Mailman 3 February 2014 - Python-checkins

cpython: Issue #20452: Fix test_time_and_call_at() of test_asyncio on Windows
by victor.stinner Jan. 31, 2014

Jan. 31, 2014

http://hg.python.org/cpython/rev/60a960434e5c changeset: 88872:60a960434e5c user: Victor Stinner <victor.stinner(a)gmail.com> date: Sat Feb 01 02:18:52 2014 +0100 summary: Issue #20452: Fix test_time_and_call_at() of test_asyncio on Windows Use the granularity to check the minimum time delta, instead of arbitrary value. files: Lib/test/test_asyncio/test_base_events.py | 15 +++++----- 1 files changed, 8 insertions(+), 7 deletions(-) diff --git a/Lib/test/test_asyncio/test_base_events.py b/Lib/test/test_asyncio/test_base_events.py --- a/Lib/test/test_asyncio/test_base_events.py +++ b/Lib/test/test_asyncio/test_base_events.py @@ -116,17 +116,18 @@ self.loop.stop() self.loop._process_events = unittest.mock.Mock() - when = self.loop.time() + 0.1 + delay = 0.1 + + when = self.loop.time() + delay self.loop.call_at(when, cb) t0 = self.loop.time() self.loop.run_forever() dt = self.loop.time() - t0 - self.assertTrue(0.09 <= dt <= 0.9, - # Issue #20452: add more info in case of failure, - # to try to investigate the bug - (dt, - self.loop._granularity, - time.get_clock_info('monotonic'))) + + self.assertGreaterEqual(dt, delay - self.loop._granularity, dt) + # tolerate a difference of +800 ms because some Python buildbots + # are really slow + self.assertLessEqual(dt, 0.9, dt) def test_run_once_in_executor_handle(self): def cb(): -- Repository URL: http://hg.python.org/cpython

1 0

cpython: Issue #20452: test_asyncio checks also the granularity
by victor.stinner Jan. 31, 2014

Jan. 31, 2014

http://hg.python.org/cpython/rev/a090804862f8 changeset: 88871:a090804862f8 user: Victor Stinner <victor.stinner(a)gmail.com> date: Sat Feb 01 02:17:54 2014 +0100 summary: Issue #20452: test_asyncio checks also the granularity files: Lib/test/test_asyncio/test_events.py | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-) diff --git a/Lib/test/test_asyncio/test_events.py b/Lib/test/test_asyncio/test_events.py --- a/Lib/test/test_asyncio/test_events.py +++ b/Lib/test/test_asyncio/test_events.py @@ -1179,6 +1179,14 @@ calls.append(self.loop._run_once_counter) self.assertEqual(calls, [1, 3, 5, 6]) + def test_granularity(self): + granularity = self.loop._granularity + self.assertGreater(granularity, 0.0) + # Worst expected granularity: 1 ms on Linux (limited by poll/epoll + # resolution), 15.6 ms on Windows (limited by time.monotonic + # resolution) + self.assertLess(granularity, 0.050) + class SubprocessTestsMixin: -- Repository URL: http://hg.python.org/cpython

1 0

cpython: Fix extra line added in 0bcf23a52d55
by brian.quinlan Jan. 31, 2014

Jan. 31, 2014

http://hg.python.org/cpython/rev/4f59b836fa54 changeset: 88870:4f59b836fa54 user: Brian Quinlan <brian(a)sweetapp.com> date: Sat Feb 01 12:07:54 2014 +1100 summary: Fix extra line added in 0bcf23a52d55 files: Misc/NEWS | 3 --- 1 files changed, 0 insertions(+), 3 deletions(-) diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -72,9 +72,6 @@ - Issue #15304: concurrent.futures.wait() can block forever even if Futures have completed. Patch by Glenn Langford. -Fix warning message when `os.chdir()` fails inside - `test.support.temp_cwd()`. Patch by Chris Jerdonek. - IDLE ---- -- Repository URL: http://hg.python.org/cpython

1 0

cpython: Issue #20319: concurrent.futures.wait() can block forever even if Futures have
by brian.quinlan Jan. 31, 2014

Jan. 31, 2014

http://hg.python.org/cpython/rev/0bcf23a52d55 changeset: 88869:0bcf23a52d55 user: Brian Quinlan <brian(a)sweetapp.com> date: Sat Feb 01 11:49:04 2014 +1100 summary: Issue #20319: concurrent.futures.wait() can block forever even if Futures have completed files: Lib/concurrent/futures/_base.py | 6 ++++-- Misc/NEWS | 6 ++++++ 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/Lib/concurrent/futures/_base.py b/Lib/concurrent/futures/_base.py --- a/Lib/concurrent/futures/_base.py +++ b/Lib/concurrent/futures/_base.py @@ -225,7 +225,8 @@ finally: for f in fs: - f._waiters.remove(waiter) + with f._condition: + f._waiters.remove(waiter) DoneAndNotDoneFutures = collections.namedtuple( 'DoneAndNotDoneFutures', 'done not_done') @@ -272,7 +273,8 @@ waiter.event.wait(timeout) for f in fs: - f._waiters.remove(waiter) + with f._condition: + f._waiters.remove(waiter) done.update(waiter.finished_futures) return DoneAndNotDoneFutures(done, set(fs) - done) diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -69,6 +69,12 @@ - Issue #17481: inspect.getfullargspec() now uses inspect.signature() API. +- Issue #15304: concurrent.futures.wait() can block forever even if + Futures have completed. Patch by Glenn Langford. + +Fix warning message when `os.chdir()` fails inside + `test.support.temp_cwd()`. Patch by Chris Jerdonek. + IDLE ---- -- Repository URL: http://hg.python.org/cpython

1 0