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 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
257 /**< Total number of queue RX packets. */
258 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
259 /**< Total number of queue TX packets. */
260 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
261 /**< Total number of successfully received queue bytes. */
262 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
263 /**< Total number of successfully transmitted queue bytes. */
264 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS];
265 /**< Total number of queue packets received that are dropped. */
269 * Device supported speeds bitmap flags
271 #define ETH_LINK_SPEED_AUTONEG (0 << 0) /**< Autonegotiate (all speeds) */
272 #define ETH_LINK_SPEED_FIXED (1 << 0) /**< Disable autoneg (fixed speed) */
273 #define ETH_LINK_SPEED_10M_HD (1 << 1) /**< 10 Mbps half-duplex */
274 #define ETH_LINK_SPEED_10M (1 << 2) /**< 10 Mbps full-duplex */
275 #define ETH_LINK_SPEED_100M_HD (1 << 3) /**< 100 Mbps half-duplex */
276 #define ETH_LINK_SPEED_100M (1 << 4) /**< 100 Mbps full-duplex */
277 #define ETH_LINK_SPEED_1G (1 << 5) /**< 1 Gbps */
278 #define ETH_LINK_SPEED_2_5G (1 << 6) /**< 2.5 Gbps */
279 #define ETH_LINK_SPEED_5G (1 << 7) /**< 5 Gbps */
280 #define ETH_LINK_SPEED_10G (1 << 8) /**< 10 Gbps */
281 #define ETH_LINK_SPEED_20G (1 << 9) /**< 20 Gbps */
282 #define ETH_LINK_SPEED_25G (1 << 10) /**< 25 Gbps */
283 #define ETH_LINK_SPEED_40G (1 << 11) /**< 40 Gbps */
284 #define ETH_LINK_SPEED_50G (1 << 12) /**< 50 Gbps */
285 #define ETH_LINK_SPEED_56G (1 << 13) /**< 56 Gbps */
286 #define ETH_LINK_SPEED_100G (1 << 14) /**< 100 Gbps */
289 * Ethernet numeric link speeds in Mbps
291 #define ETH_SPEED_NUM_NONE 0 /**< Not defined */
292 #define ETH_SPEED_NUM_10M 10 /**< 10 Mbps */
293 #define ETH_SPEED_NUM_100M 100 /**< 100 Mbps */
294 #define ETH_SPEED_NUM_1G 1000 /**< 1 Gbps */
295 #define ETH_SPEED_NUM_2_5G 2500 /**< 2.5 Gbps */
296 #define ETH_SPEED_NUM_5G 5000 /**< 5 Gbps */
297 #define ETH_SPEED_NUM_10G 10000 /**< 10 Gbps */
298 #define ETH_SPEED_NUM_20G 20000 /**< 20 Gbps */
299 #define ETH_SPEED_NUM_25G 25000 /**< 25 Gbps */
300 #define ETH_SPEED_NUM_40G 40000 /**< 40 Gbps */
301 #define ETH_SPEED_NUM_50G 50000 /**< 50 Gbps */
302 #define ETH_SPEED_NUM_56G 56000 /**< 56 Gbps */
303 #define ETH_SPEED_NUM_100G 100000 /**< 100 Gbps */
306 * A structure used to retrieve link-level information of an Ethernet port.
309 struct rte_eth_link {
310 uint32_t link_speed; /**< ETH_SPEED_NUM_ */
311 uint16_t link_duplex : 1; /**< ETH_LINK_[HALF/FULL]_DUPLEX */
312 uint16_t link_autoneg : 1; /**< ETH_LINK_[AUTONEG/FIXED] */
313 uint16_t link_status : 1; /**< ETH_LINK_[DOWN/UP] */
314 } __rte_aligned(8); /**< aligned for atomic64 read/write */
316 /* Utility constants */
317 #define ETH_LINK_HALF_DUPLEX 0 /**< Half-duplex connection (see link_duplex). */
318 #define ETH_LINK_FULL_DUPLEX 1 /**< Full-duplex connection (see link_duplex). */
319 #define ETH_LINK_DOWN 0 /**< Link is down (see link_status). */
320 #define ETH_LINK_UP 1 /**< Link is up (see link_status). */
321 #define ETH_LINK_FIXED 0 /**< No autonegotiation (see link_autoneg). */
322 #define ETH_LINK_AUTONEG 1 /**< Autonegotiated (see link_autoneg). */
325 * A structure used to configure the ring threshold registers of an RX/TX
326 * queue for an Ethernet port.
328 struct rte_eth_thresh {
329 uint8_t pthresh; /**< Ring prefetch threshold. */
330 uint8_t hthresh; /**< Ring host threshold. */
331 uint8_t wthresh; /**< Ring writeback threshold. */
335 * Simple flags are used for rte_eth_conf.rxmode.mq_mode.
337 #define ETH_MQ_RX_RSS_FLAG 0x1
338 #define ETH_MQ_RX_DCB_FLAG 0x2
339 #define ETH_MQ_RX_VMDQ_FLAG 0x4
342 * A set of values to identify what method is to be used to route
343 * packets to multiple queues.
345 enum rte_eth_rx_mq_mode {
346 /** None of DCB,RSS or VMDQ mode */
349 /** For RX side, only RSS is on */
350 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG,
351 /** For RX side,only DCB is on. */
352 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG,
353 /** Both DCB and RSS enable */
354 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG,
356 /** Only VMDQ, no RSS nor DCB */
357 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG,
358 /** RSS mode with VMDQ */
359 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG,
360 /** Use VMDQ+DCB to route traffic to queues */
361 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG,
362 /** Enable both VMDQ and DCB in VMDq */
363 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG |
368 * for rx mq mode backward compatible
370 #define ETH_RSS ETH_MQ_RX_RSS
371 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB
372 #define ETH_DCB_RX ETH_MQ_RX_DCB
375 * A set of values to identify what method is to be used to transmit
376 * packets using multi-TCs.
378 enum rte_eth_tx_mq_mode {
379 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */
380 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */
381 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */
382 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */
386 * for tx mq mode backward compatible
388 #define ETH_DCB_NONE ETH_MQ_TX_NONE
389 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB
390 #define ETH_DCB_TX ETH_MQ_TX_DCB
393 * A structure used to configure the RX features of an Ethernet port.
395 struct rte_eth_rxmode {
396 /** The multi-queue packet distribution mode to be used, e.g. RSS. */
397 enum rte_eth_rx_mq_mode mq_mode;
398 uint32_t max_rx_pkt_len; /**< Only used if JUMBO_FRAME enabled. */
399 /** Maximum allowed size of LRO aggregated packet. */
400 uint32_t max_lro_pkt_size;
401 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/
403 * Per-port Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
404 * Only offloads set on rx_offload_capa field on rte_eth_dev_info
405 * structure are allowed to be set.
409 uint64_t reserved_64s[2]; /**< Reserved for future fields */
410 void *reserved_ptrs[2]; /**< Reserved for future fields */
414 * VLAN types to indicate if it is for single VLAN, inner VLAN or outer VLAN.
415 * Note that single VLAN is treated the same as inner VLAN.
418 ETH_VLAN_TYPE_UNKNOWN = 0,
419 ETH_VLAN_TYPE_INNER, /**< Inner VLAN. */
420 ETH_VLAN_TYPE_OUTER, /**< Single VLAN, or outer VLAN. */
425 * A structure used to describe a vlan filter.
426 * If the bit corresponding to a VID is set, such VID is on.
428 struct rte_vlan_filter_conf {
433 * A structure used to configure the Receive Side Scaling (RSS) feature
434 * of an Ethernet port.
435 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points
436 * to an array holding the RSS key to use for hashing specific header
437 * fields of received packets. The length of this array should be indicated
438 * by *rss_key_len* below. Otherwise, a default random hash key is used by
441 * The *rss_key_len* field of the *rss_conf* structure indicates the length
442 * in bytes of the array pointed by *rss_key*. To be compatible, this length
443 * will be checked in i40e only. Others assume 40 bytes to be used as before.
445 * The *rss_hf* field of the *rss_conf* structure indicates the different
446 * types of IPv4/IPv6 packets to which the RSS hashing must be applied.
447 * Supplying an *rss_hf* equal to zero disables the RSS feature.
449 struct rte_eth_rss_conf {
450 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */
451 uint8_t rss_key_len; /**< hash key length in bytes. */
452 uint64_t rss_hf; /**< Hash functions to apply - see below. */
456 * A packet can be identified by hardware as different flow types. Different
457 * NIC hardware may support different flow types.
458 * Basically, the NIC hardware identifies the flow type as deep protocol as
459 * possible, and exclusively. For example, if a packet is identified as
460 * 'RTE_ETH_FLOW_NONFRAG_IPV4_TCP', it will not be any of other flow types,
461 * though it is an actual IPV4 packet.
463 #define RTE_ETH_FLOW_UNKNOWN 0
464 #define RTE_ETH_FLOW_RAW 1
465 #define RTE_ETH_FLOW_IPV4 2
466 #define RTE_ETH_FLOW_FRAG_IPV4 3
467 #define RTE_ETH_FLOW_NONFRAG_IPV4_TCP 4
468 #define RTE_ETH_FLOW_NONFRAG_IPV4_UDP 5
469 #define RTE_ETH_FLOW_NONFRAG_IPV4_SCTP 6
470 #define RTE_ETH_FLOW_NONFRAG_IPV4_OTHER 7
471 #define RTE_ETH_FLOW_IPV6 8
472 #define RTE_ETH_FLOW_FRAG_IPV6 9
473 #define RTE_ETH_FLOW_NONFRAG_IPV6_TCP 10
474 #define RTE_ETH_FLOW_NONFRAG_IPV6_UDP 11
475 #define RTE_ETH_FLOW_NONFRAG_IPV6_SCTP 12
476 #define RTE_ETH_FLOW_NONFRAG_IPV6_OTHER 13
477 #define RTE_ETH_FLOW_L2_PAYLOAD 14
478 #define RTE_ETH_FLOW_IPV6_EX 15
479 #define RTE_ETH_FLOW_IPV6_TCP_EX 16
480 #define RTE_ETH_FLOW_IPV6_UDP_EX 17
481 #define RTE_ETH_FLOW_PORT 18
482 /**< Consider device port number as a flow differentiator */
483 #define RTE_ETH_FLOW_VXLAN 19 /**< VXLAN protocol based flow */
484 #define RTE_ETH_FLOW_GENEVE 20 /**< GENEVE protocol based flow */
485 #define RTE_ETH_FLOW_NVGRE 21 /**< NVGRE protocol based flow */
486 #define RTE_ETH_FLOW_VXLAN_GPE 22 /**< VXLAN-GPE protocol based flow */
487 #define RTE_ETH_FLOW_GTPU 23 /**< GTPU protocol based flow */
488 #define RTE_ETH_FLOW_MAX 24
491 * Below macros are defined for RSS offload types, they can be used to
492 * fill rte_eth_rss_conf.rss_hf or rte_flow_action_rss.types.
494 #define ETH_RSS_IPV4 (1ULL << 2)
495 #define ETH_RSS_FRAG_IPV4 (1ULL << 3)
496 #define ETH_RSS_NONFRAG_IPV4_TCP (1ULL << 4)
497 #define ETH_RSS_NONFRAG_IPV4_UDP (1ULL << 5)
498 #define ETH_RSS_NONFRAG_IPV4_SCTP (1ULL << 6)
499 #define ETH_RSS_NONFRAG_IPV4_OTHER (1ULL << 7)
500 #define ETH_RSS_IPV6 (1ULL << 8)
501 #define ETH_RSS_FRAG_IPV6 (1ULL << 9)
502 #define ETH_RSS_NONFRAG_IPV6_TCP (1ULL << 10)
503 #define ETH_RSS_NONFRAG_IPV6_UDP (1ULL << 11)
504 #define ETH_RSS_NONFRAG_IPV6_SCTP (1ULL << 12)
505 #define ETH_RSS_NONFRAG_IPV6_OTHER (1ULL << 13)
506 #define ETH_RSS_L2_PAYLOAD (1ULL << 14)
507 #define ETH_RSS_IPV6_EX (1ULL << 15)
508 #define ETH_RSS_IPV6_TCP_EX (1ULL << 16)
509 #define ETH_RSS_IPV6_UDP_EX (1ULL << 17)
510 #define ETH_RSS_PORT (1ULL << 18)
511 #define ETH_RSS_VXLAN (1ULL << 19)
512 #define ETH_RSS_GENEVE (1ULL << 20)
513 #define ETH_RSS_NVGRE (1ULL << 21)
514 #define ETH_RSS_GTPU (1ULL << 23)
515 #define ETH_RSS_ETH (1ULL << 24)
516 #define ETH_RSS_S_VLAN (1ULL << 25)
517 #define ETH_RSS_C_VLAN (1ULL << 26)
518 #define ETH_RSS_ESP (1ULL << 27)
519 #define ETH_RSS_AH (1ULL << 28)
520 #define ETH_RSS_L2TPV3 (1ULL << 29)
521 #define ETH_RSS_PFCP (1ULL << 30)
525 * We use the following macros to combine with above ETH_RSS_* for
526 * more specific input set selection. These bits are defined starting
527 * from the high end of the 64 bits.
528 * Note: If we use above ETH_RSS_* without SRC/DST_ONLY, it represents
529 * both SRC and DST are taken into account. If SRC_ONLY and DST_ONLY of
530 * the same level are used simultaneously, it is the same case as none of
533 #define ETH_RSS_L3_SRC_ONLY (1ULL << 63)
534 #define ETH_RSS_L3_DST_ONLY (1ULL << 62)
535 #define ETH_RSS_L4_SRC_ONLY (1ULL << 61)
536 #define ETH_RSS_L4_DST_ONLY (1ULL << 60)
537 #define ETH_RSS_L2_SRC_ONLY (1ULL << 59)
538 #define ETH_RSS_L2_DST_ONLY (1ULL << 58)
541 * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
542 * the same level are used simultaneously, it is the same case as
543 * none of them are added.
