Re: [RFC KVM 03/27] KVM: x86: Introduce KVM separate virtual address space

[Alexandre Chartre]

On 5/13/19 5:45 PM, Andy Lutomirski wrote:
> On Mon, May 13, 2019 at 7:39 AM Alexandre Chartre
> <alexandre.chartre@xxxxxxxxxx> wrote:
> > From: Liran Alon <liran.alon@xxxxxxxxxx>
> >
> > Create a separate mm for KVM that will be active when KVM #VMExit
> > handlers run. Up until the point which we architectully need to
> > access host (or other VM) sensitive data.
> >
> > This patch just create kvm_mm but never makes it active yet.
> > This will be done by next commits.
>
> NAK to this whole pile of code.  KVM is not so special that it can
> duplicate core infrastructure like this.  Use copy_init_mm() or
> improve it as needed.
>
> --Andy

This was originally inspired from how efi_mm is built. If I remember
correctly copy_init_mm() or other mm init functions do initialization
we don't need in this case; we basically want a blank mm. I will have
another look at copy_init_mm().

In any case, if we really need a mm create/init function I agree it
doesn't below to kvm. For now, this part of shortcuts used for the POC.

alex.