1 ABI and API Deprecation
2 =======================
4 See the :doc:`guidelines document for details of the ABI policy </contributing/versioning>`.
5 API and ABI deprecation notices are to be posted here.
11 * eal: both declaring and identifying devices will be streamlined in v18.05.
12 New functions will appear to query a specific port from buses, classes of
13 device and device drivers. Device declaration will be made coherent with the
14 new scheme of device identification.
15 As such, ``rte_devargs`` device representation will change.
17 - removal of ``name`` and ``args`` fields.
18 - The enum ``rte_devtype`` was used to identify a bus and will disappear.
19 - The ``rte_devargs_list`` will be made private.
20 - Functions previously deprecated will change or disappear:
22 + ``rte_eal_devargs_add``
23 + ``rte_eal_devargs_type_count``
24 + ``rte_eal_parse_devargs_str``, replaced by ``rte_eal_devargs_parse``
25 + ``rte_eal_devargs_parse`` will change its format and use.
26 + all ``rte_devargs`` related functions will be renamed, changing the
27 ``rte_eal_devargs_`` prefix to ``rte_devargs_``.
29 * pci: Several exposed functions are misnamed.
30 The following functions are deprecated starting from v17.11 and are replaced:
32 - ``eal_parse_pci_BDF`` replaced by ``rte_pci_addr_parse``
33 - ``eal_parse_pci_DomBDF`` replaced by ``rte_pci_addr_parse``
34 - ``rte_eal_compare_pci_addr`` replaced by ``rte_pci_addr_cmp``
36 * eal: The semantics of the return value for the ``rte_lcore_has_role`` function
37 are planned to change in v18.05. The function currently returns 0 and <0 for
38 success and failure, respectively. This will change to 1 and 0 for true and
39 false, respectively, to make use of the function more intuitive.
41 * eal: a new set of mbuf mempool ops name APIs for user, platform and best
42 mempool names have been defined in ``rte_mbuf`` in v18.02. The uses of
43 ``rte_eal_mbuf_default_mempool_ops`` shall be replaced by
44 ``rte_mbuf_best_mempool_ops``.
45 The following function is now redundant and it is target to be deprecated
48 - ``rte_eal_mbuf_default_mempool_ops``
50 * mempool: several API and ABI changes are planned in v18.05.
51 The following functions, introduced for Xen, which is not supported
52 anymore since v17.11, are hard to use, not used anywhere else in DPDK.
53 Therefore they will be deprecated in v18.05 and removed in v18.08:
55 - ``rte_mempool_xmem_create``
56 - ``rte_mempool_xmem_size``
57 - ``rte_mempool_xmem_usage``
59 The following changes are planned:
61 - removal of ``get_capabilities`` mempool ops and related flags.
62 - substitute ``register_memory_area`` with ``populate`` ops.
63 - addition of new ops to customize required memory chunk calculation,
64 customize objects population and allocate contiguous
65 block of objects if underlying driver supports it.
67 * mbuf: The opaque ``mbuf->hash.sched`` field will be updated to support generic
68 definition in line with the ethdev TM and MTR APIs. Currently, this field
69 is defined in librte_sched in a non-generic way. The new generic format
70 will contain: queue ID, traffic class, color. Field size will not change.
72 * ethdev: a new Tx and Rx offload API was introduced on 17.11.
73 In the new API, offloads are divided into per-port and per-queue offloads.
74 Offloads are disabled by default and enabled per application request.
76 The old ethdev - drivers offload interface will be deprecated on 18.05.
78 - removal of the conversion in ethdev from new offloading API to old API for drivers.
80 In later releases the old offloading API will be deprecated, which will include:
81 - removal of ``ETH_TXQ_FLAGS_NO*`` flags.
82 - removal of ``txq_flags`` field from ``rte_eth_txconf`` struct.
83 - removal of the offloads bit-field from ``rte_eth_rxmode`` struct.
85 * ethdev: the legacy filter API, including
86 ``rte_eth_dev_filter_supported()``, ``rte_eth_dev_filter_ctrl()`` as well
87 as filter types MACVLAN, ETHERTYPE, FLEXIBLE, SYN, NTUPLE, TUNNEL, FDIR,
88 HASH and L2_TUNNEL, is superseded by the generic flow API (rte_flow) in
89 PMDs that implement the latter.
90 Target release for removal of the legacy API will be defined once most
91 PMDs have switched to rte_flow.
93 * ethdev: A new rss level field planned in 18.05.
94 The new API add rss_level field to ``rte_eth_rss_conf`` to enable a choice
95 of RSS hash calculation on outer or inner header of tunneled packet.
97 * ethdev: Currently, if the rte_eth_rx_burst() function returns a value less
98 than *nb_pkts*, the application will assume that no more packets are present.
99 Some of the hw queue based hardware can only support smaller burst for RX
100 and TX and thus break the expectation of the rx_burst API. Similar is the
101 case for TX burst as well as ring sizes. ``rte_eth_dev_info`` will be added
102 with following new parameters so as to support semantics for drivers to
103 define a preferred size for Rx/Tx burst and rings.
105 - Member ``struct preferred_size`` would be added to enclose all preferred
106 size to be fetched from driver/implementation.
107 - Members ``uint16_t rx_burst``, ``uint16_t tx_burst``, ``uint16_t rx_ring``,
108 and ``uint16_t tx_ring`` would be added to ``struct preferred_size``.
110 * ethdev: A work is being planned for 18.05 to expose VF port representors
111 as a mean to perform control and data path operation on the different VFs.
112 As VF representor is an ethdev port, new fields are needed in order to map
113 between the VF representor and the VF or the parent PF. Those new fields
114 are to be included in ``rte_eth_dev_info`` struct.
116 * ethdev: The prototype and the behavior of
117 ``dev_ops->eth_mac_addr_set()`` will change in v18.05. A return code
118 will be added to notify the caller if an error occurred in the PMD. In
119 ``rte_eth_dev_default_mac_addr_set()``, the new default MAC address
120 will be copied in ``dev->data->mac_addrs[0]`` only if the operation is
121 successful. This modification will only impact the PMDs, not the
124 * i40e: The default flexible payload configuration which extracts the first 16
125 bytes of the payload for RSS will be deprecated starting from 18.02. If
126 required the previous behavior can be configured using existing flow
127 director APIs. There is no ABI/API break. This change will just remove a
128 global configuration setting and require explicit configuration.
130 * ring: The alignment constraints on the ring structure will be relaxed
131 to one cache line instead of two, and an empty cache line padding will
132 be added between the producer and consumer structures. The size of the
133 structure and the offset of the fields will remain the same on
134 platforms with 64B cache line, but will change on other platforms.