[clang] [Clang] Reject overflow behavior types on _Atomic (PR #222179)

[Eli Friedman via cfe-commits]

https://github.com/efriedma-quic commented:

You could do the overflow computation on the result of atomicrmw, which would be pretty cheap.  I can see why you wouldn't want to do that, though; the trap wouldn't trigger until after the result is visible to other threads, which might be unintuitive.

https://github.com/llvm/llvm-project/pull/222179