Re: [Xen-ia64-devel] [RFC][PATCH] [RESEND] support special guest optimisations in the hypervisor

[Dietmar Hahn <dietmar.hahn@xxxxxxxxxxxxxxxxxxx>]

Am Montag, 2. Juli 2007 schrieb Alex Williamson:
> On Thu, 2007-06-28 at 14:52 +0200, Dietmar Hahn wrote:
> > Hi,
> >
> > attached to this mail are 2 patches:
> > - opt_feature_xen.patch
> >   implements the new hypercall  __HYPERVISOR_opt_feature in the
> > hypervisor. - opt_feature_linux.patch
> >   implements the usage of the hypercall.
> >
> > The xen - patch has real effects in vcpu_translate() on inserting
> > identity mappings in the vhpt/tlb. I'am not sure about some performance
> > impacts. But without such a patch correct future protection key handling
> > is impossible.
> >
> > Please send comments or commit!
>
> Hi Dietmar,
>
>    I have a couple comments.  First, a minor point, but isn't an
> "identity mapping" feature a little vague?  Do we specifically want a
> "region 7 identity mapping" feature bit?  Second, the HVM code recently
You are right, using a more specific name is much better. Maybe
IA64_OPT_FEATURE_IDENT_MAP_PV_REG7

> added support for the for the guest to identify itself using an ACPI
> _OSI call.  This has some similarity with what this patch is trying to
> do for PV domains.  I wonder if it might be appropriate to make the HVM
> side explicitly set feature bits so we don't have one method of setting
> features on HVM and another on PV.  Maybe set_os_type() could make the
> appropriate domain_opt_feature() calls for HVM domains(?).  The PV
> domains should of course use the hypercall interface for setting
> features.  Thanks,
I had a look at this code (didn't do this before because I have no vti 
machine). Everywhere GOS_WINDOWS() is currently used a opt_feature can be 
used.
First question: Should we add 2 opt_features, one for both region 4 and region 
5, or one for each.
Second question: would having arguments to this feature calls make sense?
Please Anthony can you comment on this!
Many thanks.

Dietmar.

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel