CVI Crashes under Win2000

[Grant_LECO]

I've had CVI 6.0 crash on me 8 or 9 times already today. It's always the same error message on the screen.

PROCESS_HAS_LOCKED_MEMORY_PAGES. Dumping Physical memory.

If anything goes wrong in my program that requires me to press break or continue then I'm kicked out of Win2000 completely and I get this low level message on the screen.

One issue that concerns me is that CVI can't be ended by Win2000 while it's in any kind of debugging mode.

"Program Debugger is running. Can't close window. Close debugger first."

That's a good idea, to close the debugger that is, but when I press stop I get kicked out of win2000 completely and I get

PROCESS_HAS_LOCKED_PAGES. Dumping physical Memory.

Why doens't CVI end when win2000
tells it to end? Is the locked pages a conflict with win2000? Are both programs striving for control? Why do I get kicked out so fast and far? I feel like I don't stand a chance. I can't end the program by hitting ctrl-alt-delete and I sure can't hit the little red stop sign.

I've used CVI4.0 and 5.0 and 5.5 and now 6.0 and haven't seen anything like this.

Any hints for me?

Grant

[Jason F.]

Hi Grant,

I actually replied to your email that you sent in to us on this same issue but I am not sure that you received it. Here is the content of the email.

I haven't been able to find any specific information on known issues that will cause this type of behavior, but possibly with some more information from you I can help you determine what it is that is causing this to happen, if its normal, and if its avoidable.

You say this happens while debugging an executable, have you narrowed it down to any certain lines of code that will cause the error to happen (by using breakpoints)? Have you been able to see this behavior on any other computers running similar setups (win2k + CVI 6.0)? Are you running any other applications at the same time that you are
debugging your executable in CVI? It is very likely that if a fatal error occurs and CVI can't save the application from its own undoing that severe system failures will occur, such as Win2K blue screening and rebooting.

If you could send us the simplest version of some code (along with all project files) that exhibits this behavior then we will be able to try and reproduce the problem in-house to see if there is something wrong with CVI or if it is something else causing this to happen and what we can do about it. Without a reproduceable example there really isn't much we can do but speculate, especially if we don't even know at what point in your code this is occuring.

So, if you could would you go ahead and try placing some strategic breakpoints in your code to see what lines of code are causing this to happen in your application and if you cannot discern anything please provide us the simple reproduceable example project that I talked about above (either reply to the email
that I sent back to you or goto www.ni.com/ask and fill out the email support request form again). Thanks!

Jason F.
Applications Engineer
National Instruments
www.ni.com/ask

[Grant_LECO]

Thanks Jason. I reinstalled CVI again and didn't ask it to change any paths. I had to shuffle stuff around, like my shortcuts and stuff, but now things appear to be stable. I haven't had time to really try and crash it but so far so good. I'll stay in touch.

Grant

[MagiQUARK]

I am having the same problem
I use CVI 5.0.1 on win2K , MIO16E4 , NIDAQ v 6.9.2 .
tried re-installing with no difference .
Quick description : Two progs running DDE Comms interacting , error occurs only when one is running under CVI and is instructed to quit the error occurs .
It doesn't occur on all machines but on the ones it does it is consistent.

Once We got DRIVER_IRQL_NOT_LESS_OR_EQUAL_TO_ZERO


Anything else we can try ?