cpython/Lib/test/libregrtest/main.py

643 lines
22 KiB
Python
Raw Normal View History

import datetime
import faulthandler
import json
2016-09-09 01:46:56 -03:00
import locale
import os
import platform
import random
import re
import sys
import sysconfig
import tempfile
import time
import unittest
from test.libregrtest.cmdline import _parse_args
from test.libregrtest.runtest import (
findtests, runtest, get_abs_module,
STDTESTS, NOTTESTS, PASSED, FAILED, ENV_CHANGED, SKIPPED, RESOURCE_DENIED,
INTERRUPTED, CHILD_ERROR, TEST_DID_NOT_RUN,
PROGRESS_MIN_TIME, format_test_result)
from test.libregrtest.setup import setup_tests
from test.libregrtest.utils import removepy, count, format_duration, printlist
from test import support
# When tests are run from the Python build directory, it is best practice
# to keep the test files in a subfolder. This eases the cleanup of leftover
# files using the "make distclean" command.
if sysconfig.is_python_build():
TEMPDIR = sysconfig.get_config_var('abs_builddir')
if TEMPDIR is None:
# bpo-30284: On Windows, only srcdir is available. Using abs_builddir
# mostly matters on UNIX when building Python out of the source tree,
# especially when the source tree is read only.
TEMPDIR = sysconfig.get_config_var('srcdir')
TEMPDIR = os.path.join(TEMPDIR, 'build')
else:
TEMPDIR = tempfile.gettempdir()
TEMPDIR = os.path.abspath(TEMPDIR)
class Regrtest:
"""Execute a test suite.
This also parses command-line options and modifies its behavior
accordingly.
tests -- a list of strings containing test names (optional)
testdir -- the directory in which to look for tests (optional)
Users other than the Python test suite will certainly want to
specify testdir; if it's omitted, the directory containing the
Python test suite is searched for.
If the tests argument is omitted, the tests listed on the
command-line will be used. If that's empty, too, then all *.py
files beginning with test_ will be used.
The other default arguments (verbose, quiet, exclude,
single, randomize, findleaks, use_resources, trace, coverdir,
print_slow, and random_seed) allow programmers calling main()
directly to set the values that would normally be set by flags
on the command line.
"""
def __init__(self):
# Namespace of command line options
self.ns = None
# tests
self.tests = []
self.selected = []
# test results
self.good = []
self.bad = []
self.skipped = []
self.resource_denieds = []
self.environment_changed = []
self.run_no_tests = []
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.rerun = []
self.first_result = None
self.interrupted = False
# used by --slow
self.test_times = []
# used by --coverage, trace.Trace instance
self.tracer = None
# used to display the progress bar "[ 3/100]"
self.start_time = time.monotonic()
self.test_count = ''
self.test_count_width = 1
# used by --single
self.next_single_test = None
self.next_single_filename = None
# used by --junit-xml
self.testsuite_xml = None
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.win_load_tracker = None
def get_executed(self):
return (set(self.good) | set(self.bad) | set(self.skipped)
| set(self.resource_denieds) | set(self.environment_changed)
| set(self.run_no_tests))
def accumulate_result(self, result, rerun=False):
test_name = result.test_name
ok = result.result
if ok not in (CHILD_ERROR, INTERRUPTED) and not rerun:
self.test_times.append((result.test_time, test_name))
if ok == PASSED:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.good.append(test_name)
elif ok in (FAILED, CHILD_ERROR):
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
if not rerun:
self.bad.append(test_name)
elif ok == ENV_CHANGED:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.environment_changed.append(test_name)
elif ok == SKIPPED:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.skipped.append(test_name)
elif ok == RESOURCE_DENIED:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.skipped.append(test_name)
self.resource_denieds.append(test_name)
elif ok == TEST_DID_NOT_RUN:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.run_no_tests.append(test_name)
elif ok == INTERRUPTED:
self.interrupted = True
else:
raise ValueError("invalid test result: %r" % ok)
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
if rerun and ok not in {FAILED, CHILD_ERROR, INTERRUPTED}:
self.bad.remove(test_name)
xml_data = result.xml_data
if xml_data:
import xml.etree.ElementTree as ET
for e in xml_data:
try:
self.testsuite_xml.append(ET.fromstring(e))
except ET.ParseError:
print(xml_data, file=sys.__stderr__)
raise
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
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}"
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
line = f"[{line}] {text}"
# add the system load prefix: "load avg: 1.80 "
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
load_avg = self.getloadavg()
if load_avg is not None:
line = f"load avg: {load_avg:.2f} {line}"
# add the timestamp prefix: "0:01:05 "
test_time = time.monotonic() - self.start_time
test_time = datetime.timedelta(seconds=int(test_time))
line = f"{test_time} {line}"
print(line, flush=True)
def parse_args(self, kwargs):
ns = _parse_args(sys.argv[1:], **kwargs)
if ns.timeout and not hasattr(faulthandler, 'dump_traceback_later'):
print("Warning: The timeout option requires "
"faulthandler.dump_traceback_later", file=sys.stderr)
ns.timeout = None
if ns.xmlpath:
support.junit_xml_list = self.testsuite_xml = []
# Strip .py extensions.
removepy(ns.args)
return ns
def find_tests(self, tests):
self.tests = tests
if self.ns.single:
self.next_single_filename = os.path.join(TEMPDIR, 'pynexttest')
try:
with open(self.next_single_filename, 'r') as fp:
next_test = fp.read().strip()
self.tests = [next_test]
except OSError:
pass
if self.ns.fromfile:
self.tests = []
# regex to match 'test_builtin' in line:
# '0:00:00 [ 4/400] test_builtin -- test_dict took 1 sec'
regex = re.compile(r'\btest_[a-zA-Z0-9_]+\b')
with open(os.path.join(support.SAVEDCWD, self.ns.fromfile)) as fp:
for line in fp:
line = line.split('#', 1)[0]
line = line.strip()
match = regex.search(line)
if match is not None:
self.tests.append(match.group())
removepy(self.tests)
stdtests = STDTESTS[:]
nottests = NOTTESTS.copy()
if self.ns.exclude:
for arg in self.ns.args:
if arg in stdtests:
stdtests.remove(arg)
nottests.add(arg)
self.ns.args = []
# if testdir is set, then we are not running the python tests suite, so
# don't add default tests to be executed or skipped (pass empty values)
if self.ns.testdir:
alltests = findtests(self.ns.testdir, list(), set())
else:
alltests = findtests(self.ns.testdir, stdtests, nottests)
if not self.ns.fromfile:
self.selected = self.tests or self.ns.args or alltests
else:
self.selected = self.tests
if self.ns.single:
self.selected = self.selected[:1]
try:
pos = alltests.index(self.selected[0])
self.next_single_test = alltests[pos + 1]
except IndexError:
pass
# Remove all the selected tests that precede start if it's set.
if self.ns.start:
try:
del self.selected[:self.selected.index(self.ns.start)]
except ValueError:
print("Couldn't find starting test (%s), using all tests"
% self.ns.start, file=sys.stderr)
if self.ns.randomize:
if self.ns.random_seed is None:
self.ns.random_seed = random.randrange(10000000)
random.seed(self.ns.random_seed)
random.shuffle(self.selected)
def list_tests(self):
for name in self.selected:
print(name)
def _list_cases(self, suite):
for test in suite:
if isinstance(test, unittest.loader._FailedTest):
continue
if isinstance(test, unittest.TestSuite):
self._list_cases(test)
elif isinstance(test, unittest.TestCase):
if support.match_test(test):
print(test.id())
def list_cases(self):
support.verbose = False
support.set_match_tests(self.ns.match_tests)
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
for test_name in self.selected:
abstest = get_abs_module(self.ns, test_name)
try:
suite = unittest.defaultTestLoader.loadTestsFromName(abstest)
self._list_cases(suite)
except unittest.SkipTest:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
self.skipped.append(test_name)
if self.skipped:
print(file=sys.stderr)
print(count(len(self.skipped), "test"), "skipped:", file=sys.stderr)
printlist(self.skipped, file=sys.stderr)
def rerun_failed_tests(self):
self.ns.verbose = True
self.ns.failfast = False
self.ns.verbose3 = False
self.first_result = self.get_tests_result()
print()
print("Re-running failed tests in verbose mode")
self.rerun = self.bad[:]
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
for test_name in self.rerun:
print(f"Re-running {test_name} in verbose mode", flush=True)
self.ns.verbose = True
result = runtest(self.ns, test_name)
self.accumulate_result(result, rerun=True)
if result.result == INTERRUPTED:
break
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
if self.bad:
print(count(len(self.bad), 'test'), "failed again:")
printlist(self.bad)
self.display_result()
def display_result(self):
# If running the test suite for PGO then no one cares about results.
if self.ns.pgo:
return
print()
print("== Tests result: %s ==" % self.get_tests_result())
if self.interrupted:
print("Test suite interrupted by signal SIGINT.")
