[3.8] bpo-36670, regrtest: Fix WindowsLoadTracker() for partial line (GH-16550) (GH-16560)

* bpo-36670, regrtest: Fix WindowsLoadTracker() for partial line (GH-16550)

WindowsLoadTracker.read_output() now uses a short buffer for
incomplete line.

(cherry picked from commit 3e04cd268e)

* bpo-36670: Enhance regrtest WindowsLoadTracker (GH-16553)

The last line is now passed to the parser even if it does not end
with a newline, but only if it's a valid value.

(cherry picked from commit c65119d5bf)

* bpo-36670: Enhance regrtest (GH-16556)

* Add log() method: add timestamp and load average prefixes
  to main messages.
* WindowsLoadTracker:

  * LOAD_FACTOR_1 is now computed using SAMPLING_INTERVAL
  * Initialize the load to the arithmetic mean of the first 5 values
    of the Processor Queue Length value (so over 5 seconds), rather
    than 0.0.
  * Handle BrokenPipeError and when typeperf exit.

* format_duration(1.5) now returns '1.5 sec', rather than
  '1 sec 500 ms'

(cherry picked from commit 098e25672f)
(cherry picked from commit de3195c937)

Co-authored-by: Victor Stinner <vstinner@python.org>
This commit is contained in:
Miss Islington (bot) 2019-10-03 08:44:52 -07:00 committed by GitHub
parent ab98cd8aee
commit f9016e5fc9
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
5 changed files with 133 additions and 73 deletions

View File

@ -138,16 +138,8 @@ class Regrtest:
print(xml_data, file=sys.__stderr__)
raise
def display_progress(self, test_index, text):
if self.ns.quiet:
return
# "[ 51/405/1] test_tcl passed"
line = f"{test_index:{self.test_count_width}}{self.test_count}"
fails = len(self.bad) + len(self.environment_changed)
if fails and not self.ns.pgo:
line = f"{line}/{fails}"
line = f"[{line}] {text}"
def log(self, line=''):
empty = not line
# add the system load prefix: "load avg: 1.80 "
load_avg = self.getloadavg()
@ -158,8 +150,23 @@ class Regrtest:
test_time = time.monotonic() - self.start_time
test_time = datetime.timedelta(seconds=int(test_time))
line = f"{test_time} {line}"
if empty:
line = line[:-1]
print(line, flush=True)
def display_progress(self, test_index, text):
if self.ns.quiet:
return
# "[ 51/405/1] test_tcl passed"
line = f"{test_index:{self.test_count_width}}{self.test_count}"
fails = len(self.bad) + len(self.environment_changed)
if fails and not self.ns.pgo:
line = f"{line}/{fails}"
self.log(f"[{line}] {text}")
def parse_args(self, kwargs):
ns = _parse_args(sys.argv[1:], **kwargs)
@ -297,11 +304,11 @@ class Regrtest:
self.first_result = self.get_tests_result()
print()
print("Re-running failed tests in verbose mode")
self.log()
self.log("Re-running failed tests in verbose mode")
self.rerun = self.bad[:]
for test_name in self.rerun:
print(f"Re-running {test_name} in verbose mode", flush=True)
self.log(f"Re-running {test_name} in verbose mode")
self.ns.verbose = True
result = runtest(self.ns, test_name)
@ -382,7 +389,7 @@ class Regrtest:
save_modules = sys.modules.keys()
print("Run tests sequentially")
self.log("Run tests sequentially")
previous_test = None
for test_index, test_name in enumerate(self.tests, 1):

