v8/tools/testrunner/standard_runner_test.py
Michael Achenbach f321ada760 [test] Let numfuzz tests automatically use all available instances
All --stress-* flags are now automatically tested. This also removes
a superfluous option that was never changed. The default value is
now inlined.

No-Try: true
Bug: v8:13113
Change-Id: If7428b383ed01ff36a93f618badababfc448db26
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3899259
Reviewed-by: Alexander Schulze <alexschulze@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/main@{#83258}
2022-09-16 10:52:10 +00:00

582 lines
19 KiB
Python

#!/usr/bin/env python3
# Copyright 2017 the V8 project authors. All rights reserved.
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.
"""
Global system tests for V8 test runners and fuzzers.
This hooks up the framework under tools/testrunner testing high-level scenarios
with different test suite extensions and build configurations.
"""
# TODO(machenbach): Mock out util.GuessOS to make these tests really platform
# independent.
# TODO(machenbach): Move coverage recording to a global test entry point to
# include other unittest suites in the coverage report.
# TODO(machenbach): Coverage data from multiprocessing doesn't work.
# TODO(majeski): Add some tests for the fuzzers.
from collections import deque
import os
import sys
import unittest
from os.path import dirname as up
from mock import patch
TOOLS_ROOT = up(up(os.path.abspath(__file__)))
sys.path.append(TOOLS_ROOT)
from testrunner import standard_runner
from testrunner import num_fuzzer
from testrunner.testproc import base
from testrunner.testproc import fuzzer
from testrunner.utils.test_utils import (
temp_base,
TestRunnerTest,
with_json_output,
FakeOSContext,
)
class StandardRunnerTest(TestRunnerTest):
def get_runner_class(self):
return standard_runner.StandardTestRunner
def testPass(self):
"""Test running only passing tests in two variants.
Also test printing durations.
"""
result = self.run_tests(
'--progress=verbose',
'--variants=default,stress',
'--time',
'sweet/bananas',
'sweet/raspberries',
)
result.stdout_includes('sweet/bananas default: PASS')
# TODO(majeski): Implement for test processors
# self.assertIn('Total time:', result.stderr, result)
# self.assertIn('sweet/bananas', result.stderr, result)
result.has_returncode(0)
def testPassHeavy(self):
"""Test running with some tests marked heavy."""
result = self.run_tests(
'--progress=verbose',
'--variants=nooptimization',
'-j2',
'sweet',
baseroot='testroot3',
)
result.stdout_includes('7 tests ran')
result.has_returncode(0)
def testShardedProc(self):
for shard in [1, 2]:
result = self.run_tests(
'--progress=verbose',
'--variants=default,stress',
'--shard-count=2',
'--shard-run=%d' % shard,
'sweet/blackberries',
'sweet/raspberries',
infra_staging=False,
)
# One of the shards gets one variant of each test.
result.stdout_includes('2 tests ran')
if shard == 1:
result.stdout_includes('sweet/raspberries default')
result.stdout_includes('sweet/raspberries stress')
result.has_returncode(0)
else:
result.stdout_includes(
'sweet/blackberries default: FAIL')
result.stdout_includes(
'sweet/blackberries stress: FAIL')
result.has_returncode(1)
@unittest.skip("incompatible with test processors")
def testSharded(self):
"""Test running a particular shard."""
for shard in [1, 2]:
result = self.run_tests(
'--progress=verbose',
'--variants=default,stress',
'--shard-count=2',
'--shard-run=%d' % shard,
'sweet/bananas',
'sweet/raspberries',
)
# One of the shards gets one variant of each test.
result.stdout_includes('Running 2 tests')
result.stdout_includes('sweet/bananas')
result.stdout_includes('sweet/raspberries')
result.has_returncode(0)
def testFail(self):
"""Test running only failing tests in two variants."""
result = self.run_tests(
'--progress=verbose',
'--variants=default,stress',
'sweet/strawberries',
infra_staging=False,
)
result.stdout_includes('sweet/strawberries default: FAIL')
result.has_returncode(1)
def testGN(self):
"""Test running only failing tests in two variants."""
result = self.run_tests('--gn',baseroot="testroot5")
result.stdout_includes('>>> Latest GN build found: build')
result.stdout_includes('Build found: ')
result.stdout_includes('v8_test_/out.gn/build')
result.has_returncode(2)
def testMalformedJsonConfig(self):
"""Test running only failing tests in two variants."""
