NVIDIA MLNX_OFED Documentation Rev 5.4-2.4.1.3
Linux Kernel Upstream Release Notes v5.17

Single Root IO Virtualization (SR-IOV)

Single Root IO Virtualization (SR-IOV) is a technology that allows a physical PCIe device to present itself multiple times through the PCIe bus. This technology enables multiple virtual instances of the device with separate resources. NVIDIA adapters are capable of exposing up to 127 virtual instances (Virtual Functions (VFs) for each port in the NVIDIA ConnectX® family cards. These virtual functions can then be provisioned separately. Each VF can be seen as an additional device connected to the Physical Function. It shares the same resources with the Physical Function, and its number of ports equals those of the Physical Function.
SR-IOV is commonly used in conjunction with an SR-IOV enabled hypervisor to provide virtual machines direct hardware access to network resources hence increasing its performance.
In this chapter we will demonstrate setup and configuration of SR-IOV in a Red Hat Linux environment using ConnectX® VPI adapter cards.

To set up an SR-IOV environment, the following is required:

  • MLNX_OFED Driver

  • A server/blade with an SR-IOV-capable motherboard BIOS

  • Hypervisor that supports SR-IOV such as: Red Hat Enterprise Linux Server Version 6

  • NVIDIA ConnectX® VPI Adapter Card family with SR-IOV capability

Depending on your system, perform the steps below to set up your BIOS. The figures used in this section are for illustration purposes only. For further information, please refer to the appropriate BIOS User Manual:

  1. Enable "SR-IOV" in the system BIOS.

    worddavb2ee67a7eb9aae5c536610e39a37dcc5.png

  2. Enable "Intel Virtualization Technology".

    worddav6931c32564b3b0c166f4a26788219144.png

  3. Install a hypervisor that supports SR-IOV.

  4. Depending on your system, update the /boot/grub/grub.conf file to include a similar command line load parameter for the Linux kernel.
    For example, to Intel systems, add:

    Copy
    Copied!
                

    default=0 timeout=5 splashimage=(hd0,0)/grub/splash.xpm.gz hiddenmenu title Red Hat Enterprise Linux Server (4.x.x) root (hd0,0) kernel /vmlinuz-4.x.x ro root=/dev/VolGroup00/LogVol00 rhgb quiet intel_iommu=on initrd /initrd-4.x.x.img

    Note: Please make sure the parameter "intel_iommu=on" exists when updating the /boot/grub/grub.conf file, otherwise SR-IOV cannot be loaded.

    Some OSs use /boot/grub2/grub.cfg file. If your server uses such file, please edit this file instead (add “intel_iommu=on” for the relevant menu entry at the end of the line that starts with "linux16").

  1. Install the MLNX_OFED driver for Linux that supports SR-IOV.

  2. Check if SR-IOV is enabled in the firmware.

    Copy
    Copied!
                

    mlxconfig -d /dev/mst/mt4115_pciconf0 q   Device #1: ----------   Device type: Connect4 PCI device: /dev/mst/mt4115_pciconf0 Configurations: Current SRIOV_EN 1 NUM_OF_VFS 8

    Warning

    If needed, use mlxconfig to set the relevant fields:
    mlxconfig -d /dev/mst/mt4115_pciconf0 set SRIOV_EN=1 NUM_OF_VFS=16

  3. Reboot the server.

  4. Write to the sysfs file the number of Virtual Functions you need to create for the PF. You can use one of the following equivalent files:
    You can use one of the following equivalent files:
    - A standard Linux kernel generated file that is available in the new kernels.

    Copy
    Copied!
                

    echo [num_vfs] > /sys/class/infiniband/mlx5_0/device/sriov_numvfs

    Note: This file will be generated only if IOMMU is set in the grub.conf file (by adding intel_iommu=on, as seen in the fourth step under “Setting Up SR-IOV”).

    - A file generated by the mlx5_core driver with the same functionality as the kernel generated one.

    Copy
    Copied!
                

    echo [num_vfs] > /sys/class/infiniband/mlx5_0/device/mlx5_num_vfs

    Note: This file is used by old kernels that do not support the standard file. In such kernels, using sriov_numvfs results in the following error: “bash: echo: write error: Function not implemented”.

