On Fri, May 25, 2018 at 12:06:10AM +0200, Andrea Parri wrote:
> Hi Mark,
>
> > As an aside, If I complete the autogeneration stuff, it'll be possible
> > to generate those. I split out the necessary barriers in [1], but I
> > still have a lot of other preparatory cleanup to do.
>
> I do grasp the rationale behind that naming:
>
> __atomic_mb_{before,after}_{acquire,release,fence}()
>
> and yet I remain puzzled by it:
>
> For example, can you imagine (using):
>
> __atomic_mb_before_acquire() ?
>
> (as your __atomic_mb_after_acquire() is whispering me "acquire-fences"...)
Yes, I really do think he means acquire-fence. It is however something I
have vague memories of not being liked much because it is the memop
itself that carries the ordering.
That said, this is only an implementation detail and not a public
interface, so maybe we can get away with it.
