.\n\n"

       "    Of course, even without QEMU, afl-fuzz can still work with binaries that are\n"
       "    instrumented at compile time with afl-gcc. It is also possible to use it as a\n"
       "    traditional \"dumb\" fuzzer by specifying '-n' in the command line.\n");

  FATAL("Failed to locate 'afl-qemu-trace'.");

}


/* Make a copy of the current command line. */

static void save_cmdline(u32 argc, char** argv) {

  u32 len = 1, i;
  u8* buf;

  for (i = 0; i < argc; i++)
    len += strlen(argv[i]) + 1;
  
  buf = orig_cmdline = ck_alloc(len);

  for (i = 0; i < argc; i++) {

    u32 l = strlen(argv[i]);

    memcpy(buf, argv[i], l);
    buf += l;

    if (i != argc - 1) *(buf++) = ' ';

  }

  *buf = 0;

}


/* Main entry point */

int main(int argc, char** argv) {

  s32 opt;
  u64 prev_queued = 0;
  u32 sync_interval_cnt = 0, seek_to;
  u8  *extras_dir = 0;
  u8  mem_limit_given = 0;

  char** use_argv;

  SAYF(cCYA "afl-fuzz " cBRI VERSION cRST " by <lcamtuf@google.com>\n");

  doc_path = access(DOC_PATH, F_OK) ? "docs" : DOC_PATH;

  while ((opt = getopt(argc, argv, "+i:o:f:m:t:T:dnCB:S:M:x:Q")) > 0)

    switch (opt) {

      case 'i':

        if (in_dir) FATAL("Multiple -i options not supported");
        in_dir = optarg;

        if (!strcmp(in_dir, "-")) in_place_resume = 1;

        break;

      case 'o': /* output dir */

        if (out_dir) FATAL("Multiple -o options not supported");
        out_dir = optarg;
        break;

      case 'M':

        force_deterministic = 1;
        /* Fall through */

      case 'S': /* sync ID */

        if (sync_id) FATAL("Multiple -S or -M options not supported");
        sync_id = optarg;
        break;

      case 'f': /* target file */

        if (out_file) FATAL("Multiple -f options not supported");
        out_file = optarg;
        break;

      case 'x':

        if (extras_dir) FATAL("Multiple -x options not supported");
        extras_dir = optarg;
        break;

      case 't': {

          u8 suffix = 0;

          if (timeout_given) FATAL("Multiple -t options not supported");

          if (sscanf(optarg, "%u%c", &exec_tmout, &suffix) < 1 ||
              optarg[0] == '-') FATAL("Bad syntax used for -t");

          if (exec_tmout < 5) FATAL("Dangerously low value of -t");

          if (suffix == '+') timeout_given = 2; else timeout_given = 1;

          break;

      }

      case 'm': {

          u8 suffix = 'M';

          if (mem_limit_given) FATAL("Multiple -m options not supported");
          mem_limit_given = 1;

          if (!strcmp(optarg, "none")) {

            mem_limit = 0;
            break;

          }

          if (sscanf(optarg, "%llu%c", &mem_limit, &suffix) < 1 ||
              optarg[0] == '-') FATAL("Bad syntax used for -m");

          switch (suffix) {

            case 'T': mem_limit *= 1024 * 1024; break;
            case 'G': mem_limit *= 1024; break;
            case 'k': mem_limit /= 1024; break;
            case 'M': break;

            default:  FATAL("Unsupported suffix or bad syntax for -m");

          }

          if (mem_limit < 5) FATAL("Dangerously low value of -m");

          if (sizeof(rlim_t) == 4 && mem_limit > 2000)
            FATAL("Value of -m out of range on 32-bit systems");

        }

        break;

      case 'd':

        if (skip_deterministic) FATAL("Multiple -d options not supported");
        skip_deterministic = 1;
        use_splicing = 1;
        break;

      case 'B':

        /* This is a secret undocumented option! It is useful if you find
           an interesting test case during a normal fuzzing process, and want
           to mutate it without rediscovering any of the test cases already
           found during an earlier run.

           To use this mode, you need to point -B to the fuzz_bitmap produced
           by an earlier run for the exact same binary... and that's it.

