1 /* SPDX-License-Identifier: BSD-3-Clause
2 * Copyright(c) 2010-2017 Intel Corporation
11 * RTE Ethernet Device API
13 * The Ethernet Device API is composed of two parts:
15 * - The application-oriented Ethernet API that includes functions to setup
16 * an Ethernet device (configure it, setup its RX and TX queues and start it),
17 * to get its MAC address, the speed and the status of its physical link,
18 * to receive and to transmit packets, and so on.
20 * - The driver-oriented Ethernet API that exports functions allowing
21 * an Ethernet Poll Mode Driver (PMD) to allocate an Ethernet device instance,
22 * create memzone for HW rings and process registered callbacks, and so on.
23 * PMDs should include ethdev_driver.h instead of this header.
25 * By default, all the functions of the Ethernet Device API exported by a PMD
26 * are lock-free functions which assume to not be invoked in parallel on
27 * different logical cores to work on the same target object. For instance,
28 * the receive function of a PMD cannot be invoked in parallel on two logical
29 * cores to poll the same RX queue [of the same port]. Of course, this function
30 * can be invoked in parallel by different logical cores on different RX queues.
31 * It is the responsibility of the upper level application to enforce this rule.
33 * If needed, parallel accesses by multiple logical cores to shared queues
34 * shall be explicitly protected by dedicated inline lock-aware functions
35 * built on top of their corresponding lock-free functions of the PMD API.
37 * In all functions of the Ethernet API, the Ethernet device is
38 * designated by an integer >= 0 named the device port identifier.
40 * At the Ethernet driver level, Ethernet devices are represented by a generic
41 * data structure of type *rte_eth_dev*.
43 * Ethernet devices are dynamically registered during the PCI probing phase
44 * performed at EAL initialization time.
45 * When an Ethernet device is being probed, an *rte_eth_dev* structure and
46 * a new port identifier are allocated for that device. Then, the eth_dev_init()
47 * function supplied by the Ethernet driver matching the probed PCI
48 * device is invoked to properly initialize the device.
50 * The role of the device init function consists of resetting the hardware,
51 * checking access to Non-volatile Memory (NVM), reading the MAC address
54 * If the device init operation is successful, the correspondence between
55 * the port identifier assigned to the new device and its associated
56 * *rte_eth_dev* structure is effectively registered.
57 * Otherwise, both the *rte_eth_dev* structure and the port identifier are
60 * The functions exported by the application Ethernet API to setup a device
61 * designated by its port identifier must be invoked in the following order:
62 * - rte_eth_dev_configure()
63 * - rte_eth_tx_queue_setup()
64 * - rte_eth_rx_queue_setup()
65 * - rte_eth_dev_start()
67 * Then, the network application can invoke, in any order, the functions
68 * exported by the Ethernet API to get the MAC address of a given device, to
69 * get the speed and the status of a device physical link, to receive/transmit
70 * [burst of] packets, and so on.
72 * If the application wants to change the configuration (i.e. call
73 * rte_eth_dev_configure(), rte_eth_tx_queue_setup(), or
74 * rte_eth_rx_queue_setup()), it must call rte_eth_dev_stop() first to stop the
75 * device and then do the reconfiguration before calling rte_eth_dev_start()
76 * again. The transmit and receive functions should not be invoked when the
79 * Please note that some configuration is not stored between calls to
80 * rte_eth_dev_stop()/rte_eth_dev_start(). The following configuration will
84 * - flow control settings
85 * - receive mode configuration (promiscuous mode, all-multicast mode,
86 * hardware checksum mode, RSS/VMDQ settings etc.)
87 * - VLAN filtering configuration
88 * - default MAC address
89 * - MAC addresses supplied to MAC address array
90 * - flow director filtering mode (but not filtering rules)
91 * - NIC queue statistics mappings
93 * Any other configuration will not be stored and will need to be re-entered
94 * before a call to rte_eth_dev_start().
96 * Finally, a network application can close an Ethernet device by invoking the
97 * rte_eth_dev_close() function.
99 * Each function of the application Ethernet API invokes a specific function
100 * of the PMD that controls the target device designated by its port
102 * For this purpose, all device-specific functions of an Ethernet driver are
103 * supplied through a set of pointers contained in a generic structure of type
105 * The address of the *eth_dev_ops* structure is stored in the *rte_eth_dev*
106 * structure by the device init function of the Ethernet driver, which is
107 * invoked during the PCI probing phase, as explained earlier.
109 * In other words, each function of the Ethernet API simply retrieves the
110 * *rte_eth_dev* structure associated with the device port identifier and
111 * performs an indirect invocation of the corresponding driver function
112 * supplied in the *eth_dev_ops* structure of the *rte_eth_dev* structure.
114 * For performance reasons, the address of the burst-oriented RX and TX
115 * functions of the Ethernet driver are not contained in the *eth_dev_ops*
116 * structure. Instead, they are directly stored at the beginning of the
117 * *rte_eth_dev* structure to avoid an extra indirect memory access during
120 * RTE ethernet device drivers do not use interrupts for transmitting or
121 * receiving. Instead, Ethernet drivers export Poll-Mode receive and transmit
122 * functions to applications.
123 * Both receive and transmit functions are packet-burst oriented to minimize
124 * their cost per packet through the following optimizations:
126 * - Sharing among multiple packets the incompressible cost of the
127 * invocation of receive/transmit functions.
129 * - Enabling receive/transmit functions to take advantage of burst-oriented
130 * hardware features (L1 cache, prefetch instructions, NIC head/tail
131 * registers) to minimize the number of CPU cycles per packet, for instance,
132 * by avoiding useless read memory accesses to ring descriptors, or by
133 * systematically using arrays of pointers that exactly fit L1 cache line
134 * boundaries and sizes.
136 * The burst-oriented receive function does not provide any error notification,
137 * to avoid the corresponding overhead. As a hint, the upper-level application
138 * might check the status of the device link once being systematically returned
139 * a 0 value by the receive function of the driver for a given number of tries.
148 /* Use this macro to check if LRO API is supported */
149 #define RTE_ETHDEV_HAS_LRO_SUPPORT
151 /* Alias RTE_LIBRTE_ETHDEV_DEBUG for backward compatibility. */
152 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
153 #define RTE_ETHDEV_DEBUG_RX
154 #define RTE_ETHDEV_DEBUG_TX
157 #include <rte_compat.h>
159 #include <rte_interrupts.h>
161 #include <rte_devargs.h>
162 #include <rte_errno.h>
163 #include <rte_common.h>
164 #include <rte_config.h>
165 #include <rte_ether.h>
166 #include <rte_power_intrinsics.h>
168 #include "rte_ethdev_trace_fp.h"
169 #include "rte_dev_info.h"
171 extern int rte_eth_dev_logtype;
173 #define RTE_ETHDEV_LOG(level, ...) \
174 rte_log(RTE_LOG_ ## level, rte_eth_dev_logtype, "" __VA_ARGS__)
179 * Initializes a device iterator.
181 * This iterator allows accessing a list of devices matching some devargs.
184 * Device iterator handle initialized by the function.
185 * The fields bus_str and cls_str might be dynamically allocated,
186 * and could be freed by calling rte_eth_iterator_cleanup().
189 * Device description string.
192 * 0 on successful initialization, negative otherwise.
194 int rte_eth_iterator_init(struct rte_dev_iterator *iter, const char *devargs);
197 * Iterates on devices with devargs filter.
198 * The ownership is not checked.
200 * The next port id is returned, and the iterator is updated.
203 * Device iterator handle initialized by rte_eth_iterator_init().
204 * Some fields bus_str and cls_str might be freed when no more port is found,
205 * by calling rte_eth_iterator_cleanup().
208 * A port id if found, RTE_MAX_ETHPORTS otherwise.
210 uint16_t rte_eth_iterator_next(struct rte_dev_iterator *iter);
213 * Free some allocated fields of the iterator.
215 * This function is automatically called by rte_eth_iterator_next()
216 * on the last iteration (i.e. when no more matching port is found).
218 * It is safe to call this function twice; it will do nothing more.
221 * Device iterator handle initialized by rte_eth_iterator_init().
222 * The fields bus_str and cls_str are freed if needed.
224 void rte_eth_iterator_cleanup(struct rte_dev_iterator *iter);
227 * Macro to iterate over all ethdev ports matching some devargs.
229 * If a break is done before the end of the loop,
230 * the function rte_eth_iterator_cleanup() must be called.
233 * Iterated port id of type uint16_t.
235 * Device parameters input as string of type char*.
237 * Iterator handle of type struct rte_dev_iterator, used internally.
239 #define RTE_ETH_FOREACH_MATCHING_DEV(id, devargs, iter) \
240 for (rte_eth_iterator_init(iter, devargs), \
241 id = rte_eth_iterator_next(iter); \
242 id != RTE_MAX_ETHPORTS; \
243 id = rte_eth_iterator_next(iter))
246 * A structure used to retrieve statistics for an Ethernet port.
247 * Not all statistics fields in struct rte_eth_stats are supported
248 * by any type of network interface card (NIC). If any statistics
249 * field is not supported, its value is 0.
250 * All byte-related statistics do not include Ethernet FCS regardless
251 * of whether these bytes have been delivered to the application
252 * (see DEV_RX_OFFLOAD_KEEP_CRC).
254 struct rte_eth_stats {
255 uint64_t ipackets; /**< Total number of successfully received packets. */
256 uint64_t opackets; /**< Total number of successfully transmitted packets.*/
257 uint64_t ibytes; /**< Total number of successfully received bytes. */
258 uint64_t obytes; /**< Total number of successfully transmitted bytes. */
260 /**< Total of RX packets dropped by the HW,
261 * because there are no available buffer (i.e. RX queues are full).
263 uint64_t ierrors; /**< Total number of erroneous received packets. */
264 uint64_t oerrors; /**< Total number of failed transmitted packets. */
265 uint64_t rx_nombuf; /**< Total number of RX mbuf allocation failures. */
266 /* Queue stats are limited to max 256 queues */
267 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
268 /**< Total number of queue RX packets. */
269 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
270 /**< Total number of queue TX packets. */
271 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
272 /**< Total number of successfully received queue bytes. */
273 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
274 /**< Total number of successfully transmitted queue bytes. */
275 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS];
276 /**< Total number of queue packets received that are dropped. */
279 /**@{@name Link speed capabilities
280 * Device supported speeds bitmap flags
282 #define ETH_LINK_SPEED_AUTONEG (0 << 0) /**< Autonegotiate (all speeds) */
283 #define ETH_LINK_SPEED_FIXED (1 << 0) /**< Disable autoneg (fixed speed) */
284 #define ETH_LINK_SPEED_10M_HD (1 << 1) /**< 10 Mbps half-duplex */
285 #define ETH_LINK_SPEED_10M (1 << 2) /**< 10 Mbps full-duplex */
286 #define ETH_LINK_SPEED_100M_HD (1 << 3) /**< 100 Mbps half-duplex */
287 #define ETH_LINK_SPEED_100M (1 << 4) /**< 100 Mbps full-duplex */
288 #define ETH_LINK_SPEED_1G (1 << 5) /**< 1 Gbps */
289 #define ETH_LINK_SPEED_2_5G (1 << 6) /**< 2.5 Gbps */
290 #define ETH_LINK_SPEED_5G (1 << 7) /**< 5 Gbps */
291 #define ETH_LINK_SPEED_10G (1 << 8) /**< 10 Gbps */
292 #define ETH_LINK_SPEED_20G (1 << 9) /**< 20 Gbps */
293 #define ETH_LINK_SPEED_25G (1 << 10) /**< 25 Gbps */
294 #define ETH_LINK_SPEED_40G (1 << 11) /**< 40 Gbps */
295 #define ETH_LINK_SPEED_50G (1 << 12) /**< 50 Gbps */
296 #define ETH_LINK_SPEED_56G (1 << 13) /**< 56 Gbps */
297 #define ETH_LINK_SPEED_100G (1 << 14) /**< 100 Gbps */
298 #define ETH_LINK_SPEED_200G (1 << 15) /**< 200 Gbps */
301 /**@{@name Link speed
302 * Ethernet numeric link speeds in Mbps
304 #define ETH_SPEED_NUM_NONE 0 /**< Not defined */
305 #define ETH_SPEED_NUM_10M 10 /**< 10 Mbps */
306 #define ETH_SPEED_NUM_100M 100 /**< 100 Mbps */
307 #define ETH_SPEED_NUM_1G 1000 /**< 1 Gbps */
308 #define ETH_SPEED_NUM_2_5G 2500 /**< 2.5 Gbps */
309 #define ETH_SPEED_NUM_5G 5000 /**< 5 Gbps */
310 #define ETH_SPEED_NUM_10G 10000 /**< 10 Gbps */
311 #define ETH_SPEED_NUM_20G 20000 /**< 20 Gbps */
312 #define ETH_SPEED_NUM_25G 25000 /**< 25 Gbps */
313 #define ETH_SPEED_NUM_40G 40000 /**< 40 Gbps */
314 #define ETH_SPEED_NUM_50G 50000 /**< 50 Gbps */
315 #define ETH_SPEED_NUM_56G 56000 /**< 56 Gbps */
316 #define ETH_SPEED_NUM_100G 100000 /**< 100 Gbps */
317 #define ETH_SPEED_NUM_200G 200000 /**< 200 Gbps */
318 #define ETH_SPEED_NUM_UNKNOWN UINT32_MAX /**< Unknown */
322 * A structure used to retrieve link-level information of an Ethernet port.
325 struct rte_eth_link {
326 uint32_t link_speed; /**< ETH_SPEED_NUM_ */
327 uint16_t link_duplex : 1; /**< ETH_LINK_[HALF/FULL]_DUPLEX */
328 uint16_t link_autoneg : 1; /**< ETH_LINK_[AUTONEG/FIXED] */
329 uint16_t link_status : 1; /**< ETH_LINK_[DOWN/UP] */
330 } __rte_aligned(8); /**< aligned for atomic64 read/write */
332 /**@{@name Link negotiation
333 * Constants used in link management.
335 #define ETH_LINK_HALF_DUPLEX 0 /**< Half-duplex connection (see link_duplex). */
336 #define ETH_LINK_FULL_DUPLEX 1 /**< Full-duplex connection (see link_duplex). */
337 #define ETH_LINK_DOWN 0 /**< Link is down (see link_status). */
338 #define ETH_LINK_UP 1 /**< Link is up (see link_status). */
339 #define ETH_LINK_FIXED 0 /**< No autonegotiation (see link_autoneg). */
340 #define ETH_LINK_AUTONEG 1 /**< Autonegotiated (see link_autoneg). */
341 #define RTE_ETH_LINK_MAX_STR_LEN 40 /**< Max length of default link string. */
345 * A structure used to configure the ring threshold registers of an RX/TX
346 * queue for an Ethernet port.
348 struct rte_eth_thresh {
349 uint8_t pthresh; /**< Ring prefetch threshold. */
350 uint8_t hthresh; /**< Ring host threshold. */
351 uint8_t wthresh; /**< Ring writeback threshold. */
354 /**@{@name Multi-queue mode
355 * @see rte_eth_conf.rxmode.mq_mode.
357 #define ETH_MQ_RX_RSS_FLAG 0x1 /**< Enable RSS. @see rte_eth_rss_conf */
358 #define ETH_MQ_RX_DCB_FLAG 0x2 /**< Enable DCB. */
359 #define ETH_MQ_RX_VMDQ_FLAG 0x4 /**< Enable VMDq. */
363 * A set of values to identify what method is to be used to route
364 * packets to multiple queues.
366 enum rte_eth_rx_mq_mode {
367 /** None of DCB,RSS or VMDQ mode */
370 /** For RX side, only RSS is on */
371 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG,
372 /** For RX side,only DCB is on. */
373 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG,
374 /** Both DCB and RSS enable */
375 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG,
377 /** Only VMDQ, no RSS nor DCB */
378 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG,
379 /** RSS mode with VMDQ */
380 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG,
381 /** Use VMDQ+DCB to route traffic to queues */
382 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG,
383 /** Enable both VMDQ and DCB in VMDq */
384 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG |
389 * for rx mq mode backward compatible
391 #define ETH_RSS ETH_MQ_RX_RSS
392 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB
393 #define ETH_DCB_RX ETH_MQ_RX_DCB
396 * A set of values to identify what method is to be used to transmit
397 * packets using multi-TCs.
399 enum rte_eth_tx_mq_mode {
400 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */
401 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */
402 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */
403 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */
407 * for tx mq mode backward compatible
409 #define ETH_DCB_NONE ETH_MQ_TX_NONE
410 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB
411 #define ETH_DCB_TX ETH_MQ_TX_DCB
414 * A structure used to configure the RX features of an Ethernet port.
416 struct rte_eth_rxmode {
417 /** The multi-queue packet distribution mode to be used, e.g. RSS. */
418 enum rte_eth_rx_mq_mode mq_mode;
419 uint32_t max_rx_pkt_len; /**< Only used if JUMBO_FRAME enabled. */
420 /** Maximum allowed size of LRO aggregated packet. */
421 uint32_t max_lro_pkt_size;
422 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/
424 * Per-port Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
425 * Only offloads set on rx_offload_capa field on rte_eth_dev_info
426 * structure are allowed to be set.
430 uint64_t reserved_64s[2]; /**< Reserved for future fields */
431 void *reserved_ptrs[2]; /**< Reserved for future fields */
435 * VLAN types to indicate if it is for single VLAN, inner VLAN or outer VLAN.
436 * Note that single VLAN is treated the same as inner VLAN.
439 ETH_VLAN_TYPE_UNKNOWN = 0,
440 ETH_VLAN_TYPE_INNER, /**< Inner VLAN. */
441 ETH_VLAN_TYPE_OUTER, /**< Single VLAN, or outer VLAN. */
446 * A structure used to describe a vlan filter.
447 * If the bit corresponding to a VID is set, such VID is on.
449 struct rte_vlan_filter_conf {
454 * A structure used to configure the Receive Side Scaling (RSS) feature
455 * of an Ethernet port.
456 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points
457 * to an array holding the RSS key to use for hashing specific header
458 * fields of received packets. The length of this array should be indicated
459 * by *rss_key_len* below. Otherwise, a default random hash key is used by
462 * The *rss_key_len* field of the *rss_conf* structure indicates the length
463 * in bytes of the array pointed by *rss_key*. To be compatible, this length
464 * will be checked in i40e only. Others assume 40 bytes to be used as before.
466 * The *rss_hf* field of the *rss_conf* structure indicates the different
467 * types of IPv4/IPv6 packets to which the RSS hashing must be applied.
468 * Supplying an *rss_hf* equal to zero disables the RSS feature.
470 struct rte_eth_rss_conf {
471 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */
472 uint8_t rss_key_len; /**< hash key length in bytes. */
473 uint64_t rss_hf; /**< Hash functions to apply - see below. */
477 * A packet can be identified by hardware as different flow types. Different
478 * NIC hardware may support different flow types.
479 * Basically, the NIC hardware identifies the flow type as deep protocol as
480 * possible, and exclusively. For example, if a packet is identified as
481 * 'RTE_ETH_FLOW_NONFRAG_IPV4_TCP', it will not be any of other flow types,
482 * though it is an actual IPV4 packet.
484 #define RTE_ETH_FLOW_UNKNOWN 0
485 #define RTE_ETH_FLOW_RAW 1
486 #define RTE_ETH_FLOW_IPV4 2
487 #define RTE_ETH_FLOW_FRAG_IPV4 3
488 #define RTE_ETH_FLOW_NONFRAG_IPV4_TCP 4
489 #define RTE_ETH_FLOW_NONFRAG_IPV4_UDP 5
490 #define RTE_ETH_FLOW_NONFRAG_IPV4_SCTP 6
491 #define RTE_ETH_FLOW_NONFRAG_IPV4_OTHER 7
492 #define RTE_ETH_FLOW_IPV6 8
493 #define RTE_ETH_FLOW_FRAG_IPV6 9
494 #define RTE_ETH_FLOW_NONFRAG_IPV6_TCP 10
495 #define RTE_ETH_FLOW_NONFRAG_IPV6_UDP 11
496 #define RTE_ETH_FLOW_NONFRAG_IPV6_SCTP 12
497 #define RTE_ETH_FLOW_NONFRAG_IPV6_OTHER 13
498 #define RTE_ETH_FLOW_L2_PAYLOAD 14
499 #define RTE_ETH_FLOW_IPV6_EX 15
500 #define RTE_ETH_FLOW_IPV6_TCP_EX 16
501 #define RTE_ETH_FLOW_IPV6_UDP_EX 17
502 #define RTE_ETH_FLOW_PORT 18
503 /**< Consider device port number as a flow differentiator */
504 #define RTE_ETH_FLOW_VXLAN 19 /**< VXLAN protocol based flow */
505 #define RTE_ETH_FLOW_GENEVE 20 /**< GENEVE protocol based flow */
506 #define RTE_ETH_FLOW_NVGRE 21 /**< NVGRE protocol based flow */
507 #define RTE_ETH_FLOW_VXLAN_GPE 22 /**< VXLAN-GPE protocol based flow */
508 #define RTE_ETH_FLOW_GTPU 23 /**< GTPU protocol based flow */
509 #define RTE_ETH_FLOW_MAX 24
512 * Below macros are defined for RSS offload types, they can be used to
513 * fill rte_eth_rss_conf.rss_hf or rte_flow_action_rss.types.
515 #define ETH_RSS_IPV4 (1ULL << 2)
516 #define ETH_RSS_FRAG_IPV4 (1ULL << 3)
517 #define ETH_RSS_NONFRAG_IPV4_TCP (1ULL << 4)
518 #define ETH_RSS_NONFRAG_IPV4_UDP (1ULL << 5)
519 #define ETH_RSS_NONFRAG_IPV4_SCTP (1ULL << 6)
520 #define ETH_RSS_NONFRAG_IPV4_OTHER (1ULL << 7)
521 #define ETH_RSS_IPV6 (1ULL << 8)
522 #define ETH_RSS_FRAG_IPV6 (1ULL << 9)
523 #define ETH_RSS_NONFRAG_IPV6_TCP (1ULL << 10)
524 #define ETH_RSS_NONFRAG_IPV6_UDP (1ULL << 11)
525 #define ETH_RSS_NONFRAG_IPV6_SCTP (1ULL << 12)
526 #define ETH_RSS_NONFRAG_IPV6_OTHER (1ULL << 13)
527 #define ETH_RSS_L2_PAYLOAD (1ULL << 14)
528 #define ETH_RSS_IPV6_EX (1ULL << 15)
529 #define ETH_RSS_IPV6_TCP_EX (1ULL << 16)
530 #define ETH_RSS_IPV6_UDP_EX (1ULL << 17)
531 #define ETH_RSS_PORT (1ULL << 18)
532 #define ETH_RSS_VXLAN (1ULL << 19)
533 #define ETH_RSS_GENEVE (1ULL << 20)
534 #define ETH_RSS_NVGRE (1ULL << 21)
535 #define ETH_RSS_GTPU (1ULL << 23)
536 #define ETH_RSS_ETH (1ULL << 24)
537 #define ETH_RSS_S_VLAN (1ULL << 25)
538 #define ETH_RSS_C_VLAN (1ULL << 26)
539 #define ETH_RSS_ESP (1ULL << 27)
540 #define ETH_RSS_AH (1ULL << 28)
541 #define ETH_RSS_L2TPV3 (1ULL << 29)
542 #define ETH_RSS_PFCP (1ULL << 30)
543 #define ETH_RSS_PPPOE (1ULL << 31)
544 #define ETH_RSS_ECPRI (1ULL << 32)
545 #define ETH_RSS_MPLS (1ULL << 33)
546 #define ETH_RSS_IPV4_CHKSUM (1ULL << 34)
549 * The ETH_RSS_L4_CHKSUM works on checksum field of any L4 header.
