1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2016 Intel Corporation.
7 The i40e PMD (librte_pmd_i40e) provides poll mode driver support for
8 10/25/40 Gbps Intel® Ethernet 700 Series Network Adapters based on
9 the Intel Ethernet Controller X710/XL710/XXV710 and Intel Ethernet
10 Connection X722 (only support part of features).
16 Features of the i40e PMD are:
18 - Multiple queues for TX and RX
19 - Receiver Side Scaling (RSS)
21 - Packet type information
25 - VLAN/QinQ stripping and inserting
29 - Port hardware statistics
31 - Link state information
33 - Mirror on port, VLAN and VSI
34 - Interrupt mode for RX
35 - Scattered and gather for TX and RX
36 - Vector Poll mode driver
41 - IEEE1588/802.1AS timestamping
42 - VF Daemon (VFD) - EXPERIMENTAL
43 - Dynamic Device Personalization (DDP)
44 - Queue region configuration
45 - Virtual Function Port Representors
46 - Malicious Device Drive event catch and notify
52 - Identifying your adapter using `Intel Support
53 <http://www.intel.com/support>`_ and get the latest NVM/FW images.
55 - Follow the DPDK :ref:`Getting Started Guide for Linux <linux_gsg>` to setup the basic DPDK environment.
57 - To get better performance on Intel platforms, please follow the "How to get best performance with NICs on Intel platforms"
58 section of the :ref:`Getting Started Guide for Linux <linux_gsg>`.
60 - Upgrade the NVM/FW version following the `Intel® Ethernet NVM Update Tool Quick Usage Guide for Linux
61 <https://www-ssl.intel.com/content/www/us/en/embedded/products/networking/nvm-update-tool-quick-linux-usage-guide.html>`_ and `Intel® Ethernet NVM Update Tool: Quick Usage Guide for EFI <https://www.intel.com/content/www/us/en/embedded/products/networking/nvm-update-tool-quick-efi-usage-guide.html>`_ if needed.
63 - For information about supported media, please refer to this document: `Intel® Ethernet Controller X710/XXV710/XL710 Feature Support Matrix
64 <http://www.intel.com/content/dam/www/public/us/en/documents/release-notes/xl710-ethernet-controller-feature-matrix.pdf>`_.
68 * Some adapters based on the Intel(R) Ethernet Controller 700 Series only
69 support Intel Ethernet Optics modules. On these adapters, other modules are not
70 supported and will not function.
72 * For connections based on Intel(R) Ethernet Controller 700 Series,
73 support is dependent on your system board. Please see your vendor for details.
75 * In all cases Intel recommends using Intel Ethernet Optics; other modules
76 may function but are not validated by Intel. Contact Intel for supported media types.
78 Recommended Matching List
79 -------------------------
81 It is highly recommended to upgrade the i40e kernel driver and firmware to
82 avoid the compatibility issues with i40e PMD. Here is the suggested matching
83 list which has been tested and verified. The detailed information can refer
84 to chapter Tested Platforms/Tested NICs in release notes.
86 For X710/XL710/XXV710,
88 +--------------+-----------------------+------------------+
89 | DPDK version | Kernel driver version | Firmware version |
90 +==============+=======================+==================+
91 | 20.08 | 2.12.6 | 7.30 |
92 +--------------+-----------------------+------------------+
93 | 20.05 | 2.11.27 | 7.30 |
94 +--------------+-----------------------+------------------+
95 | 20.02 | 2.10.19 | 7.20 |
96 +--------------+-----------------------+------------------+
97 | 19.11 | 2.9.21 | 7.00 |
98 +--------------+-----------------------+------------------+
99 | 19.08 | 2.8.43 | 7.00 |
100 +--------------+-----------------------+------------------+
101 | 19.05 | 2.7.29 | 6.80 |
102 +--------------+-----------------------+------------------+
103 | 19.02 | 2.7.26 | 6.80 |
104 +--------------+-----------------------+------------------+
105 | 18.11 | 2.4.6 | 6.01 |
106 +--------------+-----------------------+------------------+
107 | 18.08 | 2.4.6 | 6.01 |
108 +--------------+-----------------------+------------------+
109 | 18.05 | 2.4.6 | 6.01 |
110 +--------------+-----------------------+------------------+
111 | 18.02 | 2.4.3 | 6.01 |
112 +--------------+-----------------------+------------------+
113 | 17.11 | 2.1.26 | 6.01 |
114 +--------------+-----------------------+------------------+
115 | 17.08 | 2.0.19 | 6.01 |
116 +--------------+-----------------------+------------------+
117 | 17.05 | 1.5.23 | 5.05 |
118 +--------------+-----------------------+------------------+
119 | 17.02 | 1.5.23 | 5.05 |
120 +--------------+-----------------------+------------------+
121 | 16.11 | 1.5.23 | 5.05 |
122 +--------------+-----------------------+------------------+
123 | 16.07 | 1.4.25 | 5.04 |
124 +--------------+-----------------------+------------------+
125 | 16.04 | 1.4.25 | 5.02 |
126 +--------------+-----------------------+------------------+
131 +--------------+-----------------------+------------------+
132 | DPDK version | Kernel driver version | Firmware version |
133 +==============+=======================+==================+
134 | 20.08 | 2.12.6 | 4.11 |
135 +--------------+-----------------------+------------------+
136 | 20.05 | 2.11.27 | 4.11 |
137 +--------------+-----------------------+------------------+
138 | 20.02 | 2.10.19 | 4.11 |
139 +--------------+-----------------------+------------------+
140 | 19.11 | 2.9.21 | 4.10 |
141 +--------------+-----------------------+------------------+
142 | 19.08 | 2.9.21 | 4.10 |
143 +--------------+-----------------------+------------------+
144 | 19.05 | 2.7.29 | 3.33 |
145 +--------------+-----------------------+------------------+
146 | 19.02 | 2.7.26 | 3.33 |
147 +--------------+-----------------------+------------------+
148 | 18.11 | 2.4.6 | 3.33 |
149 +--------------+-----------------------+------------------+
152 Pre-Installation Configuration
153 ------------------------------
158 The following options can be modified in the ``config`` file.
159 Please note that enabling debugging options may affect system performance.
161 - ``CONFIG_RTE_LIBRTE_I40E_PMD`` (default ``y``)
163 Toggle compilation of the ``librte_pmd_i40e`` driver.
165 - ``CONFIG_RTE_LIBRTE_I40E_DEBUG_*`` (default ``n``)
167 Toggle display of generic debugging messages.
169 - ``CONFIG_RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC`` (default ``y``)
171 Toggle bulk allocation for RX.
173 - ``CONFIG_RTE_LIBRTE_I40E_INC_VECTOR`` (default ``n``)
175 Toggle the use of Vector PMD instead of normal RX/TX path.
176 To enable vPMD for RX, bulk allocation for Rx must be allowed.
178 - ``CONFIG_RTE_LIBRTE_I40E_16BYTE_RX_DESC`` (default ``n``)
180 Toggle to use a 16-byte RX descriptor, by default the RX descriptor is 32 byte.
182 - ``CONFIG_RTE_LIBRTE_I40E_QUEUE_NUM_PER_PF`` (default ``64``)
184 Number of queues reserved for PF.
186 - ``CONFIG_RTE_LIBRTE_I40E_QUEUE_NUM_PER_VM`` (default ``4``)
188 Number of queues reserved for each VMDQ Pool.
190 Runtime Config Options
191 ~~~~~~~~~~~~~~~~~~~~~~
193 - ``Reserved number of Queues per VF`` (default ``4``)