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
omitted = set(self.selected) - self.get_executed()
if omitted:
print()
print(count(len(omitted), "test"), "omitted:")
printlist(omitted)
if self.good and not self.ns.quiet:
print()
if (not self.bad
and not self.skipped
and not self.interrupted
and len(self.good) > 1):
print("All", end=' ')
print(count(len(self.good), "test"), "OK.")
if self.ns.print_slow:
self.test_times.sort(reverse=True)
print()
print("10 slowest tests:")
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
for test_time, test in self.test_times[:10]:
print("- %s: %s" % (test, format_duration(test_time)))
if self.bad:
print()
print(count(len(self.bad), "test"), "failed:")
printlist(self.bad)
if self.environment_changed:
print()
print("{} altered the execution environment:".format(
count(len(self.environment_changed), "test")))
printlist(self.environment_changed)
if self.skipped and not self.ns.quiet:
print()
print(count(len(self.skipped), "test"), "skipped:")
printlist(self.skipped)
if self.rerun:
print()
print("%s:" % count(len(self.rerun), "re-run test"))
printlist(self.rerun)
if self.run_no_tests:
print()
print(count(len(self.run_no_tests), "test"), "run no tests:")
printlist(self.run_no_tests)
def run_tests_sequential(self):
if self.ns.trace:
import trace
self.tracer = trace.Trace(trace=False, count=True)
save_modules = sys.modules.keys()
print("Run tests sequentially")
previous_test = None
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
for test_index, test_name in enumerate(self.tests, 1):
start_time = time.monotonic()
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
text = test_name
if previous_test:
text = '%s -- %s' % (text, previous_test)
self.display_progress(test_index, text)
if self.tracer:
# If we're tracing code coverage, then we don't exit with status
# if on a false return value from main.
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
cmd = ('result = runtest(self.ns, test_name); '
'self.accumulate_result(result)')
ns = dict(locals())
self.tracer.runctx(cmd, globals=globals(), locals=ns)
result = ns['result']
else:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
result = runtest(self.ns, test_name)
self.accumulate_result(result)
if result.result == INTERRUPTED:
break
previous_test = format_test_result(result)
test_time = time.monotonic() - start_time
if test_time >= PROGRESS_MIN_TIME:
previous_test = "%s in %s" % (previous_test, format_duration(test_time))
elif result[0] == PASSED:
# be quiet: say nothing if the test passed shortly
previous_test = None
# Unload the newly imported modules (best effort finalization)
for module in sys.modules.keys():
if module not in save_modules and module.startswith("test."):
support.unload(module)
if previous_test:
print(previous_test)
def _test_forever(self, tests):
while True:
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
for test_name in tests:
yield test_name
if self.bad:
return
if self.ns.fail_env_changed and self.environment_changed:
return
def display_header(self):
# Print basic platform information
print("==", platform.python_implementation(), *sys.version.split())
print("==", platform.platform(aliased=True),
"%s-endian" % sys.byteorder)
print("== cwd:", os.getcwd())
cpu_count = os.cpu_count()
if cpu_count:
print("== CPU count:", cpu_count)
print("== encodings: locale=%s, FS=%s"
% (locale.getpreferredencoding(False),
sys.getfilesystemencoding()))
def get_tests_result(self):
result = []
if self.bad:
result.append("FAILURE")
elif self.ns.fail_env_changed and self.environment_changed:
result.append("ENV CHANGED")
elif not any((self.good, self.bad, self.skipped, self.interrupted,
self.environment_changed)):
result.append("NO TEST RUN")
if self.interrupted:
result.append("INTERRUPTED")
if not result:
result.append("SUCCESS")
result = ', '.join(result)
if self.first_result:
result = '%s then %s' % (self.first_result, result)
return result
def run_tests(self):
# For a partial run, we do not need to clutter the output.
if (self.ns.header
or not(self.ns.pgo or self.ns.quiet or self.ns.single
or self.tests or self.ns.args)):
self.display_header()
if self.ns.huntrleaks:
warmup, repetitions, _ = self.ns.huntrleaks
if warmup < 3:
msg = ("WARNING: Running tests with --huntrleaks/-R and less than "
"3 warmup repetitions can give false positives!")
