Open Quuxplusone opened 1 week ago
Please submit an LWG issue; the redundancy should be removed, but there's a lot of LWG discretion involved on what exactly the result should look like.
I think the decision on explicit
default constructors is already made in LWG2510.
https://eel.is/c++draft/thread#lock.general-2 says:
But this isn't the actual correct definition of
defer_lock_t
et al.; those are normatively defined further down in https://eel.is/c++draft/thread#mutex.syn as havingexplicit
default constructors.IMHO we should figure out a way to strike the first (non-normative?) definitions entirely. At worst, the first definitions should be updated to match the correct normative definitions, so that nobody gets confused.