Olivier Deprez | 157378f | 2022-04-04 15:47:50 +0200 | [diff] [blame^] | 1 | # SPDX-License-Identifier: GPL-2.0 |
| 2 | # |
| 3 | # Parses test results from a kernel dmesg log. |
| 4 | # |
| 5 | # Copyright (C) 2019, Google LLC. |
| 6 | # Author: Felix Guo <felixguoxiuping@gmail.com> |
| 7 | # Author: Brendan Higgins <brendanhiggins@google.com> |
| 8 | |
| 9 | import re |
| 10 | |
| 11 | from collections import namedtuple |
| 12 | from datetime import datetime |
| 13 | from enum import Enum, auto |
| 14 | from functools import reduce |
| 15 | from typing import List, Optional, Tuple |
| 16 | |
| 17 | TestResult = namedtuple('TestResult', ['status','suites','log']) |
| 18 | |
| 19 | class TestSuite(object): |
| 20 | def __init__(self): |
| 21 | self.status = None |
| 22 | self.name = None |
| 23 | self.cases = [] |
| 24 | |
| 25 | def __str__(self): |
| 26 | return 'TestSuite(' + self.status + ',' + self.name + ',' + str(self.cases) + ')' |
| 27 | |
| 28 | def __repr__(self): |
| 29 | return str(self) |
| 30 | |
| 31 | class TestCase(object): |
| 32 | def __init__(self): |
| 33 | self.status = None |
| 34 | self.name = '' |
| 35 | self.log = [] |
| 36 | |
| 37 | def __str__(self): |
| 38 | return 'TestCase(' + self.status + ',' + self.name + ',' + str(self.log) + ')' |
| 39 | |
| 40 | def __repr__(self): |
| 41 | return str(self) |
| 42 | |
| 43 | class TestStatus(Enum): |
| 44 | SUCCESS = auto() |
| 45 | FAILURE = auto() |
| 46 | TEST_CRASHED = auto() |
| 47 | NO_TESTS = auto() |
| 48 | FAILURE_TO_PARSE_TESTS = auto() |
| 49 | |
| 50 | kunit_start_re = re.compile(r'TAP version [0-9]+$') |
| 51 | kunit_end_re = re.compile('(List of all partitions:|' |
| 52 | 'Kernel panic - not syncing: VFS:)') |
| 53 | |
| 54 | def isolate_kunit_output(kernel_output): |
| 55 | started = False |
| 56 | for line in kernel_output: |
| 57 | line = line.rstrip() # line always has a trailing \n |
| 58 | if kunit_start_re.search(line): |
| 59 | prefix_len = len(line.split('TAP version')[0]) |
| 60 | started = True |
| 61 | yield line[prefix_len:] if prefix_len > 0 else line |
| 62 | elif kunit_end_re.search(line): |
| 63 | break |
| 64 | elif started: |
| 65 | yield line[prefix_len:] if prefix_len > 0 else line |
| 66 | |
| 67 | def raw_output(kernel_output): |
| 68 | for line in kernel_output: |
| 69 | print(line.rstrip()) |
| 70 | |
| 71 | DIVIDER = '=' * 60 |
| 72 | |
| 73 | RESET = '\033[0;0m' |
| 74 | |
| 75 | def red(text): |
| 76 | return '\033[1;31m' + text + RESET |
| 77 | |
| 78 | def yellow(text): |
| 79 | return '\033[1;33m' + text + RESET |
| 80 | |
| 81 | def green(text): |
| 82 | return '\033[1;32m' + text + RESET |
| 83 | |
| 84 | def print_with_timestamp(message): |
| 85 | print('[%s] %s' % (datetime.now().strftime('%H:%M:%S'), message)) |
| 86 | |
| 87 | def format_suite_divider(message): |
| 88 | return '======== ' + message + ' ========' |
| 89 | |
| 90 | def print_suite_divider(message): |
| 91 | print_with_timestamp(DIVIDER) |
| 92 | print_with_timestamp(format_suite_divider(message)) |
| 93 | |
| 94 | def print_log(log): |
| 95 | for m in log: |
| 96 | print_with_timestamp(m) |
| 97 | |
| 98 | TAP_ENTRIES = re.compile(r'^(TAP|[\s]*ok|[\s]*not ok|[\s]*[0-9]+\.\.[0-9]+|[\s]*#).*$') |
| 99 | |
| 100 | def consume_non_diagnositic(lines: List[str]) -> None: |
| 101 | while lines and not TAP_ENTRIES.match(lines[0]): |
| 102 | lines.pop(0) |
| 103 | |
