Profile performance and memory window crashing

[LDHGE]

I pulled up the Profile Performance and Memory Window to check VI timings last night and it crashed on selecting start (LabView totally hung). 

I have done some investigating and found a number of things:

a) It crashes/hangs on selecting Start, regardless of whether the VIs are running or not and without any of the checkboxes selected - so looks like the problem is in generating the list of VIs & SubVis (but VI Heirarchy works okay)

b) I get no error messages - on restart LabView offers me the chance to recover files, but doesn't report any previous errors or give me a fault log.

c) The original crash is caused by something in one (or more) specific VI (I'm guessing some form of corruption) as it doesn't occur when I remove a large branch of VIs or just create a new VI.  (but see point d).

d) When it crashes there are two possible behaviours, either LabView disappears (all VIs, windows etc close) or LabView hangs.  If LabView hangs (and I kill it), then after restarting Labview the profile window will crash on any and all VIs (including a blank one) until I reboot the whole machine.

I'm trying to narrow down which VI causes the problem, but this is a very slow process as I end up needing to reboot frequently. 

While working on that, I'm wondering what to do once I identify the VI with the problem.  Anyone seen this problem before or have any idea what types of things could cause the crash? 

[Ben]

If the crash only happens with the Performance Profile screen, I believe in about version LV 8.5 there where bugs with add-on tool-kits (DB tool Kit ?) that would result in LV crashing.

Ben

[LDHGE]

LabView 2009 SP1 plus Vision module.  No additional add-ins.

I saw one old (LabView 8x) post which suggested repairing LabView, but as it seems VI specific that doesn't seem the right approach.

[Joe D.]

Hi LDHGE,

Would you mind attaching the crash log from LabVIEw so we can take a look at why it is crashing?

You can find a good KnowledgeBase Here:

Where is the Error Log File Saved After a LabVIEW Crash?

[LDHGE]

Hi Joe,

98% of the time there was no error log file, however once I started to narrow down which VIs were causing the problem I did eventually get the two attached.

I've managed to get it running by removing some existing subVI calls and recreating them - however, the window does come up with red text (see the other attachment) which seems odd.  But at least it works.

However, I'd be interested in anything you find out about what the actual cause was.

[Joe D.]

Hi LDHGE,

unfortunity the log files did not tell us much. What is the smallest possible code that you have been able to reproduce it with? Can you send that to me?

Like you I would like to get to the bottom of this as well.

[LDHGE]

Hi Joe,

It looks like changing and saving a VI with the problem sends it into hiding but doesn't necessarily actually clear it (as I've had it reoccur on a "fixed" vi) so I haven't managed to produce a cut-down copy for you which shows the issue, and it remains proprietary code I can't post on a public forum. Is there a way to use the Service Request Manager to send the file directly to you?

One other thing I have noticed - if I load a problem VI then create a blank VI and run the Performance profile, it crashes, but if I create the blank VI first, then load the problem VI it's fine which corroborates the idea of a memory issue and gives me another work-around.

[Joe D.]

LDHGE,

If you can't post the code your best option is to call in (1-866-275-6964) and open a new SR and just reference this forum. That way can con keep your code confidential and make sure you get taken care of.