546 * RSS types with SRC/DST_ONLY.
550 static inline uint64_t
551 rte_eth_rss_hf_refine(uint64_t rss_hf)
553 if ((rss_hf & ETH_RSS_L3_SRC_ONLY) && (rss_hf & ETH_RSS_L3_DST_ONLY))
554 rss_hf &= ~(ETH_RSS_L3_SRC_ONLY | ETH_RSS_L3_DST_ONLY);
556 if ((rss_hf & ETH_RSS_L4_SRC_ONLY) && (rss_hf & ETH_RSS_L4_DST_ONLY))
557 rss_hf &= ~(ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY);
562 #define ETH_RSS_IP ( \
564 ETH_RSS_FRAG_IPV4 | \
565 ETH_RSS_NONFRAG_IPV4_OTHER | \
567 ETH_RSS_FRAG_IPV6 | \
568 ETH_RSS_NONFRAG_IPV6_OTHER | \
571 #define ETH_RSS_UDP ( \
572 ETH_RSS_NONFRAG_IPV4_UDP | \
573 ETH_RSS_NONFRAG_IPV6_UDP | \
576 #define ETH_RSS_TCP ( \
577 ETH_RSS_NONFRAG_IPV4_TCP | \
578 ETH_RSS_NONFRAG_IPV6_TCP | \
581 #define ETH_RSS_SCTP ( \
582 ETH_RSS_NONFRAG_IPV4_SCTP | \
583 ETH_RSS_NONFRAG_IPV6_SCTP)
585 #define ETH_RSS_TUNNEL ( \
590 #define ETH_RSS_VLAN ( \
594 /**< Mask of valid RSS hash protocols */
595 #define ETH_RSS_PROTO_MASK ( \
597 ETH_RSS_FRAG_IPV4 | \
598 ETH_RSS_NONFRAG_IPV4_TCP | \
599 ETH_RSS_NONFRAG_IPV4_UDP | \
600 ETH_RSS_NONFRAG_IPV4_SCTP | \
601 ETH_RSS_NONFRAG_IPV4_OTHER | \
603 ETH_RSS_FRAG_IPV6 | \
604 ETH_RSS_NONFRAG_IPV6_TCP | \
605 ETH_RSS_NONFRAG_IPV6_UDP | \
606 ETH_RSS_NONFRAG_IPV6_SCTP | \
607 ETH_RSS_NONFRAG_IPV6_OTHER | \
608 ETH_RSS_L2_PAYLOAD | \
610 ETH_RSS_IPV6_TCP_EX | \
611 ETH_RSS_IPV6_UDP_EX | \
618 * Definitions used for redirection table entry size.
619 * Some RSS RETA sizes may not be supported by some drivers, check the
620 * documentation or the description of relevant functions for more details.
622 #define ETH_RSS_RETA_SIZE_64 64
623 #define ETH_RSS_RETA_SIZE_128 128
624 #define ETH_RSS_RETA_SIZE_256 256
625 #define ETH_RSS_RETA_SIZE_512 512
626 #define RTE_RETA_GROUP_SIZE 64
628 /* Definitions used for VMDQ and DCB functionality */
629 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */
630 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */
631 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */
632 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */
634 /* DCB capability defines */
635 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */
636 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */
638 /* Definitions used for VLAN Offload functionality */
639 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */
640 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */
641 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */
642 #define ETH_QINQ_STRIP_OFFLOAD 0x0008 /**< QINQ Strip On/Off */
644 /* Definitions used for mask VLAN setting */
645 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */
646 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/
647 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/
648 #define ETH_QINQ_STRIP_MASK 0x0008 /**< QINQ Strip setting mask */
649 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/
651 /* Definitions used for receive MAC address */
652 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */
654 /* Definitions used for unicast hash */
655 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */
657 /* Definitions used for VMDQ pool rx mode setting */
658 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */
659 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */
660 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */
661 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */
662 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */
664 /** Maximum nb. of vlan per mirror rule */
665 #define ETH_MIRROR_MAX_VLANS 64
667 #define ETH_MIRROR_VIRTUAL_POOL_UP 0x01 /**< Virtual Pool uplink Mirroring. */
668 #define ETH_MIRROR_UPLINK_PORT 0x02 /**< Uplink Port Mirroring. */
669 #define ETH_MIRROR_DOWNLINK_PORT 0x04 /**< Downlink Port Mirroring. */
670 #define ETH_MIRROR_VLAN 0x08 /**< VLAN Mirroring. */
671 #define ETH_MIRROR_VIRTUAL_POOL_DOWN 0x10 /**< Virtual Pool downlink Mirroring. */
674 * A structure used to configure VLAN traffic mirror of an Ethernet port.
676 struct rte_eth_vlan_mirror {
677 uint64_t vlan_mask; /**< mask for valid VLAN ID. */
678 /** VLAN ID list for vlan mirroring. */
679 uint16_t vlan_id[ETH_MIRROR_MAX_VLANS];
683 * A structure used to configure traffic mirror of an Ethernet port.
685 struct rte_eth_mirror_conf {
686 uint8_t rule_type; /**< Mirroring rule type */
687 uint8_t dst_pool; /**< Destination pool for this mirror rule. */
688 uint64_t pool_mask; /**< Bitmap of pool for pool mirroring */
689 /** VLAN ID setting for VLAN mirroring. */
690 struct rte_eth_vlan_mirror vlan;
694 * A structure used to configure 64 entries of Redirection Table of the
695 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure
696 * more than 64 entries supported by hardware, an array of this structure
699 struct rte_eth_rss_reta_entry64 {
701 /**< Mask bits indicate which entries need to be updated/queried. */
702 uint16_t reta[RTE_RETA_GROUP_SIZE];
703 /**< Group of 64 redirection table entries. */
707 * This enum indicates the possible number of traffic classes
708 * in DCB configurations
710 enum rte_eth_nb_tcs {
711 ETH_4_TCS = 4, /**< 4 TCs with DCB. */
712 ETH_8_TCS = 8 /**< 8 TCs with DCB. */
716 * This enum indicates the possible number of queue pools
717 * in VMDQ configurations.
719 enum rte_eth_nb_pools {
720 ETH_8_POOLS = 8, /**< 8 VMDq pools. */
721 ETH_16_POOLS = 16, /**< 16 VMDq pools. */
722 ETH_32_POOLS = 32, /**< 32 VMDq pools. */
723 ETH_64_POOLS = 64 /**< 64 VMDq pools. */
726 /* This structure may be extended in future. */
727 struct rte_eth_dcb_rx_conf {
728 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */
729 /** Traffic class each UP mapped to. */
730 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
733 struct rte_eth_vmdq_dcb_tx_conf {
734 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */
735 /** Traffic class each UP mapped to. */
736 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
739 struct rte_eth_dcb_tx_conf {
740 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */
741 /** Traffic class each UP mapped to. */
742 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
745 struct rte_eth_vmdq_tx_conf {
746 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */
750 * A structure used to configure the VMDQ+DCB feature
751 * of an Ethernet port.
753 * Using this feature, packets are routed to a pool of queues, based
754 * on the vlan id in the vlan tag, and then to a specific queue within
755 * that pool, using the user priority vlan tag field.
757 * A default pool may be used, if desired, to route all traffic which
758 * does not match the vlan filter rules.
760 struct rte_eth_vmdq_dcb_conf {
761 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */
762 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
763 uint8_t default_pool; /**< The default pool, if applicable */
764 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
766 uint16_t vlan_id; /**< The vlan id of the received frame */
767 uint64_t pools; /**< Bitmask of pools for packet rx */
768 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
769 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
770 /**< Selects a queue in a pool */
774 * A structure used to configure the VMDQ feature of an Ethernet port when
775 * not combined with the DCB feature.
777 * Using this feature, packets are routed to a pool of queues. By default,
778 * the pool selection is based on the MAC address, the vlan id in the
779 * vlan tag as specified in the pool_map array.
780 * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool
781 * selection using only the MAC address. MAC address to pool mapping is done
782 * using the rte_eth_dev_mac_addr_add function, with the pool parameter
783 * corresponding to the pool id.
785 * Queue selection within the selected pool will be done using RSS when
786 * it is enabled or revert to the first queue of the pool if not.
788 * A default pool may be used, if desired, to route all traffic which
789 * does not match the vlan filter rules or any pool MAC address.
791 struct rte_eth_vmdq_rx_conf {
792 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */
793 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
794 uint8_t default_pool; /**< The default pool, if applicable */
795 uint8_t enable_loop_back; /**< Enable VT loop back */
796 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
797 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */
799 uint16_t vlan_id; /**< The vlan id of the received frame */
800 uint64_t pools; /**< Bitmask of pools for packet rx */
801 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
805 * A structure used to configure the TX features of an Ethernet port.
807 struct rte_eth_txmode {
808 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */
810 * Per-port Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
811 * Only offloads set on tx_offload_capa field on rte_eth_dev_info
812 * structure are allowed to be set.
816 /* For i40e specifically */
819 uint8_t hw_vlan_reject_tagged : 1,
820 /**< If set, reject sending out tagged pkts */
821 hw_vlan_reject_untagged : 1,
822 /**< If set, reject sending out untagged pkts */
823 hw_vlan_insert_pvid : 1;
824 /**< If set, enable port based VLAN insertion */
826 uint64_t reserved_64s[2]; /**< Reserved for future fields */
827 void *reserved_ptrs[2]; /**< Reserved for future fields */
831 * A structure used to configure an RX ring of an Ethernet port.
833 struct rte_eth_rxconf {
834 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */
835 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
836 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
837 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
839 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
840 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
841 * fields on rte_eth_dev_info structure are allowed to be set.
845 uint64_t reserved_64s[2]; /**< Reserved for future fields */
846 void *reserved_ptrs[2]; /**< Reserved for future fields */
850 * A structure used to configure a TX ring of an Ethernet port.
852 struct rte_eth_txconf {
853 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */
854 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */
855 uint16_t tx_free_thresh; /**< Start freeing TX buffers if there are
856 less free descriptors than this value. */
858 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
860 * Per-queue Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
861 * Only offloads set on tx_queue_offload_capa or tx_offload_capa
862 * fields on rte_eth_dev_info structure are allowed to be set.
866 uint64_t reserved_64s[2]; /**< Reserved for future fields */
867 void *reserved_ptrs[2]; /**< Reserved for future fields */
872 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
874 * A structure used to return the hairpin capabilities that are supported.
876 struct rte_eth_hairpin_cap {
877 /** The max number of hairpin queues (different bindings). */
878 uint16_t max_nb_queues;
879 /** Max number of Rx queues to be connected to one Tx queue. */
880 uint16_t max_rx_2_tx;
881 /** Max number of Tx queues to be connected to one Rx queue. */
882 uint16_t max_tx_2_rx;
883 uint16_t max_nb_desc; /**< The max num of descriptors. */
886 #define RTE_ETH_MAX_HAIRPIN_PEERS 32
890 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
892 * A structure used to hold hairpin peer data.
894 struct rte_eth_hairpin_peer {
895 uint16_t port; /**< Peer port. */
896 uint16_t queue; /**< Peer queue. */
901 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
903 * A structure used to configure hairpin binding.
905 struct rte_eth_hairpin_conf {
906 uint16_t peer_count; /**< The number of peers. */
907 struct rte_eth_hairpin_peer peers[RTE_ETH_MAX_HAIRPIN_PEERS];
911 * A structure contains information about HW descriptor ring limitations.
913 struct rte_eth_desc_lim {
914 uint16_t nb_max; /**< Max allowed number of descriptors. */
915 uint16_t nb_min; /**< Min allowed number of descriptors. */
916 uint16_t nb_align; /**< Number of descriptors should be aligned to. */
919 * Max allowed number of segments per whole packet.
921 * - For TSO packet this is the total number of data descriptors allowed
924 * @see nb_mtu_seg_max
929 * Max number of segments per one MTU.
931 * - For non-TSO packet, this is the maximum allowed number of segments
932 * in a single transmit packet.
934 * - For TSO packet each segment within the TSO may span up to this
939 uint16_t nb_mtu_seg_max;
943 * This enum indicates the flow control mode
945 enum rte_eth_fc_mode {
946 RTE_FC_NONE = 0, /**< Disable flow control. */
947 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */
948 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */
949 RTE_FC_FULL /**< Enable flow control on both side. */
953 * A structure used to configure Ethernet flow control parameter.
954 * These parameters will be configured into the register of the NIC.
955 * Please refer to the corresponding data sheet for proper value.
957 struct rte_eth_fc_conf {
958 uint32_t high_water; /**< High threshold value to trigger XOFF */
959 uint32_t low_water; /**< Low threshold value to trigger XON */
960 uint16_t pause_time; /**< Pause quota in the Pause frame */
961 uint16_t send_xon; /**< Is XON frame need be sent */
962 enum rte_eth_fc_mode mode; /**< Link flow control mode */
963 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */
964 uint8_t autoneg; /**< Use Pause autoneg */
968 * A structure used to configure Ethernet priority flow control parameter.
969 * These parameters will be configured into the register of the NIC.
970 * Please refer to the corresponding data sheet for proper value.
972 struct rte_eth_pfc_conf {
973 struct rte_eth_fc_conf fc; /**< General flow control parameter. */
974 uint8_t priority; /**< VLAN User Priority. */
980 enum rte_eth_tunnel_type {
981 RTE_TUNNEL_TYPE_NONE = 0,
982 RTE_TUNNEL_TYPE_VXLAN,
983 RTE_TUNNEL_TYPE_GENEVE,
984 RTE_TUNNEL_TYPE_TEREDO,
985 RTE_TUNNEL_TYPE_NVGRE,
986 RTE_TUNNEL_TYPE_IP_IN_GRE,
987 RTE_L2_TUNNEL_TYPE_E_TAG,
988 RTE_TUNNEL_TYPE_VXLAN_GPE,
992 /* Deprecated API file for rte_eth_dev_filter_* functions */
993 #include "rte_eth_ctrl.h"
996 * Memory space that can be configured to store Flow Director filters
997 * in the board memory.