550 * It is similar to ETH_RSS_PORT that they don't specify the specific type of
551 * L4 header. This macro is defined to replace some specific L4 (TCP/UDP/SCTP)
552 * checksum type for constructing the use of RSS offload bits.
554 * Due to above reason, some old APIs (and configuration) don't support
555 * ETH_RSS_L4_CHKSUM. The rte_flow RSS API supports it.
557 * For the case that checksum is not used in an UDP header,
558 * it takes the reserved value 0 as input for the hash function.
560 #define ETH_RSS_L4_CHKSUM (1ULL << 35)
563 * We use the following macros to combine with above ETH_RSS_* for
564 * more specific input set selection. These bits are defined starting
565 * from the high end of the 64 bits.
566 * Note: If we use above ETH_RSS_* without SRC/DST_ONLY, it represents
567 * both SRC and DST are taken into account. If SRC_ONLY and DST_ONLY of
568 * the same level are used simultaneously, it is the same case as none of
571 #define ETH_RSS_L3_SRC_ONLY (1ULL << 63)
572 #define ETH_RSS_L3_DST_ONLY (1ULL << 62)
573 #define ETH_RSS_L4_SRC_ONLY (1ULL << 61)
574 #define ETH_RSS_L4_DST_ONLY (1ULL << 60)
575 #define ETH_RSS_L2_SRC_ONLY (1ULL << 59)
576 #define ETH_RSS_L2_DST_ONLY (1ULL << 58)
579 * Only select IPV6 address prefix as RSS input set according to
580 * https://tools.ietf.org/html/rfc6052
581 * Must be combined with ETH_RSS_IPV6, ETH_RSS_NONFRAG_IPV6_UDP,
582 * ETH_RSS_NONFRAG_IPV6_TCP, ETH_RSS_NONFRAG_IPV6_SCTP.
584 #define RTE_ETH_RSS_L3_PRE32 (1ULL << 57)
585 #define RTE_ETH_RSS_L3_PRE40 (1ULL << 56)
586 #define RTE_ETH_RSS_L3_PRE48 (1ULL << 55)
587 #define RTE_ETH_RSS_L3_PRE56 (1ULL << 54)
588 #define RTE_ETH_RSS_L3_PRE64 (1ULL << 53)
589 #define RTE_ETH_RSS_L3_PRE96 (1ULL << 52)
592 * Use the following macros to combine with the above layers
593 * to choose inner and outer layers or both for RSS computation.
594 * Bits 50 and 51 are reserved for this.
598 * level 0, requests the default behavior.
599 * Depending on the packet type, it can mean outermost, innermost,
600 * anything in between or even no RSS.
601 * It basically stands for the innermost encapsulation level RSS
602 * can be performed on according to PMD and device capabilities.
604 #define ETH_RSS_LEVEL_PMD_DEFAULT (0ULL << 50)
607 * level 1, requests RSS to be performed on the outermost packet
608 * encapsulation level.
610 #define ETH_RSS_LEVEL_OUTERMOST (1ULL << 50)
613 * level 2, requests RSS to be performed on the specified inner packet
614 * encapsulation level, from outermost to innermost (lower to higher values).
616 #define ETH_RSS_LEVEL_INNERMOST (2ULL << 50)
617 #define ETH_RSS_LEVEL_MASK (3ULL << 50)
619 #define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
622 * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
623 * the same level are used simultaneously, it is the same case as
624 * none of them are added.
627 * RSS types with SRC/DST_ONLY.
631 static inline uint64_t
632 rte_eth_rss_hf_refine(uint64_t rss_hf)
634 if ((rss_hf & ETH_RSS_L3_SRC_ONLY) && (rss_hf & ETH_RSS_L3_DST_ONLY))
635 rss_hf &= ~(ETH_RSS_L3_SRC_ONLY | ETH_RSS_L3_DST_ONLY);
637 if ((rss_hf & ETH_RSS_L4_SRC_ONLY) && (rss_hf & ETH_RSS_L4_DST_ONLY))
638 rss_hf &= ~(ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY);
643 #define ETH_RSS_IPV6_PRE32 ( \
645 RTE_ETH_RSS_L3_PRE32)
647 #define ETH_RSS_IPV6_PRE40 ( \
649 RTE_ETH_RSS_L3_PRE40)
651 #define ETH_RSS_IPV6_PRE48 ( \
653 RTE_ETH_RSS_L3_PRE48)
655 #define ETH_RSS_IPV6_PRE56 ( \
657 RTE_ETH_RSS_L3_PRE56)
659 #define ETH_RSS_IPV6_PRE64 ( \
661 RTE_ETH_RSS_L3_PRE64)
663 #define ETH_RSS_IPV6_PRE96 ( \
665 RTE_ETH_RSS_L3_PRE96)
667 #define ETH_RSS_IPV6_PRE32_UDP ( \
668 ETH_RSS_NONFRAG_IPV6_UDP | \
669 RTE_ETH_RSS_L3_PRE32)
671 #define ETH_RSS_IPV6_PRE40_UDP ( \
672 ETH_RSS_NONFRAG_IPV6_UDP | \
673 RTE_ETH_RSS_L3_PRE40)
675 #define ETH_RSS_IPV6_PRE48_UDP ( \
676 ETH_RSS_NONFRAG_IPV6_UDP | \
677 RTE_ETH_RSS_L3_PRE48)
679 #define ETH_RSS_IPV6_PRE56_UDP ( \
680 ETH_RSS_NONFRAG_IPV6_UDP | \
681 RTE_ETH_RSS_L3_PRE56)
683 #define ETH_RSS_IPV6_PRE64_UDP ( \
684 ETH_RSS_NONFRAG_IPV6_UDP | \
685 RTE_ETH_RSS_L3_PRE64)
687 #define ETH_RSS_IPV6_PRE96_UDP ( \
688 ETH_RSS_NONFRAG_IPV6_UDP | \
689 RTE_ETH_RSS_L3_PRE96)
691 #define ETH_RSS_IPV6_PRE32_TCP ( \
692 ETH_RSS_NONFRAG_IPV6_TCP | \
693 RTE_ETH_RSS_L3_PRE32)
695 #define ETH_RSS_IPV6_PRE40_TCP ( \
696 ETH_RSS_NONFRAG_IPV6_TCP | \
697 RTE_ETH_RSS_L3_PRE40)
699 #define ETH_RSS_IPV6_PRE48_TCP ( \
700 ETH_RSS_NONFRAG_IPV6_TCP | \
701 RTE_ETH_RSS_L3_PRE48)
703 #define ETH_RSS_IPV6_PRE56_TCP ( \
704 ETH_RSS_NONFRAG_IPV6_TCP | \
705 RTE_ETH_RSS_L3_PRE56)
707 #define ETH_RSS_IPV6_PRE64_TCP ( \
708 ETH_RSS_NONFRAG_IPV6_TCP | \
709 RTE_ETH_RSS_L3_PRE64)
711 #define ETH_RSS_IPV6_PRE96_TCP ( \
712 ETH_RSS_NONFRAG_IPV6_TCP | \
713 RTE_ETH_RSS_L3_PRE96)
715 #define ETH_RSS_IPV6_PRE32_SCTP ( \
716 ETH_RSS_NONFRAG_IPV6_SCTP | \
717 RTE_ETH_RSS_L3_PRE32)
719 #define ETH_RSS_IPV6_PRE40_SCTP ( \
720 ETH_RSS_NONFRAG_IPV6_SCTP | \
721 RTE_ETH_RSS_L3_PRE40)
723 #define ETH_RSS_IPV6_PRE48_SCTP ( \
724 ETH_RSS_NONFRAG_IPV6_SCTP | \
725 RTE_ETH_RSS_L3_PRE48)
727 #define ETH_RSS_IPV6_PRE56_SCTP ( \
728 ETH_RSS_NONFRAG_IPV6_SCTP | \
729 RTE_ETH_RSS_L3_PRE56)
731 #define ETH_RSS_IPV6_PRE64_SCTP ( \
732 ETH_RSS_NONFRAG_IPV6_SCTP | \
733 RTE_ETH_RSS_L3_PRE64)
735 #define ETH_RSS_IPV6_PRE96_SCTP ( \
736 ETH_RSS_NONFRAG_IPV6_SCTP | \
737 RTE_ETH_RSS_L3_PRE96)
739 #define ETH_RSS_IP ( \
741 ETH_RSS_FRAG_IPV4 | \
742 ETH_RSS_NONFRAG_IPV4_OTHER | \
744 ETH_RSS_FRAG_IPV6 | \
745 ETH_RSS_NONFRAG_IPV6_OTHER | \
748 #define ETH_RSS_UDP ( \
749 ETH_RSS_NONFRAG_IPV4_UDP | \
750 ETH_RSS_NONFRAG_IPV6_UDP | \
753 #define ETH_RSS_TCP ( \
754 ETH_RSS_NONFRAG_IPV4_TCP | \
755 ETH_RSS_NONFRAG_IPV6_TCP | \
758 #define ETH_RSS_SCTP ( \
759 ETH_RSS_NONFRAG_IPV4_SCTP | \
760 ETH_RSS_NONFRAG_IPV6_SCTP)
762 #define ETH_RSS_TUNNEL ( \
767 #define ETH_RSS_VLAN ( \
771 /**< Mask of valid RSS hash protocols */
772 #define ETH_RSS_PROTO_MASK ( \
774 ETH_RSS_FRAG_IPV4 | \
775 ETH_RSS_NONFRAG_IPV4_TCP | \
776 ETH_RSS_NONFRAG_IPV4_UDP | \
777 ETH_RSS_NONFRAG_IPV4_SCTP | \
778 ETH_RSS_NONFRAG_IPV4_OTHER | \
780 ETH_RSS_FRAG_IPV6 | \
781 ETH_RSS_NONFRAG_IPV6_TCP | \
782 ETH_RSS_NONFRAG_IPV6_UDP | \
783 ETH_RSS_NONFRAG_IPV6_SCTP | \
784 ETH_RSS_NONFRAG_IPV6_OTHER | \
785 ETH_RSS_L2_PAYLOAD | \
787 ETH_RSS_IPV6_TCP_EX | \
788 ETH_RSS_IPV6_UDP_EX | \
796 * Definitions used for redirection table entry size.
797 * Some RSS RETA sizes may not be supported by some drivers, check the
798 * documentation or the description of relevant functions for more details.
800 #define ETH_RSS_RETA_SIZE_64 64
801 #define ETH_RSS_RETA_SIZE_128 128
802 #define ETH_RSS_RETA_SIZE_256 256
803 #define ETH_RSS_RETA_SIZE_512 512
804 #define RTE_RETA_GROUP_SIZE 64
806 /**@{@name VMDq and DCB maximums */
807 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */
808 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */
809 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */
810 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */
813 /**@{@name DCB capabilities */
814 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */
815 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */
818 /**@{@name VLAN offload bits */
819 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */
820 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */
821 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */
822 #define ETH_QINQ_STRIP_OFFLOAD 0x0008 /**< QINQ Strip On/Off */
824 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */
825 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/
826 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/
827 #define ETH_QINQ_STRIP_MASK 0x0008 /**< QINQ Strip setting mask */
828 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/
831 /* Definitions used for receive MAC address */
832 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */
834 /* Definitions used for unicast hash */
835 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */
837 /**@{@name VMDq Rx mode
838 * @see rte_eth_vmdq_rx_conf.rx_mode
840 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */
841 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */
842 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */
843 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */
844 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */
847 /** Maximum nb. of vlan per mirror rule */
848 #define ETH_MIRROR_MAX_VLANS 64
850 /**@{@name Mirroring type
851 * @see rte_eth_mirror_conf.rule_type
853 #define ETH_MIRROR_VIRTUAL_POOL_UP 0x01 /**< Virtual Pool uplink Mirroring. */
854 #define ETH_MIRROR_UPLINK_PORT 0x02 /**< Uplink Port Mirroring. */
855 #define ETH_MIRROR_DOWNLINK_PORT 0x04 /**< Downlink Port Mirroring. */
856 #define ETH_MIRROR_VLAN 0x08 /**< VLAN Mirroring. */
857 #define ETH_MIRROR_VIRTUAL_POOL_DOWN 0x10 /**< Virtual Pool downlink Mirroring. */
861 * A structure used to configure VLAN traffic mirror of an Ethernet port.
863 struct rte_eth_vlan_mirror {
864 uint64_t vlan_mask; /**< mask for valid VLAN ID. */
865 /** VLAN ID list for vlan mirroring. */
866 uint16_t vlan_id[ETH_MIRROR_MAX_VLANS];
870 * A structure used to configure traffic mirror of an Ethernet port.
872 struct rte_eth_mirror_conf {
873 uint8_t rule_type; /**< Mirroring rule type */
874 uint8_t dst_pool; /**< Destination pool for this mirror rule. */
875 uint64_t pool_mask; /**< Bitmap of pool for pool mirroring */
876 /** VLAN ID setting for VLAN mirroring. */
877 struct rte_eth_vlan_mirror vlan;
881 * A structure used to configure 64 entries of Redirection Table of the
882 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure
883 * more than 64 entries supported by hardware, an array of this structure
886 struct rte_eth_rss_reta_entry64 {
888 /**< Mask bits indicate which entries need to be updated/queried. */
889 uint16_t reta[RTE_RETA_GROUP_SIZE];
890 /**< Group of 64 redirection table entries. */
894 * This enum indicates the possible number of traffic classes
895 * in DCB configurations
897 enum rte_eth_nb_tcs {
898 ETH_4_TCS = 4, /**< 4 TCs with DCB. */
899 ETH_8_TCS = 8 /**< 8 TCs with DCB. */
903 * This enum indicates the possible number of queue pools
904 * in VMDQ configurations.
906 enum rte_eth_nb_pools {
907 ETH_8_POOLS = 8, /**< 8 VMDq pools. */
908 ETH_16_POOLS = 16, /**< 16 VMDq pools. */
909 ETH_32_POOLS = 32, /**< 32 VMDq pools. */
910 ETH_64_POOLS = 64 /**< 64 VMDq pools. */
913 /* This structure may be extended in future. */
914 struct rte_eth_dcb_rx_conf {
915 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */
916 /** Traffic class each UP mapped to. */
917 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
920 struct rte_eth_vmdq_dcb_tx_conf {
921 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */
922 /** Traffic class each UP mapped to. */
923 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
926 struct rte_eth_dcb_tx_conf {
927 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */
928 /** Traffic class each UP mapped to. */
929 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
932 struct rte_eth_vmdq_tx_conf {
933 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */
937 * A structure used to configure the VMDQ+DCB feature
938 * of an Ethernet port.
940 * Using this feature, packets are routed to a pool of queues, based
941 * on the vlan id in the vlan tag, and then to a specific queue within
942 * that pool, using the user priority vlan tag field.
944 * A default pool may be used, if desired, to route all traffic which
945 * does not match the vlan filter rules.
947 struct rte_eth_vmdq_dcb_conf {
948 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */
949 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
950 uint8_t default_pool; /**< The default pool, if applicable */
951 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
953 uint16_t vlan_id; /**< The vlan id of the received frame */
954 uint64_t pools; /**< Bitmask of pools for packet rx */
955 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
956 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
957 /**< Selects a queue in a pool */
961 * A structure used to configure the VMDQ feature of an Ethernet port when
962 * not combined with the DCB feature.
964 * Using this feature, packets are routed to a pool of queues. By default,
965 * the pool selection is based on the MAC address, the vlan id in the
966 * vlan tag as specified in the pool_map array.
967 * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool
968 * selection using only the MAC address. MAC address to pool mapping is done
969 * using the rte_eth_dev_mac_addr_add function, with the pool parameter
970 * corresponding to the pool id.
972 * Queue selection within the selected pool will be done using RSS when
973 * it is enabled or revert to the first queue of the pool if not.
975 * A default pool may be used, if desired, to route all traffic which
976 * does not match the vlan filter rules or any pool MAC address.
978 struct rte_eth_vmdq_rx_conf {
979 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */
980 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
981 uint8_t default_pool; /**< The default pool, if applicable */
982 uint8_t enable_loop_back; /**< Enable VT loop back */
983 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
984 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */
986 uint16_t vlan_id; /**< The vlan id of the received frame */
987 uint64_t pools; /**< Bitmask of pools for packet rx */
988 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
992 * A structure used to configure the TX features of an Ethernet port.
994 struct rte_eth_txmode {
995 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */
997 * Per-port Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
998 * Only offloads set on tx_offload_capa field on rte_eth_dev_info
999 * structure are allowed to be set.
1005 uint8_t hw_vlan_reject_tagged : 1,
1006 /**< If set, reject sending out tagged pkts */
1007 hw_vlan_reject_untagged : 1,
1008 /**< If set, reject sending out untagged pkts */
1009 hw_vlan_insert_pvid : 1;
1010 /**< If set, enable port based VLAN insertion */
1012 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1013 void *reserved_ptrs[2]; /**< Reserved for future fields */
1018 * @b EXPERIMENTAL: this structure may change without prior notice.
1020 * A structure used to configure an Rx packet segment to split.
1022 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
1023 * the PMD will split the received packets into multiple segments
1024 * according to the specification in the description array:
1026 * - The first network buffer will be allocated from the memory pool,
1027 * specified in the first array element, the second buffer, from the
1028 * pool in the second element, and so on.
1030 * - The offsets from the segment description elements specify
1031 * the data offset from the buffer beginning except the first mbuf.
1032 * The first segment offset is added with RTE_PKTMBUF_HEADROOM.
1034 * - The lengths in the elements define the maximal data amount
1035 * being received to each segment. The receiving starts with filling
1036 * up the first mbuf data buffer up to specified length. If the
1037 * there are data remaining (packet is longer than buffer in the first
1038 * mbuf) the following data will be pushed to the next segment
1039 * up to its own length, and so on.
1041 * - If the length in the segment description element is zero
1042 * the actual buffer size will be deduced from the appropriate
1043 * memory pool properties.
1045 * - If there is not enough elements to describe the buffer for entire
1046 * packet of maximal length the following parameters will be used
1047 * for the all remaining segments:
1048 * - pool from the last valid element
1049 * - the buffer size from this pool
1052 struct rte_eth_rxseg_split {
1053 struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
1054 uint16_t length; /**< Segment data length, configures split point. */
1055 uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
1056 uint32_t reserved; /**< Reserved field. */
1061 * @b EXPERIMENTAL: this structure may change without prior notice.
1063 * A common structure used to describe Rx packet segment properties.
1065 union rte_eth_rxseg {
1066 /* The settings for buffer split offload. */
1067 struct rte_eth_rxseg_split split;
1068 /* The other features settings should be added here. */
1072 * A structure used to configure an RX ring of an Ethernet port.
1074 struct rte_eth_rxconf {
1075 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */
1076 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
1077 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
1078 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
1079 uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
1081 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
1082 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
1083 * fields on rte_eth_dev_info structure are allowed to be set.
1087 * Points to the array of segment descriptions for an entire packet.
1088 * Array elements are properties for consecutive Rx segments.
1090 * The supported capabilities of receiving segmentation is reported
1091 * in rte_eth_dev_info.rx_seg_capa field.
1093 union rte_eth_rxseg *rx_seg;
1095 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1096 void *reserved_ptrs[2]; /**< Reserved for future fields */
1100 * A structure used to configure a TX ring of an Ethernet port.
1102 struct rte_eth_txconf {
1103 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */
1104 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */
1105 uint16_t tx_free_thresh; /**< Start freeing TX buffers if there are
1106 less free descriptors than this value. */
1108 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
1110 * Per-queue Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
1111 * Only offloads set on tx_queue_offload_capa or tx_offload_capa
1112 * fields on rte_eth_dev_info structure are allowed to be set.
1116 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1117 void *reserved_ptrs[2]; /**< Reserved for future fields */
1122 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1124 * A structure used to return the hairpin capabilities that are supported.
1126 struct rte_eth_hairpin_cap {
1127 /** The max number of hairpin queues (different bindings). */
1128 uint16_t max_nb_queues;
1129 /** Max number of Rx queues to be connected to one Tx queue. */
1130 uint16_t max_rx_2_tx;
1131 /** Max number of Tx queues to be connected to one Rx queue. */
1132 uint16_t max_tx_2_rx;
1133 uint16_t max_nb_desc; /**< The max num of descriptors. */
1136 #define RTE_ETH_MAX_HAIRPIN_PEERS 32
1140 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1142 * A structure used to hold hairpin peer data.
1144 struct rte_eth_hairpin_peer {
1145 uint16_t port; /**< Peer port. */
1146 uint16_t queue; /**< Peer queue. */
1151 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1153 * A structure used to configure hairpin binding.
1155 struct rte_eth_hairpin_conf {
1156 uint32_t peer_count:16; /**< The number of peers. */
1159 * Explicit Tx flow rule mode.
1160 * One hairpin pair of queues should have the same attribute.
1162 * - When set, the user should be responsible for inserting the hairpin
1163 * Tx part flows and removing them.
1164 * - When clear, the PMD will try to handle the Tx part of the flows,
1165 * e.g., by splitting one flow into two parts.
1167 uint32_t tx_explicit:1;
1170 * Manually bind hairpin queues.
1171 * One hairpin pair of queues should have the same attribute.
1173 * - When set, to enable hairpin, the user should call the hairpin bind
1174 * function after all the queues are set up properly and the ports are
1175 * started. Also, the hairpin unbind function should be called
1176 * accordingly before stopping a port that with hairpin configured.
1177 * - When clear, the PMD will try to enable the hairpin with the queues
1178 * configured automatically during port start.
1180 uint32_t manual_bind:1;
1181 uint32_t reserved:14; /**< Reserved bits. */
1182 struct rte_eth_hairpin_peer peers[RTE_ETH_MAX_HAIRPIN_PEERS];
1186 * A structure contains information about HW descriptor ring limitations.
1188 struct rte_eth_desc_lim {
1189 uint16_t nb_max; /**< Max allowed number of descriptors. */
1190 uint16_t nb_min; /**< Min allowed number of descriptors. */
1191 uint16_t nb_align; /**< Number of descriptors should be aligned to. */
1194 * Max allowed number of segments per whole packet.
1196 * - For TSO packet this is the total number of data descriptors allowed
1199 * @see nb_mtu_seg_max
1201 uint16_t nb_seg_max;
1204 * Max number of segments per one MTU.
1206 * - For non-TSO packet, this is the maximum allowed number of segments
1207 * in a single transmit packet.
1209 * - For TSO packet each segment within the TSO may span up to this
1214 uint16_t nb_mtu_seg_max;
1218 * This enum indicates the flow control mode
1220 enum rte_eth_fc_mode {
1221 RTE_FC_NONE = 0, /**< Disable flow control. */
1222 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */
1223 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */
1224 RTE_FC_FULL /**< Enable flow control on both side. */
1228 * A structure used to configure Ethernet flow control parameter.
1229 * These parameters will be configured into the register of the NIC.
1230 * Please refer to the corresponding data sheet for proper value.
1232 struct rte_eth_fc_conf {
1233 uint32_t high_water; /**< High threshold value to trigger XOFF */
1234 uint32_t low_water; /**< Low threshold value to trigger XON */
1235 uint16_t pause_time; /**< Pause quota in the Pause frame */
1236 uint16_t send_xon; /**< Is XON frame need be sent */
1237 enum rte_eth_fc_mode mode; /**< Link flow control mode */
1238 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */
1239 uint8_t autoneg; /**< Use Pause autoneg */
1243 * A structure used to configure Ethernet priority flow control parameter.
1244 * These parameters will be configured into the register of the NIC.
1245 * Please refer to the corresponding data sheet for proper value.
1247 struct rte_eth_pfc_conf {
1248 struct rte_eth_fc_conf fc; /**< General flow control parameter. */
1249 uint8_t priority; /**< VLAN User Priority. */
1253 * Tunnel type for device-specific classifier configuration.
1254 * @see rte_eth_udp_tunnel
1256 enum rte_eth_tunnel_type {
1257 RTE_TUNNEL_TYPE_NONE = 0,
1258 RTE_TUNNEL_TYPE_VXLAN,
1259 RTE_TUNNEL_TYPE_GENEVE,
1260 RTE_TUNNEL_TYPE_TEREDO,
1261 RTE_TUNNEL_TYPE_NVGRE,
1262 RTE_TUNNEL_TYPE_IP_IN_GRE,
1263 RTE_L2_TUNNEL_TYPE_E_TAG,
1264 RTE_TUNNEL_TYPE_VXLAN_GPE,
1265 RTE_TUNNEL_TYPE_ECPRI,
1266 RTE_TUNNEL_TYPE_MAX,
1269 /* Deprecated API file for rte_eth_dev_filter_* functions */
1270 #include "rte_eth_ctrl.h"
1273 * Memory space that can be configured to store Flow Director filters
1274 * in the board memory.
1276 enum rte_fdir_pballoc_type {
1277 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */
1278 RTE_FDIR_PBALLOC_128K, /**< 128k. */
1279 RTE_FDIR_PBALLOC_256K, /**< 256k. */
1283 * Select report mode of FDIR hash information in RX descriptors.
1285 enum rte_fdir_status_mode {
1286 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */
1287 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */
1288 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */
1292 * A structure used to configure the Flow Director (FDIR) feature
1293 * of an Ethernet port.
1295 * If mode is RTE_FDIR_MODE_NONE, the pballoc value is ignored.
1297 struct rte_fdir_conf {
1298 enum rte_fdir_mode mode; /**< Flow Director mode. */
1299 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */
1300 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */
1301 /** RX queue of packets matching a "drop" filter in perfect mode. */
1303 struct rte_eth_fdir_masks mask;
1304 struct rte_eth_fdir_flex_conf flex_conf;
1305 /**< Flex payload configuration. */
1309 * UDP tunneling configuration.
1311 * Used to configure the classifier of a device,
1312 * associating an UDP port with a type of tunnel.
1314 * Some NICs may need such configuration to properly parse a tunnel
1315 * with any standard or custom UDP port.
1317 struct rte_eth_udp_tunnel {
1318 uint16_t udp_port; /**< UDP port used for the tunnel. */
1319 uint8_t prot_type; /**< Tunnel type. @see rte_eth_tunnel_type */
1323 * A structure used to enable/disable specific device interrupts.
1325 struct rte_intr_conf {
1326 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */
1328 /** enable/disable rxq interrupt. 0 (default) - disable, 1 enable */
1330 /** enable/disable rmv interrupt. 0 (default) - disable, 1 enable */
1335 * A structure used to configure an Ethernet port.
1336 * Depending upon the RX multi-queue mode, extra advanced
1337 * configuration settings may be needed.
1339 struct rte_eth_conf {
1340 uint32_t link_speeds; /**< bitmap of ETH_LINK_SPEED_XXX of speeds to be
1341 used. ETH_LINK_SPEED_FIXED disables link
1342 autonegotiation, and a unique speed shall be
1343 set. Otherwise, the bitmap defines the set of
1344 speeds to be advertised. If the special value
1345 ETH_LINK_SPEED_AUTONEG (0) is used, all speeds
1346 supported are advertised. */
1347 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */
1348 struct rte_eth_txmode txmode; /**< Port TX configuration. */
1349 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value
1350 is 0, meaning the loopback mode is disabled.
1351 Read the datasheet of given ethernet controller
1352 for details. The possible values of this field
1353 are defined in implementation of each driver. */
1355 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */
1356 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf;
1357 /**< Port vmdq+dcb configuration. */
1358 struct rte_eth_dcb_rx_conf dcb_rx_conf;
1359 /**< Port dcb RX configuration. */
1360 struct rte_eth_vmdq_rx_conf vmdq_rx_conf;
1361 /**< Port vmdq RX configuration. */
1362 } rx_adv_conf; /**< Port RX filtering configuration. */
1364 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf;
1365 /**< Port vmdq+dcb TX configuration. */
1366 struct rte_eth_dcb_tx_conf dcb_tx_conf;
1367 /**< Port dcb TX configuration. */
1368 struct rte_eth_vmdq_tx_conf vmdq_tx_conf;
1369 /**< Port vmdq TX configuration. */
1370 } tx_adv_conf; /**< Port TX DCB configuration (union). */
1371 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC
1372 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */
1373 uint32_t dcb_capability_en;
1374 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. DEPRECATED */
1375 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */
1379 * RX offload capabilities of a device.
1381 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001
1382 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002
1383 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004
1384 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008
1385 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010
1386 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020
1387 #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040
1388 #define DEV_RX_OFFLOAD_MACSEC_STRIP 0x00000080
1389 #define DEV_RX_OFFLOAD_HEADER_SPLIT 0x00000100
1390 #define DEV_RX_OFFLOAD_VLAN_FILTER 0x00000200
1391 #define DEV_RX_OFFLOAD_VLAN_EXTEND 0x00000400
1392 #define DEV_RX_OFFLOAD_JUMBO_FRAME 0x00000800
1393 #define DEV_RX_OFFLOAD_SCATTER 0x00002000
1395 * Timestamp is set by the driver in RTE_MBUF_DYNFIELD_TIMESTAMP_NAME
1396 * and RTE_MBUF_DYNFLAG_RX_TIMESTAMP_NAME is set in ol_flags.
1397 * The mbuf field and flag are registered when the offload is configured.
1399 #define DEV_RX_OFFLOAD_TIMESTAMP 0x00004000
1400 #define DEV_RX_OFFLOAD_SECURITY 0x00008000
1401 #define DEV_RX_OFFLOAD_KEEP_CRC 0x00010000
1402 #define DEV_RX_OFFLOAD_SCTP_CKSUM 0x00020000
1403 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM 0x00040000
1404 #define DEV_RX_OFFLOAD_RSS_HASH 0x00080000
1405 #define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
1407 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
1408 DEV_RX_OFFLOAD_UDP_CKSUM | \
1409 DEV_RX_OFFLOAD_TCP_CKSUM)
1410 #define DEV_RX_OFFLOAD_VLAN (DEV_RX_OFFLOAD_VLAN_STRIP | \
1411 DEV_RX_OFFLOAD_VLAN_FILTER | \
1412 DEV_RX_OFFLOAD_VLAN_EXTEND | \
1413 DEV_RX_OFFLOAD_QINQ_STRIP)
1416 * If new Rx offload capabilities are defined, they also must be
1417 * mentioned in rte_rx_offload_names in rte_ethdev.c file.
1421 * TX offload capabilities of a device.
1423 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001
1424 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002
1425 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004
1426 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008
1427 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010
1428 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020
1429 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040
1430 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */
1431 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100
1432 #define DEV_TX_OFFLOAD_VXLAN_TNL_TSO 0x00000200 /**< Used for tunneling packet. */
1433 #define DEV_TX_OFFLOAD_GRE_TNL_TSO 0x00000400 /**< Used for tunneling packet. */
1434 #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */
1435 #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */
1436 #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000
1437 #define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000
1438 /**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same
1439 * tx queue without SW lock.
1441 #define DEV_TX_OFFLOAD_MULTI_SEGS 0x00008000
1442 /**< Device supports multi segment send. */
1443 #define DEV_TX_OFFLOAD_MBUF_FAST_FREE 0x00010000
1444 /**< Device supports optimization for fast release of mbufs.
1445 * When set application must guarantee that per-queue all mbufs comes from
1446 * the same mempool and has refcnt = 1.
1448 #define DEV_TX_OFFLOAD_SECURITY 0x00020000
1450 * Device supports generic UDP tunneled packet TSO.
1451 * Application must set PKT_TX_TUNNEL_UDP and other mbuf fields required
1454 #define DEV_TX_OFFLOAD_UDP_TNL_TSO 0x00040000
1456 * Device supports generic IP tunneled packet TSO.
1457 * Application must set PKT_TX_TUNNEL_IP and other mbuf fields required
1460 #define DEV_TX_OFFLOAD_IP_TNL_TSO 0x00080000
1461 /** Device supports outer UDP checksum */
1462 #define DEV_TX_OFFLOAD_OUTER_UDP_CKSUM 0x00100000
1464 * Device sends on time read from RTE_MBUF_DYNFIELD_TIMESTAMP_NAME
1465 * if RTE_MBUF_DYNFLAG_TX_TIMESTAMP_NAME is set in ol_flags.
1466 * The mbuf field and flag are registered when the offload is configured.
1468 #define DEV_TX_OFFLOAD_SEND_ON_TIMESTAMP 0x00200000
1470 * If new Tx offload capabilities are defined, they also must be
1471 * mentioned in rte_tx_offload_names in rte_ethdev.c file.
1474 /**@{@name Device capabilities
1475 * Non-offload capabilities reported in rte_eth_dev_info.dev_capa.
1477 /** Device supports Rx queue setup after device started. */
1478 #define RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP 0x00000001
1479 /** Device supports Tx queue setup after device started. */
1480 #define RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP 0x00000002
1484 * Fallback default preferred Rx/Tx port parameters.
1485 * These are used if an application requests default parameters
1486 * but the PMD does not provide preferred values.
1488 #define RTE_ETH_DEV_FALLBACK_RX_RINGSIZE 512
1489 #define RTE_ETH_DEV_FALLBACK_TX_RINGSIZE 512
1490 #define RTE_ETH_DEV_FALLBACK_RX_NBQUEUES 1
1491 #define RTE_ETH_DEV_FALLBACK_TX_NBQUEUES 1
1494 * Preferred Rx/Tx port parameters.
1495 * There are separate instances of this structure for transmission
1496 * and reception respectively.
1498 struct rte_eth_dev_portconf {
1499 uint16_t burst_size; /**< Device-preferred burst size */
1500 uint16_t ring_size; /**< Device-preferred size of queue rings */
1501 uint16_t nb_queues; /**< Device-preferred number of queues */
1505 * Default values for switch domain id when ethdev does not support switch
1506 * domain definitions.
1508 #define RTE_ETH_DEV_SWITCH_DOMAIN_ID_INVALID (UINT16_MAX)
1511 * Ethernet device associated switch information
1513 struct rte_eth_switch_info {
1514 const char *name; /**< switch name */
1515 uint16_t domain_id; /**< switch domain id */
1518 * mapping to the devices physical switch port as enumerated from the
1519 * perspective of the embedded interconnect/switch. For SR-IOV enabled
1520 * device this may correspond to the VF_ID of each virtual function,
1521 * but each driver should explicitly define the mapping of switch
1522 * port identifier to that physical interconnect/switch
1528 * @b EXPERIMENTAL: this structure may change without prior notice.
1530 * Ethernet device Rx buffer segmentation capabilities.
1532 struct rte_eth_rxseg_capa {
1534 uint32_t multi_pools:1; /**< Supports receiving to multiple pools.*/
1535 uint32_t offset_allowed:1; /**< Supports buffer offsets. */
1536 uint32_t offset_align_log2:4; /**< Required offset alignment. */
1537 uint16_t max_nseg; /**< Maximum amount of segments to split. */
1538 uint16_t reserved; /**< Reserved field. */
1542 * Ethernet device information
1546 * Ethernet device representor port type.
1548 enum rte_eth_representor_type {
1549 RTE_ETH_REPRESENTOR_NONE, /**< not a representor. */
1550 RTE_ETH_REPRESENTOR_VF, /**< representor of Virtual Function. */
1551 RTE_ETH_REPRESENTOR_SF, /**< representor of Sub Function. */
1552 RTE_ETH_REPRESENTOR_PF, /**< representor of Physical Function. */
1556 * A structure used to retrieve the contextual information of
1557 * an Ethernet device, such as the controlling driver of the
1560 struct rte_eth_dev_info {
1561 struct rte_device *device; /** Generic device information */
1562 const char *driver_name; /**< Device Driver name. */
1563 unsigned int if_index; /**< Index to bound host interface, or 0 if none.
1564 Use if_indextoname() to translate into an interface name. */
1565 uint16_t min_mtu; /**< Minimum MTU allowed */
1566 uint16_t max_mtu; /**< Maximum MTU allowed */
1567 const uint32_t *dev_flags; /**< Device flags */
1568 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */
1569 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */
1570 /** Maximum configurable size of LRO aggregated packet. */
1571 uint32_t max_lro_pkt_size;
1572 uint16_t max_rx_queues; /**< Maximum number of RX queues. */
1573 uint16_t max_tx_queues; /**< Maximum number of TX queues. */
1574 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */
1575 uint32_t max_hash_mac_addrs;
1576 /** Maximum number of hash MAC addresses for MTA and UTA. */
1577 uint16_t max_vfs; /**< Maximum number of VFs. */
1578 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
1579 struct rte_eth_rxseg_capa rx_seg_capa; /**< Segmentation capability.*/
1580 uint64_t rx_offload_capa;
1581 /**< All RX offload capabilities including all per-queue ones */
1582 uint64_t tx_offload_capa;
1583 /**< All TX offload capabilities including all per-queue ones */
1584 uint64_t rx_queue_offload_capa;
1585 /**< Device per-queue RX offload capabilities. */
1586 uint64_t tx_queue_offload_capa;
1587 /**< Device per-queue TX offload capabilities. */
1589 /**< Device redirection table size, the total number of entries. */
1590 uint8_t hash_key_size; /**< Hash key size in bytes */
1591 /** Bit mask of RSS offloads, the bit offset also means flow type */
1592 uint64_t flow_type_rss_offloads;
1593 struct rte_eth_rxconf default_rxconf; /**< Default RX configuration */
1594 struct rte_eth_txconf default_txconf; /**< Default TX configuration */
1595 uint16_t vmdq_queue_base; /**< First queue ID for VMDQ pools. */
1596 uint16_t vmdq_queue_num; /**< Queue number for VMDQ pools. */
1597 uint16_t vmdq_pool_base; /**< First ID of VMDQ pools. */
1598 struct rte_eth_desc_lim rx_desc_lim; /**< RX descriptors limits */
1599 struct rte_eth_desc_lim tx_desc_lim; /**< TX descriptors limits */
1600 uint32_t speed_capa; /**< Supported speeds bitmap (ETH_LINK_SPEED_). */
1601 /** Configured number of rx/tx queues */
1602 uint16_t nb_rx_queues; /**< Number of RX queues. */
1603 uint16_t nb_tx_queues; /**< Number of TX queues. */
1604 /** Rx parameter recommendations */
1605 struct rte_eth_dev_portconf default_rxportconf;
1606 /** Tx parameter recommendations */
1607 struct rte_eth_dev_portconf default_txportconf;
1608 /** Generic device capabilities (RTE_ETH_DEV_CAPA_). */
1611 * Switching information for ports on a device with a
1612 * embedded managed interconnect/switch.
1614 struct rte_eth_switch_info switch_info;
1616 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1617 void *reserved_ptrs[2]; /**< Reserved for future fields */
1620 /**@{@name Rx/Tx queue states */
1621 #define RTE_ETH_QUEUE_STATE_STOPPED 0 /**< Queue stopped. */
1622 #define RTE_ETH_QUEUE_STATE_STARTED 1 /**< Queue started. */
1623 #define RTE_ETH_QUEUE_STATE_HAIRPIN 2 /**< Queue used for hairpin. */
1627 * Ethernet device RX queue information structure.
1628 * Used to retrieve information about configured queue.
1630 struct rte_eth_rxq_info {
1631 struct rte_mempool *mp; /**< mempool used by that queue. */
1632 struct rte_eth_rxconf conf; /**< queue config parameters. */
1633 uint8_t scattered_rx; /**< scattered packets RX supported. */
1634 uint8_t queue_state; /**< one of RTE_ETH_QUEUE_STATE_*. */
1635 uint16_t nb_desc; /**< configured number of RXDs. */
1636 uint16_t rx_buf_size; /**< hardware receive buffer size. */
1637 } __rte_cache_min_aligned;
1640 * Ethernet device TX queue information structure.
1641 * Used to retrieve information about configured queue.
1643 struct rte_eth_txq_info {
1644 struct rte_eth_txconf conf; /**< queue config parameters. */
1645 uint16_t nb_desc; /**< configured number of TXDs. */
1646 uint8_t queue_state; /**< one of RTE_ETH_QUEUE_STATE_*. */
1647 } __rte_cache_min_aligned;
1649 /* Generic Burst mode flag definition, values can be ORed. */
1652 * If the queues have different burst mode description, this bit will be set
1653 * by PMD, then the application can iterate to retrieve burst description for
1656 #define RTE_ETH_BURST_FLAG_PER_QUEUE (1ULL << 0)
1659 * Ethernet device RX/TX queue packet burst mode information structure.
1660 * Used to retrieve information about packet burst mode setting.
1662 struct rte_eth_burst_mode {
1663 uint64_t flags; /**< The ORed values of RTE_ETH_BURST_FLAG_xxx */
1665 #define RTE_ETH_BURST_MODE_INFO_SIZE 1024 /**< Maximum size for information */
1666 char info[RTE_ETH_BURST_MODE_INFO_SIZE]; /**< burst mode information */
1669 /** Maximum name length for extended statistics counters */
1670 #define RTE_ETH_XSTATS_NAME_SIZE 64
1673 * An Ethernet device extended statistic structure
1675 * This structure is used by rte_eth_xstats_get() to provide
1676 * statistics that are not provided in the generic *rte_eth_stats*
1678 * It maps a name id, corresponding to an index in the array returned
1679 * by rte_eth_xstats_get_names(), to a statistic value.
1681 struct rte_eth_xstat {
1682 uint64_t id; /**< The index in xstats name array. */
1683 uint64_t value; /**< The statistic counter value. */
1687 * A name element for extended statistics.
1689 * An array of this structure is returned by rte_eth_xstats_get_names().
1690 * It lists the names of extended statistics for a PMD. The *rte_eth_xstat*
1691 * structure references these names by their array index.
1693 * The xstats should follow a common naming scheme.
1694 * Some names are standardized in rte_stats_strings.
1696 * - rx_missed_errors
1698 * - tx_size_128_to_255_packets
1700 struct rte_eth_xstat_name {
1701 char name[RTE_ETH_XSTATS_NAME_SIZE]; /**< The statistic name. */
1704 #define ETH_DCB_NUM_TCS 8
1705 #define ETH_MAX_VMDQ_POOL 64
1708 * A structure used to get the information of queue and
1709 * TC mapping on both TX and RX paths.
1711 struct rte_eth_dcb_tc_queue_mapping {
1712 /** rx queues assigned to tc per Pool */
1716 } tc_rxq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1717 /** rx queues assigned to tc per Pool */
1721 } tc_txq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1725 * A structure used to get the information of DCB.
1726 * It includes TC UP mapping and queue TC mapping.
1728 struct rte_eth_dcb_info {
1729 uint8_t nb_tcs; /**< number of TCs */
1730 uint8_t prio_tc[ETH_DCB_NUM_USER_PRIORITIES]; /**< Priority to tc */
1731 uint8_t tc_bws[ETH_DCB_NUM_TCS]; /**< TX BW percentage for each TC */
1732 /** rx queues assigned to tc */
1733 struct rte_eth_dcb_tc_queue_mapping tc_queue;
1737 * This enum indicates the possible Forward Error Correction (FEC) modes
1738 * of an ethdev port.
1740 enum rte_eth_fec_mode {
1741 RTE_ETH_FEC_NOFEC = 0, /**< FEC is off */
1742 RTE_ETH_FEC_AUTO, /**< FEC autonegotiation modes */
1743 RTE_ETH_FEC_BASER, /**< FEC using common algorithm */
1744 RTE_ETH_FEC_RS, /**< FEC using RS algorithm */
1747 /* Translate from FEC mode to FEC capa */
1748 #define RTE_ETH_FEC_MODE_TO_CAPA(x) (1U << (x))
1750 /* This macro indicates FEC capa mask */
1751 #define RTE_ETH_FEC_MODE_CAPA_MASK(x) (1U << (RTE_ETH_FEC_ ## x))
1753 /* A structure used to get capabilities per link speed */
1754 struct rte_eth_fec_capa {
1755 uint32_t speed; /**< Link speed (see ETH_SPEED_NUM_*) */
1756 uint32_t capa; /**< FEC capabilities bitmask */
1759 #define RTE_ETH_ALL RTE_MAX_ETHPORTS
1761 /* Macros to check for valid port */
1762 #define RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, retval) do { \
1763 if (!rte_eth_dev_is_valid_port(port_id)) { \
1764 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1769 #define RTE_ETH_VALID_PORTID_OR_RET(port_id) do { \
1770 if (!rte_eth_dev_is_valid_port(port_id)) { \
1771 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1776 /**@{@name L2 tunnel configuration */
1777 /**< l2 tunnel enable mask */
1778 #define ETH_L2_TUNNEL_ENABLE_MASK 0x00000001
1779 /**< l2 tunnel insertion mask */
1780 #define ETH_L2_TUNNEL_INSERTION_MASK 0x00000002
1781 /**< l2 tunnel stripping mask */
1782 #define ETH_L2_TUNNEL_STRIPPING_MASK 0x00000004
1783 /**< l2 tunnel forwarding mask */
1784 #define ETH_L2_TUNNEL_FORWARDING_MASK 0x00000008
1788 * Function type used for RX packet processing packet callbacks.
1790 * The callback function is called on RX with a burst of packets that have
1791 * been received on the given port and queue.
1794 * The Ethernet port on which RX is being performed.
1796 * The queue on the Ethernet port which is being used to receive the packets.
1798 * The burst of packets that have just been received.
1800 * The number of packets in the burst pointed to by "pkts".
1802 * The max number of packets that can be stored in the "pkts" array.
1804 * The arbitrary user parameter passed in by the application when the callback
1805 * was originally configured.
