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 rte_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 #include <rte_compat.h>
153 #include <rte_interrupts.h>
155 #include <rte_devargs.h>
156 #include <rte_errno.h>
157 #include <rte_common.h>
158 #include <rte_config.h>
159 #include <rte_ether.h>
161 #include "rte_ethdev_trace_fp.h"
162 #include "rte_dev_info.h"
164 extern int rte_eth_dev_logtype;
166 #define RTE_ETHDEV_LOG(level, ...) \
167 rte_log(RTE_LOG_ ## level, rte_eth_dev_logtype, "" __VA_ARGS__)
172 * Initializes a device iterator.
174 * This iterator allows accessing a list of devices matching some devargs.
177 * Device iterator handle initialized by the function.
178 * The fields bus_str and cls_str might be dynamically allocated,
179 * and could be freed by calling rte_eth_iterator_cleanup().
182 * Device description string.
185 * 0 on successful initialization, negative otherwise.
187 int rte_eth_iterator_init(struct rte_dev_iterator *iter, const char *devargs);
190 * Iterates on devices with devargs filter.
191 * The ownership is not checked.
193 * The next port id is returned, and the iterator is updated.
196 * Device iterator handle initialized by rte_eth_iterator_init().
197 * Some fields bus_str and cls_str might be freed when no more port is found,
198 * by calling rte_eth_iterator_cleanup().
201 * A port id if found, RTE_MAX_ETHPORTS otherwise.
203 uint16_t rte_eth_iterator_next(struct rte_dev_iterator *iter);
206 * Free some allocated fields of the iterator.
208 * This function is automatically called by rte_eth_iterator_next()
209 * on the last iteration (i.e. when no more matching port is found).
211 * It is safe to call this function twice; it will do nothing more.
214 * Device iterator handle initialized by rte_eth_iterator_init().
215 * The fields bus_str and cls_str are freed if needed.
217 void rte_eth_iterator_cleanup(struct rte_dev_iterator *iter);
220 * Macro to iterate over all ethdev ports matching some devargs.
222 * If a break is done before the end of the loop,
223 * the function rte_eth_iterator_cleanup() must be called.
226 * Iterated port id of type uint16_t.
228 * Device parameters input as string of type char*.
230 * Iterator handle of type struct rte_dev_iterator, used internally.
232 #define RTE_ETH_FOREACH_MATCHING_DEV(id, devargs, iter) \
233 for (rte_eth_iterator_init(iter, devargs), \
234 id = rte_eth_iterator_next(iter); \
235 id != RTE_MAX_ETHPORTS; \
236 id = rte_eth_iterator_next(iter))
239 * A structure used to retrieve statistics for an Ethernet port.
240 * Not all statistics fields in struct rte_eth_stats are supported
241 * by any type of network interface card (NIC). If any statistics
242 * field is not supported, its value is 0.
244 struct rte_eth_stats {
245 uint64_t ipackets; /**< Total number of successfully received packets. */
246 uint64_t opackets; /**< Total number of successfully transmitted packets.*/
247 uint64_t ibytes; /**< Total number of successfully received bytes. */
248 uint64_t obytes; /**< Total number of successfully transmitted bytes. */
250 /**< Total of RX packets dropped by the HW,
251 * because there are no available buffer (i.e. RX queues are full).
253 uint64_t ierrors; /**< Total number of erroneous received packets. */
254 uint64_t oerrors; /**< Total number of failed transmitted packets. */
255 uint64_t rx_nombuf; /**< Total number of RX mbuf allocation failures. */
256 /* Queue stats are limited to max 256 queues */
257 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
258 /**< Total number of queue RX packets. */
259 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
260 /**< Total number of queue TX packets. */
261 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
262 /**< Total number of successfully received queue bytes. */
263 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
264 /**< Total number of successfully transmitted queue bytes. */
265 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS];
266 /**< Total number of queue packets received that are dropped. */
270 * Device supported speeds bitmap flags
272 #define ETH_LINK_SPEED_AUTONEG (0 << 0) /**< Autonegotiate (all speeds) */
273 #define ETH_LINK_SPEED_FIXED (1 << 0) /**< Disable autoneg (fixed speed) */
274 #define ETH_LINK_SPEED_10M_HD (1 << 1) /**< 10 Mbps half-duplex */
275 #define ETH_LINK_SPEED_10M (1 << 2) /**< 10 Mbps full-duplex */
276 #define ETH_LINK_SPEED_100M_HD (1 << 3) /**< 100 Mbps half-duplex */
277 #define ETH_LINK_SPEED_100M (1 << 4) /**< 100 Mbps full-duplex */
278 #define ETH_LINK_SPEED_1G (1 << 5) /**< 1 Gbps */
279 #define ETH_LINK_SPEED_2_5G (1 << 6) /**< 2.5 Gbps */
280 #define ETH_LINK_SPEED_5G (1 << 7) /**< 5 Gbps */
281 #define ETH_LINK_SPEED_10G (1 << 8) /**< 10 Gbps */
282 #define ETH_LINK_SPEED_20G (1 << 9) /**< 20 Gbps */
283 #define ETH_LINK_SPEED_25G (1 << 10) /**< 25 Gbps */
284 #define ETH_LINK_SPEED_40G (1 << 11) /**< 40 Gbps */
285 #define ETH_LINK_SPEED_50G (1 << 12) /**< 50 Gbps */
286 #define ETH_LINK_SPEED_56G (1 << 13) /**< 56 Gbps */
287 #define ETH_LINK_SPEED_100G (1 << 14) /**< 100 Gbps */
288 #define ETH_LINK_SPEED_200G (1 << 15) /**< 200 Gbps */
291 * Ethernet numeric link speeds in Mbps
293 #define ETH_SPEED_NUM_NONE 0 /**< Not defined */
294 #define ETH_SPEED_NUM_10M 10 /**< 10 Mbps */
295 #define ETH_SPEED_NUM_100M 100 /**< 100 Mbps */
296 #define ETH_SPEED_NUM_1G 1000 /**< 1 Gbps */
297 #define ETH_SPEED_NUM_2_5G 2500 /**< 2.5 Gbps */
298 #define ETH_SPEED_NUM_5G 5000 /**< 5 Gbps */
299 #define ETH_SPEED_NUM_10G 10000 /**< 10 Gbps */
300 #define ETH_SPEED_NUM_20G 20000 /**< 20 Gbps */
301 #define ETH_SPEED_NUM_25G 25000 /**< 25 Gbps */
302 #define ETH_SPEED_NUM_40G 40000 /**< 40 Gbps */
303 #define ETH_SPEED_NUM_50G 50000 /**< 50 Gbps */
304 #define ETH_SPEED_NUM_56G 56000 /**< 56 Gbps */
305 #define ETH_SPEED_NUM_100G 100000 /**< 100 Gbps */
306 #define ETH_SPEED_NUM_200G 200000 /**< 200 Gbps */
307 #define ETH_SPEED_NUM_UNKNOWN UINT32_MAX /**< Unknown */
310 * A structure used to retrieve link-level information of an Ethernet port.
313 struct rte_eth_link {
314 uint32_t link_speed; /**< ETH_SPEED_NUM_ */
315 uint16_t link_duplex : 1; /**< ETH_LINK_[HALF/FULL]_DUPLEX */
316 uint16_t link_autoneg : 1; /**< ETH_LINK_[AUTONEG/FIXED] */
317 uint16_t link_status : 1; /**< ETH_LINK_[DOWN/UP] */
318 } __rte_aligned(8); /**< aligned for atomic64 read/write */
320 /* Utility constants */
321 #define ETH_LINK_HALF_DUPLEX 0 /**< Half-duplex connection (see link_duplex). */
322 #define ETH_LINK_FULL_DUPLEX 1 /**< Full-duplex connection (see link_duplex). */
323 #define ETH_LINK_DOWN 0 /**< Link is down (see link_status). */
324 #define ETH_LINK_UP 1 /**< Link is up (see link_status). */
325 #define ETH_LINK_FIXED 0 /**< No autonegotiation (see link_autoneg). */
326 #define ETH_LINK_AUTONEG 1 /**< Autonegotiated (see link_autoneg). */
327 #define RTE_ETH_LINK_MAX_STR_LEN 40 /**< Max length of default link string. */
330 * A structure used to configure the ring threshold registers of an RX/TX
331 * queue for an Ethernet port.
333 struct rte_eth_thresh {
334 uint8_t pthresh; /**< Ring prefetch threshold. */
335 uint8_t hthresh; /**< Ring host threshold. */
336 uint8_t wthresh; /**< Ring writeback threshold. */
340 * Simple flags are used for rte_eth_conf.rxmode.mq_mode.
342 #define ETH_MQ_RX_RSS_FLAG 0x1
343 #define ETH_MQ_RX_DCB_FLAG 0x2
344 #define ETH_MQ_RX_VMDQ_FLAG 0x4
347 * A set of values to identify what method is to be used to route
348 * packets to multiple queues.
350 enum rte_eth_rx_mq_mode {
351 /** None of DCB,RSS or VMDQ mode */
354 /** For RX side, only RSS is on */
355 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG,
356 /** For RX side,only DCB is on. */
357 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG,
358 /** Both DCB and RSS enable */
359 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG,
361 /** Only VMDQ, no RSS nor DCB */
362 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG,
363 /** RSS mode with VMDQ */
364 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG,
365 /** Use VMDQ+DCB to route traffic to queues */
366 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG,
367 /** Enable both VMDQ and DCB in VMDq */
368 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG |
373 * for rx mq mode backward compatible
375 #define ETH_RSS ETH_MQ_RX_RSS
376 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB
377 #define ETH_DCB_RX ETH_MQ_RX_DCB
380 * A set of values to identify what method is to be used to transmit
381 * packets using multi-TCs.
383 enum rte_eth_tx_mq_mode {
384 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */
385 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */
386 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */
387 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */
391 * for tx mq mode backward compatible
393 #define ETH_DCB_NONE ETH_MQ_TX_NONE
394 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB
395 #define ETH_DCB_TX ETH_MQ_TX_DCB
398 * A structure used to configure the RX features of an Ethernet port.
400 struct rte_eth_rxmode {
401 /** The multi-queue packet distribution mode to be used, e.g. RSS. */
402 enum rte_eth_rx_mq_mode mq_mode;
403 uint32_t max_rx_pkt_len; /**< Only used if JUMBO_FRAME enabled. */
404 /** Maximum allowed size of LRO aggregated packet. */
405 uint32_t max_lro_pkt_size;
406 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/
408 * Per-port Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
409 * Only offloads set on rx_offload_capa field on rte_eth_dev_info
410 * structure are allowed to be set.
414 uint64_t reserved_64s[2]; /**< Reserved for future fields */
415 void *reserved_ptrs[2]; /**< Reserved for future fields */
419 * VLAN types to indicate if it is for single VLAN, inner VLAN or outer VLAN.
420 * Note that single VLAN is treated the same as inner VLAN.
423 ETH_VLAN_TYPE_UNKNOWN = 0,
424 ETH_VLAN_TYPE_INNER, /**< Inner VLAN. */
425 ETH_VLAN_TYPE_OUTER, /**< Single VLAN, or outer VLAN. */
430 * A structure used to describe a vlan filter.
431 * If the bit corresponding to a VID is set, such VID is on.
433 struct rte_vlan_filter_conf {
438 * A structure used to configure the Receive Side Scaling (RSS) feature
439 * of an Ethernet port.
440 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points
441 * to an array holding the RSS key to use for hashing specific header
442 * fields of received packets. The length of this array should be indicated
443 * by *rss_key_len* below. Otherwise, a default random hash key is used by
446 * The *rss_key_len* field of the *rss_conf* structure indicates the length
447 * in bytes of the array pointed by *rss_key*. To be compatible, this length
448 * will be checked in i40e only. Others assume 40 bytes to be used as before.
450 * The *rss_hf* field of the *rss_conf* structure indicates the different
451 * types of IPv4/IPv6 packets to which the RSS hashing must be applied.
452 * Supplying an *rss_hf* equal to zero disables the RSS feature.
454 struct rte_eth_rss_conf {
455 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */
456 uint8_t rss_key_len; /**< hash key length in bytes. */
457 uint64_t rss_hf; /**< Hash functions to apply - see below. */
461 * A packet can be identified by hardware as different flow types. Different
462 * NIC hardware may support different flow types.
463 * Basically, the NIC hardware identifies the flow type as deep protocol as
464 * possible, and exclusively. For example, if a packet is identified as
465 * 'RTE_ETH_FLOW_NONFRAG_IPV4_TCP', it will not be any of other flow types,
466 * though it is an actual IPV4 packet.
468 #define RTE_ETH_FLOW_UNKNOWN 0
469 #define RTE_ETH_FLOW_RAW 1
470 #define RTE_ETH_FLOW_IPV4 2
471 #define RTE_ETH_FLOW_FRAG_IPV4 3
472 #define RTE_ETH_FLOW_NONFRAG_IPV4_TCP 4
473 #define RTE_ETH_FLOW_NONFRAG_IPV4_UDP 5
474 #define RTE_ETH_FLOW_NONFRAG_IPV4_SCTP 6
475 #define RTE_ETH_FLOW_NONFRAG_IPV4_OTHER 7
476 #define RTE_ETH_FLOW_IPV6 8
477 #define RTE_ETH_FLOW_FRAG_IPV6 9
478 #define RTE_ETH_FLOW_NONFRAG_IPV6_TCP 10
479 #define RTE_ETH_FLOW_NONFRAG_IPV6_UDP 11
480 #define RTE_ETH_FLOW_NONFRAG_IPV6_SCTP 12
481 #define RTE_ETH_FLOW_NONFRAG_IPV6_OTHER 13
482 #define RTE_ETH_FLOW_L2_PAYLOAD 14
483 #define RTE_ETH_FLOW_IPV6_EX 15
484 #define RTE_ETH_FLOW_IPV6_TCP_EX 16
485 #define RTE_ETH_FLOW_IPV6_UDP_EX 17
486 #define RTE_ETH_FLOW_PORT 18
487 /**< Consider device port number as a flow differentiator */
488 #define RTE_ETH_FLOW_VXLAN 19 /**< VXLAN protocol based flow */
489 #define RTE_ETH_FLOW_GENEVE 20 /**< GENEVE protocol based flow */
490 #define RTE_ETH_FLOW_NVGRE 21 /**< NVGRE protocol based flow */
491 #define RTE_ETH_FLOW_VXLAN_GPE 22 /**< VXLAN-GPE protocol based flow */
492 #define RTE_ETH_FLOW_GTPU 23 /**< GTPU protocol based flow */
493 #define RTE_ETH_FLOW_MAX 24
496 * Below macros are defined for RSS offload types, they can be used to
497 * fill rte_eth_rss_conf.rss_hf or rte_flow_action_rss.types.
499 #define ETH_RSS_IPV4 (1ULL << 2)
500 #define ETH_RSS_FRAG_IPV4 (1ULL << 3)
501 #define ETH_RSS_NONFRAG_IPV4_TCP (1ULL << 4)
502 #define ETH_RSS_NONFRAG_IPV4_UDP (1ULL << 5)
503 #define ETH_RSS_NONFRAG_IPV4_SCTP (1ULL << 6)
504 #define ETH_RSS_NONFRAG_IPV4_OTHER (1ULL << 7)
505 #define ETH_RSS_IPV6 (1ULL << 8)
506 #define ETH_RSS_FRAG_IPV6 (1ULL << 9)
507 #define ETH_RSS_NONFRAG_IPV6_TCP (1ULL << 10)
508 #define ETH_RSS_NONFRAG_IPV6_UDP (1ULL << 11)
509 #define ETH_RSS_NONFRAG_IPV6_SCTP (1ULL << 12)
510 #define ETH_RSS_NONFRAG_IPV6_OTHER (1ULL << 13)
511 #define ETH_RSS_L2_PAYLOAD (1ULL << 14)
512 #define ETH_RSS_IPV6_EX (1ULL << 15)
513 #define ETH_RSS_IPV6_TCP_EX (1ULL << 16)
514 #define ETH_RSS_IPV6_UDP_EX (1ULL << 17)
515 #define ETH_RSS_PORT (1ULL << 18)
516 #define ETH_RSS_VXLAN (1ULL << 19)
517 #define ETH_RSS_GENEVE (1ULL << 20)
518 #define ETH_RSS_NVGRE (1ULL << 21)
519 #define ETH_RSS_GTPU (1ULL << 23)
520 #define ETH_RSS_ETH (1ULL << 24)
521 #define ETH_RSS_S_VLAN (1ULL << 25)
522 #define ETH_RSS_C_VLAN (1ULL << 26)
523 #define ETH_RSS_ESP (1ULL << 27)
524 #define ETH_RSS_AH (1ULL << 28)
525 #define ETH_RSS_L2TPV3 (1ULL << 29)
526 #define ETH_RSS_PFCP (1ULL << 30)
527 #define ETH_RSS_PPPOE (1ULL << 31)
530 * We use the following macros to combine with above ETH_RSS_* for
531 * more specific input set selection. These bits are defined starting
532 * from the high end of the 64 bits.
533 * Note: If we use above ETH_RSS_* without SRC/DST_ONLY, it represents
534 * both SRC and DST are taken into account. If SRC_ONLY and DST_ONLY of
535 * the same level are used simultaneously, it is the same case as none of
538 #define ETH_RSS_L3_SRC_ONLY (1ULL << 63)
539 #define ETH_RSS_L3_DST_ONLY (1ULL << 62)
540 #define ETH_RSS_L4_SRC_ONLY (1ULL << 61)
541 #define ETH_RSS_L4_DST_ONLY (1ULL << 60)
542 #define ETH_RSS_L2_SRC_ONLY (1ULL << 59)
543 #define ETH_RSS_L2_DST_ONLY (1ULL << 58)
546 * Only select IPV6 address prefix as RSS input set according to
547 * https://tools.ietf.org/html/rfc6052
548 * Must be combined with ETH_RSS_IPV6, ETH_RSS_NONFRAG_IPV6_UDP,
549 * ETH_RSS_NONFRAG_IPV6_TCP, ETH_RSS_NONFRAG_IPV6_SCTP.
551 #define RTE_ETH_RSS_L3_PRE32 (1ULL << 57)
552 #define RTE_ETH_RSS_L3_PRE40 (1ULL << 56)
553 #define RTE_ETH_RSS_L3_PRE48 (1ULL << 55)
554 #define RTE_ETH_RSS_L3_PRE56 (1ULL << 54)
555 #define RTE_ETH_RSS_L3_PRE64 (1ULL << 53)
556 #define RTE_ETH_RSS_L3_PRE96 (1ULL << 52)
559 * Use the following macros to combine with the above layers
560 * to choose inner and outer layers or both for RSS computation.
561 * Bits 50 and 51 are reserved for this.
565 * level 0, requests the default behavior.
566 * Depending on the packet type, it can mean outermost, innermost,
567 * anything in between or even no RSS.
568 * It basically stands for the innermost encapsulation level RSS
569 * can be performed on according to PMD and device capabilities.
571 #define ETH_RSS_LEVEL_PMD_DEFAULT (0ULL << 50)
574 * level 1, requests RSS to be performed on the outermost packet
575 * encapsulation level.
577 #define ETH_RSS_LEVEL_OUTERMOST (1ULL << 50)
580 * level 2, requests RSS to be performed on the specified inner packet
581 * encapsulation level, from outermost to innermost (lower to higher values).
583 #define ETH_RSS_LEVEL_INNERMOST (2ULL << 50)
584 #define ETH_RSS_LEVEL_MASK (3ULL << 50)
586 #define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
589 * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
590 * the same level are used simultaneously, it is the same case as
591 * none of them are added.
594 * RSS types with SRC/DST_ONLY.
