Re: [Xen-devel] 32-on-64: pvfb issue

[Markus Armbruster <armbru@xxxxxxxxxx>]

Ian Campbell <Ian.Campbell@xxxxxxxxxxxxx> writes:

> On Thu, 2007-01-18 at 18:31 +0000, Keir Fraser wrote:
>> This is even better than with blkfront/blkback because there are definitely
>> no PV-on-HVM pvfb driver frontends out in the wild (and it's HVM that makes
>> things harder).
>
> PV on HVM for fbfront is pretty hard since the kernel.org tree does not
> export zap_page_range to modules, neither do many of the older distro
> kernels I've looked at.
>
> I tried doing a compat version like with other such functions but it
> ended up pulling half of mm/memory.c into the compat layer which I don't
> like.
>
> Do I vaguely recall plans to get rid of the use of zap_page_range or is
> my memory playing tricks?
>
> Ian.

We'd like to replace zap_page_range() with unmap_mapping_range(), for
several good reasons:

* it is already fully exported,

* it deals with locking automatically via the address_space
  i_mmap_lock spinlock,

* it automatically iterates over all the vmas on the address_space
  without us having to loop over them ourselves.

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