195 The number of reserved queue per VF is determined by its host PF. If the
196 PCI address of an i40e PF is aaaa:bb.cc, the number of reserved queues per
197 VF can be configured with EAL parameter like -w aaaa:bb.cc,queue-num-per-vf=n.
198 The value n can be 1, 2, 4, 8 or 16. If no such parameter is configured, the
199 number of reserved queues per VF is 4 by default. If VF request more than
200 reserved queues per VF, PF will able to allocate max to 16 queues after a VF
204 - ``Support multiple driver`` (default ``disable``)
206 There was a multiple driver support issue during use of 700 series Ethernet
207 Adapter with both Linux kernel and DPDK PMD. To fix this issue, ``devargs``
208 parameter ``support-multi-driver`` is introduced, for example::
210 -w 84:00.0,support-multi-driver=1
212 With the above configuration, DPDK PMD will not change global registers, and
213 will switch PF interrupt from IntN to Int0 to avoid interrupt conflict between
214 DPDK and Linux Kernel.
216 - ``Support VF Port Representor`` (default ``not enabled``)
218 The i40e PF PMD supports the creation of VF port representors for the control
219 and monitoring of i40e virtual function devices. Each port representor
220 corresponds to a single virtual function of that device. Using the ``devargs``
221 option ``representor`` the user can specify which virtual functions to create
222 port representors for on initialization of the PF PMD by passing the VF IDs of
223 the VFs which are required.::
225 -w DBDF,representor=[0,1,4]
227 Currently hot-plugging of representor ports is not supported so all required
228 representors must be specified on the creation of the PF.
230 - ``Use latest supported vector`` (default ``disable``)
232 Latest supported vector path may not always get the best perf so vector path was
233 recommended to use only on later platform. But users may want the latest vector path
234 since it can get better perf in some real work loading cases. So ``devargs`` param
235 ``use-latest-supported-vec`` is introduced, for example::
237 -w 84:00.0,use-latest-supported-vec=1
239 - ``Enable validation for VF message`` (default ``not enabled``)
241 The PF counts messages from each VF. If in any period of seconds the message
242 statistic from a VF exceeds maximal limitation, the PF will ignore any new message
243 from that VF for some seconds.
244 Format -- "maximal-message@period-seconds:ignore-seconds"
247 -w 84:00.0,vf_msg_cfg=80@120:180
249 Vector RX Pre-conditions
250 ~~~~~~~~~~~~~~~~~~~~~~~~
251 For Vector RX it is assumed that the number of descriptor rings will be a power
252 of 2. With this pre-condition, the ring pointer can easily scroll back to the
253 head after hitting the tail without a conditional check. In addition Vector RX
254 can use this assumption to do a bit mask using ``ring_size - 1``.
256 Driver compilation and testing
257 ------------------------------
259 Refer to the document :ref:`compiling and testing a PMD for a NIC <pmd_build_and_test>`
263 SR-IOV: Prerequisites and sample Application Notes
264 --------------------------------------------------
266 #. Load the kernel module:
268 .. code-block:: console
272 Check the output in dmesg:
274 .. code-block:: console
276 i40e 0000:83:00.1 ens802f0: renamed from eth0
278 #. Bring up the PF ports:
280 .. code-block:: console
284 #. Create VF device(s):
286 Echo the number of VFs to be created into the ``sriov_numvfs`` sysfs entry
291 .. code-block:: console
293 echo 2 > /sys/devices/pci0000:00/0000:00:03.0/0000:81:00.0/sriov_numvfs
296 #. Assign VF MAC address:
298 Assign MAC address to the VF using iproute2 utility. The syntax is:
300 .. code-block:: console
302 ip link set <PF netdev id> vf <VF id> mac <macaddr>
306 .. code-block:: console
308 ip link set ens802f0 vf 0 mac a0:b0:c0:d0:e0:f0
310 #. Assign VF to VM, and bring up the VM.
