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
50 - Identifying your adapter using `Intel Support
51 <http://www.intel.com/support>`_ and get the latest NVM/FW images.
53 - Follow the DPDK :ref:`Getting Started Guide for Linux <linux_gsg>` to setup the basic DPDK environment.
55 - To get better performance on Intel platforms, please follow the "How to get best performance with NICs on Intel platforms"
56 section of the :ref:`Getting Started Guide for Linux <linux_gsg>`.
58 - Upgrade the NVM/FW version following the `Intel® Ethernet NVM Update Tool Quick Usage Guide for Linux
59 <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.
61 Recommended Matching List
62 -------------------------
64 It is highly recommended to upgrade the i40e kernel driver and firmware to
65 avoid the compatibility issues with i40e PMD. Here is the suggested matching
66 list which has been tested and verified. The detailed information can refer
67 to chapter Tested Platforms/Tested NICs in release notes.
69 +--------------+-----------------------+------------------+
70 | DPDK version | Kernel driver version | Firmware version |
71 +==============+=======================+==================+
72 | 19.08 | 2.9.21 | 7.00 |
73 +--------------+-----------------------+------------------+
74 | 19.05 | 2.7.29 | 6.80 |
75 +--------------+-----------------------+------------------+
76 | 19.02 | 2.7.26 | 6.80 |
77 +--------------+-----------------------+------------------+
78 | 18.11 | 2.4.6 | 6.01 |
79 +--------------+-----------------------+------------------+
80 | 18.08 | 2.4.6 | 6.01 |
81 +--------------+-----------------------+------------------+
82 | 18.05 | 2.4.6 | 6.01 |
83 +--------------+-----------------------+------------------+
84 | 18.02 | 2.4.3 | 6.01 |
85 +--------------+-----------------------+------------------+
86 | 17.11 | 2.1.26 | 6.01 |
87 +--------------+-----------------------+------------------+
88 | 17.08 | 2.0.19 | 6.01 |
89 +--------------+-----------------------+------------------+
90 | 17.05 | 1.5.23 | 5.05 |
91 +--------------+-----------------------+------------------+
92 | 17.02 | 1.5.23 | 5.05 |
93 +--------------+-----------------------+------------------+
94 | 16.11 | 1.5.23 | 5.05 |
95 +--------------+-----------------------+------------------+
96 | 16.07 | 1.4.25 | 5.04 |
97 +--------------+-----------------------+------------------+
98 | 16.04 | 1.4.25 | 5.02 |
99 +--------------+-----------------------+------------------+
101 Pre-Installation Configuration
102 ------------------------------
107 The following options can be modified in the ``config`` file.
108 Please note that enabling debugging options may affect system performance.
110 - ``CONFIG_RTE_LIBRTE_I40E_PMD`` (default ``y``)
112 Toggle compilation of the ``librte_pmd_i40e`` driver.
114 - ``CONFIG_RTE_LIBRTE_I40E_DEBUG_*`` (default ``n``)
116 Toggle display of generic debugging messages.
118 - ``CONFIG_RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC`` (default ``y``)
120 Toggle bulk allocation for RX.
122 - ``CONFIG_RTE_LIBRTE_I40E_INC_VECTOR`` (default ``n``)
124 Toggle the use of Vector PMD instead of normal RX/TX path.
125 To enable vPMD for RX, bulk allocation for Rx must be allowed.
127 - ``CONFIG_RTE_LIBRTE_I40E_16BYTE_RX_DESC`` (default ``n``)
129 Toggle to use a 16-byte RX descriptor, by default the RX descriptor is 32 byte.
131 - ``CONFIG_RTE_LIBRTE_I40E_QUEUE_NUM_PER_PF`` (default ``64``)
133 Number of queues reserved for PF.
135 - ``CONFIG_RTE_LIBRTE_I40E_QUEUE_NUM_PER_VM`` (default ``4``)
137 Number of queues reserved for each VMDQ Pool.
139 Runtime Config Options
140 ~~~~~~~~~~~~~~~~~~~~~~
142 - ``Reserved number of Queues per VF`` (default ``4``)
144 The number of reserved queue per VF is determined by its host PF. If the
145 PCI address of an i40e PF is aaaa:bb.cc, the number of reserved queues per
146 VF can be configured with EAL parameter like -w aaaa:bb.cc,queue-num-per-vf=n.
147 The value n can be 1, 2, 4, 8 or 16. If no such parameter is configured, the
148 number of reserved queues per VF is 4 by default. If VF request more than
149 reserved queues per VF, PF will able to allocate max to 16 queues after a VF
153 - ``Support multiple driver`` (default ``disable``)
155 There was a multiple driver support issue during use of 700 series Ethernet
156 Adapter with both Linux kernel and DPDK PMD. To fix this issue, ``devargs``
157 parameter ``support-multi-driver`` is introduced, for example::
159 -w 84:00.0,support-multi-driver=1
161 With the above configuration, DPDK PMD will not change global registers, and
162 will switch PF interrupt from IntN to Int0 to avoid interrupt conflict between
163 DPDK and Linux Kernel.
165 - ``Support VF Port Representor`` (default ``not enabled``)
167 The i40e PF PMD supports the creation of VF port representors for the control
168 and monitoring of i40e virtual function devices. Each port representor
169 corresponds to a single virtual function of that device. Using the ``devargs``
170 option ``representor`` the user can specify which virtual functions to create
171 port representors for on initialization of the PF PMD by passing the VF IDs of
172 the VFs which are required.::
174 -w DBDF,representor=[0,1,4]
176 Currently hot-plugging of representor ports is not supported so all required
177 representors must be specified on the creation of the PF.
179 - ``Use latest supported vector`` (default ``disable``)
181 Latest supported vector path may not always get the best perf so vector path was
182 recommended to use only on later platform. But users may want the latest vector path
183 since it can get better perf in some real work loading cases. So ``devargs`` param
184 ``use-latest-supported-vec`` is introduced, for example::
186 -w 84:00.0,use-latest-supported-vec=1
188 - ``Enable validation for VF message`` (default ``not enabled``)
190 The PF counts messages from each VF. If in any period of seconds the message
191 statistic from a VF exceeds maximal limitation, the PF will ignore any new message
192 from that VF for some seconds.
193 Format -- "maximal-message@period-seconds:ignore-seconds"
196 -w 84:00.0,vf_msg_cfg=80@120:180
198 Vector RX Pre-conditions
199 ~~~~~~~~~~~~~~~~~~~~~~~~
200 For Vector RX it is assumed that the number of descriptor rings will be a power
201 of 2. With this pre-condition, the ring pointer can easily scroll back to the
202 head after hitting the tail without a conditional check. In addition Vector RX
203 can use this assumption to do a bit mask using ``ring_size - 1``.
205 Driver compilation and testing
206 ------------------------------
208 Refer to the document :ref:`compiling and testing a PMD for a NIC <pmd_build_and_test>`
212 SR-IOV: Prerequisites and sample Application Notes
213 --------------------------------------------------
215 #. Load the kernel module:
217 .. code-block:: console
221 Check the output in dmesg:
223 .. code-block:: console
225 i40e 0000:83:00.1 ens802f0: renamed from eth0
227 #. Bring up the PF ports:
229 .. code-block:: console
233 #. Create VF device(s):
235 Echo the number of VFs to be created into the ``sriov_numvfs`` sysfs entry
240 .. code-block:: console
242 echo 2 > /sys/devices/pci0000:00/0000:00:03.0/0000:81:00.0/sriov_numvfs
245 #. Assign VF MAC address:
247 Assign MAC address to the VF using iproute2 utility. The syntax is:
249 .. code-block:: console
251 ip link set <PF netdev id> vf <VF id> mac <macaddr>
255 .. code-block:: console
257 ip link set ens802f0 vf 0 mac a0:b0:c0:d0:e0:f0
259 #. Assign VF to VM, and bring up the VM.
260 Please see the documentation for the *I40E/IXGBE/IGB Virtual Function Driver*.
264 Follow instructions available in the document
265 :ref:`compiling and testing a PMD for a NIC <pmd_build_and_test>`
270 .. code-block:: console
273 EAL: PCI device 0000:83:00.0 on NUMA socket 1
274 EAL: probe driver: 8086:1572 rte_i40e_pmd
275 EAL: PCI memory mapped at 0x7f7f80000000
276 EAL: PCI memory mapped at 0x7f7f80800000
277 PMD: eth_i40e_dev_init(): FW 5.0 API 1.5 NVM 05.00.02 eetrack 8000208a
278 Interactive-mode selected
279 Configuring Port 0 (socket 0)
282 PMD: i40e_dev_rx_queue_setup(): Rx Burst Bulk Alloc Preconditions are
283 satisfied.Rx Burst Bulk Alloc function will be used on port=0, queue=0.
