1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright 2018 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_19_02.html
26 .. This section should contain new features added in this release.
29 * **Add a title in the past tense with a full stop.**
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.
35 If the feature adds a lot of sub-features you can use a bullet list
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 Suggested order in release notes items:
44 * Core libs (EAL, mempool, ring, mbuf, buses)
45 * Device abstraction libs and PMDs
47 - cryptodev (lib, PMDs)
48 - eventdev (lib, PMDs)
51 * Apps, Examples, Tools (if significant)
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 =========================================================
57 * **Added support to free hugepages exactly as originally allocated.**
59 Some applications using memory event callbacks (especially for managing
60 RDMA memory regions) require that memory be freed back to the system
61 exactly as it was originally allocated. These applications typically
62 also require that a malloc allocation not span across two separate
63 hugepage allocations. A new ``--match-allocations`` EAL init flag has
64 been added to fulfill both of these requirements.
66 * **Added API to register external memory in DPDK.**
68 A new ``rte_extmem_register``/``rte_extmem_unregister`` API was added to allow
69 chunks of external memory to be registered with DPDK without adding them to
72 * **Support for using virtio-user without hugepages**
74 The --no-huge mode was augmented to use memfd-backed memory (on systems that
75 support memfd), to allow using virtio-user-based NICs without hugepages.
77 * **Updated the enic driver.**
79 * Added support for ``RTE_ETH_DEV_CLOSE_REMOVE`` flag.
80 * Added the handler to get firmware version string.
81 * Added support for multicast filtering.
83 * **Added dynamic queues allocation support for i40e VF.**
85 Previously, available queues of VF is reserved by PF at initialize stage.
86 Now both DPDK PF and Kernel PF (>=2.1.14) will support dynamic queue
87 allocation. At runtime, when VF request more queue number exceed the initial
88 reserved amount, PF can allocate up to 16 queues as the request after a VF
91 * **Added ICE net PMD**
93 Added the new ``ice`` net driver for IntelĀ® Ethernet Network Adapters E810.
94 See the :doc:`../nics/ice` NIC guide for more details on this new driver.
96 * **Added support for SW-assisted VDPA live migration.**
98 This SW-assisted VDPA live migration facility helps VDPA devices without
99 logging capability to perform live migration, a mediated SW relay can help
100 devices to track dirty pages caused by DMA. IFC driver has enabled this
101 SW-assisted live migration mode.
103 * **Added security checks to cryptodev symmetric session operations.**
105 Added a set of security checks to the access cryptodev symmetric session.
106 The checks include the session's user data read/write check and the
107 session private data referencing status check while freeing a session.
109 * **Updated the AESNI-MB PMD.**
111 * Add support for intel-ipsec-mb version 0.52.
112 * Add AES-GMAC algorithm support.
113 * Add Plain SHA1, SHA224, SHA256, SHA384, and SHA512 algorithms support.
115 * **Added IPsec Library.**
117 Added an experimental library ``librte_ipsec`` to provide ESP tunnel and
118 transport support for IPv4 and IPv6 packets.
120 The library provides support for AES-CBC ciphering and AES-CBC with HMAC-SHA1
121 algorithm-chaining, and AES-GCM and NULL algorithms only at present. It is
122 planned to add more algorithms in future releases.
124 See :doc:`../prog_guide/ipsec_lib` for more information.
126 * **Updated the ipsec-secgw sample application.**
128 The ``ipsec-secgw`` sample application has been updated to use the new
129 ``librte_ipsec`` library also added in this release.
130 The original functionality of ipsec-secgw is retained, a new command line
131 parameter ``-l`` has been added to ipsec-secgw to use the IPsec library,
132 instead of the existing IPsec code in the application.
134 The IPsec library does not support all the functionality of the existing
135 ipsec-secgw application, its is planned to add the outstanding functionality
138 See :doc:`../sample_app_ug/ipsec_secgw` for more information.
140 * **Enabled checksum support in the ISA-L compressdev driver.**
142 Added support for both adler and crc32 checksums in the ISA-L PMD.
143 This aids data integrity across both compression and decompression.
145 * **Added a compression performance test tool.**
147 Added a new performance test tool to test the compressdev PMD. The tool tests
148 compression ratio and compression throughput.
150 * **Added intel_pstate support to Power Management library.**
152 Previously, using the power management library required the
153 disabling of the intel_pstate kernel driver, and the enabling of the
154 acpi_cpufreq kernel driver. This is no longer the case, as the use of
155 the intel_pstate kernel driver is now supported, and automatically
156 detected by the library.
162 .. This section should contain removed items in this release. Sample format:
164 * Add a short 1-2 sentence description of the removed item
167 This section is a comment. Do not overwrite or remove it.
168 Also, make sure to start the actual text at the margin.
169 =========================================================
175 .. This section should contain API changes. Sample format:
177 * sample: Add a short 1-2 sentence description of the API change
178 which was announced in the previous releases and made in this release.
179 Start with a scope label like "ethdev:".
180 Use fixed width quotes for ``function_names`` or ``struct_names``.
183 This section is a comment. Do not overwrite or remove it.
184 Also, make sure to start the actual text at the margin.
185 =========================================================
187 * eal: Function ``rte_bsf64`` in ``rte_bitmap.h`` has been renamed to
188 ``rte_bsf64_safe`` and moved to ``rte_common.h``. A new ``rte_bsf64`` function
189 has been added in ``rte_common.h`` that follows convention set by existing
190 ``rte_bsf32`` function.
