[Xen-devel] [PATCH v3 0/2] libxl: change to qemu-xen (upstream) by default

[Ian Jackson <ian.jackson@xxxxxxxxxxxxx>]

Ian Campbell writes ("Re: [Xen-devel] [PATCH v2] libxl: use qemu-xen (upstream 
QEMU) as device model by default"):
> On Tue, 2012-12-04 at 12:58 +0000, Stefano Stabellini wrote:
> > -Use the device-model based upon the historical Xen fork of Qemu.  This
> > -device-model is currently the default.
> > +use the device-model merged into the upstream QEMU project.
> 
> Can you correct to "Use" as you move it please.

Fixed.

> > -use the device-model merged into the upstream QEMU project.  This
> > -device-model will become the default in a future version of Xen.
> > +Use the device-model based upon the historical Xen fork of Qemu.
> > +This device-model is still the default for NetBSD dom0.
> 
> Does that look like changing for 4.3?

I guess not but I don't think that should block this change.

> > -            b_info->device_model_version =
> > -                LIBXL_DEVICE_MODEL_VERSION_QEMU_XEN_TRADITIONAL;
> > +            b_info->device_model_version = libxl__default_device_model(gc);
> 
> For PV we stat() the executable before using it -- I don't remember why
> but should we be doing the same here?

Presumably in case the upstream build was disabled.  I think the same
logic applies so I have arranged to do the same.

> I think the logic needs to be (perhaps in Linux's
> libxl__default_device_model(gc) only)
>       if (libxl_defbool_val(b_info->device_model_stubdomain)
>               LIBXL_DEVICE_MODEL_VERSION_QEMU_XEN_TRADITIONAL;
>       else
>               LIBXL_DEVICE_MODEL_VERSION_QEMU_XEN;

Yes, done.

> > +  /* Return the system-wide default device model:
> > +   * qemu-xen for Linux, qemu-xen-traditional for NetBSD.
> 
> I bet you someone forgets to update this when NetBSD changes ;-)

That remark just needs to go IMO.

So we now have:
 [PATCH 1/2] libxl: move check for existence of qemuu device model
 [PATCH 2/2] libxl: use qemu-xen (upstream QEMU) as device model by default


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