    The following rules apply when writing to these files:

    - If there are no VFs assigned, the number of VFs can be changed to any valid value (0 - max #VFs as set during FW burning)

    - If there are VFs assigned to a VM, it is not possible to change the number of VFs
    - If the administrator unloads the driver on the PF while there are no VFs assigned, the driver will unload and SRI-OV will be disabled
    - If there are VFs assigned while the driver of the PF is unloaded, SR-IOV will not be disabled. This means that VFs will be visible on the VM. However, they will not be operational. This is applicable to OSs with kernels that use pci_stub and not vfio.
    - The VF driver will discover this situation and will close its resources
    - When the driver on the PF is reloaded, the VF becomes operational. The administrator of the VF will need to restart the driver in order to resume working with the VF.

  5. Load the driver. To verify that the VFs were created. Run:

    Copy
    Copied!
                

    lspci | grep Mellanox 08:00.0 Infiniband controller: Mellanox Technologies MT27700 Family [ConnectX-4] 08:00.1 Infiniband controller: Mellanox Technologies MT27700 Family [ConnectX-4] 08:00.2 Infiniband controller: Mellanox Technologies MT27700 Family [ConnectX-4 Virtual Function] 08:00.3 Infiniband controller: Mellanox Technologies MT27700 Family [ConnectX-4 Virtual Function] 08:00.4 Infiniband controller: Mellanox Technologies MT27700 Family [ConnectX-4 Virtual Function] 08:00.5 Infiniband controller: Mellanox Technologies MT27700 Family [ConnectX-4 Virtual Function]

  6. Configure the VFs.
    After VFs are created, 3 sysfs entries per VF are available under /sys/class/infiniband/mlx5_<PF INDEX>/device/sriov (shown below for VFs 0 to 2):

    Copy
    Copied!
                

    +-- 0 | +-- node | +-- policy | +-- port +-- 1 | +-- node | +-- policy | +-- port +-- 2 +-- node +-- policy +-- port

    For each Virtual Function, the following files are available:
    - Node - Node’s GUID:
    The user can set the node GUID by writing to the /sys/class/infiniband/<PF>/device/sriov/<index>/node file. The example below, shows how to set the node GUID for VF 0 of mlx5_0.

    Copy
    Copied!
                

    echo 00:11:22:33:44:55:1:0 > /sys/class/infiniband/mlx5_0/device/sriov/0/node

    - Port - Port’s GUID:
    The user can set the port GUID by writing to the /sys/class/infiniband/<PF>/device/sriov/<index>/port file. The example below, shows how to set the port GUID for VF 0 of mlx5_0.

    Copy
    Copied!
                

    echo 00:11:22:33:44:55:2:0 > /sys/class/infiniband/mlx5_0/device/sriov/0/port

    - Policy - The vport's policy. The user can set the port GUID by writing to the /sys/class/infiniband/<PF>/device/sriov/<index>/port file. The policy can be one of:
    - Down - the VPort PortState remains 'Down'
    - Up - if the current VPort PortState is 'Down', it is modified to 'Initialize'. In all other states, it is unmodified. The result is that the SM may bring the VPort up.
    - Follow - follows the PortState of the physical port. If the PortState of the physical port is 'Active', then the VPort implements the 'Up' policy. Otherwise, the VPort PortState is 'Down'.

    Notes:

    - The policy of all the vports is initialized to “Down” after the PF driver is restarted except for VPort0 for which the policy is modified to 'Follow' by the PF driver.

    - To see the VFs configuration, you must unbind and bind them or reboot the VMs if the VFs were assigned.

  7. Make sure that OpenSM supports Virtualization (Virtualization must be enabled).
    The /etc/opensm/opensm.conf file should contain the following line:

    Copy
    Copied!
                

    virt_enabled 2 

    Note: OpenSM and any other utility that uses SMP MADs (ibnetdiscover, sminfo, iblink- info, smpdump, ibqueryerr, ibdiagnet and smpquery) should run on the PF and not on the VFs. In case of multi PFs (multi-host), OpenSM should run on Host0.

VFs Initialization Note

Since the same mlx5_core driver supports both Physical and Virtual Functions, once the Virtual Functions are created, the driver of the PF will attempt to initialize them so they will be available to the OS owning the PF. If you want to assign a Virtual Function to a VM, you need to make sure the VF is not used by the PF driver. If a VF is used, you should first unbind it before assigning to a VM.

Procedure_Heading_Icon.PNG

To unbind a device use the following command:

  1. Get the full PCI address of the device.

    Copy
    Copied!
                

    lspci -D

    Example:

    Copy
    Copied!
                