print(msg, file=sys.stdout, flush=True)
if self.ns.randomize:
print("Using random seed", self.ns.random_seed)
if self.ns.forever:
self.tests = self._test_forever(list(self.selected))
self.test_count = ''
self.test_count_width = 3
else:
self.tests = iter(self.selected)
self.test_count = '/{}'.format(len(self.selected))
self.test_count_width = len(self.test_count) - 1
if self.ns.use_mp:
from test.libregrtest.runtest_mp import run_tests_multiprocess
run_tests_multiprocess(self)
else:
self.run_tests_sequential()
def finalize(self):
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
if self.win_load_tracker is not None:
self.win_load_tracker.close()
self.win_load_tracker = None
if self.next_single_filename:
if self.next_single_test:
with open(self.next_single_filename, 'w') as fp:
fp.write(self.next_single_test + '\n')
else:
os.unlink(self.next_single_filename)
if self.tracer:
r = self.tracer.results()
r.write_results(show_missing=True, summary=True,
coverdir=self.ns.coverdir)
print()
duration = time.monotonic() - self.start_time
print("Total duration: %s" % format_duration(duration))
print("Tests result: %s" % self.get_tests_result())
if self.ns.runleaks:
os.system("leaks %d" % os.getpid())
def save_xml_result(self):
if not self.ns.xmlpath and not self.testsuite_xml:
return
import xml.etree.ElementTree as ET
root = ET.Element("testsuites")
# Manually count the totals for the overall summary
totals = {'tests': 0, 'errors': 0, 'failures': 0}
for suite in self.testsuite_xml:
root.append(suite)
for k in totals:
try:
totals[k] += int(suite.get(k, 0))
except ValueError:
pass
for k, v in totals.items():
root.set(k, str(v))
xmlpath = os.path.join(support.SAVEDCWD, self.ns.xmlpath)
with open(xmlpath, 'wb') as f:
for s in ET.tostringlist(root):
f.write(s)
def main(self, tests=None, **kwargs):
global TEMPDIR
self.ns = self.parse_args(kwargs)
if self.ns.tempdir:
TEMPDIR = self.ns.tempdir
elif self.ns.worker_args:
ns_dict, _ = json.loads(self.ns.worker_args)
TEMPDIR = ns_dict.get("tempdir") or TEMPDIR
os.makedirs(TEMPDIR, exist_ok=True)
# Define a writable temp dir that will be used as cwd while running
# the tests. The name of the dir includes the pid to allow parallel
# testing (see the -j option).
test_cwd = 'test_python_{}'.format(os.getpid())
test_cwd = os.path.join(TEMPDIR, test_cwd)
# Run the tests in a context manager that temporarily changes the CWD to a
# temporary and writable directory. If it's not possible to create or
# change the CWD, the original CWD will be used. The original CWD is
# available from support.SAVEDCWD.
with support.temp_cwd(test_cwd, quiet=True):
self._main(tests, kwargs)
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
def getloadavg(self):
if self.win_load_tracker is not None:
return self.win_load_tracker.getloadavg()
if hasattr(os, 'getloadavg'):
return os.getloadavg()[0]
return None
def _main(self, tests, kwargs):
if self.ns.huntrleaks:
warmup, repetitions, _ = self.ns.huntrleaks
if warmup < 1 or repetitions < 1:
msg = ("Invalid values for the --huntrleaks/-R parameters. The "
"number of warmups and repetitions must be at least 1 "
"each (1:1).")
print(msg, file=sys.stderr, flush=True)
sys.exit(2)
if self.ns.worker_args is not None:
from test.libregrtest.runtest_mp import run_tests_worker
run_tests_worker(self.ns.worker_args)
if self.ns.wait:
input("Press any key to continue...")
