Re: Adding plain accesses and detecting data races in the LKMM

From: Alan Stern
Date: Mon Apr 15 2019 - 09:50:07 EST

On Mon, 15 Apr 2019, Paul E. McKenney wrote:

> Another question is "should the kernel permit smp_mb__{before,after}*()
> anywhere other than immediately before or after the primitive being
> strengthened?"
> Thoughts?

How would the kernel forbid other constructions?