1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright 2018 The DPDK contributors
4 ABI and API Deprecation
5 =======================
7 See the guidelines document for details of the :doc:`ABI policy
8 <../contributing/abi_policy>`. API and ABI deprecation notices are to be posted
14 * kvargs: The function ``rte_kvargs_process`` will get a new parameter
15 for returning key match count. It will ease handling of no-match case.
17 * eal: The function ``rte_eal_remote_launch`` will return new error codes
18 after read or write error on the pipe, instead of calling ``rte_panic``.
20 * rte_atomicNN_xxx: These APIs do not take memory order parameter. This does
21 not allow for writing optimized code for all the CPU architectures supported
22 in DPDK. DPDK has adopted the atomic operations from
23 https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html. These
24 operations must be used for patches that need to be merged in 20.08 onwards.
25 This change will not introduce any performance degradation.
27 * rte_smp_*mb: These APIs provide full barrier functionality. However, many
28 use cases do not require full barriers. To support such use cases, DPDK has
29 adopted atomic operations from
30 https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html. These
31 operations and a new wrapper ``rte_atomic_thread_fence`` instead of
32 ``__atomic_thread_fence`` must be used for patches that need to be merged in
33 20.08 onwards. This change will not introduce any performance degradation.
35 * mbuf: The mbuf offload flags ``PKT_*`` will be renamed as ``RTE_MBUF_F_*``.
36 A compatibility layer will be kept until DPDK 22.11, except for the flags
37 that are already deprecated (``PKT_RX_L4_CKSUM_BAD``, ``PKT_RX_IP_CKSUM_BAD``,
38 ``PKT_RX_EIP_CKSUM_BAD``, ``PKT_TX_QINQ_PKT``) which will be removed.
40 * pci: To reduce unnecessary ABIs exposed by DPDK bus driver, "rte_bus_pci.h"
41 will be made internal in 21.11 and macros/data structures/functions defined
42 in the header will not be considered as ABI anymore. This change is inspired
43 by the RFC https://patchwork.dpdk.org/project/dpdk/list/?series=17176.
45 * lib: will fix extending some enum/define breaking the ABI. There are multiple
46 samples in DPDK that enum/define terminated with a ``.*MAX.*`` value which is
47 used by iterators, and arrays holding these values are sized with this
48 ``.*MAX.*`` value. So extending this enum/define increases the ``.*MAX.*``
49 value which increases the size of the array and depending on how/where the
50 array is used this may break the ABI.
51 ``RTE_ETH_FLOW_MAX`` is one sample of the mentioned case, adding a new flow
52 type will break the ABI because of ``flex_mask[RTE_ETH_FLOW_MAX]`` array
53 usage in following public struct hierarchy:
54 ``rte_eth_fdir_flex_conf -> rte_fdir_conf -> rte_eth_conf (in the middle)``.
55 Need to identify this kind of usages and fix in 20.11, otherwise this blocks
56 us extending existing enum/define.
57 One solution can be using a fixed size array instead of ``.*MAX.*`` value.
59 * ethdev: Will add ``RTE_ETH_`` prefix to all ethdev macros/enums in v21.11.
60 Macros will be added for backward compatibility.
61 Backward compatibility macros will be removed on v22.11.
62 A few old backward compatibility macros from 2013 that does not have
63 proper prefix will be removed on v21.11.
65 * ethdev: The flow director API, including ``rte_eth_conf.fdir_conf`` field,
66 and the related structures (``rte_fdir_*`` and ``rte_eth_fdir_*``),
67 will be removed in DPDK 20.11.
69 * ethdev: New offload flags ``DEV_RX_OFFLOAD_FLOW_MARK`` will be added in 19.11.
70 This will allow application to enable or disable PMDs from updating
71 ``rte_mbuf::hash::fdir``.
72 This scheme will allow PMDs to avoid writes to ``rte_mbuf`` fields on Rx and
73 thereby improve Rx performance if application wishes do so.
74 In 19.11 PMDs will still update the field even when the offload is not
77 * ethdev: ``uint32_t max_rx_pkt_len`` field of ``struct rte_eth_rxmode``, will be
78 replaced by a new ``uint32_t mtu`` field of ``struct rte_eth_conf`` in v21.11.
79 The new ``mtu`` field will be used to configure the initial device MTU via
80 ``rte_eth_dev_configure()`` API.
81 Later MTU can be changed by ``rte_eth_dev_set_mtu()`` API as done now.
82 The existing ``(struct rte_eth_dev)->data->mtu`` variable will be used to store
83 the configured ``mtu`` value,
84 and this new ``(struct rte_eth_dev)->data->dev_conf.mtu`` variable will
85 be used to store the user configuration request.
86 Unlike ``max_rx_pkt_len``, which was valid only when ``JUMBO_FRAME`` enabled,
87 ``mtu`` field will be always valid.
88 When ``mtu`` config is not provided by the application, default ``RTE_ETHER_MTU``
90 ``(struct rte_eth_dev)->data->mtu`` should be updated after MTU set successfully,
91 either by ``rte_eth_dev_configure()`` or ``rte_eth_dev_set_mtu()``.
