4 .. **Read this first.**
6 The text in the sections below explains how to update the release notes.
8 Use proper spelling, capitalization and punctuation in all sections.
10 Variable and config names should be quoted as fixed width text:
13 Build the docs and view the output file to ensure the changes are correct::
17 xdg-open build/doc/html/guides/rel_notes/release_17_05.html
23 .. This section should contain new features added in this release. Sample
26 * **Add a title in the past tense with a full stop.**
28 Add a short 1-2 sentence description in the past tense. The description
29 should be enough to allow someone scanning the release notes to
30 understand the new feature.
32 If the feature adds a lot of sub-features you can use a bullet list like
35 * Added feature foo to do something.
36 * Enhanced feature bar to do something else.
38 Refer to the previous release notes for examples.
40 This section is a comment. do not overwrite or remove it.
41 Also, make sure to start the actual text at the margin.
42 =========================================================
44 * **Reorganized the mbuf structure.**
46 * Align fields to facilitate the writing of ``data_off``, ``refcnt``, and
47 ``nb_segs`` in one operation.
48 * Use 2 bytes for port and number of segments.
49 * Move the sequence number in the second cache line.
50 * Add a timestamp field.
51 * Set default value for ``refcnt``, ``next`` and ``nb_segs`` at mbuf free.
53 * **Added mbuf raw free API**
55 Moved ``rte_mbuf_raw_free()`` and ``rte_pktmbuf_prefree_seg()`` functions to
58 * **Added free Tx mbuf on demand API.**
60 Added a new function ``rte_eth_tx_done_cleanup()`` which allows an application
61 to request the driver to release mbufs from their Tx ring that are no longer
62 in use, independent of whether or not the ``tx_rs_thresh`` has been crossed.
64 * **Added device removal interrupt.**
66 Added a new ethdev event ``RTE_ETH_DEV_INTR_RMV`` to signify
67 the sudden removal of a device.
68 This event can be advertized by PCI drivers and enabled accordingly.
70 * **Added EAL dynamic log framework.**
72 Added new APIs to dynamically register named log types, and control
73 the level of each type independently.
75 * **Added descriptor status ethdev API.**
77 Added a new API to get the status of a descriptor.
79 For Rx, it is almost similar to the ``rx_descriptor_done`` API, except
80 it differentiates descriptors which are hold by the driver and not
81 returned to the hardware. For Tx, it is a new API.
83 * **Increased number of next hops for LPM IPv6 to 2^21.**
85 The next_hop field is extended from 8 bits to 21 bits for IPv6.
87 * **Added VFIO hotplug support.**
89 How hotplug supported with UIO and VFIO drivers.
91 * **Added powerpc support in pci probing for vfio-pci devices.**
93 sPAPR IOMMU based pci probing enabled for vfio-pci devices.
95 * **Kept consistent PMD batching behaviour.**
97 Removed the limit of fm10k/i40e/ixgbe TX burst size and vhost RX/TX burst size
98 in order to support the same policy of "make an best effort to RX/TX pkts"
101 * **Updated the ixgbe base driver.**
103 Updated the ixgbe base driver, including the following changes:
105 * Add link block check for KR.
106 * Complete HW initialization even if SFP is not present.
107 * Add VF xcast promiscuous mode.
109 * **Added powerpc support for i40e and its vector PMD .**
111 i40e PMD and its vector PMD enabled by default in powerpc.
113 * **Added VF max bandwidth setting on i40e.**
115 i40e HW supports to set the max bandwidth for a VF. Enable this capability.
117 * **Added VF TC min bandwidth setting on i40e.**
119 i40e HW supports to set the allocated bandwidth for a TC on a VF. Enable this
122 * **Added VF TC max bandwidth setting on i40e.**
124 i40e HW supports to set the max bandwidth for a TC on a VF. Enable this
127 * **Added TC strict priority mode setting on i40e.**
129 There're 2 TX scheduling modes supported for TCs by i40e HW, round ribon mode
130 and strict priority mode. By default it's round robin mode. Enable the
131 capability to change the TX scheduling mode for a TC. It's a global setting
134 * **Added i40e dynamic device personalization support.**
136 * Added dynamic device personalization processing to i40e FW.
138 * **Added Cloud Filter for QinQ steering to i40e.**
140 * Added a QinQ cloud filter on the i40e PMD, for steering traffic to a VM
141 using both VLAN tags.
142 * QinQ is not supported in Vector Mode on the i40e PMD.
143 * Vector Mode must be disabled when using the QinQ Cloud Filter.
145 * **Updated mlx5 PMD.**
147 * Supported ether type in flow item.
148 * Extended IPv6 flow item with Vtc flow, Protocol and Hop limit.
149 * Supported flag flow action.
150 * Supported RSS action flow rule.
151 * Supported TSO for tunneled and non-tunneled packets.
152 * Supported hardware checksum offloads for tunneled packets.
153 * Supported user space Rx interrupt event.
154 * Enhanced multi-packet send function for ConnectX-5.
156 * **Updated mlx4 PMD.**
158 * Supported basic flow items and actions.
159 * Supported device removal event.
161 * **Updated the sfc_efx driver.**
163 * Generic flow API support for Ethernet, VLAN, IPv4, IPv6, UDP and TCP
164 pattern items with QUEUE action for ingress traffic.
166 * Support virtual functions (VFs)
168 * **Added LiquidIO network PMD.**
170 Added poll mode driver support for Cavium LiquidIO II server adapter VFs.
172 * **Added Atomic Rules Arkville PMD.**
174 Added a new poll mode driver for the Arkville family of
175 devices from Atomic Rules. The net/ark PMD supports line-rate
176 agnostic, multi-queue data movement on Arkville core FPGA instances.
178 * **Added support for NXP DPAA2 - FSLMC bus.**
180 Added the new bus "fslmc" driver for NXP DPAA2 devices. See the
181 "Network Interface Controller Drivers" document for more details on this new
184 * **Added support for NXP DPAA2 Network PMD.**
186 Added the new "dpaa2" net driver for NXP DPAA2 devices. See the
187 "Network Interface Controller Drivers" document for more details on this new
190 * **Added support for the Wind River Systems AVP PMD.**
192 Added a new networking driver for the AVP device type. Theses devices are
193 specific to the Wind River Systems virtualization platforms.
195 * **Added vmxnet3 version 3 support.**
197 Added support for vmxnet3 version 3 which includes several
198 performance enhancements viz. configurable TX data ring, Receive
199 Data Ring, ability to register memory regions.
201 * **Updated the tap driver.**
203 * Support MTU modification.
204 * Support packet type for Rx.
205 * Support segmented packets on Rx and Tx.
206 * Speed up Rx on tap when no packets are available.
207 * Support capturing traffic from another netdevice.
208 * Dynamically change link status when the underlying interface state changes.
209 * Generic flow API support for Ethernet, VLAN, IPv4, IPv6, UDP and TCP pattern
210 items with DROP, QUEUE and PASSTHRU actions for ingress traffic.
212 * **Added MTU feature support to Virtio and Vhost.**
214 Implemented new Virtio MTU feature into Vhost and Virtio:
216 * Add ``rte_vhost_mtu_get()`` API to Vhost library.
217 * Enable Vhost PMD's MTU get feature.
218 * Get max MTU value from host in Virtio PMD
220 * **Added interrupt mode support for virtio-user.**
222 Implemented Rxq interrupt mode and LSC support for virtio-user as a virtual
223 device. Supported cases:
225 * Rxq interrupt for virtio-user + vhost-user as the backend.
226 * Rxq interrupt for virtio-user + vhost-kernel as the backend.
227 * LSC interrupt for virtio-user + vhost-user as the backend.
229 * **Added event driven programming model library (rte_eventdev).**
231 This API introduces event driven programming model.
233 In a polling model, lcores poll ethdev ports and associated
234 rx queues directly to look for packet. In an event driven model,
235 by contrast, lcores call the scheduler that selects packets for
236 them based on programmer-specified criteria. Eventdev library
237 added support for event driven programming model, which offer
238 applications automatic multicore scaling, dynamic load balancing,
239 pipelining, packet ingress order maintenance and
240 synchronization services to simplify application packet processing.
242 By introducing event driven programming model, DPDK can support
243 both polling and event driven programming models for packet processing,
244 and applications are free to choose whatever model
245 (or combination of the two) that best suits their needs.
247 * **Added Software Eventdev PMD.**
249 Added support for the software eventdev PMD. The software eventdev is a
250 software based scheduler device that implements the eventdev API. This
251 PMD allows an application to configure a pipeline using the eventdev
252 library, and run the scheduling workload on a CPU core.
254 * **Added Cavium OCTEONTX Eventdev PMD.**
256 Added the new octeontx ssovf eventdev driver for OCTEONTX devices. See the
257 "Event Device Drivers" document for more details on this new driver.
259 * **Added information metric library.**
261 A library that allows information metrics to be added and updated
262 by producers, typically other libraries, for later retrieval by
263 consumers such as applications. It is intended to provide a
264 reporting mechanism that is independent of other libraries such
267 * **Added bit-rate calculation library.**
269 A library that can be used to calculate device bit-rates. Calculated
270 bitrates are reported using the metrics library.
272 * **Added latency stats library.**
274 A library that measures packet latency. The collected statistics are jitter
275 and latency. For latency the minimum, average, and maximum is measured.
277 * **Added NXP DPAA2 SEC crypto PMD.**
279 A new "dpaa2_sec" hardware based crypto PMD for NXP DPAA2 devices has been
280 added. See the "Crypto Device Drivers" document for more details on this
283 * **Updated the Cryptodev Scheduler PMD.**
285 * Added packet-size based distribution mode, which distributes the enqueued
286 crypto operations among two slaves, based on their data lengths.
287 * Added fail-over scheduling mode, which enqueues crypto operations to a
288 primary slave first. Then, any operation that cannot be enqueued is
289 enqueued to a secondary slave.
290 * Added mode specific option support, so each scheduleing mode can
291 now be configured individually by the new added API.
293 * **Updated the QAT PMD.**
295 The QAT PMD has been updated with additional support for:
297 * AES DOCSIS BPI algorithm.
298 * DES DOCSIS BPI algorithm.
299 * ZUC EEA3/EIA3 algorithms.
301 * **Updated the AESNI MB PMD.**
303 The AESNI MB PMD has been updated with additional support for:
305 * AES DOCSIS BPI algorithm.
307 * **Updated the OpenSSL PMD.**
309 The OpenSSL PMD has been updated with additional support for:
311 * DES DOCSIS BPI algorithm.
317 .. This section should contain bug fixes added to the relevant
318 sections. Sample format:
320 * **code/section Fixed issue in the past tense with a full stop.**
322 Add a short 1-2 sentence description of the resolved issue in the past
325 The title should contain the code/lib section like a commit message.
327 Add the entries in alphabetic order in the relevant sections below.
329 This section is a comment. do not overwrite or remove it.
330 Also, make sure to start the actual text at the margin.
331 =========================================================
349 * **l2fwd-keepalive: Fixed unclean shutdowns.**
351 Added clean shutdown to l2fwd-keepalive so that it can free up
352 stale resources used for inter-process communication.
362 .. This section should contain new known issues in this release. Sample format:
364 * **Add title in present tense with full stop.**
366 Add a short 1-2 sentence description of the known issue in the present
367 tense. Add information on any known workarounds.
369 This section is a comment. do not overwrite or remove it.
370 Also, make sure to start the actual text at the margin.
371 =========================================================
373 * **LSC interrupt cannot work for virtio-user + vhost-kernel.**
375 LSC interrupt cannot be detected when setting the backend, tap device,
376 up/down as we fail to find a way to monitor such event.
382 .. This section should contain API changes. Sample format:
384 * Add a short 1-2 sentence description of the API change. Use fixed width
385 quotes for ``rte_function_names`` or ``rte_struct_names``. Use the past
388 This section is a comment. do not overwrite or remove it.
389 Also, make sure to start the actual text at the margin.
390 =========================================================
392 * The LPM ``next_hop`` field is extended from 8 bits to 21 bits for IPv6
393 while keeping ABI compatibility.
395 * **Reworked rte_ring library**
397 The rte_ring library has been reworked and updated. The following changes
398 have been made to it:
400 * removed the build-time setting ``CONFIG_RTE_RING_SPLIT_PROD_CONS``
401 * removed the build-time setting ``CONFIG_RTE_LIBRTE_RING_DEBUG``
402 * removed the build-time setting ``CONFIG_RTE_RING_PAUSE_REP_COUNT``
403 * removed the function ``rte_ring_set_water_mark`` as part of a general
404 removal of watermarks support in the library.
405 * added an extra parameter to the burst/bulk enqueue functions to
406 return the number of free spaces in the ring after enqueue. This can
407 be used by an application to implement its own watermark functionality.
408 * added an extra parameter to the burst/bulk dequeue functions to return
409 the number elements remaining in the ring after dequeue.
410 * changed the return value of the enqueue and dequeue bulk functions to
411 match that of the burst equivalents. In all cases, ring functions which
412 operate on multiple packets now return the number of elements enqueued
413 or dequeued, as appropriate. The updated functions are:
415 - ``rte_ring_mp_enqueue_bulk``
416 - ``rte_ring_sp_enqueue_bulk``
417 - ``rte_ring_enqueue_bulk``
418 - ``rte_ring_mc_dequeue_bulk``
419 - ``rte_ring_sc_dequeue_bulk``
420 - ``rte_ring_dequeue_bulk``
422 NOTE: the above functions all have different parameters as well as
423 different return values, due to the other listed changes above. This
424 means that all instances of the functions in existing code will be
425 flagged by the compiler. The return value usage should be checked
426 while fixing the compiler error due to the extra parameter.
428 * **Reworked rte_vhost library**
430 The rte_vhost library has been reworked to make it generic enough so that
431 user could build other vhost-user drivers on top of it. To achieve that,
432 following changes have been made:
434 * The following vhost-pmd APIs are removed:
436 * ``rte_eth_vhost_feature_disable``
437 * ``rte_eth_vhost_feature_enable``
438 * ``rte_eth_vhost_feature_get``
440 * The vhost API ``rte_vhost_driver_callback_register(ops)`` is reworked to
441 be per vhost-user socket file. Thus, it takes one more argument:
442 ``rte_vhost_driver_callback_register(path, ops)``.
444 * The vhost API ``rte_vhost_get_queue_num`` is deprecated, instead,
445 ``rte_vhost_get_vring_num`` should be used.
447 * Following macros are removed in ``rte_virtio_net.h``
453 * Following net specific header files are removed in ``rte_virtio_net.h``
455 * ``linux/virtio_net.h``
460 * The vhost struct ``virtio_net_device_ops`` is renamed to
463 * The vhost API ``rte_vhost_driver_session_start`` is removed. Instead,
464 ``rte_vhost_driver_start`` should be used, and no need to create a
467 * The vhost public header file ``rte_virtio_net.h`` is renamed to
470 * **Reworked rte_ethdev library**
472 * Changed set of input parameters for ``rte_eth_xstats_get`` and ``rte_eth_xstats_get_names`` functions.
474 * Added new functions ``rte_eth_xstats_get_all`` and ``rte_eth_xstats_get_names_all`` to provide backward compatibility for
475 ``rte_eth_xstats_get`` and ``rte_eth_xstats_get_names``
477 * Added new function ``rte_eth_xstats_get_id_by_name``
483 .. This section should contain ABI changes. Sample format:
485 * Add a short 1-2 sentence description of the ABI change that was announced
486 in the previous releases and made in this release. Use fixed width quotes
487 for ``rte_function_names`` or ``rte_struct_names``. Use the past tense.
489 This section is a comment. do not overwrite or remove it.
490 Also, make sure to start the actual text at the margin.
491 =========================================================
493 * **Reorganized the mbuf structure.**
495 The order and size of the fields in the ``mbuf`` structure changed,
496 as described in the `New Features`_ section.
498 * The ``rte_cryptodev_info.sym`` structure has new field ``max_nb_sessions_per_qp``
499 to support drivers which may support limited number of sessions per queue_pair.
505 .. This section should contain removed items in this release. Sample format:
507 * Add a short 1-2 sentence description of the removed item in the past
510 This section is a comment. do not overwrite or remove it.
511 Also, make sure to start the actual text at the margin.
512 =========================================================
514 * KNI vhost support removed.
516 * dpdk_qat sample application removed.
518 Shared Library Versions
519 -----------------------
521 .. Update any library version updated in this release and prepend with a ``+``
525 + librte_cfgfile.so.2
528 This section is a comment. do not overwrite or remove it.
529 =========================================================
532 The libraries prepended with a plus sign were incremented in this version.
537 + librte_bitratestats.so.1
540 librte_cryptodev.so.2
541 librte_distributor.so.1
549 + librte_latencystats.so.1
554 + librte_metrics.so.1
573 .. This section should contain a list of platforms that were tested with this
578 * <vendor> platform with <vendor> <type of devices> combinations
583 * Other relevant details...
585 This section is a comment. do not overwrite or remove it.
586 Also, make sure to start the actual text at the margin.
587 =========================================================