i#6938 sched migrate: Use annotation for fallthrough

[derekbruening]

Replaces a comment with an official compiler annotation for a switch case fallthrough, to fix a warning under some compilers as part of running scheduler_unit_tests internally.

Issue: #6938

[edeiana]

Just noticed: are these failing tests among those that recently started to fail often for AARCH64? (I have not seen them failing personally, so just double checking this change is not the culprit, although it's unlikely as it's not failing to build)

2024-09-09T22:27:00.5769792Z 	code_api|linux.fib-conflict 
2024-09-09T22:27:00.5770504Z 	code_api|linux.fib-conflict-early 
2024-09-09T22:27:00.5771253Z 	code_api|linux.thread-reset 
2024-09-09T22:27:00.5771858Z 	code_api|api.startstop

[derekbruening]

The first ones are all ignored (there are layers of scripts and a later ones marks them ignore:

2024-09-09T22:27:00.6058886Z 	(ignore: i#2417) 	code_api|linux.fib-conflict 
2024-09-09T22:27:00.6059837Z 	(ignore: i#2417) 	code_api|linux.fib-conflict-early 
2024-09-09T22:27:00.6060781Z 	(ignore: i#2417) 	code_api|linux.thread-reset 
2024-09-09T22:27:00.6061517Z 	code_api|api.startstop 
2024-09-09T22:27:00.6062369Z 	code_api|api.static_sideline_FLAKY 

So api.startstop is the only failure: I think that is #6928.
These arm suites have been very flaky in the last few weeks with multiple failures: not sure why they seem to be worse than before.