Re: [Xen-devel] [PATCH v3] docs: add PVH specification

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 26.09.14 at 02:00, <mukesh.rathor@xxxxxxxxxx> wrote:
> On Tue, 23 Sep 2014 14:16:46 +0100
> "Jan Beulich" <JBeulich@xxxxxxxx> wrote:
> 
>> >>> On 23.09.14 at 02:38, <mukesh.rathor@xxxxxxxxxx> wrote:
>> > On Thu, 18 Sep 2014 19:19:24 +0200
>> > Roger Pau Monne <roger.pau@xxxxxxxxxx> wrote:
>> > 
>> >> Introduce a document that describes the interfaces used on PVH.
>> >> This document has been designed from a guest OS point of view
>> >> (i.e.: what a guest needs to do in order to support PVH).
>> > ..... 
>> >> +
>> >> +*TODO*: we need to figure out what to do with MMIO regions, right
>> >> now Xen +sets all the holes in the native e820 to MMIO regions for
>> >> Dom0 up to 4GB. We +need to decide what to do with MMIO regions
>> >> above 4GB on Dom0, and what to do +for PVH DomUs with
>> >> pci-passthrough.
>> > 
>> > My previous comment in earlier version on this:
>> > 
>> > "We map all non-ram regions for dom0 1:1 till the highest non-ram
>> > e820 entry. If there is anything that is beyond the last e820 entry,
>> > it will remain unmapped."
>> 
>> But that's something that needs fixing rather than spelling out in
>> the documentation. I.e. Roger having this as a TODO seems quite
>> right to me.
> 
> Yes, but what Roger is saying implies we don't map above 4GB which
> is incorrect. Perhaps:
> 
> We map all non-ram regions for dom0 1:1 till the last e820 entry. If the 
> last entry ends below 4GB, then the remaining space is mapped 1:1 upto 4GB.
> This implies that if there is any region beyond the last e820 entry above
> 4GB, it is not mapped.  
> TODO: Map region beyond last e820 if it's above 4GB. Add support for domUs
> with pci passthru.

Hmm, yeah, you wording is indeed more precise, but for the vast
majority of systems they'll both end up being equivalent in effect
since memory almost always is contiguous from the 4Gb boundary
up to TOM.

Jan


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