1807 * The number of packets returned to the user.
1809 typedef uint16_t (*rte_rx_callback_fn)(uint16_t port_id, uint16_t queue,
1810 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts,
1814 * Function type used for TX packet processing packet callbacks.
1816 * The callback function is called on TX with a burst of packets immediately
1817 * before the packets are put onto the hardware queue for transmission.
1820 * The Ethernet port on which TX is being performed.
1822 * The queue on the Ethernet port which is being used to transmit the packets.
1824 * The burst of packets that are about to be transmitted.
1826 * The number of packets in the burst pointed to by "pkts".
1828 * The arbitrary user parameter passed in by the application when the callback
1829 * was originally configured.
1831 * The number of packets to be written to the NIC.
1833 typedef uint16_t (*rte_tx_callback_fn)(uint16_t port_id, uint16_t queue,
1834 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param);
1837 * Possible states of an ethdev port.
1839 enum rte_eth_dev_state {
1840 /** Device is unused before being probed. */
1841 RTE_ETH_DEV_UNUSED = 0,
1842 /** Device is attached when allocated in probing. */
1843 RTE_ETH_DEV_ATTACHED,
1844 /** Device is in removed state when plug-out is detected. */
1845 RTE_ETH_DEV_REMOVED,
1848 struct rte_eth_dev_sriov {
1849 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */
1850 uint8_t nb_q_per_pool; /**< rx queue number per pool */
1851 uint16_t def_vmdq_idx; /**< Default pool num used for PF */
1852 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */
1854 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
1856 #define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN
1858 #define RTE_ETH_DEV_NO_OWNER 0
1860 #define RTE_ETH_MAX_OWNER_NAME_LEN 64
1862 struct rte_eth_dev_owner {
1863 uint64_t id; /**< The owner unique identifier. */
1864 char name[RTE_ETH_MAX_OWNER_NAME_LEN]; /**< The owner name. */
1867 /**@{@name Device flags
1868 * Flags internally saved in rte_eth_dev_data.dev_flags
1869 * and reported in rte_eth_dev_info.dev_flags.
1871 /** PMD supports thread-safe flow operations */
1872 #define RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE 0x0001
1873 /** Device supports link state interrupt */
1874 #define RTE_ETH_DEV_INTR_LSC 0x0002
1875 /** Device is a bonded slave */
1876 #define RTE_ETH_DEV_BONDED_SLAVE 0x0004
1877 /** Device supports device removal interrupt */
1878 #define RTE_ETH_DEV_INTR_RMV 0x0008
1879 /** Device is port representor */
1880 #define RTE_ETH_DEV_REPRESENTOR 0x0010
1881 /** Device does not support MAC change after started */
1882 #define RTE_ETH_DEV_NOLIVE_MAC_ADDR 0x0020
1884 * Queue xstats filled automatically by ethdev layer.
1885 * PMDs filling the queue xstats themselves should not set this flag
1887 #define RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS 0x0040
1891 * Iterates over valid ethdev ports owned by a specific owner.
1894 * The id of the next possible valid owned port.
1896 * The owner identifier.
1897 * RTE_ETH_DEV_NO_OWNER means iterate over all valid ownerless ports.
1899 * Next valid port id owned by owner_id, RTE_MAX_ETHPORTS if there is none.
1901 uint64_t rte_eth_find_next_owned_by(uint16_t port_id,
1902 const uint64_t owner_id);
1905 * Macro to iterate over all enabled ethdev ports owned by a specific owner.
1907 #define RTE_ETH_FOREACH_DEV_OWNED_BY(p, o) \
1908 for (p = rte_eth_find_next_owned_by(0, o); \
1909 (unsigned int)p < (unsigned int)RTE_MAX_ETHPORTS; \
1910 p = rte_eth_find_next_owned_by(p + 1, o))
1913 * Iterates over valid ethdev ports.
1916 * The id of the next possible valid port.
1918 * Next valid port id, RTE_MAX_ETHPORTS if there is none.
1920 uint16_t rte_eth_find_next(uint16_t port_id);
1923 * Macro to iterate over all enabled and ownerless ethdev ports.
1925 #define RTE_ETH_FOREACH_DEV(p) \
1926 RTE_ETH_FOREACH_DEV_OWNED_BY(p, RTE_ETH_DEV_NO_OWNER)
1929 * Iterates over ethdev ports of a specified device.
1931 * @param port_id_start
1932 * The id of the next possible valid port.
1934 * The generic device behind the ports to iterate.
1936 * Next port id of the device, possibly port_id_start,
1937 * RTE_MAX_ETHPORTS if there is none.
1940 rte_eth_find_next_of(uint16_t port_id_start,
1941 const struct rte_device *parent);
1944 * Macro to iterate over all ethdev ports of a specified device.
1947 * The id of the matching port being iterated.
1949 * The rte_device pointer matching the iterated ports.
1951 #define RTE_ETH_FOREACH_DEV_OF(port_id, parent) \
1952 for (port_id = rte_eth_find_next_of(0, parent); \
1953 port_id < RTE_MAX_ETHPORTS; \
1954 port_id = rte_eth_find_next_of(port_id + 1, parent))
1957 * Iterates over sibling ethdev ports (i.e. sharing the same rte_device).
1959 * @param port_id_start
1960 * The id of the next possible valid sibling port.
1961 * @param ref_port_id
1962 * The id of a reference port to compare rte_device with.
1964 * Next sibling port id, possibly port_id_start or ref_port_id itself,
1965 * RTE_MAX_ETHPORTS if there is none.
1968 rte_eth_find_next_sibling(uint16_t port_id_start, uint16_t ref_port_id);
1971 * Macro to iterate over all ethdev ports sharing the same rte_device
1972 * as the specified port.
1973 * Note: the specified reference port is part of the loop iterations.
1976 * The id of the matching port being iterated.
1977 * @param ref_port_id
1978 * The id of the port being compared.
1980 #define RTE_ETH_FOREACH_DEV_SIBLING(port_id, ref_port_id) \
1981 for (port_id = rte_eth_find_next_sibling(0, ref_port_id); \
1982 port_id < RTE_MAX_ETHPORTS; \
1983 port_id = rte_eth_find_next_sibling(port_id + 1, ref_port_id))
1987 * @b EXPERIMENTAL: this API may change without prior notice.
1989 * Get a new unique owner identifier.
1990 * An owner identifier is used to owns Ethernet devices by only one DPDK entity
1991 * to avoid multiple management of device by different entities.
1994 * Owner identifier pointer.
1996 * Negative errno value on error, 0 on success.
1999 int rte_eth_dev_owner_new(uint64_t *owner_id);
2003 * @b EXPERIMENTAL: this API may change without prior notice.
2005 * Set an Ethernet device owner.
2008 * The identifier of the port to own.
2010 * The owner pointer.
2012 * Negative errno value on error, 0 on success.
2015 int rte_eth_dev_owner_set(const uint16_t port_id,
2016 const struct rte_eth_dev_owner *owner);
2020 * @b EXPERIMENTAL: this API may change without prior notice.
2022 * Unset Ethernet device owner to make the device ownerless.
2025 * The identifier of port to make ownerless.
2027 * The owner identifier.
2029 * 0 on success, negative errno value on error.
2032 int rte_eth_dev_owner_unset(const uint16_t port_id,
2033 const uint64_t owner_id);
2037 * @b EXPERIMENTAL: this API may change without prior notice.
2039 * Remove owner from all Ethernet devices owned by a specific owner.
2042 * The owner identifier.
2044 * 0 on success, negative errno value on error.
2047 int rte_eth_dev_owner_delete(const uint64_t owner_id);
2051 * @b EXPERIMENTAL: this API may change without prior notice.
2053 * Get the owner of an Ethernet device.
2056 * The port identifier.
2058 * The owner structure pointer to fill.
2060 * 0 on success, negative errno value on error..
2063 int rte_eth_dev_owner_get(const uint16_t port_id,
2064 struct rte_eth_dev_owner *owner);
2067 * Get the number of ports which are usable for the application.
2069 * These devices must be iterated by using the macro
2070 * ``RTE_ETH_FOREACH_DEV`` or ``RTE_ETH_FOREACH_DEV_OWNED_BY``
2071 * to deal with non-contiguous ranges of devices.
2074 * The count of available Ethernet devices.
2076 uint16_t rte_eth_dev_count_avail(void);
2079 * Get the total number of ports which are allocated.
2081 * Some devices may not be available for the application.
2084 * The total count of Ethernet devices.
2086 uint16_t rte_eth_dev_count_total(void);
2089 * Convert a numerical speed in Mbps to a bitmap flag that can be used in
2090 * the bitmap link_speeds of the struct rte_eth_conf
2093 * Numerical speed value in Mbps
2095 * ETH_LINK_[HALF/FULL]_DUPLEX (only for 10/100M speeds)
2097 * 0 if the speed cannot be mapped
2099 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex);
2102 * Get DEV_RX_OFFLOAD_* flag name.
2107 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
2109 const char *rte_eth_dev_rx_offload_name(uint64_t offload);
2112 * Get DEV_TX_OFFLOAD_* flag name.
2117 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
2119 const char *rte_eth_dev_tx_offload_name(uint64_t offload);
2122 * Configure an Ethernet device.
2123 * This function must be invoked first before any other function in the
2124 * Ethernet API. This function can also be re-invoked when a device is in the
2128 * The port identifier of the Ethernet device to configure.
2129 * @param nb_rx_queue
2130 * The number of receive queues to set up for the Ethernet device.
2131 * @param nb_tx_queue
2132 * The number of transmit queues to set up for the Ethernet device.
2134 * The pointer to the configuration data to be used for the Ethernet device.
2135 * The *rte_eth_conf* structure includes:
2136 * - the hardware offload features to activate, with dedicated fields for
2137 * each statically configurable offload hardware feature provided by
2138 * Ethernet devices, such as IP checksum or VLAN tag stripping for
2140 * The Rx offload bitfield API is obsolete and will be deprecated.
2141 * Applications should set the ignore_bitfield_offloads bit on *rxmode*
2142 * structure and use offloads field to set per-port offloads instead.
2143 * - Any offloading set in eth_conf->[rt]xmode.offloads must be within
2144 * the [rt]x_offload_capa returned from rte_eth_dev_info_get().
2145 * Any type of device supported offloading set in the input argument
2146 * eth_conf->[rt]xmode.offloads to rte_eth_dev_configure() is enabled
2147 * on all queues and it can't be disabled in rte_eth_[rt]x_queue_setup()
2148 * - the Receive Side Scaling (RSS) configuration when using multiple RX
2149 * queues per port. Any RSS hash function set in eth_conf->rss_conf.rss_hf
2150 * must be within the flow_type_rss_offloads provided by drivers via
2151 * rte_eth_dev_info_get() API.
2153 * Embedding all configuration information in a single data structure
2154 * is the more flexible method that allows the addition of new features
2155 * without changing the syntax of the API.
2157 * - 0: Success, device configured.
2158 * - <0: Error code returned by the driver configuration function.
2160 int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
2161 uint16_t nb_tx_queue, const struct rte_eth_conf *eth_conf);
2165 * @b EXPERIMENTAL: this API may change without prior notice.
2167 * Check if an Ethernet device was physically removed.
2170 * The port identifier of the Ethernet device.
2172 * 1 when the Ethernet device is removed, otherwise 0.
2176 rte_eth_dev_is_removed(uint16_t port_id);
2179 * Allocate and set up a receive queue for an Ethernet device.
2181 * The function allocates a contiguous block of memory for *nb_rx_desc*
2182 * receive descriptors from a memory zone associated with *socket_id*
2183 * and initializes each receive descriptor with a network buffer allocated
2184 * from the memory pool *mb_pool*.
2187 * The port identifier of the Ethernet device.
2188 * @param rx_queue_id
2189 * The index of the receive queue to set up.
2190 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2191 * to rte_eth_dev_configure().
2193 * The number of receive descriptors to allocate for the receive ring.
2195 * The *socket_id* argument is the socket identifier in case of NUMA.
2196 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
2197 * the DMA memory allocated for the receive descriptors of the ring.
2199 * The pointer to the configuration data to be used for the receive queue.
2200 * NULL value is allowed, in which case default RX configuration
2202 * The *rx_conf* structure contains an *rx_thresh* structure with the values
2203 * of the Prefetch, Host, and Write-Back threshold registers of the receive
2205 * In addition it contains the hardware offloads features to activate using
2206 * the DEV_RX_OFFLOAD_* flags.
2207 * If an offloading set in rx_conf->offloads
2208 * hasn't been set in the input argument eth_conf->rxmode.offloads
2209 * to rte_eth_dev_configure(), it is a new added offloading, it must be
2210 * per-queue type and it is enabled for the queue.
2211 * No need to repeat any bit in rx_conf->offloads which has already been
2212 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
2213 * at port level can't be disabled at queue level.
2214 * The configuration structure also contains the pointer to the array
2215 * of the receiving buffer segment descriptions, see rx_seg and rx_nseg
2216 * fields, this extended configuration might be used by split offloads like
2217 * RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mb_pool is not NULL,
2218 * the extended configuration fields must be set to NULL and zero.
2220 * The pointer to the memory pool from which to allocate *rte_mbuf* network
2221 * memory buffers to populate each descriptor of the receive ring. There are
2222 * two options to provide Rx buffer configuration:
2224 * mb_pool is not NULL, rx_conf.rx_nseg is 0.
2225 * - multiple segments description:
2226 * mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
2227 * Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.
2230 * - 0: Success, receive queue correctly set up.
2231 * - -EIO: if device is removed.
2232 * - -ENODEV: if *port_id* is invalid.
2233 * - -EINVAL: The memory pool pointer is null or the size of network buffers
2234 * which can be allocated from this memory pool does not fit the various
2235 * buffer sizes allowed by the device controller.
2236 * - -ENOMEM: Unable to allocate the receive ring descriptors or to
2237 * allocate network memory buffers from the memory pool when
2238 * initializing receive descriptors.
2240 int rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
2241 uint16_t nb_rx_desc, unsigned int socket_id,
2242 const struct rte_eth_rxconf *rx_conf,
2243 struct rte_mempool *mb_pool);
2247 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2249 * Allocate and set up a hairpin receive queue for an Ethernet device.
2251 * The function set up the selected queue to be used in hairpin.
2254 * The port identifier of the Ethernet device.
2255 * @param rx_queue_id
2256 * The index of the receive queue to set up.
2257 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2258 * to rte_eth_dev_configure().
2260 * The number of receive descriptors to allocate for the receive ring.
2261 * 0 means the PMD will use default value.
2263 * The pointer to the hairpin configuration.
2266 * - (0) if successful.
2267 * - (-ENODEV) if *port_id* is invalid.
2268 * - (-ENOTSUP) if hardware doesn't support.
2269 * - (-EINVAL) if bad parameter.
2270 * - (-ENOMEM) if unable to allocate the resources.
2273 int rte_eth_rx_hairpin_queue_setup
2274 (uint16_t port_id, uint16_t rx_queue_id, uint16_t nb_rx_desc,
2275 const struct rte_eth_hairpin_conf *conf);
2278 * Allocate and set up a transmit queue for an Ethernet device.
2281 * The port identifier of the Ethernet device.
2282 * @param tx_queue_id
2283 * The index of the transmit queue to set up.
2284 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2285 * to rte_eth_dev_configure().
2287 * The number of transmit descriptors to allocate for the transmit ring.
2289 * The *socket_id* argument is the socket identifier in case of NUMA.
2290 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
2291 * the DMA memory allocated for the transmit descriptors of the ring.
2293 * The pointer to the configuration data to be used for the transmit queue.
2294 * NULL value is allowed, in which case default TX configuration
2296 * The *tx_conf* structure contains the following data:
2297 * - The *tx_thresh* structure with the values of the Prefetch, Host, and
2298 * Write-Back threshold registers of the transmit ring.
2299 * When setting Write-Back threshold to the value greater then zero,
2300 * *tx_rs_thresh* value should be explicitly set to one.
2301 * - The *tx_free_thresh* value indicates the [minimum] number of network
2302 * buffers that must be pending in the transmit ring to trigger their
2303 * [implicit] freeing by the driver transmit function.
2304 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit
2305 * descriptors that must be pending in the transmit ring before setting the
2306 * RS bit on a descriptor by the driver transmit function.
2307 * The *tx_rs_thresh* value should be less or equal then
2308 * *tx_free_thresh* value, and both of them should be less then
2310 * - The *offloads* member contains Tx offloads to be enabled.
2311 * If an offloading set in tx_conf->offloads
2312 * hasn't been set in the input argument eth_conf->txmode.offloads
2313 * to rte_eth_dev_configure(), it is a new added offloading, it must be
2314 * per-queue type and it is enabled for the queue.
2315 * No need to repeat any bit in tx_conf->offloads which has already been
2316 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
2317 * at port level can't be disabled at queue level.
2319 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces
2320 * the transmit function to use default values.
2322 * - 0: Success, the transmit queue is correctly set up.
2323 * - -ENOMEM: Unable to allocate the transmit ring descriptors.
2325 int rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,
2326 uint16_t nb_tx_desc, unsigned int socket_id,
2327 const struct rte_eth_txconf *tx_conf);
2331 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2333 * Allocate and set up a transmit hairpin queue for an Ethernet device.
2336 * The port identifier of the Ethernet device.
2337 * @param tx_queue_id
2338 * The index of the transmit queue to set up.
2339 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2340 * to rte_eth_dev_configure().
2342 * The number of transmit descriptors to allocate for the transmit ring.
2343 * 0 to set default PMD value.
2345 * The hairpin configuration.
2348 * - (0) if successful.
2349 * - (-ENODEV) if *port_id* is invalid.
2350 * - (-ENOTSUP) if hardware doesn't support.
2351 * - (-EINVAL) if bad parameter.
2352 * - (-ENOMEM) if unable to allocate the resources.
2355 int rte_eth_tx_hairpin_queue_setup
2356 (uint16_t port_id, uint16_t tx_queue_id, uint16_t nb_tx_desc,
2357 const struct rte_eth_hairpin_conf *conf);
2361 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2363 * Get all the hairpin peer Rx / Tx ports of the current port.
2364 * The caller should ensure that the array is large enough to save the ports
2368 * The port identifier of the Ethernet device.
2370 * Pointer to the array to store the peer ports list.
2372 * Length of the array to store the port identifiers.
2374 * Current port to peer port direction
2375 * positive - current used as Tx to get all peer Rx ports.
2376 * zero - current used as Rx to get all peer Tx ports.
2379 * - (0 or positive) actual peer ports number.
2380 * - (-EINVAL) if bad parameter.
2381 * - (-ENODEV) if *port_id* invalid
2382 * - (-ENOTSUP) if hardware doesn't support.
2383 * - Others detailed errors from PMD drivers.
2386 int rte_eth_hairpin_get_peer_ports(uint16_t port_id, uint16_t *peer_ports,
2387 size_t len, uint32_t direction);
2391 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2393 * Bind all hairpin Tx queues of one port to the Rx queues of the peer port.
2394 * It is only allowed to call this function after all hairpin queues are
2395 * configured properly and the devices are in started state.
2398 * The identifier of the Tx port.
2400 * The identifier of peer Rx port.
2401 * RTE_MAX_ETHPORTS is allowed for the traversal of all devices.
2402 * Rx port ID could have the same value as Tx port ID.
2405 * - (0) if successful.
2406 * - (-ENODEV) if Tx port ID is invalid.
2407 * - (-EBUSY) if device is not in started state.
2408 * - (-ENOTSUP) if hardware doesn't support.
2409 * - Others detailed errors from PMD drivers.
2412 int rte_eth_hairpin_bind(uint16_t tx_port, uint16_t rx_port);
2416 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2418 * Unbind all hairpin Tx queues of one port from the Rx queues of the peer port.
2419 * This should be called before closing the Tx or Rx devices, if the bind
2420 * function is called before.
2421 * After unbinding the hairpin ports pair, it is allowed to bind them again.
2422 * Changing queues configuration should be after stopping the device(s).
2425 * The identifier of the Tx port.
2427 * The identifier of peer Rx port.
2428 * RTE_MAX_ETHPORTS is allowed for traversal of all devices.
2429 * Rx port ID could have the same value as Tx port ID.
2432 * - (0) if successful.
2433 * - (-ENODEV) if Tx port ID is invalid.
2434 * - (-EBUSY) if device is in stopped state.
2435 * - (-ENOTSUP) if hardware doesn't support.
2436 * - Others detailed errors from PMD drivers.
2439 int rte_eth_hairpin_unbind(uint16_t tx_port, uint16_t rx_port);
2442 * Return the NUMA socket to which an Ethernet device is connected
2445 * The port identifier of the Ethernet device
2447 * The NUMA socket id to which the Ethernet device is connected or
2448 * a default of zero if the socket could not be determined.
2449 * -1 is returned is the port_id value is out of range.
2451 int rte_eth_dev_socket_id(uint16_t port_id);
2454 * Check if port_id of device is attached
2457 * The port identifier of the Ethernet device
2459 * - 0 if port is out of range or not attached
2460 * - 1 if device is attached
2462 int rte_eth_dev_is_valid_port(uint16_t port_id);
2465 * Start specified RX queue of a port. It is used when rx_deferred_start
2466 * flag of the specified queue is true.
2469 * The port identifier of the Ethernet device
2470 * @param rx_queue_id
2471 * The index of the rx queue to update the ring.
2472 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2473 * to rte_eth_dev_configure().
2475 * - 0: Success, the receive queue is started.
2476 * - -ENODEV: if *port_id* is invalid.
2477 * - -EINVAL: The queue_id out of range or belong to hairpin.
2478 * - -EIO: if device is removed.
2479 * - -ENOTSUP: The function not supported in PMD driver.
2481 int rte_eth_dev_rx_queue_start(uint16_t port_id, uint16_t rx_queue_id);
2484 * Stop specified RX queue of a port
2487 * The port identifier of the Ethernet device
2488 * @param rx_queue_id
2489 * The index of the rx queue to update the ring.
2490 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2491 * to rte_eth_dev_configure().
2493 * - 0: Success, the receive queue is stopped.
2494 * - -ENODEV: if *port_id* is invalid.
2495 * - -EINVAL: The queue_id out of range or belong to hairpin.
2496 * - -EIO: if device is removed.
2497 * - -ENOTSUP: The function not supported in PMD driver.
2499 int rte_eth_dev_rx_queue_stop(uint16_t port_id, uint16_t rx_queue_id);
2502 * Start TX for specified queue of a port. It is used when tx_deferred_start
2503 * flag of the specified queue is true.
2506 * The port identifier of the Ethernet device
2507 * @param tx_queue_id
2508 * The index of the tx queue to update the ring.
2509 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2510 * to rte_eth_dev_configure().
2512 * - 0: Success, the transmit queue is started.
2513 * - -ENODEV: if *port_id* is invalid.
2514 * - -EINVAL: The queue_id out of range or belong to hairpin.
2515 * - -EIO: if device is removed.
2516 * - -ENOTSUP: The function not supported in PMD driver.
2518 int rte_eth_dev_tx_queue_start(uint16_t port_id, uint16_t tx_queue_id);
2521 * Stop specified TX queue of a port
2524 * The port identifier of the Ethernet device
2525 * @param tx_queue_id
2526 * The index of the tx queue to update the ring.
