Re: [Xen-devel] [PATCH v3 2/6] x86/hvm: Provide XEN_DMOP_add_to_physmap

[Ross Lagerwall <ross.lagerwall@xxxxxxxxxx>]

On 01/18/2018 09:24 AM, Jan Beulich wrote:
> > > > On 12.01.18 at 13:45, <ross.lagerwall@xxxxxxxxxx> wrote:
> > --- a/xen/arch/x86/hvm/dm.c
> > +++ b/xen/arch/x86/hvm/dm.c
> > @@ -640,6 +640,42 @@ static int dm_op(const struct dmop_args *op_args)
> >           break;
> >       }
> >   
> > +    case XEN_DMOP_add_to_physmap:
> > +    {
> > +        struct xen_dm_op_add_to_physmap *data =
> > +            &op.u.add_to_physmap;
> > +        struct xen_add_to_physmap xatp = {
> > +            .domid = op_args->domid,
> > +            .size = data->size,
> > +            .space = XENMAPSPACE_gmfn_range,
> > +            .idx = data->src_gfn,
> > +            .gpfn = data->dst_gfn,
> > +        };
> > +
> > +        if ( data->pad )
> > +        {
> > +            rc = -EINVAL;
> > +            break;
> > +        }
> > +
> > +        if ( xatp.size != data->size )
> > +        {
> > +            rc = -EOVERFLOW;
> > +            break;
> I guess this is fine for now, but clearly not really necessary, as the
> continuation logic could take care of the situation. Please add a
> comment to this effect.
Oh, I didn't think of using the continuation logic for that. I'll use it 
then.
> > --- a/xen/include/public/hvm/dm_op.h
> > +++ b/xen/include/public/hvm/dm_op.h
> > @@ -368,6 +368,24 @@ struct xen_dm_op_remote_shutdown {
> >                              /* (Other reason values are not blocked) */
> >   };
> >   
> > +/*
> > + * XEN_DMOP_add_to_physmap : Sets the GFNs at which a page range appears in
> > + *                           the specified guest's address space. Identical to
> > + *                           XENMEM_add_to_physmap with
> > + *                           space == XENMAPSPACE_gmfn_range.
> > + */
> > +#define XEN_DMOP_add_to_physmap 17
> > +
> > +struct xen_dm_op_add_to_physmap {
> > +    /* Number of GFNs to process. */
> > +    uint32_t size;
> > +    uint32_t pad;
> > +    /* Starting GFN of the source mapping page(s). */
> > +    uint64_aligned_t src_gfn;
> > +    /* Starting GFN where the source mapping page(s) should appear. */
> > +    uint64_aligned_t dst_gfn;
> > +};
> So why again is this "add-to-physmap"? It's moving memory around,
> isn't it? The comment also could be more clear in this regard.
> Everything else looks fine to me now.
Do you have a preference for what to call it?

--
Ross Lagerwall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel