FS#20503 - [vpython] segfaults with intel video card

Attached to Project: Arch Linux
Opened by Nicolas Bigaouette (big_gie) - Wednesday, 18 August 2010, 00:01 GMT
Last edited by Roman Kyrylych (Romashka) - Thursday, 26 August 2010, 11:18 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To No-one
Architecture All
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:
Importing "visual" (python-vpython 5.32-1) in python creates a segfault:
$ python
Python 2.6.5 (r265:79063, Aug 17 2010, 14:03:21)
[GCC 4.5.1] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import visual
Segmentation fault

I have recompiled python and vpython with debug info. Here is a stacktrace:
http://pastebin.com/zihsxktb

After talking on the mailing list, it was suggested the problem might be related to the video driver, as the problem appears on two x86_64 machines using intel-dri 7.8.2-1 and xf86-video-intel 2.12.0-1. A third x86_64 with an nvidia card does not have the problem.

Here is the discussion:
http://sourceforge.net/mailarchive/forum.php?thread_name=AANLkTi%3Dq-%3D3o%2B%2B4aWaNTRj03PkQZNUdC6KiPTKsj8Mc%3D%40mail.gmail.com&forum_name=visualpython-users
http://sourceforge.net/mailarchive/forum.php?thread_name=AANLkTi%3DKgvYRpT6UhZCpW_SqJ3L9uy4796UcPcnsWSbn%40mail.gmail.com&forum_name=visualpython-users
http://sourceforge.net/mailarchive/forum.php?thread_name=AANLkTikFj3tRqURqHgKkMEh1acqeVYL3-Q5n0-gxLAS4%40mail.gmail.com&forum_name=visualpython-users
http://sourceforge.net/mailarchive/forum.php?thread_name=AANLkTinceESHd70ainZfiLbROBK6BMhVUM_Ne_kPazbj%40mail.gmail.com&forum_name=visualpython-users

Somebody on the forum might have the same problem:
https://bbs.archlinux.org/viewtopic.php?pid=810326
This task depends upon

Closed by  Roman Kyrylych (Romashka)
Thursday, 26 August 2010, 11:18 GMT
Reason for closing:  Upstream
Comment by Jan de Groot (JGC) - Wednesday, 18 August 2010, 06:47 GMT
Looking at your sourceforge thread, I think this is rather a bug in vpython that doesn't initialize things well than a problem with mesa or one of its drivers. Importing wx before visual works for example, so it looks like vpython skips an initialisation step that is done by wx, but which doesn't trigger when the nvidia driver is used.
Comment by Nicolas Bigaouette (big_gie) - Wednesday, 18 August 2010, 13:40 GMT
Yes you are right. I would definitely say it's a vpython bug (how come it's not initialized??)
This bug can be close.

Loading...