gecko-dev/testing/web-platform/harness/wptrunner/testrunner.py

668 строки
25 KiB
Python

# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at http://mozilla.org/MPL/2.0/.
from __future__ import unicode_literals
import multiprocessing
import sys
import threading
import traceback
from Queue import Empty
from multiprocessing import Process, current_process, Queue
from mozlog import structuredlog
# Special value used as a sentinal in various commands
Stop = object()
class MessageLogger(object):
def __init__(self, message_func):
self.send_message = message_func
def _log_data(self, action, **kwargs):
self.send_message("log", action, kwargs)
def process_output(self, process, data, command):
self._log_data("process_output", process=process, data=data, command=command)
def _log_func(level_name):
def log(self, message):
self._log_data(level_name.lower(), message=message)
log.__doc__ = """Log a message with level %s
:param message: The string message to log
""" % level_name
log.__name__ = str(level_name).lower()
return log
# Create all the methods on StructuredLog for debug levels
for level_name in structuredlog.log_levels:
setattr(MessageLogger, level_name.lower(), _log_func(level_name))
class TestRunner(object):
def __init__(self, test_queue, command_queue, result_queue, executor):
"""Class implementing the main loop for running tests.
This class delegates the job of actually running a test to the executor
that is passed in.
:param test_queue: subprocess.Queue containing the tests to run
:param command_queue: subprocess.Queue used to send commands to the
process
:param result_queue: subprocess.Queue used to send results to the
parent TestManager process
:param executor: TestExecutor object that will actually run a test.
"""
self.test_queue = test_queue
self.command_queue = command_queue
self.result_queue = result_queue
self.executor = executor
self.name = current_process().name
self.logger = MessageLogger(self.send_message)
def __enter__(self):
return self
def __exit__(self, exc_type, exc_value, traceback):
self.teardown()
def setup(self):
self.executor.setup(self)
def teardown(self):
self.executor.teardown()
self.send_message("runner_teardown")
self.result_queue = None
self.command_queue = None
self.browser = None
def run(self):
"""Main loop accepting commands over the pipe and triggering
the associated methods"""
self.setup()
commands = {"run_test": self.run_test,
"stop": self.stop,
"wait": self.wait}
while True:
command, args = self.command_queue.get()
try:
rv = commands[command](*args)
except Exception:
self.send_message("error",
"Error running command %s with arguments %r:\n%s" %
(command, args, traceback.format_exc()))
else:
if rv is Stop:
break
def stop(self):
return Stop
def run_test(self):
if not self.executor.is_alive():
self.send_message("restart_runner")
return
try:
# Need to block here just to allow for contention with other processes
test = self.test_queue.get(block=True, timeout=1)
except Empty:
# If we are running tests in groups (e.g. by-dir) then this queue might be
# empty but there could be other test queues. restart_runner won't actually
# start the runner if there aren't any more tests to run
self.send_message("restart_runner")
return
else:
self.send_message("test_start", test)
try:
return self.executor.run_test(test)
except Exception:
self.logger.critical(traceback.format_exc())
raise
def wait(self):
self.executor.protocol.wait()
self.send_message("after_test_ended", True)
def send_message(self, command, *args):
self.result_queue.put((command, args))
def start_runner(test_queue, runner_command_queue, runner_result_queue,
executor_cls, executor_kwargs,
executor_browser_cls, executor_browser_kwargs,
stop_flag):
"""Launch a TestRunner in a new process"""
try:
browser = executor_browser_cls(**executor_browser_kwargs)
executor = executor_cls(browser, **executor_kwargs)
with TestRunner(test_queue, runner_command_queue, runner_result_queue, executor) as runner:
try:
runner.run()
except KeyboardInterrupt:
stop_flag.set()
except Exception:
runner_result_queue.put(("log", ("critical", {"message": traceback.format_exc()})))
print >> sys.stderr, traceback.format_exc()
stop_flag.set()
finally:
runner_command_queue = None
runner_result_queue = None
manager_count = 0
def next_manager_number():
global manager_count
local = manager_count = manager_count + 1
return local
class TestRunnerManager(threading.Thread):
init_lock = threading.Lock()
def __init__(self, suite_name, test_queue, test_source_cls, browser_cls, browser_kwargs,
executor_cls, executor_kwargs, stop_flag, pause_after_test=False,
pause_on_unexpected=False, debug_info=None):
"""Thread that owns a single TestRunner process and any processes required
by the TestRunner (e.g. the Firefox binary).
