Re: [Qemu-devel] [PATCH v2 0/4] Use of unique identifier for pairing virtio and passthrough devices...

[Venu Busireddy]

On 2018-06-27 23:12:12 +0300, Michael S. Tsirkin wrote:
> On Wed, Jun 27, 2018 at 02:59:01PM -0500, Venu Busireddy wrote:
> > On 2018-06-27 22:47:05 +0300, Michael S. Tsirkin wrote:
> > > On Wed, Jun 27, 2018 at 02:29:58PM -0500, Venu Busireddy wrote:
> > > > On 2018-06-27 15:24:58 +0300, Roman Kagan wrote:
> > > > > On Tue, Jun 26, 2018 at 10:49:30PM -0500, Venu Busireddy wrote:
> > > > > > The patch set "Enable virtio_net to act as a standby for a passthru
> > > > > > device" [1] deals with live migration of guests that use passthrough
> > > > > > devices. However, that scheme uses the MAC address for pairing
> > > > > > the virtio device and the passthrough device. The thread "netvsc:
> > > > > > refactor notifier/event handling code to use the failover framework"
> > > > > > [2] discusses an alternate mechanism, such as using an UUID, for 
> > > > > > pairing
> > > > > > the devices. Based on that discussion, proposals "Add "Group 
> > > > > > Identifier"
> > > > > > to virtio PCI capabilities." [3] and "RFC: Use of bridge devices to
> > > > > > store pairing information..." [4] were made.
> > > > > 
> > > > > I must have missed something in those threads, but where does this 
> > > > > UUID
> > > > > thing come about?  AFAICS this identifier doesn't need to be
> > > > > "universally" unique, nor persistent; it only has to be unique across
> > > > > the VM and stable throughout the VM lifetime.
> > > > 
> > > > The notion of using UUID came up in the thread
> > > > 
> > > >    https://www.spinics.net/lists/netdev/msg499011.html
> > > 
> > > That's probably because it was expected one of standard serial number 
> > > capabilities
> > > (VPD or PCI Express serial #) will be used, which are expected to be 
> > > unique.
> > > 
> > > If you are rolling your own vendor specific one, it's just an ID and
> > > does not have to be unique.
> > > 
> > > > > FWIW Hyper-V uses a 32-bit integer for this purpose, not a UUID as 
> > > > > seems
> > > > > to be implied in the thread you refer to.
> > > > 
> > > > Yes, Hyper-V uses a serial number (but I think it is 64-bit value).
> > > > However, what we are doing is similar to that. Instead of 32 bits,
> > > > we are using 128 bits.
> > > 
> > > That's OK. The name is confusing though. It's a failover group id,
> > > not a UUID.
> > 
> > Sure, we can name it whatever we want. I can change it to
> > "failover-group-id", if that is what we want to call it.
> > 
> > But what is more important is, what is represented by that name? I thought
> > we were going to use UUID. The QEMU command line changes in this patch
> > set expect the user to specify an UUID as the value for this option
> > (whatever we name it). Are we still in agreement about that, or do you
> > propose something else to be used? If so, what is it? A 32-bit number, a
> > 64-bit number, or an arbitrary string?
> > 
> > Regards,
> > 
> > Venu
> 
> If we don't really need a UUID, I'd avoid that requirement.

I don't see the need for a 128-bit UUID. I just took that approach because
UUID was mentioned in "https://www.spinics.net/lists/netdev/msg499011.html";.
Since it is unlikely to have more than 4 billion devices in the system,
a 32-bit value would be more than enough to uniquely identify devices!

I am looking for direction from you :-). Roman already opined that UUID
may be an overkill. It appears that you too are leaning that way. Would
it be acceptable if I change the group identifier ("failover-group-id")
to a 32-bit value? If you concur, I will start reworking my patch. Could
you please confirm?

Thanks,

Venu

> 
> 
> > > 
> > > > > > The current patch set includes all the feedback received for 
> > > > > > proposals [3]
> > > > > > and [4]. For the sake of completeness, patch for the virtio 
> > > > > > specification
> > > > > > is also included here. Following is the updated proposal.
> > > > > > 
> > > > > > 1. Extend the virtio specification to include a new virtio PCI 
> > > > > > capability
> > > > > >    "VIRTIO_PCI_CAP_GROUP_ID_CFG".
> > > > > > 
> > > > > > 2. Enhance the QEMU CLI to include a "uuid" option to the virtio 
> > > > > > device.
> > > > > >    The "uuid" is a string in UUID format.
> > > > > > 
> > > > > > 3. Enhance the QEMU CLI to include a "uuid" option to the bridge 
> > > > > > device.
> > > > > >    The "uuid" is a string in UUID format. Currently, PCIe bridge for
> > > > > >    the Q35 model is supported.
> > > > > > 
> > > > > > 4. The operator creates a unique identifier string using 'uuidgen'.
> > > > > > 
> > > > > > 5. When the virtio device is created, the operator uses the "uuid" 
> > > > > > option
> > > > > >    (for example, '-device virtio-net-pci,uuid="string"') and 
> > > > > > specifies
> > > > > >    the UUID created in step 4.
> > > > > > 
> > > > > >    QEMU stores the UUID in the virtio device's configuration space
> > > > > >    in the capability "VIRTIO_PCI_CAP_GROUP_ID_CFG".
> > > > > > 
> > > > > > 6. When assigning a PCI device to the guest in passthrough mode, the
> > > > > >    operator first creates a bridge using the "uuid" option (for 
> > > > > > example,
> > > > > >    '-device pcie-downstream,uuid="string"') to specify the UUID 
> > > > > > created
> > > > > >    in step 4, and then attaches the passthrough device to the 
> > > > > > bridge.
> > > > > > 
> > > > > >    QEMU stores the UUID in the configuration space of the bridge as
> > > > > >    Vendor-Specific capability (0x09). The "Vendor" here is not to be
> > > > > >    confused with a specific organization. Instead, the vendor of the
> > > > > >    bridge is QEMU. To avoid mixing up with other bridges, the bridge
> > > > > >    will be created with vendor ID 0x1b36 (PCI_VENDOR_ID_REDHAT) and
> > > > > >    device ID 0x000e (PCI_DEVICE_ID_REDHAT_PCIE_BRIDGE) if the "uuid"
> > > > > >    option is specified. Otherwise, current defaults are used.
> > > > > 
> > > > > I wonder if it makes more sense to drop the concept of failover 
> > > > > groups,
> > > > > and just refer to the standby device by device-id, like 
> > > > > 
> > > > >   -device virtio-net-pci,id=foo \
> > > > >   -device pcie-downstream,failover=foo
> > > > 
> > > > Isn't this the same as what this patch series proposes? In your
> > > > suggestion, "foo" is the entity that connects the passthrough device
> > > > and the failover device. In this patch set, that "foo" is the UUID,
> > > > and the options "id" and "failover" are replaced by "uuid". Do you 
> > > > agree?
> > > > 
> > > > Regards,
> > > > 
> > > > Venu
> > > > 
> > > > > The bridge device will then lookup the failover device, figure out the
> > > > > common identifier to expose to the guest, and defer the visibility of
> > > > > the PT device behind the bridge until the guest acknowledged the 
> > > > > support
> > > > > for failover on the PV device.
> > > > > 
> > > > > Roman.