286 Port 0: 68:05:CA:26:85:84
287 Checking link statuses...
288 Port 0 Link Up - speed 10000 Mbps - full-duplex
294 Sample Application Notes
295 ------------------------
300 Vlan filter only works when Promiscuous mode is off.
302 To start ``testpmd``, and add vlan 10 to port 0:
304 .. code-block:: console
306 ./app/testpmd -l 0-15 -n 4 -- -i --forward-mode=mac
309 testpmd> set promisc 0 off
310 testpmd> rx_vlan add 10 0
316 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
317 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.
319 The default input set of each flow type is::
321 ipv4-other : src_ip_address, dst_ip_address
322 ipv4-frag : src_ip_address, dst_ip_address
323 ipv4-tcp : src_ip_address, dst_ip_address, src_port, dst_port
324 ipv4-udp : src_ip_address, dst_ip_address, src_port, dst_port
325 ipv4-sctp : src_ip_address, dst_ip_address, src_port, dst_port,
327 ipv6-other : src_ip_address, dst_ip_address
328 ipv6-frag : src_ip_address, dst_ip_address
329 ipv6-tcp : src_ip_address, dst_ip_address, src_port, dst_port
330 ipv6-udp : src_ip_address, dst_ip_address, src_port, dst_port
331 ipv6-sctp : src_ip_address, dst_ip_address, src_port, dst_port,
333 l2_payload : ether_type
335 The flex payload is selected from offset 0 to 15 of packet's payload by default, while it is masked out from matching.
337 Start ``testpmd`` with ``--disable-rss`` and ``--pkt-filter-mode=perfect``:
339 .. code-block:: console
341 ./app/testpmd -l 0-15 -n 4 -- -i --disable-rss --pkt-filter-mode=perfect \
342 --rxq=8 --txq=8 --nb-cores=8 --nb-ports=1
344 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:
346 .. code-block:: console
348 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp \
349 src 2.2.2.3 32 dst 2.2.2.5 32 vlan 0 flexbytes () \
350 fwd pf queue 1 fd_id 1
352 Check the flow director status:
354 .. code-block:: console
356 testpmd> show port fdir 0
358 ######################## FDIR infos for port 0 ####################
360 SUPPORTED FLOW TYPE: ipv4-frag ipv4-tcp ipv4-udp ipv4-sctp ipv4-other
361 ipv6-frag ipv6-tcp ipv6-udp ipv6-sctp ipv6-other
364 max_len: 16 payload_limit: 480
365 payload_unit: 2 payload_seg: 3
366 bitmask_unit: 2 bitmask_num: 2
369 src_ipv4: 0x00000000,
370 dst_ipv4: 0x00000000,
373 src_ipv6: 0x00000000,0x00000000,0x00000000,0x00000000,
374 dst_ipv6: 0x00000000,0x00000000,0x00000000,0x00000000
375 FLEX PAYLOAD SRC OFFSET:
376 L2_PAYLOAD: 0 1 2 3 4 5 6 ...
377 L3_PAYLOAD: 0 1 2 3 4 5 6 ...
378 L4_PAYLOAD: 0 1 2 3 4 5 6 ...
380 ipv4-udp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
381 ipv4-tcp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
382 ipv4-sctp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
383 ipv4-other: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
384 ipv4-frag: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
385 ipv6-udp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
386 ipv6-tcp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
387 ipv6-sctp: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
388 ipv6-other: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
389 ipv6-frag: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
390 l2_payload: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
391 guarant_count: 1 best_count: 0
392 guarant_space: 512 best_space: 7168
399 Delete all flow director rules on a port:
401 .. code-block:: console
403 testpmd> flush_flow_director 0
408 The Intel® Ethernet 700 Series support a feature called
411 A Virtual Ethernet Bridge (VEB) is an IEEE Edge Virtual Bridging (EVB) term
412 for functionality that allows local switching between virtual endpoints within
413 a physical endpoint and also with an external bridge/network.