    0000:09:00.2

  2. Unbind the device.

    Copy
    Copied!
                

    echo 0000:09:00.2 > /sys/bus/pci/drivers/mlx5_core/unbind

  3. Bind the unbound VF.

    Copy
    Copied!
                

    echo 0000:09:00.2 > /sys/bus/pci/drivers/mlx5_core/bind

PCI BDF Mapping of PFs and VFs

PCI addresses are sequential for both of the PF and their VFs. Assuming the card's PCI slot is 05:00 and it has 2 ports, the PFs PCI address will be 05:00.0 and 05:00.1.
Given 3 VFs per PF, the VFs PCI addresses will be:

Copy
Copied!
            

05:00.2-4 for VFs 0-2 of PF 0 (mlx5_0) 05:00.5-7 for VFs 0-2 of PF 1 (mlx5_1)


Assigning a Virtual Function to a Virtual Machine

This section describes a mechanism for adding a SR-IOV VF to a Virtual Machine.

Assigning the SR-IOV Virtual Function to the Red Hat KVM VM Server

  1. Run the virt-manager.

  2. Double click on the virtual machine and open its Properties.

  3. Go to Details → Add hardware → PCI host device.

    image2019-3-8_12-50-6.png

  4. Choose a NVIDIA virtual function according to its PCI device (e.g., 00:03.1)

  5. If the Virtual Machine is up reboot it, otherwise start it.

  6. Log into the virtual machine and verify that it recognizes the NVIDIA card. Run:

    Copy
    Copied!
                

    lspci | grep Mellanox

    Example:

    Copy
    Copied!
                

    lspci | grep Mellanox 01:00.0 Infiniband controller: Mellanox Technologies MT28800 Family [ConnectX-5 Ex]

  7. Add the device to the /etc/sysconfig/network-scripts/ifcfg-ethX configuration file. The MAC address for every virtual function is configured randomly, therefore it is not necessary to add it.

Ethernet Virtual Function Configuration when Running SR-IOV

SR-IOV Virtual function configuration can be done through Hypervisor iprout2/netlink tool, if present. Otherwise, it can be done via sysfs.

Copy
Copied!
            

ip link set { dev DEVICE | group DEVGROUP } [ { up | down } ] ... [ vf NUM [ mac LLADDR ] [ vlan VLANID [ qos VLAN-QOS ] ] ... [ spoofchk { on | off} ] ] ...   sysfs configuration (ConnectX-4):   /sys/class/net/enp8s0f0/device/sriov/[VF]   +-- [VF] | +-- config | +-- link_state | +-- mac | +-- mac_list | +-- max_tx_rate | +-- min_tx_rate | +-- spoofcheck | +-- stats | +-- trunk | +-- trust | +-- vlan

VLAN Guest Tagging (VGT) and VLAN Switch Tagging (VST)

When running ETH ports on VGT, the ports may be configured to simply pass through packets as is from VFs (VLAN Guest Tagging), or the administrator may configure the Hypervisor to silently force packets to be associated with a VLAN/Qos (VLAN Switch Tagging).
In the latter case, untagged or priority-tagged outgoing packets from the guest will have the VLAN tag inserted, and incoming packets will have the VLAN tag removed.
The default behavior is VGT.

To configure VF VST mode, run:

Copy
Copied!
            

ip link set dev <PF device> vf <NUM> vlan <vlan_id> [qos <qos>]

where:

  • NUM = 0..max-vf-num

  • vlan_id = 0..4095

  • qos = 0..7

For example:

  • ip link set dev eth2 vf 2 vlan 10 qos 3 - sets VST mode for VF #2 belonging to PF eth2, with vlan_id = 10 and qos = 3

  • ip link set dev eth2 vf 2 vlan 0 - sets mode for VF 2 back to VGT

Additional Ethernet VF Configuration Options

  • Guest MAC configuration - by default, guest MAC addresses are configured to be all zeroes. If the administrator wishes the guest to always start up with the same MAC, he/she should configure guest MACs before the guest driver comes up. The guest MAC may be configured by using:

    Copy
    Copied!
                

    ip link set dev <PF device> vf <NUM> mac <LLADDR>

    For legacy and ConnectX-4 guests, which do not generate random MACs, the administrator should always configure their MAC addresses via IP link, as above.

