autotoken: splicing; splice_optout

7 files changed, 155 insertions, 11 deletions

diff --git a/custom_mutators/autotokens/Makefile b/custom_mutators/autotokens/Makefile
index 8af63635..ab1da4b6 100644
--- a/custom_mutators/autotokens/Makefile
+++ b/custom_mutators/autotokens/Makefile
@@ -1,5 +1,9 @@
 ifdef debug
-	CFLAGS += "-fsanitize=address -Wall"
+	CFLAGS += -fsanitize=address -Wall
+	CXX := clang++
+endif
+ifdef DEBUG
+	CFLAGS += -fsanitize=address -Wall
 	CXX := clang++
 endif
 
diff --git a/custom_mutators/autotokens/autotokens.cpp b/custom_mutators/autotokens/autotokens.cpp
index 7aecb010..c9ec4352 100644
--- a/custom_mutators/autotokens/autotokens.cpp
+++ b/custom_mutators/autotokens/autotokens.cpp
@@ -19,6 +19,13 @@ extern "C" {
 #define AUTOTOKENS_ALTERNATIVE_TOKENIZE 0
 #define AUTOTOKENS_CHANGE_MIN 8
 #define AUTOTOKENS_WHITESPACE " "
+#define AUTOTOKENS_SIZE_MIN 8
+#define AUTOTOKENS_SPLICE_MIN 4
+#define AUTOTOKENS_SPLICE_MAX 64
+
+#if AUTOTOKENS_SPLICE_MIN >= AUTOTOKENS_SIZE_MIN
+  #error SPLICE_MIN must be lower than SIZE_MIN
+#endif
 
 using namespace std;
 
@@ -42,6 +49,7 @@ static u32        extras_cnt, a_extras_cnt;
 static u64        all_spaces, all_tabs, all_lf, all_ws;
 static u64        all_structure_items;
 static unordered_map<string, vector<u32> *> file_mapping;
+static unordered_map<u32, vector<u32> *>    id_mapping;
 static unordered_map<string, u32>           token_to_id;
 static unordered_map<u32, string>           id_to_token;
 static string                               whitespace = AUTOTOKENS_WHITESPACE;
@@ -76,6 +84,8 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
                                   u8 **out_buf, u8 *add_buf,
                                   size_t add_buf_size, size_t max_size) {
 
+  (void)(data);
+
   if (s == NULL) {
 
     *out_buf = NULL;
@@ -92,14 +102,14 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
                                afl_ptr->havoc_div / 256));
   // DEBUG(stderr, "structure size: %lu, rounds: %u \n", m.size(), rounds);
 
-  u32 max_rand = 7;
+  u32 max_rand = 14;
 
   for (i = 0; i < rounds; ++i) {
 
     switch (rand_below(afl_ptr, max_rand)) {
 
       /* CHANGE */
-      case 0 ... 3:                                         /* fall through */
+      case 0 ... 7:                                         /* fall through */
       {
 
         u32 pos = rand_below(afl_ptr, m_size);
@@ -122,18 +132,19 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
       }
 
       /* INSERT (m_size +1 so we insert also after last place) */
-      case 4 ... 5: {
+      case 8 ... 9: {
 
         u32 new_item;
         do {
 
           new_item = rand_below(afl_ptr, current_id);
 
-        } while (!alternative_tokenize && new_item >= whitespace_ids);
+        } while (unlikely(!alternative_tokenize && new_item >= whitespace_ids));
 
         u32 pos = rand_below(afl_ptr, m_size + 1);
         m.insert(m.begin() + pos, new_item);
         ++m_size;
+        DEBUG(stderr, "INS: %u at %u\n", new_item, pos);
 
         if (likely(!alternative_tokenize)) {
 
@@ -168,8 +179,63 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
 
       }
 
+      /* SPLICING */
+      case 10 ... 11: {
+
+        u32  strategy = rand_below(afl_ptr, 4), dst_off, n;
+        auto src = id_mapping[rand_below(afl_ptr, valid_structures)];
+        u32  src_size = src->size();
+        u32  src_off = rand_below(afl_ptr, src_size - AUTOTOKENS_SPLICE_MIN);
+        u32  rand_r = 1 + MAX(AUTOTOKENS_SPLICE_MIN,
+                              MIN(AUTOTOKENS_SPLICE_MAX, src_size - src_off));
+
+        switch (strategy) {
+
+          // insert
+          case 0: {
+
+            dst_off = rand_below(afl_ptr, m_size);
+            n = AUTOTOKENS_SPLICE_MIN +
+                rand_below(afl_ptr, MIN(AUTOTOKENS_SPLICE_MAX,
+                                        rand_r - AUTOTOKENS_SPLICE_MIN));
+            m.insert(m.begin() + dst_off, src->begin() + src_off,
+                     src->begin() + src_off + n);
+            m_size += n;
+            DEBUG(stderr, "SPLICE-INS: %u at %u\n", n, dst_off);
+            break;
+
+          }
+
+          // overwrite
+          default: {
+
+            dst_off = rand_below(afl_ptr, m_size - AUTOTOKENS_SPLICE_MIN);
+            n = AUTOTOKENS_SPLICE_MIN +
+                rand_below(
+                    afl_ptr,
+                    MIN(AUTOTOKENS_SPLICE_MAX - AUTOTOKENS_SPLICE_MIN,
+                        MIN(m_size - dst_off - AUTOTOKENS_SPLICE_MIN,
+                            src_size - src_off - AUTOTOKENS_SPLICE_MIN)));
+
+            for (u32 i = 0; i < n; ++i) {
+
+              m[dst_off + i] = (*src)[src_off + i];
+
+            }
+
+            DEBUG(stderr, "SPLICE-MUT: %u at %u\n", n, dst_off);
+            break;
+
+          }
+
+        }
+
+        break;
+
+      }
+
       /* ERASE - only if large enough */
-      case 6: {
+      case 12 ... 13: {
 
         if (m_size > 8) {
 
@@ -178,7 +244,7 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
 
         } else {
 
-          max_rand = 6;
+          max_rand = 12;
 
         }
 
@@ -236,12 +302,15 @@ extern "C" size_t afl_custom_fuzz(my_mutator_t *data, u8 *buf, size_t buf_size,
 extern "C" unsigned char afl_custom_queue_get(void                *data,
                                               const unsigned char *filename) {
 
+  (void)(data);
+
   if (likely(!debug)) {
 
     if ((afl_ptr->shm.cmplog_mode && !afl_ptr->queue_cur->is_ascii) ||
         (only_fav && !afl_ptr->queue_cur->favored)) {
 
       s = NULL;
+      DEBUG(stderr, "cmplog not ascii or only_fav and not favorite\n");
       return 0;
 
     }
@@ -334,8 +403,8 @@ extern "C" unsigned char afl_custom_queue_get(void                *data,
 
       fclose(fp);
       file_mapping[fn] = structure;  // NULL ptr so we don't read the file again
-      DEBUG(stderr, "Too short (%lu) %s\n", len, filename);
       s = NULL;
+      DEBUG(stderr, "Too short (%lu) %s\n", len, filename);
       return 0;
 
     }
@@ -362,8 +431,8 @@ extern "C" unsigned char afl_custom_queue_get(void                *data,
       if (((len * AFL_TXT_MIN_PERCENT) / 100) > valid_chars) {
 
         file_mapping[fn] = NULL;
-        DEBUG(stderr, "Not text (%lu) %s\n", len, filename);
         s = NULL;
+        DEBUG(stderr, "Not text (%lu) %s\n", len, filename);
         return 0;
 
       }
@@ -766,6 +835,15 @@ extern "C" unsigned char afl_custom_queue_get(void                *data,
 
     }
 
+    if (tokens.size() < AUTOTOKENS_SIZE_MIN) {
+
+      file_mapping[fn] = NULL;
+      s = NULL;
+      DEBUG(stderr, "too few tokens\n");
+      return 0;
+
+    }
+
     /* Now we transform the tokens into an ID list and saved that */
 
     structure = new vector<u32>();
@@ -791,8 +869,9 @@ extern "C" unsigned char afl_custom_queue_get(void                *data,
 