View File

@ -104,13 +104,14 @@ class ExitThread(Exception):
class TestWorkerProcess(threading.Thread):
def __init__(self, worker_id, pending, output, ns, timeout):
def __init__(self, worker_id, runner):
super().__init__()
self.worker_id = worker_id
self.pending = pending
self.output = output
self.ns = ns
self.timeout = timeout
self.pending = runner.pending
self.output = runner.output
self.ns = runner.ns
self.timeout = runner.worker_timeout
self.regrtest = runner.regrtest
self.current_test_name = None
self.start_time = None
self._popen = None
@ -292,7 +293,8 @@ class TestWorkerProcess(threading.Thread):
if not self.is_alive():
break
dt = time.monotonic() - start_time
print(f"Waiting for {self} thread for {format_duration(dt)}", flush=True)
self.regrtest.log(f"Waiting for {self} thread "
f"for {format_duration(dt)}")
if dt > JOIN_TIMEOUT:
print_warning(f"Failed to join {self} in {format_duration(dt)}")
break
@ -314,6 +316,7 @@ def get_running(workers):
class MultiprocessTestRunner:
def __init__(self, regrtest):
self.regrtest = regrtest
self.log = self.regrtest.log
self.ns = regrtest.ns
self.output = queue.Queue()
self.pending = MultiprocessIterator(self.regrtest.tests)
@ -324,11 +327,10 @@ class MultiprocessTestRunner:
self.workers = None
def start_workers(self):
self.workers = [TestWorkerProcess(index, self.pending, self.output,
self.ns, self.worker_timeout)
self.workers = [TestWorkerProcess(index, self)
for index in range(1, self.ns.use_mp + 1)]
print("Run tests in parallel using %s child processes"
% len(self.workers))
self.log("Run tests in parallel using %s child processes"
% len(self.workers))
for worker in self.workers:
worker.start()
@ -362,7 +364,7 @@ class MultiprocessTestRunner:
# display progress
running = get_running(self.workers)
if running and not self.ns.pgo:
print('running: %s' % ', '.join(running), flush=True)
self.log('running: %s' % ', '.join(running))
def display_result(self, mp_result):
result = mp_result.result
@ -382,8 +384,7 @@ class MultiprocessTestRunner:
if item[0]:
# Thread got an exception
format_exc = item[1]
print(f"regrtest worker thread failed: {format_exc}",
file=sys.stderr, flush=True)
print_warning(f"regrtest worker thread failed: {format_exc}")
return True
self.test_index += 1

View File

@ -16,11 +16,14 @@ def format_duration(seconds):
if minutes:
parts.append('%s min' % minutes)
if seconds:
parts.append('%s sec' % seconds)
if ms:
parts.append('%s ms' % ms)
if parts:
# 2 min 1 sec
parts.append('%s sec' % seconds)
else:
# 1.0 sec
parts.append('%.1f sec' % (seconds + ms / 1000))
if not parts:
return '0 ms'
return '%s ms' % ms
parts = parts[:2]
return ' '.join(parts)

View File

@ -1,4 +1,5 @@
import _winapi
import math
import msvcrt
import os
import subprocess
@ -10,11 +11,14 @@ from test.libregrtest.utils import print_warning
# Max size of asynchronous reads
BUFSIZE = 8192
# Exponential damping factor (see below)
LOAD_FACTOR_1 = 0.9200444146293232478931553241
# Seconds per measurement
SAMPLING_INTERVAL = 5
SAMPLING_INTERVAL = 1
# Exponential damping factor to compute exponentially weighted moving average
# on 1 minute (60 seconds)
LOAD_FACTOR_1 = 1 / math.exp(SAMPLING_INTERVAL / 60)
# Initialize the load using the arithmetic mean of the first NVALUE values
# of the Processor Queue Length
NVALUE = 5
# Windows registry subkey of HKEY_LOCAL_MACHINE where the counter names
# of typeperf are registered
COUNTER_REGISTRY_KEY = (r"SOFTWARE\Microsoft\Windows NT\CurrentVersion"
@ -30,9 +34,10 @@ class WindowsLoadTracker():
"""
def __init__(self):
self.load = 0.0
self.counter_name = ''
self.popen = None
self._values = []
self._load = None
self._buffer = ''
self._popen = None
self.start()
def start(self):
@ -64,7 +69,7 @@ class WindowsLoadTracker():
# Spawn off the load monitor
counter_name = self._get_counter_name()
command = ['typeperf', counter_name, '-si', str(SAMPLING_INTERVAL)]
self.popen = subprocess.Popen(' '.join(command), stdout=command_stdout, cwd=support.SAVEDCWD)
self._popen = subprocess.Popen(' '.join(command), stdout=command_stdout, cwd=support.SAVEDCWD)
# Close our copy of the write end of the pipe
os.close(command_stdout)
@ -84,52 +89,88 @@ class WindowsLoadTracker():
process_queue_length = counters_dict['44']
return f'"\\{system}\\{process_queue_length}"'
def close(self):
if self.popen is None:
def close(self, kill=True):
if self._popen is None:
return
self.popen.kill()
self.popen.wait()
self.popen = None
self._load = None
if kill:
self._popen.kill()
self._popen.wait()
self._popen = None
def __del__(self):
self.close()
def read_output(self):
def _parse_line(self, line):
# typeperf outputs in a CSV format like this:
# "07/19/2018 01:32:26.605","3.000000"
# (date, process queue length)
tokens = line.split(',')
if len(tokens) != 2:
raise ValueError
value = tokens[1]
if not value.startswith('"') or not value.endswith('"'):
raise ValueError
value = value[1:-1]
return float(value)
def _read_lines(self):
overlapped, _ = _winapi.ReadFile(self.pipe, BUFSIZE, True)
bytes_read, res = overlapped.GetOverlappedResult(False)
if res != 0:
return
return ()
output = overlapped.getbuffer()
return output.decode('oem', 'replace')
output = output.decode('oem', 'replace')
output = self._buffer + output
lines = output.splitlines(True)
# bpo-36670: typeperf only writes a newline *before* writing a value,
# not after. Sometimes, the written line in incomplete (ex: only
# timestamp, without the process queue length). Only pass the last line
# to the parser if it's a valid value, otherwise store it in
# self._buffer.
try:
self._parse_line(lines[-1])
except ValueError:
self._buffer = lines.pop(-1)
else:
self._buffer = ''
return lines
def getloadavg(self):
typeperf_output = self.read_output()
# Nothing to update, just return the current load
if not typeperf_output:
return self.load
if self._popen is None:
return None
returncode = self._popen.poll()
if returncode is not None:
self.close(kill=False)
return None
try:
lines = self._read_lines()
except BrokenPipeError:
self.close()
return None
for line in lines:
line = line.rstrip()
# Process the backlog of load values
for line in typeperf_output.splitlines():
# Ignore the initial header:
# "(PDH-CSV 4.0)","\\\\WIN\\System\\Processor Queue Length"
if '\\\\' in line:
if 'PDH-CSV' in line:
continue
# Ignore blank lines
if not line.strip():
if not line:
continue
# typeperf outputs in a CSV format like this:
# "07/19/2018 01:32:26.605","3.000000"
# (date, process queue length)
try:
tokens = line.split(',')
if len(tokens) != 2:
raise ValueError
value = tokens[1].replace('"', '')
load = float(value)
processor_queue_length = self._parse_line(line)
except ValueError:
print_warning("Failed to parse typeperf output: %a" % line)
continue
@ -137,7 +178,13 @@ class WindowsLoadTracker():
# We use an exponentially weighted moving average, imitating the
# load calculation on Unix systems.
# https://en.wikipedia.org/wiki/Load_(computing)#Unix-style_load_calculation
new_load = self.load * LOAD_FACTOR_1 + load * (1.0 - LOAD_FACTOR_1)
self.load = new_load
# https://en.wikipedia.org/wiki/Moving_average#Exponential_moving_average
if self._load is not None:
self._load = (self._load * LOAD_FACTOR_1
+ processor_queue_length * (1.0 - LOAD_FACTOR_1))
elif len(self._values) < NVALUE:
self._values.append(processor_queue_length)
else:
self._load = sum(self._values) / len(self._values)
return self.load
return self._load

