> On Feb 11, 2022, at 7:07 AM, Taylor R Campbell <riastr...@netbsd.org> wrote: > >> Date: Fri, 11 Feb 2022 06:52:48 -0800 >> From: Jason Thorpe <thor...@me.com> >> >> I would prefer we adopt the Solaris description about a generic >> barrier that provides "lock-is-visible-before-load/store" without >> explicitly stating "load-before-load/store", and provide a new >> membar_acquire() that means "load-before-load/store". > > What does `lock-is-visible' mean? How would you ever use it > correctly, or audit correct use? What operations do we have that have > `lock-is-visible' semantics, which don't also imply an acquire > operation and thus don't need an explicit barrier anyway?
Actually, you know what, I’m just going to defer on all of this. I have a huge beef with the modern naming of various memory barrier operations (they are not intuitive at all, IMO) … so whatever. > It turns out that _even on x86_, our membar_enter fails to implement > the semantics we documented for it Sure, change the documentation rather than fix the bug, then. -- thorpej
