Internal error: "memory.cpp", line 593 with PXI 8175

[G Liagre]

Hi,

I have created a program with Labview RT 7.0 which makes acquisition, storage and display of data with PXI controllers 8175.
I used Datasocket for the data display on a deported computer.

Because of the number of the Datasocket tags used, instead of sending values all the time, I'm sending value for the deported computer only if the user needs to see it.
For example : if the user needs to see data, so only Datasocket tags for data display will be active, and if the user needs to see parameters values, so only parameters datasockets tags will be active. And, if the display program is in "standby" mode, almost all datasocket aren't active.
So when the user needs to see data values, he opens the data values window of the display program, and then, the datasockets tags for data value are now active, but : sometimes at this moment, the PXI stops with an "internal error "memory.cpp" line 593"!

The only solution is to manually reboot the PXI, the programmatically reboot doesn't work in this case of error. But it is far away from the displaying computer.

The error happens regularly after a 2 weeks period of standby. So I though about a memory overflow, I checked all the arrays, and none of them is growing to the infinity.

If someone knows about this problem, I will be very happy to learn about how to get rid of it!!!

Best Regards,
G Liagre

[RobCole]

I'll try this reply again, since my last one seems to have disappeared.

Search the NI website for "memory.cpp" and you should see that quite a few other people have run into similar problems with that module.

I can't remember everything that I found yesterday when looking this up, but there are quite a few different things that people have found.

I hope that you find the answer you need.

Rob

[Kerouac]

Hello,

Here is the starting point to troubleshoot these type of error:
* LabVIEW Internal Errors (Failure in XXX.c / XXX.cpp at line YYY)
http://digital.ni.com/public.nsf/websearch/01B186B5CEB1D07786256747000F4319?OpenDocument
Could you post an example VI that demonstrates the crash, so that we can more easily help diagnose your problem.
Please include as many details as possible for us to recreate the issue.

Thanks.

Matthieu G.
National Instruments

[G Liagre]

Hi Rob, Hi Matthieu,

Thank you for your reply.
I found in the database many reply about this error, but none of them were helpfull for me.
I tryed your link to the knowledge data base, but the error described there is for dll association in LabView and interface nodes.
I didn't used any interface node in my code.

The problem is that I cannot test my program on the PXI because I don't have access to it rightnow (only by remote desktop, and with a 56k modem..)

Let me give you a part of this huge code.
I suspect this part of the code to be the one which create the problem, but it's not sure.
If you find something obviously wrong, tell me!

Best regards,
G Liagre

[Kerouac]

Hi,

I looked at your code. I did not find any VI like "Release Notifier" to clear the notifier refnum. I think that your application is too large to be able to debug it correctly. You must isolate the issue. Try to delete some parts of your application in order to be able to localize the issue.
If you can post a "light" version of your application that reproduce the issue, it will be helpful.

Thanks.

Matthieu G.

[G Liagre]

Hi Matthieu,

Thank you for your reply.

For the "release notifier", it's ok, it's used elsewhere in the code, and it permit to finish this while-loop.

I will not be able to isolate more the issue, in fact, because I don't have PXI rightnow. I can try it on my computer, but I think that it will not have the same behaviour as builded in RT for PXI.

The best way I found rightnow is a dayly automatic reboot of the PXI to avoid this error, and I will continue with it, until I have some time to return into the code of this program...

Thanks!