  • Spoof checking - Spoof checking is currently available only on upstream kernels newer than 3.1.

    Copy
    Copied!
                

    ip link set dev <PF device> vf <NUM> spoofchk [on | off]

  • Guest Link State

    Copy
    Copied!
                

    ip link set dev <PF device> vf <UM> state [enable| disable| auto]

Virtual Function Statistics

Virtual function statistics can be queried via sysfs:

Copy
Copied!
            

cat /sys/class/infiniband/mlx5_2/device/sriov/2/stats tx_packets : 5011 tx_bytes : 4450870 tx_dropped : 0 rx_packets : 5003 rx_bytes : 4450222 rx_broadcast : 0 rx_multicast : 0 tx_broadcast : 0 tx_multicast : 8 rx_dropped : 0


Mapping VFs to Ports

Procedure_Heading_Icon.PNG

To view the VFs mapping to ports:

Use the ip link tool v2.6.34~3 and above.

Copy
Copied!
            

ip link

Output:

Copy
Copied!
            

61: p1p1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000 link/ether 00:02:c9:f1:72:e0 brd ff:ff:ff:ff:ff:ff vf 0 MAC 00:00:00:00:00:00, vlan 4095, spoof checking off, link-state auto vf 37 MAC 00:00:00:00:00:00, vlan 4095, spoof checking off, link-state auto vf 38 MAC ff:ff:ff:ff:ff:ff, vlan 65535, spoof checking off, link-state disable vf 39 MAC ff:ff:ff:ff:ff:ff, vlan 65535, spoof checking off, link-state disable

When a MAC is ff:ff:ff:ff:ff:ff, the VF is not assigned to the port of the net device it is listed under. In the example above, vf38 is not assigned to the same port as p1p1, in contrast to vf0.

However, even VFs that are not assigned to the net device, could be used to set and change its settings. For example, the following is a valid command to change the spoof check:

Copy
Copied!
            

ip link set dev p1p1 vf 38 spoofchk on

This command will affect only the vf38. The changes can be seen in ip link on the net device that this device is assigned to.

Mapping VFs to Ports using the mlnx_get_vfs.pl tool

Procedure_Heading_Icon.PNG

To map the PCI representation in BDF to the respective ports, run:

Copy
Copied!
            

mlnx_get_vfs.pl

Output:

Copy
Copied!
            

BDF 0000:04:00.0 Port 1: 2 vf0 0000:04:00.1 vf1 0000:04:00.2 Port 2: 2 vf2 0000:04:00.3 vf3 0000:04:00.4 Both: 1 vf4 0000:04:00.5


RoCE Support

RoCE is supported on Virtual Functions and VLANs may be used with it. For RoCE, the hypervisor GID table size is of 16 entries while the VFs share the remaining 112 entries. When the number of VFs is larger than 56 entries, some of them will have GID table with only a single entry which is inadequate if VF's Ethernet device is assigned with an IP address.

Virtual Guest Tagging (VGT+)

VGT+ is an advanced mode of Virtual Guest Tagging (VGT), in which a VF is allowed to tag its own packets as in VGT, but is still subject to an administrative VLAN trunk policy. The policy determines which VLAN IDs are allowed to be transmitted or received. The policy does not determine the user priority, which is left unchanged.
Packets can be sent in one of the following modes: when the VF is allowed to send/receive untagged and priority tagged traffic and when it is not. No default VLAN is defined for VGT+ port. The send packets are passed to the eSwitch only if they match the set, and the received packets are forwarded to the VF only if they match the set.

Configuration

Warning

When working in SR-IOV, the default operating mode is VGT.

Procedure_Heading_Icon.PNG

To enable VGT+ mode:

Set the corresponding port/VF (in the example below port eth5, VF0) range of allowed VLANs.

Copy
Copied!
            

echo "<add> <start_vid> <end_vid>" > /sys/class/net/eth5/device/sriov/0/trunk

Examples:

  • Adding VLAN ID range (4-15) to trunk:

    Copy
    Copied!
                

    echo add 4 15 > /sys/class/net/eth5/device/sriov/0/trunk

  • Adding a single VLAN ID to trunk:

    Copy
    Copied!
                

    echo add 17 17 > /sys/class/net/eth5/device/sriov/0/trunk

Note: When VLAN ID = 0, it indicates that untagged and priority-tagged traffics are allowed

Procedure_Heading_Icon.PNG

To disable VGT+ mode, make sure to remove all VLANs.