999 enum rte_fdir_pballoc_type {
1000 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */
1001 RTE_FDIR_PBALLOC_128K, /**< 128k. */
1002 RTE_FDIR_PBALLOC_256K, /**< 256k. */
1006 * Select report mode of FDIR hash information in RX descriptors.
1008 enum rte_fdir_status_mode {
1009 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */
1010 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */
1011 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */
1015 * A structure used to configure the Flow Director (FDIR) feature
1016 * of an Ethernet port.
1018 * If mode is RTE_FDIR_MODE_NONE, the pballoc value is ignored.
1020 struct rte_fdir_conf {
1021 enum rte_fdir_mode mode; /**< Flow Director mode. */
1022 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */
1023 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */
1024 /** RX queue of packets matching a "drop" filter in perfect mode. */
1026 struct rte_eth_fdir_masks mask;
1027 struct rte_eth_fdir_flex_conf flex_conf;
1028 /**< Flex payload configuration. */
1032 * UDP tunneling configuration.
1033 * Used to config the UDP port for a type of tunnel.
1034 * NICs need the UDP port to identify the tunnel type.
1035 * Normally a type of tunnel has a default UDP port, this structure can be used
1036 * in case if the users want to change or support more UDP port.
1038 struct rte_eth_udp_tunnel {
1039 uint16_t udp_port; /**< UDP port used for the tunnel. */
1040 uint8_t prot_type; /**< Tunnel type. Defined in rte_eth_tunnel_type. */
1044 * A structure used to enable/disable specific device interrupts.
1046 struct rte_intr_conf {
1047 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */
1049 /** enable/disable rxq interrupt. 0 (default) - disable, 1 enable */
1051 /** enable/disable rmv interrupt. 0 (default) - disable, 1 enable */
1056 * A structure used to configure an Ethernet port.
1057 * Depending upon the RX multi-queue mode, extra advanced
1058 * configuration settings may be needed.
1060 struct rte_eth_conf {
1061 uint32_t link_speeds; /**< bitmap of ETH_LINK_SPEED_XXX of speeds to be
1062 used. ETH_LINK_SPEED_FIXED disables link
1063 autonegotiation, and a unique speed shall be
1064 set. Otherwise, the bitmap defines the set of
1065 speeds to be advertised. If the special value
1066 ETH_LINK_SPEED_AUTONEG (0) is used, all speeds
1067 supported are advertised. */
1068 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */
1069 struct rte_eth_txmode txmode; /**< Port TX configuration. */
1070 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value
1071 is 0, meaning the loopback mode is disabled.
1072 Read the datasheet of given ethernet controller
1073 for details. The possible values of this field
1074 are defined in implementation of each driver. */
1076 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */
1077 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf;
1078 /**< Port vmdq+dcb configuration. */
1079 struct rte_eth_dcb_rx_conf dcb_rx_conf;
1080 /**< Port dcb RX configuration. */
1081 struct rte_eth_vmdq_rx_conf vmdq_rx_conf;
1082 /**< Port vmdq RX configuration. */
1083 } rx_adv_conf; /**< Port RX filtering configuration. */
1085 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf;
1086 /**< Port vmdq+dcb TX configuration. */
1087 struct rte_eth_dcb_tx_conf dcb_tx_conf;
1088 /**< Port dcb TX configuration. */
1089 struct rte_eth_vmdq_tx_conf vmdq_tx_conf;
1090 /**< Port vmdq TX configuration. */
1091 } tx_adv_conf; /**< Port TX DCB configuration (union). */
1092 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC
1093 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */
1094 uint32_t dcb_capability_en;
1095 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. DEPRECATED */
1096 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */
1100 * RX offload capabilities of a device.
1102 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001
1103 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002
1104 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004
1105 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008
1106 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010
1107 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020
1108 #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040
1109 #define DEV_RX_OFFLOAD_MACSEC_STRIP 0x00000080
1110 #define DEV_RX_OFFLOAD_HEADER_SPLIT 0x00000100
1111 #define DEV_RX_OFFLOAD_VLAN_FILTER 0x00000200
1112 #define DEV_RX_OFFLOAD_VLAN_EXTEND 0x00000400
1113 #define DEV_RX_OFFLOAD_JUMBO_FRAME 0x00000800
1114 #define DEV_RX_OFFLOAD_SCATTER 0x00002000
1115 #define DEV_RX_OFFLOAD_TIMESTAMP 0x00004000
1116 #define DEV_RX_OFFLOAD_SECURITY 0x00008000
1117 #define DEV_RX_OFFLOAD_KEEP_CRC 0x00010000
1118 #define DEV_RX_OFFLOAD_SCTP_CKSUM 0x00020000
1119 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM 0x00040000
1120 #define DEV_RX_OFFLOAD_RSS_HASH 0x00080000
1122 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
1123 DEV_RX_OFFLOAD_UDP_CKSUM | \
1124 DEV_RX_OFFLOAD_TCP_CKSUM)
1125 #define DEV_RX_OFFLOAD_VLAN (DEV_RX_OFFLOAD_VLAN_STRIP | \
1126 DEV_RX_OFFLOAD_VLAN_FILTER | \
1127 DEV_RX_OFFLOAD_VLAN_EXTEND | \
1128 DEV_RX_OFFLOAD_QINQ_STRIP)
1131 * If new Rx offload capabilities are defined, they also must be
1132 * mentioned in rte_rx_offload_names in rte_ethdev.c file.
1136 * TX offload capabilities of a device.
1138 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001
1139 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002
1140 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004
1141 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008
1142 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010
1143 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020
1144 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040
1145 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */
1146 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100
1147 #define DEV_TX_OFFLOAD_VXLAN_TNL_TSO 0x00000200 /**< Used for tunneling packet. */
1148 #define DEV_TX_OFFLOAD_GRE_TNL_TSO 0x00000400 /**< Used for tunneling packet. */
1149 #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */
1150 #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */
1151 #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000
1152 #define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000
1153 /**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same
1154 * tx queue without SW lock.
1156 #define DEV_TX_OFFLOAD_MULTI_SEGS 0x00008000
1157 /**< Device supports multi segment send. */
1158 #define DEV_TX_OFFLOAD_MBUF_FAST_FREE 0x00010000
1159 /**< Device supports optimization for fast release of mbufs.
1160 * When set application must guarantee that per-queue all mbufs comes from
1161 * the same mempool and has refcnt = 1.
1163 #define DEV_TX_OFFLOAD_SECURITY 0x00020000
1165 * Device supports generic UDP tunneled packet TSO.
1166 * Application must set PKT_TX_TUNNEL_UDP and other mbuf fields required
1169 #define DEV_TX_OFFLOAD_UDP_TNL_TSO 0x00040000
1171 * Device supports generic IP tunneled packet TSO.
1172 * Application must set PKT_TX_TUNNEL_IP and other mbuf fields required
1175 #define DEV_TX_OFFLOAD_IP_TNL_TSO 0x00080000
1176 /** Device supports outer UDP checksum */
1177 #define DEV_TX_OFFLOAD_OUTER_UDP_CKSUM 0x00100000
1179 #define RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP 0x00000001
1180 /**< Device supports Rx queue setup after device started*/
1181 #define RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP 0x00000002
1182 /**< Device supports Tx queue setup after device started*/
1185 * If new Tx offload capabilities are defined, they also must be
1186 * mentioned in rte_tx_offload_names in rte_ethdev.c file.
1190 * Fallback default preferred Rx/Tx port parameters.
1191 * These are used if an application requests default parameters
1192 * but the PMD does not provide preferred values.
1194 #define RTE_ETH_DEV_FALLBACK_RX_RINGSIZE 512
1195 #define RTE_ETH_DEV_FALLBACK_TX_RINGSIZE 512
1196 #define RTE_ETH_DEV_FALLBACK_RX_NBQUEUES 1
1197 #define RTE_ETH_DEV_FALLBACK_TX_NBQUEUES 1
1200 * Preferred Rx/Tx port parameters.
1201 * There are separate instances of this structure for transmission
1202 * and reception respectively.
1204 struct rte_eth_dev_portconf {
1205 uint16_t burst_size; /**< Device-preferred burst size */
1206 uint16_t ring_size; /**< Device-preferred size of queue rings */
1207 uint16_t nb_queues; /**< Device-preferred number of queues */
1211 * Default values for switch domain id when ethdev does not support switch
1212 * domain definitions.
1214 #define RTE_ETH_DEV_SWITCH_DOMAIN_ID_INVALID (UINT16_MAX)
1217 * Ethernet device associated switch information
1219 struct rte_eth_switch_info {
1220 const char *name; /**< switch name */
1221 uint16_t domain_id; /**< switch domain id */
1224 * mapping to the devices physical switch port as enumerated from the
1225 * perspective of the embedded interconnect/switch. For SR-IOV enabled
1226 * device this may correspond to the VF_ID of each virtual function,
1227 * but each driver should explicitly define the mapping of switch
1228 * port identifier to that physical interconnect/switch
1233 * Ethernet device information
1237 * A structure used to retrieve the contextual information of
1238 * an Ethernet device, such as the controlling driver of the
1241 struct rte_eth_dev_info {
1242 struct rte_device *device; /** Generic device information */
1243 const char *driver_name; /**< Device Driver name. */
1244 unsigned int if_index; /**< Index to bound host interface, or 0 if none.
1245 Use if_indextoname() to translate into an interface name. */
1246 uint16_t min_mtu; /**< Minimum MTU allowed */
1247 uint16_t max_mtu; /**< Maximum MTU allowed */
1248 const uint32_t *dev_flags; /**< Device flags */
1249 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */
1250 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */
1251 /** Maximum configurable size of LRO aggregated packet. */
1252 uint32_t max_lro_pkt_size;
1253 uint16_t max_rx_queues; /**< Maximum number of RX queues. */
1254 uint16_t max_tx_queues; /**< Maximum number of TX queues. */
1255 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */
1256 uint32_t max_hash_mac_addrs;
1257 /** Maximum number of hash MAC addresses for MTA and UTA. */
1258 uint16_t max_vfs; /**< Maximum number of VFs. */
1259 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
1260 uint64_t rx_offload_capa;
1261 /**< All RX offload capabilities including all per-queue ones */
1262 uint64_t tx_offload_capa;
1263 /**< All TX offload capabilities including all per-queue ones */
1264 uint64_t rx_queue_offload_capa;
1265 /**< Device per-queue RX offload capabilities. */
1266 uint64_t tx_queue_offload_capa;
1267 /**< Device per-queue TX offload capabilities. */
1269 /**< Device redirection table size, the total number of entries. */
1270 uint8_t hash_key_size; /**< Hash key size in bytes */
1271 /** Bit mask of RSS offloads, the bit offset also means flow type */
1272 uint64_t flow_type_rss_offloads;
1273 struct rte_eth_rxconf default_rxconf; /**< Default RX configuration */
1274 struct rte_eth_txconf default_txconf; /**< Default TX configuration */
1275 uint16_t vmdq_queue_base; /**< First queue ID for VMDQ pools. */
1276 uint16_t vmdq_queue_num; /**< Queue number for VMDQ pools. */
1277 uint16_t vmdq_pool_base; /**< First ID of VMDQ pools. */
1278 struct rte_eth_desc_lim rx_desc_lim; /**< RX descriptors limits */
1279 struct rte_eth_desc_lim tx_desc_lim; /**< TX descriptors limits */
1280 uint32_t speed_capa; /**< Supported speeds bitmap (ETH_LINK_SPEED_). */
1281 /** Configured number of rx/tx queues */
1282 uint16_t nb_rx_queues; /**< Number of RX queues. */
1283 uint16_t nb_tx_queues; /**< Number of TX queues. */
1284 /** Rx parameter recommendations */
1285 struct rte_eth_dev_portconf default_rxportconf;
1286 /** Tx parameter recommendations */
1287 struct rte_eth_dev_portconf default_txportconf;
1288 /** Generic device capabilities (RTE_ETH_DEV_CAPA_). */
1291 * Switching information for ports on a device with a
1292 * embedded managed interconnect/switch.
1294 struct rte_eth_switch_info switch_info;
1296 uint64_t reserved_64s[2]; /**< Reserved for future fields */
1297 void *reserved_ptrs[2]; /**< Reserved for future fields */
1301 * Ethernet device RX queue information structure.
1302 * Used to retrieve information about configured queue.
1304 struct rte_eth_rxq_info {
1305 struct rte_mempool *mp; /**< mempool used by that queue. */
1306 struct rte_eth_rxconf conf; /**< queue config parameters. */
1307 uint8_t scattered_rx; /**< scattered packets RX supported. */
1308 uint16_t nb_desc; /**< configured number of RXDs. */
1309 } __rte_cache_min_aligned;
1312 * Ethernet device TX queue information structure.
1313 * Used to retrieve information about configured queue.
1315 struct rte_eth_txq_info {
1316 struct rte_eth_txconf conf; /**< queue config parameters. */
1317 uint16_t nb_desc; /**< configured number of TXDs. */
1318 } __rte_cache_min_aligned;
1320 /* Generic Burst mode flag definition, values can be ORed. */
1323 * If the queues have different burst mode description, this bit will be set
1324 * by PMD, then the application can iterate to retrieve burst description for
1327 #define RTE_ETH_BURST_FLAG_PER_QUEUE (1ULL << 0)
1330 * Ethernet device RX/TX queue packet burst mode information structure.
1331 * Used to retrieve information about packet burst mode setting.
1333 struct rte_eth_burst_mode {
1334 uint64_t flags; /**< The ORed values of RTE_ETH_BURST_FLAG_xxx */
1336 #define RTE_ETH_BURST_MODE_INFO_SIZE 1024 /**< Maximum size for information */
1337 char info[RTE_ETH_BURST_MODE_INFO_SIZE]; /**< burst mode information */
1340 /** Maximum name length for extended statistics counters */
1341 #define RTE_ETH_XSTATS_NAME_SIZE 64
1344 * An Ethernet device extended statistic structure
1346 * This structure is used by rte_eth_xstats_get() to provide
1347 * statistics that are not provided in the generic *rte_eth_stats*
1349 * It maps a name id, corresponding to an index in the array returned
1350 * by rte_eth_xstats_get_names(), to a statistic value.
1352 struct rte_eth_xstat {
1353 uint64_t id; /**< The index in xstats name array. */
1354 uint64_t value; /**< The statistic counter value. */
1358 * A name element for extended statistics.
1360 * An array of this structure is returned by rte_eth_xstats_get_names().
1361 * It lists the names of extended statistics for a PMD. The *rte_eth_xstat*
1362 * structure references these names by their array index.
1364 struct rte_eth_xstat_name {
1365 char name[RTE_ETH_XSTATS_NAME_SIZE]; /**< The statistic name. */
1368 #define ETH_DCB_NUM_TCS 8
1369 #define ETH_MAX_VMDQ_POOL 64
1372 * A structure used to get the information of queue and
1373 * TC mapping on both TX and RX paths.
1375 struct rte_eth_dcb_tc_queue_mapping {
1376 /** rx queues assigned to tc per Pool */
1380 } tc_rxq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1381 /** rx queues assigned to tc per Pool */
1385 } tc_txq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1389 * A structure used to get the information of DCB.
1390 * It includes TC UP mapping and queue TC mapping.
1392 struct rte_eth_dcb_info {
1393 uint8_t nb_tcs; /**< number of TCs */
1394 uint8_t prio_tc[ETH_DCB_NUM_USER_PRIORITIES]; /**< Priority to tc */
1395 uint8_t tc_bws[ETH_DCB_NUM_TCS]; /**< TX BW percentage for each TC */
1396 /** rx queues assigned to tc */
1397 struct rte_eth_dcb_tc_queue_mapping tc_queue;
1400 #define RTE_ETH_ALL RTE_MAX_ETHPORTS
1402 /* Macros to check for valid port */
1403 #define RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, retval) do { \
1404 if (!rte_eth_dev_is_valid_port(port_id)) { \
1405 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1410 #define RTE_ETH_VALID_PORTID_OR_RET(port_id) do { \
1411 if (!rte_eth_dev_is_valid_port(port_id)) { \
1412 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1418 * l2 tunnel configuration.
1421 /**< l2 tunnel enable mask */
1422 #define ETH_L2_TUNNEL_ENABLE_MASK 0x00000001
1423 /**< l2 tunnel insertion mask */
1424 #define ETH_L2_TUNNEL_INSERTION_MASK 0x00000002
1425 /**< l2 tunnel stripping mask */
1426 #define ETH_L2_TUNNEL_STRIPPING_MASK 0x00000004
1427 /**< l2 tunnel forwarding mask */
1428 #define ETH_L2_TUNNEL_FORWARDING_MASK 0x00000008
1431 * Function type used for RX packet processing packet callbacks.
1433 * The callback function is called on RX with a burst of packets that have
1434 * been received on the given port and queue.
1437 * The Ethernet port on which RX is being performed.
1439 * The queue on the Ethernet port which is being used to receive the packets.
1441 * The burst of packets that have just been received.
1443 * The number of packets in the burst pointed to by "pkts".
1445 * The max number of packets that can be stored in the "pkts" array.
1447 * The arbitrary user parameter passed in by the application when the callback
1448 * was originally configured.
1450 * The number of packets returned to the user.
1452 typedef uint16_t (*rte_rx_callback_fn)(uint16_t port_id, uint16_t queue,
1453 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts,
1457 * Function type used for TX packet processing packet callbacks.
1459 * The callback function is called on TX with a burst of packets immediately
1460 * before the packets are put onto the hardware queue for transmission.
1463 * The Ethernet port on which TX is being performed.
1465 * The queue on the Ethernet port which is being used to transmit the packets.
1467 * The burst of packets that are about to be transmitted.
1469 * The number of packets in the burst pointed to by "pkts".
1471 * The arbitrary user parameter passed in by the application when the callback
1472 * was originally configured.
1474 * The number of packets to be written to the NIC.
1476 typedef uint16_t (*rte_tx_callback_fn)(uint16_t port_id, uint16_t queue,
1477 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param);
1480 * Possible states of an ethdev port.
1482 enum rte_eth_dev_state {
1483 /** Device is unused before being probed. */
1484 RTE_ETH_DEV_UNUSED = 0,
1485 /** Device is attached when allocated in probing. */
1486 RTE_ETH_DEV_ATTACHED,
1487 /** Device is in removed state when plug-out is detected. */
1488 RTE_ETH_DEV_REMOVED,
1491 struct rte_eth_dev_sriov {
1492 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */
1493 uint8_t nb_q_per_pool; /**< rx queue number per pool */
1494 uint16_t def_vmdq_idx; /**< Default pool num used for PF */
1495 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */
1497 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
1499 #define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN
1501 #define RTE_ETH_DEV_NO_OWNER 0
1503 #define RTE_ETH_MAX_OWNER_NAME_LEN 64
1505 struct rte_eth_dev_owner {
1506 uint64_t id; /**< The owner unique identifier. */
1507 char name[RTE_ETH_MAX_OWNER_NAME_LEN]; /**< The owner name. */
1511 * Port is released (i.e. totally freed and data erased) on close.
1512 * Temporary flag for PMD migration to new rte_eth_dev_close() behaviour.
1514 #define RTE_ETH_DEV_CLOSE_REMOVE 0x0001
1515 /** Device supports link state interrupt */
1516 #define RTE_ETH_DEV_INTR_LSC 0x0002
1517 /** Device is a bonded slave */
1518 #define RTE_ETH_DEV_BONDED_SLAVE 0x0004
1519 /** Device supports device removal interrupt */
1520 #define RTE_ETH_DEV_INTR_RMV 0x0008
1521 /** Device is port representor */
1522 #define RTE_ETH_DEV_REPRESENTOR 0x0010
1523 /** Device does not support MAC change after started */
1524 #define RTE_ETH_DEV_NOLIVE_MAC_ADDR 0x0020
1527 * Iterates over valid ethdev ports owned by a specific owner.
1530 * The id of the next possible valid owned port.
1532 * The owner identifier.
1533 * RTE_ETH_DEV_NO_OWNER means iterate over all valid ownerless ports.
1535 * Next valid port id owned by owner_id, RTE_MAX_ETHPORTS if there is none.
1537 uint64_t rte_eth_find_next_owned_by(uint16_t port_id,
1538 const uint64_t owner_id);
1541 * Macro to iterate over all enabled ethdev ports owned by a specific owner.
1543 #define RTE_ETH_FOREACH_DEV_OWNED_BY(p, o) \
1544 for (p = rte_eth_find_next_owned_by(0, o); \
1545 (unsigned int)p < (unsigned int)RTE_MAX_ETHPORTS; \
1546 p = rte_eth_find_next_owned_by(p + 1, o))
1549 * Iterates over valid ethdev ports.
1552 * The id of the next possible valid port.
1554 * Next valid port id, RTE_MAX_ETHPORTS if there is none.
1556 uint16_t rte_eth_find_next(uint16_t port_id);
1559 * Macro to iterate over all enabled and ownerless ethdev ports.
1561 #define RTE_ETH_FOREACH_DEV(p) \
1562 RTE_ETH_FOREACH_DEV_OWNED_BY(p, RTE_ETH_DEV_NO_OWNER)
1566 * @b EXPERIMENTAL: this API may change without prior notice.
1568 * Iterates over ethdev ports of a specified device.
1570 * @param port_id_start
1571 * The id of the next possible valid port.
1573 * The generic device behind the ports to iterate.
1575 * Next port id of the device, possibly port_id_start,
1576 * RTE_MAX_ETHPORTS if there is none.
1580 rte_eth_find_next_of(uint16_t port_id_start,
1581 const struct rte_device *parent);
1584 * Macro to iterate over all ethdev ports of a specified device.
1587 * The id of the matching port being iterated.
1589 * The rte_device pointer matching the iterated ports.
1591 #define RTE_ETH_FOREACH_DEV_OF(port_id, parent) \
1592 for (port_id = rte_eth_find_next_of(0, parent); \
1593 port_id < RTE_MAX_ETHPORTS; \
1594 port_id = rte_eth_find_next_of(port_id + 1, parent))
1598 * @b EXPERIMENTAL: this API may change without prior notice.
1600 * Iterates over sibling ethdev ports (i.e. sharing the same rte_device).
1602 * @param port_id_start
1603 * The id of the next possible valid sibling port.
1604 * @param ref_port_id
1605 * The id of a reference port to compare rte_device with.
1607 * Next sibling port id, possibly port_id_start or ref_port_id itself,
1608 * RTE_MAX_ETHPORTS if there is none.
1612 rte_eth_find_next_sibling(uint16_t port_id_start,
1613 uint16_t ref_port_id);
1616 * Macro to iterate over all ethdev ports sharing the same rte_device
1617 * as the specified port.
1618 * Note: the specified reference port is part of the loop iterations.
1621 * The id of the matching port being iterated.
1622 * @param ref_port_id
1623 * The id of the port being compared.
1625 #define RTE_ETH_FOREACH_DEV_SIBLING(port_id, ref_port_id) \
1626 for (port_id = rte_eth_find_next_sibling(0, ref_port_id); \
1627 port_id < RTE_MAX_ETHPORTS; \
1628 port_id = rte_eth_find_next_sibling(port_id + 1, ref_port_id))
1632 * @b EXPERIMENTAL: this API may change without prior notice.
1634 * Get a new unique owner identifier.
1635 * An owner identifier is used to owns Ethernet devices by only one DPDK entity
1636 * to avoid multiple management of device by different entities.
1639 * Owner identifier pointer.
1641 * Negative errno value on error, 0 on success.
1644 int rte_eth_dev_owner_new(uint64_t *owner_id);
1648 * @b EXPERIMENTAL: this API may change without prior notice.
1650 * Set an Ethernet device owner.
1653 * The identifier of the port to own.
1655 * The owner pointer.
1657 * Negative errno value on error, 0 on success.
1660 int rte_eth_dev_owner_set(const uint16_t port_id,
1661 const struct rte_eth_dev_owner *owner);
1665 * @b EXPERIMENTAL: this API may change without prior notice.
1667 * Unset Ethernet device owner to make the device ownerless.
1670 * The identifier of port to make ownerless.
1672 * The owner identifier.
1674 * 0 on success, negative errno value on error.
1677 int rte_eth_dev_owner_unset(const uint16_t port_id,
1678 const uint64_t owner_id);
1682 * @b EXPERIMENTAL: this API may change without prior notice.
1684 * Remove owner from all Ethernet devices owned by a specific owner.
1687 * The owner identifier.
1689 * 0 on success, negative errno value on error.
1692 int rte_eth_dev_owner_delete(const uint64_t owner_id);
1696 * @b EXPERIMENTAL: this API may change without prior notice.
1698 * Get the owner of an Ethernet device.
1701 * The port identifier.
1703 * The owner structure pointer to fill.
1705 * 0 on success, negative errno value on error..
1708 int rte_eth_dev_owner_get(const uint16_t port_id,
1709 struct rte_eth_dev_owner *owner);
1712 * Get the number of ports which are usable for the application.
1714 * These devices must be iterated by using the macro
1715 * ``RTE_ETH_FOREACH_DEV`` or ``RTE_ETH_FOREACH_DEV_OWNED_BY``
1716 * to deal with non-contiguous ranges of devices.
1719 * The count of available Ethernet devices.
1721 uint16_t rte_eth_dev_count_avail(void);
1724 * Get the total number of ports which are allocated.
1726 * Some devices may not be available for the application.
1729 * The total count of Ethernet devices.
1731 uint16_t rte_eth_dev_count_total(void);
1734 * Convert a numerical speed in Mbps to a bitmap flag that can be used in
1735 * the bitmap link_speeds of the struct rte_eth_conf
1738 * Numerical speed value in Mbps
1740 * ETH_LINK_[HALF/FULL]_DUPLEX (only for 10/100M speeds)
1742 * 0 if the speed cannot be mapped
1744 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex);
1747 * Get DEV_RX_OFFLOAD_* flag name.
1752 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
1754 const char *rte_eth_dev_rx_offload_name(uint64_t offload);
1757 * Get DEV_TX_OFFLOAD_* flag name.
1762 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
1764 const char *rte_eth_dev_tx_offload_name(uint64_t offload);
1767 * Configure an Ethernet device.
1768 * This function must be invoked first before any other function in the
1769 * Ethernet API. This function can also be re-invoked when a device is in the
1773 * The port identifier of the Ethernet device to configure.
1774 * @param nb_rx_queue
1775 * The number of receive queues to set up for the Ethernet device.
1776 * @param nb_tx_queue
1777 * The number of transmit queues to set up for the Ethernet device.
1779 * The pointer to the configuration data to be used for the Ethernet device.
1780 * The *rte_eth_conf* structure includes:
1781 * - the hardware offload features to activate, with dedicated fields for
1782 * each statically configurable offload hardware feature provided by
1783 * Ethernet devices, such as IP checksum or VLAN tag stripping for
1785 * The Rx offload bitfield API is obsolete and will be deprecated.
1786 * Applications should set the ignore_bitfield_offloads bit on *rxmode*
1787 * structure and use offloads field to set per-port offloads instead.
1788 * - Any offloading set in eth_conf->[rt]xmode.offloads must be within
1789 * the [rt]x_offload_capa returned from rte_eth_dev_info_get().
1790 * Any type of device supported offloading set in the input argument
1791 * eth_conf->[rt]xmode.offloads to rte_eth_dev_configure() is enabled
1792 * on all queues and it can't be disabled in rte_eth_[rt]x_queue_setup()
1793 * - the Receive Side Scaling (RSS) configuration when using multiple RX
1794 * queues per port. Any RSS hash function set in eth_conf->rss_conf.rss_hf
1795 * must be within the flow_type_rss_offloads provided by drivers via
1796 * rte_eth_dev_info_get() API.
