[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] Re: [PATCH] libxl: enabling upstream qemu as pure pv backend.

On Wed, 8 Jun 2011, Ian Campbell wrote:
> > > > But if we do this we have to be careful because in the stubdom case the
> > > > libxl_device_model_info that we pass to libxl__create_xenpv_qemu is NOT
> > > > the same used to build the stubdom.
> > > > The info parameter passed to libxl__create_stubdom is the
> > > > libxl_device_model_info that describes the stub qemu-for-FV.
> > > > While the info parameter that we pass to libxl__create_xenpv_qemu is the
> > > > libxl_device_model_info that describes the qemu-for-PV in dom0.
> > > 
> > > Right, my suggestion was that they potentially _could_ be the same, with
> > > the two functions doinging the right thing internally. This would avoid
> > > the need to figure out which params need to be copied from the user
> > > supplied struct to the xenpv one -- they would simply be the same.
> > > 
> > 
> > Ideally the vnc and sdl proprieties of a vfb should be moved into
> > the corresponding fields of device_model_info, because they are not
> > specific to the vfb protocol, rather a configuration of the backend
> > (qemu).
> > This way we wouldn't need libxl__build_xenpv_qemu_args anymore.
> > 
> > I wouldn't use the same device_model_info instance for both qemu-FV and
> > qemu-PV in the stubdom case, I would rather keep them separate.
> OK, then we either need a way to configure both (dupping the options) or
> one needs to be derived from the other in some useful way...

First the vfb configuration is derived from the xenfv configuration by
Then from the vfb we build xenpv in libxl__create_xenpv_qemu.


If we move the sdl and vnc options away from vfb we could derive both
the xenpv configuration and the vfb configuration in
libxl__vfb_and_vkb_from_device_model_info, that would need to be
libxl__create_stubdom would just add the vfb to xenstore and call
libxl__create_xenpv_qemu that doesn't need to do any "translation"
anymore. So I guess it would simplify the code a bit.


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.