Re: [V7 PATCH 5/7] pvh: change xsm_add_to_physmap

[Mukesh Rathor <mukesh.rathor@oracle.com>]

On Tue, 28 Jan 2014 10:31:36 +0000
"Jan Beulich" <JBeulich@suse.com> wrote:

> >>> On 28.01.14 at 02:55, Mukesh Rathor <mukesh.rathor@oracle.com>
> >>> wrote:
> > --- a/xen/common/memory.c
> > +++ b/xen/common/memory.c
....
> The only think x86-specific here is that {get,put}_pg_owner() may
> not exist on ARM. But the general operation isn't x86-specific, so
> there shouldn't be any CONFIG_X86 dependency here. Instead
> you ought to work out with the ARM maintainers whether to stub
> out those two functions, or whether the functionality is useful
> there too (and hence proper implementations would be needed).
> 
> In the latter case I would then also wonder whether the x86
> implementation shouldn't be moved into common code.

Stefano/Ian:

If you have use for get_pg_owner() I can stub it out for now and
have it return 1, as NULL would result in error. Otherwise, I can
change the function prototype to return rc with ARM always returning 
0 and not doing anything, like:

        if ( xatpb.space == XENMAPSPACE_gmfn_foreign )
        {
            if ( (rc = get_pg_owner(xatpb.foreign_domid, &fd)) )
            {
                rcu_unlock_domain(d);
                return rc;
            }
        }

which on ARM would always return 0, setting fd to NULL.

If you think it would be needed in ARM, I can just leave the function
prototype the same and you guys can implement whenever as I don't have the
insight into ARM, and if it looks the same as x86 you can commonise it too.

thanks
Mukesh