LabVIEW cought fatal signal 8.2 - Received SIGSEV

[Fusselpelusa]

Hi,
I am getting the error written in the subject line.
Following lines come after:

Reason: address not mapped to object
Attempt to reference address: 0xbae22b0

Also huge core.dump files are created and the normal procedure, Labview is intended to do is stopped until the 100MB are written. Strangly, labview does not crash but works perfetrly the next say 5 to 20 minutes until another core.dump gets created. This is really bizzare as we cannot find any related issue that might cause such a partial crash. i googled and found this site but I am not sure if this is true for Labview 8.2 as well. We are running it with mandriva 2006, the only distro that is apparently fully OK with labview.
To give a little bit more background, we tracked down that the error comes from a case structure that calls a system call which is a simple script calling 2 c-programs. Everything is running and no seg fautl is happening, and we get the results we want, but somehow these annoying seg-faults occur and prevent the data for this second to be written and actually lost.

Any help is appreciated.

Thanks
Thomas

[eek]

Hi Thomas,

This is curious behavior. I'm glad you found the KnowledgeBase article regarding this error message, as it may at least shed some light onto some troubleshooting steps we can take.

While the article is written in regards to LabVIEW 7.1, it's statement at the end of the document regarding supported processors is still current in that LabVIEW is only supported on 32-bit processors. Thus, I would ask what kind of processor are you running on? This will be regardless of what version of LabVIEW you are experiencing this error with.

I look forward to finding a solution to this issue with you.

Cheers,
Emilie K. | Applications Engineer | National Instruments 

[Fusselpelusa]

Thanks for the reply,
I am using a 32 bit machine (processor: i686 Intel(R) Celeron(R) D CPU 3.46GHz)  with Kernel 2.6.12-31mdk.
Thomas

[eek]

Hmm. Perhaps the problem shouldn't be entirely attributed to LabVIEW... I imagine that if you are writing data then your application is also acquiring data as well. Meaning, are you using any NI driver calls in your application? If so, you might want to peruse this thread. I am by no means a Linux expert but perhaps you might catch some useful tips on how to coach our drivers to work on Linux stations. Let's see where this gets us.

Cheers,
Emilie K

[Fusselpelusa]

Hm we are using this program sonce half a year now and everythig was fine,with all drivers installed and used (VISA, NIKAL and 488.2). We are acquiring data as well as writing data since then and had never problems with core dumps. However just after adding the code with the system call, these core dumps plus error messages start.
I wil look at the proposed thread and see if i can figure out whats going on.
Thanks
Thomas