Copy
Copied!
            

echo rem 0 4095 > /sys/class/net/eth5/device/sriov/0/trunk

Procedure_Heading_Icon.PNG

To remove selected VLANs.

  • Remove VLAN ID range (4-15) from trunk:

    Copy
    Copied!
                

    echo rem 4 15 > /sys/class/net/eth5/device/sriov/0/trunk

  • Remove a single VLAN ID from trunk:

    Copy
    Copied!
                

    echo rem 17 17 > /sys/class/net/eth5/device/sriov/0/trunk

SR-IOV Advanced Security Features

SR-IOV MAC Anti-Spoofing

Normally, MAC addresses are unique identifiers assigned to network interfaces, and they are fixed addresses that cannot be changed. MAC address spoofing is a technique for altering the MAC address to serve different purposes. Some of the cases in which a MAC address is altered can be legal, while others can be illegal and abuse security mechanisms or disguises a possible attacker.
The SR-IOV MAC address anti-spoofing feature, also known as MAC Spoof Check provides protection against malicious VM MAC address forging. If the network administrator assigns a MAC address to a VF (through the hypervisor) and enables spoof check on it, this will limit the end user to send traffic only from the assigned MAC address of that VF.

MAC Anti-Spoofing Configuration

Warning

MAC anti-spoofing is disabled by default.

In the configuration example below, the VM is located on VF-0 and has the following MAC address: 11:22:33:44:55:66.
There are two ways to enable or disable MAC anti-spoofing:

  1. Use the standard IP link commands - available from Kernel 3.10 and above.

    1. To enable MAC anti-spoofing, run:

      Copy
      Copied!
                  

      ip link set ens785f1 vf 0 spoofchk on

    2. To disable MAC anti-spoofing, run:

      Copy
      Copied!
                  

      ip link set ens785f1 vf 0 spoofchk off

  2. Specify echo "ON" or "OFF" to the file located under /sys/class/net/<ifname / device/sriov/<VF index>/spoofcheck.

    1. To enable MAC anti-spoofing, run:

      Copy
      Copied!
                  

      echo "ON" > /sys/class/net/ens785f1/vf/0/spoofchk

    2. To disable MAC anti-spoofing, run:

      Copy
      Copied!
                  

      echo "OFF" > /sys/class/net/ens785f1/vf/0/spoofchk

Warning

This configuration is non-persistent and does not survive driver restart.

Limit and Bandwidth Share Per VF

This feature enables rate limiting traffic per VF in SR-IOV mode. For details on how to configure rate limit per VF for ConnectX-4 and above adapter cards, please refer to HowTo Configure Rate Limit per VF for ConnectX-4/ConnectX-5/ConnectX-6 Community post.

Limit Bandwidth per Group of VFs

VFs Rate Limit for vSwitch (OVS) feature allows users to join available VFs into groups and set a rate limitation on each group. Rate limitation on a VF group ensures that the total Tx bandwidth that the VFs in this group get (altogether combined) will not exceed the given value.
With this feature, a VF can still be configured with an individual rate limit as in the past (under /sys/class/net/<ifname>/device/sriov/<vf_num>/max_tx_rate). However, the actual bandwidth limit on the VF will eventually be determined considering the VF group limitation and how many VFs are in the same group.
For example: 2 VFs (0 and 1) are attached to group 3.

Case 1: The rate limitation on the group is set to 20G. Rate limit of each VF is 15G

Result: Each VF will have a rate limit of 10G

Case 2: Group’s max rate limitation is still set to 20G. VF 0 is configured to 30G limit, while VF 1 is configured to 5G rate limit

Result: VF 0 will have 15G de-facto. VF 1 will have 5G

The rule of thumb is that the group’s bandwidth is distributed evenly between the number of VFs in the group. If there are leftovers, they will be assigned to VFs whose individual rate limit has not been met yet.

VFs Rate Limit Feature Configuration

  1. When VF rate group is supported by FW, the driver will create a new hierarchy in the SRI-OV sysfs named “groups” (/sys/class/net/<ifname>/device/sriov/groups/). It will contain all the info and the configurations allowed for VF groups.