support.PGO = self.ns.pgo
setup_tests(self.ns)
self.find_tests(tests)
if self.ns.list_tests:
self.list_tests()
sys.exit(0)
if self.ns.list_cases:
self.list_cases()
sys.exit(0)
[3.7] bpo-36719: sync regrtest with master branch (GH-12967) * Clean up code which checked presence of os.{stat,lstat,chmod} (GH-11643) (cherry picked from commit 8377cd4fcd0d51d86834c9b0518d29aac3b49e18) * bpo-36725: regrtest: add TestResult type (GH-12960) * Add TestResult and MultiprocessResult types to ensure that results always have the same fields. * runtest() now handles KeyboardInterrupt * accumulate_result() and format_test_result() now takes a TestResult * cleanup_test_droppings() is now called by runtest() and mark the test as ENV_CHANGED if the test leaks support.TESTFN file. * runtest() now includes code "around" the test in the test timing * Add print_warning() in test.libregrtest.utils to standardize how libregrtest logs warnings to ease parsing the test output. * support.unload() is now called with abstest rather than test_name * Rename 'test' variable/parameter to 'test_name' * dash_R(): remove unused the_module parameter * Remove unused imports (cherry picked from commit 4d29983185bc12ca685a1eb3873bacb8a7b67416) * bpo-36725: Refactor regrtest multiprocessing code (GH-12961) Rewrite run_tests_multiprocess() function as a new MultiprocessRunner class with multiple methods to better report errors and stop immediately when needed. Changes: * Worker processes are now killed immediately if tests are interrupted or if a test does crash (CHILD_ERROR): worker processes are killed. * Rewrite how errors in a worker thread are reported to the main thread. No longer ignore BaseException or parsing errors silently. * Remove 'finished' variable: use worker.is_alive() instead * Always compute omitted tests. Add Regrtest.get_executed() method. (cherry picked from commit 3cde440f20a9db75fb2c4e65e8e4d04a53216a2d) * bpo-36719: regrtest always detect uncollectable objects (GH-12951) regrtest now always detects uncollectable objects. Previously, the check was only enabled by --findleaks. The check now also works with -jN/--multiprocess N. --findleaks becomes a deprecated alias to --fail-env-changed. (cherry picked from commit 75120d2205af086140e5e4e2dc620eb19cdf9078) * bpo-34060: Report system load when running test suite for Windows (GH-8357) While Windows exposes the system processor queue length, the raw value used for load calculations on Unix systems, it does not provide an API to access the averaged value. Hence to calculate the load we must track and average it ourselves. We can't use multiprocessing or a thread to read it in the background while the tests run since using those would conflict with test_multiprocessing and test_xxsubprocess. Thus, we use Window's asynchronous IO API to run the tracker in the background with it sampling at the correct rate. When we wish to access the load we check to see if there's new data on the stream, if there is, we update our load values. (cherry picked from commit e16467af0bfcc9f399df251495ff2d2ad20a1669) * bpo-36719: Fix regrtest re-run (GH-12964) Properly handle a test which fail but then pass. Add test_rerun_success() unit test. (cherry picked from commit 837acc1957d86ca950433f5064fd06d09b57d23b) * bpo-36719: regrtest closes explicitly WindowsLoadTracker (GH-12965) Regrtest.finalize() now closes explicitly the WindowsLoadTracker instance. (cherry picked from commit 00db7c73af4f60df61e9df87cde7401c3ed9df69)
2019-04-26 07:16:30 -03:00
# If we're on windows and this is the parent runner (not a worker),
# track the load average.
if sys.platform == 'win32' and (self.ns.worker_args is None):
from test.libregrtest.win_utils import WindowsLoadTracker
try:
self.win_load_tracker = WindowsLoadTracker()
except FileNotFoundError as error:
# Windows IoT Core and Windows Nano Server do not provide
# typeperf.exe for x64, x86 or ARM
print(f'Failed to create WindowsLoadTracker: {error}')
self.run_tests()
self.display_result()
if self.ns.verbose2 and self.bad:
self.rerun_failed_tests()
self.finalize()
self.save_xml_result()
if self.bad:
sys.exit(2)
if self.interrupted:
sys.exit(130)
if self.ns.fail_env_changed and self.environment_changed:
sys.exit(3)
sys.exit(0)
def main(tests=None, **kwargs):
"""Run the Python suite."""
Regrtest().main(tests=tests, **kwargs)