311 Please see the documentation for the *I40E/IXGBE/IGB Virtual Function Driver*.
315 Follow instructions available in the document
316 :ref:`compiling and testing a PMD for a NIC <pmd_build_and_test>`
321 .. code-block:: console
324 EAL: PCI device 0000:83:00.0 on NUMA socket 1
325 EAL: probe driver: 8086:1572 rte_i40e_pmd
326 EAL: PCI memory mapped at 0x7f7f80000000
327 EAL: PCI memory mapped at 0x7f7f80800000
328 PMD: eth_i40e_dev_init(): FW 5.0 API 1.5 NVM 05.00.02 eetrack 8000208a
329 Interactive-mode selected
330 Configuring Port 0 (socket 0)
333 PMD: i40e_dev_rx_queue_setup(): Rx Burst Bulk Alloc Preconditions are
334 satisfied.Rx Burst Bulk Alloc function will be used on port=0, queue=0.
337 Port 0: 68:05:CA:26:85:84
338 Checking link statuses...
339 Port 0 Link Up - speed 10000 Mbps - full-duplex
345 Sample Application Notes
346 ------------------------
351 Vlan filter only works when Promiscuous mode is off.
353 To start ``testpmd``, and add vlan 10 to port 0:
355 .. code-block:: console
357 ./app/testpmd -l 0-15 -n 4 -- -i --forward-mode=mac
360 testpmd> set promisc 0 off
361 testpmd> rx_vlan add 10 0
367 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
368 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set per flow type and the flexible payload.
370 The default input set of each flow type is::
372 ipv4-other : src_ip_address, dst_ip_address
373 ipv4-frag : src_ip_address, dst_ip_address
374 ipv4-tcp : src_ip_address, dst_ip_address, src_port, dst_port
375 ipv4-udp : src_ip_address, dst_ip_address, src_port, dst_port
376 ipv4-sctp : src_ip_address, dst_ip_address, src_port, dst_port,
378 ipv6-other : src_ip_address, dst_ip_address
379 ipv6-frag : src_ip_address, dst_ip_address
380 ipv6-tcp : src_ip_address, dst_ip_address, src_port, dst_port
381 ipv6-udp : src_ip_address, dst_ip_address, src_port, dst_port
382 ipv6-sctp : src_ip_address, dst_ip_address, src_port, dst_port,
384 l2_payload : ether_type
386 The flex payload is selected from offset 0 to 15 of packet's payload by default, while it is masked out from matching.
388 Start ``testpmd`` with ``--disable-rss`` and ``--pkt-filter-mode=perfect``:
390 .. code-block:: console
392 ./app/testpmd -l 0-15 -n 4 -- -i --disable-rss --pkt-filter-mode=perfect \
393 --rxq=8 --txq=8 --nb-cores=8 --nb-ports=1
395 Add a rule to direct ``ipv4-udp`` packet whose ``dst_ip=2.2.2.5, src_ip=2.2.2.3, src_port=32, dst_port=32`` to queue 1:
397 .. code-block:: console
399 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp \
400 src 2.2.2.3 32 dst 2.2.2.5 32 vlan 0 flexbytes () \
401 fwd pf queue 1 fd_id 1
403 Check the flow director status:
405 .. code-block:: console
407 testpmd> show port fdir 0
409 ######################## FDIR infos for port 0 ####################
411 SUPPORTED FLOW TYPE: ipv4-frag ipv4-tcp ipv4-udp ipv4-sctp ipv4-other
412 ipv6-frag ipv6-tcp ipv6-udp ipv6-sctp ipv6-other
415 max_len: 16 payload_limit: 480
416 payload_unit: 2 payload_seg: 3
417 bitmask_unit: 2 bitmask_num: 2
420 src_ipv4: 0x00000000,
421 dst_ipv4: 0x00000000,
424 src_ipv6: 0x00000000,0x00000000,0x00000000,0x00000000,
425 dst_ipv6: 0x00000000,0x00000000,0x00000000,0x00000000
426 FLEX PAYLOAD SRC OFFSET:
427 L2_PAYLOAD: 0 1 2 3 4 5 6 ...
428 L3_PAYLOAD: 0 1 2 3 4 5 6 ...
429 L4_PAYLOAD: 0 1 2 3 4 5 6 ...
431 ipv4-udp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
432 ipv4-tcp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
433 ipv4-sctp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
434 ipv4-other: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
435 ipv4-frag: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
436 ipv6-udp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
437 ipv6-tcp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
438 ipv6-sctp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
439 ipv6-other: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
440 ipv6-frag: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
441 l2_payload: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
442 guarant_count: 1 best_count: 0
443 guarant_space: 512 best_space: 7168
450 Delete all flow director rules on a port:
452 .. code-block:: console
454 testpmd> flush_flow_director 0
459 The Intel® Ethernet 700 Series support a feature called
462 A Virtual Ethernet Bridge (VEB) is an IEEE Edge Virtual Bridging (EVB) term
463 for functionality that allows local switching between virtual endpoints within
464 a physical endpoint and also with an external bridge/network.
466 A "Floating" VEB doesn't have an uplink connection to the outside world so all
467 switching is done internally and remains within the host. As such, this
468 feature provides security benefits.
470 In addition, a Floating VEB overcomes a limitation of normal VEBs where they
471 cannot forward packets when the physical link is down. Floating VEBs don't need
472 to connect to the NIC port so they can still forward traffic from VF to VF
473 even when the physical link is down.
475 Therefore, with this feature enabled VFs can be limited to communicating with
476 each other but not an outside network, and they can do so even when there is
477 no physical uplink on the associated NIC port.
479 To enable this feature, the user should pass a ``devargs`` parameter to the
482 -w 84:00.0,enable_floating_veb=1
484 In this configuration the PMD will use the floating VEB feature for all the
485 VFs created by this PF device.
487 Alternatively, the user can specify which VFs need to connect to this floating
488 VEB using the ``floating_veb_list`` argument::
490 -w 84:00.0,enable_floating_veb=1,floating_veb_list=1;3-4
492 In this example ``VF1``, ``VF3`` and ``VF4`` connect to the floating VEB,
493 while other VFs connect to the normal VEB.
495 The current implementation only supports one floating VEB and one regular
496 VEB. VFs can connect to a floating VEB or a regular VEB according to the
497 configuration passed on the EAL command line.
499 The floating VEB functionality requires a NIC firmware version of 5.0
502 Dynamic Device Personalization (DDP)
503 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
505 The Intel® Ethernet 700 Series except for the Intel Ethernet Connection
506 X722 support a feature called "Dynamic Device Personalization (DDP)",
507 which is used to configure hardware by downloading a profile to support
508 protocols/filters which are not supported by default. The DDP
509 functionality requires a NIC firmware version of 6.0 or greater.
511 Current implementation supports GTP-C/GTP-U/PPPoE/PPPoL2TP/ESP,
512 steering can be used with rte_flow API.
514 GTPv1 package is released, and it can be downloaded from
515 https://downloadcenter.intel.com/download/27587.
517 PPPoE package is released, and it can be downloaded from
518 https://downloadcenter.intel.com/download/28040.
520 ESP-AH package is released, and it can be downloaded from
521 https://downloadcenter.intel.com/download/29446.
