mbedtls/tests/scripts/generate_pkcs7_tests.py

#!/usr/bin/env python3
#
#  Copyright The Mbed TLS Contributors
#  SPDX-License-Identifier: Apache-2.0
#
#  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
#
#  Unless required by applicable law or agreed to in writing, software
#  distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
#  WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
#  See the License for the specific language governing permissions and
#  limitations under the License.
#

"""
Make fuzz like testing for pkcs7 tests
Given a valid DER pkcs7 file add tests to the test_suite_pkcs7.data file
 - It is expected that the pkcs7_asn1_fail( data_t *pkcs7_buf )
    function is defined in test_suite_pkcs7.function
 - This is not meant to be portable code, if anything it is meant to serve as
   documentation for showing how those ugly tests in test_suite_pkcs7.data were created
"""


import sys
from os.path import exists

PKCS7_TEST_FILE = "../suites/test_suite_pkcs7.data"

class Test: # pylint: disable=too-few-public-methods
    """
    A instance of a test in test_suite_pkcs7.data
    """
    def __init__(self, name, depends, func_call):
        self.name = name
        self.depends = depends
        self.func_call = func_call

    # pylint: disable=no-self-use
    def to_string(self):
        return "\n" + self.name + "\n" + self.depends + "\n" + self.func_call + "\n"

class TestData:
    """
    Take in test_suite_pkcs7.data file.
    Allow for new tests to be added.
    """
    mandatory_dep = "MBEDTLS_SHA256_C"
    test_name = "PKCS7 Parse Failure Invalid ASN1"
    test_function = "pkcs7_asn1_fail:"
    def __init__(self, file_name):
        self.file_name = file_name
        self.last_test_num, self.old_tests = self.read_test_file(file_name)
        self.new_tests = []

    # pylint: disable=no-self-use
    def read_test_file(self, file):
        """
        Parse the test_suite_pkcs7.data file.
        """
        tests = []
        if not exists(file):
            print(file + " Does not exist")
            sys.exit()
        with open(file, "r", encoding='UTF-8') as fp:
            data = fp.read()
        lines = [line.strip() for line in data.split('\n') if len(line.strip()) > 1]
        i = 0
        while i < len(lines):
            if "depends" in lines[i+1]:
                tests.append(Test(lines[i], lines[i+1], lines[i+2]))
                i += 3
            else:
                tests.append(Test(lines[i], None, lines[i+1]))
                i += 2
        latest_test_num = float(tests[-1].name.split('#')[1])
        return latest_test_num, tests

    def add(self, name, func_call):
        self.last_test_num += 1
        self.new_tests.append(Test(self.test_name + ": " + name +  " #" + \
                str(self.last_test_num), "depends_on:" + self.mandatory_dep, \
                self.test_function + '"' + func_call + '"'))

    def write_changes(self):
        with open(self.file_name, 'a', encoding='UTF-8') as fw:
            fw.write("\n")
            for t in self.new_tests:
                fw.write(t.to_string())


def asn1_mutate(data):
    """
    We have been given an asn1 structure representing a pkcs7.
    We want to return an array of slightly modified versions of this data
    they should be modified in a way which makes the structure invalid

    We know that asn1 structures are:
    |---1 byte showing data type---|----byte(s) for length of data---|---data content--|
    We know that some data types can contain other data types.
    Return a dictionary of reasons and mutated data types.
    """

    # off the bat just add bytes to start and end of the buffer
    mutations = []
    reasons = []
    mutations.append(["00"] + data)
    reasons.append("Add null byte to start")
    mutations.append(data + ["00"])
    reasons.append("Add null byte to end")
    # for every asn1 entry we should attempt to:
    #    - change the data type tag
    #    - make the length longer than actual
    #    - make the length shorter than actual
    i = 0
    while i < len(data):
        tag_i = i
        leng_i = tag_i + 1
        data_i = leng_i + 1 + (int(data[leng_i][1], 16) if data[leng_i][0] == '8' else 0)
        if data[leng_i][0] == '8':
            length = int(''.join(data[leng_i + 1: data_i]), 16)
        else:
            length = int(data[leng_i], 16)

        tag = data[tag_i]
        print("Looking at ans1: offset " + str(i) + " tag = " + tag + \
                ", length = " + str(length)+ ":")
        print(''.join(data[data_i:data_i+length]))
        # change tag to something else
        if tag == "02":
            # turn integers into octet strings
            new_tag = "04"
        else:
            # turn everything else into an integer
            new_tag = "02"
        mutations.append(data[:tag_i] + [new_tag] + data[leng_i:])
        reasons.append("Change tag " + tag + " to " + new_tag)

        # change lengths to too big
        # skip any edge cases which would cause carry over
        if int(data[data_i - 1], 16) < 255:
            new_length = str(hex(int(data[data_i - 1], 16) + 1))[2:]
            if len(new_length) == 1:
                new_length = "0"+new_length
            mutations.append(data[:data_i -1] + [new_length] + data[data_i:])
            reasons.append("Change length from " + str(length) + " to " \
                    + str(length + 1))
            # we can add another test here for tags that contain other tags \
            # where they have more data than there containing tags account for
            if tag in ["30", "a0", "31"]:
                mutations.append(data[:data_i -1] + [new_length] + \
                        data[data_i:data_i + length] + ["00"] + \
                        data[data_i + length:])
                reasons.append("Change contents of tag " + tag + " to contain \
                        one unaccounted extra byte")
        # change lengths to too small
        if int(data[data_i - 1], 16) > 0:
            new_length = str(hex(int(data[data_i - 1], 16) - 1))[2:]
            if len(new_length) == 1:
                new_length = "0"+new_length
            mutations.append(data[:data_i -1] + [new_length] + data[data_i:])
            reasons.append("Change length from " + str(length) + " to " + str(length - 1))

        # some tag types contain other tag types so we should iterate into the data
        if tag in ["30", "a0", "31"]:
            i = data_i
        else:
            i = data_i + length

    return list(zip(reasons, mutations))

if __name__ == "__main__":
    if len(sys.argv) < 2:
        print("USAGE: " + sys.argv[0] + " <pkcs7_der_file>")
        sys.exit()

    DATA_FILE = sys.argv[1]
    TEST_DATA = TestData(PKCS7_TEST_FILE)
    with open(DATA_FILE, 'rb') as f:
        DATA_STR = f.read().hex()
    # make data an array of byte strings eg ['de','ad','be','ef']
    HEX_DATA = list(map(''.join, [[DATA_STR[i], DATA_STR[i+1]] for i in range(0, len(DATA_STR), \
            2)]))
    # returns tuples of test_names and modified data buffers
    MUT_ARR = asn1_mutate(HEX_DATA)

    print("made " + str(len(MUT_ARR)) + " new tests")
    for new_test in MUT_ARR:
        TEST_DATA.add(new_test[0], ''.join(new_test[1]))

    TEST_DATA.write_changes()