Skip to content

[3.8] bpo-37531: Enhance regrtest multiprocess timeout (GH-15345) #15871

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
vstinner merged 1 commit into from
Sep 10, 2019
Merged

[3.8] bpo-37531: Enhance regrtest multiprocess timeout (GH-15345) #15871

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
81 changes: 63 additions & 18 deletions Lib/test/libregrtest/runtest_mp.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -126,6 +126,38 @@ def __repr__(self):
info.append(f'pid={popen.pid}')
return '<%s>' % ' '.join(info)

def _kill(self):
dt = time.monotonic() - self.start_time

popen = self._popen
pid = popen.pid
print("Kill worker process %s running for %.1f sec" % (pid, dt),
file=sys.stderr, flush=True)

try:
popen.kill()
return True
except OSError as exc:
print("WARNING: Failed to kill worker process %s: %r" % (pid, exc),
file=sys.stderr, flush=True)
return False

def _close_wait(self):
popen = self._popen

# stdout and stderr must be closed to ensure that communicate()
# does not hang
popen.stdout.close()
popen.stderr.close()

try:
popen.wait(JOIN_TIMEOUT)
except (subprocess.TimeoutExpired, OSError) as exc:
print("WARNING: Failed to wait for worker process %s "
"completion (timeout=%.1f sec): %r"
% (popen.pid, JOIN_TIMEOUT, exc),
file=sys.stderr, flush=True)

def kill(self):
"""
Kill the current process (if any).
Expand All @@ -135,30 +167,45 @@ def kill(self):
"""
self._killed = True

popen = self._popen
if popen is None:
if self._popen is None:
return
popen.kill()
# stdout and stderr must be closed to ensure that communicate()
# does not hang
popen.stdout.close()
popen.stderr.close()
popen.wait()

if not self._kill():
return

self._close_wait()

def mp_result_error(self, test_name, error_type, stdout='', stderr='',
err_msg=None):
test_time = time.monotonic() - self.start_time
result = TestResult(test_name, error_type, test_time, None)
return MultiprocessResult(result, stdout, stderr, err_msg)

def _timedout(self, test_name):
self._kill()

stdout = sterr = ''
popen = self._popen
try:
stdout, stderr = popen.communicate(timeout=JOIN_TIMEOUT)
except (subprocess.TimeoutExpired, OSError) as exc:
print("WARNING: Failed to read worker process %s output "
"(timeout=%.1f sec): %r"
% (popen.pid, exc, timeout),
file=sys.stderr, flush=True)

self._close_wait()

return self.mp_result_error(test_name, TIMEOUT, stdout, stderr)

def _runtest(self, test_name):
try:
self.start_time = time.monotonic()
self.current_test_name = test_name

self._popen = run_test_in_subprocess(test_name, self.ns)
popen = self._popen
with popen:
try:
try:
if self._killed:
# If kill() has been called before self._popen is set,
Expand All @@ -175,12 +222,7 @@ def _runtest(self, test_name):
# on reading closed stdout/stderr
raise ExitThread

popen.kill()
stdout, stderr = popen.communicate()
self.kill()

return self.mp_result_error(test_name, TIMEOUT,
stdout, stderr)
return self._timedout(test_name)
except OSError:
if self._killed:
# kill() has been called: communicate() fails
Expand All @@ -190,8 +232,10 @@ def _runtest(self, test_name):
except:
self.kill()
raise
finally:
self._close_wait()

retcode = popen.wait()
retcode = popen.returncode
finally:
self.current_test_name = None
self._popen = None
Expand Down Expand Up @@ -286,10 +330,11 @@ def wait_workers(self):
if not worker.is_alive():
break
dt = time.monotonic() - start_time
print("Wait for regrtest worker %r for %.1f sec" % (worker, dt))
print("Wait for regrtest worker %r for %.1f sec" % (worker, dt),
flush=True)
if dt > JOIN_TIMEOUT:
print("Warning -- failed to join a regrtest worker %s"
% worker)
% worker, flush=True)
break

def _get_result(self):
Expand Down
3 changes: 3 additions & 0 deletions Misc/NEWS.d/next/Tests/2019-08-20-19-24-19.bpo-37531.wRoXfU.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,3 @@
Enhance regrtest multiprocess timeout: write a message when killing a worker
process, catch popen.kill() and popen.wait() exceptions, put a timeout on the
second call to popen.communicate().