View File

@ -24,6 +24,7 @@ from test.libregrtest import utils
Py_DEBUG = hasattr(sys, 'gettotalrefcount')
ROOT_DIR = os.path.join(os.path.dirname(__file__), '..', '..')
ROOT_DIR = os.path.abspath(os.path.normpath(ROOT_DIR))
LOG_PREFIX = r'[0-9]+:[0-9]+:[0-9]+ (?:load avg: [0-9]+\.[0-9]{2} )?'
TEST_INTERRUPTED = textwrap.dedent("""
from signal import SIGINT
@ -390,8 +391,8 @@ class BaseTestCase(unittest.TestCase):
self.assertRegex(output, regex)
def parse_executed_tests(self, output):
regex = (r'^[0-9]+:[0-9]+:[0-9]+ (?:load avg: [0-9]+\.[0-9]{2} )?\[ *[0-9]+(?:/ *[0-9]+)*\] (%s)'
% self.TESTNAME_REGEX)
regex = (r'^%s\[ *[0-9]+(?:/ *[0-9]+)*\] (%s)'
% (LOG_PREFIX, self.TESTNAME_REGEX))
parser = re.finditer(regex, output, re.MULTILINE)
return list(match.group(1) for match in parser)
@ -451,9 +452,10 @@ class BaseTestCase(unittest.TestCase):
if rerun:
regex = list_regex('%s re-run test%s', rerun)
self.check_line(output, regex)
self.check_line(output, "Re-running failed tests in verbose mode")
regex = LOG_PREFIX + r"Re-running failed tests in verbose mode"
self.check_line(output, regex)
for test_name in rerun:
regex = f"Re-running {test_name} in verbose mode"
regex = LOG_PREFIX + f"Re-running {test_name} in verbose mode"
self.check_line(output, regex)
if no_test_ran:
@ -1173,9 +1175,9 @@ class TestUtils(unittest.TestCase):
self.assertEqual(utils.format_duration(10e-3),
'10 ms')
self.assertEqual(utils.format_duration(1.5),
'1 sec 500 ms')
'1.5 sec')
self.assertEqual(utils.format_duration(1),
'1 sec')
'1.0 sec')
self.assertEqual(utils.format_duration(2 * 60),
'2 min')
self.assertEqual(utils.format_duration(2 * 60 + 1),