2527 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2528 * to rte_eth_dev_configure().
2530 * - 0: Success, the transmit queue is stopped.
2531 * - -ENODEV: if *port_id* is invalid.
2532 * - -EINVAL: The queue_id out of range or belong to hairpin.
2533 * - -EIO: if device is removed.
2534 * - -ENOTSUP: The function not supported in PMD driver.
2536 int rte_eth_dev_tx_queue_stop(uint16_t port_id, uint16_t tx_queue_id);
2539 * Start an Ethernet device.
2541 * The device start step is the last one and consists of setting the configured
2542 * offload features and in starting the transmit and the receive units of the
2545 * Device RTE_ETH_DEV_NOLIVE_MAC_ADDR flag causes MAC address to be set before
2546 * PMD port start callback function is invoked.
2548 * On success, all basic functions exported by the Ethernet API (link status,
2549 * receive/transmit, and so on) can be invoked.
2552 * The port identifier of the Ethernet device.
2554 * - 0: Success, Ethernet device started.
2555 * - <0: Error code of the driver device start function.
2557 int rte_eth_dev_start(uint16_t port_id);
2560 * Stop an Ethernet device. The device can be restarted with a call to
2561 * rte_eth_dev_start()
2564 * The port identifier of the Ethernet device.
2566 * - 0: Success, Ethernet device stopped.
2567 * - <0: Error code of the driver device stop function.
2569 int rte_eth_dev_stop(uint16_t port_id);
2572 * Link up an Ethernet device.
2574 * Set device link up will re-enable the device rx/tx
2575 * functionality after it is previously set device linked down.
2578 * The port identifier of the Ethernet device.
2580 * - 0: Success, Ethernet device linked up.
2581 * - <0: Error code of the driver device link up function.
2583 int rte_eth_dev_set_link_up(uint16_t port_id);
2586 * Link down an Ethernet device.
2587 * The device rx/tx functionality will be disabled if success,
2588 * and it can be re-enabled with a call to
2589 * rte_eth_dev_set_link_up()
2592 * The port identifier of the Ethernet device.
2594 int rte_eth_dev_set_link_down(uint16_t port_id);
2597 * Close a stopped Ethernet device. The device cannot be restarted!
2598 * The function frees all port resources.
2601 * The port identifier of the Ethernet device.
2603 * - Zero if the port is closed successfully.
2604 * - Negative if something went wrong.
2606 int rte_eth_dev_close(uint16_t port_id);
2609 * Reset a Ethernet device and keep its port id.
2611 * When a port has to be reset passively, the DPDK application can invoke
2612 * this function. For example when a PF is reset, all its VFs should also
2613 * be reset. Normally a DPDK application can invoke this function when
2614 * RTE_ETH_EVENT_INTR_RESET event is detected, but can also use it to start
2615 * a port reset in other circumstances.
2617 * When this function is called, it first stops the port and then calls the
2618 * PMD specific dev_uninit( ) and dev_init( ) to return the port to initial
2619 * state, in which no Tx and Rx queues are setup, as if the port has been
2620 * reset and not started. The port keeps the port id it had before the
2623 * After calling rte_eth_dev_reset( ), the application should use
2624 * rte_eth_dev_configure( ), rte_eth_rx_queue_setup( ),
2625 * rte_eth_tx_queue_setup( ), and rte_eth_dev_start( )
2626 * to reconfigure the device as appropriate.
2628 * Note: To avoid unexpected behavior, the application should stop calling
2629 * Tx and Rx functions before calling rte_eth_dev_reset( ). For thread
2630 * safety, all these controlling functions should be called from the same
2634 * The port identifier of the Ethernet device.
2637 * - (0) if successful.
2638 * - (-ENODEV) if *port_id* is invalid.
2639 * - (-ENOTSUP) if hardware doesn't support this function.
2640 * - (-EPERM) if not ran from the primary process.
2641 * - (-EIO) if re-initialisation failed or device is removed.
2642 * - (-ENOMEM) if the reset failed due to OOM.
2643 * - (-EAGAIN) if the reset temporarily failed and should be retried later.
2645 int rte_eth_dev_reset(uint16_t port_id);
2648 * Enable receipt in promiscuous mode for an Ethernet device.
2651 * The port identifier of the Ethernet device.
2653 * - (0) if successful.
2654 * - (-ENOTSUP) if support for promiscuous_enable() does not exist
2656 * - (-ENODEV) if *port_id* invalid.
2658 int rte_eth_promiscuous_enable(uint16_t port_id);
2661 * Disable receipt in promiscuous mode for an Ethernet device.
2664 * The port identifier of the Ethernet device.
2666 * - (0) if successful.
2667 * - (-ENOTSUP) if support for promiscuous_disable() does not exist
2669 * - (-ENODEV) if *port_id* invalid.
2671 int rte_eth_promiscuous_disable(uint16_t port_id);
2674 * Return the value of promiscuous mode for an Ethernet device.
2677 * The port identifier of the Ethernet device.
2679 * - (1) if promiscuous is enabled
2680 * - (0) if promiscuous is disabled.
2683 int rte_eth_promiscuous_get(uint16_t port_id);
2686 * Enable the receipt of any multicast frame by an Ethernet device.
2689 * The port identifier of the Ethernet device.
2691 * - (0) if successful.
2692 * - (-ENOTSUP) if support for allmulticast_enable() does not exist
2694 * - (-ENODEV) if *port_id* invalid.
2696 int rte_eth_allmulticast_enable(uint16_t port_id);
2699 * Disable the receipt of all multicast frames by an Ethernet device.
2702 * The port identifier of the Ethernet device.
2704 * - (0) if successful.
2705 * - (-ENOTSUP) if support for allmulticast_disable() does not exist
2707 * - (-ENODEV) if *port_id* invalid.
2709 int rte_eth_allmulticast_disable(uint16_t port_id);
2712 * Return the value of allmulticast mode for an Ethernet device.
2715 * The port identifier of the Ethernet device.
2717 * - (1) if allmulticast is enabled
2718 * - (0) if allmulticast is disabled.
2721 int rte_eth_allmulticast_get(uint16_t port_id);
2724 * Retrieve the link status (up/down), the duplex mode (half/full),
2725 * the negotiation (auto/fixed), and if available, the speed (Mbps).
2727 * It might need to wait up to 9 seconds.
2728 * @see rte_eth_link_get_nowait.
2731 * The port identifier of the Ethernet device.
2733 * Link information written back.
2735 * - (0) if successful.
2736 * - (-ENOTSUP) if the function is not supported in PMD driver.
2737 * - (-ENODEV) if *port_id* invalid.
2738 * - (-EINVAL) if bad parameter.
2740 int rte_eth_link_get(uint16_t port_id, struct rte_eth_link *link);
2743 * Retrieve the link status (up/down), the duplex mode (half/full),
2744 * the negotiation (auto/fixed), and if available, the speed (Mbps).
2747 * The port identifier of the Ethernet device.
2749 * Link information written back.
2751 * - (0) if successful.
2752 * - (-ENOTSUP) if the function is not supported in PMD driver.
2753 * - (-ENODEV) if *port_id* invalid.
2754 * - (-EINVAL) if bad parameter.
2756 int rte_eth_link_get_nowait(uint16_t port_id, struct rte_eth_link *link);
2760 * @b EXPERIMENTAL: this API may change without prior notice.
2762 * The function converts a link_speed to a string. It handles all special
2763 * values like unknown or none speed.
2766 * link_speed of rte_eth_link struct
2768 * Link speed in textual format. It's pointer to immutable memory.
2769 * No free is required.
2772 const char *rte_eth_link_speed_to_str(uint32_t link_speed);
2776 * @b EXPERIMENTAL: this API may change without prior notice.
2778 * The function converts a rte_eth_link struct representing a link status to
2782 * A pointer to a string to be filled with textual representation of
2783 * device status. At least ETH_LINK_MAX_STR_LEN bytes should be allocated to
2784 * store default link status text.
2786 * Length of available memory at 'str' string.
2788 * Link status returned by rte_eth_link_get function
2790 * Number of bytes written to str array or -EINVAL if bad parameter.
2793 int rte_eth_link_to_str(char *str, size_t len,
2794 const struct rte_eth_link *eth_link);
2797 * Retrieve the general I/O statistics of an Ethernet device.
2800 * The port identifier of the Ethernet device.
2802 * A pointer to a structure of type *rte_eth_stats* to be filled with
2803 * the values of device counters for the following set of statistics:
2804 * - *ipackets* with the total of successfully received packets.
2805 * - *opackets* with the total of successfully transmitted packets.
2806 * - *ibytes* with the total of successfully received bytes.
2807 * - *obytes* with the total of successfully transmitted bytes.
2808 * - *ierrors* with the total of erroneous received packets.
2809 * - *oerrors* with the total of failed transmitted packets.
2811 * Zero if successful. Non-zero otherwise.
2813 int rte_eth_stats_get(uint16_t port_id, struct rte_eth_stats *stats);
2816 * Reset the general I/O statistics of an Ethernet device.
2819 * The port identifier of the Ethernet device.
2821 * - (0) if device notified to reset stats.
2822 * - (-ENOTSUP) if hardware doesn't support.
2823 * - (-ENODEV) if *port_id* invalid.
2824 * - (<0): Error code of the driver stats reset function.
2826 int rte_eth_stats_reset(uint16_t port_id);
2829 * Retrieve names of extended statistics of an Ethernet device.
2831 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2833 * xstats_names[i].name => xstats[i].value
2835 * And the array index is same with id field of 'struct rte_eth_xstat':
2838 * This assumption makes key-value pair matching less flexible but simpler.
2841 * The port identifier of the Ethernet device.
2842 * @param xstats_names
2843 * An rte_eth_xstat_name array of at least *size* elements to
2844 * be filled. If set to NULL, the function returns the required number
2847 * The size of the xstats_names array (number of elements).
2849 * - A positive value lower or equal to size: success. The return value
2850 * is the number of entries filled in the stats table.
2851 * - A positive value higher than size: error, the given statistics table
2852 * is too small. The return value corresponds to the size that should
2853 * be given to succeed. The entries in the table are not valid and
2854 * shall not be used by the caller.
2855 * - A negative value on error (invalid port id).
2857 int rte_eth_xstats_get_names(uint16_t port_id,
2858 struct rte_eth_xstat_name *xstats_names,
2862 * Retrieve extended statistics of an Ethernet device.
2864 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2866 * xstats_names[i].name => xstats[i].value
2868 * And the array index is same with id field of 'struct rte_eth_xstat':
2871 * This assumption makes key-value pair matching less flexible but simpler.
2874 * The port identifier of the Ethernet device.
2876 * A pointer to a table of structure of type *rte_eth_xstat*
2877 * to be filled with device statistics ids and values.
2878 * This parameter can be set to NULL if n is 0.
2880 * The size of the xstats array (number of elements).
2882 * - A positive value lower or equal to n: success. The return value
2883 * is the number of entries filled in the stats table.
2884 * - A positive value higher than n: error, the given statistics table
2885 * is too small. The return value corresponds to the size that should
2886 * be given to succeed. The entries in the table are not valid and
2887 * shall not be used by the caller.
2888 * - A negative value on error (invalid port id).
2890 int rte_eth_xstats_get(uint16_t port_id, struct rte_eth_xstat *xstats,
2894 * Retrieve names of extended statistics of an Ethernet device.
2897 * The port identifier of the Ethernet device.
2898 * @param xstats_names
2899 * An rte_eth_xstat_name array of at least *size* elements to
2900 * be filled. If set to NULL, the function returns the required number
2903 * IDs array given by app to retrieve specific statistics
2905 * The size of the xstats_names array (number of elements).
2907 * - A positive value lower or equal to size: success. The return value
2908 * is the number of entries filled in the stats table.
2909 * - A positive value higher than size: error, the given statistics table
2910 * is too small. The return value corresponds to the size that should
2911 * be given to succeed. The entries in the table are not valid and
2912 * shall not be used by the caller.
2913 * - A negative value on error (invalid port id).
2916 rte_eth_xstats_get_names_by_id(uint16_t port_id,
2917 struct rte_eth_xstat_name *xstats_names, unsigned int size,
2921 * Retrieve extended statistics of an Ethernet device.
2924 * The port identifier of the Ethernet device.
2926 * A pointer to an ids array passed by application. This tells which
2927 * statistics values function should retrieve. This parameter
2928 * can be set to NULL if size is 0. In this case function will retrieve
2929 * all available statistics.
2931 * A pointer to a table to be filled with device statistics values.
2933 * The size of the ids array (number of elements).
2935 * - A positive value lower or equal to size: success. The return value
2936 * is the number of entries filled in the stats table.
2937 * - A positive value higher than size: error, the given statistics table
2938 * is too small. The return value corresponds to the size that should
2939 * be given to succeed. The entries in the table are not valid and
2940 * shall not be used by the caller.
2941 * - A negative value on error (invalid port id).
2943 int rte_eth_xstats_get_by_id(uint16_t port_id, const uint64_t *ids,
2944 uint64_t *values, unsigned int size);
2947 * Gets the ID of a statistic from its name.
2949 * This function searches for the statistics using string compares, and
2950 * as such should not be used on the fast-path. For fast-path retrieval of
2951 * specific statistics, store the ID as provided in *id* from this function,
2952 * and pass the ID to rte_eth_xstats_get()
2954 * @param port_id The port to look up statistics from
2955 * @param xstat_name The name of the statistic to return
2956 * @param[out] id A pointer to an app-supplied uint64_t which should be
2957 * set to the ID of the stat if the stat exists.
2960 * -ENODEV for invalid port_id,
2961 * -EIO if device is removed,
2962 * -EINVAL if the xstat_name doesn't exist in port_id
2963 * -ENOMEM if bad parameter.
2965 int rte_eth_xstats_get_id_by_name(uint16_t port_id, const char *xstat_name,
2969 * Reset extended statistics of an Ethernet device.
2972 * The port identifier of the Ethernet device.
2974 * - (0) if device notified to reset extended stats.
2975 * - (-ENOTSUP) if pmd doesn't support both
2976 * extended stats and basic stats reset.
2977 * - (-ENODEV) if *port_id* invalid.
2978 * - (<0): Error code of the driver xstats reset function.
2980 int rte_eth_xstats_reset(uint16_t port_id);
2983 * Set a mapping for the specified transmit queue to the specified per-queue
2984 * statistics counter.
2987 * The port identifier of the Ethernet device.
2988 * @param tx_queue_id
2989 * The index of the transmit queue for which a queue stats mapping is required.
2990 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2991 * to rte_eth_dev_configure().
2993 * The per-queue packet statistics functionality number that the transmit
2994 * queue is to be assigned.
2995 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
2996 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256.
2998 * Zero if successful. Non-zero otherwise.
3000 int rte_eth_dev_set_tx_queue_stats_mapping(uint16_t port_id,
3001 uint16_t tx_queue_id, uint8_t stat_idx);
3004 * Set a mapping for the specified receive queue to the specified per-queue
3005 * statistics counter.
3008 * The port identifier of the Ethernet device.
3009 * @param rx_queue_id
3010 * The index of the receive queue for which a queue stats mapping is required.
3011 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3012 * to rte_eth_dev_configure().
3014 * The per-queue packet statistics functionality number that the receive
3015 * queue is to be assigned.
3016 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
3017 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256.
3019 * Zero if successful. Non-zero otherwise.
3021 int rte_eth_dev_set_rx_queue_stats_mapping(uint16_t port_id,
3022 uint16_t rx_queue_id,
3026 * Retrieve the Ethernet address of an Ethernet device.
3029 * The port identifier of the Ethernet device.
3031 * A pointer to a structure of type *ether_addr* to be filled with
3032 * the Ethernet address of the Ethernet device.
3034 * - (0) if successful
3035 * - (-ENODEV) if *port_id* invalid.
3036 * - (-EINVAL) if bad parameter.
3038 int rte_eth_macaddr_get(uint16_t port_id, struct rte_ether_addr *mac_addr);
3041 * Retrieve the contextual information of an Ethernet device.
3043 * As part of this function, a number of of fields in dev_info will be
3044 * initialized as follows:
3049 * Where lim is defined within the rte_eth_dev_info_get as
3051 * const struct rte_eth_desc_lim lim = {
3052 * .nb_max = UINT16_MAX,
3055 * .nb_seg_max = UINT16_MAX,
3056 * .nb_mtu_seg_max = UINT16_MAX,
3059 * device = dev->device
3060 * min_mtu = RTE_ETHER_MIN_MTU
3061 * max_mtu = UINT16_MAX
3063 * The following fields will be populated if support for dev_infos_get()
3064 * exists for the device and the rte_eth_dev 'dev' has been populated
3065 * successfully with a call to it:
3067 * driver_name = dev->device->driver->name
3068 * nb_rx_queues = dev->data->nb_rx_queues
3069 * nb_tx_queues = dev->data->nb_tx_queues
3070 * dev_flags = &dev->data->dev_flags
3073 * The port identifier of the Ethernet device.
3075 * A pointer to a structure of type *rte_eth_dev_info* to be filled with
3076 * the contextual information of the Ethernet device.
3078 * - (0) if successful.
3079 * - (-ENOTSUP) if support for dev_infos_get() does not exist for the device.
3080 * - (-ENODEV) if *port_id* invalid.
3081 * - (-EINVAL) if bad parameter.
3083 int rte_eth_dev_info_get(uint16_t port_id, struct rte_eth_dev_info *dev_info);
3086 * Retrieve the firmware version of a device.
3089 * The port identifier of the device.
3091 * A pointer to a string array storing the firmware version of a device,
3092 * the string includes terminating null. This pointer is allocated by caller.
3094 * The size of the string array pointed by fw_version, which should be
3095 * large enough to store firmware version of the device.
3097 * - (0) if successful.
3098 * - (-ENOTSUP) if operation is not supported.
3099 * - (-ENODEV) if *port_id* invalid.
3100 * - (-EIO) if device is removed.
3101 * - (-EINVAL) if bad parameter.
3102 * - (>0) if *fw_size* is not enough to store firmware version, return
3103 * the size of the non truncated string.
3105 int rte_eth_dev_fw_version_get(uint16_t port_id,
3106 char *fw_version, size_t fw_size);
3109 * Retrieve the supported packet types of an Ethernet device.
3111 * When a packet type is announced as supported, it *must* be recognized by
3112 * the PMD. For instance, if RTE_PTYPE_L2_ETHER, RTE_PTYPE_L2_ETHER_VLAN
3113 * and RTE_PTYPE_L3_IPV4 are announced, the PMD must return the following
3114 * packet types for these packets:
3115 * - Ether/IPv4 -> RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4
3116 * - Ether/Vlan/IPv4 -> RTE_PTYPE_L2_ETHER_VLAN | RTE_PTYPE_L3_IPV4
3117 * - Ether/[anything else] -> RTE_PTYPE_L2_ETHER
3118 * - Ether/Vlan/[anything else] -> RTE_PTYPE_L2_ETHER_VLAN
3120 * When a packet is received by a PMD, the most precise type must be
3121 * returned among the ones supported. However a PMD is allowed to set
3122 * packet type that is not in the supported list, at the condition that it
3123 * is more precise. Therefore, a PMD announcing no supported packet types
3124 * can still set a matching packet type in a received packet.
3127 * Better to invoke this API after the device is already started or rx burst
3128 * function is decided, to obtain correct supported ptypes.
3130 * if a given PMD does not report what ptypes it supports, then the supported
3131 * ptype count is reported as 0.
3133 * The port identifier of the Ethernet device.
3135 * A hint of what kind of packet type which the caller is interested in.
3137 * An array pointer to store adequate packet types, allocated by caller.
3139 * Size of the array pointed by param ptypes.
3141 * - (>=0) Number of supported ptypes. If the number of types exceeds num,
3142 * only num entries will be filled into the ptypes array, but the full
3143 * count of supported ptypes will be returned.
3144 * - (-ENODEV) if *port_id* invalid.
3145 * - (-EINVAL) if bad parameter.
3147 int rte_eth_dev_get_supported_ptypes(uint16_t port_id, uint32_t ptype_mask,
3148 uint32_t *ptypes, int num);
3150 * Inform Ethernet device about reduced range of packet types to handle.
3152 * Application can use this function to set only specific ptypes that it's
3153 * interested. This information can be used by the PMD to optimize Rx path.
3155 * The function accepts an array `set_ptypes` allocated by the caller to
3156 * store the packet types set by the driver, the last element of the array
3157 * is set to RTE_PTYPE_UNKNOWN. The size of the `set_ptype` array should be
3158 * `rte_eth_dev_get_supported_ptypes() + 1` else it might only be filled
3162 * The port identifier of the Ethernet device.
3164 * The ptype family that application is interested in should be bitwise OR of
3165 * RTE_PTYPE_*_MASK or 0.
3167 * An array pointer to store set packet types, allocated by caller. The
3168 * function marks the end of array with RTE_PTYPE_UNKNOWN.
3170 * Size of the array pointed by param ptypes.
3171 * Should be rte_eth_dev_get_supported_ptypes() + 1 to accommodate the
3175 * - (-ENODEV) if *port_id* invalid.
3176 * - (-EINVAL) if *ptype_mask* is invalid (or) set_ptypes is NULL and
3179 int rte_eth_dev_set_ptypes(uint16_t port_id, uint32_t ptype_mask,
3180 uint32_t *set_ptypes, unsigned int num);
3183 * Retrieve the MTU of an Ethernet device.
3186 * The port identifier of the Ethernet device.
3188 * A pointer to a uint16_t where the retrieved MTU is to be stored.
3190 * - (0) if successful.
3191 * - (-ENODEV) if *port_id* invalid.
3192 * - (-EINVAL) if bad parameter.
3194 int rte_eth_dev_get_mtu(uint16_t port_id, uint16_t *mtu);
3197 * Change the MTU of an Ethernet device.
3200 * The port identifier of the Ethernet device.
3202 * A uint16_t for the MTU to be applied.
3204 * - (0) if successful.
3205 * - (-ENOTSUP) if operation is not supported.
3206 * - (-ENODEV) if *port_id* invalid.
3207 * - (-EIO) if device is removed.
3208 * - (-EINVAL) if *mtu* invalid, validation of mtu can occur within
3209 * rte_eth_dev_set_mtu if dev_infos_get is supported by the device or
3210 * when the mtu is set using dev->dev_ops->mtu_set.
3211 * - (-EBUSY) if operation is not allowed when the port is running
3213 int rte_eth_dev_set_mtu(uint16_t port_id, uint16_t mtu);
3216 * Enable/Disable hardware filtering by an Ethernet device of received
3217 * VLAN packets tagged with a given VLAN Tag Identifier.
3220 * The port identifier of the Ethernet device.
3222 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
3224 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*.
3225 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*.
3227 * - (0) if successful.
3228 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
3229 * - (-ENODEV) if *port_id* invalid.
3230 * - (-EIO) if device is removed.
3231 * - (-ENOSYS) if VLAN filtering on *port_id* disabled.
3232 * - (-EINVAL) if *vlan_id* > 4095.
3234 int rte_eth_dev_vlan_filter(uint16_t port_id, uint16_t vlan_id, int on);
3237 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device.
