The existing virNodeDeviceDettach() assumes that there is only a
single PCI device assignment backend driver appropriate for any
hypervisor. This is no longer true, as the qemu driver is getting
support for PCI device assigment via VFIO. The new API
virNodeDeviceDetachFlags adds a driverName arg
On 04/24/2013 01:26 PM, Laine Stump wrote:
The existing virNodeDeviceDettach() assumes that there is only a
single PCI device assignment backend driver appropriate for any
hypervisor. This is no longer true, as the qemu driver is getting
support for PCI device assigment via VFIO. The new API
