Fatal Internal Error: "memory.cpp", line 638

[Xu]

Hi,

Attatched are three test VIs. They are almost the same except the content of "FP IO Point Constant" arrays. They were tested in LabVIEW 7.1 development environment on Windows XP.

noerror.vi runs normally;

alwayserror.vi generates an internal error:
'Fatal Internal Error: "memory.cpp", line 638';

seconderror.vi runs normally the first time after loaded from disk. But it generates the same error in second run.

Does anybody know what could have happened? Is it a bug? Thanks,

Xu

[Ben]

I beleive your suspicions are correct!

The differences between the examples you posted that crask and do not crash looks to be dependent on what type of FP tag is passed to the typecast function in what order!

If you only do emties, Fine.

If you only do non-empties, Fine.

Do an empty followed by a non-empty ---> TROUBLE!

The above explains why the various examples you posted work/don't etc.

Please see the attached zip to explore more.

If you run my version of "allwayserror.vi" with the VI server stuff shutdown (see boolean) it demonstrates that as long as the typcast always gets the same type (i.e. empty vs non-empty) the typecast is OK.

If you turn on the VI server part of the code, you will see that if the typecast is in DIFFERENT VI's it still crashes.

I vote BUG!

Work-around: try using the "Flatten to string".

Ben

[Ben]

The attachments I forgot!

Ben

[Ben]

Reported as SR#= 635103

Ben

[Xu]

Hi Ben,

Thanks for your example showed that even the typecast is in different VI's it still crashes.

What does it mean in your last post "Reported as SR#= 635103"?

Xu

[Ben]

"What does it mean in your last post "Reported as SR#= 635103"?"

That was a note to myslef and the rest of the world that I have reported this issue to NI.

The Service Request Number was 635103 (SR# = 635103).

Thanks for sharing your findings!

Ben

[Xu]

Got it. thanks,
Xu

[Ben]

This issue has been confirmed and a Corrective Action Report (CAR) was generated.

Car# = 3GKCEKSN

Ben