1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright 2017 The DPDK contributors
7 .. **Read this first.**
9 The text in the sections below explains how to update the release notes.
11 Use proper spelling, capitalization and punctuation in all sections.
13 Variable and config names should be quoted as fixed width text:
16 Build the docs and view the output file to ensure the changes are correct::
20 xdg-open build/doc/html/guides/rel_notes/release_17_05.html
26 .. This section should contain new features added in this release. Sample
29 * **Add a title in the past tense with a full stop.**
31 Add a short 1-2 sentence description in the past tense. The description
32 should be enough to allow someone scanning the release notes to
33 understand the new feature.
35 If the feature adds a lot of sub-features you can use a bullet list like
38 * Added feature foo to do something.
39 * Enhanced feature bar to do something else.
41 Refer to the previous release notes for examples.
43 This section is a comment. do not overwrite or remove it.
44 Also, make sure to start the actual text at the margin.
45 =========================================================
47 * **Reorganized mbuf structure.**
49 The mbuf structure has been reorganized as follows:
51 * Align fields to facilitate the writing of ``data_off``, ``refcnt``, and
52 ``nb_segs`` in one operation.
53 * Use 2 bytes for port and number of segments.
54 * Move the sequence number to the second cache line.
55 * Add a timestamp field.
56 * Set default value for ``refcnt``, ``next`` and ``nb_segs`` at mbuf free.
58 * **Added mbuf raw free API.**
60 Moved ``rte_mbuf_raw_free()`` and ``rte_pktmbuf_prefree_seg()`` functions to
63 * **Added free Tx mbuf on demand API.**
65 Added a new function ``rte_eth_tx_done_cleanup()`` which allows an
66 application to request the driver to release mbufs that are no longer in use
67 from a Tx ring, independent of whether or not the ``tx_rs_thresh`` has been
70 * **Added device removal interrupt.**
72 Added a new ethdev event ``RTE_ETH_DEV_INTR_RMV`` to signify
73 the sudden removal of a device.
74 This event can be advertised by PCI drivers and enabled accordingly.
76 * **Added EAL dynamic log framework.**
78 Added new APIs to dynamically register named log types, and control
79 the level of each type independently.
81 * **Added descriptor status ethdev API.**
83 Added a new API to get the status of a descriptor.
85 For Rx, it is almost similar to the ``rx_descriptor_done`` API, except
86 it differentiates descriptors which are held by the driver and not
87 returned to the hardware. For Tx, it is a new API.
89 * **Increased number of next hops for LPM IPv6 to 2^21.**
91 The next_hop field has been extended from 8 bits to 21 bits for IPv6.
93 * **Added VFIO hotplug support.**
95 Added hotplug support for VFIO in addition to the existing UIO support.
97 * **Added PowerPC support to pci probing for vfio-pci devices.**
99 Enabled sPAPR IOMMU based pci probing for vfio-pci devices.
101 * **Kept consistent PMD batching behavior.**
103 Removed the limit of fm10k/i40e/ixgbe Tx burst size and vhost Rx/Tx burst size
104 in order to support the same policy of "make an best effort to Rx/Tx pkts"
107 * **Updated the ixgbe base driver.**
109 Updated the ixgbe base driver, including the following changes:
111 * Add link block check for KR.
112 * Complete HW initialization even if SFP is not present.
113 * Add VF xcast promiscuous mode.
115 * **Added PowerPC support for i40e and its vector PMD.**
117 Enabled i40e PMD and its vector PMD by default in PowerPC.
119 * **Added VF max bandwidth setting in i40e.**
121 Enabled capability to set the max bandwidth for a VF in i40e.
123 * **Added VF TC min and max bandwidth setting in i40e.**
125 Enabled capability to set the min and max allocated bandwidth for a TC on a
128 * **Added TC strict priority mode setting on i40e.**
130 There are 2 Tx scheduling modes supported for TCs by i40e HW: round robin
131 mode and strict priority mode. By default the round robin mode is used. It
132 is now possible to change the Tx scheduling mode for a TC. This is a global
133 setting on a physical port.
