Messages (9) msg123433 - (view) Author: Brian Quinlan (bquinlan) * Date: 2010-12-05 18:53

multiprocessing generates fatal error "Invalid thread state for this thread" in PyThreadState_Swap

This seems to happen on RHEL 5 and Centos 5.5

Here is the minimal repro:
>>> import multiprocessing.managers
>>> mpp = multiprocessing.Pool(4)
>>> sm = multiprocessing.managers.SyncManager()
>>> sm.start()

See http://bugs.python.org/issue10517 for more details

msg132410 - (view) Author: Sandro Tosi (sandro.tosi) * Date: 2011-03-28 18:17

Hi Brian,
are you able to replicate it? I just build 3.2 and default (3.3) on a Debian unstable system and wasn't able to replicate it (with a busy loop like while date ; do ./python -c "import multiprocessing.managers ; mpp = multiprocessing.Pool(4); sm = multiprocessing.managers.SyncManager(); sm.start()" ; done )

msg132418 - (view) Author: Brian Quinlan (bquinlan) * Date: 2011-03-28 20:20

No, I wasn't able to replicate.

msg132429 - (view) Author: Sandro Tosi (sandro.tosi) * Date: 2011-03-28 21:58

can we close this issue then?

msg132634 - (view) Author: Senthil Kumaran (orsenthil) * Date: 2011-03-31 04:08

I wonder if this is CentOS and RHEL specific. Unable to reproduce this on Ubuntu.

msg132681 - (view) Author: Dave Malcolm (dmalcolm) Date: 2011-03-31 15:43

This is indeed looking like it's RHEL-specific.

I was about to close it out as a duplicate of issue 10517, but I'm wondering why Brian chose to open it as a separate bug.

msg132696 - (view) Author: Brian Quinlan (bquinlan) * Date: 2011-03-31 19:52

Filing a new bug might have been a mistake. Once the investigation in issue 10517 isolated the failure as being in a different module, I thought it best to file a new bug with a minimal repro case.

Fill free to cleanup.

msg134590 - (view) Author: Charles-François Natali (neologix) * Date: 2011-04-27 16:07

It's a duplicate of http://bugs.python.org/issue10517

msg134597 - (view) Author: Jesse Noller (jnoller) * Date: 2011-04-27 17:11

Dupe of issue10517