     // save the token structure to the file mapping
     file_mapping[fn] = structure;
-    s = structure;
+    id_mapping[valid_structures] = structure;
     ++valid_structures;
+    s = structure;
     all_structure_items += structure->size();
 
     // we are done!
@@ -897,6 +976,12 @@ extern "C" my_mutator_t *afl_custom_init(afl_state *afl, unsigned int seed) {
 
 }
 
+extern "C" void afl_custom_splice_optout(my_mutator_t *data) {
+
+  (void)(data);
+
+}
+
 extern "C" void afl_custom_deinit(my_mutator_t *data) {
 
   /* we use this to print statistics at exit :-)
diff --git a/docs/custom_mutators.md b/docs/custom_mutators.md
index 4ffeda7a..322caa5b 100644
--- a/docs/custom_mutators.md
+++ b/docs/custom_mutators.md
@@ -48,6 +48,7 @@ C/C++:
 ```c
 void *afl_custom_init(afl_state_t *afl, unsigned int seed);
 unsigned int afl_custom_fuzz_count(void *data, const unsigned char *buf, size_t buf_size);
+void afl_custom_splice_optout(void *data);
 size_t afl_custom_fuzz(void *data, unsigned char *buf, size_t buf_size, unsigned char **out_buf, unsigned char *add_buf, size_t add_buf_size, size_t max_size);
 const char *afl_custom_describe(void *data, size_t max_description_len);
 size_t afl_custom_post_process(void *data, unsigned char *buf, size_t buf_size, unsigned char **out_buf);
@@ -72,6 +73,9 @@ def init(seed):
 def fuzz_count(buf):
     return cnt
 
+def splice_optout()
+    pass
+
 def fuzz(buf, add_buf, max_size):
     return mutated_out
 
@@ -132,6 +136,13 @@ def deinit():  # optional for Python
     for a specific queue entry, use this function. This function is most useful
     if `AFL_CUSTOM_MUTATOR_ONLY` is **not** used.
 
+- `splice_optout` (optional):
+
+    If this function is present, no splicing target is passed to the `fuzz`
+    function. This saves time if splicing data is not needed by the custom
+    fuzzing function.
+    This function is never called, just needs to be present to activate.
+
 - `fuzz` (optional):
 
     This method performs custom mutations on a given input. It also accepts an
diff --git a/include/afl-fuzz.h b/include/afl-fuzz.h
index 69fea579..1e8d085d 100644
--- a/include/afl-fuzz.h
+++ b/include/afl-fuzz.h
@@ -344,6 +344,7 @@ enum {
   /* 12 */ PY_FUNC_INTROSPECTION,
   /* 13 */ PY_FUNC_DESCRIBE,
   /* 14 */ PY_FUNC_FUZZ_SEND,
+  /* 15 */ PY_FUNC_SPLICE_OPTOUT,
   PY_FUNC_COUNT
 
 };
@@ -495,6 +496,7 @@ typedef struct afl_state {
       no_unlink,                        /* do not unlink cur_input          */
       debug,                            /* Debug mode                       */
       custom_only,                      /* Custom mutator only mode         */
+      custom_splice_optout,             /* Custom mutator no splice buffer  */
       is_main_node,                     /* if this is the main node         */
       is_secondary_node,                /* if this is a secondary instance  */
       pizza_is_served;                  /* pizza mode                       */
@@ -829,6 +831,17 @@ struct custom_mutator {
   u32 (*afl_custom_fuzz_count)(void *data, const u8 *buf, size_t buf_size);
 
   /**
+   * Opt-out of a splicing input for the fuzz mutator
+   *
+   * Empty dummy function. It's presence tells afl-fuzz not to pass a
+   * splice data pointer and len.