415 A "Floating" VEB doesn't have an uplink connection to the outside world so all
416 switching is done internally and remains within the host. As such, this
417 feature provides security benefits.
419 In addition, a Floating VEB overcomes a limitation of normal VEBs where they
420 cannot forward packets when the physical link is down. Floating VEBs don't need
421 to connect to the NIC port so they can still forward traffic from VF to VF
422 even when the physical link is down.
424 Therefore, with this feature enabled VFs can be limited to communicating with
425 each other but not an outside network, and they can do so even when there is
426 no physical uplink on the associated NIC port.
428 To enable this feature, the user should pass a ``devargs`` parameter to the
431 -w 84:00.0,enable_floating_veb=1
433 In this configuration the PMD will use the floating VEB feature for all the
434 VFs created by this PF device.
436 Alternatively, the user can specify which VFs need to connect to this floating
437 VEB using the ``floating_veb_list`` argument::
439 -w 84:00.0,enable_floating_veb=1,floating_veb_list=1;3-4
441 In this example ``VF1``, ``VF3`` and ``VF4`` connect to the floating VEB,
442 while other VFs connect to the normal VEB.
444 The current implementation only supports one floating VEB and one regular
445 VEB. VFs can connect to a floating VEB or a regular VEB according to the
446 configuration passed on the EAL command line.
448 The floating VEB functionality requires a NIC firmware version of 5.0
451 Dynamic Device Personalization (DDP)
452 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
454 The Intel® Ethernet 700 Series except for the Intel Ethernet Connection
455 X722 support a feature called "Dynamic Device Personalization (DDP)",
456 which is used to configure hardware by downloading a profile to support
457 protocols/filters which are not supported by default. The DDP
458 functionality requires a NIC firmware version of 6.0 or greater.
460 Current implementation supports GTP-C/GTP-U/PPPoE/PPPoL2TP,
461 steering can be used with rte_flow API.
463 GTPv1 package is released, and it can be downloaded from
464 https://downloadcenter.intel.com/download/27587.
466 PPPoE package is released, and it can be downloaded from
467 https://downloadcenter.intel.com/download/28040.
469 Load a profile which supports GTP and store backup profile:
471 .. code-block:: console
473 testpmd> ddp add 0 ./gtp.pkgo,./backup.pkgo
475 Delete a GTP profile and restore backup profile:
477 .. code-block:: console
479 testpmd> ddp del 0 ./backup.pkgo
481 Get loaded DDP package info list:
483 .. code-block:: console
485 testpmd> ddp get list 0
487 Display information about a GTP profile:
489 .. code-block:: console
491 testpmd> ddp get info ./gtp.pkgo
493 Input set configuration
494 ~~~~~~~~~~~~~~~~~~~~~~~
495 Input set for any PCTYPE can be configured with user defined configuration,
496 For example, to use only 48bit prefix for IPv6 src address for IPv6 TCP RSS:
498 .. code-block:: console
500 testpmd> port config 0 pctype 43 hash_inset clear all
501 testpmd> port config 0 pctype 43 hash_inset set field 13
502 testpmd> port config 0 pctype 43 hash_inset set field 14
503 testpmd> port config 0 pctype 43 hash_inset set field 15
505 Queue region configuration
506 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
507 The Intel® Ethernet 700 Series supports a feature of queue regions
508 configuration for RSS in the PF, so that different traffic classes or
509 different packet classification types can be separated to different
510 queues in different queue regions. There is an API for configuration
511 of queue regions in RSS with a command line. It can parse the parameters
512 of the region index, queue number, queue start index, user priority, traffic
513 classes and so on. Depending on commands from the command line, it will call
514 i40e private APIs and start the process of setting or flushing the queue
515 region configuration. As this feature is specific for i40e only private
516 APIs are used. These new ``test_pmd`` commands are as shown below. For
517 details please refer to :doc:`../testpmd_app_ug/index`.