  2. All VFs are placed in group 0 by default since it is the only existing group following the initial driver start. It would be the only group available under /sys/class/net/<ifname>/device/sriov/groups/

  3. The VF can be moved to a different group by writing to the group file -> echo $GROUP_ID > /sys/class/net/<ifname>/device/sriov/<vf_id>/group

  4. The group IDs allowed are 0-255

  5. Only when there is at least 1 VF in a group, there will be a group configuration available under /sys/class/net/<ifname>/device/sriov/groups/ (Except for group 0, which is always available even when it’s empty).

  6. Once the group is created (by moving at least 1 VF to that group), users can configure the group’s rate limit. For example:

    1. echo 10000 > /sys/class/net/<ifname>/device/sriov/5/max_tx_rate – setting individual rate limitation of VF 5 to 10G (Optional)

    2. echo 7 > /sys/class/net/<ifname>/device/sriov/5/group – moving VF 5 to group 7

    3. echo 5000 > /sys/class/net/<ifname>/device/sriov/groups/7/max_tx_rate – setting group 7 with rate limitation of 5G

    4. When running traffic via VF 5 now, it will be limited to 5G because of the group rate limit even though the VF itself is limited to 10G

    5. echo 3 > /sys/class/net/<ifname>/device/sriov/5/group – moving VF 5 to group 3

    6. Group 7 will now disappear from /sys/class/net/<ifname>/device/sriov/groups since there are 0 VFs in it. Group 3 will now appear. Since there’s no rate limit on group 3, VF 5 can transmit at 10G (thanks to its individual configuration)

Notes:

  • You can see to which group the VF belongs to in the ‘stats’ sysfs (cat /sys/class/net/<ifname>/device/sriov/<vf_num>/stats)

  • You can see the current rate limit and number of attached VFs to a group in the group’s ‘config’ sysfs (cat /sys/class/net/<ifname>/device/sriov/groups/<group_id>/config)

Bandwidth Guarantee per Group of VFs

Bandwidth guarantee (minimum BW) can be set on a group of VFs to ensure this group is able to transmit at least the amount of bandwidth specified on the wire.

Note the following:

  • The minimum BW settings on VF groups determine how the groups share the total BW between themselves. It does not impact an individual VF’s rate settings.

  • The total minimum BW that is set on the VF groups should not exceed the total line rate. Otherwise, results are unexpected.

  • It is still possible to set minimum BW on the individual VFs inside the group. This will determine how the VFs share the group’s minimum BW between themselves. The total minimum BW of the VF member should not exceed the minimum BW of the group.

For instruction on how to create groups of VFs, see Limit Bandwidth per Group of VFs above.

Example

With a 40Gb link speed, assuming 4 groups and default group 0 have been created:

Copy
Copied!
            

echo 20000 > /sys/class/net/<ifname>/device/sriov/group/1/min_tx_rate echo 5000 > /sys/class/net/<ifname>/device/sriov/group/2/min_tx_rate echo 15000 > /sys/class/net/<ifname>/device/sriov/group/3/min_tx_rate

Copy
Copied!
            

Group 0(default) : 0 - No BW guarantee is configured. Group 1 : 20000 - This is the maximum min rate among groups Group 2 : 5000 which is 25% of the maximum min rate Group 3 : 15000 which is 75% of the maximum min rate Group 4 : 0 - No BW guarantee is configured.

Assuming there are VFs attempting to transmit in full line rate in all groups, the results would look like: In which case, the minimum BW allocation would be:

Copy
Copied!
            

Group0 – Will have no BW to use since no BW guarantee was set on it while other groups do have such settings. Group1 – Will transmit at 20Gb/s Group2 – Will transmit at 5Gb/s Group3 – Will transmit at 15Gb/s Group4 - Will have no BW to use since no BW guarantee was set on it while other groups do have such settings.

Privileged VFs

In case a malicious driver is running over one of the VFs, and in case that VF's permissions are not restricted, this may open security holes. However, VFs can be marked as trusted and can thus receive an exclusive subset of physical function privileges or permissions. For example, in case of allowing all VFs, rather than specific VFs, to enter a promiscuous mode as a privilege, this will enable malicious users to sniff and monitor the entire physical port for incoming traffic, including traffic targeting other VFs, which is considered a severe security hole.