3240 * The port identifier of the Ethernet device.
3241 * @param rx_queue_id
3242 * The index of the receive queue for which a queue stats mapping is required.
3243 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3244 * to rte_eth_dev_configure().
3246 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port.
3247 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port.
3249 * - (0) if successful.
3250 * - (-ENOTSUP) if hardware-assisted VLAN stripping not configured.
3251 * - (-ENODEV) if *port_id* invalid.
3252 * - (-EINVAL) if *rx_queue_id* invalid.
3254 int rte_eth_dev_set_vlan_strip_on_queue(uint16_t port_id, uint16_t rx_queue_id,
3258 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to
3262 * The port identifier of the Ethernet device.
3266 * The Tag Protocol ID
3268 * - (0) if successful.
3269 * - (-ENOTSUP) if hardware-assisted VLAN TPID setup is not supported.
3270 * - (-ENODEV) if *port_id* invalid.
3271 * - (-EIO) if device is removed.
3273 int rte_eth_dev_set_vlan_ether_type(uint16_t port_id,
3274 enum rte_vlan_type vlan_type,
3278 * Set VLAN offload configuration on an Ethernet device.
3281 * The port identifier of the Ethernet device.
3282 * @param offload_mask
3283 * The VLAN Offload bit mask can be mixed use with "OR"
3284 * ETH_VLAN_STRIP_OFFLOAD
3285 * ETH_VLAN_FILTER_OFFLOAD
3286 * ETH_VLAN_EXTEND_OFFLOAD
3287 * ETH_QINQ_STRIP_OFFLOAD
3289 * - (0) if successful.
3290 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
3291 * - (-ENODEV) if *port_id* invalid.
3292 * - (-EIO) if device is removed.
3294 int rte_eth_dev_set_vlan_offload(uint16_t port_id, int offload_mask);
3297 * Read VLAN Offload configuration from an Ethernet device
3300 * The port identifier of the Ethernet device.
3302 * - (>0) if successful. Bit mask to indicate
3303 * ETH_VLAN_STRIP_OFFLOAD
3304 * ETH_VLAN_FILTER_OFFLOAD
3305 * ETH_VLAN_EXTEND_OFFLOAD
3306 * ETH_QINQ_STRIP_OFFLOAD
3307 * - (-ENODEV) if *port_id* invalid.
3309 int rte_eth_dev_get_vlan_offload(uint16_t port_id);
3312 * Set port based TX VLAN insertion on or off.
3315 * The port identifier of the Ethernet device.
3317 * Port based TX VLAN identifier together with user priority.
3319 * Turn on or off the port based TX VLAN insertion.
3322 * - (0) if successful.
3323 * - negative if failed.
3325 int rte_eth_dev_set_vlan_pvid(uint16_t port_id, uint16_t pvid, int on);
3327 typedef void (*buffer_tx_error_fn)(struct rte_mbuf **unsent, uint16_t count,
3331 * Structure used to buffer packets for future TX
3332 * Used by APIs rte_eth_tx_buffer and rte_eth_tx_buffer_flush
3334 struct rte_eth_dev_tx_buffer {
3335 buffer_tx_error_fn error_callback;
3336 void *error_userdata;
3337 uint16_t size; /**< Size of buffer for buffered tx */
3338 uint16_t length; /**< Number of packets in the array */
3339 struct rte_mbuf *pkts[];
3340 /**< Pending packets to be sent on explicit flush or when full */
3344 * Calculate the size of the tx buffer.
3347 * Number of stored packets.
3349 #define RTE_ETH_TX_BUFFER_SIZE(sz) \
3350 (sizeof(struct rte_eth_dev_tx_buffer) + (sz) * sizeof(struct rte_mbuf *))
3353 * Initialize default values for buffered transmitting
3356 * Tx buffer to be initialized.
3363 rte_eth_tx_buffer_init(struct rte_eth_dev_tx_buffer *buffer, uint16_t size);
3366 * Configure a callback for buffered packets which cannot be sent
3368 * Register a specific callback to be called when an attempt is made to send
3369 * all packets buffered on an ethernet port, but not all packets can
3370 * successfully be sent. The callback registered here will be called only
3371 * from calls to rte_eth_tx_buffer() and rte_eth_tx_buffer_flush() APIs.
3372 * The default callback configured for each queue by default just frees the
3373 * packets back to the calling mempool. If additional behaviour is required,
3374 * for example, to count dropped packets, or to retry transmission of packets
3375 * which cannot be sent, this function should be used to register a suitable
3376 * callback function to implement the desired behaviour.
3377 * The example callback "rte_eth_count_unsent_packet_callback()" is also
3378 * provided as reference.
3381 * The port identifier of the Ethernet device.
3383 * The function to be used as the callback.
3385 * Arbitrary parameter to be passed to the callback function
3387 * 0 on success, or -EINVAL if bad parameter
3390 rte_eth_tx_buffer_set_err_callback(struct rte_eth_dev_tx_buffer *buffer,
3391 buffer_tx_error_fn callback, void *userdata);
3394 * Callback function for silently dropping unsent buffered packets.
3396 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
3397 * adjust the default behavior when buffered packets cannot be sent. This
3398 * function drops any unsent packets silently and is used by tx buffered
3399 * operations as default behavior.
3401 * NOTE: this function should not be called directly, instead it should be used
3402 * as a callback for packet buffering.
3404 * NOTE: when configuring this function as a callback with
3405 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
3406 * should point to an uint64_t value.
3409 * The previously buffered packets which could not be sent
3411 * The number of unsent packets in the pkts array
3416 rte_eth_tx_buffer_drop_callback(struct rte_mbuf **pkts, uint16_t unsent,
3420 * Callback function for tracking unsent buffered packets.
3422 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
3423 * adjust the default behavior when buffered packets cannot be sent. This
3424 * function drops any unsent packets, but also updates a user-supplied counter
3425 * to track the overall number of packets dropped. The counter should be an
3426 * uint64_t variable.
3428 * NOTE: this function should not be called directly, instead it should be used
3429 * as a callback for packet buffering.
3431 * NOTE: when configuring this function as a callback with
3432 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
3433 * should point to an uint64_t value.
3436 * The previously buffered packets which could not be sent
3438 * The number of unsent packets in the pkts array
3440 * Pointer to an uint64_t value, which will be incremented by unsent
3443 rte_eth_tx_buffer_count_callback(struct rte_mbuf **pkts, uint16_t unsent,
3447 * Request the driver to free mbufs currently cached by the driver. The
3448 * driver will only free the mbuf if it is no longer in use. It is the
3449 * application's responsibility to ensure rte_eth_tx_buffer_flush(..) is
3453 * The port identifier of the Ethernet device.
3455 * The index of the transmit queue through which output packets must be
3457 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
3458 * to rte_eth_dev_configure().
3460 * Maximum number of packets to free. Use 0 to indicate all possible packets
3461 * should be freed. Note that a packet may be using multiple mbufs.
3464 * -ENODEV: Invalid interface
3465 * -EIO: device is removed
3466 * -ENOTSUP: Driver does not support function
3468 * 0-n: Number of packets freed. More packets may still remain in ring that
3472 rte_eth_tx_done_cleanup(uint16_t port_id, uint16_t queue_id, uint32_t free_cnt);
3475 * Subtypes for IPsec offload event(@ref RTE_ETH_EVENT_IPSEC) raised by
3478 enum rte_eth_event_ipsec_subtype {
3479 RTE_ETH_EVENT_IPSEC_UNKNOWN = 0,
3480 /**< Unknown event type */
3481 RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW,
3482 /**< Sequence number overflow */
3483 RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY,
3484 /**< Soft time expiry of SA */
3485 RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY,
3486 /**< Soft byte expiry of SA */
3487 RTE_ETH_EVENT_IPSEC_MAX
3488 /**< Max value of this enum */
3492 * Descriptor for @ref RTE_ETH_EVENT_IPSEC event. Used by eth dev to send extra
3493 * information of the IPsec offload event.
3495 struct rte_eth_event_ipsec_desc {
3496 enum rte_eth_event_ipsec_subtype subtype;
3497 /**< Type of RTE_ETH_EVENT_IPSEC_* event */
3499 /**< Event specific metadata
3501 * For the following events, *userdata* registered
3502 * with the *rte_security_session* would be returned
3505 * - @ref RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW
3506 * - @ref RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY
3507 * - @ref RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY
3509 * @see struct rte_security_session_conf
3515 * The eth device event type for interrupt, and maybe others in the future.
3517 enum rte_eth_event_type {
3518 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */
3519 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */
3520 RTE_ETH_EVENT_QUEUE_STATE,
3521 /**< queue state event (enabled/disabled) */
3522 RTE_ETH_EVENT_INTR_RESET,
3523 /**< reset interrupt event, sent to VF on PF reset */
3524 RTE_ETH_EVENT_VF_MBOX, /**< message from the VF received by PF */
3525 RTE_ETH_EVENT_MACSEC, /**< MACsec offload related event */
3526 RTE_ETH_EVENT_INTR_RMV, /**< device removal event */
3527 RTE_ETH_EVENT_NEW, /**< port is probed */
3528 RTE_ETH_EVENT_DESTROY, /**< port is released */
3529 RTE_ETH_EVENT_IPSEC, /**< IPsec offload related event */
3530 RTE_ETH_EVENT_FLOW_AGED,/**< New aged-out flows is detected */
3531 RTE_ETH_EVENT_MAX /**< max value of this enum */
3534 typedef int (*rte_eth_dev_cb_fn)(uint16_t port_id,
3535 enum rte_eth_event_type event, void *cb_arg, void *ret_param);
3536 /**< user application callback to be registered for interrupts */
3539 * Register a callback function for port event.
3543 * RTE_ETH_ALL means register the event for all port ids.
3547 * User supplied callback function to be called.
3549 * Pointer to the parameters for the registered callback.
3552 * - On success, zero.
3553 * - On failure, a negative value.
3555 int rte_eth_dev_callback_register(uint16_t port_id,
3556 enum rte_eth_event_type event,
3557 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3560 * Unregister a callback function for port event.
3564 * RTE_ETH_ALL means unregister the event for all port ids.
3568 * User supplied callback function to be called.
3570 * Pointer to the parameters for the registered callback. -1 means to
3571 * remove all for the same callback address and same event.
3574 * - On success, zero.
3575 * - On failure, a negative value.
3577 int rte_eth_dev_callback_unregister(uint16_t port_id,
3578 enum rte_eth_event_type event,
3579 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3582 * When there is no rx packet coming in Rx Queue for a long time, we can
3583 * sleep lcore related to RX Queue for power saving, and enable rx interrupt
3584 * to be triggered when Rx packet arrives.
3586 * The rte_eth_dev_rx_intr_enable() function enables rx queue
3587 * interrupt on specific rx queue of a port.
3590 * The port identifier of the Ethernet device.
3592 * The index of the receive queue from which to retrieve input packets.
3593 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3594 * to rte_eth_dev_configure().
3596 * - (0) if successful.
3597 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3599 * - (-ENODEV) if *port_id* invalid.
3600 * - (-EIO) if device is removed.
3602 int rte_eth_dev_rx_intr_enable(uint16_t port_id, uint16_t queue_id);
3605 * When lcore wakes up from rx interrupt indicating packet coming, disable rx
3606 * interrupt and returns to polling mode.
3608 * The rte_eth_dev_rx_intr_disable() function disables rx queue
3609 * interrupt on specific rx queue of a port.
3612 * The port identifier of the Ethernet device.
3614 * The index of the receive queue from which to retrieve input packets.
3615 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3616 * to rte_eth_dev_configure().
3618 * - (0) if successful.
3619 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3621 * - (-ENODEV) if *port_id* invalid.
3622 * - (-EIO) if device is removed.
3624 int rte_eth_dev_rx_intr_disable(uint16_t port_id, uint16_t queue_id);
3627 * RX Interrupt control per port.
3630 * The port identifier of the Ethernet device.
3632 * Epoll instance fd which the intr vector associated to.
3633 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3635 * The operation be performed for the vector.
3636 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3640 * - On success, zero.
3641 * - On failure, a negative value.
3643 int rte_eth_dev_rx_intr_ctl(uint16_t port_id, int epfd, int op, void *data);
3646 * RX Interrupt control per queue.
3649 * The port identifier of the Ethernet device.
3651 * The index of the receive queue from which to retrieve input packets.
3652 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3653 * to rte_eth_dev_configure().
3655 * Epoll instance fd which the intr vector associated to.
3656 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3658 * The operation be performed for the vector.
3659 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3663 * - On success, zero.
3664 * - On failure, a negative value.
3666 int rte_eth_dev_rx_intr_ctl_q(uint16_t port_id, uint16_t queue_id,
3667 int epfd, int op, void *data);
3670 * Get interrupt fd per Rx queue.
3673 * The port identifier of the Ethernet device.
3675 * The index of the receive queue from which to retrieve input packets.
3676 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3677 * to rte_eth_dev_configure().
3679 * - (>=0) the interrupt fd associated to the requested Rx queue if
3684 rte_eth_dev_rx_intr_ctl_q_get_fd(uint16_t port_id, uint16_t queue_id);
3687 * Turn on the LED on the Ethernet device.
3688 * This function turns on the LED on the Ethernet device.
3691 * The port identifier of the Ethernet device.
3693 * - (0) if successful.
3694 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3696 * - (-ENODEV) if *port_id* invalid.
3697 * - (-EIO) if device is removed.
3699 int rte_eth_led_on(uint16_t port_id);
3702 * Turn off the LED on the Ethernet device.
3703 * This function turns off the LED on the Ethernet device.
3706 * The port identifier of the Ethernet device.
3708 * - (0) if successful.
3709 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3711 * - (-ENODEV) if *port_id* invalid.
3712 * - (-EIO) if device is removed.
3714 int rte_eth_led_off(uint16_t port_id);
3718 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
3720 * Get Forward Error Correction(FEC) capability.
3723 * The port identifier of the Ethernet device.
3724 * @param speed_fec_capa
3725 * speed_fec_capa is out only with per-speed capabilities.
3726 * If set to NULL, the function returns the required number
3727 * of required array entries.
3729 * a number of elements in an speed_fec_capa array.
3732 * - A non-negative value lower or equal to num: success. The return value
3733 * is the number of entries filled in the fec capa array.
3734 * - A non-negative value higher than num: error, the given fec capa array
3735 * is too small. The return value corresponds to the num that should
3736 * be given to succeed. The entries in fec capa array are not valid and
3737 * shall not be used by the caller.
3738 * - (-ENOTSUP) if underlying hardware OR driver doesn't support.
3740 * - (-EIO) if device is removed.
3741 * - (-ENODEV) if *port_id* invalid.
3742 * - (-EINVAL) if *num* or *speed_fec_capa* invalid
3745 int rte_eth_fec_get_capability(uint16_t port_id,
3746 struct rte_eth_fec_capa *speed_fec_capa,
3751 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
3753 * Get current Forward Error Correction(FEC) mode.
3754 * If link is down and AUTO is enabled, AUTO is returned, otherwise,
3755 * configured FEC mode is returned.
3756 * If link is up, current FEC mode is returned.
3759 * The port identifier of the Ethernet device.
3761 * A bitmask of enabled FEC modes. If AUTO bit is set, other
3762 * bits specify FEC modes which may be negotiated. If AUTO
3763 * bit is clear, specify FEC modes to be used (only one valid
3764 * mode per speed may be set).
3766 * - (0) if successful.
3767 * - (-ENOTSUP) if underlying hardware OR driver doesn't support.
3769 * - (-EIO) if device is removed.
3770 * - (-ENODEV) if *port_id* invalid.
3773 int rte_eth_fec_get(uint16_t port_id, uint32_t *fec_capa);
3777 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
3779 * Set Forward Error Correction(FEC) mode.
3782 * The port identifier of the Ethernet device.
3784 * A bitmask of allowed FEC modes. If AUTO bit is set, other
3785 * bits specify FEC modes which may be negotiated. If AUTO
3786 * bit is clear, specify FEC modes to be used (only one valid
3787 * mode per speed may be set).
3789 * - (0) if successful.
3790 * - (-EINVAL) if the FEC mode is not valid.
3791 * - (-ENOTSUP) if underlying hardware OR driver doesn't support.
3792 * - (-EIO) if device is removed.
3793 * - (-ENODEV) if *port_id* invalid.
3796 int rte_eth_fec_set(uint16_t port_id, uint32_t fec_capa);
3799 * Get current status of the Ethernet link flow control for Ethernet device
3802 * The port identifier of the Ethernet device.
3804 * The pointer to the structure where to store the flow control parameters.
3806 * - (0) if successful.
3807 * - (-ENOTSUP) if hardware doesn't support flow control.
3808 * - (-ENODEV) if *port_id* invalid.
3809 * - (-EIO) if device is removed.
3810 * - (-EINVAL) if bad parameter.
3812 int rte_eth_dev_flow_ctrl_get(uint16_t port_id,
3813 struct rte_eth_fc_conf *fc_conf);
3816 * Configure the Ethernet link flow control for Ethernet device
3819 * The port identifier of the Ethernet device.
3821 * The pointer to the structure of the flow control parameters.
3823 * - (0) if successful.
3824 * - (-ENOTSUP) if hardware doesn't support flow control mode.
3825 * - (-ENODEV) if *port_id* invalid.
3826 * - (-EINVAL) if bad parameter
3827 * - (-EIO) if flow control setup failure or device is removed.
3829 int rte_eth_dev_flow_ctrl_set(uint16_t port_id,
3830 struct rte_eth_fc_conf *fc_conf);
3833 * Configure the Ethernet priority flow control under DCB environment
3834 * for Ethernet device.
3837 * The port identifier of the Ethernet device.
3839 * The pointer to the structure of the priority flow control parameters.
3841 * - (0) if successful.
3842 * - (-ENOTSUP) if hardware doesn't support priority flow control mode.
3843 * - (-ENODEV) if *port_id* invalid.
3844 * - (-EINVAL) if bad parameter
3845 * - (-EIO) if flow control setup failure or device is removed.
3847 int rte_eth_dev_priority_flow_ctrl_set(uint16_t port_id,
3848 struct rte_eth_pfc_conf *pfc_conf);
3851 * Add a MAC address to the set used for filtering incoming packets.
3854 * The port identifier of the Ethernet device.
3856 * The MAC address to add.
3858 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is
3859 * not enabled, this should be set to 0.
3861 * - (0) if successfully added or *mac_addr* was already added.
3862 * - (-ENOTSUP) if hardware doesn't support this feature.
3863 * - (-ENODEV) if *port* is invalid.
3864 * - (-EIO) if device is removed.
3865 * - (-ENOSPC) if no more MAC addresses can be added.
3866 * - (-EINVAL) if MAC address is invalid.
3868 int rte_eth_dev_mac_addr_add(uint16_t port_id, struct rte_ether_addr *mac_addr,
3872 * Remove a MAC address from the internal array of addresses.
3875 * The port identifier of the Ethernet device.
3877 * MAC address to remove.
3879 * - (0) if successful, or *mac_addr* didn't exist.
3880 * - (-ENOTSUP) if hardware doesn't support.
3881 * - (-ENODEV) if *port* invalid.
3882 * - (-EADDRINUSE) if attempting to remove the default MAC address.
3883 * - (-EINVAL) if MAC address is invalid.
3885 int rte_eth_dev_mac_addr_remove(uint16_t port_id,
3886 struct rte_ether_addr *mac_addr);
3889 * Set the default MAC address.
3892 * The port identifier of the Ethernet device.
3894 * New default MAC address.
3896 * - (0) if successful, or *mac_addr* didn't exist.
3897 * - (-ENOTSUP) if hardware doesn't support.
3898 * - (-ENODEV) if *port* invalid.
3899 * - (-EINVAL) if MAC address is invalid.
3901 int rte_eth_dev_default_mac_addr_set(uint16_t port_id,
3902 struct rte_ether_addr *mac_addr);
3905 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3908 * The port identifier of the Ethernet device.
3912 * Redirection table size. The table size can be queried by
3913 * rte_eth_dev_info_get().
3915 * - (0) if successful.
3916 * - (-ENODEV) if *port_id* is invalid.
3917 * - (-ENOTSUP) if hardware doesn't support.
3918 * - (-EINVAL) if bad parameter.
3919 * - (-EIO) if device is removed.
3921 int rte_eth_dev_rss_reta_update(uint16_t port_id,
3922 struct rte_eth_rss_reta_entry64 *reta_conf,
3923 uint16_t reta_size);
3926 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3929 * The port identifier of the Ethernet device.
3931 * RETA to query. For each requested reta entry, corresponding bit
3932 * in mask must be set.
3934 * Redirection table size. The table size can be queried by
3935 * rte_eth_dev_info_get().
3937 * - (0) if successful.
3938 * - (-ENODEV) if *port_id* is invalid.
3939 * - (-ENOTSUP) if hardware doesn't support.
3940 * - (-EINVAL) if bad parameter.
3941 * - (-EIO) if device is removed.
3943 int rte_eth_dev_rss_reta_query(uint16_t port_id,
3944 struct rte_eth_rss_reta_entry64 *reta_conf,
3945 uint16_t reta_size);
3948 * Updates unicast hash table for receiving packet with the given destination
3949 * MAC address, and the packet is routed to all VFs for which the RX mode is
3950 * accept packets that match the unicast hash table.
3953 * The port identifier of the Ethernet device.
3955 * Unicast MAC address.
3957 * 1 - Set an unicast hash bit for receiving packets with the MAC address.
3958 * 0 - Clear an unicast hash bit.
3960 * - (0) if successful.
3961 * - (-ENOTSUP) if hardware doesn't support.
3962 * - (-ENODEV) if *port_id* invalid.
3963 * - (-EIO) if device is removed.
3964 * - (-EINVAL) if bad parameter.
3966 int rte_eth_dev_uc_hash_table_set(uint16_t port_id, struct rte_ether_addr *addr,
3970 * Updates all unicast hash bitmaps for receiving packet with any Unicast
3971 * Ethernet MAC addresses,the packet is routed to all VFs for which the RX
3972 * mode is accept packets that match the unicast hash table.
3975 * The port identifier of the Ethernet device.
3977 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet
3979 * 0 - Clear all unicast hash bitmaps
3981 * - (0) if successful.
3982 * - (-ENOTSUP) if hardware doesn't support.
3983 * - (-ENODEV) if *port_id* invalid.
3984 * - (-EIO) if device is removed.
3985 * - (-EINVAL) if bad parameter.
3987 int rte_eth_dev_uc_all_hash_table_set(uint16_t port_id, uint8_t on);
3990 * Set a traffic mirroring rule on an Ethernet device
3993 * The port identifier of the Ethernet device.
3994 * @param mirror_conf
3995 * The pointer to the traffic mirroring structure describing the mirroring rule.
3996 * The *rte_eth_vm_mirror_conf* structure includes the type of mirroring rule,
3997 * destination pool and the value of rule if enable vlan or pool mirroring.
4000 * The index of traffic mirroring rule, we support four separated rules.
4002 * 1 - Enable a mirroring rule.
4003 * 0 - Disable a mirroring rule.
4005 * - (0) if successful.