| 104 | def save_non_diagnositic(lines: List[str], test_case: TestCase) -> None: |
| 105 | while lines and not TAP_ENTRIES.match(lines[0]): |
| 106 | test_case.log.append(lines[0]) |
| 107 | lines.pop(0) |
| 108 | |
| 109 | OkNotOkResult = namedtuple('OkNotOkResult', ['is_ok','description', 'text']) |
| 110 | |
| 111 | OK_NOT_OK_SUBTEST = re.compile(r'^[\s]+(ok|not ok) [0-9]+ - (.*)$') |
| 112 | |
| 113 | OK_NOT_OK_MODULE = re.compile(r'^(ok|not ok) ([0-9]+) - (.*)$') |
| 114 | |
| 115 | def parse_ok_not_ok_test_case(lines: List[str], test_case: TestCase) -> bool: |
| 116 | save_non_diagnositic(lines, test_case) |
| 117 | if not lines: |
| 118 | test_case.status = TestStatus.TEST_CRASHED |
| 119 | return True |
| 120 | line = lines[0] |
| 121 | match = OK_NOT_OK_SUBTEST.match(line) |
| 122 | while not match and lines: |
| 123 | line = lines.pop(0) |
| 124 | match = OK_NOT_OK_SUBTEST.match(line) |
| 125 | if match: |
| 126 | test_case.log.append(lines.pop(0)) |
| 127 | test_case.name = match.group(2) |
| 128 | if test_case.status == TestStatus.TEST_CRASHED: |
| 129 | return True |
| 130 | if match.group(1) == 'ok': |
| 131 | test_case.status = TestStatus.SUCCESS |
| 132 | else: |
| 133 | test_case.status = TestStatus.FAILURE |
| 134 | return True |
| 135 | else: |
| 136 | return False |
| 137 | |
| 138 | SUBTEST_DIAGNOSTIC = re.compile(r'^[\s]+# .*?: (.*)$') |
| 139 | DIAGNOSTIC_CRASH_MESSAGE = 'kunit test case crashed!' |
| 140 | |
| 141 | def parse_diagnostic(lines: List[str], test_case: TestCase) -> bool: |
| 142 | save_non_diagnositic(lines, test_case) |
| 143 | if not lines: |
| 144 | return False |
| 145 | line = lines[0] |
| 146 | match = SUBTEST_DIAGNOSTIC.match(line) |
| 147 | if match: |
| 148 | test_case.log.append(lines.pop(0)) |
| 149 | if match.group(1) == DIAGNOSTIC_CRASH_MESSAGE: |
| 150 | test_case.status = TestStatus.TEST_CRASHED |
| 151 | return True |
| 152 | else: |
| 153 | return False |
| 154 | |
| 155 | def parse_test_case(lines: List[str]) -> Optional[TestCase]: |
| 156 | test_case = TestCase() |
| 157 | save_non_diagnositic(lines, test_case) |
| 158 | while parse_diagnostic(lines, test_case): |
| 159 | pass |
| 160 | if parse_ok_not_ok_test_case(lines, test_case): |
| 161 | return test_case |
| 162 | else: |
| 163 | return None |
| 164 | |
| 165 | SUBTEST_HEADER = re.compile(r'^[\s]+# Subtest: (.*)$') |
| 166 | |
| 167 | def parse_subtest_header(lines: List[str]) -> Optional[str]: |
| 168 | consume_non_diagnositic(lines) |
| 169 | if not lines: |
| 170 | return None |
| 171 | match = SUBTEST_HEADER.match(lines[0]) |
| 172 | if match: |
| 173 | lines.pop(0) |
| 174 | return match.group(1) |
| 175 | else: |
| 176 | return None |
| 177 | |
| 178 | SUBTEST_PLAN = re.compile(r'[\s]+[0-9]+\.\.([0-9]+)') |
| 179 | |
| 180 | def parse_subtest_plan(lines: List[str]) -> Optional[int]: |
| 181 | consume_non_diagnositic(lines) |
| 182 | match = SUBTEST_PLAN.match(lines[0]) |
| 183 | if match: |
| 184 | lines.pop(0) |
| 185 | return int(match.group(1)) |
| 186 | else: |
| 187 | return None |
| 188 | |
| 189 | def max_status(left: TestStatus, right: TestStatus) -> TestStatus: |
| 190 | if left == TestStatus.TEST_CRASHED or right == TestStatus.TEST_CRASHED: |
| 191 | return TestStatus.TEST_CRASHED |
| 192 | elif left == TestStatus.FAILURE or right == TestStatus.FAILURE: |
| 193 | return TestStatus.FAILURE |
| 194 | elif left != TestStatus.SUCCESS: |
