Message11192
This has to do with the fix for #2500. I compared proposed change in https://github.com/jythontools/jython/pull/40/files with
https://hg.python.org/jython/rev/eb676b781ab2
@darjus: If I remove your additional changes regarding wait-option, i.e. apply PR 40 as is, test_socket_jy goes well again. (Which is why I didn't notice this when I some weeks ago asserted that PR 40 doesn't break regrtests)
Maybe you need to adjust the test somehow?
Unfortunately I don't have much clue what this stuff is all about, just observing changes etc.
@James Don't think it's related; the hanging is annoying though. It's probably more related with hanging of test_logging etc #2536, #2537, maybe even with #2519. There also is usually socket and netty in the game. Probably also gc. I guess socket.py has some __del__ finalizers that clean up netty and depending on when gc kicks in, some unfortunate tests can yield deadlock. |
|
Date |
User |
Action |
Args |
2017-03-06 21:22:54 | stefan.richthofer | set | messageid: <1488835374.47.0.727608415103.issue2564@psf.upfronthosting.co.za> |
2017-03-06 21:22:54 | stefan.richthofer | set | recipients:
+ stefan.richthofer, darjus, jamesmudd |
2017-03-06 21:22:54 | stefan.richthofer | link | issue2564 messages |
2017-03-06 21:22:53 | stefan.richthofer | create | |
|