Is there a limit to the number of mnu files that can be loaded?

[boomer]

LabVIEW 7.1 fatal error (image.cpp) occurs when trying to expand any drop-down menus, even in the main menu bar of a blank, new VI.  This crash can also occur in the LabVIEW welcome screen when no VIs are open.   For example, if I create a new blank VI, then try to show the block diagram by selecting Window-->show Block diagram, this causes a fatal error.   Cleaning and then reinstalling LabVIEW doesn't help.  Installing new video drivers and simplifying the video mode doesn't help.  The fix, that has worked, is to remove over 200 user-defined mnu files.   This behavior occurs with LabVIEW 7.1, and WinXP, on a PXI-8196 controller.  One clue was that the "Init Menus" state during LabVIEWstartup takes about 5 minutes to complete.  Is there a limit to the number of mnu files that can be loaded?

[mikeporter]

Two hundred menu files? All I can say is wow!

In terms of the problem you are having, you should have started suspecting something when it began taking more than 5 minutes to launch LV . If something hurts that's often a good sign that there is something bad going on. What do you need that many menus for anyway?

Mike...

[boomer]

Yeah...  well, that part of it (the 200-plus mnu's), is not my doing.  That is what you get when you use our (needless to say, large) engineering group's installation script, which includes reusable code.  At least, that is where I assume the mnu's came from, as I didn't put this system together initially.  I guess somebody had the idea to make ALL the reusable code available through the palette.

...-..-.

[Craig_D]

Boomer,

I am not aware of an explicit limit to the number of mnu files that can be loaded and we do not have anything documented on cpp errors that result from this that I could find.  A few questions:

1. What is the exact line number of the cpp error?--A log file of the stack trace would be helpful as well to see if something similar has been seen before.
2. Have you seen a "magic number" of mnu files past which these errors occur?
3. Are the errors reproduceable?
4. Since they may be the result of an installation script, does this happen on multiple systems which have these mnu files on them?

Answers to these questions would be very helpful in finding out what went wrong.

Please let me know if you have any more questions.

Regards,

Craig D
Applications Engineer
National Instruments

[boomer]

Craig,

It was image.cpp line 2770.  I am told (by another person in the company that has seen this) that it has been seen on several machines, always PXI controllers, but not on all PXI controllers.  I am told that it is not always image.cpp that reports the error, but for me it was always the same line of code identified by the error message.  I did not have a chance to do any experimenting to find a particular threshold number of mnu files.  I may be able to find error logs.  I currently sent the controller to our IT department, doing a windows XP repair, just in case there was something corrupted in the system itself.

The error was reproducible.  This installation at one time must have included real-time / FPGA, because on the opening screen I ge the drop-down ring for selecting a target.  If I click on this drop down ring, I get the crash on line 2770, every time.  If, instead I open an empty new VI, I am able to switch to the block diagram by selecting the window directly with the mouse, but if I try to pull down on the "window" menu to do "show block diagram", I get the crash on line 2770, every time.  Because the user-defined mnu files are now deleted, I would have to create them again, or find a way to reinstall them, to reproduce the error at this point.

Vinny

[Craig_D]

Vinny,

Thanks for the clarification.  I could not find another case which had the same error and line number as a result of mnu files.

I would suggest creating an error report if you see this error (or any cpp error) again and include the generated error log and any files you could pass along to us that would help us re-create this on our end, find out what is going wrong, and ultimately fix it.

Thanks for letting us know about this!

Please let me know if you have any more questions.

Regards,

Craig D

Applications Engineer

National Instruments