598 static inline uint64_t
599 rte_eth_rss_hf_refine(uint64_t rss_hf)
601 if ((rss_hf & ETH_RSS_L3_SRC_ONLY) && (rss_hf & ETH_RSS_L3_DST_ONLY))
602 rss_hf &= ~(ETH_RSS_L3_SRC_ONLY | ETH_RSS_L3_DST_ONLY);
604 if ((rss_hf & ETH_RSS_L4_SRC_ONLY) && (rss_hf & ETH_RSS_L4_DST_ONLY))
605 rss_hf &= ~(ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY);
610 #define ETH_RSS_IPV6_PRE32 ( \
612 RTE_ETH_RSS_L3_PRE32)
614 #define ETH_RSS_IPV6_PRE40 ( \
616 RTE_ETH_RSS_L3_PRE40)
618 #define ETH_RSS_IPV6_PRE48 ( \
620 RTE_ETH_RSS_L3_PRE48)
622 #define ETH_RSS_IPV6_PRE56 ( \
624 RTE_ETH_RSS_L3_PRE56)
626 #define ETH_RSS_IPV6_PRE64 ( \
628 RTE_ETH_RSS_L3_PRE64)
630 #define ETH_RSS_IPV6_PRE96 ( \
632 RTE_ETH_RSS_L3_PRE96)
634 #define ETH_RSS_IPV6_PRE32_UDP ( \
635 ETH_RSS_NONFRAG_IPV6_UDP | \
636 RTE_ETH_RSS_L3_PRE32)
638 #define ETH_RSS_IPV6_PRE40_UDP ( \
639 ETH_RSS_NONFRAG_IPV6_UDP | \
640 RTE_ETH_RSS_L3_PRE40)
642 #define ETH_RSS_IPV6_PRE48_UDP ( \
643 ETH_RSS_NONFRAG_IPV6_UDP | \
644 RTE_ETH_RSS_L3_PRE48)
646 #define ETH_RSS_IPV6_PRE56_UDP ( \
647 ETH_RSS_NONFRAG_IPV6_UDP | \
648 RTE_ETH_RSS_L3_PRE56)
650 #define ETH_RSS_IPV6_PRE64_UDP ( \
651 ETH_RSS_NONFRAG_IPV6_UDP | \
652 RTE_ETH_RSS_L3_PRE64)
654 #define ETH_RSS_IPV6_PRE96_UDP ( \
655 ETH_RSS_NONFRAG_IPV6_UDP | \
656 RTE_ETH_RSS_L3_PRE96)
658 #define ETH_RSS_IPV6_PRE32_TCP ( \
659 ETH_RSS_NONFRAG_IPV6_TCP | \
660 RTE_ETH_RSS_L3_PRE32)
662 #define ETH_RSS_IPV6_PRE40_TCP ( \
663 ETH_RSS_NONFRAG_IPV6_TCP | \
664 RTE_ETH_RSS_L3_PRE40)
666 #define ETH_RSS_IPV6_PRE48_TCP ( \
667 ETH_RSS_NONFRAG_IPV6_TCP | \
668 RTE_ETH_RSS_L3_PRE48)
670 #define ETH_RSS_IPV6_PRE56_TCP ( \
671 ETH_RSS_NONFRAG_IPV6_TCP | \
672 RTE_ETH_RSS_L3_PRE56)
674 #define ETH_RSS_IPV6_PRE64_TCP ( \
675 ETH_RSS_NONFRAG_IPV6_TCP | \
676 RTE_ETH_RSS_L3_PRE64)
678 #define ETH_RSS_IPV6_PRE96_TCP ( \
679 ETH_RSS_NONFRAG_IPV6_TCP | \
680 RTE_ETH_RSS_L3_PRE96)
682 #define ETH_RSS_IPV6_PRE32_SCTP ( \
683 ETH_RSS_NONFRAG_IPV6_SCTP | \
684 RTE_ETH_RSS_L3_PRE32)
686 #define ETH_RSS_IPV6_PRE40_SCTP ( \
687 ETH_RSS_NONFRAG_IPV6_SCTP | \
688 RTE_ETH_RSS_L3_PRE40)
690 #define ETH_RSS_IPV6_PRE48_SCTP ( \
691 ETH_RSS_NONFRAG_IPV6_SCTP | \
692 RTE_ETH_RSS_L3_PRE48)
694 #define ETH_RSS_IPV6_PRE56_SCTP ( \
695 ETH_RSS_NONFRAG_IPV6_SCTP | \
696 RTE_ETH_RSS_L3_PRE56)
698 #define ETH_RSS_IPV6_PRE64_SCTP ( \
699 ETH_RSS_NONFRAG_IPV6_SCTP | \
700 RTE_ETH_RSS_L3_PRE64)
702 #define ETH_RSS_IPV6_PRE96_SCTP ( \
703 ETH_RSS_NONFRAG_IPV6_SCTP | \
704 RTE_ETH_RSS_L3_PRE96)
706 #define ETH_RSS_IP ( \
708 ETH_RSS_FRAG_IPV4 | \
709 ETH_RSS_NONFRAG_IPV4_OTHER | \
711 ETH_RSS_FRAG_IPV6 | \
712 ETH_RSS_NONFRAG_IPV6_OTHER | \
715 #define ETH_RSS_UDP ( \
716 ETH_RSS_NONFRAG_IPV4_UDP | \
717 ETH_RSS_NONFRAG_IPV6_UDP | \
720 #define ETH_RSS_TCP ( \
721 ETH_RSS_NONFRAG_IPV4_TCP | \
722 ETH_RSS_NONFRAG_IPV6_TCP | \
725 #define ETH_RSS_SCTP ( \
726 ETH_RSS_NONFRAG_IPV4_SCTP | \
727 ETH_RSS_NONFRAG_IPV6_SCTP)
729 #define ETH_RSS_TUNNEL ( \
734 #define ETH_RSS_VLAN ( \
738 /**< Mask of valid RSS hash protocols */
739 #define ETH_RSS_PROTO_MASK ( \
741 ETH_RSS_FRAG_IPV4 | \
742 ETH_RSS_NONFRAG_IPV4_TCP | \
743 ETH_RSS_NONFRAG_IPV4_UDP | \
744 ETH_RSS_NONFRAG_IPV4_SCTP | \
745 ETH_RSS_NONFRAG_IPV4_OTHER | \
747 ETH_RSS_FRAG_IPV6 | \
748 ETH_RSS_NONFRAG_IPV6_TCP | \
749 ETH_RSS_NONFRAG_IPV6_UDP | \
750 ETH_RSS_NONFRAG_IPV6_SCTP | \
751 ETH_RSS_NONFRAG_IPV6_OTHER | \
752 ETH_RSS_L2_PAYLOAD | \
754 ETH_RSS_IPV6_TCP_EX | \
755 ETH_RSS_IPV6_UDP_EX | \
762 * Definitions used for redirection table entry size.
763 * Some RSS RETA sizes may not be supported by some drivers, check the
764 * documentation or the description of relevant functions for more details.
766 #define ETH_RSS_RETA_SIZE_64 64
767 #define ETH_RSS_RETA_SIZE_128 128
768 #define ETH_RSS_RETA_SIZE_256 256
769 #define ETH_RSS_RETA_SIZE_512 512
770 #define RTE_RETA_GROUP_SIZE 64
772 /* Definitions used for VMDQ and DCB functionality */
773 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */
774 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */
775 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */
776 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */
778 /* DCB capability defines */
779 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */
780 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */
782 /* Definitions used for VLAN Offload functionality */
783 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */
784 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */
785 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */
786 #define ETH_QINQ_STRIP_OFFLOAD 0x0008 /**< QINQ Strip On/Off */
788 /* Definitions used for mask VLAN setting */
789 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */
790 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/
791 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/
792 #define ETH_QINQ_STRIP_MASK 0x0008 /**< QINQ Strip setting mask */
793 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/
795 /* Definitions used for receive MAC address */
796 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */
798 /* Definitions used for unicast hash */
799 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */
801 /* Definitions used for VMDQ pool rx mode setting */
802 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */
803 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */
804 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */
805 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */
806 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */
808 /** Maximum nb. of vlan per mirror rule */
809 #define ETH_MIRROR_MAX_VLANS 64
811 #define ETH_MIRROR_VIRTUAL_POOL_UP 0x01 /**< Virtual Pool uplink Mirroring. */
812 #define ETH_MIRROR_UPLINK_PORT 0x02 /**< Uplink Port Mirroring. */
813 #define ETH_MIRROR_DOWNLINK_PORT 0x04 /**< Downlink Port Mirroring. */
814 #define ETH_MIRROR_VLAN 0x08 /**< VLAN Mirroring. */
815 #define ETH_MIRROR_VIRTUAL_POOL_DOWN 0x10 /**< Virtual Pool downlink Mirroring. */
818 * A structure used to configure VLAN traffic mirror of an Ethernet port.
820 struct rte_eth_vlan_mirror {
821 uint64_t vlan_mask; /**< mask for valid VLAN ID. */
822 /** VLAN ID list for vlan mirroring. */
823 uint16_t vlan_id[ETH_MIRROR_MAX_VLANS];
827 * A structure used to configure traffic mirror of an Ethernet port.
829 struct rte_eth_mirror_conf {
830 uint8_t rule_type; /**< Mirroring rule type */
831 uint8_t dst_pool; /**< Destination pool for this mirror rule. */
832 uint64_t pool_mask; /**< Bitmap of pool for pool mirroring */
833 /** VLAN ID setting for VLAN mirroring. */
834 struct rte_eth_vlan_mirror vlan;
838 * A structure used to configure 64 entries of Redirection Table of the
839 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure
840 * more than 64 entries supported by hardware, an array of this structure
843 struct rte_eth_rss_reta_entry64 {
845 /**< Mask bits indicate which entries need to be updated/queried. */
846 uint16_t reta[RTE_RETA_GROUP_SIZE];
847 /**< Group of 64 redirection table entries. */
851 * This enum indicates the possible number of traffic classes
852 * in DCB configurations
854 enum rte_eth_nb_tcs {
855 ETH_4_TCS = 4, /**< 4 TCs with DCB. */
856 ETH_8_TCS = 8 /**< 8 TCs with DCB. */
860 * This enum indicates the possible number of queue pools
861 * in VMDQ configurations.
863 enum rte_eth_nb_pools {
864 ETH_8_POOLS = 8, /**< 8 VMDq pools. */
865 ETH_16_POOLS = 16, /**< 16 VMDq pools. */
866 ETH_32_POOLS = 32, /**< 32 VMDq pools. */
867 ETH_64_POOLS = 64 /**< 64 VMDq pools. */
870 /* This structure may be extended in future. */
871 struct rte_eth_dcb_rx_conf {
872 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */
873 /** Traffic class each UP mapped to. */
874 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
877 struct rte_eth_vmdq_dcb_tx_conf {
878 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */
879 /** Traffic class each UP mapped to. */
880 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
883 struct rte_eth_dcb_tx_conf {
884 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */
885 /** Traffic class each UP mapped to. */
886 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
889 struct rte_eth_vmdq_tx_conf {
890 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */
894 * A structure used to configure the VMDQ+DCB feature
895 * of an Ethernet port.
897 * Using this feature, packets are routed to a pool of queues, based
898 * on the vlan id in the vlan tag, and then to a specific queue within
899 * that pool, using the user priority vlan tag field.
901 * A default pool may be used, if desired, to route all traffic which
902 * does not match the vlan filter rules.
904 struct rte_eth_vmdq_dcb_conf {
905 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */
906 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
907 uint8_t default_pool; /**< The default pool, if applicable */
908 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
910 uint16_t vlan_id; /**< The vlan id of the received frame */
911 uint64_t pools; /**< Bitmask of pools for packet rx */
912 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
913 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
914 /**< Selects a queue in a pool */
918 * A structure used to configure the VMDQ feature of an Ethernet port when
919 * not combined with the DCB feature.
921 * Using this feature, packets are routed to a pool of queues. By default,
922 * the pool selection is based on the MAC address, the vlan id in the
923 * vlan tag as specified in the pool_map array.
924 * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool
925 * selection using only the MAC address. MAC address to pool mapping is done
926 * using the rte_eth_dev_mac_addr_add function, with the pool parameter
927 * corresponding to the pool id.
929 * Queue selection within the selected pool will be done using RSS when
930 * it is enabled or revert to the first queue of the pool if not.
932 * A default pool may be used, if desired, to route all traffic which
933 * does not match the vlan filter rules or any pool MAC address.
935 struct rte_eth_vmdq_rx_conf {
936 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */
937 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
938 uint8_t default_pool; /**< The default pool, if applicable */
939 uint8_t enable_loop_back; /**< Enable VT loop back */
940 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
941 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */
943 uint16_t vlan_id; /**< The vlan id of the received frame */
944 uint64_t pools; /**< Bitmask of pools for packet rx */
945 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
949 * A structure used to configure the TX features of an Ethernet port.
951 struct rte_eth_txmode {
952 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */
954 * Per-port Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
955 * Only offloads set on tx_offload_capa field on rte_eth_dev_info
956 * structure are allowed to be set.
962 uint8_t hw_vlan_reject_tagged : 1,
963 /**< If set, reject sending out tagged pkts */
964 hw_vlan_reject_untagged : 1,
965 /**< If set, reject sending out untagged pkts */
966 hw_vlan_insert_pvid : 1;
967 /**< If set, enable port based VLAN insertion */
969 uint64_t reserved_64s[2]; /**< Reserved for future fields */
970 void *reserved_ptrs[2]; /**< Reserved for future fields */
975 * @b EXPERIMENTAL: this structure may change without prior notice.
977 * A structure used to configure an Rx packet segment to split.
979 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field,
980 * the PMD will split the received packets into multiple segments
981 * according to the specification in the description array:
983 * - The first network buffer will be allocated from the memory pool,
984 * specified in the first array element, the second buffer, from the
985 * pool in the second element, and so on.
987 * - The offsets from the segment description elements specify
988 * the data offset from the buffer beginning except the first mbuf.
989 * The first segment offset is added with RTE_PKTMBUF_HEADROOM.
991 * - The lengths in the elements define the maximal data amount
992 * being received to each segment. The receiving starts with filling
993 * up the first mbuf data buffer up to specified length. If the
994 * there are data remaining (packet is longer than buffer in the first
995 * mbuf) the following data will be pushed to the next segment
996 * up to its own length, and so on.
998 * - If the length in the segment description element is zero
999 * the actual buffer size will be deduced from the appropriate
1000 * memory pool properties.
1002 * - If there is not enough elements to describe the buffer for entire
1003 * packet of maximal length the following parameters will be used
1004 * for the all remaining segments:
1005 * - pool from the last valid element
1006 * - the buffer size from this pool
1009 struct rte_eth_rxseg_split {
1010 struct rte_mempool *mp; /**< Memory pool to allocate segment from. */
1011 uint16_t length; /**< Segment data length, configures split point. */
1012 uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */
1013 uint32_t reserved; /**< Reserved field. */
1018 * @b EXPERIMENTAL: this structure may change without prior notice.
1020 * A common structure used to describe Rx packet segment properties.
1022 union rte_eth_rxseg {
1023 /* The settings for buffer split offload. */
1024 struct rte_eth_rxseg_split split;
1025 /* The other features settings should be added here. */
1029 * A structure used to configure an RX ring of an Ethernet port.
1031 struct rte_eth_rxconf {
1032 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */
1033 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
1034 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
1035 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
1036 uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */
1038 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
1039 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
1040 * fields on rte_eth_dev_info structure are allowed to be set.
1044 * Points to the array of segment descriptions for an entire packet.
1045 * Array elements are properties for consecutive Rx segments.
1047 * The supported capabilities of receiving segmentation is reported
1048 * in rte_eth_dev_info.rx_seg_capa field.
1050 union rte_eth_rxseg *rx_seg;
1052 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1053 void *reserved_ptrs[2]; /**< Reserved for future fields */
1057 * A structure used to configure a TX ring of an Ethernet port.
1059 struct rte_eth_txconf {
1060 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */
1061 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */
1062 uint16_t tx_free_thresh; /**< Start freeing TX buffers if there are
1063 less free descriptors than this value. */
1065 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
1067 * Per-queue Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
1068 * Only offloads set on tx_queue_offload_capa or tx_offload_capa
1069 * fields on rte_eth_dev_info structure are allowed to be set.
1073 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1074 void *reserved_ptrs[2]; /**< Reserved for future fields */
1079 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1081 * A structure used to return the hairpin capabilities that are supported.
1083 struct rte_eth_hairpin_cap {
1084 /** The max number of hairpin queues (different bindings). */
1085 uint16_t max_nb_queues;
1086 /** Max number of Rx queues to be connected to one Tx queue. */
1087 uint16_t max_rx_2_tx;
1088 /** Max number of Tx queues to be connected to one Rx queue. */
1089 uint16_t max_tx_2_rx;
1090 uint16_t max_nb_desc; /**< The max num of descriptors. */
1093 #define RTE_ETH_MAX_HAIRPIN_PEERS 32
1097 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1099 * A structure used to hold hairpin peer data.
1101 struct rte_eth_hairpin_peer {
1102 uint16_t port; /**< Peer port. */
1103 uint16_t queue; /**< Peer queue. */
1108 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1110 * A structure used to configure hairpin binding.
1112 struct rte_eth_hairpin_conf {
1113 uint32_t peer_count:16; /**< The number of peers. */
1116 * Explicit Tx flow rule mode.
1117 * One hairpin pair of queues should have the same attribute.
1119 * - When set, the user should be responsible for inserting the hairpin
1120 * Tx part flows and removing them.
1121 * - When clear, the PMD will try to handle the Tx part of the flows,
1122 * e.g., by splitting one flow into two parts.
1124 uint32_t tx_explicit:1;
1127 * Manually bind hairpin queues.
1128 * One hairpin pair of queues should have the same attribute.
1130 * - When set, to enable hairpin, the user should call the hairpin bind
1131 * function after all the queues are set up properly and the ports are
1132 * started. Also, the hairpin unbind function should be called
1133 * accordingly before stopping a port that with hairpin configured.
1134 * - When clear, the PMD will try to enable the hairpin with the queues
1135 * configured automatically during port start.
1137 uint32_t manual_bind:1;
1138 uint32_t reserved:14; /**< Reserved bits. */
1139 struct rte_eth_hairpin_peer peers[RTE_ETH_MAX_HAIRPIN_PEERS];
1143 * A structure contains information about HW descriptor ring limitations.
1145 struct rte_eth_desc_lim {
1146 uint16_t nb_max; /**< Max allowed number of descriptors. */
1147 uint16_t nb_min; /**< Min allowed number of descriptors. */
1148 uint16_t nb_align; /**< Number of descriptors should be aligned to. */
1151 * Max allowed number of segments per whole packet.
1153 * - For TSO packet this is the total number of data descriptors allowed
1156 * @see nb_mtu_seg_max
1158 uint16_t nb_seg_max;
1161 * Max number of segments per one MTU.
1163 * - For non-TSO packet, this is the maximum allowed number of segments
1164 * in a single transmit packet.
1166 * - For TSO packet each segment within the TSO may span up to this
1171 uint16_t nb_mtu_seg_max;
1175 * This enum indicates the flow control mode
1177 enum rte_eth_fc_mode {
1178 RTE_FC_NONE = 0, /**< Disable flow control. */
1179 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */
1180 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */
1181 RTE_FC_FULL /**< Enable flow control on both side. */
1185 * A structure used to configure Ethernet flow control parameter.
1186 * These parameters will be configured into the register of the NIC.
1187 * Please refer to the corresponding data sheet for proper value.
1189 struct rte_eth_fc_conf {
1190 uint32_t high_water; /**< High threshold value to trigger XOFF */
1191 uint32_t low_water; /**< Low threshold value to trigger XON */
1192 uint16_t pause_time; /**< Pause quota in the Pause frame */
1193 uint16_t send_xon; /**< Is XON frame need be sent */
1194 enum rte_eth_fc_mode mode; /**< Link flow control mode */
1195 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */
1196 uint8_t autoneg; /**< Use Pause autoneg */
1200 * A structure used to configure Ethernet priority flow control parameter.
1201 * These parameters will be configured into the register of the NIC.
1202 * Please refer to the corresponding data sheet for proper value.
1204 struct rte_eth_pfc_conf {
1205 struct rte_eth_fc_conf fc; /**< General flow control parameter. */
1206 uint8_t priority; /**< VLAN User Priority. */
1212 enum rte_eth_tunnel_type {
1213 RTE_TUNNEL_TYPE_NONE = 0,
1214 RTE_TUNNEL_TYPE_VXLAN,
1215 RTE_TUNNEL_TYPE_GENEVE,
1216 RTE_TUNNEL_TYPE_TEREDO,
1217 RTE_TUNNEL_TYPE_NVGRE,
1218 RTE_TUNNEL_TYPE_IP_IN_GRE,
1219 RTE_L2_TUNNEL_TYPE_E_TAG,
1220 RTE_TUNNEL_TYPE_VXLAN_GPE,
1221 RTE_TUNNEL_TYPE_MAX,
1224 /* Deprecated API file for rte_eth_dev_filter_* functions */
1225 #include "rte_eth_ctrl.h"
1228 * Memory space that can be configured to store Flow Director filters
1229 * in the board memory.
1231 enum rte_fdir_pballoc_type {
1232 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */
1233 RTE_FDIR_PBALLOC_128K, /**< 128k. */
1234 RTE_FDIR_PBALLOC_256K, /**< 256k. */
1238 * Select report mode of FDIR hash information in RX descriptors.
1240 enum rte_fdir_status_mode {
1241 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */
1242 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */
1243 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */
1247 * A structure used to configure the Flow Director (FDIR) feature
1248 * of an Ethernet port.
1250 * If mode is RTE_FDIR_MODE_NONE, the pballoc value is ignored.
1252 struct rte_fdir_conf {
1253 enum rte_fdir_mode mode; /**< Flow Director mode. */
1254 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */
1255 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */
1256 /** RX queue of packets matching a "drop" filter in perfect mode. */
1258 struct rte_eth_fdir_masks mask;
1259 struct rte_eth_fdir_flex_conf flex_conf;
1260 /**< Flex payload configuration. */
1264 * UDP tunneling configuration.
1265 * Used to config the UDP port for a type of tunnel.
1266 * NICs need the UDP port to identify the tunnel type.
1267 * Normally a type of tunnel has a default UDP port, this structure can be used
1268 * in case if the users want to change or support more UDP port.
1270 struct rte_eth_udp_tunnel {
1271 uint16_t udp_port; /**< UDP port used for the tunnel. */
1272 uint8_t prot_type; /**< Tunnel type. Defined in rte_eth_tunnel_type. */
1276 * A structure used to enable/disable specific device interrupts.
1278 struct rte_intr_conf {
1279 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */
1281 /** enable/disable rxq interrupt. 0 (default) - disable, 1 enable */
1283 /** enable/disable rmv interrupt. 0 (default) - disable, 1 enable */
1288 * A structure used to configure an Ethernet port.
1289 * Depending upon the RX multi-queue mode, extra advanced
1290 * configuration settings may be needed.
1292 struct rte_eth_conf {
1293 uint32_t link_speeds; /**< bitmap of ETH_LINK_SPEED_XXX of speeds to be
1294 used. ETH_LINK_SPEED_FIXED disables link
1295 autonegotiation, and a unique speed shall be
1296 set. Otherwise, the bitmap defines the set of
1297 speeds to be advertised. If the special value
1298 ETH_LINK_SPEED_AUTONEG (0) is used, all speeds
1299 supported are advertised. */
1300 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */
1301 struct rte_eth_txmode txmode; /**< Port TX configuration. */
1302 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value
1303 is 0, meaning the loopback mode is disabled.
1304 Read the datasheet of given ethernet controller
1305 for details. The possible values of this field
1306 are defined in implementation of each driver. */
1308 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */
1309 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf;
1310 /**< Port vmdq+dcb configuration. */
1311 struct rte_eth_dcb_rx_conf dcb_rx_conf;
1312 /**< Port dcb RX configuration. */
1313 struct rte_eth_vmdq_rx_conf vmdq_rx_conf;
1314 /**< Port vmdq RX configuration. */
1315 } rx_adv_conf; /**< Port RX filtering configuration. */
1317 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf;
1318 /**< Port vmdq+dcb TX configuration. */
1319 struct rte_eth_dcb_tx_conf dcb_tx_conf;
1320 /**< Port dcb TX configuration. */
1321 struct rte_eth_vmdq_tx_conf vmdq_tx_conf;
1322 /**< Port vmdq TX configuration. */
1323 } tx_adv_conf; /**< Port TX DCB configuration (union). */
1324 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC
1325 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */
1326 uint32_t dcb_capability_en;
1327 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. DEPRECATED */
1328 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */
1332 * RX offload capabilities of a device.
1334 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001
1335 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002
1336 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004
1337 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008
1338 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010
1339 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020
1340 #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040
1341 #define DEV_RX_OFFLOAD_MACSEC_STRIP 0x00000080
1342 #define DEV_RX_OFFLOAD_HEADER_SPLIT 0x00000100
1343 #define DEV_RX_OFFLOAD_VLAN_FILTER 0x00000200
1344 #define DEV_RX_OFFLOAD_VLAN_EXTEND 0x00000400
1345 #define DEV_RX_OFFLOAD_JUMBO_FRAME 0x00000800
1346 #define DEV_RX_OFFLOAD_SCATTER 0x00002000
1348 * Timestamp is set by the driver in RTE_MBUF_DYNFIELD_TIMESTAMP_NAME
1349 * and RTE_MBUF_DYNFLAG_RX_TIMESTAMP_NAME is set in ol_flags.
1350 * The mbuf field and flag are registered when the offload is configured.
1352 #define DEV_RX_OFFLOAD_TIMESTAMP 0x00004000
1353 #define DEV_RX_OFFLOAD_SECURITY 0x00008000
1354 #define DEV_RX_OFFLOAD_KEEP_CRC 0x00010000
1355 #define DEV_RX_OFFLOAD_SCTP_CKSUM 0x00020000
1356 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM 0x00040000
1357 #define DEV_RX_OFFLOAD_RSS_HASH 0x00080000
1358 #define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000
1360 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
1361 DEV_RX_OFFLOAD_UDP_CKSUM | \
1362 DEV_RX_OFFLOAD_TCP_CKSUM)
1363 #define DEV_RX_OFFLOAD_VLAN (DEV_RX_OFFLOAD_VLAN_STRIP | \
1364 DEV_RX_OFFLOAD_VLAN_FILTER | \
1365 DEV_RX_OFFLOAD_VLAN_EXTEND | \
1366 DEV_RX_OFFLOAD_QINQ_STRIP)
1369 * If new Rx offload capabilities are defined, they also must be
1370 * mentioned in rte_rx_offload_names in rte_ethdev.c file.
1374 * TX offload capabilities of a device.
1376 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001
1377 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002
1378 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004
1379 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008
1380 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010
1381 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020
1382 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040
1383 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */
1384 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100
1385 #define DEV_TX_OFFLOAD_VXLAN_TNL_TSO 0x00000200 /**< Used for tunneling packet. */
1386 #define DEV_TX_OFFLOAD_GRE_TNL_TSO 0x00000400 /**< Used for tunneling packet. */
1387 #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */
1388 #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */
1389 #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000
1390 #define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000
1391 /**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same
1392 * tx queue without SW lock.
1394 #define DEV_TX_OFFLOAD_MULTI_SEGS 0x00008000
1395 /**< Device supports multi segment send. */
1396 #define DEV_TX_OFFLOAD_MBUF_FAST_FREE 0x00010000
1397 /**< Device supports optimization for fast release of mbufs.
1398 * When set application must guarantee that per-queue all mbufs comes from
1399 * the same mempool and has refcnt = 1.
1401 #define DEV_TX_OFFLOAD_SECURITY 0x00020000
1403 * Device supports generic UDP tunneled packet TSO.
1404 * Application must set PKT_TX_TUNNEL_UDP and other mbuf fields required
1407 #define DEV_TX_OFFLOAD_UDP_TNL_TSO 0x00040000
1409 * Device supports generic IP tunneled packet TSO.
1410 * Application must set PKT_TX_TUNNEL_IP and other mbuf fields required
1413 #define DEV_TX_OFFLOAD_IP_TNL_TSO 0x00080000
1414 /** Device supports outer UDP checksum */
1415 #define DEV_TX_OFFLOAD_OUTER_UDP_CKSUM 0x00100000
1417 * Device sends on time read from RTE_MBUF_DYNFIELD_TIMESTAMP_NAME
1418 * if RTE_MBUF_DYNFLAG_TX_TIMESTAMP_NAME is set in ol_flags.
1419 * The mbuf field and flag are registered when the offload is configured.
1421 #define DEV_TX_OFFLOAD_SEND_ON_TIMESTAMP 0x00200000
1424 #define RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP 0x00000001
1425 /**< Device supports Rx queue setup after device started*/
1426 #define RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP 0x00000002
1427 /**< Device supports Tx queue setup after device started*/
1430 * If new Tx offload capabilities are defined, they also must be
1431 * mentioned in rte_tx_offload_names in rte_ethdev.c file.
1435 * Fallback default preferred Rx/Tx port parameters.
1436 * These are used if an application requests default parameters
1437 * but the PMD does not provide preferred values.
1439 #define RTE_ETH_DEV_FALLBACK_RX_RINGSIZE 512
1440 #define RTE_ETH_DEV_FALLBACK_TX_RINGSIZE 512
1441 #define RTE_ETH_DEV_FALLBACK_RX_NBQUEUES 1
1442 #define RTE_ETH_DEV_FALLBACK_TX_NBQUEUES 1
1445 * Preferred Rx/Tx port parameters.
1446 * There are separate instances of this structure for transmission
1447 * and reception respectively.
1449 struct rte_eth_dev_portconf {
1450 uint16_t burst_size; /**< Device-preferred burst size */
1451 uint16_t ring_size; /**< Device-preferred size of queue rings */
1452 uint16_t nb_queues; /**< Device-preferred number of queues */
1456 * Default values for switch domain id when ethdev does not support switch
1457 * domain definitions.
1459 #define RTE_ETH_DEV_SWITCH_DOMAIN_ID_INVALID (UINT16_MAX)
1462 * Ethernet device associated switch information
1464 struct rte_eth_switch_info {
1465 const char *name; /**< switch name */
1466 uint16_t domain_id; /**< switch domain id */
1469 * mapping to the devices physical switch port as enumerated from the
1470 * perspective of the embedded interconnect/switch. For SR-IOV enabled
1471 * device this may correspond to the VF_ID of each virtual function,
1472 * but each driver should explicitly define the mapping of switch
1473 * port identifier to that physical interconnect/switch
1479 * @b EXPERIMENTAL: this structure may change without prior notice.
1481 * Ethernet device Rx buffer segmentation capabilities.
1483 struct rte_eth_rxseg_capa {
1485 uint32_t multi_pools:1; /**< Supports receiving to multiple pools.*/
1486 uint32_t offset_allowed:1; /**< Supports buffer offsets. */
1487 uint32_t offset_align_log2:4; /**< Required offset alignment. */
1488 uint16_t max_nseg; /**< Maximum amount of segments to split. */
1489 uint16_t reserved; /**< Reserved field. */
1493 * Ethernet device information
1497 * A structure used to retrieve the contextual information of
1498 * an Ethernet device, such as the controlling driver of the
1501 struct rte_eth_dev_info {
1502 struct rte_device *device; /** Generic device information */
1503 const char *driver_name; /**< Device Driver name. */
1504 unsigned int if_index; /**< Index to bound host interface, or 0 if none.
1505 Use if_indextoname() to translate into an interface name. */
1506 uint16_t min_mtu; /**< Minimum MTU allowed */
1507 uint16_t max_mtu; /**< Maximum MTU allowed */
1508 const uint32_t *dev_flags; /**< Device flags */
1509 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */
1510 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */
1511 /** Maximum configurable size of LRO aggregated packet. */
1512 uint32_t max_lro_pkt_size;
1513 uint16_t max_rx_queues; /**< Maximum number of RX queues. */
1514 uint16_t max_tx_queues; /**< Maximum number of TX queues. */
1515 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */
1516 uint32_t max_hash_mac_addrs;
1517 /** Maximum number of hash MAC addresses for MTA and UTA. */
1518 uint16_t max_vfs; /**< Maximum number of VFs. */
1519 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
1520 struct rte_eth_rxseg_capa rx_seg_capa; /**< Segmentation capability.*/
1521 uint64_t rx_offload_capa;
1522 /**< All RX offload capabilities including all per-queue ones */
1523 uint64_t tx_offload_capa;
1524 /**< All TX offload capabilities including all per-queue ones */
1525 uint64_t rx_queue_offload_capa;
1526 /**< Device per-queue RX offload capabilities. */
1527 uint64_t tx_queue_offload_capa;
1528 /**< Device per-queue TX offload capabilities. */
1530 /**< Device redirection table size, the total number of entries. */
1531 uint8_t hash_key_size; /**< Hash key size in bytes */
1532 /** Bit mask of RSS offloads, the bit offset also means flow type */
1533 uint64_t flow_type_rss_offloads;
1534 struct rte_eth_rxconf default_rxconf; /**< Default RX configuration */
1535 struct rte_eth_txconf default_txconf; /**< Default TX configuration */
1536 uint16_t vmdq_queue_base; /**< First queue ID for VMDQ pools. */
1537 uint16_t vmdq_queue_num; /**< Queue number for VMDQ pools. */
1538 uint16_t vmdq_pool_base; /**< First ID of VMDQ pools. */
1539 struct rte_eth_desc_lim rx_desc_lim; /**< RX descriptors limits */
1540 struct rte_eth_desc_lim tx_desc_lim; /**< TX descriptors limits */
1541 uint32_t speed_capa; /**< Supported speeds bitmap (ETH_LINK_SPEED_). */
1542 /** Configured number of rx/tx queues */
1543 uint16_t nb_rx_queues; /**< Number of RX queues. */
1544 uint16_t nb_tx_queues; /**< Number of TX queues. */
1545 /** Rx parameter recommendations */
1546 struct rte_eth_dev_portconf default_rxportconf;
1547 /** Tx parameter recommendations */
1548 struct rte_eth_dev_portconf default_txportconf;
1549 /** Generic device capabilities (RTE_ETH_DEV_CAPA_). */
1552 * Switching information for ports on a device with a
1553 * embedded managed interconnect/switch.
1555 struct rte_eth_switch_info switch_info;
1557 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1558 void *reserved_ptrs[2]; /**< Reserved for future fields */
1562 * Ethernet device RX queue information structure.
1563 * Used to retrieve information about configured queue.
1565 struct rte_eth_rxq_info {
1566 struct rte_mempool *mp; /**< mempool used by that queue. */
1567 struct rte_eth_rxconf conf; /**< queue config parameters. */
1568 uint8_t scattered_rx; /**< scattered packets RX supported. */
1569 uint16_t nb_desc; /**< configured number of RXDs. */
1570 uint16_t rx_buf_size; /**< hardware receive buffer size. */
1571 } __rte_cache_min_aligned;
1574 * Ethernet device TX queue information structure.
1575 * Used to retrieve information about configured queue.
1577 struct rte_eth_txq_info {
1578 struct rte_eth_txconf conf; /**< queue config parameters. */
1579 uint16_t nb_desc; /**< configured number of TXDs. */
1580 } __rte_cache_min_aligned;
1582 /* Generic Burst mode flag definition, values can be ORed. */
1585 * If the queues have different burst mode description, this bit will be set
1586 * by PMD, then the application can iterate to retrieve burst description for
1589 #define RTE_ETH_BURST_FLAG_PER_QUEUE (1ULL << 0)
1592 * Ethernet device RX/TX queue packet burst mode information structure.
1593 * Used to retrieve information about packet burst mode setting.
1595 struct rte_eth_burst_mode {
1596 uint64_t flags; /**< The ORed values of RTE_ETH_BURST_FLAG_xxx */
1598 #define RTE_ETH_BURST_MODE_INFO_SIZE 1024 /**< Maximum size for information */
1599 char info[RTE_ETH_BURST_MODE_INFO_SIZE]; /**< burst mode information */
1602 /** Maximum name length for extended statistics counters */
1603 #define RTE_ETH_XSTATS_NAME_SIZE 64
1606 * An Ethernet device extended statistic structure
1608 * This structure is used by rte_eth_xstats_get() to provide
1609 * statistics that are not provided in the generic *rte_eth_stats*
1611 * It maps a name id, corresponding to an index in the array returned
1612 * by rte_eth_xstats_get_names(), to a statistic value.
1614 struct rte_eth_xstat {
1615 uint64_t id; /**< The index in xstats name array. */
1616 uint64_t value; /**< The statistic counter value. */
1620 * A name element for extended statistics.
1622 * An array of this structure is returned by rte_eth_xstats_get_names().
1623 * It lists the names of extended statistics for a PMD. The *rte_eth_xstat*
1624 * structure references these names by their array index.
1626 * The xstats should follow a common naming scheme.
1627 * Some names are standardized in rte_stats_strings.
1629 * - rx_missed_errors
1631 * - tx_size_128_to_255_packets
1633 struct rte_eth_xstat_name {
1634 char name[RTE_ETH_XSTATS_NAME_SIZE]; /**< The statistic name. */
1637 #define ETH_DCB_NUM_TCS 8
1638 #define ETH_MAX_VMDQ_POOL 64
1641 * A structure used to get the information of queue and
1642 * TC mapping on both TX and RX paths.
1644 struct rte_eth_dcb_tc_queue_mapping {
1645 /** rx queues assigned to tc per Pool */
1649 } tc_rxq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1650 /** rx queues assigned to tc per Pool */
1654 } tc_txq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1658 * A structure used to get the information of DCB.
1659 * It includes TC UP mapping and queue TC mapping.
1661 struct rte_eth_dcb_info {
1662 uint8_t nb_tcs; /**< number of TCs */
1663 uint8_t prio_tc[ETH_DCB_NUM_USER_PRIORITIES]; /**< Priority to tc */
1664 uint8_t tc_bws[ETH_DCB_NUM_TCS]; /**< TX BW percentage for each TC */
1665 /** rx queues assigned to tc */
1666 struct rte_eth_dcb_tc_queue_mapping tc_queue;
1670 * This enum indicates the possible Forward Error Correction (FEC) modes
1671 * of an ethdev port.
1673 enum rte_eth_fec_mode {
1674 RTE_ETH_FEC_NOFEC = 0, /**< FEC is off */
1675 RTE_ETH_FEC_AUTO, /**< FEC autonegotiation modes */
1676 RTE_ETH_FEC_BASER, /**< FEC using common algorithm */
1677 RTE_ETH_FEC_RS, /**< FEC using RS algorithm */
1680 /* Translate from FEC mode to FEC capa */
1681 #define RTE_ETH_FEC_MODE_TO_CAPA(x) (1U << (x))
1683 /* This macro indicates FEC capa mask */
1684 #define RTE_ETH_FEC_MODE_CAPA_MASK(x) (1U << (RTE_ETH_FEC_ ## x))
1686 /* A structure used to get capabilities per link speed */
1687 struct rte_eth_fec_capa {
1688 uint32_t speed; /**< Link speed (see ETH_SPEED_NUM_*) */
1689 uint32_t capa; /**< FEC capabilities bitmask */
1692 #define RTE_ETH_ALL RTE_MAX_ETHPORTS
1694 /* Macros to check for valid port */
1695 #define RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, retval) do { \
1696 if (!rte_eth_dev_is_valid_port(port_id)) { \
1697 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1702 #define RTE_ETH_VALID_PORTID_OR_RET(port_id) do { \
1703 if (!rte_eth_dev_is_valid_port(port_id)) { \
1704 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1710 * l2 tunnel configuration.
1713 /**< l2 tunnel enable mask */
1714 #define ETH_L2_TUNNEL_ENABLE_MASK 0x00000001
1715 /**< l2 tunnel insertion mask */
1716 #define ETH_L2_TUNNEL_INSERTION_MASK 0x00000002
1717 /**< l2 tunnel stripping mask */
1718 #define ETH_L2_TUNNEL_STRIPPING_MASK 0x00000004
1719 /**< l2 tunnel forwarding mask */
1720 #define ETH_L2_TUNNEL_FORWARDING_MASK 0x00000008
1723 * Function type used for RX packet processing packet callbacks.
1725 * The callback function is called on RX with a burst of packets that have
1726 * been received on the given port and queue.
1729 * The Ethernet port on which RX is being performed.
1731 * The queue on the Ethernet port which is being used to receive the packets.
1733 * The burst of packets that have just been received.
1735 * The number of packets in the burst pointed to by "pkts".
1737 * The max number of packets that can be stored in the "pkts" array.
1739 * The arbitrary user parameter passed in by the application when the callback
1740 * was originally configured.
1742 * The number of packets returned to the user.
1744 typedef uint16_t (*rte_rx_callback_fn)(uint16_t port_id, uint16_t queue,
1745 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts,
1749 * Function type used for TX packet processing packet callbacks.
1751 * The callback function is called on TX with a burst of packets immediately
1752 * before the packets are put onto the hardware queue for transmission.
1755 * The Ethernet port on which TX is being performed.
1757 * The queue on the Ethernet port which is being used to transmit the packets.
1759 * The burst of packets that are about to be transmitted.
1761 * The number of packets in the burst pointed to by "pkts".
1763 * The arbitrary user parameter passed in by the application when the callback
1764 * was originally configured.
1766 * The number of packets to be written to the NIC.
1768 typedef uint16_t (*rte_tx_callback_fn)(uint16_t port_id, uint16_t queue,
1769 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param);
1772 * Possible states of an ethdev port.
1774 enum rte_eth_dev_state {
1775 /** Device is unused before being probed. */
1776 RTE_ETH_DEV_UNUSED = 0,
1777 /** Device is attached when allocated in probing. */
1778 RTE_ETH_DEV_ATTACHED,
1779 /** Device is in removed state when plug-out is detected. */
1780 RTE_ETH_DEV_REMOVED,
1783 struct rte_eth_dev_sriov {
1784 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */
1785 uint8_t nb_q_per_pool; /**< rx queue number per pool */
1786 uint16_t def_vmdq_idx; /**< Default pool num used for PF */
1787 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */
1789 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
1791 #define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN
1793 #define RTE_ETH_DEV_NO_OWNER 0
1795 #define RTE_ETH_MAX_OWNER_NAME_LEN 64
1797 struct rte_eth_dev_owner {
1798 uint64_t id; /**< The owner unique identifier. */
1799 char name[RTE_ETH_MAX_OWNER_NAME_LEN]; /**< The owner name. */
1802 /** PMD supports thread-safe flow operations */
1803 #define RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE 0x0001
1804 /** Device supports link state interrupt */
1805 #define RTE_ETH_DEV_INTR_LSC 0x0002
1806 /** Device is a bonded slave */
1807 #define RTE_ETH_DEV_BONDED_SLAVE 0x0004
1808 /** Device supports device removal interrupt */
1809 #define RTE_ETH_DEV_INTR_RMV 0x0008
1810 /** Device is port representor */
1811 #define RTE_ETH_DEV_REPRESENTOR 0x0010
1812 /** Device does not support MAC change after started */
1813 #define RTE_ETH_DEV_NOLIVE_MAC_ADDR 0x0020
1815 * Queue xstats filled automatically by ethdev layer.
1816 * PMDs filling the queue xstats themselves should not set this flag
1818 #define RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS 0x0040
1821 * Iterates over valid ethdev ports owned by a specific owner.
1824 * The id of the next possible valid owned port.
1826 * The owner identifier.
1827 * RTE_ETH_DEV_NO_OWNER means iterate over all valid ownerless ports.
1829 * Next valid port id owned by owner_id, RTE_MAX_ETHPORTS if there is none.
1831 uint64_t rte_eth_find_next_owned_by(uint16_t port_id,
1832 const uint64_t owner_id);
1835 * Macro to iterate over all enabled ethdev ports owned by a specific owner.
1837 #define RTE_ETH_FOREACH_DEV_OWNED_BY(p, o) \
1838 for (p = rte_eth_find_next_owned_by(0, o); \
1839 (unsigned int)p < (unsigned int)RTE_MAX_ETHPORTS; \
1840 p = rte_eth_find_next_owned_by(p + 1, o))
1843 * Iterates over valid ethdev ports.
1846 * The id of the next possible valid port.
1848 * Next valid port id, RTE_MAX_ETHPORTS if there is none.
1850 uint16_t rte_eth_find_next(uint16_t port_id);
1853 * Macro to iterate over all enabled and ownerless ethdev ports.
1855 #define RTE_ETH_FOREACH_DEV(p) \
1856 RTE_ETH_FOREACH_DEV_OWNED_BY(p, RTE_ETH_DEV_NO_OWNER)
1860 * @b EXPERIMENTAL: this API may change without prior notice.
1862 * Iterates over ethdev ports of a specified device.
1864 * @param port_id_start
1865 * The id of the next possible valid port.
1867 * The generic device behind the ports to iterate.
1869 * Next port id of the device, possibly port_id_start,
1870 * RTE_MAX_ETHPORTS if there is none.
1874 rte_eth_find_next_of(uint16_t port_id_start,
1875 const struct rte_device *parent);
1878 * Macro to iterate over all ethdev ports of a specified device.
1881 * The id of the matching port being iterated.
1883 * The rte_device pointer matching the iterated ports.
1885 #define RTE_ETH_FOREACH_DEV_OF(port_id, parent) \
1886 for (port_id = rte_eth_find_next_of(0, parent); \
1887 port_id < RTE_MAX_ETHPORTS; \
1888 port_id = rte_eth_find_next_of(port_id + 1, parent))
1892 * @b EXPERIMENTAL: this API may change without prior notice.
1894 * Iterates over sibling ethdev ports (i.e. sharing the same rte_device).
1896 * @param port_id_start
1897 * The id of the next possible valid sibling port.
1898 * @param ref_port_id
1899 * The id of a reference port to compare rte_device with.
1901 * Next sibling port id, possibly port_id_start or ref_port_id itself,
1902 * RTE_MAX_ETHPORTS if there is none.
1906 rte_eth_find_next_sibling(uint16_t port_id_start, uint16_t ref_port_id);
1909 * Macro to iterate over all ethdev ports sharing the same rte_device
1910 * as the specified port.
1911 * Note: the specified reference port is part of the loop iterations.
1914 * The id of the matching port being iterated.
1915 * @param ref_port_id
1916 * The id of the port being compared.
1918 #define RTE_ETH_FOREACH_DEV_SIBLING(port_id, ref_port_id) \
1919 for (port_id = rte_eth_find_next_sibling(0, ref_port_id); \
1920 port_id < RTE_MAX_ETHPORTS; \
1921 port_id = rte_eth_find_next_sibling(port_id + 1, ref_port_id))
1925 * @b EXPERIMENTAL: this API may change without prior notice.
1927 * Get a new unique owner identifier.
1928 * An owner identifier is used to owns Ethernet devices by only one DPDK entity
1929 * to avoid multiple management of device by different entities.
1932 * Owner identifier pointer.
1934 * Negative errno value on error, 0 on success.
1937 int rte_eth_dev_owner_new(uint64_t *owner_id);
1941 * @b EXPERIMENTAL: this API may change without prior notice.
1943 * Set an Ethernet device owner.
1946 * The identifier of the port to own.
1948 * The owner pointer.
1950 * Negative errno value on error, 0 on success.
1953 int rte_eth_dev_owner_set(const uint16_t port_id,
1954 const struct rte_eth_dev_owner *owner);
1958 * @b EXPERIMENTAL: this API may change without prior notice.
1960 * Unset Ethernet device owner to make the device ownerless.
1963 * The identifier of port to make ownerless.
1965 * The owner identifier.
1967 * 0 on success, negative errno value on error.
1970 int rte_eth_dev_owner_unset(const uint16_t port_id,
1971 const uint64_t owner_id);
1975 * @b EXPERIMENTAL: this API may change without prior notice.
1977 * Remove owner from all Ethernet devices owned by a specific owner.
1980 * The owner identifier.
1982 * 0 on success, negative errno value on error.
1985 int rte_eth_dev_owner_delete(const uint64_t owner_id);
1989 * @b EXPERIMENTAL: this API may change without prior notice.
1991 * Get the owner of an Ethernet device.
1994 * The port identifier.
1996 * The owner structure pointer to fill.
1998 * 0 on success, negative errno value on error..
2001 int rte_eth_dev_owner_get(const uint16_t port_id,
2002 struct rte_eth_dev_owner *owner);
2005 * Get the number of ports which are usable for the application.
2007 * These devices must be iterated by using the macro
2008 * ``RTE_ETH_FOREACH_DEV`` or ``RTE_ETH_FOREACH_DEV_OWNED_BY``
2009 * to deal with non-contiguous ranges of devices.
2012 * The count of available Ethernet devices.
2014 uint16_t rte_eth_dev_count_avail(void);
2017 * Get the total number of ports which are allocated.
2019 * Some devices may not be available for the application.
2022 * The total count of Ethernet devices.
2024 uint16_t rte_eth_dev_count_total(void);
2027 * Convert a numerical speed in Mbps to a bitmap flag that can be used in
2028 * the bitmap link_speeds of the struct rte_eth_conf
2031 * Numerical speed value in Mbps
2033 * ETH_LINK_[HALF/FULL]_DUPLEX (only for 10/100M speeds)
2035 * 0 if the speed cannot be mapped
2037 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex);
2040 * Get DEV_RX_OFFLOAD_* flag name.
2045 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
2047 const char *rte_eth_dev_rx_offload_name(uint64_t offload);
2050 * Get DEV_TX_OFFLOAD_* flag name.
2055 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
2057 const char *rte_eth_dev_tx_offload_name(uint64_t offload);
2060 * Configure an Ethernet device.
2061 * This function must be invoked first before any other function in the
2062 * Ethernet API. This function can also be re-invoked when a device is in the
2066 * The port identifier of the Ethernet device to configure.
2067 * @param nb_rx_queue
2068 * The number of receive queues to set up for the Ethernet device.
2069 * @param nb_tx_queue
2070 * The number of transmit queues to set up for the Ethernet device.
2072 * The pointer to the configuration data to be used for the Ethernet device.
2073 * The *rte_eth_conf* structure includes:
2074 * - the hardware offload features to activate, with dedicated fields for
2075 * each statically configurable offload hardware feature provided by
2076 * Ethernet devices, such as IP checksum or VLAN tag stripping for
2078 * The Rx offload bitfield API is obsolete and will be deprecated.
2079 * Applications should set the ignore_bitfield_offloads bit on *rxmode*
2080 * structure and use offloads field to set per-port offloads instead.
2081 * - Any offloading set in eth_conf->[rt]xmode.offloads must be within
2082 * the [rt]x_offload_capa returned from rte_eth_dev_info_get().
2083 * Any type of device supported offloading set in the input argument
2084 * eth_conf->[rt]xmode.offloads to rte_eth_dev_configure() is enabled
2085 * on all queues and it can't be disabled in rte_eth_[rt]x_queue_setup()
2086 * - the Receive Side Scaling (RSS) configuration when using multiple RX
2087 * queues per port. Any RSS hash function set in eth_conf->rss_conf.rss_hf
2088 * must be within the flow_type_rss_offloads provided by drivers via
2089 * rte_eth_dev_info_get() API.
2091 * Embedding all configuration information in a single data structure
2092 * is the more flexible method that allows the addition of new features
2093 * without changing the syntax of the API.
2095 * - 0: Success, device configured.
2096 * - <0: Error code returned by the driver configuration function.
2098 int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
2099 uint16_t nb_tx_queue, const struct rte_eth_conf *eth_conf);
2103 * @b EXPERIMENTAL: this API may change without prior notice.
2105 * Check if an Ethernet device was physically removed.
2108 * The port identifier of the Ethernet device.
2110 * 1 when the Ethernet device is removed, otherwise 0.
2114 rte_eth_dev_is_removed(uint16_t port_id);
2117 * Allocate and set up a receive queue for an Ethernet device.
2119 * The function allocates a contiguous block of memory for *nb_rx_desc*
2120 * receive descriptors from a memory zone associated with *socket_id*
2121 * and initializes each receive descriptor with a network buffer allocated
2122 * from the memory pool *mb_pool*.
2125 * The port identifier of the Ethernet device.
2126 * @param rx_queue_id
2127 * The index of the receive queue to set up.
2128 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2129 * to rte_eth_dev_configure().
2131 * The number of receive descriptors to allocate for the receive ring.
2133 * The *socket_id* argument is the socket identifier in case of NUMA.
2134 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
2135 * the DMA memory allocated for the receive descriptors of the ring.
2137 * The pointer to the configuration data to be used for the receive queue.
2138 * NULL value is allowed, in which case default RX configuration
2140 * The *rx_conf* structure contains an *rx_thresh* structure with the values
2141 * of the Prefetch, Host, and Write-Back threshold registers of the receive
2143 * In addition it contains the hardware offloads features to activate using
2144 * the DEV_RX_OFFLOAD_* flags.
2145 * If an offloading set in rx_conf->offloads
2146 * hasn't been set in the input argument eth_conf->rxmode.offloads
2147 * to rte_eth_dev_configure(), it is a new added offloading, it must be
2148 * per-queue type and it is enabled for the queue.
2149 * No need to repeat any bit in rx_conf->offloads which has already been
2150 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
2151 * at port level can't be disabled at queue level.
2152 * The configuration structure also contains the pointer to the array
2153 * of the receiving buffer segment descriptions, see rx_seg and rx_nseg
2154 * fields, this extended configuration might be used by split offloads like
2155 * RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL,
2156 * the extended configuration fields must be set to NULL and zero.
2158 * The pointer to the memory pool from which to allocate *rte_mbuf* network
2159 * memory buffers to populate each descriptor of the receive ring. There are
2160 * two options to provide Rx buffer configuration:
2162 * mb_pool is not NULL, rx_conf.rx_nseg is 0.
2163 * - multiple segments description:
2164 * mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0.
2165 * Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads.
2168 * - 0: Success, receive queue correctly set up.
2169 * - -EIO: if device is removed.
2170 * - -ENODEV: if *port_id* is invalid.
2171 * - -EINVAL: The memory pool pointer is null or the size of network buffers
2172 * which can be allocated from this memory pool does not fit the various
2173 * buffer sizes allowed by the device controller.
2174 * - -ENOMEM: Unable to allocate the receive ring descriptors or to
2175 * allocate network memory buffers from the memory pool when
2176 * initializing receive descriptors.
2178 int rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
2179 uint16_t nb_rx_desc, unsigned int socket_id,
2180 const struct rte_eth_rxconf *rx_conf,
2181 struct rte_mempool *mb_pool);
2185 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2187 * Allocate and set up a hairpin receive queue for an Ethernet device.
2189 * The function set up the selected queue to be used in hairpin.
2192 * The port identifier of the Ethernet device.
2193 * @param rx_queue_id
2194 * The index of the receive queue to set up.
2195 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2196 * to rte_eth_dev_configure().
2198 * The number of receive descriptors to allocate for the receive ring.
2199 * 0 means the PMD will use default value.
2201 * The pointer to the hairpin configuration.
2204 * - (0) if successful.
2205 * - (-ENODEV) if *port_id* is invalid.
2206 * - (-ENOTSUP) if hardware doesn't support.
2207 * - (-EINVAL) if bad parameter.
2208 * - (-ENOMEM) if unable to allocate the resources.
2211 int rte_eth_rx_hairpin_queue_setup
2212 (uint16_t port_id, uint16_t rx_queue_id, uint16_t nb_rx_desc,
2213 const struct rte_eth_hairpin_conf *conf);
2216 * Allocate and set up a transmit queue for an Ethernet device.
2219 * The port identifier of the Ethernet device.
2220 * @param tx_queue_id
2221 * The index of the transmit queue to set up.
2222 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2223 * to rte_eth_dev_configure().
2225 * The number of transmit descriptors to allocate for the transmit ring.
2227 * The *socket_id* argument is the socket identifier in case of NUMA.
2228 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
2229 * the DMA memory allocated for the transmit descriptors of the ring.
2231 * The pointer to the configuration data to be used for the transmit queue.
2232 * NULL value is allowed, in which case default TX configuration
2234 * The *tx_conf* structure contains the following data:
2235 * - The *tx_thresh* structure with the values of the Prefetch, Host, and
2236 * Write-Back threshold registers of the transmit ring.
2237 * When setting Write-Back threshold to the value greater then zero,
2238 * *tx_rs_thresh* value should be explicitly set to one.
2239 * - The *tx_free_thresh* value indicates the [minimum] number of network
2240 * buffers that must be pending in the transmit ring to trigger their
2241 * [implicit] freeing by the driver transmit function.
2242 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit
2243 * descriptors that must be pending in the transmit ring before setting the
2244 * RS bit on a descriptor by the driver transmit function.
2245 * The *tx_rs_thresh* value should be less or equal then
2246 * *tx_free_thresh* value, and both of them should be less then
2248 * - The *offloads* member contains Tx offloads to be enabled.
2249 * If an offloading set in tx_conf->offloads
2250 * hasn't been set in the input argument eth_conf->txmode.offloads
2251 * to rte_eth_dev_configure(), it is a new added offloading, it must be
2252 * per-queue type and it is enabled for the queue.
2253 * No need to repeat any bit in tx_conf->offloads which has already been
2254 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
2255 * at port level can't be disabled at queue level.
2257 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces
2258 * the transmit function to use default values.
2260 * - 0: Success, the transmit queue is correctly set up.
2261 * - -ENOMEM: Unable to allocate the transmit ring descriptors.
2263 int rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,
2264 uint16_t nb_tx_desc, unsigned int socket_id,
2265 const struct rte_eth_txconf *tx_conf);
2269 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2271 * Allocate and set up a transmit hairpin queue for an Ethernet device.
2274 * The port identifier of the Ethernet device.
2275 * @param tx_queue_id
2276 * The index of the transmit queue to set up.
2277 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2278 * to rte_eth_dev_configure().
2280 * The number of transmit descriptors to allocate for the transmit ring.
2281 * 0 to set default PMD value.
2283 * The hairpin configuration.
2286 * - (0) if successful.
2287 * - (-ENODEV) if *port_id* is invalid.
2288 * - (-ENOTSUP) if hardware doesn't support.
2289 * - (-EINVAL) if bad parameter.
2290 * - (-ENOMEM) if unable to allocate the resources.
2293 int rte_eth_tx_hairpin_queue_setup
2294 (uint16_t port_id, uint16_t tx_queue_id, uint16_t nb_tx_desc,
2295 const struct rte_eth_hairpin_conf *conf);
2299 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2301 * Get all the hairpin peer Rx / Tx ports of the current port.
2302 * The caller should ensure that the array is large enough to save the ports
2306 * The port identifier of the Ethernet device.
2308 * Pointer to the array to store the peer ports list.
2310 * Length of the array to store the port identifiers.
2312 * Current port to peer port direction
2313 * positive - current used as Tx to get all peer Rx ports.
2314 * zero - current used as Rx to get all peer Tx ports.
2317 * - (0 or positive) actual peer ports number.
2318 * - (-EINVAL) if bad parameter.
2319 * - (-ENODEV) if *port_id* invalid
2320 * - (-ENOTSUP) if hardware doesn't support.
2321 * - Others detailed errors from PMD drivers.
2324 int rte_eth_hairpin_get_peer_ports(uint16_t port_id, uint16_t *peer_ports,
2325 size_t len, uint32_t direction);
2329 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2331 * Bind all hairpin Tx queues of one port to the Rx queues of the peer port.
2332 * It is only allowed to call this function after all hairpin queues are
2333 * configured properly and the devices are in started state.
2336 * The identifier of the Tx port.
2338 * The identifier of peer Rx port.
2339 * RTE_MAX_ETHPORTS is allowed for the traversal of all devices.
2340 * Rx port ID could have the same value as Tx port ID.
2343 * - (0) if successful.
2344 * - (-ENODEV) if Tx port ID is invalid.
2345 * - (-EBUSY) if device is not in started state.
2346 * - (-ENOTSUP) if hardware doesn't support.
2347 * - Others detailed errors from PMD drivers.
2350 int rte_eth_hairpin_bind(uint16_t tx_port, uint16_t rx_port);
2354 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
2356 * Unbind all hairpin Tx queues of one port from the Rx queues of the peer port.
2357 * This should be called before closing the Tx or Rx devices, if the bind
2358 * function is called before.
2359 * After unbinding the hairpin ports pair, it is allowed to bind them again.
2360 * Changing queues configuration should be after stopping the device(s).
2363 * The identifier of the Tx port.
2365 * The identifier of peer Rx port.
2366 * RTE_MAX_ETHPORTS is allowed for traversal of all devices.
2367 * Rx port ID could have the same value as Tx port ID.
2370 * - (0) if successful.
2371 * - (-ENODEV) if Tx port ID is invalid.
2372 * - (-EBUSY) if device is in stopped state.
2373 * - (-ENOTSUP) if hardware doesn't support.
2374 * - Others detailed errors from PMD drivers.
2377 int rte_eth_hairpin_unbind(uint16_t tx_port, uint16_t rx_port);
2380 * Return the NUMA socket to which an Ethernet device is connected
2383 * The port identifier of the Ethernet device
2385 * The NUMA socket id to which the Ethernet device is connected or
2386 * a default of zero if the socket could not be determined.
2387 * -1 is returned is the port_id value is out of range.
2389 int rte_eth_dev_socket_id(uint16_t port_id);
2392 * Check if port_id of device is attached
2395 * The port identifier of the Ethernet device
2397 * - 0 if port is out of range or not attached
2398 * - 1 if device is attached
2400 int rte_eth_dev_is_valid_port(uint16_t port_id);
2403 * Start specified RX queue of a port. It is used when rx_deferred_start
2404 * flag of the specified queue is true.
2407 * The port identifier of the Ethernet device
2408 * @param rx_queue_id
2409 * The index of the rx queue to update the ring.
2410 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2411 * to rte_eth_dev_configure().
2413 * - 0: Success, the receive queue is started.
2414 * - -ENODEV: if *port_id* is invalid.
2415 * - -EINVAL: The queue_id out of range or belong to hairpin.
2416 * - -EIO: if device is removed.
2417 * - -ENOTSUP: The function not supported in PMD driver.
2419 int rte_eth_dev_rx_queue_start(uint16_t port_id, uint16_t rx_queue_id);
2422 * Stop specified RX queue of a port
2425 * The port identifier of the Ethernet device
2426 * @param rx_queue_id
2427 * The index of the rx queue to update the ring.
2428 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2429 * to rte_eth_dev_configure().
2431 * - 0: Success, the receive queue is stopped.
2432 * - -ENODEV: if *port_id* is invalid.
2433 * - -EINVAL: The queue_id out of range or belong to hairpin.
2434 * - -EIO: if device is removed.
2435 * - -ENOTSUP: The function not supported in PMD driver.
2437 int rte_eth_dev_rx_queue_stop(uint16_t port_id, uint16_t rx_queue_id);
2440 * Start TX for specified queue of a port. It is used when tx_deferred_start
2441 * flag of the specified queue is true.
2444 * The port identifier of the Ethernet device
2445 * @param tx_queue_id
2446 * The index of the tx queue to update the ring.
2447 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2448 * to rte_eth_dev_configure().
2450 * - 0: Success, the transmit queue is started.
2451 * - -ENODEV: if *port_id* is invalid.
2452 * - -EINVAL: The queue_id out of range or belong to hairpin.
2453 * - -EIO: if device is removed.
2454 * - -ENOTSUP: The function not supported in PMD driver.
2456 int rte_eth_dev_tx_queue_start(uint16_t port_id, uint16_t tx_queue_id);
2459 * Stop specified TX queue of a port
2462 * The port identifier of the Ethernet device
2463 * @param tx_queue_id
2464 * The index of the tx queue to update the ring.
2465 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2466 * to rte_eth_dev_configure().
2468 * - 0: Success, the transmit queue is stopped.
2469 * - -ENODEV: if *port_id* is invalid.
2470 * - -EINVAL: The queue_id out of range or belong to hairpin.
2471 * - -EIO: if device is removed.
2472 * - -ENOTSUP: The function not supported in PMD driver.
2474 int rte_eth_dev_tx_queue_stop(uint16_t port_id, uint16_t tx_queue_id);
2477 * Start an Ethernet device.
2479 * The device start step is the last one and consists of setting the configured
2480 * offload features and in starting the transmit and the receive units of the
2483 * Device RTE_ETH_DEV_NOLIVE_MAC_ADDR flag causes MAC address to be set before
2484 * PMD port start callback function is invoked.
2486 * On success, all basic functions exported by the Ethernet API (link status,
2487 * receive/transmit, and so on) can be invoked.
2490 * The port identifier of the Ethernet device.
2492 * - 0: Success, Ethernet device started.
2493 * - <0: Error code of the driver device start function.
2495 int rte_eth_dev_start(uint16_t port_id);
2498 * Stop an Ethernet device. The device can be restarted with a call to
2499 * rte_eth_dev_start()
2502 * The port identifier of the Ethernet device.
2504 * - 0: Success, Ethernet device stopped.
2505 * - <0: Error code of the driver device stop function.
2507 int rte_eth_dev_stop(uint16_t port_id);
2510 * Link up an Ethernet device.
2512 * Set device link up will re-enable the device rx/tx
2513 * functionality after it is previously set device linked down.
2516 * The port identifier of the Ethernet device.
2518 * - 0: Success, Ethernet device linked up.
2519 * - <0: Error code of the driver device link up function.
2521 int rte_eth_dev_set_link_up(uint16_t port_id);
2524 * Link down an Ethernet device.
2525 * The device rx/tx functionality will be disabled if success,
2526 * and it can be re-enabled with a call to
2527 * rte_eth_dev_set_link_up()
2530 * The port identifier of the Ethernet device.
2532 int rte_eth_dev_set_link_down(uint16_t port_id);
2535 * Close a stopped Ethernet device. The device cannot be restarted!
2536 * The function frees all port resources.
2539 * The port identifier of the Ethernet device.
2541 * - Zero if the port is closed successfully.
2542 * - Negative if something went wrong.
2544 int rte_eth_dev_close(uint16_t port_id);
2547 * Reset a Ethernet device and keep its port id.
2549 * When a port has to be reset passively, the DPDK application can invoke
2550 * this function. For example when a PF is reset, all its VFs should also
2551 * be reset. Normally a DPDK application can invoke this function when
2552 * RTE_ETH_EVENT_INTR_RESET event is detected, but can also use it to start
2553 * a port reset in other circumstances.
2555 * When this function is called, it first stops the port and then calls the
2556 * PMD specific dev_uninit( ) and dev_init( ) to return the port to initial
2557 * state, in which no Tx and Rx queues are setup, as if the port has been
2558 * reset and not started. The port keeps the port id it had before the
2561 * After calling rte_eth_dev_reset( ), the application should use
2562 * rte_eth_dev_configure( ), rte_eth_rx_queue_setup( ),
2563 * rte_eth_tx_queue_setup( ), and rte_eth_dev_start( )
2564 * to reconfigure the device as appropriate.