523 Load a profile which supports GTP and store backup profile:
525 .. code-block:: console
527 testpmd> ddp add 0 ./gtp.pkgo,./backup.pkgo
529 Delete a GTP profile and restore backup profile:
531 .. code-block:: console
533 testpmd> ddp del 0 ./backup.pkgo
535 Get loaded DDP package info list:
537 .. code-block:: console
539 testpmd> ddp get list 0
541 Display information about a GTP profile:
543 .. code-block:: console
545 testpmd> ddp get info ./gtp.pkgo
547 Input set configuration
548 ~~~~~~~~~~~~~~~~~~~~~~~
549 Input set for any PCTYPE can be configured with user defined configuration,
550 For example, to use only 48bit prefix for IPv6 src address for IPv6 TCP RSS:
552 .. code-block:: console
554 testpmd> port config 0 pctype 43 hash_inset clear all
555 testpmd> port config 0 pctype 43 hash_inset set field 13
556 testpmd> port config 0 pctype 43 hash_inset set field 14
557 testpmd> port config 0 pctype 43 hash_inset set field 15
559 Queue region configuration
560 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
561 The Intel® Ethernet 700 Series supports a feature of queue regions
562 configuration for RSS in the PF, so that different traffic classes or
563 different packet classification types can be separated to different
564 queues in different queue regions. There is an API for configuration
565 of queue regions in RSS with a command line. It can parse the parameters
566 of the region index, queue number, queue start index, user priority, traffic
567 classes and so on. Depending on commands from the command line, it will call
568 i40e private APIs and start the process of setting or flushing the queue
569 region configuration. As this feature is specific for i40e only private
570 APIs are used. These new ``test_pmd`` commands are as shown below. For
571 details please refer to :doc:`../testpmd_app_ug/index`.
573 .. code-block:: console
575 testpmd> set port (port_id) queue-region region_id (value) \
576 queue_start_index (value) queue_num (value)
577 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
578 testpmd> set port (port_id) queue-region UP (value) region_id (value)
579 testpmd> set port (port_id) queue-region flush (on|off)
580 testpmd> show port (port_id) queue-region
587 RSS Flow supports to set hash input set, hash function, enable hash
588 and configure queue region.
590 Configure queue region as queue 0, 1, 2, 3.
592 .. code-block:: console
594 testpmd> flow create 0 ingress pattern end actions rss types end \
595 queues 0 1 2 3 end / end
597 Enable hash and set input set for ipv4-tcp.
599 .. code-block:: console
601 testpmd> flow create 0 ingress pattern eth / ipv4 / tcp / end \
602 actions rss types ipv4-tcp l3-src-only end queues end / end
604 Set symmetric hash enable for flow type ipv4-tcp.
606 .. code-block:: console
608 testpmd> flow create 0 ingress pattern eth / ipv4 / tcp / end \
609 actions rss types ipv4-tcp end queues end func symmetric_toeplitz / end
611 Set hash function as simple xor.
613 .. code-block:: console
615 testpmd> flow create 0 ingress pattern end actions rss types end \
616 queues end func simple_xor / end
618 Limitations or Known issues
619 ---------------------------
621 MPLS packet classification
622 ~~~~~~~~~~~~~~~~~~~~~~~~~~
624 For firmware versions prior to 5.0, MPLS packets are not recognized by the NIC.
625 The L2 Payload flow type in flow director can be used to classify MPLS packet
626 by using a command in testpmd like:
628 testpmd> flow_director_filter 0 mode IP add flow l2_payload ether \
629 0x8847 flexbytes () fwd pf queue <N> fd_id <M>
631 With the NIC firmware version 5.0 or greater, some limited MPLS support
632 is added: Native MPLS (MPLS in Ethernet) skip is implemented, while no
633 new packet type, no classification or offload are possible. With this change,
634 L2 Payload flow type in flow director cannot be used to classify MPLS packet
635 as with previous firmware versions. Meanwhile, the Ethertype filter can be
636 used to classify MPLS packet by using a command in testpmd like:
638 testpmd> ethertype_filter 0 add mac_ignr 00:00:00:00:00:00 ethertype \
641 16 Byte RX Descriptor setting on DPDK VF
642 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
644 Currently the VF's RX descriptor mode is decided by PF. There's no PF-VF
645 interface for VF to request the RX descriptor mode, also no interface to notify
646 VF its own RX descriptor mode.
647 For all available versions of the i40e driver, these drivers don't support 16
648 byte RX descriptor. If the Linux i40e kernel driver is used as host driver,
649 while DPDK i40e PMD is used as the VF driver, DPDK cannot choose 16 byte receive
650 descriptor. The reason is that the RX descriptor is already set to 32 byte by
651 the i40e kernel driver. That is to say, user should keep
652 ``CONFIG_RTE_LIBRTE_I40E_16BYTE_RX_DESC=n`` in config file.
