1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
|
import _thread
import contextlib
import functools
import sys
import threading
import time
from test import support
#=======================================================================
# Threading support to prevent reporting refleaks when running regrtest.py -R
# NOTE: we use thread._count() rather than threading.enumerate() (or the
# moral equivalent thereof) because a threading.Thread object is still alive
# until its __bootstrap() method has returned, even after it has been
# unregistered from the threading module.
# thread._count(), on the other hand, only gets decremented *after* the
# __bootstrap() method has returned, which gives us reliable reference counts
# at the end of a test run.
def threading_setup():
return _thread._count(), threading._dangling.copy()
def threading_cleanup(*original_values):
_MAX_COUNT = 100
for count in range(_MAX_COUNT):
values = _thread._count(), threading._dangling
if values == original_values:
break
if not count:
# Display a warning at the first iteration
support.environment_altered = True
dangling_threads = values[1]
support.print_warning(f"threading_cleanup() failed to cleanup "
f"{values[0] - original_values[0]} threads "
f"(count: {values[0]}, "
f"dangling: {len(dangling_threads)})")
for thread in dangling_threads:
support.print_warning(f"Dangling thread: {thread!r}")
# Don't hold references to threads
dangling_threads = None
values = None
time.sleep(0.01)
support.gc_collect()
def reap_threads(func):
"""Use this function when threads are being used. This will
ensure that the threads are cleaned up even when the test fails.
"""
@functools.wraps(func)
def decorator(*args):
key = threading_setup()
try:
return func(*args)
finally:
threading_cleanup(*key)
return decorator
@contextlib.contextmanager
def wait_threads_exit(timeout=None):
"""
bpo-31234: Context manager to wait until all threads created in the with
statement exit.
Use _thread.count() to check if threads exited. Indirectly, wait until
threads exit the internal t_bootstrap() C function of the _thread module.
threading_setup() and threading_cleanup() are designed to emit a warning
if a test leaves running threads in the background. This context manager
is designed to cleanup threads started by the _thread.start_new_thread()
which doesn't allow to wait for thread exit, whereas thread.Thread has a
join() method.
"""
if timeout is None:
timeout = support.SHORT_TIMEOUT
old_count = _thread._count()
try:
yield
finally:
start_time = time.monotonic()
deadline = start_time + timeout
while True:
count = _thread._count()
if count <= old_count:
break
if time.monotonic() > deadline:
dt = time.monotonic() - start_time
msg = (f"wait_threads() failed to cleanup {count - old_count} "
f"threads after {dt:.1f} seconds "
f"(count: {count}, old count: {old_count})")
raise AssertionError(msg)
time.sleep(0.010)
support.gc_collect()
def join_thread(thread, timeout=None):
"""Join a thread. Raise an AssertionError if the thread is still alive
after timeout seconds.
"""
if timeout is None:
timeout = support.SHORT_TIMEOUT
thread.join(timeout)
if thread.is_alive():
msg = f"failed to join the thread in {timeout:.1f} seconds"
raise AssertionError(msg)
@contextlib.contextmanager
def start_threads(threads, unlock=None):
import faulthandler
threads = list(threads)
started = []
try:
try:
for t in threads:
t.start()
started.append(t)
except:
if support.verbose:
print("Can't start %d threads, only %d threads started" %
(len(threads), len(started)))
raise
yield
finally:
try:
if unlock:
unlock()
endtime = time.monotonic()
for timeout in range(1, 16):
endtime += 60
for t in started:
t.join(max(endtime - time.monotonic(), 0.01))
started = [t for t in started if t.is_alive()]
if not started:
break
if support.verbose:
print('Unable to join %d threads during a period of '
'%d minutes' % (len(started), timeout))
finally:
started = [t for t in started if t.is_alive()]
if started:
faulthandler.dump_traceback(sys.stdout)
raise AssertionError('Unable to join %d threads' % len(started))
class catch_threading_exception:
"""
Context manager catching threading.Thread exception using
threading.excepthook.
Attributes set when an exception is caught:
* exc_type
* exc_value
* exc_traceback
* thread
See threading.excepthook() documentation for these attributes.
These attributes are deleted at the context manager exit.
Usage:
with threading_helper.catch_threading_exception() as cm:
# code spawning a thread which raises an exception
...
# check the thread exception, use cm attributes:
# exc_type, exc_value, exc_traceback, thread
...
# exc_type, exc_value, exc_traceback, thread attributes of cm no longer
# exists at this point
# (to avoid reference cycles)
"""
def __init__(self):
self.exc_type = None
self.exc_value = None
self.exc_traceback = None
self.thread = None
self._old_hook = None
def _hook(self, args):
self.exc_type = args.exc_type
self.exc_value = args.exc_value
self.exc_traceback = args.exc_traceback
self.thread = args.thread
def __enter__(self):
self._old_hook = threading.excepthook
threading.excepthook = self._hook
return self
def __exit__(self, *exc_info):
threading.excepthook = self._old_hook
del self.exc_type
del self.exc_value
del self.exc_traceback
del self.thread
|