192 * eal: Segment fd API on Linux now sets error code to ``ENOTSUP`` in more cases
193 where segment fd API is not expected to be supported:
195 - On attempt to get segment fd for an externally allocated memory segment
196 - In cases where memfd support would have been required to provide segment
197 fd's (such as in-memory or no-huge mode)
199 * eal: Functions ``rte_malloc_dump_stats()``, ``rte_malloc_dump_heaps()`` and
200 ``rte_malloc_get_socket_stats()`` are no longer safe to call concurrently with
201 ``rte_malloc_heap_create()`` or ``rte_malloc_heap_destroy()`` function calls.
203 * mbuf: ``RTE_MBUF_INDIRECT()``, which was deprecated in 18.05, was replaced
204 with ``RTE_MBUF_CLONED()`` and removed in 19.02.
206 * sched: As result of the new format of the mbuf sched field, the
207 functions ``rte_sched_port_pkt_write()`` and
208 ``rte_sched_port_pkt_read_tree_path()`` got an additional parameter of
209 type ``struct rte_sched_port``.
211 * pdump: The ``rte_pdump_set_socket_dir()``, the parameter ``path`` of
212 ``rte_pdump_init()`` and enum ``rte_pdump_socktype`` were deprecated
213 since 18.05 and are removed in this release.
215 * cryptodev: The parameter ``session_pool`` in the function
216 ``rte_cryptodev_queue_pair_setup()`` is removed.
218 * cryptodev: a new function ``rte_cryptodev_sym_session_pool_create()`` is
219 introduced. This function is now mandatory when creating symmetric session
220 header mempool. Please note all crypto applications are required to use this
221 function from now on. Failed to do so will cause
222 ``rte_cryptodev_sym_session_create()`` function call return error.
228 .. This section should contain ABI changes. Sample format:
230 * sample: Add a short 1-2 sentence description of the ABI change
231 which was announced in the previous releases and made in this release.
232 Start with a scope label like "ethdev:".
233 Use fixed width quotes for ``function_names`` or ``struct_names``.
236 This section is a comment. Do not overwrite or remove it.
237 Also, make sure to start the actual text at the margin.
238 =========================================================
240 * mbuf: The format of the sched field of ``rte_mbuf`` has been changed
241 to include the following fields: ``queue ID``, ``traffic class``, ``color``.
243 * cryptodev: as shown in the the 18.11 deprecation notice, the structure
244 ``rte_cryptodev_qp_conf`` has been added two parameters of symmetric session
245 mempool and symmetric session private data mempool.
247 * cryptodev: as shown in the the 18.11 deprecation notice, the structure
248 ``rte_cryptodev_sym_session`` has been updated to contain more information
249 to ensure safely accessing the session and session private data.
251 * security: New field ``uint64_t opaque_data`` is added into
252 ``rte_security_session`` structure. That would allow upper layer to easily
253 associate/de-associate some user defined data with the security session.
256 Shared Library Versions
257 -----------------------
259 .. Update any library version updated in this release
260 and prepend with a ``+`` sign, like this:
266 This section is a comment. Do not overwrite or remove it.
267 =========================================================
269 The libraries prepended with a plus sign were incremented in this version.
275 librte_bitratestats.so.2
278 librte_bus_fslmc.so.2
279 librte_bus_ifpga.so.2
282 librte_bus_vmbus.so.2
285 librte_compressdev.so.1
286 + librte_cryptodev.so.6
287 librte_distributor.so.1
292 librte_flow_classify.so.1
300 librte_latencystats.so.1
314 librte_pmd_ixgbe.so.2
315 librte_pmd_dpaa2_qdma.so.1
317 librte_pmd_softnic.so.1
318 librte_pmd_vhost.so.2
325 + librte_security.so.2
334 .. This section should contain new known issues in this release. Sample format:
336 * **Add title in present tense with full stop.**
338 Add a short 1-2 sentence description of the known issue
339 in the present tense. Add information on any known workarounds.
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 =========================================================
345 * ``AVX-512`` support has been disabled for ``GCC`` builds when ``binutils 2.30``
346 is detected [1] because of a crash [2]. This can affect ``native`` machine type
347 build targets on the platforms that support ``AVX512F`` like ``Intel Skylake``
348 processors, and can cause a possible performance drop. The immediate workaround
349 is to use ``clang`` compiler on these platforms.
350 Initial workaround in DPDK v18.11 was to disable ``AVX-512`` support for ``GCC``
351 completely, but based on information on defect submitted to GCC community [3],
352 issue has been identified as ``binutils 2.30`` issue. Since currently only GCC
353 generates ``AVX-512`` instructions, the scope is limited to ``GCC`` and
356 - [1]: Commit ("mk: fix scope of disabling AVX512F support")
357 - [2]: https://bugs.dpdk.org/show_bug.cgi?id=97
358 - [3]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88096
364 .. This section should contain a list of platforms that were tested
369 * <vendor> platform with <vendor> <type of devices> combinations
374 * Other relevant details...
376 This section is a comment. Do not overwrite or remove it.
377 Also, make sure to start the actual text at the margin.
378 =========================================================