Why is the size of my FPGA VI increasing?

[RPluth]

I am using LABview 8.5 FPGA Module.  I noticed that the size ov my FPGA VI file keeps increasing when it shouldn't.  As a test, I added one simple control to my FPGA VI file and saved it.  As expected, the size of the file increased.  I then went in and did nothing more than delete this control I just added, then saved the file.  The size of the file INCREASED again!  Why would it increase when I delete something from the file!?

Anyone else see this happen?  Is there some Labview option I need to set or un set to stop this??

Thanks,
Rick

[Wiebe@CARYA]

What amounts are we talking about? Bytes, kiB, MiB? The history might grow,
but it sould be off by default.

Does the number of slices the compiled VI takes also grow? Or just the VI
size?

I haven't noticed this myself, but I never checked. I'll look at it some
day.

Regards,

Wiebe.

[RPluth]

Thanks for your interest, Wiebe.

The history is off as far as I know. I think the slices are the same.  As amatter of fact, one experiment I did was I checked the VI Properties->Memory Usage, added a simple control, saved the VI, deleted the control, saved it, and looked at the Properties->Memory Usage again.  The numbers between the two times I checked the Memory Usage were the same EXCEPT for the "Total VI Size On Disk", which had increased.

Here's a detailed description of something I did yesterday.  My FPGA VI had grown to a 140Meg.  I know, pretty big.  I do have a lot in it because of what my application demands.  I created a new empty VI and painstakingly copied all the Labview code (eg. front panel items, block diagrams, etc) from the 140Meg VI to this new one.  I saved the new one.  Bottom line hey had the axact same "code".  However, the new VI file size was only 20Meg!

I'm beginning to believe that the Save command stored previous versions of my Labview code within the VI file, perhaps for purposes of being able to revert to older revisions???  I know I grasping at straws, but this is becoming an issue.  Any NI engineer listening that knows about this.

[TCPlomp]

Pfew, a factor 7, that's quite a lot.

What happens if you do a 'force recompile' (ctrl-shift-run)?

Ton

[RPluth]

Tom,

How would a recompile affect the size of the VI file that's on my hard disk?  It's the File.VI file that I'm concerned with, not the bit file.

Am I missing something?


Regards,
Rick

[Wiebe@CARYA]

"code".  However, the new VI file size was only 20Meg!I'm beginning to
believe that the Save command stored previous versions

20 Meg is still huge! I don't think any of my VI's ever came over 2 MiB...
And that is including normal (PC) VI's. I am currently working on a huge
FPGA project, that fills up 85% of a 3 MiB FPGA, and the VI is 1041 KiB. The
VI is saved hundreds of times...

My guess is that you don't use any sub VI's... I'm not saying that you
should (but it would be better for compile time, load time and readability),
but it might explain why nobody else has seen this. In other words: the bug
is related to the size of your diagram, or a large number of objects in a
bad combination.

Regards,

Wiebe.

[RPluth]

Wiebe,

Forget the huge size of my VI files.  It is not the problem.  As a matter of fact, what's happening is the reason my file size is so huge.

In an experioment, I added two FPGA I/O Nodes to the block diagram of an empty FPGA VI, and saved it.  The file size was 83KB.  I then added a 16 bit integer control to the front panel, and saved the file.  The size was now 99KB.  Next, I deleted the control I just added, and save the FPGA VI file.  The size was now 112KB.  As you can see, the size increased.  I repeated adding a control, saving deleting the control, and saving.  The file size was now 141 KB.

When I did the this experiment starting of with one FPGA I/I Node and a 16 bit integer control.  The size increased when I added a second control, and decreased when I deleted it, like I'd expect.

So, can anyone explain this to me?  If you perform the simple experiment I described above, do you get the same results?

Thanks,
Rick

[Wiebe@CARYA]

> If you perform the simple experiment I described above, do you get the
same results?

I'll try, but can't let you know untill over a few days.

Regards,

Wiebe.