iavf FreeBSD* Base Driver for Adaptive Virtual Functions ======================================================== March 14, 2022 Contents ======== - Overview - Identifying Your Adapter - The VF Driver - Building and Installation - Configuration and Tuning - Known Issues/Troubleshooting Overview ======== This file describes the FreeBSD* driver for Intel(R) Ethernet. This driver has been developed for use with all community-supported versions of FreeBSD. For questions related to hardware requirements, refer to the documentation supplied with your Intel Ethernet Adapter. All hardware requirements listed apply to use with FreeBSD. The associated Physical Function (PF) drivers for this VF driver are: - ice - ixl Identifying Your Adapter ======================== This driver is compatible with virtual functions bound to devices based on the following: * Intel(R) Ethernet Controller E810-C * Intel(R) Ethernet Controller E810-XXV * Intel(R) Ethernet Connection E822-C * Intel(R) Ethernet Connection E822-L * Intel(R) Ethernet Connection E823-C * Intel(R) Ethernet Connection E823-L * Intel(R) Ethernet Controller X710 * Intel(R) Ethernet Controller XL710 * Intel(R) Ethernet Network Connection X722 * Intel(R) Ethernet Controller XXV710 * Intel(R) Ethernet Controller V710 For information on how to identify your adapter, and for the latest Intel network drivers, refer to the Intel Support website: http://www.intel.com/support The VF Driver ============= The VF driver is normally used in a virtualized environment where a host driver manages SRIOV, and provides a VF device to the guest. In the FreeBSD guest, the iavf driver would be loaded and will function using the VF device assigned to it. The VF driver provides most of the same functionality as the core driver, but is actually a subordinate to the host. Access to many controls is accomplished by a request to the host via what is called the "Admin queue." These are startup and initialization events, however; once in operation, the device is self-contained and should achieve near native performance. Some notable limitations of the VF environment: * The PF can configure the VF to allow promiscuous mode, when using iovctl. * Media info is not available from the PF, so it will always appear as auto. Adaptive Virtual Function ------------------------- Adaptive Virtual Function (AVF) allows the virtual function driver, or VF, to adapt to changing feature sets of the physical function driver (PF) with which it is associated. This allows system administrators to update a PF without having to update all the VFs associated with it. All AVFs have a single common device ID and branding string. AVFs have a minimum set of features known as "base mode," but may provide additional features depending on what features are available in the PF with which the AVF is associated. The following are base mode features: - 4 Queue Pairs (QP) and associated Configuration Status Registers (CSRs) for Tx/Rx - iavf descriptors and ring format - Descriptor write-back completion - 1 control queue, with iavf descriptors, CSRs and ring format - 5 MSI-X interrupt vectors and corresponding iavf CSRs - 1 Interrupt Throttle Rate (ITR) index - 1 Virtual Station Interface (VSI) per VF - 1 Traffic Class (TC), TC0 - Receive Side Scaling (RSS) with 64 entry indirection table and key, configured through the PF - 1 unicast MAC address reserved per VF - 16 MAC address filters for each VF - Stateless offloads - non-tunneled checksums - AVF device ID - HW mailbox is used for VF to PF communications (including on Windows) Building and Installation ========================= NOTE: This driver package is to be used only as a standalone archive and the user should not attempt to incorporate it into the kernel source tree. In the instructions below, x.x.x is the driver version as indicated in the name of the driver tar file. 1. Move the base driver tar file to the directory of your choice. For example, use /home/username/iavf or /usr/local/src/iavf. 2. Untar/unzip the archive: # tar xzf iavf-x.x.x.tar.gz This will create the iavf-x.x.x directory. 