1798 * Embedding all configuration information in a single data structure
1799 * is the more flexible method that allows the addition of new features
1800 * without changing the syntax of the API.
1802 * - 0: Success, device configured.
1803 * - <0: Error code returned by the driver configuration function.
1805 int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
1806 uint16_t nb_tx_queue, const struct rte_eth_conf *eth_conf);
1810 * @b EXPERIMENTAL: this API may change without prior notice.
1812 * Check if an Ethernet device was physically removed.
1815 * The port identifier of the Ethernet device.
1817 * 1 when the Ethernet device is removed, otherwise 0.
1821 rte_eth_dev_is_removed(uint16_t port_id);
1824 * Allocate and set up a receive queue for an Ethernet device.
1826 * The function allocates a contiguous block of memory for *nb_rx_desc*
1827 * receive descriptors from a memory zone associated with *socket_id*
1828 * and initializes each receive descriptor with a network buffer allocated
1829 * from the memory pool *mb_pool*.
1832 * The port identifier of the Ethernet device.
1833 * @param rx_queue_id
1834 * The index of the receive queue to set up.
1835 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1836 * to rte_eth_dev_configure().
1838 * The number of receive descriptors to allocate for the receive ring.
1840 * The *socket_id* argument is the socket identifier in case of NUMA.
1841 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1842 * the DMA memory allocated for the receive descriptors of the ring.
1844 * The pointer to the configuration data to be used for the receive queue.
1845 * NULL value is allowed, in which case default RX configuration
1847 * The *rx_conf* structure contains an *rx_thresh* structure with the values
1848 * of the Prefetch, Host, and Write-Back threshold registers of the receive
1850 * In addition it contains the hardware offloads features to activate using
1851 * the DEV_RX_OFFLOAD_* flags.
1852 * If an offloading set in rx_conf->offloads
1853 * hasn't been set in the input argument eth_conf->rxmode.offloads
1854 * to rte_eth_dev_configure(), it is a new added offloading, it must be
1855 * per-queue type and it is enabled for the queue.
1856 * No need to repeat any bit in rx_conf->offloads which has already been
1857 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
1858 * at port level can't be disabled at queue level.
1860 * The pointer to the memory pool from which to allocate *rte_mbuf* network
1861 * memory buffers to populate each descriptor of the receive ring.
1863 * - 0: Success, receive queue correctly set up.
1864 * - -EIO: if device is removed.
1865 * - -EINVAL: The memory pool pointer is null or the size of network buffers
1866 * which can be allocated from this memory pool does not fit the various
1867 * buffer sizes allowed by the device controller.
1868 * - -ENOMEM: Unable to allocate the receive ring descriptors or to
1869 * allocate network memory buffers from the memory pool when
1870 * initializing receive descriptors.
1872 int rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
1873 uint16_t nb_rx_desc, unsigned int socket_id,
1874 const struct rte_eth_rxconf *rx_conf,
1875 struct rte_mempool *mb_pool);
1879 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1881 * Allocate and set up a hairpin receive queue for an Ethernet device.
1883 * The function set up the selected queue to be used in hairpin.
1886 * The port identifier of the Ethernet device.
1887 * @param rx_queue_id
1888 * The index of the receive queue to set up.
1889 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1890 * to rte_eth_dev_configure().
1892 * The number of receive descriptors to allocate for the receive ring.
1893 * 0 means the PMD will use default value.
1895 * The pointer to the hairpin configuration.
1898 * - (0) if successful.
1899 * - (-ENOTSUP) if hardware doesn't support.
1900 * - (-EINVAL) if bad parameter.
1901 * - (-ENOMEM) if unable to allocate the resources.
1904 int rte_eth_rx_hairpin_queue_setup
1905 (uint16_t port_id, uint16_t rx_queue_id, uint16_t nb_rx_desc,
1906 const struct rte_eth_hairpin_conf *conf);
1909 * Allocate and set up a transmit queue for an Ethernet device.
1912 * The port identifier of the Ethernet device.
1913 * @param tx_queue_id
1914 * The index of the transmit queue to set up.
1915 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1916 * to rte_eth_dev_configure().
1918 * The number of transmit descriptors to allocate for the transmit ring.
1920 * The *socket_id* argument is the socket identifier in case of NUMA.
1921 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1922 * the DMA memory allocated for the transmit descriptors of the ring.
1924 * The pointer to the configuration data to be used for the transmit queue.
1925 * NULL value is allowed, in which case default TX configuration
1927 * The *tx_conf* structure contains the following data:
1928 * - The *tx_thresh* structure with the values of the Prefetch, Host, and
1929 * Write-Back threshold registers of the transmit ring.
1930 * When setting Write-Back threshold to the value greater then zero,
1931 * *tx_rs_thresh* value should be explicitly set to one.
1932 * - The *tx_free_thresh* value indicates the [minimum] number of network
1933 * buffers that must be pending in the transmit ring to trigger their
1934 * [implicit] freeing by the driver transmit function.
1935 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit
1936 * descriptors that must be pending in the transmit ring before setting the
1937 * RS bit on a descriptor by the driver transmit function.
1938 * The *tx_rs_thresh* value should be less or equal then
1939 * *tx_free_thresh* value, and both of them should be less then
1941 * - The *offloads* member contains Tx offloads to be enabled.
1942 * If an offloading set in tx_conf->offloads
1943 * hasn't been set in the input argument eth_conf->txmode.offloads
1944 * to rte_eth_dev_configure(), it is a new added offloading, it must be
1945 * per-queue type and it is enabled for the queue.
1946 * No need to repeat any bit in tx_conf->offloads which has already been
1947 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
1948 * at port level can't be disabled at queue level.
1950 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces
1951 * the transmit function to use default values.
1953 * - 0: Success, the transmit queue is correctly set up.
1954 * - -ENOMEM: Unable to allocate the transmit ring descriptors.
1956 int rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,
1957 uint16_t nb_tx_desc, unsigned int socket_id,
1958 const struct rte_eth_txconf *tx_conf);
1962 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
1964 * Allocate and set up a transmit hairpin queue for an Ethernet device.
1967 * The port identifier of the Ethernet device.
1968 * @param tx_queue_id
1969 * The index of the transmit queue to set up.
1970 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1971 * to rte_eth_dev_configure().
1973 * The number of transmit descriptors to allocate for the transmit ring.
1974 * 0 to set default PMD value.
1976 * The hairpin configuration.
1979 * - (0) if successful.
1980 * - (-ENOTSUP) if hardware doesn't support.
1981 * - (-EINVAL) if bad parameter.
1982 * - (-ENOMEM) if unable to allocate the resources.
1985 int rte_eth_tx_hairpin_queue_setup
1986 (uint16_t port_id, uint16_t tx_queue_id, uint16_t nb_tx_desc,
1987 const struct rte_eth_hairpin_conf *conf);
1990 * Return the NUMA socket to which an Ethernet device is connected
1993 * The port identifier of the Ethernet device
1995 * The NUMA socket id to which the Ethernet device is connected or
1996 * a default of zero if the socket could not be determined.
1997 * -1 is returned is the port_id value is out of range.
1999 int rte_eth_dev_socket_id(uint16_t port_id);
2002 * Check if port_id of device is attached
2005 * The port identifier of the Ethernet device
2007 * - 0 if port is out of range or not attached
2008 * - 1 if device is attached
2010 int rte_eth_dev_is_valid_port(uint16_t port_id);
2013 * Start specified RX queue of a port. It is used when rx_deferred_start
2014 * flag of the specified queue is true.
2017 * The port identifier of the Ethernet device
2018 * @param rx_queue_id
2019 * The index of the rx queue to update the ring.
2020 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2021 * to rte_eth_dev_configure().
2023 * - 0: Success, the receive queue is started.
2024 * - -EINVAL: The port_id or the queue_id out of range or belong to hairpin.
2025 * - -EIO: if device is removed.
2026 * - -ENOTSUP: The function not supported in PMD driver.
2028 int rte_eth_dev_rx_queue_start(uint16_t port_id, uint16_t rx_queue_id);
2031 * Stop specified RX queue of a port
2034 * The port identifier of the Ethernet device
2035 * @param rx_queue_id
2036 * The index of the rx queue to update the ring.
2037 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2038 * to rte_eth_dev_configure().
2040 * - 0: Success, the receive queue is stopped.
2041 * - -EINVAL: The port_id or the queue_id out of range or belong to hairpin.
2042 * - -EIO: if device is removed.
2043 * - -ENOTSUP: The function not supported in PMD driver.
2045 int rte_eth_dev_rx_queue_stop(uint16_t port_id, uint16_t rx_queue_id);
2048 * Start TX for specified queue of a port. It is used when tx_deferred_start
2049 * flag of the specified queue is true.
2052 * The port identifier of the Ethernet device
2053 * @param tx_queue_id
2054 * The index of the tx queue to update the ring.
2055 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2056 * to rte_eth_dev_configure().
2058 * - 0: Success, the transmit queue is started.
2059 * - -EINVAL: The port_id or the queue_id out of range or belong to hairpin.
2060 * - -EIO: if device is removed.
2061 * - -ENOTSUP: The function not supported in PMD driver.
2063 int rte_eth_dev_tx_queue_start(uint16_t port_id, uint16_t tx_queue_id);
2066 * Stop specified TX queue of a port
2069 * The port identifier of the Ethernet device
2070 * @param tx_queue_id
2071 * The index of the tx queue to update the ring.
2072 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2073 * to rte_eth_dev_configure().
2075 * - 0: Success, the transmit queue is stopped.
2076 * - -EINVAL: The port_id or the queue_id out of range or belong to hairpin.
2077 * - -EIO: if device is removed.
2078 * - -ENOTSUP: The function not supported in PMD driver.
2080 int rte_eth_dev_tx_queue_stop(uint16_t port_id, uint16_t tx_queue_id);
2083 * Start an Ethernet device.
2085 * The device start step is the last one and consists of setting the configured
2086 * offload features and in starting the transmit and the receive units of the
2089 * Device RTE_ETH_DEV_NOLIVE_MAC_ADDR flag causes MAC address to be set before
2090 * PMD port start callback function is invoked.
2092 * On success, all basic functions exported by the Ethernet API (link status,
2093 * receive/transmit, and so on) can be invoked.
2096 * The port identifier of the Ethernet device.
2098 * - 0: Success, Ethernet device started.
2099 * - <0: Error code of the driver device start function.
2101 int rte_eth_dev_start(uint16_t port_id);
2104 * Stop an Ethernet device. The device can be restarted with a call to
2105 * rte_eth_dev_start()
2108 * The port identifier of the Ethernet device.
2110 void rte_eth_dev_stop(uint16_t port_id);
2113 * Link up an Ethernet device.
2115 * Set device link up will re-enable the device rx/tx
2116 * functionality after it is previously set device linked down.
2119 * The port identifier of the Ethernet device.
2121 * - 0: Success, Ethernet device linked up.
2122 * - <0: Error code of the driver device link up function.
2124 int rte_eth_dev_set_link_up(uint16_t port_id);
2127 * Link down an Ethernet device.
2128 * The device rx/tx functionality will be disabled if success,
2129 * and it can be re-enabled with a call to
2130 * rte_eth_dev_set_link_up()
2133 * The port identifier of the Ethernet device.
2135 int rte_eth_dev_set_link_down(uint16_t port_id);
2138 * Close a stopped Ethernet device. The device cannot be restarted!
2139 * The function frees all port resources if the driver supports
2140 * the flag RTE_ETH_DEV_CLOSE_REMOVE.
2143 * The port identifier of the Ethernet device.
2145 void rte_eth_dev_close(uint16_t port_id);
2148 * Reset a Ethernet device and keep its port id.
2150 * When a port has to be reset passively, the DPDK application can invoke
2151 * this function. For example when a PF is reset, all its VFs should also
2152 * be reset. Normally a DPDK application can invoke this function when
2153 * RTE_ETH_EVENT_INTR_RESET event is detected, but can also use it to start
2154 * a port reset in other circumstances.
2156 * When this function is called, it first stops the port and then calls the
2157 * PMD specific dev_uninit( ) and dev_init( ) to return the port to initial
2158 * state, in which no Tx and Rx queues are setup, as if the port has been
2159 * reset and not started. The port keeps the port id it had before the
2162 * After calling rte_eth_dev_reset( ), the application should use
2163 * rte_eth_dev_configure( ), rte_eth_rx_queue_setup( ),
2164 * rte_eth_tx_queue_setup( ), and rte_eth_dev_start( )
2165 * to reconfigure the device as appropriate.
2167 * Note: To avoid unexpected behavior, the application should stop calling
2168 * Tx and Rx functions before calling rte_eth_dev_reset( ). For thread
2169 * safety, all these controlling functions should be called from the same
2173 * The port identifier of the Ethernet device.
2176 * - (0) if successful.
2177 * - (-EINVAL) if port identifier is invalid.
2178 * - (-ENOTSUP) if hardware doesn't support this function.
2179 * - (-EPERM) if not ran from the primary process.
2180 * - (-EIO) if re-initialisation failed or device is removed.
2181 * - (-ENOMEM) if the reset failed due to OOM.
2182 * - (-EAGAIN) if the reset temporarily failed and should be retried later.
2184 int rte_eth_dev_reset(uint16_t port_id);
2187 * Enable receipt in promiscuous mode for an Ethernet device.
2190 * The port identifier of the Ethernet device.
2192 * - (0) if successful.
2193 * - (-ENOTSUP) if support for promiscuous_enable() does not exist
2195 * - (-ENODEV) if *port_id* invalid.
2197 int rte_eth_promiscuous_enable(uint16_t port_id);
2200 * Disable receipt in promiscuous mode for an Ethernet device.
2203 * The port identifier of the Ethernet device.
2205 * - (0) if successful.
2206 * - (-ENOTSUP) if support for promiscuous_disable() does not exist
2208 * - (-ENODEV) if *port_id* invalid.