| 195 | return left |
| 196 | elif right != TestStatus.SUCCESS: |
| 197 | return right |
| 198 | else: |
| 199 | return TestStatus.SUCCESS |
| 200 | |
| 201 | def parse_ok_not_ok_test_suite(lines: List[str], |
| 202 | test_suite: TestSuite, |
| 203 | expected_suite_index: int) -> bool: |
| 204 | consume_non_diagnositic(lines) |
| 205 | if not lines: |
| 206 | test_suite.status = TestStatus.TEST_CRASHED |
| 207 | return False |
| 208 | line = lines[0] |
| 209 | match = OK_NOT_OK_MODULE.match(line) |
| 210 | if match: |
| 211 | lines.pop(0) |
| 212 | if match.group(1) == 'ok': |
| 213 | test_suite.status = TestStatus.SUCCESS |
| 214 | else: |
| 215 | test_suite.status = TestStatus.FAILURE |
| 216 | suite_index = int(match.group(2)) |
| 217 | if suite_index != expected_suite_index: |
| 218 | print_with_timestamp( |
| 219 | red('[ERROR] ') + 'expected_suite_index ' + |
| 220 | str(expected_suite_index) + ', but got ' + |
| 221 | str(suite_index)) |
| 222 | return True |
| 223 | else: |
| 224 | return False |
| 225 | |
| 226 | def bubble_up_errors(to_status, status_container_list) -> TestStatus: |
| 227 | status_list = map(to_status, status_container_list) |
| 228 | return reduce(max_status, status_list, TestStatus.SUCCESS) |
| 229 | |
| 230 | def bubble_up_test_case_errors(test_suite: TestSuite) -> TestStatus: |
| 231 | max_test_case_status = bubble_up_errors(lambda x: x.status, test_suite.cases) |
| 232 | return max_status(max_test_case_status, test_suite.status) |
| 233 | |
| 234 | def parse_test_suite(lines: List[str], expected_suite_index: int) -> Optional[TestSuite]: |
| 235 | if not lines: |
| 236 | return None |
| 237 | consume_non_diagnositic(lines) |
| 238 | test_suite = TestSuite() |
| 239 | test_suite.status = TestStatus.SUCCESS |
| 240 | name = parse_subtest_header(lines) |
| 241 | if not name: |
| 242 | return None |
| 243 | test_suite.name = name |
| 244 | expected_test_case_num = parse_subtest_plan(lines) |
| 245 | if expected_test_case_num is None: |
| 246 | return None |
| 247 | while expected_test_case_num > 0: |
| 248 | test_case = parse_test_case(lines) |
| 249 | if not test_case: |
| 250 | break |
| 251 | test_suite.cases.append(test_case) |
| 252 | expected_test_case_num -= 1 |
| 253 | if parse_ok_not_ok_test_suite(lines, test_suite, expected_suite_index): |
| 254 | test_suite.status = bubble_up_test_case_errors(test_suite) |
| 255 | return test_suite |
| 256 | elif not lines: |
| 257 | print_with_timestamp(red('[ERROR] ') + 'ran out of lines before end token') |
| 258 | return test_suite |
| 259 | else: |
| 260 | print('failed to parse end of suite' + lines[0]) |
| 261 | return None |
| 262 | |
| 263 | TAP_HEADER = re.compile(r'^TAP version 14$') |
| 264 | |
| 265 | def parse_tap_header(lines: List[str]) -> bool: |
| 266 | consume_non_diagnositic(lines) |
| 267 | if TAP_HEADER.match(lines[0]): |
| 268 | lines.pop(0) |
| 269 | return True |
| 270 | else: |
| 271 | return False |
| 272 | |
| 273 | TEST_PLAN = re.compile(r'[0-9]+\.\.([0-9]+)') |
| 274 | |
| 275 | def parse_test_plan(lines: List[str]) -> Optional[int]: |
| 276 | consume_non_diagnositic(lines) |
| 277 | match = TEST_PLAN.match(lines[0]) |
| 278 | if match: |
| 279 | lines.pop(0) |
| 280 | return int(match.group(1)) |
| 281 | else: |
| 282 | return None |
| 283 | |
| 284 | def bubble_up_suite_errors(test_suite_list: List[TestSuite]) -> TestStatus: |