TestRunnerManagers are responsible for launching the browser process and the
runner process, and for logging the test progress. The actual test running
is done by the TestRunner. In particular they:
* Start the binary of the program under test
* Start the TestRunner
* Tell the TestRunner to start a test, if any
* Log that the test started
* Log the test results
* Take any remedial action required e.g. restart crashed or hung
processes
"""
self.suite_name = suite_name
self.test_queue = test_queue
self.test_source_cls = test_source_cls
self.browser_cls = browser_cls
self.browser_kwargs = browser_kwargs
self.executor_cls = executor_cls
self.executor_kwargs = executor_kwargs
self.test_source = None
self.browser = None
self.browser_pid = None
self.browser_started = False
# Flags used to shut down this thread if we get a sigint
self.parent_stop_flag = stop_flag
self.child_stop_flag = multiprocessing.Event()
self.pause_after_test = pause_after_test
self.pause_on_unexpected = pause_on_unexpected
self.debug_info = debug_info
self.manager_number = next_manager_number()
self.command_queue = Queue()
self.remote_queue = Queue()
self.test_runner_proc = None
threading.Thread.__init__(self, name="Thread-TestrunnerManager-%i" % self.manager_number)
# This is started in the actual new thread
self.logger = None
# The test that is currently running
self.test = None
self.unexpected_count = 0
# This may not really be what we want
self.daemon = True
self.init_fail_count = 0
self.max_init_fails = 5
self.init_timer = None
self.restart_count = 0
self.max_restarts = 5
def run(self):
"""Main loop for the TestManager.
TestManagers generally receive commands from their
TestRunner updating them on the status of a test. They
may also have a stop flag set by the main thread indicating
that the manager should shut down the next time the event loop
spins."""
self.logger = structuredlog.StructuredLogger(self.suite_name)
with self.browser_cls(self.logger, **self.browser_kwargs) as browser, self.test_source_cls(self.test_queue) as test_source:
self.browser = browser
self.test_source = test_source
try:
if self.init() is Stop:
return
while True:
commands = {"init_succeeded": self.init_succeeded,
"init_failed": self.init_failed,
"test_start": self.test_start,
"test_ended": self.test_ended,
"after_test_ended": self.after_test_ended,
"restart_runner": self.restart_runner,
"runner_teardown": self.runner_teardown,
"log": self.log,
"error": self.error}
try:
command, data = self.command_queue.get(True, 1)
except IOError:
if not self.should_stop():
self.logger.error("Got IOError from poll")
self.restart_count += 1
if self.restart_runner() is Stop:
break
except Empty:
command = None
if self.should_stop():
self.logger.debug("A flag was set; stopping")
break
if command is not None:
self.restart_count = 0
if commands[command](*data) is Stop:
break
else:
if (self.debug_info and self.debug_info.interactive and
self.browser_started and not browser.is_alive()):
self.logger.debug("Debugger exited")
break
if not self.test_runner_proc.is_alive():
if not self.command_queue.empty():
# We got a new message so process that
continue
# If we got to here the runner presumably shut down
# unexpectedly
self.logger.info("Test runner process shut down")
if self.test is not None:
# This could happen if the test runner crashed for some other
# reason
# Need to consider the unlikely case where one test causes the
# runner process to repeatedly die
self.logger.critical("Last test did not complete")
break
self.logger.warning(
"More tests found, but runner process died, restarting")
self.restart_count += 1
if self.restart_runner() is Stop:
break
finally:
self.logger.debug("TestRunnerManager main loop terminating, starting cleanup")
self.stop_runner()
self.teardown()
self.logger.debug("TestRunnerManager main loop terminated")
def should_stop(self):
return self.child_stop_flag.is_set() or self.parent_stop_flag.is_set()
def init(self):
"""Launch the browser that is being tested,
and the TestRunner process that will run the tests."""
