[cpp-threads] Yet another visibility question

[Nelson, Clark]

> But Ben's objection still stands.  So I think we want 
> 
> "All modifications to a particular atomic memory location or lock L
> occur in some 
> particular total order, referred to as the modification order for L."

Another concern I had about your proposal was that you seemed to be
saying that even non-atomic operations on a specific memory location had
to have some total ordering. Now you seem to be backing away from that.

But perhaps I've been reading too much into one formulation or the
other, so I'll just ask: do you mean to require a total order of
non-atomic side effects to a given memory location? (Now that I think
about it some more, I'm guessing not; wouldn't any case in which it
would matter be a data race, and therefore undefined behavior?)

Clark