[MODERATED] Re: [PATCH v2 3/8] MDSv2 5

[Andi Kleen <ak@linux.intel.com>]

On Thu, Dec 13, 2018 at 12:48:28PM -0800, speck for Andi Kleen wrote:
> > This is actually fairly ok, because if the verw *really* doesn't do
> > anything, it also isn't very expensive (ie on the order of just 40
> > cycles instead of 400 cycles).  So I'd actually be inclined to say
> > "let's do that for the *initial* patches and back-ports" just to give
> > any vmware users the protection.
> 
> I think that's fine
> 
> 40 cycles will be hard to measure even in the most extreme 
> syscall micro benchmarks, and is unlikely to impact
> anything real.

I looked into this.

One problem that if we ignore MB_CLEAR there is no way to tell the user
whether they are mitigated or not in 

/sys/devices/system/cpu/vulnerabilities/mds

I fear while it would cause a lot of confusion.

So even if VMWare users were fixed this way there would be no
way for them to find out in Linux. And for other non VMWare
users there would be also no way to find out, which would
be quite bad.

Based on that I think we should use MB_CLEAR as in the original
patches.

Potentially have some kind of VMWare specific quirk to override 
it, but it still seems fairly bogus.

-Andi