> -----Original Message-----
> From: Stefano Stabellini [mailto:stefano.stabell...@eu.citrix.com]
> Sent: Wednesday, May 28, 2014 8:30 PM
> To: Chen, Tiejun
> Cc: Stefano Stabellini; anthony.per...@citrix.com; m...@redhat.com;
> kelly.zyta...@amd.com; qemu-devel@nongnu.org;
> xen-de...@lists.xensource.com; peter.mayd...@linaro.org;
> anth...@codemonkey.ws; Kay, Allen M; Zhang, Yang Z
> Subject: RE: [v3][PATCH 3/5] xen, gfx passthrough: support Intel IGD
> passthrough with VT-D
>
[snip]
> > > > +/*
> > > > + * Currently we just pass this physical host bridge for IGD, 00:02.0.
> > > > + */
> > > > +static int is_igd_passthrough(PCIDevice *pci_dev) {
> > > > + PCIDevice *f = pci_dev->bus->devices[PCI_DEVFN(2, 0)];
> > > > + if (pci_dev->bus->devices[PCI_DEVFN(2, 0)]) {
> > >
> > > Isn't the purpose of this function to check that the *current*
> > > device is the graphic passthrough device?
> >
> > No.
> >
> > > In that case, shouldn't it just be:
> > >
> > > if (pci_dev) {
> > >
> >
> > Here pci_dev is just that host bridge, so here we have to get that real
> passthrough device by that given devfn to further confirm.
>
> I understand now, thanks for the explanation. Maybe you want to expand the
> comment on top of is_igd_passthrough.
Sure, its easy to do :)
Thanks
Tiejun
