1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright 2018 The DPDK contributors 3 4DPDK Release 18.11 5================== 6 7.. **Read this first.** 8 9 The text in the sections below explains how to update the release notes. 10 11 Use proper spelling, capitalization and punctuation in all sections. 12 13 Variable and config names should be quoted as fixed width text: 14 ``LIKE_THIS``. 15 16 Build the docs and view the output file to ensure the changes are correct:: 17 18 make doc-guides-html 19 20 xdg-open build/doc/html/guides/rel_notes/release_18_11.html 21 22 23New Features 24------------ 25 26.. This section should contain new features added in this release. 27 Sample format: 28 29 * **Add a title in the past tense with a full stop.** 30 31 Add a short 1-2 sentence description in the past tense. 32 The description should be enough to allow someone scanning 33 the release notes to understand the new feature. 34 35 If the feature adds a lot of sub-features you can use a bullet list 36 like this: 37 38 * Added feature foo to do something. 39 * Enhanced feature bar to do something else. 40 41 Refer to the previous release notes for examples. 42 43 Suggested order in release notes items: 44 * Core libs (EAL, mempool, ring, mbuf, buses) 45 * Device abstraction libs and PMDs 46 - ethdev (lib, PMDs) 47 - cryptodev (lib, PMDs) 48 - eventdev (lib, PMDs) 49 - etc 50 * Other libs 51 * Apps, Examples, Tools (if significative) 52 53 This section is a comment. Do not overwrite or remove it. 54 Also, make sure to start the actual text at the margin. 55 ========================================================= 56 57* **Added support for using externally allocated memory in DPDK.** 58 59 DPDK has gained support for creating new ``rte_malloc`` heaps referencing 60 memory that was created outside of DPDK's own page allocator, and using that 61 memory natively with any other DPDK library or data structure. 62 63* **Added check for ensuring allocated memory addressable by devices.** 64 65 Some devices can have addressing limitations so a new function, 66 ``rte_mem_check_dma_mask``, has been added for checking allocated memory is 67 not out of the device range. Because now memory can be dynamically allocated 68 after initialization, a dma mask is kept and any new allocated memory will be 69 checked out against that dma mask and rejected if out of range. If more than 70 one device has addressing limitations, the dma mask is the more restricted one. 71 72* **Updated the C11 memory model version of ring library.** 73 74 The latency is decreased for architectures using the C11 memory model 75 version of the ring library. 76 77 On Cavium ThunderX2 platform, the changes decreased latency by 27~29% 78 and 3~15% for MPMC and SPSC cases respectively (with 2 lcores). The 79 real improvements may vary with the number of contending lcores and 80 the size of ring. 81 82* **Added hot-unplug handle mechanism.** 83 84 ``rte_dev_hotplug_handle_enable`` and ``rte_dev_hotplug_handle_disable`` are 85 for enabling or disabling hotplug handle mechanism. 86 87* **Support device multi-process hotplug.** 88 89 Hotplug and hot-unplug for devices will now be supported in multiprocessing 90 scenario. Any ethdev devices created in the primary process will be regarded 91 as shared and will be available for all DPDK processes. Synchronization 92 between processes will be done using DPDK IPC. 93 94* **Added new Flow API actions to rewrite fields in packet headers.** 95 96 Added new Flow API actions to: 97 98 * Modify source and destination IP addresses in the outermost IPv4/IPv6 99 headers. 100 * Modify source and destination port numbers in the outermost TCP/UDP 101 headers. 102 103* **Added new Flow API action to swap MAC addresses in Ethernet header.** 104 105 Added new Flow API action to swap the source and destination MAC 106 addresses in the outermost Ethernet header. 107 108* **Add support to offload more flow match and actions for CXGBE PMD** 109 110 Flow API support has been enhanced for CXGBE Poll Mode Driver to offload: 111 112 * Match items: destination MAC address. 113 * Action items: push/pop/rewrite vlan header, 114 rewrite IP addresses in outermost IPv4/IPv6 header, 115 rewrite port numbers in outermost TCP/UDP header, 116 swap MAC addresses in outermost Ethernet header. 117 118* **Added a devarg to use the latest supported vector path in i40e.** 119 A new devarg ``use-latest-supported-vec`` was introduced to allow users to 120 choose the latest vector path that the platform supported. For example, users 121 can use AVX2 vector path on BDW/HSW to get better performance. 122 123* **Added support for SR-IOV in netvsc PMD.** 124 125 The ``netvsc`` poll mode driver now supports the Accelerated Networking 126 SR-IOV option in Hyper-V and Azure. This is an alternative to the previous 127 vdev_netvsc, tap, and failsafe drivers combination. 128 129* **Added a new net driver for Marvell Armada 3k device.** 130 131 Added the new ``mvneta`` net driver for Marvell Armada 3k device. See the 132 :doc:`../nics/mvneta` NIC guide for more details on this new driver. 133 134* **Added NXP ENETC PMD.** 135 136 Added the new enetc driver for NXP enetc platform. See the 137 "ENETC Poll Mode Driver" document for more details on this new driver. 138 139* **Added Ethernet poll mode driver for Aquantia aQtion family of 10G devices.** 140 141 Added the new ``atlantic`` ethernet poll mode driver for Aquantia XGBE devices. 142 See the :doc:`../nics/atlantic` nic driver guide for more details on this 143 driver. 144 145* **Updated Solarflare network PMD.** 146 147 Updated the sfc_efx driver including the following changes: 148 149 * Added support for Rx scatter in EF10 datapath implementation. 150 * Added support for Rx descriptor status API in EF10 datapath implementation. 151 * Added support for TSO in EF10 datapath implementation. 152 * Added support for Tx descriptor status API in EF10 (ef10 and ef10_simple) 153 datapaths implementation. 154 155* **Updated the enic driver.** 156 157 * Added AVX2-based vectorized Rx handler. 158 * Added VLAN and checksum offloads to the simple Tx handler. 159 * Added the count flow action. 160 * Enabled the virtual address IOVA mode. 161 162* **Updated failsafe driver.** 163 164 Updated the failsafe driver including the following changes: 165 166 * Support for Rx and Tx queues start and stop. 167 * Support for Rx and Tx queues deferred start. 168 * Support for runtime Rx and Tx queues setup. 169 * Support multicast MAC address set. 170 171* **Added a devarg to use PCAP interface physical MAC address.** 172 A new devarg ``phy_mac`` was introduced to allow users to use physical 173 MAC address of the selected PCAP interface. 174 175* **Added TAP Rx/Tx queues sharing with a secondary process.** 176 177 A secondary process can attach a TAP device created in the primary process, 178 probe the queues, and process Rx/Tx in a secondary process. 179 180* **Added classification and metering support to SoftNIC PMD.** 181 182 Added support for flow classification (rte_flow API), and metering and 183 policing (rte_mtr API) to the SoftNIC PMD. 184 185* **Added Crypto support to Softnic PMD.** 186 187 The Softnic is now capable of processing symmetric crypto workloads such 188 as cipher, cipher-authentication chaining, and aead encryption and 189 decryption. This is achieved by calling DPDK Cryptodev APIs. 190 191* **Added cryptodev port to port library.** 192 193 Cryptodev port is a shim layer in the port library that interacts with DPDK 194 Cryptodev PMDs including burst enqueuing and dequeuing crypto operations. 195 196* **Added symmetric cryptographic actions to the pipeline library.** 197 198 In the pipeline library an added symmetric crypto action parsing and action 199 handler are implemented. The action allows automatically preparing the crypto 200 operation with the rules specified such as algorithm, key, and IV, etc for 201 the cryptodev port to process. 202 203* **Added support for GEN3 devices to Intel QAT driver .