Miss Dada  🏳️‍⚧️

having read Ralf Jung's post, it occurs to me: but wasn't the MutexGuard bug caused by the compiler relying on struct members being Send %

% (which is incorrect) instead of them being Sync when inferring impl Sync for MutexGuard?

No, the compiler relies (only) on stuff being Sync to impl Sync. There is additionally an impl of &T: Send for T:Sync.

"if you follow through", this is a combination of features interacting together

MutexGuard contains an &Mutex. Mutex itself is Sync+Send when T is Send, due to an impl. So is &Mutex, then. So is MutexGuard, then.

rustc only inferred the second two steps. The rest was explicitly provided to it; we *told* the compiler "T: Send => Mutex<T>: Sync+Send"

relevant line: unsafe impl<T: ?Sized + Send> Sync for Mutex<T> { }