+   *
+   * @param data pointer returned in afl_custom_init by this custom mutator
+   * @noreturn
+   */
+  void (*afl_custom_splice_optout)(void *data);
+
+  /**
    * Perform custom mutations on a given input
    *
    * (Optional for now. Required in the future)
@@ -1057,6 +1070,7 @@ u8          havoc_mutation_probability_py(void *);
 u8          queue_get_py(void *, const u8 *);
 const char *introspection_py(void *);
 u8          queue_new_entry_py(void *, const u8 *, const u8 *);
+void        splice_optout(void *);
 void        deinit_py(void *);
 
 #endif
diff --git a/src/afl-fuzz-mutators.c b/src/afl-fuzz-mutators.c
index 22e5262e..ce43064a 100644
--- a/src/afl-fuzz-mutators.c
+++ b/src/afl-fuzz-mutators.c
@@ -358,6 +358,19 @@ struct custom_mutator *load_custom_mutator(afl_state_t *afl, const char *fn) {
 
   }
 
+  /* "afl_custom_splice_optout", optional, never called */
+  mutator->afl_custom_splice_optout = dlsym(dh, "afl_custom_splice_optout");
+  if (!mutator->afl_custom_splice_optout) {
+
+    ACTF("optional symbol 'afl_custom_splice_optout' not found.");
+
+  } else {
+
+    OKF("Found 'afl_custom_splice_optout'.");
+    afl->custom_splice_optout = 1;
+
+  }
+
   /* "afl_custom_fuzz_send", optional */
   mutator->afl_custom_fuzz_send = dlsym(dh, "afl_custom_fuzz_send");
   if (!mutator->afl_custom_fuzz_send) {
diff --git a/src/afl-fuzz-one.c b/src/afl-fuzz-one.c
index eaf65987..5e352dcb 100644
--- a/src/afl-fuzz-one.c
+++ b/src/afl-fuzz-one.c
@@ -1954,7 +1954,8 @@ custom_mutator_stage:
           u32                 target_len = 0;
 
           /* check if splicing makes sense yet (enough entries) */
-          if (likely(afl->ready_for_splicing_count > 1)) {
+          if (likely(!afl->custom_splice_optout &&
+                     afl->ready_for_splicing_count > 1)) {
 
             /* Pick a random other queue entry for passing to external API
                that has the necessary length */
diff --git a/src/afl-fuzz-python.c b/src/afl-fuzz-python.c
index b509b936..69c305f7 100644
--- a/src/afl-fuzz-python.c
+++ b/src/afl-fuzz-python.c
@@ -248,6 +248,8 @@ static py_mutator_t *init_py_module(afl_state_t *afl, u8 *module_name) {
         PyObject_GetAttrString(py_module, "queue_get");
     py_functions[PY_FUNC_FUZZ_SEND] =
         PyObject_GetAttrString(py_module, "fuzz_send");
+    py_functions[PY_FUNC_SPLICE_OPTOUT] =
+        PyObject_GetAttrString(py_module, "splice_optout");
     py_functions[PY_FUNC_QUEUE_NEW_ENTRY] =
         PyObject_GetAttrString(py_module, "queue_new_entry");
     py_functions[PY_FUNC_INTROSPECTION] =
@@ -394,6 +396,13 @@ void deinit_py(void *py_mutator) {
 
 }
 
+void splice_optout_py(void *py_mutator) {
+
+  // this is never called
+  (void)(py_mutator);
+
+}
+
 struct custom_mutator *load_custom_mutator_py(afl_state_t *afl,
                                               char        *module_name) {
 
@@ -474,6 +483,13 @@ struct custom_mutator *load_custom_mutator_py(afl_state_t *afl,
 
   }
 
+  if (py_functions[PY_FUNC_SPLICE_OPTOUT]) {
+
+    mutator->afl_custom_splice_optout = splice_optout_py;
+    afl->custom_splice_optout = 1;
+
+  }
+
   if (py_functions[PY_FUNC_QUEUE_NEW_ENTRY]) {
 
     mutator->afl_custom_queue_new_entry = queue_new_entry_py;