3. To install man page: # cd iavf-x.x.x # gzip -c iavf.4 > /usr/share/man/man4/iavf.4.gz 4. To load the driver onto a running system: # cd iavf-x.x.x/src # make # kldload ./if_iavf.ko To install the driver without using iflib: # cd iavf-x.x.x/src # make legacy # kldload ./if_iavf.ko 5. To assign an IP address to the interface, enter the following, where X is the interface number for the device: # ifconfig iavfX 6. Verify that the interface works. Enter the following, where is the IP address for another machine on the same subnet as the interface that is being tested: # ping 7. If you want the driver to load automatically when the system is booted: # cd iavf-x.x.x/src # make # make install Edit /boot/loader.conf, and add the following line: if_iavf_load="YES" Edit /etc/rc.conf, and create the appropriate ifconfig_iavfX entry: ifconfig_iavfX="" Example usage: ifconfig_iavf0="inet 192.168.10.1 netmask 255.255.255.0" NOTE: For assistance, see the ifconfig man page. Configuration and Tuning ======================== Important System Configuration Changes -------------------------------------- - Change the file /etc/sysctl.conf, and add the line: hw.intr_storm_threshold: 0 (the default is 1000) - Best throughput results are seen with a large MTU; use 9706 if possible. The default number of descriptors per ring is 1024. Increasing this may improve performance, depending on your use case. Configuring for iflib --------------------- Iflib is a common framework for network interface drivers for FreeBSD that uses a shared set of sysctl names. The iflib driver works best in FreeBSD 11.3 and later. See the iflib man page for more information. Jumbo Frames ------------ Jumbo Frames support is enabled by changing the Maximum Transmission Unit (MTU) to a value larger than the default value of 1500. Use the ifconfig command to increase the MTU size. For example, enter the following where X is the interface number: # ifconfig iavfX mtu 9000 To confirm an interface's MTU value, use the ifconfig command. To confirm the MTU used between two specific devices, use: # route get NOTE: The maximum MTU setting for jumbo frames is 9706. This corresponds to the maximum jumbo frame size of 9728 bytes. NOTE: This driver will attempt to use multiple page sized buffers to receive each jumbo packet. This should help to avoid buffer starvation issues when allocating receive packets. NOTE: Packet loss may have a greater impact on throughput when you use jumbo frames. If you observe a drop in performance after enabling jumbo frames, enabling flow control may mitigate the issue. VLANS ----- To create a new VLAN interface: # ifconfig create To associate the VLAN interface with a physical interface and assign a VLAN ID, IP address, and netmask: # ifconfig netmask vlan vlandev Example: # ifconfig vlan10 10.0.0.1 netmask 255.255.255.0 vlan 10 vlandev iavf0 In this example, all packets will be marked on egress with 802.1Q VLAN tags, specifying a VLAN ID of 10. To remove a VLAN interface: # ifconfig destroy Checksum Offload ---------------- Checksum offloading supports both TCP and UDP packets and is supported for both transmit and receive. Checksum offloading can be enabled or disabled using ifconfig. To enable checksum offloading: # ifconfig iavfX rxcsum rxcsum6 # ifconfig iavfX txcsum txcsum6 To disable checksum offloading: # ifconfig iavfX -rxcsum -rxcsum6 # ifconfig iavfX -txcsum -txcsum6 To confirm the current setting: # ifconfig iavfX Look for the presence or absence of the following line: options=3 See the ifconfig man page for further information. TSO --- TSO (TCP Segmentation Offload) supports both IPv4 and IPv6. TSO can be disabled and enabled using the ifconfig utility or sysctl. NOTE: TSO requires Tx checksum, if Tx checksum is disabled, TSO will also be disabled. To enable/disable TSO in the stack: # sysctl net.inet.tcp.tso=0 (or 1 to enable it) Doing this disables/enables TSO in the stack and affects all installed adapters. To disable BOTH TSO IPv4 and IPv6, where X is the number of the interface in use: # ifconfig iavfX -tso To enable BOTH TSO IPv4 and IPv6: # ifconfig iavfX tso