93 An application may need to configure device for a specific Rx packet size, like for
94 cases ``DEV_RX_OFFLOAD_SCATTER`` is not supported and device received packet size
95 can't be bigger than Rx buffer size.
96 To cover these cases an application needs to know the device packet overhead to be
97 able to calculate the ``mtu`` corresponding to a Rx buffer size, for this
98 ``(struct rte_eth_dev_info).max_rx_pktlen`` will be kept,
99 the device packet overhead can be calculated as:
100 ``(struct rte_eth_dev_info).max_rx_pktlen - (struct rte_eth_dev_info).max_mtu``
102 * ethdev: ``rx_descriptor_done`` dev_ops and ``rte_eth_rx_descriptor_done``
103 will be removed in 21.11.
104 Existing ``rte_eth_rx_descriptor_status`` and ``rte_eth_tx_descriptor_status``
105 APIs can be used as replacement.
107 * ethdev: The port mirroring API can be replaced with a more fine grain flow API.
108 The structs ``rte_eth_mirror_conf``, ``rte_eth_vlan_mirror`` and the functions
109 ``rte_eth_mirror_rule_set``, ``rte_eth_mirror_rule_reset`` will be marked
110 as deprecated in DPDK 20.11, along with the associated macros ``ETH_MIRROR_*``.
111 This API will be fully removed in DPDK 21.11.
113 * ethdev: Attribute ``shared`` of the ``struct rte_flow_action_count``
114 is deprecated and will be removed in DPDK 21.11. Shared counters should
115 be managed using shared actions API (``rte_flow_shared_action_create`` etc).
117 * ethdev: The flow API matching pattern structures, ``struct rte_flow_item_*``,
118 should start with relevant protocol header.
119 Some matching pattern structures implements this by duplicating protocol header
120 fields in the struct. To clarify the intention and to be sure protocol header
121 is intact, will replace those fields with relevant protocol header struct.
122 In v21.02 both individual protocol header fields and the protocol header struct
123 will be added as union, target is switch usage to the protocol header by time.
124 In v21.11 LTS, protocol header fields will be cleaned and only protocol header
127 * ethdev: Queue specific stats fields will be removed from ``struct rte_eth_stats``.
128 Mentioned fields are: ``q_ipackets``, ``q_opackets``, ``q_ibytes``, ``q_obytes``,
130 Instead queue stats will be received via xstats API. Current method support
131 will be limited to maximum 256 queues.
132 Also compile time flag ``RTE_ETHDEV_QUEUE_STAT_CNTRS`` will be removed.
134 * ethdev: The offload flag ``PKT_RX_EIP_CKSUM_BAD`` will be removed and
135 replaced by the new flag ``PKT_RX_OUTER_IP_CKSUM_BAD``. The new name is more
136 consistent with existing outer header checksum status flag naming, which
137 should help in reducing confusion about its usage.
139 * i40e: As there are both i40evf and iavf pmd, the functions of them are
140 duplicated. And now more and more advanced features are developed on iavf.
141 To keep consistent with kernel driver's name
142 (https://patchwork.ozlabs.org/patch/970154/), i40evf is no need to maintain.
143 Starting from 21.05, the default VF driver of i40e will be iavf, but i40evf
144 can still be used if users specify the devarg "driver=i40evf". I40evf will
145 be deleted in DPDK 21.11.
147 * net: ``s_addr`` and ``d_addr`` fields of ``rte_ether_hdr`` structure
148 will be renamed in DPDK 21.11 to avoid conflict with Windows Sockets headers.
150 * compressdev: ``min`` and ``max`` fields of ``rte_param_log2_range`` structure
151 will be renamed in DPDK 21.11 to avoid conflict with Windows Sockets headers.
153 * cryptodev: ``min`` and ``max`` fields of ``rte_crypto_param_range`` structure
154 will be renamed in DPDK 21.11 to avoid conflict with Windows Sockets headers.
156 * cryptodev: The field ``dataunit_len`` of the ``struct rte_crypto_cipher_xform``
157 has a limited size ``uint16_t``.
158 It will be moved and extended as ``uint32_t`` in DPDK 21.11.
160 * eventdev: The structure ``rte_event_eth_rx_adapter_queue_conf`` will be
161 extended to include ``rte_event_eth_rx_adapter_event_vector_config`` elements
162 and the function ``rte_event_eth_rx_adapter_queue_event_vector_config`` will
163 be removed in DPDK 21.11.
165 An application can enable event vectorization by passing the desired vector
166 values to the function ``rte_event_eth_rx_adapter_queue_add`` using
167 the structure ``rte_event_eth_rx_adapter_queue_add``.
169 * metrics: The function ``rte_metrics_init`` will have a non-void return
170 in order to notify errors instead of calling ``rte_exit``.
172 * cmdline: ``cmdline`` structure will be made opaque to hide platform-specific
173 content. On Linux and FreeBSD, supported prior to DPDK 20.11,
174 original structure will be kept until DPDK 21.11.
176 * security: The functions ``rte_security_set_pkt_metadata`` and
177 ``rte_security_get_userdata`` will be made inline functions and additional
178 flags will be added in structure ``rte_security_ctx`` in DPDK 21.11.