           I only used this once or twice to get variants of a particular
           file, so I'm not making this an official setting. */

        if (in_bitmap) FATAL("Multiple -B options not supported");

        in_bitmap = optarg;
        read_bitmap(in_bitmap);
        break;

      case 'C':

        if (crash_mode) FATAL("Multiple -C options not supported");
        crash_mode = FAULT_CRASH;
        break;

      case 'n':

        if (dumb_mode) FATAL("Multiple -n options not supported");
        if (getenv("AFL_DUMB_FORKSRV")) dumb_mode = 2 ; else dumb_mode = 1;

        break;

      case 'T':

        if (use_banner) FATAL("Multiple -T options not supported");
        use_banner = optarg;
        break;

      case 'Q':

        if (qemu_mode) FATAL("Multiple -Q options not supported");
        qemu_mode = 1;

        if (!mem_limit_given) mem_limit = MEM_LIMIT_QEMU;

        break;

      default:

        usage(argv[0]);

    }

  if (optind == argc || !in_dir || !out_dir) usage(argv[0]);

  setup_signal_handlers();
  check_asan_opts();

  if (sync_id) fix_up_sync();

  if (!strcmp(in_dir, out_dir))
    FATAL("Input and output directories can't be the same");

  if (dumb_mode) {

    if (crash_mode) FATAL("-C and -n are mutually exclusive");
    if (qemu_mode)  FATAL("-Q and -n are mutually exclusive");

  }

  if (getenv("AFL_NO_FORKSRV"))   no_forkserver    = 1;
  if (getenv("AFL_NO_CPU_RED"))   no_cpu_meter_red = 1;
  if (getenv("AFL_NO_VAR_CHECK")) no_var_check     = 1;

  if (dumb_mode == 2 && no_forkserver)
    FATAL("AFL_DUMB_FORKSRV and AFL_NO_FORKSRV are mutually exclusive");

  save_cmdline(argc, argv);

  fix_up_banner(argv[optind]);

  check_if_tty();

  get_core_count();
  check_crash_handling();
  check_cpu_governor();

  setup_post();
  setup_shm();

  setup_dirs_fds();
  read_testcases();
  load_auto();

  pivot_inputs();

  if (extras_dir) load_extras(extras_dir);

  if (!timeout_given) find_timeout();

  detect_file_args(argv + optind + 1);

  if (!out_file) setup_stdio_file();

  check_binary(argv[optind]);

  start_time = get_cur_time();

  if (qemu_mode)
    use_argv = get_qemu_argv(argv[0], argv + optind, argc - optind);
  else
    use_argv = argv + optind;

  perform_dry_run(use_argv);

  cull_queue();

  show_init_stats();

  seek_to = find_start_position();

  write_stats_file(0, 0);
  save_auto();

  if (stop_soon) goto stop_fuzzing;

  /* Woop woop woop */

  if (!not_on_tty) {
    sleep(4);
    start_time += 4000;
    if (stop_soon) goto stop_fuzzing;
  }

  while (1) {

    u8 skipped_fuzz;

    cull_queue();

    if (!queue_cur) {

      queue_cycle++;
      current_entry     = 0;
      cur_skipped_paths = 0;
      queue_cur         = queue;

      while (seek_to) {
        current_entry++;
        seek_to--;
        queue_cur = queue_cur->next;
      }

      show_stats();

      if (not_on_tty) {
        ACTF("Entering queue cycle %llu.", queue_cycle);
        fflush(stdout);
      }

      /* If we had a full queue cycle with no new finds, try
         recombination strategies next. */

      if (queued_paths == prev_queued) {

        if (use_splicing) cycles_wo_finds++; else use_splicing = 1;

      } else cycles_wo_finds = 0;

      prev_queued = queued_paths;

      if (sync_id && queue_cycle == 1 && getenv("AFL_IMPORT_FIRST"))
        sync_fuzzers(use_argv);

    }

    skipped_fuzz = fuzz_one(use_argv);

    if (!stop_soon && sync_id && !skipped_fuzz) {
      
      if (!(sync_interval_cnt++ % SYNC_INTERVAL))
        sync_fuzzers(use_argv);

    }

    if (stop_soon) break;

    queue_cur = queue_cur->next;
    current_entry++;

  }

  if (queue_cur) show_stats();

  write_bitmap();
  write_stats_file(0, 0);
  save_auto();

stop_fuzzing:

  SAYF(CURSOR_SHOW cLRD "\n\n+++ Testing aborted by user +++\n" cRST);

  /* Running for more than 30 minutes but still doing first cycle? */

  if (queue_cycle == 1 && get_cur_time() - start_time > 30 * 60 * 1000) {

    SAYF("\n" cYEL "[!] " cRST
           "Stopped during the first cycle, results may be incomplete.\n"
           "    (For info on resuming, see %s/README.)\n", doc_path);

  }

  fclose(plot_file);
  destroy_queue();
  destroy_extras();
  ck_free(target_path);

  alloc_report();

  OKF("We're done here. Have a nice day!\n");

  exit(0);

}