2210 int rte_eth_promiscuous_disable(uint16_t port_id);
2213 * Return the value of promiscuous mode for an Ethernet device.
2216 * The port identifier of the Ethernet device.
2218 * - (1) if promiscuous is enabled
2219 * - (0) if promiscuous is disabled.
2222 int rte_eth_promiscuous_get(uint16_t port_id);
2225 * Enable the receipt of any multicast frame by an Ethernet device.
2228 * The port identifier of the Ethernet device.
2230 * - (0) if successful.
2231 * - (-ENOTSUP) if support for allmulticast_enable() does not exist
2233 * - (-ENODEV) if *port_id* invalid.
2235 int rte_eth_allmulticast_enable(uint16_t port_id);
2238 * Disable the receipt of all multicast frames by an Ethernet device.
2241 * The port identifier of the Ethernet device.
2243 * - (0) if successful.
2244 * - (-ENOTSUP) if support for allmulticast_disable() does not exist
2246 * - (-ENODEV) if *port_id* invalid.
2248 int rte_eth_allmulticast_disable(uint16_t port_id);
2251 * Return the value of allmulticast mode for an Ethernet device.
2254 * The port identifier of the Ethernet device.
2256 * - (1) if allmulticast is enabled
2257 * - (0) if allmulticast is disabled.
2260 int rte_eth_allmulticast_get(uint16_t port_id);
2263 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2264 * or FULL-DUPLEX) of the physical link of an Ethernet device. It might need
2265 * to wait up to 9 seconds in it.
2268 * The port identifier of the Ethernet device.
2270 * A pointer to an *rte_eth_link* structure to be filled with
2271 * the status, the speed and the mode of the Ethernet device link.
2273 * - (0) if successful.
2274 * - (-ENOTSUP) if the function is not supported in PMD driver.
2275 * - (-ENODEV) if *port_id* invalid.
2277 int rte_eth_link_get(uint16_t port_id, struct rte_eth_link *link);
2280 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2281 * or FULL-DUPLEX) of the physical link of an Ethernet device. It is a no-wait
2282 * version of rte_eth_link_get().
2285 * The port identifier of the Ethernet device.
2287 * A pointer to an *rte_eth_link* structure to be filled with
2288 * the status, the speed and the mode of the Ethernet device link.
2290 * - (0) if successful.
2291 * - (-ENOTSUP) if the function is not supported in PMD driver.
2292 * - (-ENODEV) if *port_id* invalid.
2294 int rte_eth_link_get_nowait(uint16_t port_id, struct rte_eth_link *link);
2297 * Retrieve the general I/O statistics of an Ethernet device.
2300 * The port identifier of the Ethernet device.
2302 * A pointer to a structure of type *rte_eth_stats* to be filled with
2303 * the values of device counters for the following set of statistics:
2304 * - *ipackets* with the total of successfully received packets.
2305 * - *opackets* with the total of successfully transmitted packets.
2306 * - *ibytes* with the total of successfully received bytes.
2307 * - *obytes* with the total of successfully transmitted bytes.
2308 * - *ierrors* with the total of erroneous received packets.
2309 * - *oerrors* with the total of failed transmitted packets.
2311 * Zero if successful. Non-zero otherwise.
2313 int rte_eth_stats_get(uint16_t port_id, struct rte_eth_stats *stats);
2316 * Reset the general I/O statistics of an Ethernet device.
2319 * The port identifier of the Ethernet device.
2321 * - (0) if device notified to reset stats.
2322 * - (-ENOTSUP) if hardware doesn't support.
2323 * - (-ENODEV) if *port_id* invalid.
2324 * - (<0): Error code of the driver stats reset function.
2326 int rte_eth_stats_reset(uint16_t port_id);
2329 * Retrieve names of extended statistics of an Ethernet device.
2331 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2333 * xstats_names[i].name => xstats[i].value
2335 * And the array index is same with id field of 'struct rte_eth_xstat':
2338 * This assumption makes key-value pair matching less flexible but simpler.
2341 * The port identifier of the Ethernet device.
2342 * @param xstats_names
2343 * An rte_eth_xstat_name array of at least *size* elements to
2344 * be filled. If set to NULL, the function returns the required number
2347 * The size of the xstats_names array (number of elements).
2349 * - A positive value lower or equal to size: success. The return value
2350 * is the number of entries filled in the stats table.
2351 * - A positive value higher than size: error, the given statistics table
2352 * is too small. The return value corresponds to the size that should
2353 * be given to succeed. The entries in the table are not valid and
2354 * shall not be used by the caller.
2355 * - A negative value on error (invalid port id).
2357 int rte_eth_xstats_get_names(uint16_t port_id,
2358 struct rte_eth_xstat_name *xstats_names,
2362 * Retrieve extended statistics of an Ethernet device.
2364 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2366 * xstats_names[i].name => xstats[i].value
2368 * And the array index is same with id field of 'struct rte_eth_xstat':
2371 * This assumption makes key-value pair matching less flexible but simpler.
2374 * The port identifier of the Ethernet device.
2376 * A pointer to a table of structure of type *rte_eth_xstat*
2377 * to be filled with device statistics ids and values.
2378 * This parameter can be set to NULL if n is 0.
2380 * The size of the xstats array (number of elements).
2382 * - A positive value lower or equal to n: success. The return value
2383 * is the number of entries filled in the stats table.
2384 * - A positive value higher than n: error, the given statistics table
2385 * is too small. The return value corresponds to the size that should
2386 * be given to succeed. The entries in the table are not valid and
2387 * shall not be used by the caller.
2388 * - A negative value on error (invalid port id).
2390 int rte_eth_xstats_get(uint16_t port_id, struct rte_eth_xstat *xstats,
2394 * Retrieve names of extended statistics of an Ethernet device.
2397 * The port identifier of the Ethernet device.
2398 * @param xstats_names
2399 * An rte_eth_xstat_name array of at least *size* elements to
2400 * be filled. If set to NULL, the function returns the required number
2403 * IDs array given by app to retrieve specific statistics
2405 * The size of the xstats_names array (number of elements).
2407 * - A positive value lower or equal to size: success. The return value
2408 * is the number of entries filled in the stats table.
2409 * - A positive value higher than size: error, the given statistics table
2410 * is too small. The return value corresponds to the size that should
2411 * be given to succeed. The entries in the table are not valid and
2412 * shall not be used by the caller.
2413 * - A negative value on error (invalid port id).
2416 rte_eth_xstats_get_names_by_id(uint16_t port_id,
2417 struct rte_eth_xstat_name *xstats_names, unsigned int size,
2421 * Retrieve extended statistics of an Ethernet device.
2424 * The port identifier of the Ethernet device.
2426 * A pointer to an ids array passed by application. This tells which
2427 * statistics values function should retrieve. This parameter
2428 * can be set to NULL if size is 0. In this case function will retrieve
2429 * all available statistics.
2431 * A pointer to a table to be filled with device statistics values.
2433 * The size of the ids array (number of elements).
2435 * - A positive value lower or equal to size: success. The return value
2436 * is the number of entries filled in the stats table.
2437 * - A positive value higher than size: error, the given statistics table
2438 * is too small. The return value corresponds to the size that should
2439 * be given to succeed. The entries in the table are not valid and
2440 * shall not be used by the caller.
2441 * - A negative value on error (invalid port id).
2443 int rte_eth_xstats_get_by_id(uint16_t port_id, const uint64_t *ids,
2444 uint64_t *values, unsigned int size);
2447 * Gets the ID of a statistic from its name.
2449 * This function searches for the statistics using string compares, and
2450 * as such should not be used on the fast-path. For fast-path retrieval of
2451 * specific statistics, store the ID as provided in *id* from this function,
2452 * and pass the ID to rte_eth_xstats_get()
2454 * @param port_id The port to look up statistics from
2455 * @param xstat_name The name of the statistic to return
2456 * @param[out] id A pointer to an app-supplied uint64_t which should be
2457 * set to the ID of the stat if the stat exists.
2460 * -ENODEV for invalid port_id,
2461 * -EIO if device is removed,
2462 * -EINVAL if the xstat_name doesn't exist in port_id
2464 int rte_eth_xstats_get_id_by_name(uint16_t port_id, const char *xstat_name,
2468 * Reset extended statistics of an Ethernet device.
2471 * The port identifier of the Ethernet device.
2473 * - (0) if device notified to reset extended stats.
2474 * - (-ENOTSUP) if pmd doesn't support both
2475 * extended stats and basic stats reset.
2476 * - (-ENODEV) if *port_id* invalid.
2477 * - (<0): Error code of the driver xstats reset function.
2479 int rte_eth_xstats_reset(uint16_t port_id);
2482 * Set a mapping for the specified transmit queue to the specified per-queue
2483 * statistics counter.
2486 * The port identifier of the Ethernet device.
2487 * @param tx_queue_id
2488 * The index of the transmit queue for which a queue stats mapping is required.
2489 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2490 * to rte_eth_dev_configure().
2492 * The per-queue packet statistics functionality number that the transmit
2493 * queue is to be assigned.
2494 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
2496 * Zero if successful. Non-zero otherwise.
2498 int rte_eth_dev_set_tx_queue_stats_mapping(uint16_t port_id,
2499 uint16_t tx_queue_id, uint8_t stat_idx);
2502 * Set a mapping for the specified receive queue to the specified per-queue
2503 * statistics counter.
2506 * The port identifier of the Ethernet device.
2507 * @param rx_queue_id
2508 * The index of the receive queue for which a queue stats mapping is required.
2509 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2510 * to rte_eth_dev_configure().
2512 * The per-queue packet statistics functionality number that the receive
2513 * queue is to be assigned.
2514 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
2516 * Zero if successful. Non-zero otherwise.
2518 int rte_eth_dev_set_rx_queue_stats_mapping(uint16_t port_id,
2519 uint16_t rx_queue_id,
2523 * Retrieve the Ethernet address of an Ethernet device.
2526 * The port identifier of the Ethernet device.
2528 * A pointer to a structure of type *ether_addr* to be filled with
2529 * the Ethernet address of the Ethernet device.
2531 * - (0) if successful
2532 * - (-ENODEV) if *port_id* invalid.
2534 int rte_eth_macaddr_get(uint16_t port_id, struct rte_ether_addr *mac_addr);
2537 * Retrieve the contextual information of an Ethernet device.
2539 * As part of this function, a number of of fields in dev_info will be
2540 * initialized as follows:
2545 * Where lim is defined within the rte_eth_dev_info_get as
2547 * const struct rte_eth_desc_lim lim = {
2548 * .nb_max = UINT16_MAX,
2551 * .nb_seg_max = UINT16_MAX,
2552 * .nb_mtu_seg_max = UINT16_MAX,
2555 * device = dev->device
2556 * min_mtu = RTE_ETHER_MIN_MTU
2557 * max_mtu = UINT16_MAX
2559 * The following fields will be populated if support for dev_infos_get()
2560 * exists for the device and the rte_eth_dev 'dev' has been populated
2561 * successfully with a call to it:
2563 * driver_name = dev->device->driver->name
2564 * nb_rx_queues = dev->data->nb_rx_queues
2565 * nb_tx_queues = dev->data->nb_tx_queues
2566 * dev_flags = &dev->data->dev_flags
2569 * The port identifier of the Ethernet device.
2571 * A pointer to a structure of type *rte_eth_dev_info* to be filled with
2572 * the contextual information of the Ethernet device.
2574 * - (0) if successful.
2575 * - (-ENOTSUP) if support for dev_infos_get() does not exist for the device.
2576 * - (-ENODEV) if *port_id* invalid.
2578 int rte_eth_dev_info_get(uint16_t port_id, struct rte_eth_dev_info *dev_info);
2581 * Retrieve the firmware version of a device.
2584 * The port identifier of the device.
2586 * A pointer to a string array storing the firmware version of a device,
2587 * the string includes terminating null. This pointer is allocated by caller.
2589 * The size of the string array pointed by fw_version, which should be
2590 * large enough to store firmware version of the device.
2592 * - (0) if successful.
2593 * - (-ENOTSUP) if operation is not supported.
2594 * - (-ENODEV) if *port_id* invalid.
2595 * - (-EIO) if device is removed.
2596 * - (>0) if *fw_size* is not enough to store firmware version, return
2597 * the size of the non truncated string.
2599 int rte_eth_dev_fw_version_get(uint16_t port_id,
2600 char *fw_version, size_t fw_size);
2603 * Retrieve the supported packet types of an Ethernet device.
2605 * When a packet type is announced as supported, it *must* be recognized by
2606 * the PMD. For instance, if RTE_PTYPE_L2_ETHER, RTE_PTYPE_L2_ETHER_VLAN
2607 * and RTE_PTYPE_L3_IPV4 are announced, the PMD must return the following
2608 * packet types for these packets:
2609 * - Ether/IPv4 -> RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4
2610 * - Ether/Vlan/IPv4 -> RTE_PTYPE_L2_ETHER_VLAN | RTE_PTYPE_L3_IPV4
2611 * - Ether/[anything else] -> RTE_PTYPE_L2_ETHER
2612 * - Ether/Vlan/[anything else] -> RTE_PTYPE_L2_ETHER_VLAN
2614 * When a packet is received by a PMD, the most precise type must be
2615 * returned among the ones supported. However a PMD is allowed to set
2616 * packet type that is not in the supported list, at the condition that it
2617 * is more precise. Therefore, a PMD announcing no supported packet types
2618 * can still set a matching packet type in a received packet.
2621 * Better to invoke this API after the device is already started or rx burst
2622 * function is decided, to obtain correct supported ptypes.
2624 * if a given PMD does not report what ptypes it supports, then the supported
2625 * ptype count is reported as 0.
2627 * The port identifier of the Ethernet device.
2629 * A hint of what kind of packet type which the caller is interested in.
2631 * An array pointer to store adequate packet types, allocated by caller.
2633 * Size of the array pointed by param ptypes.
2635 * - (>=0) Number of supported ptypes. If the number of types exceeds num,
2636 * only num entries will be filled into the ptypes array, but the full
2637 * count of supported ptypes will be returned.
