This is a series of two patches:

The first fixes FTBFS when --enable-xen and --disable-xen-pci-passthrough
configure options are set with when building for the linux target os.

The second fixes a regression that was introduced many years ago with the
upgrade from the Qemu traditional device model that is still available
from xenbits.xen.org and based on very old Qemu version 0.10.2.

The regression is that the Qemu traditional device model reserves slot 2
for the Intel IGD on the PCI bus when the Intel IGD is passed through
to a Xen HVM domain, but the current Qemu upsream device model does not
and in fact results in a different slot assigned to the Intel IGD.

This behavior does not conform to the requirement that the Intel IGD must
be assigned to slot 2, as noted in docs/igd-assign.txt in the Qemu source
code: "IGD must be given address 02.0 on the PCI root bus in the VM."

I have used the second patch of the series for the past two years with
no problems. Without the patch, the reliability of PCI passthrough of the
Intel IGD to a Xen HVM guest is very poor, and in some cases the guest
fails to start without the patch.

v2: Remove From: <email address> tag at top of message

v3: No change to this cover letter since v2

v4: Use brchuckz@aol.com instead of brchuckz@netscape.net for the author's
    email address to match the address used by the same author in commits
    be9c61da and c0e86b76 

Chuck Zmudzinski (2):
  xen/pt: fix syntax error that causes FTBFS in some configurations
  xen/pt: reserve PCI slot 2 for Intel igd-passthru

 hw/i386/pc_piix.c    |  3 +++
 hw/xen/meson.build   |  2 +-
 hw/xen/xen_pt.c      | 25 +++++++++++++++++++++++++
 hw/xen/xen_pt.h      | 16 ++++++++++++++++
 hw/xen/xen_pt_stub.c |  4 ++++
 5 files changed, 49 insertions(+), 1 deletion(-)

-- 
2.37.2