Re: [PATCH 03/18] PVH xen: create domctl_memory_mapping() function

["Jan Beulich" <JBeulich@suse.com>]

>>> On 05.06.13 at 02:47, Mukesh Rathor <mukesh.rathor@oracle.com> wrote:
> On Fri, 31 May 2013 10:46:45 +0100
> "Jan Beulich" <JBeulich@suse.com> wrote:
> 
>> >>> On 25.05.13 at 03:25, Mukesh Rathor <mukesh.rathor@oracle.com>
>> >>> wrote:
>> > Changes in V5:
>> >   - Move iomem_access_permitted check to case statment from the
>> > function as current doesn't point to dom0 during construct_dom0.
>> > 
>> > Changes in V6:
>> >   - Move iomem_access_permitted back to domctl_memory_mapping() as
>> > it should be after the sanity and wrap checks of mfns/gfns.
>> 
>> So you're undoing what you previously did? How can it work then?
>> Or is the whole patch perhaps no longer needed with Dom0 code
>> dropped for the time being?
>> 
>> Otherwise, rather than undoing what you did in v5, I'd think you'd
>> want to keep the sanity/wrap checks in the caller as well, since if
>> you need the function to be split out just for Dom0 building, you
>> can assume the inputs to be sane. And if so, the XSM check could
>> remain in the caller too. That would also make sure there really is
>> no change in functionality:
> 
> Well, thats where I had it originally, but one of the reviewers early on
> suggested moving it to the function, in case there's another caller in
> future. 
> 
> It's only for dom0, so I can drop the check ' !is_idle_vcpu(current) ' and
> add it during dom0 construction patch, or drop the whole patch and add it
> to construct dom0 patch series, now phase 1.5. Please lmk.

If you don't need it for the current phase, best would indeed seem
to be to drop it from the current series.

Jan