4006 * - (-ENOTSUP) if hardware doesn't support this feature.
4007 * - (-ENODEV) if *port_id* invalid.
4008 * - (-EIO) if device is removed.
4009 * - (-EINVAL) if the mr_conf information is not correct.
4011 int rte_eth_mirror_rule_set(uint16_t port_id,
4012 struct rte_eth_mirror_conf *mirror_conf,
4017 * Reset a traffic mirroring rule on an Ethernet device.
4020 * The port identifier of the Ethernet device.
4022 * The index of traffic mirroring rule, we support four separated rules.
4024 * - (0) if successful.
4025 * - (-ENOTSUP) if hardware doesn't support this feature.
4026 * - (-ENODEV) if *port_id* invalid.
4027 * - (-EIO) if device is removed.
4028 * - (-EINVAL) if bad parameter.
4030 int rte_eth_mirror_rule_reset(uint16_t port_id,
4034 * Set the rate limitation for a queue on an Ethernet device.
4037 * The port identifier of the Ethernet device.
4041 * The tx rate in Mbps. Allocated from the total port link speed.
4043 * - (0) if successful.
4044 * - (-ENOTSUP) if hardware doesn't support this feature.
4045 * - (-ENODEV) if *port_id* invalid.
4046 * - (-EIO) if device is removed.
4047 * - (-EINVAL) if bad parameter.
4049 int rte_eth_set_queue_rate_limit(uint16_t port_id, uint16_t queue_idx,
4053 * Configuration of Receive Side Scaling hash computation of Ethernet device.
4056 * The port identifier of the Ethernet device.
4058 * The new configuration to use for RSS hash computation on the port.
4060 * - (0) if successful.
4061 * - (-ENODEV) if port identifier is invalid.
4062 * - (-EIO) if device is removed.
4063 * - (-ENOTSUP) if hardware doesn't support.
4064 * - (-EINVAL) if bad parameter.
4066 int rte_eth_dev_rss_hash_update(uint16_t port_id,
4067 struct rte_eth_rss_conf *rss_conf);
4070 * Retrieve current configuration of Receive Side Scaling hash computation
4071 * of Ethernet device.
4074 * The port identifier of the Ethernet device.
4076 * Where to store the current RSS hash configuration of the Ethernet device.
4078 * - (0) if successful.
4079 * - (-ENODEV) if port identifier is invalid.
4080 * - (-EIO) if device is removed.
4081 * - (-ENOTSUP) if hardware doesn't support RSS.
4082 * - (-EINVAL) if bad parameter.
4085 rte_eth_dev_rss_hash_conf_get(uint16_t port_id,
4086 struct rte_eth_rss_conf *rss_conf);
4089 * Add UDP tunneling port for a type of tunnel.
4091 * Some NICs may require such configuration to properly parse a tunnel
4092 * with any standard or custom UDP port.
4093 * The packets with this UDP port will be parsed for this type of tunnel.
4094 * The device parser will also check the rest of the tunnel headers
4095 * before classifying the packet.
4097 * With some devices, this API will affect packet classification, i.e.:
4098 * - mbuf.packet_type reported on Rx
4099 * - rte_flow rules with tunnel items
4102 * The port identifier of the Ethernet device.
4104 * UDP tunneling configuration.
4107 * - (0) if successful.
4108 * - (-ENODEV) if port identifier is invalid.
4109 * - (-EIO) if device is removed.
4110 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4113 rte_eth_dev_udp_tunnel_port_add(uint16_t port_id,
4114 struct rte_eth_udp_tunnel *tunnel_udp);
4117 * Delete UDP tunneling port for a type of tunnel.
4119 * The packets with this UDP port will not be classified as this type of tunnel
4120 * anymore if the device use such mapping for tunnel packet classification.
4122 * @see rte_eth_dev_udp_tunnel_port_add
4125 * The port identifier of the Ethernet device.
4127 * UDP tunneling configuration.
4130 * - (0) if successful.
4131 * - (-ENODEV) if port identifier is invalid.
4132 * - (-EIO) if device is removed.
4133 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4136 rte_eth_dev_udp_tunnel_port_delete(uint16_t port_id,
4137 struct rte_eth_udp_tunnel *tunnel_udp);
4140 * Get DCB information on an Ethernet device.
4143 * The port identifier of the Ethernet device.
4147 * - (0) if successful.
4148 * - (-ENODEV) if port identifier is invalid.
4149 * - (-EIO) if device is removed.
4150 * - (-ENOTSUP) if hardware doesn't support.
4151 * - (-EINVAL) if bad parameter.
4153 int rte_eth_dev_get_dcb_info(uint16_t port_id,
4154 struct rte_eth_dcb_info *dcb_info);
4156 struct rte_eth_rxtx_callback;
4159 * Add a callback to be called on packet RX on a given port and queue.
4161 * This API configures a function to be called for each burst of
4162 * packets received on a given NIC port queue. The return value is a pointer
4163 * that can be used to later remove the callback using
4164 * rte_eth_remove_rx_callback().
4166 * Multiple functions are called in the order that they are added.
4169 * The port identifier of the Ethernet device.
4171 * The queue on the Ethernet device on which the callback is to be added.
4173 * The callback function
4175 * A generic pointer parameter which will be passed to each invocation of the
4176 * callback function on this port and queue. Inter-thread synchronization
4177 * of any user data changes is the responsibility of the user.
4181 * On success, a pointer value which can later be used to remove the callback.
4183 const struct rte_eth_rxtx_callback *
4184 rte_eth_add_rx_callback(uint16_t port_id, uint16_t queue_id,
4185 rte_rx_callback_fn fn, void *user_param);
4188 * Add a callback that must be called first on packet RX on a given port
4191 * This API configures a first function to be called for each burst of
4192 * packets received on a given NIC port queue. The return value is a pointer
4193 * that can be used to later remove the callback using
4194 * rte_eth_remove_rx_callback().
4196 * Multiple functions are called in the order that they are added.
4199 * The port identifier of the Ethernet device.
4201 * The queue on the Ethernet device on which the callback is to be added.
4203 * The callback function
4205 * A generic pointer parameter which will be passed to each invocation of the
4206 * callback function on this port and queue. Inter-thread synchronization
4207 * of any user data changes is the responsibility of the user.
4211 * On success, a pointer value which can later be used to remove the callback.
4213 const struct rte_eth_rxtx_callback *
4214 rte_eth_add_first_rx_callback(uint16_t port_id, uint16_t queue_id,
4215 rte_rx_callback_fn fn, void *user_param);
4218 * Add a callback to be called on packet TX on a given port and queue.
4220 * This API configures a function to be called for each burst of
4221 * packets sent on a given NIC port queue. The return value is a pointer
4222 * that can be used to later remove the callback using
4223 * rte_eth_remove_tx_callback().
4225 * Multiple functions are called in the order that they are added.
4228 * The port identifier of the Ethernet device.
4230 * The queue on the Ethernet device on which the callback is to be added.
4232 * The callback function
4234 * A generic pointer parameter which will be passed to each invocation of the
4235 * callback function on this port and queue. Inter-thread synchronization
4236 * of any user data changes is the responsibility of the user.
4240 * On success, a pointer value which can later be used to remove the callback.
4242 const struct rte_eth_rxtx_callback *
4243 rte_eth_add_tx_callback(uint16_t port_id, uint16_t queue_id,
4244 rte_tx_callback_fn fn, void *user_param);
4247 * Remove an RX packet callback from a given port and queue.
4249 * This function is used to removed callbacks that were added to a NIC port
4250 * queue using rte_eth_add_rx_callback().
4252 * Note: the callback is removed from the callback list but it isn't freed
4253 * since the it may still be in use. The memory for the callback can be
4254 * subsequently freed back by the application by calling rte_free():
4256 * - Immediately - if the port is stopped, or the user knows that no
4257 * callbacks are in flight e.g. if called from the thread doing RX/TX
4260 * - After a short delay - where the delay is sufficient to allow any
4261 * in-flight callbacks to complete. Alternately, the RCU mechanism can be
4262 * used to detect when data plane threads have ceased referencing the
4266 * The port identifier of the Ethernet device.
4268 * The queue on the Ethernet device from which the callback is to be removed.
4270 * User supplied callback created via rte_eth_add_rx_callback().
4273 * - 0: Success. Callback was removed.
4274 * - -ENODEV: If *port_id* is invalid.
4275 * - -ENOTSUP: Callback support is not available.
4276 * - -EINVAL: The queue_id is out of range, or the callback
4277 * is NULL or not found for the port/queue.
4279 int rte_eth_remove_rx_callback(uint16_t port_id, uint16_t queue_id,
4280 const struct rte_eth_rxtx_callback *user_cb);
4283 * Remove a TX packet callback from a given port and queue.
4285 * This function is used to removed callbacks that were added to a NIC port
4286 * queue using rte_eth_add_tx_callback().
4288 * Note: the callback is removed from the callback list but it isn't freed
4289 * since the it may still be in use. The memory for the callback can be
4290 * subsequently freed back by the application by calling rte_free():
4292 * - Immediately - if the port is stopped, or the user knows that no
4293 * callbacks are in flight e.g. if called from the thread doing RX/TX
4296 * - After a short delay - where the delay is sufficient to allow any
4297 * in-flight callbacks to complete. Alternately, the RCU mechanism can be
4298 * used to detect when data plane threads have ceased referencing the
4302 * The port identifier of the Ethernet device.
4304 * The queue on the Ethernet device from which the callback is to be removed.
4306 * User supplied callback created via rte_eth_add_tx_callback().
4309 * - 0: Success. Callback was removed.
4310 * - -ENODEV: If *port_id* is invalid.
4311 * - -ENOTSUP: Callback support is not available.
4312 * - -EINVAL: The queue_id is out of range, or the callback
4313 * is NULL or not found for the port/queue.
4315 int rte_eth_remove_tx_callback(uint16_t port_id, uint16_t queue_id,
4316 const struct rte_eth_rxtx_callback *user_cb);
4319 * Retrieve information about given port's RX queue.
4322 * The port identifier of the Ethernet device.
4324 * The RX queue on the Ethernet device for which information
4325 * will be retrieved.
4327 * A pointer to a structure of type *rte_eth_rxq_info_info* to be filled with
4328 * the information of the Ethernet device.
4332 * - -ENODEV: If *port_id* is invalid.
4333 * - -ENOTSUP: routine is not supported by the device PMD.
4334 * - -EINVAL: The queue_id is out of range, or the queue
4337 int rte_eth_rx_queue_info_get(uint16_t port_id, uint16_t queue_id,
4338 struct rte_eth_rxq_info *qinfo);
4341 * Retrieve information about given port's TX queue.
4344 * The port identifier of the Ethernet device.
4346 * The TX queue on the Ethernet device for which information
4347 * will be retrieved.
4349 * A pointer to a structure of type *rte_eth_txq_info_info* to be filled with
4350 * the information of the Ethernet device.
4354 * - -ENODEV: If *port_id* is invalid.
4355 * - -ENOTSUP: routine is not supported by the device PMD.
4356 * - -EINVAL: The queue_id is out of range, or the queue
4359 int rte_eth_tx_queue_info_get(uint16_t port_id, uint16_t queue_id,
4360 struct rte_eth_txq_info *qinfo);
4363 * Retrieve information about the Rx packet burst mode.
4366 * The port identifier of the Ethernet device.
4368 * The Rx queue on the Ethernet device for which information
4369 * will be retrieved.
4371 * A pointer to a structure of type *rte_eth_burst_mode* to be filled
4372 * with the information of the packet burst mode.
4376 * - -ENODEV: If *port_id* is invalid.
4377 * - -ENOTSUP: routine is not supported by the device PMD.
4378 * - -EINVAL: The queue_id is out of range.
4380 int rte_eth_rx_burst_mode_get(uint16_t port_id, uint16_t queue_id,
4381 struct rte_eth_burst_mode *mode);
4384 * Retrieve information about the Tx packet burst mode.
4387 * The port identifier of the Ethernet device.
4389 * The Tx queue on the Ethernet device for which information
4390 * will be retrieved.
4392 * A pointer to a structure of type *rte_eth_burst_mode* to be filled
4393 * with the information of the packet burst mode.
4397 * - -ENODEV: If *port_id* is invalid.
4398 * - -ENOTSUP: routine is not supported by the device PMD.
4399 * - -EINVAL: The queue_id is out of range.
4401 int rte_eth_tx_burst_mode_get(uint16_t port_id, uint16_t queue_id,
4402 struct rte_eth_burst_mode *mode);
4406 * @b EXPERIMENTAL: this API may change without prior notice.
4408 * Retrieve the monitor condition for a given receive queue.
4411 * The port identifier of the Ethernet device.
4413 * The Rx queue on the Ethernet device for which information
4414 * will be retrieved.
4416 * The pointer to power-optimized monitoring condition structure.
4420 * -ENOTSUP: Operation not supported.
4421 * -EINVAL: Invalid parameters.
4422 * -ENODEV: Invalid port ID.
4425 int rte_eth_get_monitor_addr(uint16_t port_id, uint16_t queue_id,
4426 struct rte_power_monitor_cond *pmc);
4429 * Retrieve device registers and register attributes (number of registers and
4433 * The port identifier of the Ethernet device.
4435 * Pointer to rte_dev_reg_info structure to fill in. If info->data is
4436 * NULL the function fills in the width and length fields. If non-NULL
4437 * the registers are put into the buffer pointed at by the data field.
4439 * - (0) if successful.
4440 * - (-ENOTSUP) if hardware doesn't support.
4441 * - (-EINVAL) if bad parameter.
4442 * - (-ENODEV) if *port_id* invalid.
4443 * - (-EIO) if device is removed.
4444 * - others depends on the specific operations implementation.
4446 int rte_eth_dev_get_reg_info(uint16_t port_id, struct rte_dev_reg_info *info);
4449 * Retrieve size of device EEPROM
4452 * The port identifier of the Ethernet device.
4454 * - (>=0) EEPROM size if successful.
4455 * - (-ENOTSUP) if hardware doesn't support.
4456 * - (-ENODEV) if *port_id* invalid.
4457 * - (-EIO) if device is removed.
4458 * - others depends on the specific operations implementation.
4460 int rte_eth_dev_get_eeprom_length(uint16_t port_id);
4463 * Retrieve EEPROM and EEPROM attribute
4466 * The port identifier of the Ethernet device.
4468 * The template includes buffer for return EEPROM data and
4469 * EEPROM attributes to be filled.
4471 * - (0) if successful.
4472 * - (-ENOTSUP) if hardware doesn't support.
4473 * - (-EINVAL) if bad parameter.
4474 * - (-ENODEV) if *port_id* invalid.
4475 * - (-EIO) if device is removed.
4476 * - others depends on the specific operations implementation.
4478 int rte_eth_dev_get_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
4481 * Program EEPROM with provided data
4484 * The port identifier of the Ethernet device.
4486 * The template includes EEPROM data for programming and
4487 * EEPROM attributes to be filled
4489 * - (0) if successful.
4490 * - (-ENOTSUP) if hardware doesn't support.
4491 * - (-ENODEV) if *port_id* invalid.
4492 * - (-EINVAL) if bad parameter.
4493 * - (-EIO) if device is removed.
4494 * - others depends on the specific operations implementation.
4496 int rte_eth_dev_set_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
4500 * @b EXPERIMENTAL: this API may change without prior notice.
4502 * Retrieve the type and size of plugin module EEPROM
4505 * The port identifier of the Ethernet device.
4507 * The type and size of plugin module EEPROM.
4509 * - (0) if successful.
4510 * - (-ENOTSUP) if hardware doesn't support.
4511 * - (-ENODEV) if *port_id* invalid.
4512 * - (-EINVAL) if bad parameter.
4513 * - (-EIO) if device is removed.
4514 * - others depends on the specific operations implementation.
4518 rte_eth_dev_get_module_info(uint16_t port_id,
4519 struct rte_eth_dev_module_info *modinfo);
4523 * @b EXPERIMENTAL: this API may change without prior notice.
4525 * Retrieve the data of plugin module EEPROM
4528 * The port identifier of the Ethernet device.
4530 * The template includes the plugin module EEPROM attributes, and the
4531 * buffer for return plugin module EEPROM data.
4533 * - (0) if successful.
4534 * - (-ENOTSUP) if hardware doesn't support.
4535 * - (-EINVAL) if bad parameter.
4536 * - (-ENODEV) if *port_id* invalid.
4537 * - (-EIO) if device is removed.
4538 * - others depends on the specific operations implementation.
4542 rte_eth_dev_get_module_eeprom(uint16_t port_id,
4543 struct rte_dev_eeprom_info *info);
4546 * Set the list of multicast addresses to filter on an Ethernet device.
4549 * The port identifier of the Ethernet device.
4550 * @param mc_addr_set
4551 * The array of multicast addresses to set. Equal to NULL when the function
4552 * is invoked to flush the set of filtered addresses.
4554 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0
4555 * when the function is invoked to flush the set of filtered addresses.
4557 * - (0) if successful.
4558 * - (-ENODEV) if *port_id* invalid.
4559 * - (-EIO) if device is removed.
4560 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering.
4561 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources.
4562 * - (-EINVAL) if bad parameter.
4564 int rte_eth_dev_set_mc_addr_list(uint16_t port_id,
4565 struct rte_ether_addr *mc_addr_set,
4566 uint32_t nb_mc_addr);
4569 * Enable IEEE1588/802.1AS timestamping for an Ethernet device.
4572 * The port identifier of the Ethernet device.
4576 * - -ENODEV: The port ID is invalid.
4577 * - -EIO: if device is removed.
4578 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4580 int rte_eth_timesync_enable(uint16_t port_id);
4583 * Disable IEEE1588/802.1AS timestamping for an Ethernet device.
4586 * The port identifier of the Ethernet device.
4590 * - -ENODEV: The port ID is invalid.
4591 * - -EIO: if device is removed.
4592 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4594 int rte_eth_timesync_disable(uint16_t port_id);
4597 * Read an IEEE1588/802.1AS RX timestamp from an Ethernet device.
4600 * The port identifier of the Ethernet device.
4602 * Pointer to the timestamp struct.
4604 * Device specific flags. Used to pass the RX timesync register index to
4605 * i40e. Unused in igb/ixgbe, pass 0 instead.
4609 * - -EINVAL: No timestamp is available.
4610 * - -ENODEV: The port ID is invalid.
4611 * - -EIO: if device is removed.
4612 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4614 int rte_eth_timesync_read_rx_timestamp(uint16_t port_id,
4615 struct timespec *timestamp, uint32_t flags);
4618 * Read an IEEE1588/802.1AS TX timestamp from an Ethernet device.
4621 * The port identifier of the Ethernet device.
4623 * Pointer to the timestamp struct.
4627 * - -EINVAL: No timestamp is available.
4628 * - -ENODEV: The port ID is invalid.
4629 * - -EIO: if device is removed.
4630 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4632 int rte_eth_timesync_read_tx_timestamp(uint16_t port_id,
4633 struct timespec *timestamp);
4636 * Adjust the timesync clock on an Ethernet device.
4638 * This is usually used in conjunction with other Ethdev timesync functions to
4639 * synchronize the device time using the IEEE1588/802.1AS protocol.
4642 * The port identifier of the Ethernet device.
4644 * The adjustment in nanoseconds.
4648 * - -ENODEV: The port ID is invalid.
4649 * - -EIO: if device is removed.
4650 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4652 int rte_eth_timesync_adjust_time(uint16_t port_id, int64_t delta);
4655 * Read the time from the timesync clock on an Ethernet device.
4657 * This is usually used in conjunction with other Ethdev timesync functions to
4658 * synchronize the device time using the IEEE1588/802.1AS protocol.
4661 * The port identifier of the Ethernet device.
4663 * Pointer to the timespec struct that holds the time.
4667 * - -EINVAL: Bad parameter.
4669 int rte_eth_timesync_read_time(uint16_t port_id, struct timespec *time);
4672 * Set the time of the timesync clock on an Ethernet device.
4674 * This is usually used in conjunction with other Ethdev timesync functions to
4675 * synchronize the device time using the IEEE1588/802.1AS protocol.
4678 * The port identifier of the Ethernet device.
4680 * Pointer to the timespec struct that holds the time.
4684 * - -EINVAL: No timestamp is available.
4685 * - -ENODEV: The port ID is invalid.
4686 * - -EIO: if device is removed.
4687 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4689 int rte_eth_timesync_write_time(uint16_t port_id, const struct timespec *time);
4693 * @b EXPERIMENTAL: this API may change without prior notice.
4695 * Read the current clock counter of an Ethernet device
4697 * This returns the current raw clock value of an Ethernet device. It is
4698 * a raw amount of ticks, with no given time reference.
4699 * The value returned here is from the same clock than the one
4700 * filling timestamp field of Rx packets when using hardware timestamp
4701 * offload. Therefore it can be used to compute a precise conversion of
4702 * the device clock to the real time.
4704 * E.g, a simple heuristic to derivate the frequency would be:
4705 * uint64_t start, end;
4706 * rte_eth_read_clock(port, start);
4707 * rte_delay_ms(100);
4708 * rte_eth_read_clock(port, end);
4709 * double freq = (end - start) * 10;
4711 * Compute a common reference with:
4712 * uint64_t base_time_sec = current_time();
4713 * uint64_t base_clock;
4714 * rte_eth_read_clock(port, base_clock);
4716 * Then, convert the raw mbuf timestamp with:
4717 * base_time_sec + (double)(*timestamp_dynfield(mbuf) - base_clock) / freq;
4719 * This simple example will not provide a very good accuracy. One must
4720 * at least measure multiple times the frequency and do a regression.
4721 * To avoid deviation from the system time, the common reference can
4722 * be repeated from time to time. The integer division can also be
4723 * converted by a multiplication and a shift for better performance.
4726 * The port identifier of the Ethernet device.
4728 * Pointer to the uint64_t that holds the raw clock value.
4732 * - -ENODEV: The port ID is invalid.
4733 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4734 * - -EINVAL: if bad parameter.
4738 rte_eth_read_clock(uint16_t port_id, uint64_t *clock);
4741 * Get the port id from device name. The device name should be specified
4743 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:2:00.0
4744 * - SoC device name, for example- fsl-gmac0
4745 * - vdev dpdk name, for example- net_[pcap0|null0|tap0]
4748 * pci address or name of the device
4750 * pointer to port identifier of the device
4752 * - (0) if successful and port_id is filled.
4753 * - (-ENODEV or -EINVAL) on failure.
4756 rte_eth_dev_get_port_by_name(const char *name, uint16_t *port_id);
4759 * Get the device name from port id. The device name is specified as below:
4760 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:02:00.0
4761 * - SoC device name, for example- fsl-gmac0
4762 * - vdev dpdk name, for example- net_[pcap0|null0|tun0|tap0]
4765 * Port identifier of the device.
4767 * Buffer of size RTE_ETH_NAME_MAX_LEN to store the name.
4769 * - (0) if successful.
4770 * - (-ENODEV) if *port_id* is invalid.
4771 * - (-EINVAL) on failure.
4774 rte_eth_dev_get_name_by_port(uint16_t port_id, char *name);
4777 * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from
4778 * the ethernet device information, otherwise adjust them to boundaries.
4781 * The port identifier of the Ethernet device.
