LabVIEW
Stop a sub VI anytime
It is difficult to give specific advice without seeing the code, but your stop button will only get read once per while loop iteration and the loop cannot complete until everything in it has completed.
Can you explain what the loop is doing that prevents is from reading the stop condition at regular intervals? For example if there is a lengthy inner loop (FOR or while), the terminal would belong there. If it is waiting forever for some timeout, that needs to be broken. What does the subVI actually do? Does it really need to stop or would it be sufficient to just hide the panel?
I would guess that you have a relatively poor architecture and whatever you need the program to do could be done in much better and different ways.
Once you show us a simplified version of your code and tell us how the program should function as it is operated (step-by-step details!), I am sure we we can point you to a much better solution.
When you want to stop parallel loops, you need a way to "get inside the loop". In the "old days", there were "References" that could be passed, but they were clumsy, and did not maintain "data flow" (you had no "wire" for the reference). In LabVIEW 2016, the Asynchronous Channel Wire was introduced, and makes a very intuitive way to pass things "asynchronously" (here meaning "instantaneously") from one loop to another. Here's a simple example with a Stream Channel wire:
The Stream Channel Wire is created by a "Channel Writer", the rectangular item in the first Loop that takes the Boolean "Stop" control in and exports a "Green Pipe" -- green for Boolean, pipe as a metaphor for Asynchronous Channel Wire which passes over (not tunneling through) the "wall" of the While Loop. The second (receiving) Loop has a corresponding Channel Reader that takes the Boolean out of the Stream and sends it to the Stop Control of the second loop.
The Stream Channel Wire is like a Queue -- everything put in one end is sequentially read at the other end. If you run this code, you may be surprised at the results -- the second loop will not stop until it reads all the commands from the first Loop, namely "Go, Go, Go, ..., Stop", and as it is the slower loop, it finishes later, and shows it looped as many times as the first loop.
To get the two loops to stop at the same time, replace the Stream Channel Wire with a "Tag" Channel Wire, where the Reader only reads the most recent item (if any) that is passed in:
This behaves as you would expect -- the two loops stop at the same time (with the second loop showing fewer iterations).
I've put both VIs right here on the diagram. What if the second Loop was a sub-VI and I wanted to pass it the Channel Wire? Simple -- select the second Loop (including the Tag Channel Wire), go to the Edit Menu and choose "Create Sub-VI". Try that. You'll see that the sub-VI has a Channel Wire Reference as an input, so you can pass Channel Wires into and out of sub-VIs.
See if this works for you.
Bob Schor
@MaSta wrote:
I made a simplified VI. example1 is the one that works, but doesn't exit the sub VI immediately, depending on where the code in the FOR loop currently is. example2 is supposed to exit the subVI anytime, but doesn't do it.
This code is too simple! What does the FOR loop actually do? If it does a hard calculation that takes 10 seconds, it needs to complete. If it just does something quick every 10 seconds, use an inner loop to poll the button. And why is is a FOR loop instead of a while loop? I really doubt that waiting for 100000x 10 second iterations (11+ days!!!!) for a natural termination of the loop is reasonable in any scenario.
Here's one possibility (might need some tweaks, depending on if you want the first task a 0s or 10s, etc.):
