Re: [RFC 0/1] virtio-net: add support for SR-IOV emulation

[Yui Washizu <yui.washidu@gmail.com>]

On 2023/07/24 15:58, Jason Wang wrote:
> On Mon, Jul 24, 2023 at 10:32 AM Yui Washizu <yui.washidu@gmail.com> wrote:
>>
>> On 2023/07/20 11:20, Jason Wang wrote:
>>> On Wed, Jul 19, 2023 at 9:59 AM Yui Washizu <yui.washidu@gmail.com> wrote:
>>>> This patch series is the first step towards enabling
>>>> hardware offloading of the L2 packet switching feature on virtio-net device to host machine.
>>>> We are considering that this hardware offloading enables
>>>> the use of high-performance networks in virtual infrastructures,
>>>> such as container infrastructures on VMs.
>>>>
>>>> To enable L2 packet switching by SR-IOV VFs, we are considering the following:
>>>> - making the guest recognize virtio-net devices as SR-IOV PF devices
>>>>     (archived with this patch series)
>>>> - allowing virtio-net devices to connect SR-IOV VFs to the backend networks,
>>>>     leaving the L2 packet switching feature to the management layer like libvirt
>>> Could you please show the qemu command line you want to propose here?
>>
>> I am considering how to specify the properties of VFs to connect SR-IOV
>> VFs to the backend networks.
>>
>>
>> For example:
>>
>>
>> qemu-system-x86_64 -device
>> pcie-root-port,port=8,chassis=8,id=pci.8,bus=pcie.0,multifunction=on
>>                      -netdev tap,id=hostnet0,vhost=on
>>                      -netdev tap,id=vfnet1,vhost=on # backend network for
>> SR-IOV VF 1
>>                      -netdev tap,id=vfnet2,vhost=on # backend network for
>> SR-IOV VF 2
>>                      -device
>> virtio-net-pci,netdev=hostnet0,sriov_max_vfs=2,sriov_netdev=vfnet1:vfnet2,...
>>
>>
>> In this example, we can specify multiple backend networks to the VFs
>> by adding "sriov_netdev" and separating them with ":".
> This seems what is in my mind as well, more below
>
>> Additionally, when passing properties like "rx_queue_size" to VFs, we
>> can utilize new properties,
>> such as "sriov_rx_queue_size_per_vfs," to ensure that the same value is
>> passed to all VFs.
> Or we can introduce new device like:
>
> -netdev tap,id=hn0 \
> -device virtio-net-pci,netdev=hn0,id=vnet_pf \
> -netdev tap,netdev=hn1 \
> -device virtio-net-pci-vf,netdev=hn1,id=vf0,pf=vnet_pf,rx_queue_size=XYZ ... \
>
> This allows us to reuse the codes for configuring vf parameters.


Yes, I was also considering this method was one of the option.


I was concerned that the implementation
for it might be different from other devices specified by the "-device" 
option
because VF devices are created when the guest creates VFs from PFs,
not when the guest starts,
and the implementation may be complicated.
So I dropped the idea then.


However, I think it is a cleaner way for specifying VFs than my 
suggested one.


I will further consider whether the implementation will be really 
difficult or not.


Thank you.


> But
> note that rx_queue_size doesn't make too much sense to vhost-vDPA, as
> qemu can perform nothing more than a simple sanity test.
>
> Thanks
>
>> I'm still considering about how to specify it, so please give me any
>> comments if you have any.
>>
>>
>>>>     - This makes hardware offloading of L2 packet switching possible.
>>>>       For example, when using vDPA devices, it allows the guest
>>>>       to utilize SR-IOV NIC embedded switch of hosts.
>>> This would be interesting.
>>>
>>> Thanks
>>>
>>>> This patch series aims to enable SR-IOV emulation on virtio-net devices.
>>>> With this series, the guest can identify the virtio-net device as an SR-IOV PF device.
>>>> The newly added property 'sriov_max_vfs' allows us to enable the SR-IOV feature
>>>> on the virtio-net device.
>>>> Currently, we are unable to specify the properties of a VF created from the guest.
>>>> The properties are set to their default values.
>>>> In the future, we plan to allow users to set the properties.
>>>>
>>>> qemu-system-x86_64 --device virtio-net,sriov_max_vfs=<num>
>>>> # when 'sriov_max_vfs' is present, the SR-IOV feature will be automatically enabled
>>>> # <num> means the max number of VF on guest
>>>>
>>>> Example commands to create VFs in virtio-net device from the guest:
>>>>
>>>> guest% readlink -f /sys/class/net/eth1/device
>>>>    /sys/devices/pci0000:00/0000:00:02.0/0000:01:00.0/virtio1
>>>> guest% echo "2" > /sys/devices/pci0000:00/0000:00:02.0/0000:01:00.0/sriov_numvfs
>>>> guest% ip link show
>>>>    eth0: ....
>>>>    eth1: ....
>>>>    eth2: .... #virtual VF created
>>>>    eth3: .... #virtual VF created
>>>>
>>>> Please note that communication between VF and PF/VF is not possible by this patch series itself.
>>>>
>>>> Yui Washizu (1):
>>>>     virtio-pci: add SR-IOV capability
>>>>
>>>>    hw/pci/msix.c                  |  8 +++--
>>>>    hw/pci/pci.c                   |  4 +++
>>>>    hw/virtio/virtio-pci.c         | 62 ++++++++++++++++++++++++++++++----
>>>>    include/hw/virtio/virtio-pci.h |  1 +
>>>>    4 files changed, 66 insertions(+), 9 deletions(-)
>>>>
>>>> --
>>>> 2.39.3
>>>>