Labview (8.5) Crash by adding item to typedef

[DonRoth]

Hi Folks -

I have a VI that is fairly large and complicated.  A strict typedef is used to add cases in a producer-consumer event-driven state machine.  I have 87 items in the strict typedef now.  When I go to add the 88th item, no matter what I name it, a broken run arrow appears on the toolbar and when I click on that it, no errors show up in the error window, but a few seconds later, LabVIEW crashes.  I have tried starting over several times with backed up versions of the code to make sure I don't have a corrupted VI but always get the same result.  When I remove the 88th item, the run arrow is no longer broken and I am back to the VI working fine.  I have also rebooted the computer but get the same results.  I wonder if I am running into some sort of memory issue.

This VI is not something I can simplify or post.  I would have to make an LLB or ZIP out of it and send to an engineer to look at.  I wanted to know if anyone else has run into something like this before - and can maybe point me to something not obvious that I need to look at.

Sincerely,

Don

[Ben]

Try using the old style "Flatten to string" to get the type descriptor for that cluster. Check the size of the type descriptor to see how large it is.

Ben

[DonRoth]

I should have said I am using a strict typedef enum.  I am attaching it here.

Don

[Ben]

No crash when adding item number 88 using LV 2009 SP1.

Ben

[DonRoth]

Right, but it is not being added within the context of my application.  I have 64-LabVIEW 2009 installed on another computer w/ 64-bit Win7 with my application partially working right now.  I don't have a 32-bit LabVIEW 2009 (w/ ASP + Vision) installation right now.

I do appreciate your trying.

Thanks,

Don   

[DonRoth]

I found the problem.  It is an odd-ball type of issue that will be hard to explain here.  Basically, I had created a reference to the typedef enum control and it has to be updated when I added the new item.  I don't know why I did not get an error.  There was a broken wire from the reference on the block diagram after updating, but it was hidden from the default view that I was looking at.  Anyhow, the problem, which could have been major, is fixed.

Thanks,

Don