@AeroSoul wrote:
So i have a problem with Quit LABVIEW function in my app.
As a last resort if something goes wrong, i call this function to shut down the executable, however sometimes it still keeps a subVI running in the background.
From what i understand from function details, this function should close everything, but for some reason it doesn't, anyone knows why?
The subVI in question is called asynchroniously with a call and forget setting, in case this matters.
I have three answers, one for every paragraph in your Message.
- The "problem" is that you are using "Quit LabVIEW", which should almost never be used. Instead, you need to examine your code for "parallel loops" and make sure to stop each parallel loop.
- Both the Stop function and the Quit LabVIEW functions only affect functions that are in the "execution Tree" of the VI in which it is called, analogous to pushing the "Abort Execution" button on the Main VI's Front Panel tool bar (which, by default, is not shown if you designate the VI as a Top Level VI, for precisely the reason that a well-designed LabVIEW Program should not have a "Panic Stop", but rather an "Orderly Shutdown" procedure.
- VIs called asynchronously are running "in their own world" (environment) -- stopping their caller will not "automatically" affect them.
As I recall, you did not attach your Project to your request for help, so I don't know the basic design you are using, other than it involves some "Start Asynchronous" calls (I'm going to use the term "Clone VIs" for this sort of code). LabVIEW, as you know, supports multiple loops running in parallel -- in order to stop a LabVIEW program, all of the parallel loops need to stop. One "crude, but effective" way to do this when no Clones are involved is to have a Global Variable, perhaps labeled "Panic Stop" (a good name, because that's what it is) that is OR'ed into the Stop terminal of every loop (including FOR loops -- if you are being "gross", you need to make sure the For loops can be stopped by turning on the Conditional Stop).
A much better way is to use some form of "signaling" -- when a Stop condition is met, messages are sent to all the parallel loops to ask them to do an orderly shutdown, closing files, releasing resources, etc.
But what about Clones? I assume your Clones have inputs from the host routine that spawned them. One of those inputs could, and probably should, be used to tell the Clone to do an orderly shutdown. So when the Top Level VI decides to Quit, it sends the Shutdown "command" to all of its clones, and they all shut down (including shutting down all the sub-VIs they might have running.
This works. About a decade ago, I wrote a "monitoring" routine that spawned up to 24 clones to monitor animal behavior for a few hours. There was a top-level "Stop" button (mainly used during trial runs, otherwise it ran for a set amount of time before doing Shutdown). On Shutdown, each clone was given a "Shutdown" command (the Host could send Commands to the Clones). Initially, the Clones were "Call and Forget", but I think I ultimately made them "Call and Collect" just to be able to have the Host "wait" for all the Clones to exit before exiting itself, raising a Warning to the Operator if it took too long (I don't remember what "too long" meant, but it was probably something like 30 seconds).
Bob Schor