Privileged VFs Configuration

In the configuration example below, the VM is located on VF-0 and has the following MAC address: 11:22:33:44:55:66.
There are two ways to enable or disable trust:

  1. Use the standard IP link commands - available from Kernel 4.5 and above.

    1. To enable trust for a specific VF, run:

      Copy
      Copied!
                  

      ip link set ens785f1 vf 0 trust on

    2. To disable trust for a specific VF, run:

      Copy
      Copied!
                  

      ip link set ens785f1 vf 0 trust off

  2. Specify echo "ON" or "OFF" to the file located under /sys/class/net/<ETH_IF_NAME> / device/sriov/<VF index>/trust.

  1. To enable trust for a specific VF, run:

    Copy
    Copied!
                

    echo "ON" > /sys/class/net/ens785f1/device/sriov/0/trust

  2. To disable trust for a specific VF, run:

    Copy
    Copied!
                

    echo "OFF" > /sys/class/net/ens785f1/device/sriov/0/trust

Probed VFs

Probing Virtual Functions (VFs) after SR-IOV is enabled might consume the adapter cards' resources. Therefore, it is recommended not to enable probing of VFs when no monitoring of the VM is needed.
VF probing can be disabled in two ways, depending on the kernel version installed on your server:

  1. If the kernel version installed is v4.12 or above, it is recommended to use the PCI sysfs interface sriov_drivers_autoprobe. For more information, see linux-next branch .

  2. If the kernel version installed is older than v4.12, it is recommended to use the mlx5_core module parameter probe_vf with driver version 4.1 or above.

Example:

Copy
Copied!
            

echo 0 > /sys/module/mlx5_core/parameters/probe_vf

For more information on how to probe VFs, see HowTo Configure and Probe VFs on mlx5 Driv ers Community post.

VF Promiscuous Rx Modes

VF Promiscuous Mode

VFs can enter a promiscuous mode that enables receiving the unmatched traffic and all the multicast traffic that reaches the physical port in addition to the traffic originally targeted to the VF. The unmatched traffic is any traffic's DMAC that does not match any of the VFs' or PFs' MAC addresses.

Note: Only privileged/trusted VFs can enter the VF promiscuous mode.

Procedure_Heading_Icon.PNG

To set the promiscuous mode on for a VF, run:

Copy
Copied!
            

ifconfig eth2 promisc

To exit the promiscuous mode, run:

Procedure_Heading_Icon.PNG

Copy
Copied!
            

ifconfig eth2 –promisc


VF All-Multi Mode

VFs can enter an all-multi mode that enables receiving all the multicast traffic sent from/to the other functions on the same physical port in addition to the traffic originally targeted to the VF.

Note: Only privileged/trusted VFs can enter the all-multi RX mode.

To set the all-multi mode on for a VF, run:

Procedure_Heading_Icon.PNG

Copy
Copied!
            

ifconfig eth2 allmulti

To exit the all-multi mode, run:

Procedure_Heading_Icon.PNG

Copy
Copied!
            

#ifconfig eth2 –allmulti

To uninstall SR-IOV driver, perform the following:

Procedure_Heading_Icon.PNG

  1. For Hypervisors, detach all the Virtual Functions (VF) from all the Virtual Machines (VM) or stop the Virtual Machines that use the Virtual Functions.

    Please be aware that stopping the driver when there are VMs that use the VFs, will cause machine to hang.

  2. Run the script below. Please be aware, uninstalling the driver deletes the entire driver's file, but does not unload the driver.

    Copy
    Copied!
                

    [root@swl022 ~]# /usr/sbin/ofed_uninstall.sh This program will uninstall all OFED packages on your machine. Do you want to continue?[y/N]:y Running /usr/sbin/vendor_pre_uninstall.sh Removing OFED Software installations Running /bin/rpm -e --allmatches kernel-ib kernel-ib-devel libibverbs libibverbs-devel libibverbs-devel-static libibverbs-utils libmlx4 libmlx4-devel libibcm libibcm-devel libibumad libibumad-devel libibumad-static libibmad libibmad-devel libibmad-static librdmacm librdmacm-utils librdmacm-devel ibacm opensm-libs opensm-devel perftest compat-dapl compat-dapl-devel dapl dapl-devel dapl-devel-static dapl-utils srptools infiniband-diags-guest ofed-scripts opensm-devel warning: /etc/infiniband/openib.conf saved as /etc/infiniband/openib.conf.rpmsave Running /tmp/2818-ofed_vendor_post_uninstall.sh

  3. Restart the server.

© Copyright 2023, NVIDIA. Last updated on Oct 23, 2023.