Labview 6.0.2 runtime error causing Labview crash

[MichaelS]

When reshaping a 1d array in a for loop with following autoindexing to 2d array, LabView 6.02 crashes when reshaping was done to a NaN size (the NaN is produced by threshold array output inside the loop). Message: datasupp.cpp line 754, and additional message "instruction at 0x69514114 ref memory at 0x00000525 memory cannot be read.

NI folks: let me know when the corrected version will be available.

[Mark A. Hanning-Lee]

MichaelS wrote:
>
> When reshaping a 1d array in a for loop with following autoindexing to
> 2d array, LabView 6.02 crashes when reshaping was done to a NaN size
> (the NaN is produced by threshold array output inside the loop).
> Message: datasupp.cpp line 754, and additional message "instruction at
> 0x69514114 ref memory at 0x00000525 memory cannot be read.
>
> NI folks: let me know when the corrected version will be available.

Usually you don't need a new version.

If you get a failure with a line number, start by going to
www.ni.com/failure . That gives the cause of a lot of failures.

If you want to take it further, submit a small code sample via
www.ni.com/ask . Only then will it get properly logged, and the engineer
can see what is happening.

Ma
rk

[MichaelS]

1. LabView crashes can *only* be fixed by installing an upgrade that hopefully addresses the problem.

2. www.ni.com/failure does not have a description of the error I encountered.

3. I am not going to waste my time "logging" anything. I am not paid for LabView code debugging. If NI engineers care of their own product and reputation - they do that. I already went an extra mile describing the problem in fare details for them.

Thank you for your input.

Michael

[Mark A. Hanning-Lee]

MichaelS wrote:
>
> 1. LabView crashes can *only* be fixed by installing an upgrade that
> hopefully addresses the problem.

Incorrect. www.ni.com/failure lists many cases where a crash can be
fixed by changing your video driver or other system components. That's
why I pointed you to that page.

> 2. www.ni.com/failure does not have a description of the error I
> encountered.

Sorry to hear that.

> 3. I am not going to waste my time "logging" anything. I am not paid
> for LabView code debugging. If NI engineers care of their own product
> and reputation - they do that. I already went an extra mile describing
> the problem in fare details for them.

As a developer, I know that seeing an actual sample of code is worth a
lot of description.


If they can run it and reproduce the error, we all stand a better chance
of having it fixed in the next upgrade.

Best wishes, Mark

[Ben]

On a positive note:

The product ownership you desire has been demonstated by Aaron Gelfand(sp?) who supports FP in all its flavors.

I bet if the anomaly you encountered involved FP, he would have been right on it.

His dedication to FP is a big plus in my book.

Unfortunately, his performance is not the norm these days. It would be nice if every aspect of NI's hardware and software had a similar champion like Aaron.

But now I am dreaming,

Ben

[Greg_McKaskle2]

> When reshaping a 1d array in a for loop with following autoindexing to
> 2d array, LabView 6.02 crashes when reshaping was done to a NaN size
> (the NaN is produced by threshold array output inside the loop).
> Message: datasupp.cpp line 754, and additional message "instruction at
> 0x69514114 ref memory at 0x00000525 memory cannot be read.
>
> NI folks: let me know when the corrected version will be available.

Nan cast to an integer results in MaxInt or 2 billion and change. Assuming
that the other size is nonzero, this will take a bit of memory.

This definitely shouldn't crash, but the input to the loop probably
needs to be filtered to whatever value you really want.

Greg McKaskle