possible bug in NSRunloopScheduler,

[fschwiet]

Describe the bug

Regarding the implementation of NSRunloopScheduler https://github.com/reactiveui/ReactiveUI/blob/master/src/ReactiveUI/Platforms/apple-common/NSRunloopScheduler.cs, NSRunLoopScheduler uses NSTimer.Invalidate in the disposable returned for Schedule() calls: https://github.com/reactiveui/ReactiveUI/blob/master/src/ReactiveUI/Platforms/apple-common/NSRunloopScheduler.cs#L68.

The Apple developer documentation https://developer.apple.com/documentation/foundation/nstimer/1415405-invalidate?language=objc states:

You must send this message from the thread on which the timer was installed. If you send this message from another thread, the input source associated with the timer may not be removed from its run loop, which could prevent the thread from exiting properly.

The Xamarin documentation for NSTimer.Invalidate() does not mention this requirement, but I suspect the documentation in incomplete: https://developer.xamarin.com/api/type/MonoMac.Foundation.NSTimer/.

I am not sure there is a problem here, TBH. Just noting the concern.

[open-collective-bot]

Hey @fschwiet 👋,

Thank you for opening an issue. We will get back to you as soon as we can. Also, check out our Open Collective and consider backing us.

https://opencollective.com/reactiveui

PS.: We offer priority support for all backers. Don't forget to add priority label when you start backing us 😄

ReactiveUI - Open Collective

An advanced, composable, functional reactive model-view-viewmodel framework for all .NET platforms!

[glennawatson]

hi @fschwiet thanks for giving a detail report. Seems nstimer.invalidate calls directly into the native code (checked with ilspy) so you're probably right.

Realistically i won't have a chance to look into this until later next week though so if you feel like providing a PR in the mean time I will look over and review it. Thanks

[fschwiet]

This issue applies to the RxApp.MainThread for iOS and Mac

[fschwiet]

Thanks @glennawatson. I have an alternative DispatchQueue IScheduler wrapper implementation I posted in #1531 that I believe would address the issue, assuming I didn't botch that implementation.

[glennawatson]

Want to put that into a PR and we'll roll it into a test version for a little bit?

[fschwiet]

Yes but I am so busy too hehe. I'll see what I can do.

[fschwiet]

Also I don't understand why WaitForDispatcherScheduler() was used to wrap the creation of NSRunLookScheduler. That seems unnecessary, but not a bug.

[fschwiet]

Just noting that in the discussion in #1531 its mentioned that someone is working on a PR.

[RLittlesII]

@fschwiet I reproduced your solution. I am still looking into this, because of the nature of this issue, I am trying to make sure that we fully understand the impact.