653 In the future, if the Linux i40e driver supports 16 byte RX descriptor, user
654 should make sure the DPDK VF uses the same RX descriptor mode, 16 byte or 32
655 byte, as the PF driver.
657 The same rule for DPDK PF + DPDK VF. The PF and VF should use the same RX
658 descriptor mode. Or the VF RX will not work.
660 Receive packets with Ethertype 0x88A8
661 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
663 Due to the FW limitation, PF can receive packets with Ethertype 0x88A8
664 only when floating VEB is disabled.
666 Incorrect Rx statistics when packet is oversize
667 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
669 When a packet is over maximum frame size, the packet is dropped.
670 However, the Rx statistics, when calling `rte_eth_stats_get` incorrectly
671 shows it as received.
673 VF & TC max bandwidth setting
674 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
676 The per VF max bandwidth and per TC max bandwidth cannot be enabled in parallel.
677 The behavior is different when handling per VF and per TC max bandwidth setting.
678 When enabling per VF max bandwidth, SW will check if per TC max bandwidth is
679 enabled. If so, return failure.
680 When enabling per TC max bandwidth, SW will check if per VF max bandwidth
681 is enabled. If so, disable per VF max bandwidth and continue with per TC max
684 TC TX scheduling mode setting
685 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
687 There are 2 TX scheduling modes for TCs, round robin and strict priority mode.
688 If a TC is set to strict priority mode, it can consume unlimited bandwidth.
689 It means if APP has set the max bandwidth for that TC, it comes to no
691 It's suggested to set the strict priority mode for a TC that is latency
692 sensitive but no consuming much bandwidth.
694 VF performance is impacted by PCI extended tag setting
695 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
697 To reach maximum NIC performance in the VF the PCI extended tag must be
698 enabled. The DPDK i40e PF driver will set this feature during initialization,
699 but the kernel PF driver does not. So when running traffic on a VF which is
700 managed by the kernel PF driver, a significant NIC performance downgrade has
701 been observed (for 64 byte packets, there is about 25% line-rate downgrade for
702 a 25GbE device and about 35% for a 40GbE device).
704 For kernel version >= 4.11, the kernel's PCI driver will enable the extended
705 tag if it detects that the device supports it. So by default, this is not an
706 issue. For kernels <= 4.11 or when the PCI extended tag is disabled it can be
707 enabled using the steps below.
709 #. Get the current value of the PCI configure register::
711 setpci -s <XX:XX.X> a8.w
715 value = value | 0x100
717 #. Set the PCI configure register with new value::
719 setpci -s <XX:XX.X> a8.w=<value>
724 The VF vlan strip function is only supported in the i40e kernel driver >= 2.1.26.
729 DCB works only when RSS is enabled.
731 Global configuration warning
732 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
734 I40E PMD will set some global registers to enable some function or set some
735 configure. Then when using different ports of the same NIC with Linux kernel
736 and DPDK, the port with Linux kernel will be impacted by the port with DPDK.
737 For example, register I40E_GL_SWT_L2TAGCTRL is used to control L2 tag, i40e
738 PMD uses I40E_GL_SWT_L2TAGCTRL to set vlan TPID. If setting TPID in port A
739 with DPDK, then the configuration will also impact port B in the NIC with
740 kernel driver, which don't want to use the TPID.
741 So PMD reports warning to clarify what is changed by writing global register.
746 When programming cloud filters for IPv4/6_UDP/TCP/SCTP with SRC port only or DST port only,
747 it will make any cloud filter using inner_vlan or tunnel key invalid. Default configuration will be
748 recovered only by NIC core reset.
750 High Performance of Small Packets on 40GbE NIC
751 ----------------------------------------------
753 As there might be firmware fixes for performance enhancement in latest version
754 of firmware image, the firmware update might be needed for getting high performance.