2566 * Note: To avoid unexpected behavior, the application should stop calling
2567 * Tx and Rx functions before calling rte_eth_dev_reset( ). For thread
2568 * safety, all these controlling functions should be called from the same
2572 * The port identifier of the Ethernet device.
2575 * - (0) if successful.
2576 * - (-ENODEV) if *port_id* is invalid.
2577 * - (-ENOTSUP) if hardware doesn't support this function.
2578 * - (-EPERM) if not ran from the primary process.
2579 * - (-EIO) if re-initialisation failed or device is removed.
2580 * - (-ENOMEM) if the reset failed due to OOM.
2581 * - (-EAGAIN) if the reset temporarily failed and should be retried later.
2583 int rte_eth_dev_reset(uint16_t port_id);
2586 * Enable receipt in promiscuous mode for an Ethernet device.
2589 * The port identifier of the Ethernet device.
2591 * - (0) if successful.
2592 * - (-ENOTSUP) if support for promiscuous_enable() does not exist
2594 * - (-ENODEV) if *port_id* invalid.
2596 int rte_eth_promiscuous_enable(uint16_t port_id);
2599 * Disable receipt in promiscuous mode for an Ethernet device.
2602 * The port identifier of the Ethernet device.
2604 * - (0) if successful.
2605 * - (-ENOTSUP) if support for promiscuous_disable() does not exist
2607 * - (-ENODEV) if *port_id* invalid.
2609 int rte_eth_promiscuous_disable(uint16_t port_id);
2612 * Return the value of promiscuous mode for an Ethernet device.
2615 * The port identifier of the Ethernet device.
2617 * - (1) if promiscuous is enabled
2618 * - (0) if promiscuous is disabled.
2621 int rte_eth_promiscuous_get(uint16_t port_id);
2624 * Enable the receipt of any multicast frame by an Ethernet device.
2627 * The port identifier of the Ethernet device.
2629 * - (0) if successful.
2630 * - (-ENOTSUP) if support for allmulticast_enable() does not exist
2632 * - (-ENODEV) if *port_id* invalid.
2634 int rte_eth_allmulticast_enable(uint16_t port_id);
2637 * Disable the receipt of all multicast frames by an Ethernet device.
2640 * The port identifier of the Ethernet device.
2642 * - (0) if successful.
2643 * - (-ENOTSUP) if support for allmulticast_disable() does not exist
2645 * - (-ENODEV) if *port_id* invalid.
2647 int rte_eth_allmulticast_disable(uint16_t port_id);
2650 * Return the value of allmulticast mode for an Ethernet device.
2653 * The port identifier of the Ethernet device.
2655 * - (1) if allmulticast is enabled
2656 * - (0) if allmulticast is disabled.
2659 int rte_eth_allmulticast_get(uint16_t port_id);
2662 * Retrieve the link status (up/down), the duplex mode (half/full),
2663 * the negotiation (auto/fixed), and if available, the speed (Mbps).
2665 * It might need to wait up to 9 seconds.
2666 * @see rte_eth_link_get_nowait.
2669 * The port identifier of the Ethernet device.
2671 * Link information written back.
2673 * - (0) if successful.
2674 * - (-ENOTSUP) if the function is not supported in PMD driver.
2675 * - (-ENODEV) if *port_id* invalid.
2677 int rte_eth_link_get(uint16_t port_id, struct rte_eth_link *link);
2680 * Retrieve the link status (up/down), the duplex mode (half/full),
2681 * the negotiation (auto/fixed), and if available, the speed (Mbps).
2684 * The port identifier of the Ethernet device.
2686 * Link information written back.
2688 * - (0) if successful.
2689 * - (-ENOTSUP) if the function is not supported in PMD driver.
2690 * - (-ENODEV) if *port_id* invalid.
2692 int rte_eth_link_get_nowait(uint16_t port_id, struct rte_eth_link *link);
2696 * @b EXPERIMENTAL: this API may change without prior notice.
2698 * The function converts a link_speed to a string. It handles all special
2699 * values like unknown or none speed.
2702 * link_speed of rte_eth_link struct
2704 * Link speed in textual format. It's pointer to immutable memory.
2705 * No free is required.
2708 const char *rte_eth_link_speed_to_str(uint32_t link_speed);
2712 * @b EXPERIMENTAL: this API may change without prior notice.
2714 * The function converts a rte_eth_link struct representing a link status to
2718 * A pointer to a string to be filled with textual representation of
2719 * device status. At least ETH_LINK_MAX_STR_LEN bytes should be allocated to
2720 * store default link status text.
2722 * Length of available memory at 'str' string.
2724 * Link status returned by rte_eth_link_get function
2726 * Number of bytes written to str array.
2729 int rte_eth_link_to_str(char *str, size_t len,
2730 const struct rte_eth_link *eth_link);
2733 * Retrieve the general I/O statistics of an Ethernet device.
2736 * The port identifier of the Ethernet device.
2738 * A pointer to a structure of type *rte_eth_stats* to be filled with
2739 * the values of device counters for the following set of statistics:
2740 * - *ipackets* with the total of successfully received packets.
2741 * - *opackets* with the total of successfully transmitted packets.
2742 * - *ibytes* with the total of successfully received bytes.
2743 * - *obytes* with the total of successfully transmitted bytes.
2744 * - *ierrors* with the total of erroneous received packets.
2745 * - *oerrors* with the total of failed transmitted packets.
2747 * Zero if successful. Non-zero otherwise.
2749 int rte_eth_stats_get(uint16_t port_id, struct rte_eth_stats *stats);
2752 * Reset the general I/O statistics of an Ethernet device.
2755 * The port identifier of the Ethernet device.
2757 * - (0) if device notified to reset stats.
2758 * - (-ENOTSUP) if hardware doesn't support.
2759 * - (-ENODEV) if *port_id* invalid.
2760 * - (<0): Error code of the driver stats reset function.
2762 int rte_eth_stats_reset(uint16_t port_id);
2765 * Retrieve names of extended statistics of an Ethernet device.
2767 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2769 * xstats_names[i].name => xstats[i].value
2771 * And the array index is same with id field of 'struct rte_eth_xstat':
2774 * This assumption makes key-value pair matching less flexible but simpler.
2777 * The port identifier of the Ethernet device.
2778 * @param xstats_names
2779 * An rte_eth_xstat_name array of at least *size* elements to
2780 * be filled. If set to NULL, the function returns the required number
2783 * The size of the xstats_names array (number of elements).
2785 * - A positive value lower or equal to size: success. The return value
2786 * is the number of entries filled in the stats table.
2787 * - A positive value higher than size: error, the given statistics table
2788 * is too small. The return value corresponds to the size that should
2789 * be given to succeed. The entries in the table are not valid and
2790 * shall not be used by the caller.
2791 * - A negative value on error (invalid port id).
2793 int rte_eth_xstats_get_names(uint16_t port_id,
2794 struct rte_eth_xstat_name *xstats_names,
2798 * Retrieve extended statistics of an Ethernet device.
2800 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2802 * xstats_names[i].name => xstats[i].value
2804 * And the array index is same with id field of 'struct rte_eth_xstat':
2807 * This assumption makes key-value pair matching less flexible but simpler.
2810 * The port identifier of the Ethernet device.
2812 * A pointer to a table of structure of type *rte_eth_xstat*
2813 * to be filled with device statistics ids and values.
2814 * This parameter can be set to NULL if n is 0.
2816 * The size of the xstats array (number of elements).
2818 * - A positive value lower or equal to n: success. The return value
2819 * is the number of entries filled in the stats table.
2820 * - A positive value higher than n: error, the given statistics table
2821 * is too small. The return value corresponds to the size that should
2822 * be given to succeed. The entries in the table are not valid and
2823 * shall not be used by the caller.
2824 * - A negative value on error (invalid port id).
2826 int rte_eth_xstats_get(uint16_t port_id, struct rte_eth_xstat *xstats,
2830 * Retrieve names of extended statistics of an Ethernet device.
2833 * The port identifier of the Ethernet device.
2834 * @param xstats_names
2835 * An rte_eth_xstat_name array of at least *size* elements to
2836 * be filled. If set to NULL, the function returns the required number
2839 * IDs array given by app to retrieve specific statistics
2841 * The size of the xstats_names array (number of elements).
2843 * - A positive value lower or equal to size: success. The return value
2844 * is the number of entries filled in the stats table.
2845 * - A positive value higher than size: error, the given statistics table
2846 * is too small. The return value corresponds to the size that should
2847 * be given to succeed. The entries in the table are not valid and
2848 * shall not be used by the caller.
2849 * - A negative value on error (invalid port id).
2852 rte_eth_xstats_get_names_by_id(uint16_t port_id,
2853 struct rte_eth_xstat_name *xstats_names, unsigned int size,
2857 * Retrieve extended statistics of an Ethernet device.
2860 * The port identifier of the Ethernet device.
2862 * A pointer to an ids array passed by application. This tells which
2863 * statistics values function should retrieve. This parameter
2864 * can be set to NULL if size is 0. In this case function will retrieve
2865 * all available statistics.
2867 * A pointer to a table to be filled with device statistics values.
2869 * The size of the ids array (number of elements).
2871 * - A positive value lower or equal to size: success. The return value
2872 * is the number of entries filled in the stats table.
2873 * - A positive value higher than size: error, the given statistics table
2874 * is too small. The return value corresponds to the size that should
2875 * be given to succeed. The entries in the table are not valid and
2876 * shall not be used by the caller.
2877 * - A negative value on error (invalid port id).
2879 int rte_eth_xstats_get_by_id(uint16_t port_id, const uint64_t *ids,
2880 uint64_t *values, unsigned int size);
2883 * Gets the ID of a statistic from its name.
2885 * This function searches for the statistics using string compares, and
2886 * as such should not be used on the fast-path. For fast-path retrieval of
2887 * specific statistics, store the ID as provided in *id* from this function,
2888 * and pass the ID to rte_eth_xstats_get()
2890 * @param port_id The port to look up statistics from
2891 * @param xstat_name The name of the statistic to return
2892 * @param[out] id A pointer to an app-supplied uint64_t which should be
2893 * set to the ID of the stat if the stat exists.
2896 * -ENODEV for invalid port_id,
2897 * -EIO if device is removed,
2898 * -EINVAL if the xstat_name doesn't exist in port_id
2900 int rte_eth_xstats_get_id_by_name(uint16_t port_id, const char *xstat_name,
2904 * Reset extended statistics of an Ethernet device.
2907 * The port identifier of the Ethernet device.
2909 * - (0) if device notified to reset extended stats.
2910 * - (-ENOTSUP) if pmd doesn't support both
2911 * extended stats and basic stats reset.
2912 * - (-ENODEV) if *port_id* invalid.
2913 * - (<0): Error code of the driver xstats reset function.
2915 int rte_eth_xstats_reset(uint16_t port_id);
2918 * Set a mapping for the specified transmit queue to the specified per-queue
2919 * statistics counter.
2922 * The port identifier of the Ethernet device.
2923 * @param tx_queue_id
2924 * The index of the transmit queue for which a queue stats mapping is required.
2925 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2926 * to rte_eth_dev_configure().
2928 * The per-queue packet statistics functionality number that the transmit
2929 * queue is to be assigned.
2930 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
2931 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256.
2933 * Zero if successful. Non-zero otherwise.
2935 int rte_eth_dev_set_tx_queue_stats_mapping(uint16_t port_id,
2936 uint16_t tx_queue_id, uint8_t stat_idx);
2939 * Set a mapping for the specified receive queue to the specified per-queue
2940 * statistics counter.
2943 * The port identifier of the Ethernet device.
2944 * @param rx_queue_id
2945 * The index of the receive queue for which a queue stats mapping is required.
2946 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2947 * to rte_eth_dev_configure().
2949 * The per-queue packet statistics functionality number that the receive
2950 * queue is to be assigned.
2951 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
2952 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256.
2954 * Zero if successful. Non-zero otherwise.
2956 int rte_eth_dev_set_rx_queue_stats_mapping(uint16_t port_id,
2957 uint16_t rx_queue_id,
2961 * Retrieve the Ethernet address of an Ethernet device.
2964 * The port identifier of the Ethernet device.
2966 * A pointer to a structure of type *ether_addr* to be filled with
2967 * the Ethernet address of the Ethernet device.
2969 * - (0) if successful
2970 * - (-ENODEV) if *port_id* invalid.
2972 int rte_eth_macaddr_get(uint16_t port_id, struct rte_ether_addr *mac_addr);
2975 * Retrieve the contextual information of an Ethernet device.
2977 * As part of this function, a number of of fields in dev_info will be
2978 * initialized as follows:
2983 * Where lim is defined within the rte_eth_dev_info_get as
2985 * const struct rte_eth_desc_lim lim = {
2986 * .nb_max = UINT16_MAX,
2989 * .nb_seg_max = UINT16_MAX,
2990 * .nb_mtu_seg_max = UINT16_MAX,
2993 * device = dev->device
2994 * min_mtu = RTE_ETHER_MIN_MTU
2995 * max_mtu = UINT16_MAX
2997 * The following fields will be populated if support for dev_infos_get()
2998 * exists for the device and the rte_eth_dev 'dev' has been populated
2999 * successfully with a call to it:
3001 * driver_name = dev->device->driver->name
3002 * nb_rx_queues = dev->data->nb_rx_queues
3003 * nb_tx_queues = dev->data->nb_tx_queues
3004 * dev_flags = &dev->data->dev_flags
3007 * The port identifier of the Ethernet device.
3009 * A pointer to a structure of type *rte_eth_dev_info* to be filled with
3010 * the contextual information of the Ethernet device.
3012 * - (0) if successful.
3013 * - (-ENOTSUP) if support for dev_infos_get() does not exist for the device.
3014 * - (-ENODEV) if *port_id* invalid.
3016 int rte_eth_dev_info_get(uint16_t port_id, struct rte_eth_dev_info *dev_info);
3019 * Retrieve the firmware version of a device.
3022 * The port identifier of the device.
3024 * A pointer to a string array storing the firmware version of a device,
3025 * the string includes terminating null. This pointer is allocated by caller.
3027 * The size of the string array pointed by fw_version, which should be
3028 * large enough to store firmware version of the device.
3030 * - (0) if successful.
3031 * - (-ENOTSUP) if operation is not supported.
3032 * - (-ENODEV) if *port_id* invalid.
3033 * - (-EIO) if device is removed.
3034 * - (>0) if *fw_size* is not enough to store firmware version, return
3035 * the size of the non truncated string.
3037 int rte_eth_dev_fw_version_get(uint16_t port_id,
3038 char *fw_version, size_t fw_size);
3041 * Retrieve the supported packet types of an Ethernet device.
3043 * When a packet type is announced as supported, it *must* be recognized by
3044 * the PMD. For instance, if RTE_PTYPE_L2_ETHER, RTE_PTYPE_L2_ETHER_VLAN
3045 * and RTE_PTYPE_L3_IPV4 are announced, the PMD must return the following
3046 * packet types for these packets:
3047 * - Ether/IPv4 -> RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4
3048 * - Ether/Vlan/IPv4 -> RTE_PTYPE_L2_ETHER_VLAN | RTE_PTYPE_L3_IPV4
3049 * - Ether/[anything else] -> RTE_PTYPE_L2_ETHER
3050 * - Ether/Vlan/[anything else] -> RTE_PTYPE_L2_ETHER_VLAN
3052 * When a packet is received by a PMD, the most precise type must be
3053 * returned among the ones supported. However a PMD is allowed to set
3054 * packet type that is not in the supported list, at the condition that it
3055 * is more precise. Therefore, a PMD announcing no supported packet types
3056 * can still set a matching packet type in a received packet.
3059 * Better to invoke this API after the device is already started or rx burst
3060 * function is decided, to obtain correct supported ptypes.
3062 * if a given PMD does not report what ptypes it supports, then the supported
3063 * ptype count is reported as 0.
3065 * The port identifier of the Ethernet device.
3067 * A hint of what kind of packet type which the caller is interested in.
3069 * An array pointer to store adequate packet types, allocated by caller.
3071 * Size of the array pointed by param ptypes.
3073 * - (>=0) Number of supported ptypes. If the number of types exceeds num,
3074 * only num entries will be filled into the ptypes array, but the full
3075 * count of supported ptypes will be returned.
3076 * - (-ENODEV) if *port_id* invalid.
3078 int rte_eth_dev_get_supported_ptypes(uint16_t port_id, uint32_t ptype_mask,
3079 uint32_t *ptypes, int num);
3082 * @b EXPERIMENTAL: this API may change without prior notice.
3084 * Inform Ethernet device about reduced range of packet types to handle.
3086 * Application can use this function to set only specific ptypes that it's
3087 * interested. This information can be used by the PMD to optimize Rx path.
3089 * The function accepts an array `set_ptypes` allocated by the caller to
3090 * store the packet types set by the driver, the last element of the array
3091 * is set to RTE_PTYPE_UNKNOWN. The size of the `set_ptype` array should be
3092 * `rte_eth_dev_get_supported_ptypes() + 1` else it might only be filled
3096 * The port identifier of the Ethernet device.
3098 * The ptype family that application is interested in should be bitwise OR of
3099 * RTE_PTYPE_*_MASK or 0.
3101 * An array pointer to store set packet types, allocated by caller. The
3102 * function marks the end of array with RTE_PTYPE_UNKNOWN.
3104 * Size of the array pointed by param ptypes.
3105 * Should be rte_eth_dev_get_supported_ptypes() + 1 to accommodate the
3109 * - (-ENODEV) if *port_id* invalid.
3110 * - (-EINVAL) if *ptype_mask* is invalid (or) set_ptypes is NULL and
3114 int rte_eth_dev_set_ptypes(uint16_t port_id, uint32_t ptype_mask,
3115 uint32_t *set_ptypes, unsigned int num);
3118 * Retrieve the MTU of an Ethernet device.
3121 * The port identifier of the Ethernet device.
3123 * A pointer to a uint16_t where the retrieved MTU is to be stored.
3125 * - (0) if successful.
3126 * - (-ENODEV) if *port_id* invalid.
3128 int rte_eth_dev_get_mtu(uint16_t port_id, uint16_t *mtu);
3131 * Change the MTU of an Ethernet device.
3134 * The port identifier of the Ethernet device.
3136 * A uint16_t for the MTU to be applied.
3138 * - (0) if successful.
3139 * - (-ENOTSUP) if operation is not supported.
3140 * - (-ENODEV) if *port_id* invalid.
3141 * - (-EIO) if device is removed.
3142 * - (-EINVAL) if *mtu* invalid, validation of mtu can occur within
3143 * rte_eth_dev_set_mtu if dev_infos_get is supported by the device or
3144 * when the mtu is set using dev->dev_ops->mtu_set.
3145 * - (-EBUSY) if operation is not allowed when the port is running
3147 int rte_eth_dev_set_mtu(uint16_t port_id, uint16_t mtu);
3150 * Enable/Disable hardware filtering by an Ethernet device of received
3151 * VLAN packets tagged with a given VLAN Tag Identifier.
3154 * The port identifier of the Ethernet device.
3156 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
3158 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*.
3159 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*.
3161 * - (0) if successful.
3162 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
3163 * - (-ENODEV) if *port_id* invalid.
3164 * - (-EIO) if device is removed.
3165 * - (-ENOSYS) if VLAN filtering on *port_id* disabled.
3166 * - (-EINVAL) if *vlan_id* > 4095.
3168 int rte_eth_dev_vlan_filter(uint16_t port_id, uint16_t vlan_id, int on);
3171 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device.
3174 * The port identifier of the Ethernet device.
3175 * @param rx_queue_id
3176 * The index of the receive queue for which a queue stats mapping is required.
3177 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3178 * to rte_eth_dev_configure().
3180 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port.
3181 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port.
3183 * - (0) if successful.
3184 * - (-ENOTSUP) if hardware-assisted VLAN stripping not configured.
3185 * - (-ENODEV) if *port_id* invalid.
3186 * - (-EINVAL) if *rx_queue_id* invalid.
3188 int rte_eth_dev_set_vlan_strip_on_queue(uint16_t port_id, uint16_t rx_queue_id,
3192 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to
3196 * The port identifier of the Ethernet device.
3200 * The Tag Protocol ID
3202 * - (0) if successful.
3203 * - (-ENOTSUP) if hardware-assisted VLAN TPID setup is not supported.
3204 * - (-ENODEV) if *port_id* invalid.
3205 * - (-EIO) if device is removed.
3207 int rte_eth_dev_set_vlan_ether_type(uint16_t port_id,
3208 enum rte_vlan_type vlan_type,
3212 * Set VLAN offload configuration on an Ethernet device.
3215 * The port identifier of the Ethernet device.
3216 * @param offload_mask
3217 * The VLAN Offload bit mask can be mixed use with "OR"
3218 * ETH_VLAN_STRIP_OFFLOAD
3219 * ETH_VLAN_FILTER_OFFLOAD
3220 * ETH_VLAN_EXTEND_OFFLOAD
3221 * ETH_QINQ_STRIP_OFFLOAD
3223 * - (0) if successful.
3224 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
3225 * - (-ENODEV) if *port_id* invalid.
3226 * - (-EIO) if device is removed.
3228 int rte_eth_dev_set_vlan_offload(uint16_t port_id, int offload_mask);
3231 * Read VLAN Offload configuration from an Ethernet device
3234 * The port identifier of the Ethernet device.
3236 * - (>0) if successful. Bit mask to indicate
3237 * ETH_VLAN_STRIP_OFFLOAD
3238 * ETH_VLAN_FILTER_OFFLOAD
3239 * ETH_VLAN_EXTEND_OFFLOAD
3240 * ETH_QINQ_STRIP_OFFLOAD
3241 * - (-ENODEV) if *port_id* invalid.
3243 int rte_eth_dev_get_vlan_offload(uint16_t port_id);
3246 * Set port based TX VLAN insertion on or off.
3249 * The port identifier of the Ethernet device.
3251 * Port based TX VLAN identifier together with user priority.
