Created on 2011-11-08 22:11 by stacktic, last changed 2011-11-09 23:47 by pitrou. This issue is now closed.
| Files | ||||
|---|---|---|---|---|
| File name | Uploaded | Description | Edit | |
| multiprocessing_queues.patch | stacktic, 2011-11-08 22:11 | Patch for Lib/multiprocessing/queues.py against HEAD | review | |
| Messages (3) | |||
|---|---|---|---|
| msg147319 - (view) | Author: Arnaud Ysmal (stacktic) * | Date: 2011-11-08 22:11 | |
Using get(timeout=1) on a multiprocessing.Queue sometimes leads to a blocking get. It seems that there is no check whether the timeout has expired after acquiring the lock but before the time.time(), which can cause a call to poll() with a negative timeout. (patch attached) |
|||
| msg147385 - (view) | Author: Roundup Robot (python-dev) | Date: 2011-11-09 23:45 | |
New changeset e7b6dca28a2f by Antoine Pitrou in branch '2.7': Issue #13373: multiprocessing.Queue.get() could sometimes block indefinitely http://hg.python.org/cpython/rev/e7b6dca28a2f New changeset 9328080a19c0 by Antoine Pitrou in branch '3.2': Issue #13373: multiprocessing.Queue.get() could sometimes block indefinitely http://hg.python.org/cpython/rev/9328080a19c0 New changeset f34b4e250b44 by Antoine Pitrou in branch 'default': Issue #13373: multiprocessing.Queue.get() could sometimes block indefinitely http://hg.python.org/cpython/rev/f34b4e250b44 |
|||
| msg147386 - (view) | Author: Antoine Pitrou (pitrou) *  |
Date: 2011-11-09 23:47 | |
Patch committed, thank you for contributing! |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2011-11-09 23:47:57 | pitrou | set | status: open -> closed resolution: fixed messages: + msg147386 stage: patch review -> resolved |
| 2011-11-09 23:45:32 | python-dev | set | nosy:
+ python-dev messages: + msg147385 |
| 2011-11-08 22:25:17 | pitrou | set | nosy:
+ pitrou |
| 2011-11-08 22:23:03 | pitrou | set | nosy:
+ neologix stage: patch review versions: + Python 2.7, Python 3.2, Python 3.3 |
| 2011-11-08 22:11:27 | stacktic | create | |