Re: [PATCH 02/10] kvm: Check userspace_addr when modifying a memory slot

[Jason Baron <jbaron@redhat.com>]

On Fri, Dec 07, 2012 at 11:32:08AM -0700, Alex Williamson wrote:
> On Fri, 2012-12-07 at 13:17 -0500, Jason Baron wrote:
> > On Thu, Dec 06, 2012 at 03:20:37PM -0700, Alex Williamson wrote:
> > > The API documents that only flags and guest physical memory space can
> > > be modified on an existing slot, but we don't enforce that the
> > > userspace address cannot be modified.  Instead we just ignore it.
> > > This means that a user may think they've successfully moved both the
> > > guest and user addresses, when in fact only the guest address changed.
> > > Check and error instead.
> > > 
> > > Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
> > > ---
> > >  virt/kvm/kvm_main.c |    8 +++++++-
> > >  1 file changed, 7 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> > > index e426704..93213e1 100644
> > > --- a/virt/kvm/kvm_main.c
> > > +++ b/virt/kvm/kvm_main.c
> > > @@ -779,13 +779,19 @@ int __kvm_set_memory_region(struct kvm *kvm,
> > >  
> > >  	r = -ENOMEM;
> > >  
> > > -	/* Allocate if a slot is being created */
> > > +	/*
> > > +	 * Allocate if a slot is being created.  If modifying a slot,
> > > +	 * the userspace_addr cannot change.
> > > +	 */
> > >  	if (!old.npages) {
> > >  		new.user_alloc = user_alloc;
> > >  		new.userspace_addr = mem->userspace_addr;
> > >  
> > >  		if (kvm_arch_create_memslot(&new, npages))
> > >  			goto out_free;
> > > +	} else if (mem->userspace_addr != old.userspace_addr) {
> > > +		r = -EINVAL;
> > > +		goto out_free;
> > >  	}
> > >  
> > >  	/* Allocate page dirty bitmap if needed */
> > > 
> > 
> > hmmm...does this mean that on a 'destroy', where npages is 0, the user
> > has to set up userspace_addr correctly? If so, that would appear to be
> > an unwanted change in semantics here.
> 
> Good point, it does change that.  We could make this be (npages &&
> mem->userspace_addr != old.userspace_addr) to avoid that case.  Thanks,
> 
> Alex
> 

Yup, that works for me.

Thanks,

-Jason