4783 * A pointer to a uint16_t where the number of receive
4784 * descriptors stored.
4786 * A pointer to a uint16_t where the number of transmit
4787 * descriptors stored.
4789 * - (0) if successful.
4790 * - (-ENOTSUP, -ENODEV or -EINVAL) on failure.
4792 int rte_eth_dev_adjust_nb_rx_tx_desc(uint16_t port_id,
4793 uint16_t *nb_rx_desc,
4794 uint16_t *nb_tx_desc);
4797 * Test if a port supports specific mempool ops.
4800 * Port identifier of the Ethernet device.
4802 * The name of the pool operations to test.
4804 * - 0: best mempool ops choice for this port.
4805 * - 1: mempool ops are supported for this port.
4806 * - -ENOTSUP: mempool ops not supported for this port.
4807 * - -ENODEV: Invalid port Identifier.
4808 * - -EINVAL: Pool param is null.
4811 rte_eth_dev_pool_ops_supported(uint16_t port_id, const char *pool);
4814 * Get the security context for the Ethernet device.
4817 * Port identifier of the Ethernet device
4820 * - pointer to security context on success.
4823 rte_eth_dev_get_sec_ctx(uint16_t port_id);
4827 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
4829 * Query the device hairpin capabilities.
4832 * The port identifier of the Ethernet device.
4834 * Pointer to a structure that will hold the hairpin capabilities.
4836 * - (0) if successful.
4837 * - (-ENOTSUP) if hardware doesn't support.
4838 * - (-EINVAL) if bad parameter.
4841 int rte_eth_dev_hairpin_capability_get(uint16_t port_id,
4842 struct rte_eth_hairpin_cap *cap);
4846 * @b EXPERIMENTAL: this structure may change without prior notice.
4848 * ethernet device representor ID range entry
4850 struct rte_eth_representor_range {
4851 enum rte_eth_representor_type type; /**< Representor type */
4852 int controller; /**< Controller index */
4853 int pf; /**< Physical function index */
4856 int vf; /**< VF start index */
4857 int sf; /**< SF start index */
4859 uint32_t id_base; /**< Representor ID start index */
4860 uint32_t id_end; /**< Representor ID end index */
4861 char name[RTE_DEV_NAME_MAX_LEN]; /**< Representor name */
4866 * @b EXPERIMENTAL: this structure may change without prior notice.
4868 * Ethernet device representor information
4870 struct rte_eth_representor_info {
4871 uint16_t controller; /**< Controller ID of caller device. */
4872 uint16_t pf; /**< Physical function ID of caller device. */
4873 uint32_t nb_ranges_alloc; /**< Size of the ranges array. */
4874 uint32_t nb_ranges; /**< Number of initialized ranges. */
4875 struct rte_eth_representor_range ranges[];/**< Representor ID range. */
4879 * Retrieve the representor info of the device.
4881 * Get device representor info to be able to calculate a unique
4882 * representor ID. @see rte_eth_representor_id_get helper.
4885 * The port identifier of the device.
4887 * A pointer to a representor info structure.
4888 * NULL to return number of range entries and allocate memory
4889 * for next call to store detail.
4890 * The number of ranges that were written into this structure
4891 * will be placed into its nb_ranges field. This number cannot be
4892 * larger than the nb_ranges_alloc that by the user before calling
4893 * this function. It can be smaller than the value returned by the
4894 * function, however.
4896 * - (-ENOTSUP) if operation is not supported.
4897 * - (-ENODEV) if *port_id* invalid.
4898 * - (-EIO) if device is removed.
4899 * - (>=0) number of available representor range entries.
4902 int rte_eth_representor_info_get(uint16_t port_id,
4903 struct rte_eth_representor_info *info);
4905 #include <rte_ethdev_core.h>
4909 * Retrieve a burst of input packets from a receive queue of an Ethernet
4910 * device. The retrieved packets are stored in *rte_mbuf* structures whose
4911 * pointers are supplied in the *rx_pkts* array.
4913 * The rte_eth_rx_burst() function loops, parsing the RX ring of the
4914 * receive queue, up to *nb_pkts* packets, and for each completed RX
4915 * descriptor in the ring, it performs the following operations:
4917 * - Initialize the *rte_mbuf* data structure associated with the
4918 * RX descriptor according to the information provided by the NIC into
4919 * that RX descriptor.
4921 * - Store the *rte_mbuf* data structure into the next entry of the
4924 * - Replenish the RX descriptor with a new *rte_mbuf* buffer
4925 * allocated from the memory pool associated with the receive queue at
4926 * initialization time.
4928 * When retrieving an input packet that was scattered by the controller
4929 * into multiple receive descriptors, the rte_eth_rx_burst() function
4930 * appends the associated *rte_mbuf* buffers to the first buffer of the
4933 * The rte_eth_rx_burst() function returns the number of packets
4934 * actually retrieved, which is the number of *rte_mbuf* data structures
4935 * effectively supplied into the *rx_pkts* array.
4936 * A return value equal to *nb_pkts* indicates that the RX queue contained
4937 * at least *rx_pkts* packets, and this is likely to signify that other
4938 * received packets remain in the input queue. Applications implementing
4939 * a "retrieve as much received packets as possible" policy can check this
4940 * specific case and keep invoking the rte_eth_rx_burst() function until
4941 * a value less than *nb_pkts* is returned.
4943 * This receive method has the following advantages:
4945 * - It allows a run-to-completion network stack engine to retrieve and
4946 * to immediately process received packets in a fast burst-oriented
4947 * approach, avoiding the overhead of unnecessary intermediate packet
4948 * queue/dequeue operations.
4950 * - Conversely, it also allows an asynchronous-oriented processing
4951 * method to retrieve bursts of received packets and to immediately
4952 * queue them for further parallel processing by another logical core,
4953 * for instance. However, instead of having received packets being
4954 * individually queued by the driver, this approach allows the caller
4955 * of the rte_eth_rx_burst() function to queue a burst of retrieved
4956 * packets at a time and therefore dramatically reduce the cost of
4957 * enqueue/dequeue operations per packet.
4959 * - It allows the rte_eth_rx_burst() function of the driver to take
4960 * advantage of burst-oriented hardware features (CPU cache,
4961 * prefetch instructions, and so on) to minimize the number of CPU
4962 * cycles per packet.
4964 * To summarize, the proposed receive API enables many
4965 * burst-oriented optimizations in both synchronous and asynchronous
4966 * packet processing environments with no overhead in both cases.
4969 * Some drivers using vector instructions require that *nb_pkts* is
4970 * divisible by 4 or 8, depending on the driver implementation.
4972 * The rte_eth_rx_burst() function does not provide any error
4973 * notification to avoid the corresponding overhead. As a hint, the
4974 * upper-level application might check the status of the device link once
4975 * being systematically returned a 0 value for a given number of tries.
4978 * The port identifier of the Ethernet device.
4980 * The index of the receive queue from which to retrieve input packets.
4981 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
4982 * to rte_eth_dev_configure().
4984 * The address of an array of pointers to *rte_mbuf* structures that
4985 * must be large enough to store *nb_pkts* pointers in it.
4987 * The maximum number of packets to retrieve.
4988 * The value must be divisible by 8 in order to work with any driver.
4990 * The number of packets actually retrieved, which is the number
4991 * of pointers to *rte_mbuf* structures effectively supplied to the
4994 static inline uint16_t
4995 rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id,
4996 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
4998 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
5001 #ifdef RTE_ETHDEV_DEBUG_RX
5002 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
5003 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_pkt_burst, 0);
5005 if (queue_id >= dev->data->nb_rx_queues) {
5006 RTE_ETHDEV_LOG(ERR, "Invalid RX queue_id=%u\n", queue_id);
5010 nb_rx = (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id],
5013 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
5014 struct rte_eth_rxtx_callback *cb;
5016 /* __ATOMIC_RELEASE memory order was used when the
5017 * call back was inserted into the list.
5018 * Since there is a clear dependency between loading
5019 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is
5022 cb = __atomic_load_n(&dev->post_rx_burst_cbs[queue_id],
5025 if (unlikely(cb != NULL)) {
5027 nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
5028 nb_pkts, cb->param);
5030 } while (cb != NULL);
5034 rte_ethdev_trace_rx_burst(port_id, queue_id, (void **)rx_pkts, nb_rx);
5039 * Get the number of used descriptors of a rx queue
5042 * The port identifier of the Ethernet device.
5044 * The queue id on the specific port.
5046 * The number of used descriptors in the specific queue, or:
5047 * - (-ENODEV) if *port_id* is invalid.
5048 * (-EINVAL) if *queue_id* is invalid
5049 * (-ENOTSUP) if the device does not support this function
5052 rte_eth_rx_queue_count(uint16_t port_id, uint16_t queue_id)
5054 struct rte_eth_dev *dev;
5056 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
5057 dev = &rte_eth_devices[port_id];
5058 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_queue_count, -ENOTSUP);
5059 if (queue_id >= dev->data->nb_rx_queues ||
5060 dev->data->rx_queues[queue_id] == NULL)
5063 return (int)(*dev->rx_queue_count)(dev, queue_id);
5067 * Check if the DD bit of the specific RX descriptor in the queue has been set
5070 * The port identifier of the Ethernet device.
5072 * The queue id on the specific port.
5074 * The offset of the descriptor ID from tail.
5076 * - (1) if the specific DD bit is set.
5077 * - (0) if the specific DD bit is not set.
5078 * - (-ENODEV) if *port_id* invalid.
5079 * - (-ENOTSUP) if the device does not support this function
5083 rte_eth_rx_descriptor_done(uint16_t port_id, uint16_t queue_id, uint16_t offset)
5085 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
5086 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
5087 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_descriptor_done, -ENOTSUP);
5088 return (*dev->rx_descriptor_done)(dev->data->rx_queues[queue_id], offset);
5091 /**@{@name Rx hardware descriptor states
5092 * @see rte_eth_rx_descriptor_status
5094 #define RTE_ETH_RX_DESC_AVAIL 0 /**< Desc available for hw. */
5095 #define RTE_ETH_RX_DESC_DONE 1 /**< Desc done, filled by hw. */
5096 #define RTE_ETH_RX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
5100 * Check the status of a Rx descriptor in the queue
5102 * It should be called in a similar context than the Rx function:
5103 * - on a dataplane core
5104 * - not concurrently on the same queue
5106 * Since it's a dataplane function, no check is performed on port_id and
5107 * queue_id. The caller must therefore ensure that the port is enabled
5108 * and the queue is configured and running.
5110 * Note: accessing to a random descriptor in the ring may trigger cache
5111 * misses and have a performance impact.
5114 * A valid port identifier of the Ethernet device which.
5116 * A valid Rx queue identifier on this port.
5118 * The offset of the descriptor starting from tail (0 is the next
5119 * packet to be received by the driver).
5122 * - (RTE_ETH_RX_DESC_AVAIL): Descriptor is available for the hardware to
5124 * - (RTE_ETH_RX_DESC_DONE): Descriptor is done, it is filled by hw, but
5125 * not yet processed by the driver (i.e. in the receive queue).
5126 * - (RTE_ETH_RX_DESC_UNAVAIL): Descriptor is unavailable, either hold by
5127 * the driver and not yet returned to hw, or reserved by the hw.
5128 * - (-EINVAL) bad descriptor offset.
5129 * - (-ENOTSUP) if the device does not support this function.
5130 * - (-ENODEV) bad port or queue (only if compiled with debug).
5133 rte_eth_rx_descriptor_status(uint16_t port_id, uint16_t queue_id,
5136 struct rte_eth_dev *dev;
5139 #ifdef RTE_ETHDEV_DEBUG_RX
5140 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
5142 dev = &rte_eth_devices[port_id];
5143 #ifdef RTE_ETHDEV_DEBUG_RX
5144 if (queue_id >= dev->data->nb_rx_queues)
5147 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_descriptor_status, -ENOTSUP);
5148 rxq = dev->data->rx_queues[queue_id];
5150 return (*dev->rx_descriptor_status)(rxq, offset);
5153 /**@{@name Tx hardware descriptor states
5154 * @see rte_eth_tx_descriptor_status
5156 #define RTE_ETH_TX_DESC_FULL 0 /**< Desc filled for hw, waiting xmit. */
5157 #define RTE_ETH_TX_DESC_DONE 1 /**< Desc done, packet is transmitted. */
5158 #define RTE_ETH_TX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
5162 * Check the status of a Tx descriptor in the queue.
5164 * It should be called in a similar context than the Tx function:
5165 * - on a dataplane core
5166 * - not concurrently on the same queue
5168 * Since it's a dataplane function, no check is performed on port_id and
5169 * queue_id. The caller must therefore ensure that the port is enabled
5170 * and the queue is configured and running.
5172 * Note: accessing to a random descriptor in the ring may trigger cache
5173 * misses and have a performance impact.
5176 * A valid port identifier of the Ethernet device which.
5178 * A valid Tx queue identifier on this port.
5180 * The offset of the descriptor starting from tail (0 is the place where
5181 * the next packet will be send).
5184 * - (RTE_ETH_TX_DESC_FULL) Descriptor is being processed by the hw, i.e.
5185 * in the transmit queue.
5186 * - (RTE_ETH_TX_DESC_DONE) Hardware is done with this descriptor, it can
5187 * be reused by the driver.
5188 * - (RTE_ETH_TX_DESC_UNAVAIL): Descriptor is unavailable, reserved by the
5189 * driver or the hardware.
5190 * - (-EINVAL) bad descriptor offset.
5191 * - (-ENOTSUP) if the device does not support this function.
5192 * - (-ENODEV) bad port or queue (only if compiled with debug).
5194 static inline int rte_eth_tx_descriptor_status(uint16_t port_id,
5195 uint16_t queue_id, uint16_t offset)
5197 struct rte_eth_dev *dev;
5200 #ifdef RTE_ETHDEV_DEBUG_TX
5201 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
5203 dev = &rte_eth_devices[port_id];
5204 #ifdef RTE_ETHDEV_DEBUG_TX
5205 if (queue_id >= dev->data->nb_tx_queues)
5208 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_descriptor_status, -ENOTSUP);
5209 txq = dev->data->tx_queues[queue_id];
5211 return (*dev->tx_descriptor_status)(txq, offset);
5215 * Send a burst of output packets on a transmit queue of an Ethernet device.
5217 * The rte_eth_tx_burst() function is invoked to transmit output packets
5218 * on the output queue *queue_id* of the Ethernet device designated by its
5220 * The *nb_pkts* parameter is the number of packets to send which are
5221 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
5222 * allocated from a pool created with rte_pktmbuf_pool_create().
5223 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets,
5224 * up to the number of transmit descriptors available in the TX ring of the
5226 * For each packet to send, the rte_eth_tx_burst() function performs
5227 * the following operations:
5229 * - Pick up the next available descriptor in the transmit ring.
5231 * - Free the network buffer previously sent with that descriptor, if any.
5233 * - Initialize the transmit descriptor with the information provided
5234 * in the *rte_mbuf data structure.
5236 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers,
5237 * the rte_eth_tx_burst() function uses several transmit descriptors
5240 * The rte_eth_tx_burst() function returns the number of packets it
5241 * actually sent. A return value equal to *nb_pkts* means that all packets
5242 * have been sent, and this is likely to signify that other output packets
5243 * could be immediately transmitted again. Applications that implement a
5244 * "send as many packets to transmit as possible" policy can check this
5245 * specific case and keep invoking the rte_eth_tx_burst() function until
5246 * a value less than *nb_pkts* is returned.
5248 * It is the responsibility of the rte_eth_tx_burst() function to
5249 * transparently free the memory buffers of packets previously sent.
5250 * This feature is driven by the *tx_free_thresh* value supplied to the
5251 * rte_eth_dev_configure() function at device configuration time.
5252 * When the number of free TX descriptors drops below this threshold, the
5253 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers
5254 * of those packets whose transmission was effectively completed.
5256 * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can
5257 * invoke this function concurrently on the same tx queue without SW lock.
5258 * @see rte_eth_dev_info_get, struct rte_eth_txconf::offloads
5260 * @see rte_eth_tx_prepare to perform some prior checks or adjustments
5264 * The port identifier of the Ethernet device.
5266 * The index of the transmit queue through which output packets must be
5268 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5269 * to rte_eth_dev_configure().
5271 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
5272 * which contain the output packets.
5274 * The maximum number of packets to transmit.
5276 * The number of output packets actually stored in transmit descriptors of
5277 * the transmit ring. The return value can be less than the value of the
5278 * *tx_pkts* parameter when the transmit ring is full or has been filled up.
5280 static inline uint16_t
5281 rte_eth_tx_burst(uint16_t port_id, uint16_t queue_id,
5282 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
5284 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
5286 #ifdef RTE_ETHDEV_DEBUG_TX
5287 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
5288 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_pkt_burst, 0);
5290 if (queue_id >= dev->data->nb_tx_queues) {
5291 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
5296 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
5297 struct rte_eth_rxtx_callback *cb;
5299 /* __ATOMIC_RELEASE memory order was used when the
5300 * call back was inserted into the list.
5301 * Since there is a clear dependency between loading
5302 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is
5305 cb = __atomic_load_n(&dev->pre_tx_burst_cbs[queue_id],
5308 if (unlikely(cb != NULL)) {
5310 nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
5313 } while (cb != NULL);
5317 rte_ethdev_trace_tx_burst(port_id, queue_id, (void **)tx_pkts,
5319 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts);
5323 * Process a burst of output packets on a transmit queue of an Ethernet device.
5325 * The rte_eth_tx_prepare() function is invoked to prepare output packets to be
5326 * transmitted on the output queue *queue_id* of the Ethernet device designated
5328 * The *nb_pkts* parameter is the number of packets to be prepared which are
5329 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
5330 * allocated from a pool created with rte_pktmbuf_pool_create().
5331 * For each packet to send, the rte_eth_tx_prepare() function performs
5332 * the following operations:
5334 * - Check if packet meets devices requirements for tx offloads.
5336 * - Check limitations about number of segments.
5338 * - Check additional requirements when debug is enabled.
5340 * - Update and/or reset required checksums when tx offload is set for packet.
5342 * Since this function can modify packet data, provided mbufs must be safely
5343 * writable (e.g. modified data cannot be in shared segment).
5345 * The rte_eth_tx_prepare() function returns the number of packets ready to be
5346 * sent. A return value equal to *nb_pkts* means that all packets are valid and
5347 * ready to be sent, otherwise stops processing on the first invalid packet and
5348 * leaves the rest packets untouched.
5350 * When this functionality is not implemented in the driver, all packets are
5351 * are returned untouched.
5354 * The port identifier of the Ethernet device.
5355 * The value must be a valid port id.
5357 * The index of the transmit queue through which output packets must be
5359 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5360 * to rte_eth_dev_configure().
5362 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
5363 * which contain the output packets.
5365 * The maximum number of packets to process.
5367 * The number of packets correct and ready to be sent. The return value can be
5368 * less than the value of the *tx_pkts* parameter when some packet doesn't
5369 * meet devices requirements with rte_errno set appropriately:
5370 * - EINVAL: offload flags are not correctly set
5371 * - ENOTSUP: the offload feature is not supported by the hardware
5372 * - ENODEV: if *port_id* is invalid (with debug enabled only)
5376 #ifndef RTE_ETHDEV_TX_PREPARE_NOOP
5378 static inline uint16_t
5379 rte_eth_tx_prepare(uint16_t port_id, uint16_t queue_id,
5380 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
5382 struct rte_eth_dev *dev;
5384 #ifdef RTE_ETHDEV_DEBUG_TX
5385 if (!rte_eth_dev_is_valid_port(port_id)) {
5386 RTE_ETHDEV_LOG(ERR, "Invalid TX port_id=%u\n", port_id);
5392 dev = &rte_eth_devices[port_id];
5394 #ifdef RTE_ETHDEV_DEBUG_TX
5395 if (queue_id >= dev->data->nb_tx_queues) {
5396 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
5402 if (!dev->tx_pkt_prepare)
5405 return (*dev->tx_pkt_prepare)(dev->data->tx_queues[queue_id],
5412 * Native NOOP operation for compilation targets which doesn't require any
5413 * preparations steps, and functional NOOP may introduce unnecessary performance
5416 * Generally this is not a good idea to turn it on globally and didn't should
5417 * be used if behavior of tx_preparation can change.
5420 static inline uint16_t
5421 rte_eth_tx_prepare(__rte_unused uint16_t port_id,
5422 __rte_unused uint16_t queue_id,
5423 __rte_unused struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
5431 * Send any packets queued up for transmission on a port and HW queue
5433 * This causes an explicit flush of packets previously buffered via the
5434 * rte_eth_tx_buffer() function. It returns the number of packets successfully
5435 * sent to the NIC, and calls the error callback for any unsent packets. Unless
5436 * explicitly set up otherwise, the default callback simply frees the unsent
5437 * packets back to the owning mempool.
5440 * The port identifier of the Ethernet device.
5442 * The index of the transmit queue through which output packets must be
5444 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5445 * to rte_eth_dev_configure().
5447 * Buffer of packets to be transmit.
5449 * The number of packets successfully sent to the Ethernet device. The error
5450 * callback is called for any packets which could not be sent.
5452 static inline uint16_t
5453 rte_eth_tx_buffer_flush(uint16_t port_id, uint16_t queue_id,
5454 struct rte_eth_dev_tx_buffer *buffer)
5457 uint16_t to_send = buffer->length;
5462 sent = rte_eth_tx_burst(port_id, queue_id, buffer->pkts, to_send);
5466 /* All packets sent, or to be dealt with by callback below */
5467 if (unlikely(sent != to_send))
5468 buffer->error_callback(&buffer->pkts[sent],
5469 (uint16_t)(to_send - sent),
5470 buffer->error_userdata);
5476 * Buffer a single packet for future transmission on a port and queue
5478 * This function takes a single mbuf/packet and buffers it for later
5479 * transmission on the particular port and queue specified. Once the buffer is
5480 * full of packets, an attempt will be made to transmit all the buffered
5481 * packets. In case of error, where not all packets can be transmitted, a
5482 * callback is called with the unsent packets as a parameter. If no callback
5483 * is explicitly set up, the unsent packets are just freed back to the owning
5484 * mempool. The function returns the number of packets actually sent i.e.
5485 * 0 if no buffer flush occurred, otherwise the number of packets successfully
5489 * The port identifier of the Ethernet device.
5491 * The index of the transmit queue through which output packets must be
5493 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5494 * to rte_eth_dev_configure().
5496 * Buffer used to collect packets to be sent.
5498 * Pointer to the packet mbuf to be sent.
5500 * 0 = packet has been buffered for later transmission
5501 * N > 0 = packet has been buffered, and the buffer was subsequently flushed,
5502 * causing N packets to be sent, and the error callback to be called for
5505 static __rte_always_inline uint16_t
5506 rte_eth_tx_buffer(uint16_t port_id, uint16_t queue_id,
5507 struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt)
5509 buffer->pkts[buffer->length++] = tx_pkt;
5510 if (buffer->length < buffer->size)
5513 return rte_eth_tx_buffer_flush(port_id, queue_id, buffer);
5520 #endif /* _RTE_ETHDEV_H_ */