You can also enable/disable IPv4 TSO or IPv6 TSO individually. Simply replace tso|-tso in the above command with tso4 or tso6. For example, to disable TSO IPv4: # ifconfig iavfX -tso4 To disable TSO IPv6: # ifconfig iavfX -tso6 LRO --- LRO (Large Receive Offload) may provide Rx performance improvement. However, it is incompatible with packet-forwarding workloads. You should carefully evaluate the environment and enable LRO when possible. To enable: # ifconfig iavfX lro It can be disabled by using: # ifconfig iavfX -lro Rx and Tx Descriptor Rings -------------------------- Allows you to set the Rx and Tx descriptor rings independently. The tunables are: hw.iavf.rx_ring_size hw.iavf.tx_ring_size The valid range is 32-4096 in increments of 32. Use kenv to configure the descriptor rings. Changes will take effect on the next driver reload. For example: # kenv hw.iavf.rx_ring_size=1024 # kenv hw.iavf.rx_ring_size=1280 NOTE: When you are handling a large number of connections in a VF, we recommend setting the number of Rx descriptors to 1024 or above. You can verify the descriptor ring size by using the following sysctls: # sysctl dev.iavf..rx_ring_size # sysctl dev.iavf..tx_ring_size If you are using iflib, use the following sysctls instead: # sysctl dev.iavf..iflib.override_nrxds # sysctl dev.iavf..iflib.override_ntxds Link-Level Flow Control (LFC) ----------------------------- The VF driver does not have access to flow control. It must be managed from the host side. Known Issues/Troubleshooting ============================ Driver Buffer Overflow Fix -------------------------- The fix to resolve CVE-2016-8105, referenced in Intel SA-00069 , is included in this and future versions of the driver. Network Memory Buffer Allocation -------------------------------- FreeBSD may have a low number of network memory buffers (mbufs) by default. If your mbuf value is too low, it may cause the driver to fail to initialize and/or cause the system to become unresponsive. You can check to see if the system is mbuf-starved by running 'netstat -m'. Increase the number of mbufs by editing the lines below in /etc/sysctl.conf: kern.ipc.nmbclusters kern.ipc.nmbjumbop kern.ipc.nmbjumbo9 kern.ipc.nmbjumbo16 kern.ipc.nmbufs The amount of memory that you allocate is system specific, and may require some trial and error. Also, increasing the following in /etc/sysctl.conf could help increase network performance: kern.ipc.maxsockbuf net.inet.tcp.sendspace net.inet.tcp.recvspace net.inet.udp.maxdgram net.inet.udp.recvspace UDP Stress Test Dropped Packet Issue ------------------------------------ Under small packet UDP stress with the iavf driver, the system may drop UDP packets due to socket buffers being full. Setting the driver Intel Ethernet Flow Control variables to the minimum may resolve the issue. You may also try increasing the kernel's default buffer sizes by changing the values in /proc/sys/net/core/rmem_default and rmem_max Disable LRO when routing/bridging --------------------------------- LRO must be turned off when forwarding traffic. Throughput lower than expected ------------------------------ In FreeBSD 11.3, you may observe lower than expected throughput. This is due to an underlying OS limitation in FreeBSD 11.3. Using FreeBSD 12.0 or newer should resolve the issue. If your Rx throughput is lower than expected in FreeBSD 11.3 or 12.1, you can also adjust the iflib sysctl variable 'rx_budget.' We have seen performance benefits by increasing that value to at least 85. For example: # sysctl dev.iavf.0.iflib.rx_budget=85 Support ======= For general information, go to the Intel support website at: http://www.intel.com/support/ If an issue is identified with the released source code on a supported kernel with a supported adapter, email the specific information related to the issue to freebsd@intel.com Copyright(c) 2018 - 2022 Intel Corporation. Trademarks ========== Intel is a trademark or registered trademark of Intel Corporation or its subsidiaries in the United States and/or other countries. * Other names and brands may be claimed as the property of others.