3253 * Turn on or off the port based TX VLAN insertion.
3256 * - (0) if successful.
3257 * - negative if failed.
3259 int rte_eth_dev_set_vlan_pvid(uint16_t port_id, uint16_t pvid, int on);
3261 typedef void (*buffer_tx_error_fn)(struct rte_mbuf **unsent, uint16_t count,
3265 * Structure used to buffer packets for future TX
3266 * Used by APIs rte_eth_tx_buffer and rte_eth_tx_buffer_flush
3268 struct rte_eth_dev_tx_buffer {
3269 buffer_tx_error_fn error_callback;
3270 void *error_userdata;
3271 uint16_t size; /**< Size of buffer for buffered tx */
3272 uint16_t length; /**< Number of packets in the array */
3273 struct rte_mbuf *pkts[];
3274 /**< Pending packets to be sent on explicit flush or when full */
3278 * Calculate the size of the tx buffer.
3281 * Number of stored packets.
3283 #define RTE_ETH_TX_BUFFER_SIZE(sz) \
3284 (sizeof(struct rte_eth_dev_tx_buffer) + (sz) * sizeof(struct rte_mbuf *))
3287 * Initialize default values for buffered transmitting
3290 * Tx buffer to be initialized.
3297 rte_eth_tx_buffer_init(struct rte_eth_dev_tx_buffer *buffer, uint16_t size);
3300 * Configure a callback for buffered packets which cannot be sent
3302 * Register a specific callback to be called when an attempt is made to send
3303 * all packets buffered on an ethernet port, but not all packets can
3304 * successfully be sent. The callback registered here will be called only
3305 * from calls to rte_eth_tx_buffer() and rte_eth_tx_buffer_flush() APIs.
3306 * The default callback configured for each queue by default just frees the
3307 * packets back to the calling mempool. If additional behaviour is required,
3308 * for example, to count dropped packets, or to retry transmission of packets
3309 * which cannot be sent, this function should be used to register a suitable
3310 * callback function to implement the desired behaviour.
3311 * The example callback "rte_eth_count_unsent_packet_callback()" is also
3312 * provided as reference.
3315 * The port identifier of the Ethernet device.
3317 * The function to be used as the callback.
3319 * Arbitrary parameter to be passed to the callback function
3321 * 0 on success, or -1 on error with rte_errno set appropriately
3324 rte_eth_tx_buffer_set_err_callback(struct rte_eth_dev_tx_buffer *buffer,
3325 buffer_tx_error_fn callback, void *userdata);
3328 * Callback function for silently dropping unsent buffered packets.
3330 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
3331 * adjust the default behavior when buffered packets cannot be sent. This
3332 * function drops any unsent packets silently and is used by tx buffered
3333 * operations as default behavior.
3335 * NOTE: this function should not be called directly, instead it should be used
3336 * as a callback for packet buffering.
3338 * NOTE: when configuring this function as a callback with
3339 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
3340 * should point to an uint64_t value.
3343 * The previously buffered packets which could not be sent
3345 * The number of unsent packets in the pkts array
3350 rte_eth_tx_buffer_drop_callback(struct rte_mbuf **pkts, uint16_t unsent,
3354 * Callback function for tracking unsent buffered packets.
3356 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
3357 * adjust the default behavior when buffered packets cannot be sent. This
3358 * function drops any unsent packets, but also updates a user-supplied counter
3359 * to track the overall number of packets dropped. The counter should be an
3360 * uint64_t variable.
3362 * NOTE: this function should not be called directly, instead it should be used
3363 * as a callback for packet buffering.
3365 * NOTE: when configuring this function as a callback with
3366 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
3367 * should point to an uint64_t value.
3370 * The previously buffered packets which could not be sent
3372 * The number of unsent packets in the pkts array
3374 * Pointer to an uint64_t value, which will be incremented by unsent
3377 rte_eth_tx_buffer_count_callback(struct rte_mbuf **pkts, uint16_t unsent,
3381 * Request the driver to free mbufs currently cached by the driver. The
3382 * driver will only free the mbuf if it is no longer in use. It is the
3383 * application's responsibility to ensure rte_eth_tx_buffer_flush(..) is
3387 * The port identifier of the Ethernet device.
3389 * The index of the transmit queue through which output packets must be
3391 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
3392 * to rte_eth_dev_configure().
3394 * Maximum number of packets to free. Use 0 to indicate all possible packets
3395 * should be freed. Note that a packet may be using multiple mbufs.
3398 * -ENODEV: Invalid interface
3399 * -EIO: device is removed
3400 * -ENOTSUP: Driver does not support function
3402 * 0-n: Number of packets freed. More packets may still remain in ring that
3406 rte_eth_tx_done_cleanup(uint16_t port_id, uint16_t queue_id, uint32_t free_cnt);
3409 * Subtypes for IPsec offload event(@ref RTE_ETH_EVENT_IPSEC) raised by
3412 enum rte_eth_event_ipsec_subtype {
3413 RTE_ETH_EVENT_IPSEC_UNKNOWN = 0,
3414 /**< Unknown event type */
3415 RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW,
3416 /**< Sequence number overflow */
3417 RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY,
3418 /**< Soft time expiry of SA */
3419 RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY,
3420 /**< Soft byte expiry of SA */
3421 RTE_ETH_EVENT_IPSEC_MAX
3422 /**< Max value of this enum */
3426 * Descriptor for @ref RTE_ETH_EVENT_IPSEC event. Used by eth dev to send extra
3427 * information of the IPsec offload event.
3429 struct rte_eth_event_ipsec_desc {
3430 enum rte_eth_event_ipsec_subtype subtype;
3431 /**< Type of RTE_ETH_EVENT_IPSEC_* event */
3433 /**< Event specific metadata
3435 * For the following events, *userdata* registered
3436 * with the *rte_security_session* would be returned
3439 * - @ref RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW
3440 * - @ref RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY
3441 * - @ref RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY
3443 * @see struct rte_security_session_conf
3449 * The eth device event type for interrupt, and maybe others in the future.
3451 enum rte_eth_event_type {
3452 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */
3453 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */
3454 RTE_ETH_EVENT_QUEUE_STATE,
3455 /**< queue state event (enabled/disabled) */
3456 RTE_ETH_EVENT_INTR_RESET,
3457 /**< reset interrupt event, sent to VF on PF reset */
3458 RTE_ETH_EVENT_VF_MBOX, /**< message from the VF received by PF */
3459 RTE_ETH_EVENT_MACSEC, /**< MACsec offload related event */
3460 RTE_ETH_EVENT_INTR_RMV, /**< device removal event */
3461 RTE_ETH_EVENT_NEW, /**< port is probed */
3462 RTE_ETH_EVENT_DESTROY, /**< port is released */
3463 RTE_ETH_EVENT_IPSEC, /**< IPsec offload related event */
3464 RTE_ETH_EVENT_FLOW_AGED,/**< New aged-out flows is detected */
3465 RTE_ETH_EVENT_MAX /**< max value of this enum */
3468 typedef int (*rte_eth_dev_cb_fn)(uint16_t port_id,
3469 enum rte_eth_event_type event, void *cb_arg, void *ret_param);
3470 /**< user application callback to be registered for interrupts */
3473 * Register a callback function for port event.
3477 * RTE_ETH_ALL means register the event for all port ids.
3481 * User supplied callback function to be called.
3483 * Pointer to the parameters for the registered callback.
3486 * - On success, zero.
3487 * - On failure, a negative value.
3489 int rte_eth_dev_callback_register(uint16_t port_id,
3490 enum rte_eth_event_type event,
3491 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3494 * Unregister a callback function for port event.
3498 * RTE_ETH_ALL means unregister the event for all port ids.
3502 * User supplied callback function to be called.
3504 * Pointer to the parameters for the registered callback. -1 means to
3505 * remove all for the same callback address and same event.
3508 * - On success, zero.
3509 * - On failure, a negative value.
3511 int rte_eth_dev_callback_unregister(uint16_t port_id,
3512 enum rte_eth_event_type event,
3513 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3516 * When there is no rx packet coming in Rx Queue for a long time, we can
3517 * sleep lcore related to RX Queue for power saving, and enable rx interrupt
3518 * to be triggered when Rx packet arrives.
3520 * The rte_eth_dev_rx_intr_enable() function enables rx queue
3521 * interrupt on specific rx queue of a port.
3524 * The port identifier of the Ethernet device.
3526 * The index of the receive queue from which to retrieve input packets.
3527 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3528 * to rte_eth_dev_configure().
3530 * - (0) if successful.
3531 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3533 * - (-ENODEV) if *port_id* invalid.
3534 * - (-EIO) if device is removed.
3536 int rte_eth_dev_rx_intr_enable(uint16_t port_id, uint16_t queue_id);
3539 * When lcore wakes up from rx interrupt indicating packet coming, disable rx
3540 * interrupt and returns to polling mode.
3542 * The rte_eth_dev_rx_intr_disable() function disables rx queue
3543 * interrupt on specific rx queue of a port.
3546 * The port identifier of the Ethernet device.
3548 * The index of the receive queue from which to retrieve input packets.
3549 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3550 * to rte_eth_dev_configure().
3552 * - (0) if successful.
3553 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3555 * - (-ENODEV) if *port_id* invalid.
3556 * - (-EIO) if device is removed.
3558 int rte_eth_dev_rx_intr_disable(uint16_t port_id, uint16_t queue_id);
3561 * RX Interrupt control per port.
3564 * The port identifier of the Ethernet device.
3566 * Epoll instance fd which the intr vector associated to.
3567 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3569 * The operation be performed for the vector.
3570 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3574 * - On success, zero.
3575 * - On failure, a negative value.
3577 int rte_eth_dev_rx_intr_ctl(uint16_t port_id, int epfd, int op, void *data);
3580 * RX Interrupt control per queue.
3583 * The port identifier of the Ethernet device.
3585 * The index of the receive queue from which to retrieve input packets.
3586 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3587 * to rte_eth_dev_configure().
3589 * Epoll instance fd which the intr vector associated to.
3590 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3592 * The operation be performed for the vector.
3593 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3597 * - On success, zero.
3598 * - On failure, a negative value.
3600 int rte_eth_dev_rx_intr_ctl_q(uint16_t port_id, uint16_t queue_id,
3601 int epfd, int op, void *data);
3605 * @b EXPERIMENTAL: this API may change without prior notice.
3607 * Get interrupt fd per Rx queue.
3610 * The port identifier of the Ethernet device.
3612 * The index of the receive queue from which to retrieve input packets.
3613 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3614 * to rte_eth_dev_configure().
3616 * - (>=0) the interrupt fd associated to the requested Rx queue if
3622 rte_eth_dev_rx_intr_ctl_q_get_fd(uint16_t port_id, uint16_t queue_id);
3625 * Turn on the LED on the Ethernet device.
3626 * This function turns on the LED on the Ethernet device.
3629 * The port identifier of the Ethernet device.
3631 * - (0) if successful.
3632 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3634 * - (-ENODEV) if *port_id* invalid.
3635 * - (-EIO) if device is removed.
3637 int rte_eth_led_on(uint16_t port_id);
3640 * Turn off the LED on the Ethernet device.
3641 * This function turns off the LED on the Ethernet device.
3644 * The port identifier of the Ethernet device.
3646 * - (0) if successful.
3647 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3649 * - (-ENODEV) if *port_id* invalid.
3650 * - (-EIO) if device is removed.
3652 int rte_eth_led_off(uint16_t port_id);
3656 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
3658 * Get Forward Error Correction(FEC) capability.
3661 * The port identifier of the Ethernet device.
3662 * @param speed_fec_capa
3663 * speed_fec_capa is out only with per-speed capabilities.
3664 * If set to NULL, the function returns the required number
3665 * of required array entries.
3667 * a number of elements in an speed_fec_capa array.
3670 * - A non-negative value lower or equal to num: success. The return value
3671 * is the number of entries filled in the fec capa array.
3672 * - A non-negative value higher than num: error, the given fec capa array
3673 * is too small. The return value corresponds to the num that should
3674 * be given to succeed. The entries in fec capa array are not valid and
3675 * shall not be used by the caller.
3676 * - (-ENOTSUP) if underlying hardware OR driver doesn't support.
3678 * - (-EIO) if device is removed.
3679 * - (-ENODEV) if *port_id* invalid.
3680 * - (-EINVAL) if *num* or *speed_fec_capa* invalid
3683 int rte_eth_fec_get_capability(uint16_t port_id,
3684 struct rte_eth_fec_capa *speed_fec_capa,
3689 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
3691 * Get current Forward Error Correction(FEC) mode.
3692 * If link is down and AUTO is enabled, AUTO is returned, otherwise,
3693 * configured FEC mode is returned.
3694 * If link is up, current FEC mode is returned.
3697 * The port identifier of the Ethernet device.
3699 * A bitmask of enabled FEC modes. If AUTO bit is set, other
3700 * bits specify FEC modes which may be negotiated. If AUTO
3701 * bit is clear, specify FEC modes to be used (only one valid
3702 * mode per speed may be set).
3704 * - (0) if successful.
3705 * - (-ENOTSUP) if underlying hardware OR driver doesn't support.
3707 * - (-EIO) if device is removed.
3708 * - (-ENODEV) if *port_id* invalid.
3711 int rte_eth_fec_get(uint16_t port_id, uint32_t *fec_capa);
3715 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
3717 * Set Forward Error Correction(FEC) mode.
3720 * The port identifier of the Ethernet device.
3722 * A bitmask of allowed FEC modes. If AUTO bit is set, other
3723 * bits specify FEC modes which may be negotiated. If AUTO
3724 * bit is clear, specify FEC modes to be used (only one valid
3725 * mode per speed may be set).
3727 * - (0) if successful.
3728 * - (-EINVAL) if the FEC mode is not valid.
3729 * - (-ENOTSUP) if underlying hardware OR driver doesn't support.
3730 * - (-EIO) if device is removed.
3731 * - (-ENODEV) if *port_id* invalid.
3734 int rte_eth_fec_set(uint16_t port_id, uint32_t fec_capa);
3737 * Get current status of the Ethernet link flow control for Ethernet device
3740 * The port identifier of the Ethernet device.
3742 * The pointer to the structure where to store the flow control parameters.
3744 * - (0) if successful.
3745 * - (-ENOTSUP) if hardware doesn't support flow control.
3746 * - (-ENODEV) if *port_id* invalid.
3747 * - (-EIO) if device is removed.
3749 int rte_eth_dev_flow_ctrl_get(uint16_t port_id,
3750 struct rte_eth_fc_conf *fc_conf);
3753 * Configure the Ethernet link flow control for Ethernet device
3756 * The port identifier of the Ethernet device.
3758 * The pointer to the structure of the flow control parameters.
3760 * - (0) if successful.
3761 * - (-ENOTSUP) if hardware doesn't support flow control mode.
3762 * - (-ENODEV) if *port_id* invalid.
3763 * - (-EINVAL) if bad parameter
3764 * - (-EIO) if flow control setup failure or device is removed.
3766 int rte_eth_dev_flow_ctrl_set(uint16_t port_id,
3767 struct rte_eth_fc_conf *fc_conf);
3770 * Configure the Ethernet priority flow control under DCB environment
3771 * for Ethernet device.
3774 * The port identifier of the Ethernet device.
3776 * The pointer to the structure of the priority flow control parameters.
3778 * - (0) if successful.
3779 * - (-ENOTSUP) if hardware doesn't support priority flow control mode.
3780 * - (-ENODEV) if *port_id* invalid.
3781 * - (-EINVAL) if bad parameter
3782 * - (-EIO) if flow control setup failure or device is removed.
3784 int rte_eth_dev_priority_flow_ctrl_set(uint16_t port_id,
3785 struct rte_eth_pfc_conf *pfc_conf);
3788 * Add a MAC address to the set used for filtering incoming packets.
3791 * The port identifier of the Ethernet device.
3793 * The MAC address to add.
3795 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is
3796 * not enabled, this should be set to 0.
3798 * - (0) if successfully added or *mac_addr* was already added.
3799 * - (-ENOTSUP) if hardware doesn't support this feature.
3800 * - (-ENODEV) if *port* is invalid.
3801 * - (-EIO) if device is removed.
3802 * - (-ENOSPC) if no more MAC addresses can be added.
3803 * - (-EINVAL) if MAC address is invalid.
3805 int rte_eth_dev_mac_addr_add(uint16_t port_id, struct rte_ether_addr *mac_addr,
3809 * Remove a MAC address from the internal array of addresses.
3812 * The port identifier of the Ethernet device.
3814 * MAC address to remove.
3816 * - (0) if successful, or *mac_addr* didn't exist.
3817 * - (-ENOTSUP) if hardware doesn't support.
3818 * - (-ENODEV) if *port* invalid.
3819 * - (-EADDRINUSE) if attempting to remove the default MAC address
3821 int rte_eth_dev_mac_addr_remove(uint16_t port_id,
3822 struct rte_ether_addr *mac_addr);
3825 * Set the default MAC address.
3828 * The port identifier of the Ethernet device.
3830 * New default MAC address.
3832 * - (0) if successful, or *mac_addr* didn't exist.
3833 * - (-ENOTSUP) if hardware doesn't support.
3834 * - (-ENODEV) if *port* invalid.
3835 * - (-EINVAL) if MAC address is invalid.
3837 int rte_eth_dev_default_mac_addr_set(uint16_t port_id,
3838 struct rte_ether_addr *mac_addr);
3841 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3844 * The port identifier of the Ethernet device.
3848 * Redirection table size. The table size can be queried by
3849 * rte_eth_dev_info_get().
3851 * - (0) if successful.
3852 * - (-ENODEV) if *port_id* is invalid.
3853 * - (-ENOTSUP) if hardware doesn't support.
3854 * - (-EINVAL) if bad parameter.
3855 * - (-EIO) if device is removed.
3857 int rte_eth_dev_rss_reta_update(uint16_t port_id,
3858 struct rte_eth_rss_reta_entry64 *reta_conf,
3859 uint16_t reta_size);
3862 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3865 * The port identifier of the Ethernet device.
3867 * RETA to query. For each requested reta entry, corresponding bit
3868 * in mask must be set.
3870 * Redirection table size. The table size can be queried by
3871 * rte_eth_dev_info_get().
3873 * - (0) if successful.
3874 * - (-ENODEV) if *port_id* is invalid.
3875 * - (-ENOTSUP) if hardware doesn't support.
3876 * - (-EINVAL) if bad parameter.
3877 * - (-EIO) if device is removed.
3879 int rte_eth_dev_rss_reta_query(uint16_t port_id,
3880 struct rte_eth_rss_reta_entry64 *reta_conf,
3881 uint16_t reta_size);
3884 * Updates unicast hash table for receiving packet with the given destination
3885 * MAC address, and the packet is routed to all VFs for which the RX mode is
3886 * accept packets that match the unicast hash table.
3889 * The port identifier of the Ethernet device.
3891 * Unicast MAC address.
3893 * 1 - Set an unicast hash bit for receiving packets with the MAC address.
3894 * 0 - Clear an unicast hash bit.
3896 * - (0) if successful.
3897 * - (-ENOTSUP) if hardware doesn't support.
3898 * - (-ENODEV) if *port_id* invalid.
3899 * - (-EIO) if device is removed.
3900 * - (-EINVAL) if bad parameter.
3902 int rte_eth_dev_uc_hash_table_set(uint16_t port_id, struct rte_ether_addr *addr,
3906 * Updates all unicast hash bitmaps for receiving packet with any Unicast
3907 * Ethernet MAC addresses,the packet is routed to all VFs for which the RX
3908 * mode is accept packets that match the unicast hash table.
3911 * The port identifier of the Ethernet device.
3913 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet
3915 * 0 - Clear all unicast hash bitmaps
3917 * - (0) if successful.
3918 * - (-ENOTSUP) if hardware doesn't support.
3919 * - (-ENODEV) if *port_id* invalid.
3920 * - (-EIO) if device is removed.
3921 * - (-EINVAL) if bad parameter.
3923 int rte_eth_dev_uc_all_hash_table_set(uint16_t port_id, uint8_t on);
3926 * Set a traffic mirroring rule on an Ethernet device
3929 * The port identifier of the Ethernet device.
3930 * @param mirror_conf
3931 * The pointer to the traffic mirroring structure describing the mirroring rule.
3932 * The *rte_eth_vm_mirror_conf* structure includes the type of mirroring rule,
3933 * destination pool and the value of rule if enable vlan or pool mirroring.
3936 * The index of traffic mirroring rule, we support four separated rules.
3938 * 1 - Enable a mirroring rule.
3939 * 0 - Disable a mirroring rule.
3941 * - (0) if successful.
3942 * - (-ENOTSUP) if hardware doesn't support this feature.
3943 * - (-ENODEV) if *port_id* invalid.
3944 * - (-EIO) if device is removed.
3945 * - (-EINVAL) if the mr_conf information is not correct.
3947 int rte_eth_mirror_rule_set(uint16_t port_id,
3948 struct rte_eth_mirror_conf *mirror_conf,
3953 * Reset a traffic mirroring rule on an Ethernet device.
3956 * The port identifier of the Ethernet device.
3958 * The index of traffic mirroring rule, we support four separated rules.
3960 * - (0) if successful.
3961 * - (-ENOTSUP) if hardware doesn't support this feature.
3962 * - (-ENODEV) if *port_id* invalid.
3963 * - (-EIO) if device is removed.
3964 * - (-EINVAL) if bad parameter.
3966 int rte_eth_mirror_rule_reset(uint16_t port_id,
3970 * Set the rate limitation for a queue on an Ethernet device.
3973 * The port identifier of the Ethernet device.
3977 * The tx rate in Mbps. Allocated from the total port link speed.
3979 * - (0) if successful.
3980 * - (-ENOTSUP) if hardware doesn't support this feature.
3981 * - (-ENODEV) if *port_id* invalid.