755 Check the Intel support website for the latest firmware updates.
756 Users should consult the release notes specific to a DPDK release to identify
757 the validated firmware version for a NIC using the i40e driver.
759 Use 16 Bytes RX Descriptor Size
760 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
762 As i40e PMD supports both 16 and 32 bytes RX descriptor sizes, and 16 bytes size can provide helps to high performance of small packets.
763 Configuration of ``CONFIG_RTE_LIBRTE_I40E_16BYTE_RX_DESC`` in config files can be changed to use 16 bytes size RX descriptors.
765 Input set requirement of each pctype for FDIR
766 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
768 Each PCTYPE can only have one specific FDIR input set at one time.
769 For example, if creating 2 rte_flow rules with different input set for one PCTYPE,
770 it will fail and return the info "Conflict with the first rule's input set",
771 which means the current rule's input set conflicts with the first rule's.
772 Remove the first rule if want to change the input set of the PCTYPE.
774 Example of getting best performance with l3fwd example
775 ------------------------------------------------------
777 The following is an example of running the DPDK ``l3fwd`` sample application to get high performance with a
778 server with Intel Xeon processors and Intel Ethernet CNA XL710.
780 The example scenario is to get best performance with two Intel Ethernet CNA XL710 40GbE ports.
781 See :numref:`figure_intel_perf_test_setup` for the performance test setup.
783 .. _figure_intel_perf_test_setup:
785 .. figure:: img/intel_perf_test_setup.*
787 Performance Test Setup
790 1. Add two Intel Ethernet CNA XL710 to the platform, and use one port per card to get best performance.
791 The reason for using two NICs is to overcome a PCIe v3.0 limitation since it cannot provide 80GbE bandwidth
792 for two 40GbE ports, but two different PCIe v3.0 x8 slot can.
793 Refer to the sample NICs output above, then we can select ``82:00.0`` and ``85:00.0`` as test ports::
795 82:00.0 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
796 85:00.0 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
798 2. Connect the ports to the traffic generator. For high speed testing, it's best to use a hardware traffic generator.
800 3. Check the PCI devices numa node (socket id) and get the cores number on the exact socket id.
801 In this case, ``82:00.0`` and ``85:00.0`` are both in socket 1, and the cores on socket 1 in the referenced platform
803 Note: Don't use 2 logical cores on the same core (e.g core18 has 2 logical cores, core18 and core54), instead, use 2 logical
804 cores from different cores (e.g core18 and core19).
806 4. Bind these two ports to igb_uio.
808 5. As to Intel Ethernet CNA XL710 40GbE port, we need at least two queue pairs to achieve best performance, then two queues per port
809 will be required, and each queue pair will need a dedicated CPU core for receiving/transmitting packets.
811 6. The DPDK sample application ``l3fwd`` will be used for performance testing, with using two ports for bi-directional forwarding.
812 Compile the ``l3fwd sample`` with the default lpm mode.
814 7. The command line of running l3fwd would be something like the following::
816 ./l3fwd -l 18-21 -n 4 -w 82:00.0 -w 85:00.0 \
817 -- -p 0x3 --config '(0,0,18),(0,1,19),(1,0,20),(1,1,21)'
819 This means that the application uses core 18 for port 0, queue pair 0 forwarding, core 19 for port 0, queue pair 1 forwarding,
820 core 20 for port 1, queue pair 0 forwarding, and core 21 for port 1, queue pair 1 forwarding.
822 8. Configure the traffic at a traffic generator.
824 * Start creating a stream on packet generator.
826 * Set the Ethernet II type to 0x0800.
828 Tx bytes affected by the link status change
829 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
831 For firmware versions prior to 6.01 for X710 series and 3.33 for X722 series, the tx_bytes statistics data is affected by
832 the link down event. Each time the link status changes to down, the tx_bytes decreases 110 bytes.