2638 * - (-ENODEV) if *port_id* invalid.
2640 int rte_eth_dev_get_supported_ptypes(uint16_t port_id, uint32_t ptype_mask,
2641 uint32_t *ptypes, int num);
2644 * @b EXPERIMENTAL: this API may change without prior notice.
2646 * Inform Ethernet device about reduced range of packet types to handle.
2648 * Application can use this function to set only specific ptypes that it's
2649 * interested. This information can be used by the PMD to optimize Rx path.
2651 * The function accepts an array `set_ptypes` allocated by the caller to
2652 * store the packet types set by the driver, the last element of the array
2653 * is set to RTE_PTYPE_UNKNOWN. The size of the `set_ptype` array should be
2654 * `rte_eth_dev_get_supported_ptypes() + 1` else it might only be filled
2658 * The port identifier of the Ethernet device.
2660 * The ptype family that application is interested in should be bitwise OR of
2661 * RTE_PTYPE_*_MASK or 0.
2663 * An array pointer to store set packet types, allocated by caller. The
2664 * function marks the end of array with RTE_PTYPE_UNKNOWN.
2666 * Size of the array pointed by param ptypes.
2667 * Should be rte_eth_dev_get_supported_ptypes() + 1 to accommodate the
2671 * - (-ENODEV) if *port_id* invalid.
2672 * - (-EINVAL) if *ptype_mask* is invalid (or) set_ptypes is NULL and
2676 int rte_eth_dev_set_ptypes(uint16_t port_id, uint32_t ptype_mask,
2677 uint32_t *set_ptypes, unsigned int num);
2680 * Retrieve the MTU of an Ethernet device.
2683 * The port identifier of the Ethernet device.
2685 * A pointer to a uint16_t where the retrieved MTU is to be stored.
2687 * - (0) if successful.
2688 * - (-ENODEV) if *port_id* invalid.
2690 int rte_eth_dev_get_mtu(uint16_t port_id, uint16_t *mtu);
2693 * Change the MTU of an Ethernet device.
2696 * The port identifier of the Ethernet device.
2698 * A uint16_t for the MTU to be applied.
2700 * - (0) if successful.
2701 * - (-ENOTSUP) if operation is not supported.
2702 * - (-ENODEV) if *port_id* invalid.
2703 * - (-EIO) if device is removed.
2704 * - (-EINVAL) if *mtu* invalid, validation of mtu can occur within
2705 * rte_eth_dev_set_mtu if dev_infos_get is supported by the device or
2706 * when the mtu is set using dev->dev_ops->mtu_set.
2707 * - (-EBUSY) if operation is not allowed when the port is running
2709 int rte_eth_dev_set_mtu(uint16_t port_id, uint16_t mtu);
2712 * Enable/Disable hardware filtering by an Ethernet device of received
2713 * VLAN packets tagged with a given VLAN Tag Identifier.
2716 * The port identifier of the Ethernet device.
2718 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
2720 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*.
2721 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*.
2723 * - (0) if successful.
2724 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
2725 * - (-ENODEV) if *port_id* invalid.
2726 * - (-EIO) if device is removed.
2727 * - (-ENOSYS) if VLAN filtering on *port_id* disabled.
2728 * - (-EINVAL) if *vlan_id* > 4095.
2730 int rte_eth_dev_vlan_filter(uint16_t port_id, uint16_t vlan_id, int on);
2733 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device.
2734 * 82599/X540/X550 can support VLAN stripping at the rx queue level
2737 * The port identifier of the Ethernet device.
2738 * @param rx_queue_id
2739 * The index of the receive queue for which a queue stats mapping is required.
2740 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2741 * to rte_eth_dev_configure().
2743 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port.
2744 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port.
2746 * - (0) if successful.
2747 * - (-ENOTSUP) if hardware-assisted VLAN stripping not configured.
2748 * - (-ENODEV) if *port_id* invalid.
2749 * - (-EINVAL) if *rx_queue_id* invalid.
2751 int rte_eth_dev_set_vlan_strip_on_queue(uint16_t port_id, uint16_t rx_queue_id,
2755 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to
2756 * the VLAN Header. This is a register setup available on some Intel NIC, not
2757 * but all, please check the data sheet for availability.
2760 * The port identifier of the Ethernet device.
2764 * The Tag Protocol ID
2766 * - (0) if successful.
2767 * - (-ENOTSUP) if hardware-assisted VLAN TPID setup is not supported.
2768 * - (-ENODEV) if *port_id* invalid.
2769 * - (-EIO) if device is removed.
2771 int rte_eth_dev_set_vlan_ether_type(uint16_t port_id,
2772 enum rte_vlan_type vlan_type,
2776 * Set VLAN offload configuration on an Ethernet device
2777 * Enable/Disable Extended VLAN by an Ethernet device, This is a register setup
2778 * available on some Intel NIC, not but all, please check the data sheet for
2780 * Enable/Disable VLAN Strip can be done on rx queue for certain NIC, but here
2781 * the configuration is applied on the port level.
2784 * The port identifier of the Ethernet device.
2785 * @param offload_mask
2786 * The VLAN Offload bit mask can be mixed use with "OR"
2787 * ETH_VLAN_STRIP_OFFLOAD
2788 * ETH_VLAN_FILTER_OFFLOAD
2789 * ETH_VLAN_EXTEND_OFFLOAD
2790 * ETH_QINQ_STRIP_OFFLOAD
2792 * - (0) if successful.
2793 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
2794 * - (-ENODEV) if *port_id* invalid.
2795 * - (-EIO) if device is removed.
2797 int rte_eth_dev_set_vlan_offload(uint16_t port_id, int offload_mask);
2800 * Read VLAN Offload configuration from an Ethernet device
2803 * The port identifier of the Ethernet device.
2805 * - (>0) if successful. Bit mask to indicate
2806 * ETH_VLAN_STRIP_OFFLOAD
2807 * ETH_VLAN_FILTER_OFFLOAD
2808 * ETH_VLAN_EXTEND_OFFLOAD
2809 * ETH_QINQ_STRIP_OFFLOAD
2810 * - (-ENODEV) if *port_id* invalid.
2812 int rte_eth_dev_get_vlan_offload(uint16_t port_id);
2815 * Set port based TX VLAN insertion on or off.
2818 * The port identifier of the Ethernet device.
2820 * Port based TX VLAN identifier together with user priority.
2822 * Turn on or off the port based TX VLAN insertion.
2825 * - (0) if successful.
2826 * - negative if failed.
2828 int rte_eth_dev_set_vlan_pvid(uint16_t port_id, uint16_t pvid, int on);
2830 typedef void (*buffer_tx_error_fn)(struct rte_mbuf **unsent, uint16_t count,
2834 * Structure used to buffer packets for future TX
2835 * Used by APIs rte_eth_tx_buffer and rte_eth_tx_buffer_flush
2837 struct rte_eth_dev_tx_buffer {
2838 buffer_tx_error_fn error_callback;
2839 void *error_userdata;
2840 uint16_t size; /**< Size of buffer for buffered tx */
2841 uint16_t length; /**< Number of packets in the array */
2842 struct rte_mbuf *pkts[];
2843 /**< Pending packets to be sent on explicit flush or when full */
2847 * Calculate the size of the tx buffer.
2850 * Number of stored packets.
2852 #define RTE_ETH_TX_BUFFER_SIZE(sz) \
2853 (sizeof(struct rte_eth_dev_tx_buffer) + (sz) * sizeof(struct rte_mbuf *))
2856 * Initialize default values for buffered transmitting
2859 * Tx buffer to be initialized.
2866 rte_eth_tx_buffer_init(struct rte_eth_dev_tx_buffer *buffer, uint16_t size);
2869 * Configure a callback for buffered packets which cannot be sent
2871 * Register a specific callback to be called when an attempt is made to send
2872 * all packets buffered on an ethernet port, but not all packets can
2873 * successfully be sent. The callback registered here will be called only
2874 * from calls to rte_eth_tx_buffer() and rte_eth_tx_buffer_flush() APIs.
2875 * The default callback configured for each queue by default just frees the
2876 * packets back to the calling mempool. If additional behaviour is required,
2877 * for example, to count dropped packets, or to retry transmission of packets
2878 * which cannot be sent, this function should be used to register a suitable
2879 * callback function to implement the desired behaviour.
2880 * The example callback "rte_eth_count_unsent_packet_callback()" is also
2881 * provided as reference.
2884 * The port identifier of the Ethernet device.
2886 * The function to be used as the callback.
2888 * Arbitrary parameter to be passed to the callback function
2890 * 0 on success, or -1 on error with rte_errno set appropriately
2893 rte_eth_tx_buffer_set_err_callback(struct rte_eth_dev_tx_buffer *buffer,
2894 buffer_tx_error_fn callback, void *userdata);
2897 * Callback function for silently dropping unsent buffered packets.
2899 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
2900 * adjust the default behavior when buffered packets cannot be sent. This
2901 * function drops any unsent packets silently and is used by tx buffered
2902 * operations as default behavior.
2904 * NOTE: this function should not be called directly, instead it should be used
2905 * as a callback for packet buffering.
2907 * NOTE: when configuring this function as a callback with
2908 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
2909 * should point to an uint64_t value.
2912 * The previously buffered packets which could not be sent
2914 * The number of unsent packets in the pkts array
2919 rte_eth_tx_buffer_drop_callback(struct rte_mbuf **pkts, uint16_t unsent,
2923 * Callback function for tracking unsent buffered packets.
2925 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
2926 * adjust the default behavior when buffered packets cannot be sent. This
2927 * function drops any unsent packets, but also updates a user-supplied counter
2928 * to track the overall number of packets dropped. The counter should be an
2929 * uint64_t variable.
2931 * NOTE: this function should not be called directly, instead it should be used
2932 * as a callback for packet buffering.
2934 * NOTE: when configuring this function as a callback with
2935 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
2936 * should point to an uint64_t value.
2939 * The previously buffered packets which could not be sent
2941 * The number of unsent packets in the pkts array
2943 * Pointer to an uint64_t value, which will be incremented by unsent
2946 rte_eth_tx_buffer_count_callback(struct rte_mbuf **pkts, uint16_t unsent,
2950 * Request the driver to free mbufs currently cached by the driver. The
2951 * driver will only free the mbuf if it is no longer in use. It is the
2952 * application's responsibility to ensure rte_eth_tx_buffer_flush(..) is
2956 * The port identifier of the Ethernet device.
2958 * The index of the transmit queue through which output packets must be
2960 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2961 * to rte_eth_dev_configure().
2963 * Maximum number of packets to free. Use 0 to indicate all possible packets
2964 * should be freed. Note that a packet may be using multiple mbufs.
2967 * -ENODEV: Invalid interface
2968 * -EIO: device is removed
2969 * -ENOTSUP: Driver does not support function
2971 * 0-n: Number of packets freed. More packets may still remain in ring that
2975 rte_eth_tx_done_cleanup(uint16_t port_id, uint16_t queue_id, uint32_t free_cnt);
2978 * Subtypes for IPsec offload event(@ref RTE_ETH_EVENT_IPSEC) raised by
2981 enum rte_eth_event_ipsec_subtype {
2982 RTE_ETH_EVENT_IPSEC_UNKNOWN = 0,
2983 /**< Unknown event type */
2984 RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW,
2985 /**< Sequence number overflow */
2986 RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY,
2987 /**< Soft time expiry of SA */
2988 RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY,
2989 /**< Soft byte expiry of SA */
2990 RTE_ETH_EVENT_IPSEC_MAX
2991 /**< Max value of this enum */
2995 * Descriptor for @ref RTE_ETH_EVENT_IPSEC event. Used by eth dev to send extra
2996 * information of the IPsec offload event.
2998 struct rte_eth_event_ipsec_desc {
2999 enum rte_eth_event_ipsec_subtype subtype;
3000 /**< Type of RTE_ETH_EVENT_IPSEC_* event */
3002 /**< Event specific metadata
3004 * For the following events, *userdata* registered
3005 * with the *rte_security_session* would be returned
3008 * - @ref RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW
3009 * - @ref RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY
3010 * - @ref RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY
3012 * @see struct rte_security_session_conf
3018 * The eth device event type for interrupt, and maybe others in the future.
3020 enum rte_eth_event_type {
3021 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */
3022 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */
3023 RTE_ETH_EVENT_QUEUE_STATE,
3024 /**< queue state event (enabled/disabled) */
3025 RTE_ETH_EVENT_INTR_RESET,
3026 /**< reset interrupt event, sent to VF on PF reset */
3027 RTE_ETH_EVENT_VF_MBOX, /**< message from the VF received by PF */
3028 RTE_ETH_EVENT_MACSEC, /**< MACsec offload related event */
3029 RTE_ETH_EVENT_INTR_RMV, /**< device removal event */
3030 RTE_ETH_EVENT_NEW, /**< port is probed */
3031 RTE_ETH_EVENT_DESTROY, /**< port is released */
3032 RTE_ETH_EVENT_IPSEC, /**< IPsec offload related event */
3033 RTE_ETH_EVENT_FLOW_AGED,/**< New aged-out flows is detected */
3034 RTE_ETH_EVENT_MAX /**< max value of this enum */
3037 typedef int (*rte_eth_dev_cb_fn)(uint16_t port_id,
3038 enum rte_eth_event_type event, void *cb_arg, void *ret_param);
3039 /**< user application callback to be registered for interrupts */
3042 * Register a callback function for port event.
3046 * RTE_ETH_ALL means register the event for all port ids.
3050 * User supplied callback function to be called.
3052 * Pointer to the parameters for the registered callback.
3055 * - On success, zero.
3056 * - On failure, a negative value.
3058 int rte_eth_dev_callback_register(uint16_t port_id,
3059 enum rte_eth_event_type event,
3060 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3063 * Unregister a callback function for port event.
3067 * RTE_ETH_ALL means unregister the event for all port ids.
3071 * User supplied callback function to be called.
3073 * Pointer to the parameters for the registered callback. -1 means to
3074 * remove all for the same callback address and same event.