result = self.run_tests(baseroot="testroot4")
result.stdout_includes('contains invalid json')
result.stdout_includes('Failed to load build config')
result.stderr_includes('testrunner.base_runner.TestRunnerError')
result.has_returncode(5)
def testFailWithRerunAndJSON(self):
"""Test re-running a failing test and output to json."""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'--rerun-failures-count=2',
'--random-seed=123',
'--json-test-results', with_json_output,
'sweet/strawberries',
infra_staging=False,
)
result.stdout_includes('sweet/strawberries default: FAIL')
# With test processors we don't count reruns as separated failures.
# TODO(majeski): fix it?
result.stdout_includes('1 tests failed')
result.has_returncode(0)
# TODO(majeski): Previously we only reported the variant flags in the
# flags field of the test result.
# After recent changes we report all flags, including the file names.
# This is redundant to the command. Needs investigation.
result.json_content_equals('expected_test_results1.json')
def testFlakeWithRerunAndJSON(self):
"""Test re-running a failing test and output to json."""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'--rerun-failures-count=2',
'--random-seed=123',
'--json-test-results', with_json_output,
'sweet',
baseroot='testroot2',
infra_staging=False,
)
result.stdout_includes('sweet/bananaflakes default: FAIL PASS')
result.stdout_includes('=== sweet/bananaflakes (flaky) ===')
result.stdout_includes('1 tests failed')
result.stdout_includes('1 tests were flaky')
result.has_returncode(0)
result.json_content_equals('expected_test_results2.json')
def testAutoDetect(self):
"""Fake a build with several auto-detected options.
Using all those options at once doesn't really make much sense. This is
merely for getting coverage.
"""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'sweet/bananas',
config_overrides=dict(
dcheck_always_on=True, is_asan=True, is_cfi=True,
is_msan=True, is_tsan=True, is_ubsan_vptr=True, target_cpu='x86',
v8_enable_i18n_support=False, v8_target_cpu='x86',
v8_enable_verify_csa=False, v8_enable_lite_mode=False,
v8_enable_pointer_compression=False,
v8_enable_pointer_compression_shared_cage=False,
v8_enable_shared_ro_heap=False,
v8_enable_sandbox=False
)
)
expect_text = (
'>>> Autodetected:\n'
'asan\n'
'cfi_vptr\n'
'dcheck_always_on\n'
'msan\n'
'no_i18n\n'
'tsan\n'
'ubsan_vptr\n'
'webassembly\n'
'>>> Running tests for ia32.release')
result.stdout_includes(expect_text)
result.has_returncode(0)
# TODO(machenbach): Test some more implications of the auto-detected
# options, e.g. that the right env variables are set.
def testLimitedPreloading(self):
result = self.run_tests('--progress=verbose', '--variants=default', '-j1',
'sweet/*')
result.stdout_includes("sweet/mangoes default: PASS")
result.stdout_includes("sweet/cherries default: FAIL")
result.stdout_includes("sweet/apples default: FAIL")
result.stdout_includes("sweet/strawberries default: FAIL")
result.stdout_includes("sweet/bananas default: PASS")
result.stdout_includes("sweet/blackberries default: FAIL")
result.stdout_includes("sweet/raspberries default: PASS")
def testWithFakeContext(self):
with patch(
'testrunner.local.context.find_os_context_factory',
return_value=FakeOSContext):
result = self.run_tests(
'--progress=verbose',
'sweet',
)
result.stdout_includes('===>Starting stuff\n'
'>>> Running tests for x64.release\n'
'>>> Running with test processors\n')
result.stdout_includes('--- stdout ---\nfake stdout 1')
result.stdout_includes('--- stderr ---\nfake stderr 1')
result.stdout_includes('=== sweet/raspberries ===')
result.stdout_includes('=== sweet/cherries ===')
result.stdout_includes('=== sweet/apples ===')
result.stdout_includes('Command: fake_wrapper ')
result.stdout_includes(
'===\n'
'=== 4 tests failed\n'
'===\n'
'>>> 7 base tests produced 7 (100%) non-filtered tests\n'
'>>> 7 tests ran\n'
'<===Stopping stuff\n')
def testSkips(self):
"""Test skipping tests in status file for a specific variant."""
result = self.run_tests(
'--progress=verbose',
'--variants=nooptimization',
'sweet/strawberries',
infra_staging=False,
)
result.stdout_includes('0 tests ran')
result.has_returncode(2)
def testRunSkips(self):
"""Inverse the above. Test parameter to keep running skipped tests."""
result = self.run_tests(
'--progress=verbose',
'--variants=nooptimization',
'--run-skipped',
'sweet/strawberries',
)
result.stdout_includes('1 tests failed')
result.stdout_includes('1 tests ran')
result.has_returncode(1)
def testDefault(self):
"""Test using default test suites, though no tests are run since they don't
exist in a test setting.