** 204 205 Added support for the third generation of Intel QuickAssist devices. 206 207* **Updated the QAT PMD.** 208 209 The QAT PMD was updated with additional support for: 210 211 * AES-CMAC algorithm. 212 213* **Updated the AESNI MB PMD.** 214 215 The AESNI MB PMD has been updated with additional support for AES-GCM 216 algorithm support. 217 218* **Added NXP CAAM JR PMD.** 219 220 Added the new caam job ring driver for NXP platforms. See the 221 "NXP CAAM JOB RING (caam_jr)" document for more details on this new driver. 222 223* **Added support for Dynamic Huffman Encoding to Intel QAT comp PMD.** 224 225 The Intel QuickAssist (QAT) compression PMD has been updated with support 226 for Dynamic Huffman Encoding for the Deflate algorithm. 227 228* **Added Event Ethernet Tx Adapter.** 229 230 Added event ethernet Tx adapter library that provides configuration and 231 data path APIs for the ethernet transmit stage of an event driven packet 232 processing application. These APIs abstract the implementation of the 233 transmit stage and allow the application to use eventdev PMD support or 234 a common implementation. 235 236* **Added Distributed Software Eventdev PMD.** 237 238 Added the new Distributed Software Event Device (DSW), which is a 239 pure-software eventdev driver distributing the work of scheduling 240 among all eventdev ports and the lcores using them. DSW, compared to 241 the SW eventdev PMD, sacrifices load balancing performance to 242 gain better event scheduling throughput and scalability. 243 244* **Added extendable bucket feature to hash library (rte_hash).** 245 246 This new “extendable bucket” feature provides 100% insertion guarantee to 247 the capacity specified by the user by extending hash table with extra 248 buckets when needed to accommodate the unlikely event of intensive hash 249 collisions. In addition, the internal hashing algorithm was changed to use 250 partial-key hashing to improve memory efficiency and lookup performance. 251 252* **Added lock free reader/writer concurrency to hash library (rte_hash).** 253 254 Lock free reader/writer concurrency prevents the readers from getting 255 blocked due to a pre-empted writer thread. This allows the hash library 256 to be used in scenarios where the writer thread runs on control plane. 257 258* **Added Traffic Pattern Aware Power Control Library** 259 260 Added an experimental library. This extend Power Library and provide 261 empty_poll APIs. This feature measure how many times empty_poll are 262 executed per core, use the number of empty polls as a hint for system 263 power management. 264 265 See the :doc:`../prog_guide/power_man` section of the DPDK Programmers 266 Guide document for more information. 267 268* **Added JSON power policy interface for containers.** 269 270 Extended the Power Library and vm_power_manager sample app to allow power 271 policies to be submitted via a FIFO using JSON formatted strings. Previously 272 limited to Virtual Machines, this feature extends power policy functionality 273 to containers and host applications that need to have their cores frequency 274 controlled based on the rules contained in the policy. 275 276* **Added Telemetry API.** 277 278 Added the telemetry API which allows applications to transparently expose 279 their telemetry via a UNIX socket in JSON. The JSON can be consumed by any 280 Service Assurance agent, such as CollectD. 281 282* **Updated KNI kernel module, rte_kni library, and KNI sample application.** 283 284 Updated the KNI kernel module with a new kernel module parameter, 285 ``carrier=[on|off]`` to allow the user to control the default carrier 286 state of KNI kernel network interfaces. The default carrier state 287 is now set to ``off``, so the interfaces cannot be used until the 288 carrier state is set to ``on`` via ``rte_kni_update_link`` or 289 by writing ``1`` to ``/sys/devices/virtual/net/<iface>/carrier``. 290 In previous versions the default carrier state was left undefined. 291 See :doc:`../prog_guide/kernel_nic_interface` for more information. 292 293 Added the new API function ``rte_kni_update_link`` to allow the user 294 to set the carrier state of the KNI kernel network interface. 295 296 Added a new command line flag ``-m`` to the KNI sample application to 297 monitor and automatically reflect the physical NIC carrier state to the 298 KNI kernel network interface with the new ``rte_kni_update_link`` API. 299 See :doc:`../sample_app_ug/kernel_nic_interface` for more information. 300 301* **Added ability to switch queue deferred start flag on testpmd app.** 302 303 Added a console command to testpmd app, giving ability to switch 304 ``rx_deferred_start`` or ``tx_deferred_start`` flag of the specified queue of 305 the specified port. The port must be stopped before the command call in order 306 to reconfigure queues. 307 308* **Add a new sample for vDPA** 309 310 The vdpa sample application creates vhost-user sockets by using the 311 vDPA backend. vDPA stands for vhost Data Path Acceleration which utilizes 312 virtio ring compatible devices to serve virtio driver directly to enable 313 datapath acceleration. As vDPA driver can help to set up vhost datapath, 314 this application doesn't need to launch dedicated worker threads for vhost 315 enqueue/dequeue operations. 316 317* **Added cryptodev FIPS validation example application.** 318 319 Added an example application to parse and perform symmetric cryptography 320 computation to the NIST Cryptographic Algorithm Validation Program (CAVP) 321 test vectors. 322 323* **Allow unit test binary to take parameters from the environment** 324 325 The unit test "test", or "dpdk-test", binary is often called from scripts, 326 which can make passing additional parameters, such as a coremask, to it more 327 awkward. Support has been added to the application to allow it to take 328 additional command-line parameter values from the "DPDK_TEST_PARAMS" 329 environment variable to make this application easier to use. 330 331 332API Changes 333----------- 334 335.. This section should contain API changes. Sample format: 336 337 * Add a short 1-2 sentence description of the API change. 338 Use fixed width quotes for ``function_names`` or ``struct_names``. 339 Use the past tense. 340 341 This section is a comment. Do not overwrite or remove it. 342 Also, make sure to start the actual text at the margin. 343 ========================================================= 344 345* eal: ``rte_memseg_list`` structure now has an additional flag indicating 346 whether the memseg list is externally allocated. This will have implications 347 for any users of memseg-walk-related functions, as they will now have to skip 348 externally allocated segments in most cases if the intent is to only iterate 349 over internal DPDK memory. 350 ``socket_id`` parameter across the entire DPDK has gained additional meaning, 351 as some socket ID's will now be representing externally allocated memory. No 352 changes will be required for existing code as backwards compatibility will be 353 kept, and those who do not use this feature will not see these extra socket 354 ID's. Any new API's must not check socket ID parameters themselves, and must 355 instead leave it to the memory subsystem to decide whether socket ID is a 356 valid one. 357 358* eal: The following devargs functions, which were deprecated in 18.05, 359 were removed in 18.11: 360 ``rte_eal_parse_devargs_str()``, ``rte_eal_devargs_add()``, 361 ``rte_eal_devargs_type_count()``, and ``rte_eal_devargs_dump()``. 362 363* eal: The parameters of the function ``rte_devargs_remove()`` have changed 364 from bus and device names to ``struct rte_devargs``. 365 366* eal: The deprecated functions attach/detach were removed in 18.11. 367 ``rte_eal_dev_attach`` can be replaced by 368 ``rte_dev_probe`` or ``rte_eal_hotplug_add``. 369 ``rte_eal_dev_detach`` can be replaced by 370 ``rte_dev_remove`` or ``rte_eal_hotplug_remove``. 