Re: [Xen-users] Have you ever met this behavior with IGD passthrough?

[Pasi Kärkkäinen <pasik@xxxxxx>]

On Wed, Aug 07, 2013 at 09:26:38PM +0800, G.R. wrote:
> On Wed, Aug 7, 2013 at 9:12 PM, Pasi Kärkkäinen <pasik@xxxxxx> wrote:
> > On Wed, Aug 07, 2013 at 09:07:40PM +0800, G.R. wrote:
> >> Actually I appreciate if you can share your experience.
> >> I really want to make sure if the issue is specific to my HW.
> >> More below.
> >>
> >
> > I think I have a similar CPU available, so I can try with that in a week or 
> > so..
> >
> 
> Do you mean that you don't see this in your current config?
> So what's your HW spec?
> 

It has been a while since I last used IGD passthru, so I'm not sure.

I started setting up the testbox for IGD passthru to Win7 last weekend,
but it seems harddisk is broken in the testbox :(

So it'll take a while until I can try it again.. I'll let you know then!


> >
> >> On Wed, Aug 7, 2013 at 6:25 PM, Pasi Kärkkäinen <pasik@xxxxxx> wrote:
> >> > On Wed, Aug 07, 2013 at 06:16:31PM +0800, G.R. wrote:
> >> >> Hi guys, any comment on this?
> >> >>
> >> >
> >> > Can you please remind about your configuration:
> >> >
> >> > - Which CPU? Which IDG/GPU?
> >> I7 3770 with IGD 4000.
> >>
> >> > - HVM guest Windows version? 32b/64b?
> >> win7 64 bit version.
> >> > - Xen version
> >> All version from xen 4.1.x to xen 4.3
> >> > - I assume you're using qemu-traditional.. which additional patches?
> >> You should know the patch set I used -- the patch series I created and
> >> tried to submit (you pinged me several times on it).
> >>
> >
> > Yep, I do :) but I mostly meant if you have any other patches in addition 
> > to those 3 patches from the IGD passthru series.
> >
> >> > - dom0 kernel version
> >> All kernel versions I ever used. 3.6.x 3.8.7 3.9.8
> >>
> >> > - xm or xl?
> >> xl
> >> > - HVM guest configfile
> >>
> >> name = 'gamebox'
> >> builder = 'hvm'
> >> vcpus = '4'
> >> cpus = "0-4"
> >> memory = '3072'
> >>
> >
> > Did you try with only 2GB? or 1GB? Does it affect anything?
> >
> I used to stick on 2048 and only increased to this value recently.
> This appear to have something to do with the IO mmap.
> But as long as it boots, it has nothing to do with the behavior I
> mentioned in my thread.
> 

OK.


-- Pasi


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