What is memory.cpp at line 1038?

[fgperry]

We have been running the same test in LabVIEW for a few years. Now all of a sudden, we are starting to get a memory.cpp error at line 1038. The test normally runs days at a time (~2 weeks), but we are getting this error after about 18-24 hours. Any ideas of what could cause this? Could this be caused by any Windows XP hotfixes? We are using LabVIEW 7.1 on a windows xp machine.

[Enrique]

fgperry;

According to the Internal Error Support Page, it seems that "it is likely that the problem you experienced has not been reported" to NI yet.

Check the following page: it includes additional information on that type of errors:

LabVIEW Internal Errors (Failure in XXX.c / XXX.cpp at line YYY)

Regards;
Enrique

[fgperry]

I've already checked out that particular site for info. And we do open and close files in our tests. And we do have quite a few loops. But it has worked with no issues for a few years and much longer time spans (and also more open/close files and loops). So after the error occurred on our only two machines (with exact set up) that run LabVIEW 7.1 (the rest are still on 6), we rolled one back to a date before all of the hotfixes occured and we have had no problems since. The second machine had an issue with rolling back so IT came and reinstalled XP. The problem is that they install an updated version and we cannot install an older one. It has Service Pack 3 in it and I think there are some hotfixes that came along with it that are the source of the problem. But I can't gaurantee it. Hopefully we'll figure it out soon.

[fgperry]

So Windows does not have an SP3 as I found out. But the hotfixes were listed in the registry under \HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Updates\Windows XP\SP3 which made me think it was a service pack 3 (which does not exist as I've learned). So anyone know if any of the hotfixes listed in this directory of the registy cause any problems?

[fgperry]

Anyone think that this could cause any issues? This is the only hotfix I could find that would have any effect on malloc, but I don't see that this will directly affect LabVIEW.

KB873333 : Vulnerability in OLE and COM Could Allow Remote Code Execution
An application that implements the IMallocSpy debugging interface may experience heap corruption after you install security update 873333

[Ankita]

Hi,

I would suggest that you email NI support team and send the error log file as an attachment while you are troubleshooting.

Ankita

[cbauer]

fgperry,

Did you ever get an answer from NI regaurding this issue?  We have been dealing with this as well and have had not luck.

Thank you,
Chris Bauer

[fgperry]

Unfortunately, we have not figured out the problem.  We upgraded to the latest LabVIEW version as NI suggested and it still fails.  Honestly, we got tired of trying to troubleshoot it and just began to live with it.  If you get anywhere with it, please update this post so that I will get an email showing it.

[fgperry]

Oh, one more thing.  We first saw this in Jan of 2005.  At that point, we talked with NI about the issue and they told us we need to mass compile everything to the latest version of LabVIEW (7.1 at the time I think).  We have it in 7.1 now and still have the problem.  We have not tried 8.2 yet. 

[cbauer]

Unfortunately "living with it" is not an option for us.  This is not an internal application, but a sellable product.  We can not have it crashing on our customers all of the time.  I have tried contacting NI and am awaiting a reply.  When I hear anything, I will post here so that you know what they said.  Also if we are able to discover the root cause, I will post as well.  Thank you for your background.