| 285 | return bubble_up_errors(lambda x: x.status, test_suite_list) |
| 286 | |
| 287 | def parse_test_result(lines: List[str]) -> TestResult: |
| 288 | consume_non_diagnositic(lines) |
| 289 | if not lines or not parse_tap_header(lines): |
| 290 | return TestResult(TestStatus.NO_TESTS, [], lines) |
| 291 | expected_test_suite_num = parse_test_plan(lines) |
| 292 | if not expected_test_suite_num: |
| 293 | return TestResult(TestStatus.FAILURE_TO_PARSE_TESTS, [], lines) |
| 294 | test_suites = [] |
| 295 | for i in range(1, expected_test_suite_num + 1): |
| 296 | test_suite = parse_test_suite(lines, i) |
| 297 | if test_suite: |
| 298 | test_suites.append(test_suite) |
| 299 | else: |
| 300 | print_with_timestamp( |
| 301 | red('[ERROR] ') + ' expected ' + |
| 302 | str(expected_test_suite_num) + |
| 303 | ' test suites, but got ' + str(i - 2)) |
| 304 | break |
| 305 | test_suite = parse_test_suite(lines, -1) |
| 306 | if test_suite: |
| 307 | print_with_timestamp(red('[ERROR] ') + |
| 308 | 'got unexpected test suite: ' + test_suite.name) |
| 309 | if test_suites: |
| 310 | return TestResult(bubble_up_suite_errors(test_suites), test_suites, lines) |
| 311 | else: |
| 312 | return TestResult(TestStatus.NO_TESTS, [], lines) |
| 313 | |
| 314 | def print_and_count_results(test_result: TestResult) -> Tuple[int, int, int]: |
| 315 | total_tests = 0 |
| 316 | failed_tests = 0 |
| 317 | crashed_tests = 0 |
| 318 | for test_suite in test_result.suites: |
| 319 | if test_suite.status == TestStatus.SUCCESS: |
| 320 | print_suite_divider(green('[PASSED] ') + test_suite.name) |
| 321 | elif test_suite.status == TestStatus.TEST_CRASHED: |
| 322 | print_suite_divider(red('[CRASHED] ' + test_suite.name)) |
| 323 | else: |
| 324 | print_suite_divider(red('[FAILED] ') + test_suite.name) |
| 325 | for test_case in test_suite.cases: |
| 326 | total_tests += 1 |
| 327 | if test_case.status == TestStatus.SUCCESS: |
| 328 | print_with_timestamp(green('[PASSED] ') + test_case.name) |
| 329 | elif test_case.status == TestStatus.TEST_CRASHED: |
| 330 | crashed_tests += 1 |
| 331 | print_with_timestamp(red('[CRASHED] ' + test_case.name)) |
| 332 | print_log(map(yellow, test_case.log)) |
| 333 | print_with_timestamp('') |
| 334 | else: |
| 335 | failed_tests += 1 |
| 336 | print_with_timestamp(red('[FAILED] ') + test_case.name) |
| 337 | print_log(map(yellow, test_case.log)) |
| 338 | print_with_timestamp('') |
| 339 | return total_tests, failed_tests, crashed_tests |
| 340 | |
| 341 | def parse_run_tests(kernel_output) -> TestResult: |
| 342 | total_tests = 0 |
| 343 | failed_tests = 0 |
| 344 | crashed_tests = 0 |
| 345 | test_result = parse_test_result(list(isolate_kunit_output(kernel_output))) |
| 346 | if test_result.status == TestStatus.NO_TESTS: |
| 347 | print(red('[ERROR] ') + yellow('no tests run!')) |
| 348 | elif test_result.status == TestStatus.FAILURE_TO_PARSE_TESTS: |
| 349 | print(red('[ERROR] ') + yellow('could not parse test results!')) |
| 350 | else: |
| 351 | (total_tests, |
| 352 | failed_tests, |
| 353 | crashed_tests) = print_and_count_results(test_result) |
| 354 | print_with_timestamp(DIVIDER) |
| 355 | fmt = green if test_result.status == TestStatus.SUCCESS else red |
| 356 | print_with_timestamp( |
| 357 | fmt('Testing complete. %d tests run. %d failed. %d crashed.' % |
| 358 | (total_tests, failed_tests, crashed_tests))) |
| 359 | return test_result |