> On Feb 11, 2022, at 5:33 AM, Taylor R Campbell <riastr...@netbsd.org> wrote: > > So I propose to change the membar_enter documentation to match the > definitions and usage (and change the riscv definition), making it > instead: > > membar_enter() > Any load preceding membar_enter() will happen before all memory > operations following it. > > This will also let us delete the obnoxious text I added to > atomic_loadstore(9) warning about membar_enter semantics.
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”. -- thorpej