3982 * - (-EIO) if device is removed.
3983 * - (-EINVAL) if bad parameter.
3985 int rte_eth_set_queue_rate_limit(uint16_t port_id, uint16_t queue_idx,
3989 * Configuration of Receive Side Scaling hash computation of Ethernet device.
3992 * The port identifier of the Ethernet device.
3994 * The new configuration to use for RSS hash computation on the port.
3996 * - (0) if successful.
3997 * - (-ENODEV) if port identifier is invalid.
3998 * - (-EIO) if device is removed.
3999 * - (-ENOTSUP) if hardware doesn't support.
4000 * - (-EINVAL) if bad parameter.
4002 int rte_eth_dev_rss_hash_update(uint16_t port_id,
4003 struct rte_eth_rss_conf *rss_conf);
4006 * Retrieve current configuration of Receive Side Scaling hash computation
4007 * of Ethernet device.
4010 * The port identifier of the Ethernet device.
4012 * Where to store the current RSS hash configuration of the Ethernet device.
4014 * - (0) if successful.
4015 * - (-ENODEV) if port identifier is invalid.
4016 * - (-EIO) if device is removed.
4017 * - (-ENOTSUP) if hardware doesn't support RSS.
4020 rte_eth_dev_rss_hash_conf_get(uint16_t port_id,
4021 struct rte_eth_rss_conf *rss_conf);
4024 * Add UDP tunneling port for a specific type of tunnel.
4025 * The packets with this UDP port will be identified as this type of tunnel.
4026 * Before enabling any offloading function for a tunnel, users can call this API
4027 * to change or add more UDP port for the tunnel. So the offloading function
4028 * can take effect on the packets with the specific UDP port.
4031 * The port identifier of the Ethernet device.
4033 * UDP tunneling configuration.
4036 * - (0) if successful.
4037 * - (-ENODEV) if port identifier is invalid.
4038 * - (-EIO) if device is removed.
4039 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4042 rte_eth_dev_udp_tunnel_port_add(uint16_t port_id,
4043 struct rte_eth_udp_tunnel *tunnel_udp);
4046 * Delete UDP tunneling port a specific type of tunnel.
4047 * The packets with this UDP port will not be identified as this type of tunnel
4049 * Before enabling any offloading function for a tunnel, users can call this API
4050 * to delete a UDP port for the tunnel. So the offloading function will not take
4051 * effect on the packets with the specific UDP port.
4054 * The port identifier of the Ethernet device.
4056 * UDP tunneling configuration.
4059 * - (0) if successful.
4060 * - (-ENODEV) if port identifier is invalid.
4061 * - (-EIO) if device is removed.
4062 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4065 rte_eth_dev_udp_tunnel_port_delete(uint16_t port_id,
4066 struct rte_eth_udp_tunnel *tunnel_udp);
4069 * Check whether the filter type is supported on an Ethernet device.
4070 * All the supported filter types are defined in 'rte_eth_ctrl.h'.
4073 * The port identifier of the Ethernet device.
4074 * @param filter_type
4077 * - (0) if successful.
4078 * - (-ENOTSUP) if hardware doesn't support this filter type.
4079 * - (-ENODEV) if *port_id* invalid.
4080 * - (-EIO) if device is removed.
4083 int rte_eth_dev_filter_supported(uint16_t port_id,
4084 enum rte_filter_type filter_type);
4087 * Take operations to assigned filter type on an Ethernet device.
4088 * All the supported operations and filter types are defined in 'rte_eth_ctrl.h'.
4091 * The port identifier of the Ethernet device.
4092 * @param filter_type
4095 * Type of operation.
4097 * A pointer to arguments defined specifically for the operation.
4099 * - (0) if successful.
4100 * - (-ENOTSUP) if hardware doesn't support.
4101 * - (-ENODEV) if *port_id* invalid.
4102 * - (-EIO) if device is removed.
4103 * - others depends on the specific operations implementation.
4106 int rte_eth_dev_filter_ctrl(uint16_t port_id, enum rte_filter_type filter_type,
4107 enum rte_filter_op filter_op, void *arg);
4110 * Get DCB information on an Ethernet device.
4113 * The port identifier of the Ethernet device.
4117 * - (0) if successful.
4118 * - (-ENODEV) if port identifier is invalid.
4119 * - (-EIO) if device is removed.
4120 * - (-ENOTSUP) if hardware doesn't support.
4122 int rte_eth_dev_get_dcb_info(uint16_t port_id,
4123 struct rte_eth_dcb_info *dcb_info);
4125 struct rte_eth_rxtx_callback;
4128 * Add a callback to be called on packet RX on a given port and queue.
4130 * This API configures a function to be called for each burst of
4131 * packets received on a given NIC port queue. The return value is a pointer
4132 * that can be used to later remove the callback using
4133 * rte_eth_remove_rx_callback().
4135 * Multiple functions are called in the order that they are added.
4138 * The port identifier of the Ethernet device.
4140 * The queue on the Ethernet device on which the callback is to be added.
4142 * The callback function
4144 * A generic pointer parameter which will be passed to each invocation of the
4145 * callback function on this port and queue. Inter-thread synchronization
4146 * of any user data changes is the responsibility of the user.
4150 * On success, a pointer value which can later be used to remove the callback.
4152 const struct rte_eth_rxtx_callback *
4153 rte_eth_add_rx_callback(uint16_t port_id, uint16_t queue_id,
4154 rte_rx_callback_fn fn, void *user_param);
4157 * Add a callback that must be called first on packet RX on a given port
4160 * This API configures a first function to be called for each burst of
4161 * packets received on a given NIC port queue. The return value is a pointer
4162 * that can be used to later remove the callback using
4163 * rte_eth_remove_rx_callback().
4165 * Multiple functions are called in the order that they are added.
4168 * The port identifier of the Ethernet device.
4170 * The queue on the Ethernet device on which the callback is to be added.
4172 * The callback function
4174 * A generic pointer parameter which will be passed to each invocation of the
4175 * callback function on this port and queue. Inter-thread synchronization
4176 * of any user data changes is the responsibility of the user.
4180 * On success, a pointer value which can later be used to remove the callback.
4182 const struct rte_eth_rxtx_callback *
4183 rte_eth_add_first_rx_callback(uint16_t port_id, uint16_t queue_id,
4184 rte_rx_callback_fn fn, void *user_param);
4187 * Add a callback to be called on packet TX on a given port and queue.
4189 * This API configures a function to be called for each burst of
4190 * packets sent on a given NIC port queue. The return value is a pointer
4191 * that can be used to later remove the callback using
4192 * rte_eth_remove_tx_callback().
4194 * Multiple functions are called in the order that they are added.
4197 * The port identifier of the Ethernet device.
4199 * The queue on the Ethernet device on which the callback is to be added.
4201 * The callback function
4203 * A generic pointer parameter which will be passed to each invocation of the
4204 * callback function on this port and queue. Inter-thread synchronization
4205 * of any user data changes is the responsibility of the user.
4209 * On success, a pointer value which can later be used to remove the callback.
4211 const struct rte_eth_rxtx_callback *
4212 rte_eth_add_tx_callback(uint16_t port_id, uint16_t queue_id,
4213 rte_tx_callback_fn fn, void *user_param);
4216 * Remove an RX packet callback from a given port and queue.
4218 * This function is used to removed callbacks that were added to a NIC port
4219 * queue using rte_eth_add_rx_callback().
4221 * Note: the callback is removed from the callback list but it isn't freed
4222 * since the it may still be in use. The memory for the callback can be
4223 * subsequently freed back by the application by calling rte_free():
4225 * - Immediately - if the port is stopped, or the user knows that no
4226 * callbacks are in flight e.g. if called from the thread doing RX/TX
4229 * - After a short delay - where the delay is sufficient to allow any
4230 * in-flight callbacks to complete. Alternately, the RCU mechanism can be
4231 * used to detect when data plane threads have ceased referencing the
4235 * The port identifier of the Ethernet device.
4237 * The queue on the Ethernet device from which the callback is to be removed.
4239 * User supplied callback created via rte_eth_add_rx_callback().
4242 * - 0: Success. Callback was removed.
4243 * - -ENODEV: If *port_id* is invalid.
4244 * - -ENOTSUP: Callback support is not available.
4245 * - -EINVAL: The queue_id is out of range, or the callback
4246 * is NULL or not found for the port/queue.
4248 int rte_eth_remove_rx_callback(uint16_t port_id, uint16_t queue_id,
4249 const struct rte_eth_rxtx_callback *user_cb);
4252 * Remove a TX packet callback from a given port and queue.
4254 * This function is used to removed callbacks that were added to a NIC port
4255 * queue using rte_eth_add_tx_callback().
4257 * Note: the callback is removed from the callback list but it isn't freed
4258 * since the it may still be in use. The memory for the callback can be
4259 * subsequently freed back by the application by calling rte_free():
4261 * - Immediately - if the port is stopped, or the user knows that no
4262 * callbacks are in flight e.g. if called from the thread doing RX/TX
4265 * - After a short delay - where the delay is sufficient to allow any
4266 * in-flight callbacks to complete. Alternately, the RCU mechanism can be
4267 * used to detect when data plane threads have ceased referencing the
4271 * The port identifier of the Ethernet device.
4273 * The queue on the Ethernet device from which the callback is to be removed.
4275 * User supplied callback created via rte_eth_add_tx_callback().
4278 * - 0: Success. Callback was removed.
4279 * - -ENODEV: If *port_id* is invalid.
4280 * - -ENOTSUP: Callback support is not available.
4281 * - -EINVAL: The queue_id is out of range, or the callback
4282 * is NULL or not found for the port/queue.
4284 int rte_eth_remove_tx_callback(uint16_t port_id, uint16_t queue_id,
4285 const struct rte_eth_rxtx_callback *user_cb);
4288 * Retrieve information about given port's RX queue.
4291 * The port identifier of the Ethernet device.
4293 * The RX queue on the Ethernet device for which information
4294 * will be retrieved.
4296 * A pointer to a structure of type *rte_eth_rxq_info_info* to be filled with
4297 * the information of the Ethernet device.
4301 * - -ENODEV: If *port_id* is invalid.
4302 * - -ENOTSUP: routine is not supported by the device PMD.
4303 * - -EINVAL: The queue_id is out of range, or the queue
4306 int rte_eth_rx_queue_info_get(uint16_t port_id, uint16_t queue_id,
4307 struct rte_eth_rxq_info *qinfo);
4310 * Retrieve information about given port's TX queue.
4313 * The port identifier of the Ethernet device.
4315 * The TX queue on the Ethernet device for which information
4316 * will be retrieved.
4318 * A pointer to a structure of type *rte_eth_txq_info_info* to be filled with
4319 * the information of the Ethernet device.
4323 * - -ENODEV: If *port_id* is invalid.
4324 * - -ENOTSUP: routine is not supported by the device PMD.
4325 * - -EINVAL: The queue_id is out of range, or the queue
4328 int rte_eth_tx_queue_info_get(uint16_t port_id, uint16_t queue_id,
4329 struct rte_eth_txq_info *qinfo);
4332 * Retrieve information about the Rx packet burst mode.
4335 * The port identifier of the Ethernet device.
4337 * The Rx queue on the Ethernet device for which information
4338 * will be retrieved.
4340 * A pointer to a structure of type *rte_eth_burst_mode* to be filled
4341 * with the information of the packet burst mode.
4345 * - -ENODEV: If *port_id* is invalid.
4346 * - -ENOTSUP: routine is not supported by the device PMD.
4347 * - -EINVAL: The queue_id is out of range.
4350 int rte_eth_rx_burst_mode_get(uint16_t port_id, uint16_t queue_id,
4351 struct rte_eth_burst_mode *mode);
4354 * Retrieve information about the Tx packet burst mode.
4357 * The port identifier of the Ethernet device.
4359 * The Tx queue on the Ethernet device for which information
4360 * will be retrieved.
4362 * A pointer to a structure of type *rte_eth_burst_mode* to be filled
4363 * with the information of the packet burst mode.
4367 * - -ENODEV: If *port_id* is invalid.
4368 * - -ENOTSUP: routine is not supported by the device PMD.
4369 * - -EINVAL: The queue_id is out of range.
4372 int rte_eth_tx_burst_mode_get(uint16_t port_id, uint16_t queue_id,
4373 struct rte_eth_burst_mode *mode);
4376 * Retrieve device registers and register attributes (number of registers and
4380 * The port identifier of the Ethernet device.
4382 * Pointer to rte_dev_reg_info structure to fill in. If info->data is
4383 * NULL the function fills in the width and length fields. If non-NULL
4384 * the registers are put into the buffer pointed at by the data field.
4386 * - (0) if successful.
4387 * - (-ENOTSUP) if hardware doesn't support.
4388 * - (-ENODEV) if *port_id* invalid.
4389 * - (-EIO) if device is removed.
4390 * - others depends on the specific operations implementation.
4392 int rte_eth_dev_get_reg_info(uint16_t port_id, struct rte_dev_reg_info *info);
4395 * Retrieve size of device EEPROM
4398 * The port identifier of the Ethernet device.
4400 * - (>=0) EEPROM size if successful.
4401 * - (-ENOTSUP) if hardware doesn't support.
4402 * - (-ENODEV) if *port_id* invalid.
4403 * - (-EIO) if device is removed.
4404 * - others depends on the specific operations implementation.
4406 int rte_eth_dev_get_eeprom_length(uint16_t port_id);
4409 * Retrieve EEPROM and EEPROM attribute
4412 * The port identifier of the Ethernet device.
4414 * The template includes buffer for return EEPROM data and
4415 * EEPROM attributes to be filled.
4417 * - (0) if successful.
4418 * - (-ENOTSUP) if hardware doesn't support.
4419 * - (-ENODEV) if *port_id* invalid.
4420 * - (-EIO) if device is removed.
4421 * - others depends on the specific operations implementation.
4423 int rte_eth_dev_get_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
4426 * Program EEPROM with provided data
4429 * The port identifier of the Ethernet device.
4431 * The template includes EEPROM data for programming and
4432 * EEPROM attributes to be filled
4434 * - (0) if successful.
4435 * - (-ENOTSUP) if hardware doesn't support.
4436 * - (-ENODEV) if *port_id* invalid.
4437 * - (-EIO) if device is removed.
4438 * - others depends on the specific operations implementation.
4440 int rte_eth_dev_set_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
4444 * @b EXPERIMENTAL: this API may change without prior notice.
4446 * Retrieve the type and size of plugin module EEPROM
4449 * The port identifier of the Ethernet device.
4451 * The type and size of plugin module EEPROM.
4453 * - (0) if successful.
4454 * - (-ENOTSUP) if hardware doesn't support.
4455 * - (-ENODEV) if *port_id* invalid.
4456 * - (-EIO) if device is removed.
4457 * - others depends on the specific operations implementation.
4461 rte_eth_dev_get_module_info(uint16_t port_id,
4462 struct rte_eth_dev_module_info *modinfo);
4466 * @b EXPERIMENTAL: this API may change without prior notice.
4468 * Retrieve the data of plugin module EEPROM
4471 * The port identifier of the Ethernet device.
4473 * The template includes the plugin module EEPROM attributes, and the
4474 * buffer for return plugin module EEPROM data.
4476 * - (0) if successful.
4477 * - (-ENOTSUP) if hardware doesn't support.
4478 * - (-ENODEV) if *port_id* invalid.
4479 * - (-EIO) if device is removed.
4480 * - others depends on the specific operations implementation.
4484 rte_eth_dev_get_module_eeprom(uint16_t port_id,
4485 struct rte_dev_eeprom_info *info);
4488 * Set the list of multicast addresses to filter on an Ethernet device.
4491 * The port identifier of the Ethernet device.
4492 * @param mc_addr_set
4493 * The array of multicast addresses to set. Equal to NULL when the function
4494 * is invoked to flush the set of filtered addresses.
4496 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0
4497 * when the function is invoked to flush the set of filtered addresses.
4499 * - (0) if successful.
4500 * - (-ENODEV) if *port_id* invalid.
4501 * - (-EIO) if device is removed.
4502 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering.
4503 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources.
4505 int rte_eth_dev_set_mc_addr_list(uint16_t port_id,
4506 struct rte_ether_addr *mc_addr_set,
4507 uint32_t nb_mc_addr);
4510 * Enable IEEE1588/802.1AS timestamping for an Ethernet device.
4513 * The port identifier of the Ethernet device.
4517 * - -ENODEV: The port ID is invalid.
4518 * - -EIO: if device is removed.
4519 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4521 int rte_eth_timesync_enable(uint16_t port_id);
4524 * Disable IEEE1588/802.1AS timestamping for an Ethernet device.
4527 * The port identifier of the Ethernet device.
4531 * - -ENODEV: The port ID is invalid.
4532 * - -EIO: if device is removed.
4533 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4535 int rte_eth_timesync_disable(uint16_t port_id);
4538 * Read an IEEE1588/802.1AS RX timestamp from an Ethernet device.
4541 * The port identifier of the Ethernet device.
4543 * Pointer to the timestamp struct.
4545 * Device specific flags. Used to pass the RX timesync register index to
4546 * i40e. Unused in igb/ixgbe, pass 0 instead.
4550 * - -EINVAL: No timestamp is available.
4551 * - -ENODEV: The port ID is invalid.
4552 * - -EIO: if device is removed.
4553 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4555 int rte_eth_timesync_read_rx_timestamp(uint16_t port_id,
4556 struct timespec *timestamp, uint32_t flags);
4559 * Read an IEEE1588/802.1AS TX timestamp from an Ethernet device.
4562 * The port identifier of the Ethernet device.
4564 * Pointer to the timestamp struct.
4568 * - -EINVAL: No timestamp is available.
4569 * - -ENODEV: The port ID is invalid.
4570 * - -EIO: if device is removed.
4571 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4573 int rte_eth_timesync_read_tx_timestamp(uint16_t port_id,
4574 struct timespec *timestamp);
4577 * Adjust the timesync clock on an Ethernet device.
4579 * This is usually used in conjunction with other Ethdev timesync functions to
4580 * synchronize the device time using the IEEE1588/802.1AS protocol.
4583 * The port identifier of the Ethernet device.
4585 * The adjustment in nanoseconds.
4589 * - -ENODEV: The port ID is invalid.
4590 * - -EIO: if device is removed.
4591 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4593 int rte_eth_timesync_adjust_time(uint16_t port_id, int64_t delta);
4596 * Read the time from the timesync clock on an Ethernet device.
4598 * This is usually used in conjunction with other Ethdev timesync functions to
4599 * synchronize the device time using the IEEE1588/802.1AS protocol.
4602 * The port identifier of the Ethernet device.
4604 * Pointer to the timespec struct that holds the time.
4609 int rte_eth_timesync_read_time(uint16_t port_id, struct timespec *time);
4612 * Set the time of the timesync clock on an Ethernet device.
4614 * This is usually used in conjunction with other Ethdev timesync functions to
4615 * synchronize the device time using the IEEE1588/802.1AS protocol.
4618 * The port identifier of the Ethernet device.
4620 * Pointer to the timespec struct that holds the time.
4624 * - -EINVAL: No timestamp is available.
4625 * - -ENODEV: The port ID is invalid.
4626 * - -EIO: if device is removed.
4627 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4629 int rte_eth_timesync_write_time(uint16_t port_id, const struct timespec *time);
4633 * @b EXPERIMENTAL: this API may change without prior notice.
4635 * Read the current clock counter of an Ethernet device
4637 * This returns the current raw clock value of an Ethernet device. It is
4638 * a raw amount of ticks, with no given time reference.
4639 * The value returned here is from the same clock than the one
4640 * filling timestamp field of Rx packets when using hardware timestamp
4641 * offload. Therefore it can be used to compute a precise conversion of
4642 * the device clock to the real time.
4644 * E.g, a simple heuristic to derivate the frequency would be:
4645 * uint64_t start, end;
4646 * rte_eth_read_clock(port, start);
4647 * rte_delay_ms(100);
4648 * rte_eth_read_clock(port, end);
4649 * double freq = (end - start) * 10;
4651 * Compute a common reference with:
4652 * uint64_t base_time_sec = current_time();
4653 * uint64_t base_clock;
4654 * rte_eth_read_clock(port, base_clock);
4656 * Then, convert the raw mbuf timestamp with:
4657 * base_time_sec + (double)(*timestamp_dynfield(mbuf) - base_clock) / freq;
4659 * This simple example will not provide a very good accuracy. One must
4660 * at least measure multiple times the frequency and do a regression.
4661 * To avoid deviation from the system time, the common reference can
4662 * be repeated from time to time. The integer division can also be
4663 * converted by a multiplication and a shift for better performance.
4666 * The port identifier of the Ethernet device.
4668 * Pointer to the uint64_t that holds the raw clock value.
4672 * - -ENODEV: The port ID is invalid.
4673 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4677 rte_eth_read_clock(uint16_t port_id, uint64_t *clock);
4680 * Config l2 tunnel ether type of an Ethernet device for filtering specific
4681 * tunnel packets by ether type.
4684 * The port identifier of the Ethernet device.
4686 * l2 tunnel configuration.
4689 * - (0) if successful.
4690 * - (-ENODEV) if port identifier is invalid.
4691 * - (-EIO) if device is removed.
4692 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4695 rte_eth_dev_l2_tunnel_eth_type_conf(uint16_t port_id,
4696 struct rte_eth_l2_tunnel_conf *l2_tunnel);
4699 * Enable/disable l2 tunnel offload functions. Include,
4700 * 1, The ability of parsing a type of l2 tunnel of an Ethernet device.
4701 * Filtering, forwarding and offloading this type of tunnel packets depend on
4703 * 2, Stripping the l2 tunnel tag.
4704 * 3, Insertion of the l2 tunnel tag.
4705 * 4, Forwarding the packets based on the l2 tunnel tag.
4708 * The port identifier of the Ethernet device.
4710 * l2 tunnel parameters.
