[libvirt PATCH 0/2] Document basic VFIO variant driver support
update the manpage for the virsh nodedev-detach --driver option, and add a blurb to the NEWS file for the upcoming release Laine Stump (2): docs: update description of virsh nodedev-detach --driver option NEWS: document support for VFIO variant drivers NEWS.rst | 11 +++++++++++ docs/manpages/virsh.rst | 25 +++++++++++++++++-------- 2 files changed, 28 insertions(+), 8 deletions(-) -- 2.41.0
--driver can now be used to specify a specific driver to bind to the device being detached from the host driver (e.g. vfio-pci-igbvf), not just the *type* of driver (e.g. "vfio" or "xen", which are unnecessary anyway, since they are implicit in which hypervisor driver is in use) Signed-off-by: Laine Stump <laine@redhat.com> --- docs/manpages/virsh.rst | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-) diff --git a/docs/manpages/virsh.rst b/docs/manpages/virsh.rst index f4e5a0bd62..ec7c823602 100644 --- a/docs/manpages/virsh.rst +++ b/docs/manpages/virsh.rst @@ -5388,14 +5388,23 @@ nodedev-detach nodedev-detach nodedev [--driver backend_driver] -Detach *nodedev* from the host, so that it can safely be used by -guests via <hostdev> passthrough. This is reversed with -``nodedev-reattach``, and is done automatically for managed devices. - -Different backend drivers expect the device to be bound to different -dummy devices. For example, QEMU's "vfio" backend driver expects the -device to be bound to vfio-pci. The *--driver* parameter can be used -to specify the desired backend driver. +Detach *nodedev* from the host driver and bind it to a special driver +that provides the API needed by the hypervisor for assigning the +device to a virtual machine (using <hostdev> in the domain XML +definition). This is reversed with ``nodedev-reattach``, and is done +automatically by the hypervisor driver for managed devices (those +devices with "managed='yes'" in their XML definition). + +Different hypervisors expect the device being assigned to be bound to +different drivers. For example, QEMU's "vfio" backend requires the +device to be bound to the driver "vfio-pci" or to a "VFIO variant" +driver (this is a driver that supports the full API provided by +vfio-pci, plus some other APIs to support things like live +migration). The *--driver* parameter can be used to specify a +particular driver (e.g. a device-specific VFIO variant driver) the +device should be bound to. When *--driver* is omitted, the default +driver for the hypervisor is used ("vfio-pci" for QEMU, "pciback" for +Xen). nodedev-dumpxml -- 2.41.0
Signed-off-by: Laine Stump <laine@redhat.com> --- NEWS.rst | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 3571c01b29..e40c8ac259 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -17,6 +17,17 @@ v9.7.0 (unreleased) * **New features** + * qemu: basic support for use of "VFIO variant" drivers + + A VFIO variant driver is a device-specific driver that can + be used in place of the generic vfio-pci driver, and provides + extra functionality to support things like live migration of + guests with vfio-assigned devices. It can currently be used by: + + 1) setting ``managed='no'`` in the XML configuration for the device + 2) pre-binding the variant driver using the ``--driver`` option of + ``virsh nodedev-detach``. + * **Improvements** * **Bug fixes** -- 2.41.0
On 8/25/23 06:23, Laine Stump wrote:
update the manpage for the virsh nodedev-detach --driver option, and add a blurb to the NEWS file for the upcoming release
Laine Stump (2): docs: update description of virsh nodedev-detach --driver option NEWS: document support for VFIO variant drivers
NEWS.rst | 11 +++++++++++ docs/manpages/virsh.rst | 25 +++++++++++++++++-------- 2 files changed, 28 insertions(+), 8 deletions(-)
Reviewed-by: Michal Privoznik <mprivozn@redhat.com> Michal
participants (2)
-
Laine Stump -
Michal Prívozník