Thank you (again) for all the work that you are putting into making this SDK so complete. I think that I found an issue where the partial shutdown of the stack can cause a problem and hope that this fix is appropriate. If it is, I am more than happy to add testing. I just didn't want to spam you with additional changes unless this fix is on the right track.

PR #38046: Size comparison from b2a86bd to 01a8990

Thank you (again) for all the work that you are putting into making this SDK so complete. I think that I found an issue where the partial shutdown of the stack can cause a problem and hope that this fix is appropriate. If it is, I am more than happy to add testing. I just didn't want to spam you with additional changes unless this fix is on the right track.

Ideally we would have some description of what fails before the PR and succeeds after. This can serve both as a description of the issue and the test that proves that the issue was fixed.

Thank you (again) for all the work that you are putting into making this SDK so complete. I think that I found an issue where the partial shutdown of the stack can cause a problem and hope that this fix is appropriate. If it is, I am more than happy to add testing. I just didn't want to spam you with additional changes unless this fix is on the right track.

Ideally we would have some description of what fails before the PR and succeeds after. This can serve both as a description of the issue and the test that proves that the issue was fixed.

Of course!

The "problem" is re-running any command implemented as a derive class from CHIPCommand. For example, if a user modified the code of main of chip-tool in way that the user's command was executed twice (in the same thread and process), the second time that MaybeSetupStack is called, assertions will fail when (re)initializing parts of the stack because they are not properly shutdown in the MaybeShutdownStack function. I know that use case is "silly", however, I have leveraged the code written for chip-tool so that I can use it as a library for invoking commands and would like to be able to run commands repeatedly in the same thread/process.

I will investigate a way to write a unit test to demonstrate this odd behavior. If, however, you have any suggestions or guidance (as an expert!) I would very much appreciate it!

Again, thank you for all the work that you are doing on Matter and the SDK!
Will