Back to the main page.
Bug 1396 - One of the test scripts crashes matlab
| Status | CLOSED INVALID |
| Reported | 2012-03-28 11:39:00 +0200 |
| Modified | 2019-08-10 11:56:24 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | All |
| Operating System: | Linux |
| Importance: | P4 major |
| Assigned to: | |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Boris Reuderink - 2012-03-28 11:39:29 +0200
On mentat301, where the automatic tests are run, matlab crash dumps appear. The stacktrace is not very informative, see attachment.
Boris Reuderink - 2012-03-28 11:40:28 +0200
Created attachment 242 Crash dump with stack trace.
Robert Oostenveld - 2012-03-28 12:39:10 +0200
indeed not very informative. We have had cases like this before, where the only solution was to implement a workaround for a - in itself syntactically correct - piece of MATLAB code to prevent the crash from happening. So I suggest to try and hunt down the crash to a particular piece of code and see whether the crash can be reproduced for that piece in isolation.
Jörn M. Horschig - 2012-03-28 12:59:59 +0200
does it always happen? I encounter such problems usually when there is a memory problem, mostly during a call to a mex-file. An out of memory error should be caught by Matlab, right? So maybe there is something else wrong (I would still bet on a memory issue)
Boris Reuderink - 2012-03-30 13:14:03 +0200
I think it happens every time all the tests are run, so probably it is one of the tests that triggers this problem.
Boris Reuderink - 2012-10-25 15:10:10 +0200
Could this perhaps be caused by a resource limitation imposed by Torque?
Boris Reuderink - 2012-11-02 13:29:53 +0100
I am no longer working on FieldTrip. Hence, I donate all my bugs to the joint development user.
Jan-Mathijs Schoffelen - 2012-11-07 13:45:58 +0100
This bug report is too unspecific for us to take over from Boris. In general, the bugs on his list that were reassigned to the unassigned-list are not sufficiently documented for us to follow up on. This is a pity. The only thing we can do is close them as invalid, unless Boris can infuse them with some of his wisdom, or hopefully even fix them (even if he is not 'officially' on the team anymore).