135 * **Added i40e dynamic device personalization support.**
137 * Added dynamic device personalization processing to i40e firmware.
139 * **Updated i40e driver to support MPLSoUDP/MPLSoGRE.**
141 Updated i40e PMD to support MPLSoUDP/MPLSoGRE with MPLSoUDP/MPLSoGRE
142 supporting profiles which can be programmed by dynamic device personalization
145 * **Added Cloud Filter for QinQ steering to i40e.**
147 * Added a QinQ cloud filter on the i40e PMD, for steering traffic to a VM
148 using both VLAN tags. Note, this feature is not supported in Vector Mode.
150 * **Updated mlx5 PMD.**
152 Updated the mlx5 driver, including the following changes:
154 * Added Generic flow API support for classification according to ether type.
155 * Extended Generic flow API support for classification of IPv6 flow
156 according to Vtc flow, Protocol and Hop limit.
157 * Added Generic flow API support for FLAG action.
158 * Added Generic flow API support for RSS action.
159 * Added support for TSO for non-tunneled and VXLAN packets.
160 * Added support for hardware Tx checksum offloads for VXLAN packets.
161 * Added support for user space Rx interrupt mode.
162 * Improved ConnectX-5 single core and maximum performance.
164 * **Updated mlx4 PMD.**
166 Updated the mlx4 driver, including the following changes:
168 * Added support for Generic flow API basic flow items and actions.
169 * Added support for device removal event.
171 * **Updated the sfc_efx driver.**
173 * Added Generic Flow API support for Ethernet, VLAN, IPv4, IPv6, UDP and TCP
174 pattern items with QUEUE action for ingress traffic.
176 * Added support for virtual functions (VFs).
178 * **Added LiquidIO network PMD.**
180 Added poll mode driver support for Cavium LiquidIO II server adapter VFs.
182 * **Added Atomic Rules Arkville PMD.**
184 Added a new poll mode driver for the Arkville family of
185 devices from Atomic Rules. The net/ark PMD supports line-rate
186 agnostic, multi-queue data movement on Arkville core FPGA instances.
188 * **Added support for NXP DPAA2 - FSLMC bus.**
190 Added the new bus "fslmc" driver for NXP DPAA2 devices. See the
191 "Network Interface Controller Drivers" document for more details of this new
194 * **Added support for NXP DPAA2 Network PMD.**
196 Added the new "dpaa2" net driver for NXP DPAA2 devices. See the
197 "Network Interface Controller Drivers" document for more details of this new
200 * **Added support for the Wind River Systems AVP PMD.**
202 Added a new networking driver for the AVP device type. Theses devices are
203 specific to the Wind River Systems virtualization platforms.
205 * **Added vmxnet3 version 3 support.**
207 Added support for vmxnet3 version 3 which includes several
208 performance enhancements such as configurable Tx data ring, Receive
209 Data Ring, and the ability to register memory regions.
211 * **Updated the TAP driver.**
213 Updated the TAP PMD to:
215 * Support MTU modification.
216 * Support packet type for Rx.
217 * Support segmented packets on Rx and Tx.
218 * Speed up Rx on TAP when no packets are available.
219 * Support capturing traffic from another netdevice.
220 * Dynamically change link status when the underlying interface state changes.
221 * Added Generic Flow API support for Ethernet, VLAN, IPv4, IPv6, UDP and
222 TCP pattern items with DROP, QUEUE and PASSTHRU actions for ingress
225 * **Added MTU feature support to Virtio and Vhost.**
227 Implemented new Virtio MTU feature in Vhost and Virtio:
229 * Add ``rte_vhost_mtu_get()`` API to Vhost library.
230 * Enable Vhost PMD's MTU get feature.
231 * Get max MTU value from host in Virtio PMD
233 * **Added interrupt mode support for virtio-user.**
235 Implemented Rxq interrupt mode and LSC support for virtio-user as a virtual
236 device. Supported cases:
238 * Rxq interrupt for virtio-user + vhost-user as the backend.
239 * Rxq interrupt for virtio-user + vhost-kernel as the backend.
240 * LSC interrupt for virtio-user + vhost-user as the backend.
242 * **Added event driven programming model library (rte_eventdev).**
244 This API introduces an event driven programming model.
246 In a polling model, lcores poll ethdev ports and associated
247 Rx queues directly to look for a packet. By contrast in an event
248 driven model, lcores call the scheduler that selects packets for
249 them based on programmer-specified criteria. The Eventdev library
250 adds support for an event driven programming model, which offers
251 applications automatic multicore scaling, dynamic load balancing,
252 pipelining, packet ingress order maintenance and
253 synchronization services to simplify application packet processing.
255 By introducing an event driven programming model, DPDK can support
256 both polling and event driven programming models for packet processing,
257 and applications are free to choose whatever model
258 (or combination of the two) best suits their needs.
260 * **Added Software Eventdev PMD.**
262 Added support for the software eventdev PMD. The software eventdev is a
263 software based scheduler device that implements the eventdev API. This
264 PMD allows an application to configure a pipeline using the eventdev
265 library, and run the scheduling workload on a CPU core.
267 * **Added Cavium OCTEONTX Eventdev PMD.**
269 Added the new octeontx ssovf eventdev driver for OCTEONTX devices. See the
270 "Event Device Drivers" document for more details on this new driver.
272 * **Added information metrics library.**
274 Added a library that allows information metrics to be added and updated
275 by producers, typically other libraries, for later retrieval by
276 consumers such as applications. It is intended to provide a
277 reporting mechanism that is independent of other libraries such
280 * **Added bit-rate calculation library.**
282 Added a library that can be used to calculate device bit-rates. Calculated
283 bitrates are reported using the metrics library.
285 * **Added latency stats library.**
287 Added a library that measures packet latency. The collected statistics are
288 jitter and latency. For latency the minimum, average, and maximum is
291 * **Added NXP DPAA2 SEC crypto PMD.**
293 A new "dpaa2_sec" hardware based crypto PMD for NXP DPAA2 devices has been
294 added. See the "Crypto Device Drivers" document for more details on this
297 * **Updated the Cryptodev Scheduler PMD.**
299 * Added a packet-size based distribution mode, which distributes the enqueued
300 crypto operations among two slaves, based on their data lengths.
301 * Added fail-over scheduling mode, which enqueues crypto operations to a
302 primary slave first. Then, any operation that cannot be enqueued is
303 enqueued to a secondary slave.
304 * Added mode specific option support, so each scheduling mode can
305 now be configured individually by the new API.
307 * **Updated the QAT PMD.**
309 The QAT PMD has been updated with additional support for:
311 * AES DOCSIS BPI algorithm.
312 * DES DOCSIS BPI algorithm.
313 * ZUC EEA3/EIA3 algorithms.
315 * **Updated the AESNI MB PMD.**
317 The AESNI MB PMD has been updated with additional support for:
319 * AES DOCSIS BPI algorithm.
321 * **Updated the OpenSSL PMD.**
323 The OpenSSL PMD has been updated with additional support for:
325 * DES DOCSIS BPI algorithm.
331 .. This section should contain bug fixes added to the relevant
332 sections. Sample format:
334 * **code/section Fixed issue in the past tense with a full stop.**
336 Add a short 1-2 sentence description of the resolved issue in the past
339 The title should contain the code/lib section like a commit message.
341 Add the entries in alphabetic order in the relevant sections below.
343 This section is a comment. do not overwrite or remove it.
344 Also, make sure to start the actual text at the margin.
345 =========================================================
348 * **l2fwd-keepalive: Fixed unclean shutdowns.**
350 Added clean shutdown to l2fwd-keepalive so that it can free up
351 stale resources used for inter-process communication.
357 .. This section should contain new known issues in this release. Sample format:
359 * **Add title in present tense with full stop.**
361 Add a short 1-2 sentence description of the known issue in the present
362 tense. Add information on any known workarounds.
364 This section is a comment. do not overwrite or remove it.
365 Also, make sure to start the actual text at the margin.
366 =========================================================
368 * **LSC interrupt doesn't work for virtio-user + vhost-kernel.**
370 LSC interrupt cannot be detected when setting the backend, tap device,
371 up/down as we fail to find a way to monitor such event.
377 .. This section should contain API changes. Sample format:
379 * Add a short 1-2 sentence description of the API change. Use fixed width
380 quotes for ``rte_function_names`` or ``rte_struct_names``. Use the past
383 This section is a comment. do not overwrite or remove it.
384 Also, make sure to start the actual text at the margin.
385 =========================================================
387 * The LPM ``next_hop`` field is extended from 8 bits to 21 bits for IPv6
388 while keeping ABI compatibility.
390 * **Reworked rte_ring library.**
392 The rte_ring library has been reworked and updated. The following changes
393 have been made to it:
395 * Removed the build-time setting ``CONFIG_RTE_RING_SPLIT_PROD_CONS``.
396 * Removed the build-time setting ``CONFIG_RTE_LIBRTE_RING_DEBUG``.
397 * Removed the build-time setting ``CONFIG_RTE_RING_PAUSE_REP_COUNT``.
398 * Removed the function ``rte_ring_set_water_mark`` as part of a general
399 removal of watermarks support in the library.
400 * Added an extra parameter to the burst/bulk enqueue functions to
401 return the number of free spaces in the ring after enqueue. This can
402 be used by an application to implement its own watermark functionality.
403 * Added an extra parameter to the burst/bulk dequeue functions to return
404 the number elements remaining in the ring after dequeue.
405 * Changed the return value of the enqueue and dequeue bulk functions to
406 match that of the burst equivalents. In all cases, ring functions which
407 operate on multiple packets now return the number of elements enqueued
408 or dequeued, as appropriate. The updated functions are:
410 - ``rte_ring_mp_enqueue_bulk``
411 - ``rte_ring_sp_enqueue_bulk``
412 - ``rte_ring_enqueue_bulk``
413 - ``rte_ring_mc_dequeue_bulk``
414 - ``rte_ring_sc_dequeue_bulk``
415 - ``rte_ring_dequeue_bulk``
417 NOTE: the above functions all have different parameters as well as
418 different return values, due to the other listed changes above. This
419 means that all instances of the functions in existing code will be
420 flagged by the compiler. The return value usage should be checked
421 while fixing the compiler error due to the extra parameter.
423 * **Reworked rte_vhost library.**
425 The rte_vhost library has been reworked to make it generic enough so that
426 the user could build other vhost-user drivers on top of it. To achieve this
427 the following changes have been made:
429 * The following vhost-pmd APIs are removed:
431 * ``rte_eth_vhost_feature_disable``
432 * ``rte_eth_vhost_feature_enable``
433 * ``rte_eth_vhost_feature_get``
435 * The vhost API ``rte_vhost_driver_callback_register(ops)`` is reworked to
436 be per vhost-user socket file. Thus, it takes one more argument:
437 ``rte_vhost_driver_callback_register(path, ops)``.
439 * The vhost API ``rte_vhost_get_queue_num`` is deprecated, instead,
440 ``rte_vhost_get_vring_num`` should be used.
442 * The following macros are removed in ``rte_virtio_net.h``
448 * The following net specific header files are removed in ``rte_virtio_net.h``
450 * ``linux/virtio_net.h``
455 * The vhost struct ``virtio_net_device_ops`` is renamed to
458 * The vhost API ``rte_vhost_driver_session_start`` is removed. Instead,
459 ``rte_vhost_driver_start`` should be used, and there is no need to create
462 * The vhost public header file ``rte_virtio_net.h`` is renamed to
469 .. This section should contain ABI changes. Sample format:
471 * Add a short 1-2 sentence description of the ABI change that was announced
472 in the previous releases and made in this release. Use fixed width quotes
473 for ``rte_function_names`` or ``rte_struct_names``. Use the past tense.
475 This section is a comment. do not overwrite or remove it.
476 Also, make sure to start the actual text at the margin.
477 =========================================================
479 * **Reorganized the mbuf structure.**
481 The order and size of the fields in the ``mbuf`` structure changed,
482 as described in the `New Features`_ section.
484 * The ``rte_cryptodev_info.sym`` structure has a new field ``max_nb_sessions_per_qp``
485 to support drivers which may support a limited number of sessions per queue_pair.
491 .. This section should contain removed items in this release. Sample format:
493 * Add a short 1-2 sentence description of the removed item in the past
496 This section is a comment. do not overwrite or remove it.
497 Also, make sure to start the actual text at the margin.
498 =========================================================
500 * KNI vhost support has been removed.
502 * The dpdk_qat sample application has been removed.
504 Shared Library Versions
505 -----------------------
507 .. Update any library version updated in this release and prepend with a ``+``
511 + librte_cfgfile.so.2
514 This section is a comment. do not overwrite or remove it.
515 =========================================================
518 The libraries prepended with a plus sign were incremented in this version.
523 + librte_bitratestats.so.1
526 librte_cryptodev.so.2
527 librte_distributor.so.1
530 + librte_eventdev.so.1
536 + librte_latencystats.so.1
541 + librte_metrics.so.1
560 .. This section should contain a list of platforms that were tested with this
565 * <vendor> platform with <vendor> <type of devices> combinations
570 * Other relevant details...
572 This section is a comment. do not overwrite or remove it.
573 Also, make sure to start the actual text at the margin.
574 =========================================================
576 * Intel(R) platforms with Intel(R) NICs combinations
580 * Intel(R) Atom(TM) CPU C2758 @ 2.40GHz
581 * Intel(R) Xeon(R) CPU D-1540 @ 2.00GHz
582 * Intel(R) Xeon(R) CPU E5-4667 v3 @ 2.00GHz
583 * Intel(R) Xeon(R) CPU E5-2680 v2 @ 2.80GHz
584 * Intel(R) Xeon(R) CPU E5-2699 v3 @ 2.30GHz
585 * Intel(R) Xeon(R) CPU E5-2695 v4 @ 2.10GHz
586 * Intel(R) Xeon(R) CPU E5-2658 v2 @ 2.40GHz
587 * Intel(R) Xeon(R) CPU E5-2658 v3 @ 2.20GHz
594 * Red Hat Enterprise Linux Server release 7.3
595 * SUSE Enterprise Linux 12
602 * Intel(R) 82599ES 10 Gigabit Ethernet Controller
604 * Firmware version: 0x61bf0001
605 * Device id (pf/vf): 8086:10fb / 8086:10ed
606 * Driver version: 4.0.1-k (ixgbe)
608 * Intel(R) Corporation Ethernet Connection X552/X557-AT 10GBASE-T
610 * Firmware version: 0x800001cf
611 * Device id (pf/vf): 8086:15ad / 8086:15a8
612 * Driver version: 4.2.5 (ixgbe)
614 * Intel(R) Ethernet Converged Network Adapter X710-DA4 (4x10G)
616 * Firmware version: 5.05
617 * Device id (pf/vf): 8086:1572 / 8086:154c
618 * Driver version: 1.5.23 (i40e)
620 * Intel(R) Ethernet Converged Network Adapter X710-DA2 (2x10G)
622 * Firmware version: 5.05
623 * Device id (pf/vf): 8086:1572 / 8086:154c
624 * Driver version: 1.5.23 (i40e)
626 * Intel(R) Ethernet Converged Network Adapter XL710-QDA1 (1x40G)
628 * Firmware version: 5.05
629 * Device id (pf/vf): 8086:1584 / 8086:154c
630 * Driver version: 1.5.23 (i40e)
632 * Intel(R) Ethernet Converged Network Adapter XL710-QDA2 (2X40G)
634 * Firmware version: 5.05
635 * Device id (pf/vf): 8086:1583 / 8086:154c
636 * Driver version: 1.5.23 (i40e)
638 * Intel(R) Corporation I350 Gigabit Network Connection
640 * Firmware version: 1.48, 0x800006e7
641 * Device id (pf/vf): 8086:1521 / 8086:1520
642 * Driver version: 5.2.13-k (igb)
644 * Intel(R) platforms with Mellanox(R) NICs combinations
648 * Intel(R) Xeon(R) CPU E5-2697A v4 @ 2.60GHz
649 * Intel(R) Xeon(R) CPU E5-2697 v3 @ 2.60GHz
650 * Intel(R) Xeon(R) CPU E5-2680 v2 @ 2.80GHz
651 * Intel(R) Xeon(R) CPU E5-2640 @ 2.50GHz
655 * Red Hat Enterprise Linux Server release 7.3 (Maipo)
656 * Red Hat Enterprise Linux Server release 7.2 (Maipo)
661 * MLNX_OFED: 4.0-2.0.0.0
665 * Mellanox(R) ConnectX(R)-3 Pro 40G MCX354A-FCC_Ax (2x40G)
667 * Host interface: PCI Express 3.0 x8
668 * Device ID: 15b3:1007
669 * Firmware version: 2.40.5030
671 * Mellanox(R) ConnectX(R)-4 10G MCX4111A-XCAT (1x10G)
673 * Host interface: PCI Express 3.0 x8
674 * Device ID: 15b3:1013
675 * Firmware version: 12.18.2000
677 * Mellanox(R) ConnectX(R)-4 10G MCX4121A-XCAT (2x10G)
679 * Host interface: PCI Express 3.0 x8
680 * Device ID: 15b3:1013
681 * Firmware version: 12.18.2000
683 * Mellanox(R) ConnectX(R)-4 25G MCX4111A-ACAT (1x25G)
685 * Host interface: PCI Express 3.0 x8
686 * Device ID: 15b3:1013
687 * Firmware version: 12.18.2000
689 * Mellanox(R) ConnectX(R)-4 25G MCX4121A-ACAT (2x25G)
691 * Host interface: PCI Express 3.0 x8
692 * Device ID: 15b3:1013
693 * Firmware version: 12.18.2000
695 * Mellanox(R) ConnectX(R)-4 40G MCX4131A-BCAT/MCX413A-BCAT (1x40G)
697 * Host interface: PCI Express 3.0 x8
698 * Device ID: 15b3:1013
699 * Firmware version: 12.18.2000
701 * Mellanox(R) ConnectX(R)-4 40G MCX415A-BCAT (1x40G)
703 * Host interface: PCI Express 3.0 x16
704 * Device ID: 15b3:1013
705 * Firmware version: 12.18.2000
707 * Mellanox(R) ConnectX(R)-4 50G MCX4131A-GCAT/MCX413A-GCAT (1x50G)
709 * Host interface: PCI Express 3.0 x8
710 * Device ID: 15b3:1013
711 * Firmware version: 12.18.2000
713 * Mellanox(R) ConnectX(R)-4 50G MCX414A-BCAT (2x50G)
715 * Host interface: PCI Express 3.0 x8
716 * Device ID: 15b3:1013
717 * Firmware version: 12.18.2000
719 * Mellanox(R) ConnectX(R)-4 50G MCX415A-GCAT/MCX416A-BCAT/MCX416A-GCAT (2x50G)
721 * Host interface: PCI Express 3.0 x16
722 * Device ID: 15b3:1013
723 * Firmware version: 12.18.2000
725 * Mellanox(R) ConnectX(R)-4 50G MCX415A-CCAT (1x100G)
727 * Host interface: PCI Express 3.0 x16
728 * Device ID: 15b3:1013
729 * Firmware version: 12.18.2000
731 * Mellanox(R) ConnectX(R)-4 100G MCX416A-CCAT (2x100G)
733 * Host interface: PCI Express 3.0 x16
734 * Device ID: 15b3:1013
735 * Firmware version: 12.18.2000
737 * Mellanox(R) ConnectX(R)-4 Lx 10G MCX4121A-XCAT (2x10G)
739 * Host interface: PCI Express 3.0 x8
740 * Device ID: 15b3:1015
741 * Firmware version: 14.18.2000
743 * Mellanox(R) ConnectX(R)-4 Lx 25G MCX4121A-ACAT (2x25G)
745 * Host interface: PCI Express 3.0 x8
746 * Device ID: 15b3:1015
747 * Firmware version: 14.18.2000
749 * Mellanox(R) ConnectX(R)-5 100G MCX556A-ECAT (2x100G)
751 * Host interface: PCI Express 3.0 x16
752 * Device ID: 15b3:1017
753 * Firmware version: 16.19.1200
755 * Mellanox(R) ConnectX-5 Ex EN 100G MCX516A-CDAT (2x100G)
757 * Host interface: PCI Express 4.0 x16
758 * Device ID: 15b3:1019
759 * Firmware version: 16.19.1200
761 * IBM(R) Power8(R) with Mellanox(R) NICs combinations
765 * Processor: POWER8E (raw), AltiVec supported
766 * type-model: 8247-22L
767 * Firmware FW810.21 (SV810_108)
769 * OS: Ubuntu 16.04 LTS PPC le
771 * MLNX_OFED: 4.0-2.0.0.0
775 * Mellanox(R) ConnectX(R)-4 10G MCX4111A-XCAT (1x10G)
777 * Host interface: PCI Express 3.0 x8
778 * Device ID: 15b3:1013
779 * Firmware version: 12.18.2000
781 * Mellanox(R) ConnectX(R)-4 10G MCX4121A-XCAT (2x10G)
783 * Host interface: PCI Express 3.0 x8
784 * Device ID: 15b3:1013
785 * Firmware version: 12.18.2000
787 * Mellanox(R) ConnectX(R)-4 25G MCX4111A-ACAT (1x25G)
789 * Host interface: PCI Express 3.0 x8
790 * Device ID: 15b3:1013
791 * Firmware version: 12.18.2000
793 * Mellanox(R) ConnectX(R)-4 25G MCX4121A-ACAT (2x25G)
795 * Host interface: PCI Express 3.0 x8
796 * Device ID: 15b3:1013
797 * Firmware version: 12.18.2000
799 * Mellanox(R) ConnectX(R)-4 40G MCX4131A-BCAT/MCX413A-BCAT (1x40G)
801 * Host interface: PCI Express 3.0 x8
802 * Device ID: 15b3:1013
803 * Firmware version: 12.18.2000
805 * Mellanox(R) ConnectX(R)-4 40G MCX415A-BCAT (1x40G)
807 * Host interface: PCI Express 3.0 x16
808 * Device ID: 15b3:1013
809 * Firmware version: 12.18.2000
811 * Mellanox(R) ConnectX(R)-4 50G MCX4131A-GCAT/MCX413A-GCAT (1x50G)
813 * Host interface: PCI Express 3.0 x8
814 * Device ID: 15b3:1013
815 * Firmware version: 12.18.2000
817 * Mellanox(R) ConnectX(R)-4 50G MCX414A-BCAT (2x50G)
819 * Host interface: PCI Express 3.0 x8
820 * Device ID: 15b3:1013
821 * Firmware version: 12.18.2000
823 * Mellanox(R) ConnectX(R)-4 50G MCX415A-GCAT/MCX416A-BCAT/MCX416A-GCAT (2x50G)
825 * Host interface: PCI Express 3.0 x16
826 * Device ID: 15b3:1013
827 * Firmware version: 12.18.2000
829 * Mellanox(R) ConnectX(R)-4 50G MCX415A-CCAT (1x100G)
831 * Host interface: PCI Express 3.0 x16
832 * Device ID: 15b3:1013
833 * Firmware version: 12.18.2000
835 * Mellanox(R) ConnectX(R)-4 100G MCX416A-CCAT (2x100G)
837 * Host interface: PCI Express 3.0 x16
838 * Device ID: 15b3:1013
839 * Firmware version: 12.18.2000