Standalone Executable makes FPGA disappear

[durr]

Hey everyone!

I repost this in this subforum in the hope that more people see it here. I did a project using a myRio-1900 which works fine when deployed through Labview. However, the client doesn't have all the necessary toolboxes, so I thought it would be a good idea to build the executable and deploy it through NI MAX instead.

When having a fresh myRio, the first step was to load the FPGA bitfile through NI MAX (as described here: http://digital.ni.com/public.nsf/allkb/BAAA6D86CDDD583C8625729E00572C8B). This step works fine, and if the project is afterwards deployed using Labview, everything works fine.

The next step is to install the executable using NI MAX. I followed the steps described here: http://zone.ni.com/reference/en-XX/help/370622K-01/lvrthowto/rt_building_rt_app/ under the section "Deploying from outside of the Labview development environment". The application installs and runs correctly, but for some reason the FPGA seems to have vanished. It is no longer visible in NI MAX, and trying to access it (e.g. using the Encoder express VI to read the value) results in an error.

Installing the default accelerometer example reverts everything back to normal. I am quite in over my head with this, and not sure whether I need to change something in the build settings of the executable, or if it is a more fundamental problem. Any help would be much appreciated!

Thanks and cheers, Christoph

[Michael_Strain]

This sounds like a known issue that should be fixed in the next software release.  If you talk to NI Support, you can reference CAR 645349.

Basically the problem is that your RT APP that runs on boot is using a good amount of memory (and the myRIO only has 256 MB of RAM).  When MAX attempts to discover devices, the operation to discover FPGA Devices is temporarily attempting to allocate more memory than it can get and that error isn't getting propagated well.

For the next release, we've optimized the way we discover devices on that type of target so it requires significantly less available memory.

A workaround for now might be to programmatically use NI System Configuration to discover FPGA devices when your app starts up.

[durr]

Thanks for the reply! So does this also mean the FGPA is not working properly during normal standalone operation (because that is what is happening)? Or does it only concern the visibility in NI MAX?

[Michael_Strain]

For your case it should only concern the visibility in NI MAX.

If communication with the FPGA isn't working properly you will get an error when trying to use it.  As long are you are checking error terminals and you don't see any errors then you should be good.

[durr]

Yes sadly I also get an error during standalone operation when trying to use the FPGA (so nothing to do with NI MAX), so there seems to be something else that is amiss.

Are there maybe any additional options I would have to check during the Build of the executable when working with FPGAs ?

[Michael_Strain]

There aren't any additional options I'm familiar with.  What error are you getting?