Home → Problems →

MWServer

• Edit

Shuyang

14 Oct, 2020 07:50 PM

Hi MWorks team,

I'm trying to run a new experiment in MWorks Version 0.10 (2020.09.08), but the MWServer just crashes every time right after I click the green start button, without showing any error. My xml is attached below.
However, when I try to run the same experiment on version 0.8.dev-20171117 (5261d5c1) on another Mac, it works well without any problem.
Could you please help me see if there's any problem with my code or if there's anything I can do to fix this?

Thank you very much for your help!

1. Support Staff 1 Posted by Christopher Sta... on 15 Oct, 2020 02:14 PM

   

   Hello,

   I see that your experiment uses a custom plugin (for the LabJackU6 device). Custom plugins must be recompiled any time you move to a new version of MWorks. Do you know if the LabJackU6 plugin you're using has been compiled against MWorks 0.10? If it hasn't, then that's almost certainly the cause of the crash.

   If you have recompiled the plugin, and the 0.10 version of MWServer still crashes, then it would be helpful if you could send me the crash report. This is the info you see when you click the "Report..." button in the "MWServer quit unexpectedly" dialog.

   Cheers,
   Chris Stawarz

   • Edit

2. 2 Posted by Shuyang on 15 Oct, 2020 03:26 PM

   

   Hi Chris,

   I believe our custom plugins have been recompiled. And this seems to be experiment specific, some other experiment can run in MWorks 0.10 just fine. Here's the crash report:

   • Edit

3. Support Staff 3 Posted by Christopher Sta... on 15 Oct, 2020 06:03 PM

   

   Thanks for sharing the crash report. It shows that MWServer is crashing inside the LabJackU6 plugin, specifically in LabJackU6Device::interp1, called via LabJackU6Device::ljU6WriteLaser.

   When I compile the plugin source code against MWorks 0.10 and attempt to run your experiment, I get a crash at the same location. The issue is that the function findNearestNeighbourIndex is returning -1 (presumably to indicate "not found"). This value is then used to index two arrays, which in turn causes a memory access error and the crash.

   It looks like findNearestNeighbourIndex uses data from a calibration file (/Users/hullglick/Documents/Calibration_Table/lasergreen.txt in your experiment). Is it possible that this file is missing on the computer you're using? Obviously, it isn't present on my machine, either.

   Chris

   • Edit

4. 4 Posted by Shuyang on 15 Oct, 2020 10:04 PM

   

   Hi Chris,

   You're right! The calibration file is missing! It works now after I added that file on this Mac. Thank you so much for your help!

   Best,
   Shuyang

   • Edit

5. Christopher Stawarz closed this discussion on 16 Oct, 2020 02:26 PM.