vDPA Scalable Function setup for VMs

This article describes how to configure Scalable Function (referred to below as "SF") on Debian.
SF is a technology implemented using the Sub Function of the Linux Kernel, and a similar technology is Intel Scalable IOV.

This article will describe how to set up inter-VM communication using SF instead of SR-IOV VF in VMs using vhost vDPA module.

Requirement:

Hardware: MLX 6

Software:

kernel: 5.7+

qemu 5.1+

libvirt 6.9.0+

Setup:

1. Kernel Config Dependence

MLX5_ESWITCH
MLX5_SF
VDPA

QEMU Config

../configure --enable-vhost-vdpa

2. MLX Setup

Changing the operating mode

devlink dev eswitch set pci/0000:31:00.0 mode switchdev &&¥
devlink dev eswitch set pci/0000:31:00.1 mode switchdev

Checking the operation mode

devlink dev eswitch show pci/0000:31:00.0
devlink dev eswitch show pci/0000:31:00.1
devlink port show

lyan@s2r5node80:~$ devlink dev eswitch show pci/0000:31:00.0
kernel answers: Operation not permitted
lyan@s2r5node80:~$ sudo devlink dev eswitch show pci/0000:31:00.0
pci/0000:31:00.0: mode switchdev inline-mode none encap-mode basic
lyan@s2r5node80:~$ sudo devlink dev eswitch show pci/0000:31:00.1
pci/0000:31:00.1: mode switchdev inline-mode none encap-mode basic
lyan@s2r5node80:~$ devlink port show
pci/0000:31:00.0/32774: type eth netdev tapext9867445 flavour pcisf controller 0 pfnum 0 sfnum 6 splittable false
  function:
    hw_addr f2:c4:a2:17:67:95 state active opstate attached
pci/0000:31:00.0/32775: type eth netdev tapint9867445 flavour pcisf controller 0 pfnum 0 sfnum 7 splittable false
  function:
    hw_addr 36:7c:8a:1b:57:d7 state active opstate attached
pci/0000:31:00.0/32776: type eth netdev tapext9870923 flavour pcisf controller 0 pfnum 0 sfnum 8 splittable false
  function:
    hw_addr c6:ed:f1:cc:3f:54 state active opstate attached
pci/0000:31:00.0/32777: type eth netdev tapint9870923 flavour pcisf controller 0 pfnum 0 sfnum 9 splittable false
  function:
    hw_addr ba:88:5a:f8:1e:43 state active opstate attached
pci/0000:31:00.1/131071: type eth netdev eth1 flavour physical port 1 splittable false

Configure the Scalable Function

devlink port add pci/0000:31:00.0 flavour pcisf pfnum 0 sfnum 11
devlink port function set pci/0000:31:00.0/32768 hw_addr CA:FE:C0:FF:EE:11
devlink port function set pci/0000:31:00.0/32768 state active
devlink port add pci/0000:31:00.1 flavour pcisf pfnum 1 sfnum 12
devlink port function set pci/0000:31:00.1/98304 hw_addr CA:FE:C0:FF:EE:12
devlink port function set pci/0000:31:00.1/98304 state active

lshw -businfo -c network
lyan@s2r5node80:~$ sudo lshw -businfo -c network
Bus info          Device           Class          Description
=============================================================
pci@0000:31:00.0  eth0             network        MT2892 Famil
pci@0000:31:00.1  eth1             network        MT2892 Famil
pci@0000:31:00.0  tapext9862715    network        Ethernet int
pci@0000:31:00.0  tapint9862715    network        Ethernet int
pci@0000:31:00.0  tapext9866222    network        Ethernet int
pci@0000:31:00.0  tapint9866222    network        Ethernet int


devlink port show en7f0pf0sf11 -jp
devlink port show en7f1pf1sf12 -jp

3. vdpa setup

modprobe vdpa
modprobe vhost_vdpa
echo mlx5_core.rdma.2 > /sys/bus/auxiliary/devices/mlx5_core.sf.2/mlx5_core.rdma.2/driver/unbind
echo mlx5_core.eth.2 > /sys/bus/auxiliary/devices/mlx5_core.sf.2/mlx5_core.eth.2/driver/unbind
echo mlx5_core.rdma.3 > /sys/bus/auxiliary/devices/mlx5_core.sf.3/mlx5_core.rdma.3/driver/unbind
echo mlx5_core.eth.3 > /sys/bus/auxiliary/devices/mlx5_core.sf.3/mlx5_core.eth.3/driver/unbind
vdpa dev add name vdpa0 mgmtdev auxiliary/mlx5_core.sf.2
vdpa dev add name vdpa1 mgmtdev auxiliary/mlx5_core.sf.3


4. ovs setup

systemctl start openvswitch
ovs-vsctl set Open_vSwitch . other_config:hw-offload=true other_config:tc-policy=none
systemctl restart openvswitch

ovs-vsctl add-br br30-ovs
ovs-vsctl add-port br30-ovs ens2f0
ovs-vsctl add-port br30-ovs en7f0pf0sf11

5. VM setup

  <interface type='vdpa'>
    <source dev='/dev/vhost-vdpa-0'/>
  </interface>

Ref:
1. https://docs.nvidia.com/networking/display/BlueFieldDPUOSv390/Scalable%20Functions
2. https://docs.nvidia.com/networking/display/MLNXOFEDv53100143/Introduction