Re: [PATCH] x86: vmx: Allow direct access to MSR_IA32_SPEC_CTRL

[David Woodhouse]

On Sun, 2018-01-28 at 12:40 -0800, Andy Lutomirski wrote:
> 
> Do you mean that the host would intercept the guest WRMSR and do
> WRMSR itself? I would suggest that doing so is inconsistent with the
> docs. As specified, doing WRMSR to write 1 to IBRS does *not*
> protect the guest.

I believe it does. Guest kernel is protected from any guest userspace
predictions learned before IBRS was last set to 1 in *any* mode,
including host.

> For that matter, what are the semantics of VMRESUME doing a write to
> IBRS as part of its MSR switch? Is it treated as IBRS=1 from guest
> context?

Why does it matter? We *have* confirmed, FWIW, that VMRESUME writing 1
to IBRS as part of its MSR switch when it was already 1 is not
optimised away and *is* treated as writing IBRS=1 again.

Attachment: smime.p7s
Description: S/MIME cryptographic signature