"""
result = self.run_tests(
infra_staging=False,
)
result.stdout_includes('0 tests ran')
result.has_returncode(2)
def testNoBuildConfig(self):
"""Test failing run when build config is not found."""
result = self.run_tests(baseroot='wrong_path')
result.stdout_includes('Failed to load build config')
result.has_returncode(5)
def testInconsistentArch(self):
"""Test failing run when attempting to wrongly override the arch."""
result = self.run_tests('--arch=ia32')
result.stdout_includes(
'--arch value (ia32) inconsistent with build config (x64).')
result.has_returncode(5)
def testWrongVariant(self):
"""Test using a bogus variant."""
result = self.run_tests('--variants=meh')
result.has_returncode(5)
def testModeFromBuildConfig(self):
"""Test auto-detection of mode from build config."""
result = self.run_tests('--outdir=out/build', 'sweet/bananas')
result.stdout_includes('Running tests for x64.release')
result.has_returncode(0)
def testPredictable(self):
"""Test running a test in verify-predictable mode.
The test will fail because of missing allocation output. We verify that and
that the predictable flags are passed and printed after failure.
"""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'sweet/bananas',
infra_staging=False,
config_overrides=dict(v8_enable_verify_predictable=True),
)
result.stdout_includes('1 tests ran')
result.stdout_includes('sweet/bananas default: FAIL')
result.stdout_includes('Test had no allocation output')
result.stdout_includes('--predictable --verify-predictable')
result.has_returncode(1)
def testSlowArch(self):
"""Test timeout factor manipulation on slow architecture."""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'sweet/bananas',
config_overrides=dict(v8_target_cpu='arm64'),
)
# TODO(machenbach): We don't have a way for testing if the correct
# timeout was used.
result.has_returncode(0)
def testRandomSeedStressWithDefault(self):
"""Test using random-seed-stress feature has the right number of tests."""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'--random-seed-stress-count=2',
'sweet/bananas',
infra_staging=False,
)
result.stdout_includes('2 tests ran')
result.has_returncode(0)
def testRandomSeedStressWithSeed(self):
"""Test using random-seed-stress feature passing a random seed."""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'--random-seed-stress-count=2',
'--random-seed=123',
'sweet/strawberries',
)
result.stdout_includes('2 tests ran')
# We use a failing test so that the command is printed and we can verify
# that the right random seed was passed.
result.stdout_includes('--random-seed=123')
result.has_returncode(1)
def testSpecificVariants(self):
"""Test using NO_VARIANTS modifiers in status files skips the desire tests.
The test runner cmd line configures 4 tests to run (2 tests * 2 variants).
But the status file applies a modifier to each skipping one of the
variants.
"""
result = self.run_tests(
'--progress=verbose',
'--variants=default,stress',
'sweet/bananas',
'sweet/raspberries',
config_overrides=dict(is_asan=True),
)
# Both tests are either marked as running in only default or only
# slow variant.
result.stdout_includes('2 tests ran')
result.has_returncode(0)
def testDotsProgress(self):
result = self.run_tests(
'--progress=dots',
'sweet/cherries',
'sweet/bananas',
'--no-sorting', '-j1', # make results order deterministic
infra_staging=False,
)
result.stdout_includes('2 tests ran')
result.stdout_includes('F.')
result.has_returncode(1)
def testMonoProgress(self):
self._testCompactProgress('mono')
def testColorProgress(self):
self._testCompactProgress('color')
def _testCompactProgress(self, name):
result = self.run_tests(
'--progress=%s' % name,
'sweet/cherries',
'sweet/bananas',
infra_staging=False,
)
if name == 'color':
expected = ('\033[34m% 28\033[0m|'
'\033[32m+ 1\033[0m|'
'\033[31m- 1\033[0m]: Done')
else:
expected = '% 28|+ 1|- 1]: Done'
result.stdout_includes(expected)
result.stdout_includes('sweet/cherries')
result.stdout_includes('sweet/bananas')
result.has_returncode(1)
def testCompactErrorDetails(self):
result = self.run_tests(
'--progress=mono',
'sweet/cherries',
'sweet/strawberries',
infra_staging=False,
)
result.stdout_includes('sweet/cherries')
result.stdout_includes('sweet/strawberries')
result.stdout_includes('+Mock diff')
result.has_returncode(1)
def testExitAfterNFailures(self):
result = self.run_tests(
'--progress=verbose',
'--exit-after-n-failures=2',
'-j1',
'sweet/mangoes', # PASS
'sweet/strawberries', # FAIL
'sweet/blackberries', # FAIL
'sweet/raspberries', # should not run
)
result.stdout_includes('sweet/mangoes default: PASS')
result.stdout_includes('sweet/strawberries default: FAIL')
result.stdout_includes('Too many failures, exiting...')
result.stdout_includes('sweet/blackberries default: FAIL')
result.stdout_excludes('sweet/raspberries')
result.stdout_includes('2 tests failed')
result.stdout_includes('3 tests ran')
result.has_returncode(1)
def testHeavySequence(self):
"""Test a configuration with 2 heavy tests.
One heavy test will get buffered before being run.
"""
appended = 0
popped = 0
def mock_deque():
class MockQ():
def __init__(self):
self.buffer = deque()
def append(self, test):
nonlocal appended
self.buffer.append(test)
appended += 1
def popleft(self):
nonlocal popped
popped += 1
return self.buffer.popleft()
def __bool__(self):
return bool(self.buffer)
return MockQ()
# Swarming option will trigger a cleanup routine on the bot
def mock_kill():
pass
with patch('testrunner.testproc.sequence.deque', side_effect=mock_deque), \
patch('testrunner.testproc.util.kill_processes_linux', side_effect=mock_kill):
result = self.run_tests(
'--variants=default', '--swarming', 'fat', baseroot="testroot6")
result.has_returncode(1)
self.assertEqual(1, appended)
self.assertEqual(1, popped)
def testRunnerFlags(self):
"""Test that runner-specific flags are passed to tests."""
result = self.run_tests(
'--progress=verbose',
'--variants=default',
'--random-seed=42',
'sweet/bananas',
'-v',
)
result.stdout_includes(
'--test bananas --random-seed=42 --nohard-abort --testing-d8-test-runner')
result.has_returncode(0)
class FakeTimeoutProc(base.TestProcObserver):
"""Fake of the total-timeout observer that just stops after counting
"count" number of test or result events.
"""
def __init__(self, count):
super(FakeTimeoutProc, self).__init__()
self._n = 0
self._count = count
def _on_next_test(self, test):
self.__on_event()
def _on_result_for(self, test, result):
self.__on_event()
def __on_event(self):
if self._n >= self._count:
self.stop()
self._n += 1
class NumFuzzerTest(TestRunnerTest):
def get_runner_class(self):
return num_fuzzer.NumFuzzer
def testNumFuzzer(self):
fuzz_flags = [
f'{flag}=1' for flag in self.get_runner_options()
if flag.startswith('--stress-')
]
self.assertEqual(len(fuzz_flags), len(fuzzer.FUZZERS))
for fuzz_flag in fuzz_flags:
# The fake timeout observer above will stop after proessing the 10th
# test. This still executes an 11th. Each test causes a test- and a
# result event internally. We test both paths here.
for event_count in (19, 20):
with self.subTest(f'fuzz_flag={fuzz_flag} event_count={event_count}'):
with patch(
'testrunner.testproc.timeout.TimeoutProc.create',
lambda x: FakeTimeoutProc(event_count)):
result = self.run_tests(
'--command-prefix', sys.executable,
'--outdir', 'out/build',
'--variants=default',
'--fuzzer-random-seed=12345',
'--total-timeout-sec=60',
fuzz_flag,
'--progress=verbose',
'sweet/bananas',
)
result.has_returncode(0)
result.stdout_includes('>>> Autodetected')
result.stdout_includes('11 tests ran')
class OtherTest(TestRunnerTest):
def testStatusFilePresubmit(self):
"""Test that the fake status file is well-formed."""
with temp_base() as basedir:
from testrunner.local import statusfile
self.assertTrue(statusfile.PresubmitCheck(
os.path.join(basedir, 'test', 'sweet', 'sweet.status')))
if __name__ == '__main__':
unittest.main()