1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
|
#!/usr/bin/env python3
# Copyright 2023 The Chromium Authors
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.
"""Download the fuzzing corpora associated with Chromium/V8 fuzz test.
For libfuzzer and centipede, assumes that fuzzer targets are already built
and reside in the BUILD_DIR directory. For fuzzilli download corpora from
gs bucket based on arch.
* Example usage: download_fuzz_corpora.py --download-dir [DOWNLOAD_DIR]
--build-dir [BUILD_DIR] --corpora-type [CORPORA_TYPE]
"""
LIBFUZZER_CORPORA_TYPE = 'libfuzzer'
CENTIPEDE_CORPORA_TYPE = 'centipede'
FUZZILLI_CORPORA_TYPE = 'fuzzilli'
ALL_CORPORA_TYPES = [
LIBFUZZER_CORPORA_TYPE, CENTIPEDE_CORPORA_TYPE, FUZZILLI_CORPORA_TYPE
]
CORPORA_BUCKET_BASE_URL_BY_TYPE = {
LIBFUZZER_CORPORA_TYPE:
'gs://clusterfuzz-libfuzzer-backup/corpus/libfuzzer/',
CENTIPEDE_CORPORA_TYPE:
'gs://clusterfuzz-centipede-backup/corpus/centipede/',
FUZZILLI_CORPORA_TYPE: 'gs://autozilli/',
}
import argparse
import coverage_consts
import logging
from multiprocessing import cpu_count, Pool
import os
import re
import subprocess
import sys
def _gsutil(cmd, cwd):
subprocess.run(cmd, cwd=cwd)
def _get_fuzzilli_corpora(arch):
output = subprocess.check_output(
['gsutil', 'ls',
CORPORA_BUCKET_BASE_URL_BY_TYPE[FUZZILLI_CORPORA_TYPE]]).decode('utf-8')
regex = {
'x64': 'autozilli-[0-9]+\.tgz',
'x86': 'autozilli-x86-[0-9]+\.tgz',
'arm64': 'autozilli-arm64-[0-9]+\.tgz',
}[arch]
return re.findall(regex, output)
def _download_corpus(args):
target = args[0]
download_dir = args[1]
corpora_type = args[2]
url = CORPORA_BUCKET_BASE_URL_BY_TYPE[corpora_type]
if corpora_type == FUZZILLI_CORPORA_TYPE:
# For a corpora file autozilli-1.tgz, it will be downloaded to
# [DOWNLOAD_DIR]/autozilli-1/autozilli-1.tgz
corpus_dir, _ = os.path.splitext(target)
corpus_url = os.path.join(url, target)
else:
corpus_dir = target
corpus_url = os.path.join(url, target, 'latest.zip')
subprocess.run(['mkdir', corpus_dir], cwd=download_dir)
cmd = ['gsutil', 'cp', corpus_url, corpus_dir]
_gsutil(cmd, download_dir)
def _unzip_corpus(args):
target = args[0]
download_dir = args[1]
target_folder = os.path.join(download_dir, target)
subprocess.run(['unzip', 'latest.zip'], cwd=target_folder)
subprocess.run(['rm', 'latest.zip'], cwd=target_folder)
try:
# Unzipping the corpora often also contains a "regressions" folder, which
# is a subset of the total corpus, so can be ignored/removed
subprocess.run(['rm', '-rf', 'regressions'], cwd=target_folder)
except:
pass
def _unzip_fuzzilli_corpus(args):
corpus = args[0]
download_dir = args[1]
base, _ = os.path.splitext(corpus)
corpus_dir = os.path.join(download_dir, base)
subprocess.run(['tar', 'xzvf', corpus], cwd=corpus_dir)
subprocess.run(['rm', corpus], cwd=corpus_dir)
def _ParseCommandArguments():
"""Adds and parses relevant arguments for tool comands.
Returns:
A dictionary representing the arguments.
"""
arg_parser = argparse.ArgumentParser()
arg_parser.usage = __doc__
arg_parser.add_argument('--download-dir',
type=str,
required=True,
help='Directory into which corpora are downloaded.')
arg_parser.add_argument('--build-dir',
required=True,
type=str,
help='Directory where fuzzers were built.')
arg_parser.add_argument('--corpora-type',
choices=ALL_CORPORA_TYPES,
default=LIBFUZZER_CORPORA_TYPE,
help='The type of corpora to download.')
arg_parser.add_argument(
'--arch',
choices=['x64', 'x86', 'arm64'],
default='x64',
help='The cpu architecture of the target. Fuzzilli only.')
args = arg_parser.parse_args()
return args
def Main():
args = _ParseCommandArguments()
exit
if not args.download_dir:
logging.error("No download_dir given")
exit
if not os.path.isdir(args.download_dir):
logging.error("%s does not exist or is not a directory" % args.download_dir)
exit
if not args.build_dir:
logging.error("No build_dir given")
exit
if not os.path.isdir(args.build_dir):
logging.error("%s does not exist or is not a directory" % args.build_dir)
exit
if args.corpora_type == FUZZILLI_CORPORA_TYPE:
corpora_to_download = _get_fuzzilli_corpora(args.arch)
else:
corpora_to_download = []
for target in os.listdir(args.build_dir):
if target.endswith('_fuzzer'):
corpora_to_download.append(target)
print("Corpora to download: " + str(corpora_to_download))
with Pool(cpu_count()) as p:
results = p.map(_download_corpus,
[(corpus, args.download_dir, args.corpora_type)
for corpus in corpora_to_download])
unzip_func = (_unzip_fuzzilli_corpus if args.corpora_type
== FUZZILLI_CORPORA_TYPE else _unzip_corpus)
with Pool(cpu_count()) as p:
results = p.map(unzip_func, [(corpus, args.download_dir)
for corpus in corpora_to_download])
if __name__ == '__main__':
sys.exit(Main())
|