4712 * Indicate the offload function.
4714 * Enable or disable this function.
4717 * - (0) if successful.
4718 * - (-ENODEV) if port identifier is invalid.
4719 * - (-EIO) if device is removed.
4720 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4723 rte_eth_dev_l2_tunnel_offload_set(uint16_t port_id,
4724 struct rte_eth_l2_tunnel_conf *l2_tunnel,
4729 * Get the port id from device name. The device name should be specified
4731 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:2:00.0
4732 * - SoC device name, for example- fsl-gmac0
4733 * - vdev dpdk name, for example- net_[pcap0|null0|tap0]
4736 * pci address or name of the device
4738 * pointer to port identifier of the device
4740 * - (0) if successful and port_id is filled.
4741 * - (-ENODEV or -EINVAL) on failure.
4744 rte_eth_dev_get_port_by_name(const char *name, uint16_t *port_id);
4747 * Get the device name from port id. The device name is specified as below:
4748 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:02:00.0
4749 * - SoC device name, for example- fsl-gmac0
4750 * - vdev dpdk name, for example- net_[pcap0|null0|tun0|tap0]
4753 * Port identifier of the device.
4755 * Buffer of size RTE_ETH_NAME_MAX_LEN to store the name.
4757 * - (0) if successful.
4758 * - (-ENODEV) if *port_id* is invalid.
4759 * - (-EINVAL) on failure.
4762 rte_eth_dev_get_name_by_port(uint16_t port_id, char *name);
4765 * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from
4766 * the ethernet device information, otherwise adjust them to boundaries.
4769 * The port identifier of the Ethernet device.
4771 * A pointer to a uint16_t where the number of receive
4772 * descriptors stored.
4774 * A pointer to a uint16_t where the number of transmit
4775 * descriptors stored.
4777 * - (0) if successful.
4778 * - (-ENOTSUP, -ENODEV or -EINVAL) on failure.
4780 int rte_eth_dev_adjust_nb_rx_tx_desc(uint16_t port_id,
4781 uint16_t *nb_rx_desc,
4782 uint16_t *nb_tx_desc);
4785 * Test if a port supports specific mempool ops.
4788 * Port identifier of the Ethernet device.
4790 * The name of the pool operations to test.
4792 * - 0: best mempool ops choice for this port.
4793 * - 1: mempool ops are supported for this port.
4794 * - -ENOTSUP: mempool ops not supported for this port.
4795 * - -ENODEV: Invalid port Identifier.
4796 * - -EINVAL: Pool param is null.
4799 rte_eth_dev_pool_ops_supported(uint16_t port_id, const char *pool);
4802 * Get the security context for the Ethernet device.
4805 * Port identifier of the Ethernet device
4808 * - pointer to security context on success.
4811 rte_eth_dev_get_sec_ctx(uint16_t port_id);
4815 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
4817 * Query the device hairpin capabilities.
4820 * The port identifier of the Ethernet device.
4822 * Pointer to a structure that will hold the hairpin capabilities.
4824 * - (0) if successful.
4825 * - (-ENOTSUP) if hardware doesn't support.
4828 int rte_eth_dev_hairpin_capability_get(uint16_t port_id,
4829 struct rte_eth_hairpin_cap *cap);
4831 #include <rte_ethdev_core.h>
4835 * Retrieve a burst of input packets from a receive queue of an Ethernet
4836 * device. The retrieved packets are stored in *rte_mbuf* structures whose
4837 * pointers are supplied in the *rx_pkts* array.
4839 * The rte_eth_rx_burst() function loops, parsing the RX ring of the
4840 * receive queue, up to *nb_pkts* packets, and for each completed RX
4841 * descriptor in the ring, it performs the following operations:
4843 * - Initialize the *rte_mbuf* data structure associated with the
4844 * RX descriptor according to the information provided by the NIC into
4845 * that RX descriptor.
4847 * - Store the *rte_mbuf* data structure into the next entry of the
4850 * - Replenish the RX descriptor with a new *rte_mbuf* buffer
4851 * allocated from the memory pool associated with the receive queue at
4852 * initialization time.
4854 * When retrieving an input packet that was scattered by the controller
4855 * into multiple receive descriptors, the rte_eth_rx_burst() function
4856 * appends the associated *rte_mbuf* buffers to the first buffer of the
4859 * The rte_eth_rx_burst() function returns the number of packets
4860 * actually retrieved, which is the number of *rte_mbuf* data structures
4861 * effectively supplied into the *rx_pkts* array.
4862 * A return value equal to *nb_pkts* indicates that the RX queue contained
4863 * at least *rx_pkts* packets, and this is likely to signify that other
4864 * received packets remain in the input queue. Applications implementing
4865 * a "retrieve as much received packets as possible" policy can check this
4866 * specific case and keep invoking the rte_eth_rx_burst() function until
4867 * a value less than *nb_pkts* is returned.
4869 * This receive method has the following advantages:
4871 * - It allows a run-to-completion network stack engine to retrieve and
4872 * to immediately process received packets in a fast burst-oriented
4873 * approach, avoiding the overhead of unnecessary intermediate packet
4874 * queue/dequeue operations.
4876 * - Conversely, it also allows an asynchronous-oriented processing
4877 * method to retrieve bursts of received packets and to immediately
4878 * queue them for further parallel processing by another logical core,
4879 * for instance. However, instead of having received packets being
4880 * individually queued by the driver, this approach allows the caller
4881 * of the rte_eth_rx_burst() function to queue a burst of retrieved
4882 * packets at a time and therefore dramatically reduce the cost of
4883 * enqueue/dequeue operations per packet.
4885 * - It allows the rte_eth_rx_burst() function of the driver to take
4886 * advantage of burst-oriented hardware features (CPU cache,
4887 * prefetch instructions, and so on) to minimize the number of CPU
4888 * cycles per packet.
4890 * To summarize, the proposed receive API enables many
4891 * burst-oriented optimizations in both synchronous and asynchronous
4892 * packet processing environments with no overhead in both cases.
4894 * The rte_eth_rx_burst() function does not provide any error
4895 * notification to avoid the corresponding overhead. As a hint, the
4896 * upper-level application might check the status of the device link once
4897 * being systematically returned a 0 value for a given number of tries.
4900 * The port identifier of the Ethernet device.
4902 * The index of the receive queue from which to retrieve input packets.
4903 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
4904 * to rte_eth_dev_configure().
4906 * The address of an array of pointers to *rte_mbuf* structures that
4907 * must be large enough to store *nb_pkts* pointers in it.
4909 * The maximum number of packets to retrieve.
4911 * The number of packets actually retrieved, which is the number
4912 * of pointers to *rte_mbuf* structures effectively supplied to the
4915 static inline uint16_t
4916 rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id,
4917 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
4919 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
4922 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4923 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
4924 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_pkt_burst, 0);
4926 if (queue_id >= dev->data->nb_rx_queues) {
4927 RTE_ETHDEV_LOG(ERR, "Invalid RX queue_id=%u\n", queue_id);
4931 nb_rx = (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id],
4934 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
4935 struct rte_eth_rxtx_callback *cb;
4937 /* __ATOMIC_RELEASE memory order was used when the
4938 * call back was inserted into the list.
4939 * Since there is a clear dependency between loading
4940 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is
4943 cb = __atomic_load_n(&dev->post_rx_burst_cbs[queue_id],
4946 if (unlikely(cb != NULL)) {
4948 nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
4949 nb_pkts, cb->param);
4951 } while (cb != NULL);
4955 rte_ethdev_trace_rx_burst(port_id, queue_id, (void **)rx_pkts, nb_rx);
4960 * Get the number of used descriptors of a rx queue
4963 * The port identifier of the Ethernet device.
4965 * The queue id on the specific port.
4967 * The number of used descriptors in the specific queue, or:
4968 * - (-ENODEV) if *port_id* is invalid.
4969 * (-EINVAL) if *queue_id* is invalid
4970 * (-ENOTSUP) if the device does not support this function
4973 rte_eth_rx_queue_count(uint16_t port_id, uint16_t queue_id)
4975 struct rte_eth_dev *dev;
4977 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
4978 dev = &rte_eth_devices[port_id];
4979 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_queue_count, -ENOTSUP);
4980 if (queue_id >= dev->data->nb_rx_queues ||
4981 dev->data->rx_queues[queue_id] == NULL)
4984 return (int)(*dev->rx_queue_count)(dev, queue_id);
4988 * Check if the DD bit of the specific RX descriptor in the queue has been set
4991 * The port identifier of the Ethernet device.
4993 * The queue id on the specific port.
4995 * The offset of the descriptor ID from tail.
4997 * - (1) if the specific DD bit is set.
4998 * - (0) if the specific DD bit is not set.
4999 * - (-ENODEV) if *port_id* invalid.
5000 * - (-ENOTSUP) if the device does not support this function
5004 rte_eth_rx_descriptor_done(uint16_t port_id, uint16_t queue_id, uint16_t offset)
5006 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
5007 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
5008 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_descriptor_done, -ENOTSUP);
5009 return (*dev->rx_descriptor_done)(dev->data->rx_queues[queue_id], offset);
5012 #define RTE_ETH_RX_DESC_AVAIL 0 /**< Desc available for hw. */
5013 #define RTE_ETH_RX_DESC_DONE 1 /**< Desc done, filled by hw. */
5014 #define RTE_ETH_RX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
5017 * Check the status of a Rx descriptor in the queue
5019 * It should be called in a similar context than the Rx function:
5020 * - on a dataplane core
5021 * - not concurrently on the same queue
5023 * Since it's a dataplane function, no check is performed on port_id and
5024 * queue_id. The caller must therefore ensure that the port is enabled
5025 * and the queue is configured and running.
5027 * Note: accessing to a random descriptor in the ring may trigger cache
5028 * misses and have a performance impact.
5031 * A valid port identifier of the Ethernet device which.
5033 * A valid Rx queue identifier on this port.
5035 * The offset of the descriptor starting from tail (0 is the next
5036 * packet to be received by the driver).
5039 * - (RTE_ETH_RX_DESC_AVAIL): Descriptor is available for the hardware to
5041 * - (RTE_ETH_RX_DESC_DONE): Descriptor is done, it is filled by hw, but
5042 * not yet processed by the driver (i.e. in the receive queue).
5043 * - (RTE_ETH_RX_DESC_UNAVAIL): Descriptor is unavailable, either hold by
5044 * the driver and not yet returned to hw, or reserved by the hw.
5045 * - (-EINVAL) bad descriptor offset.
5046 * - (-ENOTSUP) if the device does not support this function.
5047 * - (-ENODEV) bad port or queue (only if compiled with debug).
5050 rte_eth_rx_descriptor_status(uint16_t port_id, uint16_t queue_id,
5053 struct rte_eth_dev *dev;
5056 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
5057 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
5059 dev = &rte_eth_devices[port_id];
5060 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
5061 if (queue_id >= dev->data->nb_rx_queues)
5064 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_descriptor_status, -ENOTSUP);
5065 rxq = dev->data->rx_queues[queue_id];
5067 return (*dev->rx_descriptor_status)(rxq, offset);
5070 #define RTE_ETH_TX_DESC_FULL 0 /**< Desc filled for hw, waiting xmit. */
5071 #define RTE_ETH_TX_DESC_DONE 1 /**< Desc done, packet is transmitted. */
5072 #define RTE_ETH_TX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
5075 * Check the status of a Tx descriptor in the queue.
5077 * It should be called in a similar context than the Tx function:
5078 * - on a dataplane core
5079 * - not concurrently on the same queue
5081 * Since it's a dataplane function, no check is performed on port_id and
5082 * queue_id. The caller must therefore ensure that the port is enabled
5083 * and the queue is configured and running.
5085 * Note: accessing to a random descriptor in the ring may trigger cache
5086 * misses and have a performance impact.
5089 * A valid port identifier of the Ethernet device which.
5091 * A valid Tx queue identifier on this port.
5093 * The offset of the descriptor starting from tail (0 is the place where
5094 * the next packet will be send).
5097 * - (RTE_ETH_TX_DESC_FULL) Descriptor is being processed by the hw, i.e.
5098 * in the transmit queue.
5099 * - (RTE_ETH_TX_DESC_DONE) Hardware is done with this descriptor, it can
5100 * be reused by the driver.
5101 * - (RTE_ETH_TX_DESC_UNAVAIL): Descriptor is unavailable, reserved by the
5102 * driver or the hardware.
5103 * - (-EINVAL) bad descriptor offset.
5104 * - (-ENOTSUP) if the device does not support this function.
5105 * - (-ENODEV) bad port or queue (only if compiled with debug).
5107 static inline int rte_eth_tx_descriptor_status(uint16_t port_id,
5108 uint16_t queue_id, uint16_t offset)
5110 struct rte_eth_dev *dev;
5113 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
5114 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
5116 dev = &rte_eth_devices[port_id];
5117 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
5118 if (queue_id >= dev->data->nb_tx_queues)
5121 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_descriptor_status, -ENOTSUP);
5122 txq = dev->data->tx_queues[queue_id];
5124 return (*dev->tx_descriptor_status)(txq, offset);
5128 * Send a burst of output packets on a transmit queue of an Ethernet device.
5130 * The rte_eth_tx_burst() function is invoked to transmit output packets
5131 * on the output queue *queue_id* of the Ethernet device designated by its
5133 * The *nb_pkts* parameter is the number of packets to send which are
5134 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
5135 * allocated from a pool created with rte_pktmbuf_pool_create().
5136 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets,
5137 * up to the number of transmit descriptors available in the TX ring of the
5139 * For each packet to send, the rte_eth_tx_burst() function performs
5140 * the following operations:
5142 * - Pick up the next available descriptor in the transmit ring.
5144 * - Free the network buffer previously sent with that descriptor, if any.
5146 * - Initialize the transmit descriptor with the information provided
5147 * in the *rte_mbuf data structure.
5149 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers,
5150 * the rte_eth_tx_burst() function uses several transmit descriptors
5153 * The rte_eth_tx_burst() function returns the number of packets it
5154 * actually sent. A return value equal to *nb_pkts* means that all packets
5155 * have been sent, and this is likely to signify that other output packets
5156 * could be immediately transmitted again. Applications that implement a
5157 * "send as many packets to transmit as possible" policy can check this
5158 * specific case and keep invoking the rte_eth_tx_burst() function until
5159 * a value less than *nb_pkts* is returned.
5161 * It is the responsibility of the rte_eth_tx_burst() function to
5162 * transparently free the memory buffers of packets previously sent.
5163 * This feature is driven by the *tx_free_thresh* value supplied to the
5164 * rte_eth_dev_configure() function at device configuration time.
5165 * When the number of free TX descriptors drops below this threshold, the
5166 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers
5167 * of those packets whose transmission was effectively completed.
5169 * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can
5170 * invoke this function concurrently on the same tx queue without SW lock.
5171 * @see rte_eth_dev_info_get, struct rte_eth_txconf::offloads
5173 * @see rte_eth_tx_prepare to perform some prior checks or adjustments
5177 * The port identifier of the Ethernet device.
5179 * The index of the transmit queue through which output packets must be
5181 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5182 * to rte_eth_dev_configure().
5184 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
5185 * which contain the output packets.
5187 * The maximum number of packets to transmit.
5189 * The number of output packets actually stored in transmit descriptors of
5190 * the transmit ring. The return value can be less than the value of the
5191 * *tx_pkts* parameter when the transmit ring is full or has been filled up.
5193 static inline uint16_t
5194 rte_eth_tx_burst(uint16_t port_id, uint16_t queue_id,
5195 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
5197 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
5199 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
5200 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
5201 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_pkt_burst, 0);
5203 if (queue_id >= dev->data->nb_tx_queues) {
5204 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
5209 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
5210 struct rte_eth_rxtx_callback *cb;
5212 /* __ATOMIC_RELEASE memory order was used when the
5213 * call back was inserted into the list.
5214 * Since there is a clear dependency between loading
5215 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is
5218 cb = __atomic_load_n(&dev->pre_tx_burst_cbs[queue_id],
5221 if (unlikely(cb != NULL)) {
5223 nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
5226 } while (cb != NULL);
5230 rte_ethdev_trace_tx_burst(port_id, queue_id, (void **)tx_pkts,
5232 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts);
5236 * Process a burst of output packets on a transmit queue of an Ethernet device.
5238 * The rte_eth_tx_prepare() function is invoked to prepare output packets to be
5239 * transmitted on the output queue *queue_id* of the Ethernet device designated
5241 * The *nb_pkts* parameter is the number of packets to be prepared which are
5242 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
5243 * allocated from a pool created with rte_pktmbuf_pool_create().
5244 * For each packet to send, the rte_eth_tx_prepare() function performs
5245 * the following operations:
5247 * - Check if packet meets devices requirements for tx offloads.
5249 * - Check limitations about number of segments.
5251 * - Check additional requirements when debug is enabled.
5253 * - Update and/or reset required checksums when tx offload is set for packet.
5255 * Since this function can modify packet data, provided mbufs must be safely
5256 * writable (e.g. modified data cannot be in shared segment).
5258 * The rte_eth_tx_prepare() function returns the number of packets ready to be
5259 * sent. A return value equal to *nb_pkts* means that all packets are valid and
5260 * ready to be sent, otherwise stops processing on the first invalid packet and
5261 * leaves the rest packets untouched.
5263 * When this functionality is not implemented in the driver, all packets are
5264 * are returned untouched.
5267 * The port identifier of the Ethernet device.
5268 * The value must be a valid port id.
5270 * The index of the transmit queue through which output packets must be
5272 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5273 * to rte_eth_dev_configure().
5275 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
5276 * which contain the output packets.
5278 * The maximum number of packets to process.
5280 * The number of packets correct and ready to be sent. The return value can be
5281 * less than the value of the *tx_pkts* parameter when some packet doesn't
5282 * meet devices requirements with rte_errno set appropriately:
5283 * - EINVAL: offload flags are not correctly set
5284 * - ENOTSUP: the offload feature is not supported by the hardware
5285 * - ENODEV: if *port_id* is invalid (with debug enabled only)
5289 #ifndef RTE_ETHDEV_TX_PREPARE_NOOP
5291 static inline uint16_t
5292 rte_eth_tx_prepare(uint16_t port_id, uint16_t queue_id,
5293 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
5295 struct rte_eth_dev *dev;
5297 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
5298 if (!rte_eth_dev_is_valid_port(port_id)) {
5299 RTE_ETHDEV_LOG(ERR, "Invalid TX port_id=%u\n", port_id);
5305 dev = &rte_eth_devices[port_id];
5307 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
5308 if (queue_id >= dev->data->nb_tx_queues) {
5309 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
5315 if (!dev->tx_pkt_prepare)
5318 return (*dev->tx_pkt_prepare)(dev->data->tx_queues[queue_id],
5325 * Native NOOP operation for compilation targets which doesn't require any
5326 * preparations steps, and functional NOOP may introduce unnecessary performance
5329 * Generally this is not a good idea to turn it on globally and didn't should
5330 * be used if behavior of tx_preparation can change.
5333 static inline uint16_t
5334 rte_eth_tx_prepare(__rte_unused uint16_t port_id,
5335 __rte_unused uint16_t queue_id,
5336 __rte_unused struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
5344 * Send any packets queued up for transmission on a port and HW queue
5346 * This causes an explicit flush of packets previously buffered via the
5347 * rte_eth_tx_buffer() function. It returns the number of packets successfully
5348 * sent to the NIC, and calls the error callback for any unsent packets. Unless
5349 * explicitly set up otherwise, the default callback simply frees the unsent
5350 * packets back to the owning mempool.
5353 * The port identifier of the Ethernet device.
5355 * The index of the transmit queue through which output packets must be
5357 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5358 * to rte_eth_dev_configure().
5360 * Buffer of packets to be transmit.
5362 * The number of packets successfully sent to the Ethernet device. The error
5363 * callback is called for any packets which could not be sent.
5365 static inline uint16_t
5366 rte_eth_tx_buffer_flush(uint16_t port_id, uint16_t queue_id,
5367 struct rte_eth_dev_tx_buffer *buffer)
5370 uint16_t to_send = buffer->length;
5375 sent = rte_eth_tx_burst(port_id, queue_id, buffer->pkts, to_send);
5379 /* All packets sent, or to be dealt with by callback below */
5380 if (unlikely(sent != to_send))
5381 buffer->error_callback(&buffer->pkts[sent],
5382 (uint16_t)(to_send - sent),
5383 buffer->error_userdata);
5389 * Buffer a single packet for future transmission on a port and queue
5391 * This function takes a single mbuf/packet and buffers it for later
5392 * transmission on the particular port and queue specified. Once the buffer is
5393 * full of packets, an attempt will be made to transmit all the buffered
5394 * packets. In case of error, where not all packets can be transmitted, a
5395 * callback is called with the unsent packets as a parameter. If no callback
5396 * is explicitly set up, the unsent packets are just freed back to the owning
5397 * mempool. The function returns the number of packets actually sent i.e.
5398 * 0 if no buffer flush occurred, otherwise the number of packets successfully
5402 * The port identifier of the Ethernet device.
5404 * The index of the transmit queue through which output packets must be
5406 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
5407 * to rte_eth_dev_configure().
5409 * Buffer used to collect packets to be sent.
5411 * Pointer to the packet mbuf to be sent.
5413 * 0 = packet has been buffered for later transmission
5414 * N > 0 = packet has been buffered, and the buffer was subsequently flushed,
5415 * causing N packets to be sent, and the error callback to be called for
5418 static __rte_always_inline uint16_t
5419 rte_eth_tx_buffer(uint16_t port_id, uint16_t queue_id,
5420 struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt)
5422 buffer->pkts[buffer->length++] = tx_pkt;
5423 if (buffer->length < buffer->size)
5426 return rte_eth_tx_buffer_flush(port_id, queue_id, buffer);
5433 #endif /* _RTE_ETHDEV_H_ */