519 .. code-block:: console
521 testpmd> set port (port_id) queue-region region_id (value) \
522 queue_start_index (value) queue_num (value)
523 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
524 testpmd> set port (port_id) queue-region UP (value) region_id (value)
525 testpmd> set port (port_id) queue-region flush (on|off)
526 testpmd> show port (port_id) queue-region
528 Limitations or Known issues
529 ---------------------------
531 MPLS packet classification
532 ~~~~~~~~~~~~~~~~~~~~~~~~~~
534 For firmware versions prior to 5.0, MPLS packets are not recognized by the NIC.
535 The L2 Payload flow type in flow director can be used to classify MPLS packet
536 by using a command in testpmd like:
538 testpmd> flow_director_filter 0 mode IP add flow l2_payload ether \
539 0x8847 flexbytes () fwd pf queue <N> fd_id <M>
541 With the NIC firmware version 5.0 or greater, some limited MPLS support
542 is added: Native MPLS (MPLS in Ethernet) skip is implemented, while no
543 new packet type, no classification or offload are possible. With this change,
544 L2 Payload flow type in flow director cannot be used to classify MPLS packet
545 as with previous firmware versions. Meanwhile, the Ethertype filter can be
546 used to classify MPLS packet by using a command in testpmd like:
548 testpmd> ethertype_filter 0 add mac_ignr 00:00:00:00:00:00 ethertype \
551 16 Byte RX Descriptor setting on DPDK VF
552 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
554 Currently the VF's RX descriptor mode is decided by PF. There's no PF-VF
555 interface for VF to request the RX descriptor mode, also no interface to notify
556 VF its own RX descriptor mode.
557 For all available versions of the i40e driver, these drivers don't support 16
558 byte RX descriptor. If the Linux i40e kernel driver is used as host driver,
559 while DPDK i40e PMD is used as the VF driver, DPDK cannot choose 16 byte receive
560 descriptor. The reason is that the RX descriptor is already set to 32 byte by
561 the i40e kernel driver. That is to say, user should keep
562 ``CONFIG_RTE_LIBRTE_I40E_16BYTE_RX_DESC=n`` in config file.
563 In the future, if the Linux i40e driver supports 16 byte RX descriptor, user
564 should make sure the DPDK VF uses the same RX descriptor mode, 16 byte or 32
565 byte, as the PF driver.
567 The same rule for DPDK PF + DPDK VF. The PF and VF should use the same RX
568 descriptor mode. Or the VF RX will not work.
570 Receive packets with Ethertype 0x88A8
571 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
573 Due to the FW limitation, PF can receive packets with Ethertype 0x88A8
574 only when floating VEB is disabled.
576 Incorrect Rx statistics when packet is oversize
577 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
579 When a packet is over maximum frame size, the packet is dropped.
580 However, the Rx statistics, when calling `rte_eth_stats_get` incorrectly
581 shows it as received.
583 VF & TC max bandwidth setting
584 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
586 The per VF max bandwidth and per TC max bandwidth cannot be enabled in parallel.
587 The behavior is different when handling per VF and per TC max bandwidth setting.
588 When enabling per VF max bandwidth, SW will check if per TC max bandwidth is
589 enabled. If so, return failure.
590 When enabling per TC max bandwidth, SW will check if per VF max bandwidth
591 is enabled. If so, disable per VF max bandwidth and continue with per TC max
594 TC TX scheduling mode setting
595 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
597 There are 2 TX scheduling modes for TCs, round robin and strict priority mode.
598 If a TC is set to strict priority mode, it can consume unlimited bandwidth.
599 It means if APP has set the max bandwidth for that TC, it comes to no
601 It's suggested to set the strict priority mode for a TC that is latency
602 sensitive but no consuming much bandwidth.
604 VF performance is impacted by PCI extended tag setting
605 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
607 To reach maximum NIC performance in the VF the PCI extended tag must be
608 enabled. The DPDK i40e PF driver will set this feature during initialization,
609 but the kernel PF driver does not. So when running traffic on a VF which is
610 managed by the kernel PF driver, a significant NIC performance downgrade has
611 been observed (for 64 byte packets, there is about 25% line-rate downgrade for
612 a 25GbE device and about 35% for a 40GbE device).
614 For kernel version >= 4.11, the kernel's PCI driver will enable the extended
615 tag if it detects that the device supports it. So by default, this is not an
616 issue. For kernels <= 4.11 or when the PCI extended tag is disabled it can be
617 enabled using the steps below.
619 #. Get the current value of the PCI configure register::
621 setpci -s <XX:XX.X> a8.w
625 value = value | 0x100
627 #. Set the PCI configure register with new value::
629 setpci -s <XX:XX.X> a8.w=<value>
634 The VF vlan strip function is only supported in the i40e kernel driver >= 2.1.26.
639 DCB works only when RSS is enabled.
641 Global configuration warning
642 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
644 I40E PMD will set some global registers to enable some function or set some
645 configure. Then when using different ports of the same NIC with Linux kernel
646 and DPDK, the port with Linux kernel will be impacted by the port with DPDK.
647 For example, register I40E_GL_SWT_L2TAGCTRL is used to control L2 tag, i40e
648 PMD uses I40E_GL_SWT_L2TAGCTRL to set vlan TPID. If setting TPID in port A
649 with DPDK, then the configuration will also impact port B in the NIC with
650 kernel driver, which don't want to use the TPID.
651 So PMD reports warning to clarify what is changed by writing global register.
653 High Performance of Small Packets on 40GbE NIC
654 ----------------------------------------------
656 As there might be firmware fixes for performance enhancement in latest version
657 of firmware image, the firmware update might be needed for getting high performance.
658 Check the Intel support website for the latest firmware updates.
659 Users should consult the release notes specific to a DPDK release to identify
660 the validated firmware version for a NIC using the i40e driver.
662 Use 16 Bytes RX Descriptor Size
663 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
665 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.
666 Configuration of ``CONFIG_RTE_LIBRTE_I40E_16BYTE_RX_DESC`` in config files can be changed to use 16 bytes size RX descriptors.
668 Example of getting best performance with l3fwd example
669 ------------------------------------------------------
671 The following is an example of running the DPDK ``l3fwd`` sample application to get high performance with a
672 server with Intel Xeon processors and Intel Ethernet CNA XL710.
674 The example scenario is to get best performance with two Intel Ethernet CNA XL710 40GbE ports.
675 See :numref:`figure_intel_perf_test_setup` for the performance test setup.
677 .. _figure_intel_perf_test_setup:
679 .. figure:: img/intel_perf_test_setup.*
681 Performance Test Setup
684 1. Add two Intel Ethernet CNA XL710 to the platform, and use one port per card to get best performance.
685 The reason for using two NICs is to overcome a PCIe v3.0 limitation since it cannot provide 80GbE bandwidth
686 for two 40GbE ports, but two different PCIe v3.0 x8 slot can.
687 Refer to the sample NICs output above, then we can select ``82:00.0`` and ``85:00.0`` as test ports::
689 82:00.0 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
690 85:00.0 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
692 2. Connect the ports to the traffic generator. For high speed testing, it's best to use a hardware traffic generator.
694 3. Check the PCI devices numa node (socket id) and get the cores number on the exact socket id.
695 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
697 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
698 cores from different cores (e.g core18 and core19).
700 4. Bind these two ports to igb_uio.
702 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
703 will be required, and each queue pair will need a dedicated CPU core for receiving/transmitting packets.
705 6. The DPDK sample application ``l3fwd`` will be used for performance testing, with using two ports for bi-directional forwarding.
706 Compile the ``l3fwd sample`` with the default lpm mode.
708 7. The command line of running l3fwd would be something like the following::
710 ./l3fwd -l 18-21 -n 4 -w 82:00.0 -w 85:00.0 \
711 -- -p 0x3 --config '(0,0,18),(0,1,19),(1,0,20),(1,1,21)'
713 This means that the application uses core 18 for port 0, queue pair 0 forwarding, core 19 for port 0, queue pair 1 forwarding,
714 core 20 for port 1, queue pair 0 forwarding, and core 21 for port 1, queue pair 1 forwarding.
716 8. Configure the traffic at a traffic generator.
718 * Start creating a stream on packet generator.
720 * Set the Ethernet II type to 0x0800.
722 Tx bytes affected by the link status change
723 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
725 For firmware versions prior to 6.01 for X710 series and 3.33 for X722 series, the tx_bytes statistics data is affected by
726 the link down event. Each time the link status changes to down, the tx_bytes decreases 110 bytes.