371 372* eal: The scope of ``rte_eal_hotplug_add()``/``rte_dev_probe()`` 373 and ``rte_eal_hotplug_remove()``/``rte_dev_remove()`` is extended. 374 In multi-process model, they will guarantee that the device is 375 attached or detached on all processes. 376 377* mbuf: The ``__rte_mbuf_raw_free()`` and ``__rte_pktmbuf_prefree_seg()`` 378 functions were deprecated since 17.05 and are replaced by 379 ``rte_mbuf_raw_free()`` and ``rte_pktmbuf_prefree_seg()``. 380 381* ethdev: The deprecated functions attach/detach were removed in 18.11. 382 ``rte_eth_dev_attach`` can be replaced by ``RTE_ETH_FOREACH_MATCHING_DEV`` 383 and ``rte_dev_probe`` or ``rte_eal_hotplug_add``. 384 ``rte_eth_dev_detach`` can be replaced by 385 ``rte_dev_remove`` or ``rte_eal_hotplug_remove``. 386 387* ethdev: A call to ``rte_eth_dev_release_port()`` has been added in 388 ``rte_eth_dev_close()``. As a consequence, a closed port is freed 389 and seen as invalid because of its state ``RTE_ETH_DEV_UNUSED``. 390 This new behaviour is enabled per driver for a migration period. 391 392* A new device flag, RTE_ETH_DEV_NOLIVE_MAC_ADDR, changes the order of 393 actions inside rte_eth_dev_start regarding MAC set. Some NICs do not 394 support MAC changes once the port has started and with this new device 395 flag the MAC can be properly configured in any case. This is particularly 396 important for bonding. 397 398* The default behaviour of CRC strip offload changed. Without any specific Rx 399 offload flag, default behavior by PMD is now to strip CRC. 400 DEV_RX_OFFLOAD_CRC_STRIP offload flag has been removed. 401 To request keeping CRC, application should set ``DEV_RX_OFFLOAD_KEEP_CRC`` Rx 402 offload. 403 404* eventdev: Type of 2nd parameter to ``rte_event_eth_rx_adapter_caps_get()`` 405 has been changed from uint8_t to uint16_t. 406 407* kni: By default, interface carrier status is ``off`` which means there won't 408 be any traffic. It can be set to ``on`` via ``rte_kni_update_link()`` API 409 or via ``sysfs`` interface: 410 ``echo 1 > /sys/class/net/vEth0/carrier``. 411 Note interface should be ``up`` to be able to read/write sysfs interface. 412 When KNI sample application is used, ``-m`` parameter can be used to 413 automatically update the carrier status for the interface. 414 415* kni: When ethtool support enabled (``CONFIG_RTE_KNI_KMOD_ETHTOOL=y``) 416 ethtool commands ``ETHTOOL_GSET & ETHTOOL_SSET`` are no more supported for the 417 kernels that has ``ETHTOOL_GLINKSETTINGS & ETHTOOL_SLINKSETTINGS`` support. 418 This means ``ethtool "-a|--show-pause", "-s|--change"`` won't work, and 419 ``ethtool <iface>`` output will have less information. 420 421 422ABI Changes 423----------- 424 425.. This section should contain ABI changes. Sample format: 426 427 * Add a short 1-2 sentence description of the ABI change 428 that was announced in the previous releases and made in this release. 429 Use fixed width quotes for ``function_names`` or ``struct_names``. 430 Use the past tense. 431 432 This section is a comment. Do not overwrite or remove it. 433 Also, make sure to start the actual text at the margin. 434 ========================================================= 435 436* eal: added ``legacy_mem`` and ``single_file_segments`` values to 437 ``rte_config`` structure on account of improving DPDK usability when 438 using either ``--legacy-mem`` or ``--single-file-segments`` flags. 439 440* eal: EAL library ABI version was changed due to previously announced work on 441 supporting external memory in DPDK: 442 - structure ``rte_memseg_list`` now has a new field indicating length 443 of memory addressed by the segment list 444 - structure ``rte_memseg_list`` now has a new flag indicating whether 445 the memseg list refers to external memory 446 - structure ``rte_malloc_heap`` now has a new field indicating socket 447 ID the malloc heap belongs to 448 - structure ``rte_mem_config`` has had its ``malloc_heaps`` array 449 resized from ``RTE_MAX_NUMA_NODES`` to ``RTE_MAX_HEAPS`` value 450 - structure ``rte_malloc_heap`` now has a ``heap_name`` member 451 - structure ``rte_eal_memconfig`` has been extended to contain next 452 socket ID for externally allocated segments 453 454* eal: Added ``dma_maskbits`` to ``rte_mem_config`` for keeping more restricted 455 dma mask based on devices addressing limitations. 456 457* eal: The structure ``rte_device`` got a new field to reference a ``rte_bus``. 458 It is changing the size of the ``struct rte_device`` and the inherited 459 device structures of all buses. 460 461 462Removed Items 463------------- 464 465.. This section should contain removed items in this release. Sample format: 466 467 * Add a short 1-2 sentence description of the removed item 468 in the past tense. 469 470 This section is a comment. Do not overwrite or remove it. 471 Also, make sure to start the actual text at the margin. 472 ========================================================= 473 474 475Shared Library Versions 476----------------------- 477 478.. Update any library version updated in this release 479 and prepend with a ``+`` sign, like this: 480 481 librte_acl.so.2 482 + librte_cfgfile.so.2 483 librte_cmdline.so.2 484 485 This section is a comment. Do not overwrite or remove it. 486 ========================================================= 487 488The libraries prepended with a plus sign were incremented in this version. 489 490.. code-block:: diff 491 492 librte_acl.so.2 493 librte_bbdev.so.1 494 librte_bitratestats.so.2 495 librte_bpf.so.1 496 + librte_bus_dpaa.so.2 497 + librte_bus_fslmc.so.2 498 + librte_bus_ifpga.so.2 499 + librte_bus_pci.so.2 500 + librte_bus_vdev.so.2 501 + librte_bus_vmbus.so.2 502 librte_cfgfile.so.2 503 librte_cmdline.so.2 504 librte_compressdev.so.1 505 librte_cryptodev.so.5 506 librte_distributor.so.1 507 + librte_eal.so.9 508 librte_efd.so.1 509 + librte_ethdev.so.11 510 + librte_eventdev.so.6 511 librte_flow_classify.so.1 512 librte_gro.so.1 513 librte_gso.so.1 514 librte_hash.so.2 515 librte_ip_frag.so.1 516 librte_jobstats.so.1 517 librte_kni.so.2 518 librte_kvargs.so.1 519 librte_latencystats.so.1 520 librte_lpm.so.2 521 librte_mbuf.so.4 522 librte_member.so.1 523 librte_mempool.so.5 524 librte_meter.so.2 525 librte_metrics.so.1 526 librte_net.so.1 527 librte_pci.so.1 528 librte_pdump.so.2 529 librte_pipeline.so.3 530 librte_pmd_bnxt.so.2 531 librte_pmd_bond.so.2 532 librte_pmd_i40e.so.2 533 librte_pmd_ixgbe.so.2 534 librte_pmd_dpaa2_qdma.so.1 535 librte_pmd_ring.so.2 536 librte_pmd_softnic.so.1 537 librte_pmd_vhost.so.2 538 librte_port.so.3 539 librte_power.so.1 540 librte_rawdev.so.1 541 librte_reorder.so.1 542 librte_ring.so.2 543 librte_sched.so.1 544 librte_security.so.1 545 librte_table.so.3 546 librte_timer.so.1 547 librte_vhost.so.4 548 549 550Known Issues 551------------ 552 553.. This section should contain new known issues in this release. Sample format: 554 555 * **Add title in present tense with full stop.** 556 557 Add a short 1-2 sentence description of the known issue 558 in the present tense. Add information on any known workarounds. 559 560 This section is a comment. Do not overwrite or remove it. 561 Also, make sure to start the actual text at the margin. 562 ========================================================= 563 564* When using SR-IOV (VF) support with netvsc PMD and the Mellanox mlx5 bifurcated 565 driver; the Linux netvsc device must be brought up before the netvsc device is 566 unbound and passed to the DPDK. 567 568* IBM Power8 is not supported by this release of DPDK. IBM Power9 is supported. 569 570 571Tested Platforms 572---------------- 573 574.. This section should contain a list of platforms that were tested 575 with this release. 576 577 The format is: 578 579 * <vendor> platform with <vendor> <type of devices> combinations 580 581 * List of CPU 582 * List of OS 583 * List of devices 584 * Other relevant details... 585 586 This section is a comment. Do not overwrite or remove it. 587 Also, make sure to start the actual text at the margin. 588 ========================================================= 589