As Gerd notes, the current "model" for LabVIEW RT (including RT with FPGA) is for the RT Target to run its code automatically when it starts up, and to keep running its code (essentially forever). At the same time, you have Host code that you run from the PC as a "normal" program, one that you Start by double-clicking its Icon (if an executable) or hitting the Run Arrow (if in LabVIEW Development Mode).
One of the first things the Host has to do is to establish connection with the Remote RT program through some communication method (which you presumably have already worked out). Often RT systems work by the Host sending "commands" to the RT and collecting "Results". At some point, the Host decides to Stop (you push the Stop button) -- the Host needs to first tell the RT "I'm stopping, reset yourself", and then it can go through an orderly shutdown process, closing files, destroying Queues, stopping Asynchronous Call VIs, etc.
The RT Target, when it gets the notification that the Host is stopping, needs to put itself back in a state where it can again respond to a new "Initial Host Command". Sometimes the easiest way to do this is to simply execute the commands on the Target to cause it to reboot, which will force the Startup routine to run again. Otherwise, you can programmatically "close down" the running code and go back to whatever State the Target was in when it first began executing (if your target is very slow to reboot, then this latter option might be preferable, but the reboot is surely the simplest ...).
Bob Schor
