Re: [PATCH v2 for-next v2 8/8] x86/mm: split HVM grant table code to hvm/mm.c

[Wei Liu <wei.liu2@citrix.com>]

On Fri, Apr 21, 2017 at 04:45:35AM -0600, Jan Beulich wrote:
> >>> On 21.04.17 at 12:41, <wei.liu2@citrix.com> wrote:
> > On Wed, Apr 19, 2017 at 09:35:50AM -0600, Jan Beulich wrote:
> >> >>> On 03.04.17 at 13:22, <wei.liu2@citrix.com> wrote:
> >> > PG_external requires hardware support so the code guarded by that is HVM
> >> > only.
> >> > 
> >> > Signed-off-by: Wei Liu <wei.liu2@citrix.com>
> >> > ---
> >> >  xen/arch/x86/hvm/Makefile         |  1 +
> >> >  xen/arch/x86/hvm/mm.c             | 85 
> > +++++++++++++++++++++++++++++++++++++++
> >> 
> >> I think the moved code would better go under x86/mm/, possibly (but
> >> not necessarily) into p2m.c.
> > 
> > I disagree here. What good is there to keep this code when hvm support
> > is compiled out?
> 
> Hmm, do you mean to move e.g. mm/p2m-*.c and mm/hap/ to
> hvm/mm/ then too? I think we shouldn't go overboard with this
> shuffling.
> 

Maybe, I haven't quite made up my mind.

I take it you will be against splitting / moving p2m code?

It has lower priority on my list because p2m isn't really guest facing,
while here the grant table code is (Obviously the boundary is a bit
blurred and subject to personal taste).

Wei.

> Jan
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel