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 descriptor status ethdev API.**
66 Added a new API to get the status of a descriptor.
68 For Rx, it is almost similar to the ``rx_descriptor_done`` API, except
69 it differentiates descriptors which are hold by the driver and not
70 returned to the hardware. For Tx, it is a new API.
72 * **Increased number of next hops for LPM IPv6 to 2^21.**
74 The next_hop field is extended from 8 bits to 21 bits for IPv6.
76 * **Added VFIO hotplug support.**
78 How hotplug supported with UIO and VFIO drivers.
80 * **Added powerpc support in pci probing for vfio-pci devices.**
82 sPAPR IOMMU based pci probing enabled for vfio-pci devices.
84 * **Kept consistent PMD batching behaviour.**
86 Removed the limit of fm10k/i40e/ixgbe TX burst size and vhost RX/TX burst size
87 in order to support the same policy of "make an best effort to RX/TX pkts"
90 * **Updated the ixgbe base driver.**
92 Updated the ixgbe base driver, including the following changes:
94 * Add link block check for KR.
95 * Complete HW initialization even if SFP is not present.
96 * Add VF xcast promiscuous mode.
98 * **Added powerpc support for i40e and its vector PMD .**
100 i40e PMD and its vector PMD enabled by default in powerpc.
102 * **Added VF max bandwidth setting on i40e.**
104 i40e HW supports to set the max bandwidth for a VF. Enable this capability.
106 * **Added VF TC min bandwidth setting on i40e.**
108 i40e HW supports to set the allocated bandwidth for a TC on a VF. Enable this
111 * **Added VF TC max bandwidth setting on i40e.**
113 i40e HW supports to set the max bandwidth for a TC on a VF. Enable this
116 * **Added TC strict priority mode setting on i40e.**
118 There're 2 TX scheduling modes supported for TCs by i40e HW, round ribon mode
119 and strict priority mode. By default it's round robin mode. Enable the
120 capability to change the TX scheduling mode for a TC. It's a global setting
123 * **Added i40e dynamic device personalization support.**
125 * Added dynamic device personalization processing to i40e FW.
127 * **Added TSO support for tunneled and non-tunneled packets on mlx5 driver.**
129 Added support for Hardware TSO for tunneled and non-tunneled packets.
130 Tunneling protocols supported are GRE and VXLAN.
132 * **Added support for Rx interrupts on mlx5 driver.**
134 Rx queues can be armed with an interrupt which will trigger on the
137 * **Updated the sfc_efx driver.**
139 * Generic flow API support for Ethernet, VLAN, IPv4, IPv6, UDP and TCP
140 pattern items with QUEUE action for ingress traffic.
142 * Support virtual functions (VFs)
144 * **Added LiquidIO network PMD.**
146 Added poll mode driver support for Cavium LiquidIO II server adapter VFs.
148 * **Added support for the Wind River Systems AVP PMD.**
150 Added a new networking driver for the AVP device type. Theses devices are
151 specific to the Wind River Systems virtualization platforms.
153 * **Added vmxnet3 version 3 support.**
155 Added support for vmxnet3 version 3 which includes several
156 performance enhancements viz. configurable TX data ring, Receive
157 Data Ring, ability to register memory regions.
159 * **Updated the tap driver.**
161 * Support MTU modification.
162 * Support packet type for Rx.
163 * Support segmented packets on Rx and Tx.
164 * Speed up Rx on tap when no packets are available.
165 * Support capturing traffic from another netdevice.
166 * Dynamically change link status when the underlying interface state changes.
167 * Generic flow API support for Ethernet, VLAN, IPv4, IPv6, UDP and TCP pattern
168 items with DROP, QUEUE and PASSTHRU actions for ingress traffic.
170 * **Added MTU feature support to Virtio and Vhost.**
172 Implemented new Virtio MTU feature into Vhost and Virtio:
174 * Add ``rte_vhost_mtu_get()`` API to Vhost library.
175 * Enable Vhost PMD's MTU get feature.
176 * Get max MTU value from host in Virtio PMD
178 * **Added interrupt mode support for virtio-user.**
180 Implemented Rxq interrupt mode and LSC support for virtio-user as a virtual
181 device. Supported cases:
183 * Rxq interrupt for virtio-user + vhost-user as the backend.
184 * Rxq interrupt for virtio-user + vhost-kernel as the backend.
185 * LSC interrupt for virtio-user + vhost-user as the backend.
187 * **Added event driven programming model library (rte_eventdev).**
189 This API introduces event driven programming model.
191 In a polling model, lcores poll ethdev ports and associated
192 rx queues directly to look for packet. In an event driven model,
193 by contrast, lcores call the scheduler that selects packets for
194 them based on programmer-specified criteria. Eventdev library
195 added support for event driven programming model, which offer
196 applications automatic multicore scaling, dynamic load balancing,
197 pipelining, packet ingress order maintenance and
198 synchronization services to simplify application packet processing.
200 By introducing event driven programming model, DPDK can support
201 both polling and event driven programming models for packet processing,
202 and applications are free to choose whatever model
203 (or combination of the two) that best suits their needs.
205 * **Added Software Eventdev PMD.**
207 Added support for the software eventdev PMD. The software eventdev is a
208 software based scheduler device that implements the eventdev API. This
209 PMD allows an application to configure a pipeline using the eventdev
210 library, and run the scheduling workload on a CPU core.
212 * **Added Cavium OCTEONTX Eventdev PMD.**
214 Added the new octeontx ssovf eventdev driver for OCTEONTX devices. See the
215 "Event Device Drivers" document for more details on this new driver.
221 .. This section should contain bug fixes added to the relevant
222 sections. Sample format:
224 * **code/section Fixed issue in the past tense with a full stop.**
226 Add a short 1-2 sentence description of the resolved issue in the past
229 The title should contain the code/lib section like a commit message.
231 Add the entries in alphabetic order in the relevant sections below.
233 This section is a comment. do not overwrite or remove it.
234 Also, make sure to start the actual text at the margin.
235 =========================================================
261 .. This section should contain new known issues in this release. Sample format:
263 * **Add title in present tense with full stop.**
265 Add a short 1-2 sentence description of the known issue in the present
266 tense. Add information on any known workarounds.
268 This section is a comment. do not overwrite or remove it.
269 Also, make sure to start the actual text at the margin.
270 =========================================================
272 * **LSC interrupt cannot work for virtio-user + vhost-kernel.**
274 LSC interrupt cannot be detected when setting the backend, tap device,
275 up/down as we fail to find a way to monitor such event.
281 .. This section should contain API changes. Sample format:
283 * Add a short 1-2 sentence description of the API change. Use fixed width
284 quotes for ``rte_function_names`` or ``rte_struct_names``. Use the past
287 This section is a comment. do not overwrite or remove it.
288 Also, make sure to start the actual text at the margin.
289 =========================================================
291 * The LPM ``next_hop`` field is extended from 8 bits to 21 bits for IPv6
292 while keeping ABI compatibility.
294 * **Reworked rte_ring library**
296 The rte_ring library has been reworked and updated. The following changes
297 have been made to it:
299 * removed the build-time setting ``CONFIG_RTE_RING_SPLIT_PROD_CONS``
300 * removed the build-time setting ``CONFIG_RTE_LIBRTE_RING_DEBUG``
301 * removed the build-time setting ``CONFIG_RTE_RING_PAUSE_REP_COUNT``
302 * removed the function ``rte_ring_set_water_mark`` as part of a general
303 removal of watermarks support in the library.
304 * added an extra parameter to the burst/bulk enqueue functions to
305 return the number of free spaces in the ring after enqueue. This can
306 be used by an application to implement its own watermark functionality.
307 * added an extra parameter to the burst/bulk dequeue functions to return
308 the number elements remaining in the ring after dequeue.
309 * changed the return value of the enqueue and dequeue bulk functions to
310 match that of the burst equivalents. In all cases, ring functions which
311 operate on multiple packets now return the number of elements enqueued
312 or dequeued, as appropriate. The updated functions are:
314 - ``rte_ring_mp_enqueue_bulk``
315 - ``rte_ring_sp_enqueue_bulk``
316 - ``rte_ring_enqueue_bulk``
317 - ``rte_ring_mc_dequeue_bulk``
318 - ``rte_ring_sc_dequeue_bulk``
319 - ``rte_ring_dequeue_bulk``
321 NOTE: the above functions all have different parameters as well as
322 different return values, due to the other listed changes above. This
323 means that all instances of the functions in existing code will be
324 flagged by the compiler. The return value usage should be checked
325 while fixing the compiler error due to the extra parameter.
327 * **Reworked rte_vhost library**
329 The rte_vhost library has been reworked to make it generic enough so that
330 user could build other vhost-user drivers on top of it. To achieve that,
331 following changes have been made:
333 * The following vhost-pmd APIs are removed:
335 * ``rte_eth_vhost_feature_disable``
336 * ``rte_eth_vhost_feature_enable``
337 * ``rte_eth_vhost_feature_get``
339 * The vhost API ``rte_vhost_driver_callback_register(ops)`` is reworked to
340 be per vhost-user socket file. Thus, it takes one more argument:
341 ``rte_vhost_driver_callback_register(path, ops)``.
343 * The vhost API ``rte_vhost_get_queue_num`` is deprecated, instead,
344 ``rte_vhost_get_vring_num`` should be used.
346 * Following macros are removed in ``rte_virtio_net.h``
352 * Following net specific header files are removed in ``rte_virtio_net.h``
354 * ``linux/virtio_net.h``
359 * The vhost struct ``virtio_net_device_ops`` is renamed to
362 * The vhost API ``rte_vhost_driver_session_start`` is removed. Instead,
363 ``rte_vhost_driver_start`` should be used, and no need to create a
366 * The vhost public header file ``rte_virtio_net.h`` is renamed to
373 .. This section should contain ABI changes. Sample format:
375 * Add a short 1-2 sentence description of the ABI change that was announced
376 in the previous releases and made in this release. Use fixed width quotes
377 for ``rte_function_names`` or ``rte_struct_names``. Use the past tense.
379 This section is a comment. do not overwrite or remove it.
380 Also, make sure to start the actual text at the margin.
381 =========================================================
383 * **Reorganized the mbuf structure.**
385 The order and size of the fields in the ``mbuf`` structure changed,
386 as described in the `New Features`_ section.
392 .. This section should contain removed items in this release. Sample format:
394 * Add a short 1-2 sentence description of the removed item in the past
397 This section is a comment. do not overwrite or remove it.
398 Also, make sure to start the actual text at the margin.
399 =========================================================
401 * KNI vhost support removed.
404 Shared Library Versions
405 -----------------------
407 .. Update any library version updated in this release and prepend with a ``+``
411 + librte_cfgfile.so.2
414 This section is a comment. do not overwrite or remove it.
415 =========================================================
418 The libraries prepended with a plus sign were incremented in this version.
425 librte_cryptodev.so.2
426 librte_distributor.so.1
456 .. This section should contain a list of platforms that were tested with this
461 * <vendor> platform with <vendor> <type of devices> combinations
466 * Other relevant details...
468 This section is a comment. do not overwrite or remove it.
469 Also, make sure to start the actual text at the margin.
470 =========================================================