16.2. Using SR-IOV

This section covers attaching Virtual Function to a guest as an additional network device.
SR-IOV requires Intel VT-d support.

Important

Xen requires additional kernel arguments to use SR-IOV. Modify the /boot/grub/grub.conf file to enable SR-IOV. To enable SR-IOV with Xen for Intel systems append the pci_pt_e820_access=on parameter to the kernel.
default=0
timeout=5
splashimage=(hd0,0)/grub/splash.xpm.gz
hiddenmenu
title Red Hat Enterprise Linux Server (2.6.18-192.el5xen)
   root (hd0,0)
   kernel /xen.gz-2.6.18-192.el5 iommu=1
   module /vmlinuz-2.6.18-192.el5xen ro root=/dev/VolGroup00/LogVol00 pci_pt_e820_access=on
   module /initrd-2.6.18-192.el5xen.img

Procedure 16.1. Attach an SR-IOV network device

  1. Enable Intel VT-d in BIOS and in the kernel

    Enable Intel VT-D in BIOS. See Procedure 15.1, “Preparing an Intel system for PCI passthrough” for more information on enabling Intel VT-d in BIOS and the kernel, or see your system manufacturer's documentation for specific instructions.
  2. Verify support

    Verify if the PCI device with SR-IOV capabilities are detected. This example lists an Intel 82576 network interface card which supports SR-IOV. Use the lspci command to verify if the device was detected.
    # lspci
    03:00.0 Ethernet controller: Intel Corporation 82576 Gigabit Network Connection (rev 01)
    03:00.1 Ethernet controller: Intel Corporation 82576 Gigabit Network Connection (rev 01)
    

    Note

    Note that the output has been modified to remove all other devices.
  3. Start the SR-IOV kernel modules

    If the device is supported the driver kernel module should be loaded automatically by the kernel. Optional parameters can be passed to the module using the modprobe command. The Intel 82576 network interface card uses the igb driver kernel module.
    # modprobe igb [<option>=<VAL1>,<VAL2>,]
    # lsmod |grep igb
    igb    87592  0
    dca    6708    1 igb
    
  4. Activate Virtual Functions

    The max_vfs parameter of the igb module allocates the maximum number of Virtual Functions. The max_vfs parameter causes the driver to spawn, up to the value of the parameter in, Virtual Functions. For this particular card the valid range is 0 to 7.
    Remove the module to change the variable.
    # modprobe -r igb
    Restart the module with the max_vfs set to 1 or any number of Virtual Functions up to the maximum supported by your device.
    # modprobe igb max_vfs=1
    
  5. Inspect the new Virtual Functions

    Using the lspci command, list the newly added Virtual Functions attached to the Intel 82576 network device.
    # lspci | grep 82576
    03:00.0 Ethernet controller: Intel Corporation 82576 Gigabit Network Connection (rev 01)
    03:00.1 Ethernet controller: Intel Corporation 82576 Gigabit Network Connection (rev 01)
    03:10.0 Ethernet controller: Intel Corporation 82576 Virtual Function (rev 01)
    03:10.1 Ethernet controller: Intel Corporation 82576 Virtual Function (rev 01)
    
    The identifier for the PCI device is found with the -n parameter of the lspci command.
    # lspci -n | grep 03:00.0
    03:00.0 0200: 8086:10c9 (rev 01)
    # lspci -n | grep 03:10.0
    03:10.0 0200: 8086:10ca (rev 01)
    
    The Physical Function corresponds to 8086:10c9 and the Virtual Function to 8086:10ca.
  6. Find the devices with virsh

    The libvirt service must find the device to add a device to a guest. Use the virsh nodedev-list command to list available host devices.
    # virsh nodedev-list | grep 8086
    pci_8086_10c9
    pci_8086_10c9_0
    pci_8086_10ca
    pci_8086_10ca_0
    [output truncated]
    The serial numbers for the Virtual Functions and Physical Functions should be in the list.
  7. Get advanced details

    The pci_8086_10c9 is one of the Physical Functions and pci_8086_10ca_0 is the first corresponding Virtual Function for that Physical Function. Use the virsh nodedev-dumpxml command to get advanced output for both devices.
    # virsh nodedev-dumpxml pci_8086_10ca
    # virsh nodedev-dumpxml pci_8086_10ca_0
    <device>
      <name>pci_8086_10ca_0</name>
      <parent>pci_8086_3408</parent>
      <driver>
        <name>igbvf</name>
      </driver>
      <capability type='pci'>
        <domain>0</domain>
        <bus>3</bus>
        <slot>16</slot>
        <function>1</function>
        <product id='0x10ca'>82576 Virtual Function</product>
        <vendor id='0x8086'>Intel Corporation</vendor>
      </capability>
    </device>
    
    This example adds the Virtual Function pci_8086_10ca_0 to the guest in Step 8. Note the bus, slot and function parameters of the Virtual Function, these are required for adding the device.
  8. Add the Virtual Function to the guest

    1. Shut down the guest.
    2. Use the output from the virsh nodedev-dumpxml pci_8086_10ca_0 command to calculate the values for the configuration file. Convert slot and function values to hexadecimal values (from decimal) to get the PCI bus addresses. Append "0x" to the beginning of the output to tell the computer that the value is a hexadecimal number.
      The example device has the following values: bus = 3, slot = 16 and function = 1. Use the printf utility to convert decimal values to hexadecimal values.
      $ printf %x 3
      3
      $ printf %x 16
      10
      $ printf %x 1
      1
      This example would use the following values in the configuration file:
      bus='0x03'
      slot='0x10'
      function='0x01'
    3. Open the XML configuration file with the virsh edit command. This example edits a guest named MyGuest.
      # virsh edit MyGuest
    4. The default text editor will open the libvirt configuration file for the guest. Add the new device to the devices section of the XML configuration file.
      <hostdev mode='subsystem' type='pci' managed='yes'>
            <source>
              <address bus='0x03' slot='0x10' function='0x01'/>
            </source>
      </hostdev>
      
    5. Save the configuration.
  9. Restart

    Restart the guest to complete the installation.
    # virsh start MyGuest
The guest should start successfully and detect a new network interface card. This new card is the Virtual Function of the SR-IOV device.