Re: [PATCHv3 0/2] xen/privcmd: support for paged-out frames

[Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>]

On Wed, Sep 05, 2012 at 03:51:03PM -0400, Andres Lagar-Cavilla wrote:
> 
> On Sep 5, 2012, at 2:57 PM, Konrad Rzeszutek Wilk wrote:
> 
> > On Thu, Aug 30, 2012 at 04:12:03PM -0400, Andres Lagar-Cavilla wrote:
> >> On Aug 30, 2012, at 4:05 PM, Konrad Rzeszutek Wilk wrote:
> >> 
> >>>> I think this should probably get a "Tested-by" Andres or someone else
> >>>> who uses paging before being applied.
> >>> 
> >>> How do I test it? Is there an easy explanation/tutorial Andres?
> >> 
> >> I have a unit test lying somewhere but I'll need a day to dig it out.
> > 
> > If you can send that that would be super..
> Yeah. Sorry, forgot. Will take me a bit though.
> >> 
> >> Or you can start an hvm, pause it, fire up xenpaging with debug output so it tells you what it paged out, and then trace libxc as you xc_map_foreign_bulk a known batch of evicted pfns. On the first try it should get rc == -1, errno == ENOENT, error field for the pfn == ENOENT. On the second try it should all be dandy.
> >> 
> > 
> > This works well with Xen 4.1 or should I use Xen 4.2?
> 
> Xen-4.1 is advertised to work with this but I've never tried. Xen-4.2 is what I use all the time

OK, Well I can also hoist the testing of this to you.
So let me make sure it works overnight and then push out the patches to my #linux-next
tree and you can test it and make sure nothing is amiss.


> 
> Andres
> 
> >> Thanks
> >> Andres