# It seems that this lock is helpful to prevent some race that otherwise
# sometimes stops the spawned processes initalising correctly, and
# leaves this thread hung
if self.init_timer is not None:
self.init_timer.cancel()
self.logger.debug("Init called, starting browser and runner")
def init_failed():
# This is called from a seperate thread, so we send a message to the
# main loop so we get back onto the manager thread
self.logger.debug("init_failed called from timer")
if self.command_queue:
self.command_queue.put(("init_failed", ()))
else:
self.logger.debug("Setting child stop flag in init_failed")
self.child_stop_flag.set()
with self.init_lock:
# Guard against problems initialising the browser or the browser
# remote control method
if self.debug_info is None:
self.init_timer = threading.Timer(self.browser.init_timeout, init_failed)
test_queue = self.test_source.get_queue()
if test_queue is None:
self.logger.info("No more tests")
return Stop
try:
if self.init_timer is not None:
self.init_timer.start()
self.browser.start()
self.browser_pid = self.browser.pid()
self.start_test_runner(test_queue)
except:
self.logger.warning("Failure during init %s" % traceback.format_exc())
if self.init_timer is not None:
self.init_timer.cancel()
self.logger.error(traceback.format_exc())
succeeded = False
else:
succeeded = True
self.browser_started = True
# This has to happen after the lock is released
if not succeeded:
self.init_failed()
def init_succeeded(self):
"""Callback when we have started the browser, started the remote
control connection, and we are ready to start testing."""
self.logger.debug("Init succeeded")
if self.init_timer is not None:
self.init_timer.cancel()
self.init_fail_count = 0
self.start_next_test()
def init_failed(self):
"""Callback when starting the browser or the remote control connect
fails."""
self.init_fail_count += 1
self.logger.warning("Init failed %i" % self.init_fail_count)
if self.init_timer is not None:
self.init_timer.cancel()
if self.init_fail_count < self.max_init_fails:
self.restart_runner()
else:
self.logger.critical("Test runner failed to initialise correctly; shutting down")
return Stop
def start_test_runner(self, test_queue):
# Note that we need to be careful to start the browser before the
# test runner to ensure that any state set when the browser is started
# can be passed in to the test runner.
assert self.command_queue is not None
assert self.remote_queue is not None
self.logger.info("Starting runner")
executor_browser_cls, executor_browser_kwargs = self.browser.executor_browser()
args = (test_queue,
self.remote_queue,
self.command_queue,
self.executor_cls,
self.executor_kwargs,
executor_browser_cls,
executor_browser_kwargs,
self.child_stop_flag)
self.test_runner_proc = Process(target=start_runner,
args=args,
name="Thread-TestRunner-%i" % self.manager_number)
self.test_runner_proc.start()
self.logger.debug("Test runner started")
def send_message(self, command, *args):
self.remote_queue.put((command, args))
def cleanup(self):
if self.init_timer is not None:
self.init_timer.cancel()
self.logger.debug("TestManager cleanup")
while True:
try:
self.logger.warning(" ".join(map(repr, self.command_queue.get_nowait())))
except Empty:
break
while True:
try:
self.logger.warning(" ".join(map(repr, self.remote_queue.get_nowait())))
except Empty:
break
def teardown(self):
self.logger.debug("teardown in testrunnermanager")
self.test_runner_proc = None
self.command_queue.close()
self.remote_queue.close()
self.command_queue = None
self.remote_queue = None
def ensure_runner_stopped(self):
if self.test_runner_proc is None:
return
self.test_runner_proc.join(10)
if self.test_runner_proc.is_alive():
# This might leak a file handle from the queue
self.logger.warning("Forcibly terminating runner process")
self.test_runner_proc.terminate()
self.test_runner_proc.join(10)
else:
self.logger.debug("Testrunner exited with code %i" % self.test_runner_proc.exitcode)
def runner_teardown(self):
self.ensure_runner_stopped()
return Stop
def stop_runner(self):
"""Stop the TestRunner and the Firefox binary."""
self.logger.debug("Stopping runner")
if self.test_runner_proc is None:
return
try:
self.browser.stop()
self.browser_started = False
if self.test_runner_proc.is_alive():
self.send_message("stop")
self.ensure_runner_stopped()
finally:
self.cleanup()
def start_next_test(self):
self.send_message("run_test")
def test_start(self, test):
self.test = test
self.logger.test_start(test.id)
def test_ended(self, test, results):
"""Handle the end of a test.
Output the result of each subtest, and the result of the overall
harness to the logs.
"""
assert test == self.test
# Write the result of each subtest
file_result, test_results = results
subtest_unexpected = False
for result in test_results:
if test.disabled(result.name):
continue
expected = test.expected(result.name)
is_unexpected = expected != result.status
if is_unexpected:
self.unexpected_count += 1
self.logger.debug("Unexpected count in this thread %i" % self.unexpected_count)
subtest_unexpected = True
self.logger.test_status(test.id,
result.name,
result.status,
message=result.message,
expected=expected,
stack=result.stack)
# TODO: consider changing result if there is a crash dump file
# Write the result of the test harness
expected = test.expected()
status = file_result.status if file_result.status != "EXTERNAL-TIMEOUT" else "TIMEOUT"
is_unexpected = expected != status
if is_unexpected:
self.unexpected_count += 1
self.logger.debug("Unexpected count in this thread %i" % self.unexpected_count)
if status == "CRASH":
self.browser.log_crash(process=self.browser_pid, test=test.id)
self.logger.test_end(test.id,
status,
message=file_result.message,
expected=expected,
extra=file_result.extra)
self.test = None
restart_before_next = (test.restart_after or
file_result.status in ("CRASH", "EXTERNAL-TIMEOUT") or
subtest_unexpected or is_unexpected)
if (self.pause_after_test or
(self.pause_on_unexpected and (subtest_unexpected or is_unexpected))):
self.logger.info("Pausing until the browser exits")
self.send_message("wait")
else:
self.after_test_ended(restart_before_next)
def after_test_ended(self, restart_before_next):
# Handle starting the next test, with a runner restart if required
if restart_before_next:
return self.restart_runner()
else:
return self.start_next_test()
def restart_runner(self):
"""Stop and restart the TestRunner"""
if self.restart_count >= self.max_restarts:
return Stop
self.stop_runner()
return self.init()
def log(self, action, kwargs):
getattr(self.logger, action)(**kwargs)
def error(self, message):
self.logger.error(message)
self.restart_runner()
class TestQueue(object):
def __init__(self, test_source_cls, test_type, tests, **kwargs):
self.queue = None
self.test_source_cls = test_source_cls
self.test_type = test_type
self.tests = tests
self.kwargs = kwargs
def __enter__(self):
if not self.tests[self.test_type]:
return None
self.queue = Queue()
has_tests = self.test_source_cls.queue_tests(self.queue,
self.test_type,
self.tests,
**self.kwargs)
# There is a race condition that means sometimes we continue
# before the tests have been written to the underlying pipe.
# Polling the pipe for data here avoids that
self.queue._reader.poll(10)
assert not self.queue.empty()
return self.queue
def __exit__(self, *args, **kwargs):
if self.queue is not None:
self.queue.close()
self.queue = None
class ManagerGroup(object):
def __init__(self, suite_name, size, test_source_cls, test_source_kwargs,
browser_cls, browser_kwargs,
executor_cls, executor_kwargs,
pause_after_test=False,
pause_on_unexpected=False,
debug_info=None):
"""Main thread object that owns all the TestManager threads."""
self.suite_name = suite_name
self.size = size
self.test_source_cls = test_source_cls
self.test_source_kwargs = test_source_kwargs
self.browser_cls = browser_cls
self.browser_kwargs = browser_kwargs
self.executor_cls = executor_cls
self.executor_kwargs = executor_kwargs
self.pause_after_test = pause_after_test
self.pause_on_unexpected = pause_on_unexpected
self.debug_info = debug_info
self.pool = set()
# Event that is polled by threads so that they can gracefully exit in the face
# of sigint
self.stop_flag = threading.Event()
self.logger = structuredlog.StructuredLogger(suite_name)
self.test_queue = None
def __enter__(self):
return self
def __exit__(self, exc_type, exc_val, exc_tb):
self.stop()
def run(self, test_type, tests):
"""Start all managers in the group"""
self.logger.debug("Using %i processes" % self.size)
self.test_queue = TestQueue(self.test_source_cls,
test_type,
tests,
**self.test_source_kwargs)
with self.test_queue as test_queue:
if test_queue is None:
self.logger.info("No %s tests to run" % test_type)
return
for _ in range(self.size):
manager = TestRunnerManager(self.suite_name,
test_queue,
self.test_source_cls,
self.browser_cls,
self.browser_kwargs,
self.executor_cls,
self.executor_kwargs,
self.stop_flag,
self.pause_after_test,
self.pause_on_unexpected,
self.debug_info)
manager.start()
self.pool.add(manager)
self.wait()
def is_alive(self):
"""Boolean indicating whether any manager in the group is still alive"""
return any(manager.is_alive() for manager in self.pool)
def wait(self):
"""Wait for all the managers in the group to finish"""
for item in self.pool:
item.join()
def stop(self):
"""Set the stop flag so that all managers in the group stop as soon
as possible"""
self.stop_flag.set()
self.logger.debug("Stop flag set in ManagerGroup")
def unexpected_count(self):
return sum(item.unexpected_count for item in self.pool)