Just to make it easier to keep straight what has been pushed, I'm resending the last two patches that still need an ACK. Doug had reviewed 1/2, but didn't ACK because he wanted to get more opinions on whether it's better to a) have a special case that doesn't specify any id for the primary sata controller (relying on qemu to imply the proper controller) or b) have a special case that uses "ide.0" for the primary sata controller (which we've been told will work by qemu people). There was a 3rd patch that we've decided not to patch, waiting for a qemu fix to be pushed instead - the patch that changed all i82801b11-bridge controllers into pci-bridge controllers to work around a qemu bug. Laine Stump (2): qemu: enable using implicit sata controller in q35 machines qemu: improve error reporting during PCI address validation src/qemu/qemu_command.c | 247 +++++++++++++++--------- tests/qemuxml2argvdata/qemuxml2argv-q35.args | 2 + tests/qemuxml2argvdata/qemuxml2argv-q35.xml | 5 + tests/qemuxml2argvtest.c | 2 +- tests/qemuxml2xmloutdata/qemuxml2xmlout-q35.xml | 5 + 5 files changed, 171 insertions(+), 90 deletions(-) -- 1.7.11.7 -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list