Re: [PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

[Thomas Gleixner]

On Mon, 28 Jan 2019, Zhenzhong Duan wrote:
> On 2019/1/26 2:03, Thomas Gleixner wrote:
> > Bah, nonsense. Brain was clearly still out for lunch and I confused IBPB
> > and STIBP for a moment. cond_ibpb() is the thing issues in switch_mm() and
> > that is not leaving a stale MSR around because we only write to it when we
> > need the barrier. The bit is not stale because the barrier is only issued
> > with the write. The bit has not to be cleared.
> > 
> > So the only 'issue' what happens is that switch_to() either issues a
> > barrier too much or misses one. That's really not a problem.
> 
> Ok, yes, the purpose of this patch is to avoid the one missed barrier.

And that missed barrier is not worth it to do extra work in switch_to/mm
simply because it's a one off event and there is no way to exploit that
reliably.

Thanks,

	tglx