# Re: \[PATCH v4 5/5\] x86: drop mfence in favor of lock+addl

- From: Michael S. Tsirkin \<mst@redhat.com\>
- Date: 2016-01-27 17:56:45 
- List: virtualization
- Page: https://ratatoskr.run/virtualization/2016/01/4670384
- Thread: https://ratatoskr.run/lkml/2016/01/4670237/t.md

```text
On Wed, Jan 27, 2016 at 09:37:45AM -0800, Linus Torvalds wrote:
> On Wed, Jan 27, 2016 at 7:10 AM, Michael S. Tsirkin [ref] wrote:
> >
> > -#define __smp_mb()     mb()
> > +#define __smp_mb()     asm volatile("lock; addl $0,-4(%%esp)" ::: "memory", "cc")
> 
> So this doesn't look right for x86-64. Using %esp rather than %rsp.
> How did that even work for you?
> 
>                 Linus

Oops, didn't test this version properly.
Pls disregard, I'll repost after some proper testing.

-- 
MST
```