3077 * - On success, zero.
3078 * - On failure, a negative value.
3080 int rte_eth_dev_callback_unregister(uint16_t port_id,
3081 enum rte_eth_event_type event,
3082 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3085 * When there is no rx packet coming in Rx Queue for a long time, we can
3086 * sleep lcore related to RX Queue for power saving, and enable rx interrupt
3087 * to be triggered when Rx packet arrives.
3089 * The rte_eth_dev_rx_intr_enable() function enables rx queue
3090 * interrupt on specific rx queue of a port.
3093 * The port identifier of the Ethernet device.
3095 * The index of the receive queue from which to retrieve input packets.
3096 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3097 * to rte_eth_dev_configure().
3099 * - (0) if successful.
3100 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3102 * - (-ENODEV) if *port_id* invalid.
3103 * - (-EIO) if device is removed.
3105 int rte_eth_dev_rx_intr_enable(uint16_t port_id, uint16_t queue_id);
3108 * When lcore wakes up from rx interrupt indicating packet coming, disable rx
3109 * interrupt and returns to polling mode.
3111 * The rte_eth_dev_rx_intr_disable() function disables rx queue
3112 * interrupt on specific rx queue of a port.
3115 * The port identifier of the Ethernet device.
3117 * The index of the receive queue from which to retrieve input packets.
3118 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3119 * to rte_eth_dev_configure().
3121 * - (0) if successful.
3122 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3124 * - (-ENODEV) if *port_id* invalid.
3125 * - (-EIO) if device is removed.
3127 int rte_eth_dev_rx_intr_disable(uint16_t port_id, uint16_t queue_id);
3130 * RX Interrupt control per port.
3133 * The port identifier of the Ethernet device.
3135 * Epoll instance fd which the intr vector associated to.
3136 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3138 * The operation be performed for the vector.
3139 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3143 * - On success, zero.
3144 * - On failure, a negative value.
3146 int rte_eth_dev_rx_intr_ctl(uint16_t port_id, int epfd, int op, void *data);
3149 * RX Interrupt control per queue.
3152 * The port identifier of the Ethernet device.
3154 * The index of the receive queue from which to retrieve input packets.
3155 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3156 * to rte_eth_dev_configure().
3158 * Epoll instance fd which the intr vector associated to.
3159 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3161 * The operation be performed for the vector.
3162 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3166 * - On success, zero.
3167 * - On failure, a negative value.
3169 int rte_eth_dev_rx_intr_ctl_q(uint16_t port_id, uint16_t queue_id,
3170 int epfd, int op, void *data);
3174 * @b EXPERIMENTAL: this API may change without prior notice.
3176 * Get interrupt fd per Rx queue.
3179 * The port identifier of the Ethernet device.
3181 * The index of the receive queue from which to retrieve input packets.
3182 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3183 * to rte_eth_dev_configure().
3185 * - (>=0) the interrupt fd associated to the requested Rx queue if
3191 rte_eth_dev_rx_intr_ctl_q_get_fd(uint16_t port_id, uint16_t queue_id);
3194 * Turn on the LED on the Ethernet device.
3195 * This function turns on the LED on the Ethernet device.
3198 * The port identifier of the Ethernet device.
3200 * - (0) if successful.
3201 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3203 * - (-ENODEV) if *port_id* invalid.
3204 * - (-EIO) if device is removed.
3206 int rte_eth_led_on(uint16_t port_id);
3209 * Turn off the LED on the Ethernet device.
3210 * This function turns off the LED on the Ethernet device.
3213 * The port identifier of the Ethernet device.
3215 * - (0) if successful.
3216 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3218 * - (-ENODEV) if *port_id* invalid.
3219 * - (-EIO) if device is removed.
3221 int rte_eth_led_off(uint16_t port_id);
3224 * Get current status of the Ethernet link flow control for Ethernet device
3227 * The port identifier of the Ethernet device.
3229 * The pointer to the structure where to store the flow control parameters.
3231 * - (0) if successful.
3232 * - (-ENOTSUP) if hardware doesn't support flow control.
3233 * - (-ENODEV) if *port_id* invalid.
3234 * - (-EIO) if device is removed.
3236 int rte_eth_dev_flow_ctrl_get(uint16_t port_id,
3237 struct rte_eth_fc_conf *fc_conf);
3240 * Configure the Ethernet link flow control for Ethernet device
3243 * The port identifier of the Ethernet device.
3245 * The pointer to the structure of the flow control parameters.
3247 * - (0) if successful.
3248 * - (-ENOTSUP) if hardware doesn't support flow control mode.
3249 * - (-ENODEV) if *port_id* invalid.
3250 * - (-EINVAL) if bad parameter
3251 * - (-EIO) if flow control setup failure or device is removed.
3253 int rte_eth_dev_flow_ctrl_set(uint16_t port_id,
3254 struct rte_eth_fc_conf *fc_conf);
3257 * Configure the Ethernet priority flow control under DCB environment
3258 * for Ethernet device.
3261 * The port identifier of the Ethernet device.
3263 * The pointer to the structure of the priority flow control parameters.
3265 * - (0) if successful.
3266 * - (-ENOTSUP) if hardware doesn't support priority flow control mode.
3267 * - (-ENODEV) if *port_id* invalid.
3268 * - (-EINVAL) if bad parameter
3269 * - (-EIO) if flow control setup failure or device is removed.
3271 int rte_eth_dev_priority_flow_ctrl_set(uint16_t port_id,
3272 struct rte_eth_pfc_conf *pfc_conf);
3275 * Add a MAC address to an internal array of addresses used to enable whitelist
3276 * filtering to accept packets only if the destination MAC address matches.
3279 * The port identifier of the Ethernet device.
3281 * The MAC address to add.
3283 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is
3284 * not enabled, this should be set to 0.
3286 * - (0) if successfully added or *mac_addr* was already added.
3287 * - (-ENOTSUP) if hardware doesn't support this feature.
3288 * - (-ENODEV) if *port* is invalid.
3289 * - (-EIO) if device is removed.
3290 * - (-ENOSPC) if no more MAC addresses can be added.
3291 * - (-EINVAL) if MAC address is invalid.
3293 int rte_eth_dev_mac_addr_add(uint16_t port_id, struct rte_ether_addr *mac_addr,
3297 * Remove a MAC address from the internal array of addresses.
3300 * The port identifier of the Ethernet device.
3302 * MAC address to remove.
3304 * - (0) if successful, or *mac_addr* didn't exist.
3305 * - (-ENOTSUP) if hardware doesn't support.
3306 * - (-ENODEV) if *port* invalid.
3307 * - (-EADDRINUSE) if attempting to remove the default MAC address
3309 int rte_eth_dev_mac_addr_remove(uint16_t port_id,
3310 struct rte_ether_addr *mac_addr);
3313 * Set the default MAC address.
3316 * The port identifier of the Ethernet device.
3318 * New default MAC address.
3320 * - (0) if successful, or *mac_addr* didn't exist.
3321 * - (-ENOTSUP) if hardware doesn't support.
3322 * - (-ENODEV) if *port* invalid.
3323 * - (-EINVAL) if MAC address is invalid.
3325 int rte_eth_dev_default_mac_addr_set(uint16_t port_id,
3326 struct rte_ether_addr *mac_addr);
3329 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3332 * The port identifier of the Ethernet device.
3336 * Redirection table size. The table size can be queried by
3337 * rte_eth_dev_info_get().
3339 * - (0) if successful.
3340 * - (-ENOTSUP) if hardware doesn't support.
3341 * - (-EINVAL) if bad parameter.
3342 * - (-EIO) if device is removed.
3344 int rte_eth_dev_rss_reta_update(uint16_t port_id,
3345 struct rte_eth_rss_reta_entry64 *reta_conf,
3346 uint16_t reta_size);
3349 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3352 * The port identifier of the Ethernet device.
3354 * RETA to query. For each requested reta entry, corresponding bit
3355 * in mask must be set.
3357 * Redirection table size. The table size can be queried by
3358 * rte_eth_dev_info_get().
3360 * - (0) if successful.
3361 * - (-ENOTSUP) if hardware doesn't support.
3362 * - (-EINVAL) if bad parameter.
3363 * - (-EIO) if device is removed.
3365 int rte_eth_dev_rss_reta_query(uint16_t port_id,
3366 struct rte_eth_rss_reta_entry64 *reta_conf,
3367 uint16_t reta_size);
3370 * Updates unicast hash table for receiving packet with the given destination
3371 * MAC address, and the packet is routed to all VFs for which the RX mode is
3372 * accept packets that match the unicast hash table.
3375 * The port identifier of the Ethernet device.
3377 * Unicast MAC address.
3379 * 1 - Set an unicast hash bit for receiving packets with the MAC address.
3380 * 0 - Clear an unicast hash bit.
3382 * - (0) if successful.
3383 * - (-ENOTSUP) if hardware doesn't support.
3384 * - (-ENODEV) if *port_id* invalid.
3385 * - (-EIO) if device is removed.
3386 * - (-EINVAL) if bad parameter.
3388 int rte_eth_dev_uc_hash_table_set(uint16_t port_id, struct rte_ether_addr *addr,
3392 * Updates all unicast hash bitmaps for receiving packet with any Unicast
3393 * Ethernet MAC addresses,the packet is routed to all VFs for which the RX
3394 * mode is accept packets that match the unicast hash table.
3397 * The port identifier of the Ethernet device.
3399 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet
3401 * 0 - Clear all unicast hash bitmaps
3403 * - (0) if successful.
3404 * - (-ENOTSUP) if hardware doesn't support.
3405 * - (-ENODEV) if *port_id* invalid.
3406 * - (-EIO) if device is removed.
3407 * - (-EINVAL) if bad parameter.
3409 int rte_eth_dev_uc_all_hash_table_set(uint16_t port_id, uint8_t on);
3412 * Set a traffic mirroring rule on an Ethernet device
3415 * The port identifier of the Ethernet device.
3416 * @param mirror_conf
3417 * The pointer to the traffic mirroring structure describing the mirroring rule.
3418 * The *rte_eth_vm_mirror_conf* structure includes the type of mirroring rule,
3419 * destination pool and the value of rule if enable vlan or pool mirroring.
3422 * The index of traffic mirroring rule, we support four separated rules.
3424 * 1 - Enable a mirroring rule.
3425 * 0 - Disable a mirroring rule.
3427 * - (0) if successful.
3428 * - (-ENOTSUP) if hardware doesn't support this feature.
3429 * - (-ENODEV) if *port_id* invalid.
3430 * - (-EIO) if device is removed.
3431 * - (-EINVAL) if the mr_conf information is not correct.
3433 int rte_eth_mirror_rule_set(uint16_t port_id,
3434 struct rte_eth_mirror_conf *mirror_conf,
3439 * Reset a traffic mirroring rule on an Ethernet device.
3442 * The port identifier of the Ethernet device.
3444 * The index of traffic mirroring rule, we support four separated rules.
3446 * - (0) if successful.
3447 * - (-ENOTSUP) if hardware doesn't support this feature.
3448 * - (-ENODEV) if *port_id* invalid.
3449 * - (-EIO) if device is removed.
3450 * - (-EINVAL) if bad parameter.
3452 int rte_eth_mirror_rule_reset(uint16_t port_id,
3456 * Set the rate limitation for a queue on an Ethernet device.
3459 * The port identifier of the Ethernet device.
3463 * The tx rate in Mbps. Allocated from the total port link speed.
3465 * - (0) if successful.
3466 * - (-ENOTSUP) if hardware doesn't support this feature.
3467 * - (-ENODEV) if *port_id* invalid.
3468 * - (-EIO) if device is removed.
3469 * - (-EINVAL) if bad parameter.
3471 int rte_eth_set_queue_rate_limit(uint16_t port_id, uint16_t queue_idx,
3475 * Configuration of Receive Side Scaling hash computation of Ethernet device.
3478 * The port identifier of the Ethernet device.
3480 * The new configuration to use for RSS hash computation on the port.
3482 * - (0) if successful.
3483 * - (-ENODEV) if port identifier is invalid.
3484 * - (-EIO) if device is removed.
3485 * - (-ENOTSUP) if hardware doesn't support.
3486 * - (-EINVAL) if bad parameter.
3488 int rte_eth_dev_rss_hash_update(uint16_t port_id,
3489 struct rte_eth_rss_conf *rss_conf);
3492 * Retrieve current configuration of Receive Side Scaling hash computation
3493 * of Ethernet device.
3496 * The port identifier of the Ethernet device.
3498 * Where to store the current RSS hash configuration of the Ethernet device.
3500 * - (0) if successful.
3501 * - (-ENODEV) if port identifier is invalid.
3502 * - (-EIO) if device is removed.
3503 * - (-ENOTSUP) if hardware doesn't support RSS.
3506 rte_eth_dev_rss_hash_conf_get(uint16_t port_id,
3507 struct rte_eth_rss_conf *rss_conf);
3510 * Add UDP tunneling port for a specific type of tunnel.
3511 * The packets with this UDP port will be identified as this type of tunnel.
3512 * Before enabling any offloading function for a tunnel, users can call this API
3513 * to change or add more UDP port for the tunnel. So the offloading function
3514 * can take effect on the packets with the specific UDP port.
3517 * The port identifier of the Ethernet device.
3519 * UDP tunneling configuration.
3522 * - (0) if successful.
3523 * - (-ENODEV) if port identifier is invalid.
3524 * - (-EIO) if device is removed.
3525 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3528 rte_eth_dev_udp_tunnel_port_add(uint16_t port_id,
3529 struct rte_eth_udp_tunnel *tunnel_udp);
3532 * Delete UDP tunneling port a specific type of tunnel.
3533 * The packets with this UDP port will not be identified as this type of tunnel
3535 * Before enabling any offloading function for a tunnel, users can call this API
3536 * to delete a UDP port for the tunnel. So the offloading function will not take
3537 * effect on the packets with the specific UDP port.
3540 * The port identifier of the Ethernet device.
3542 * UDP tunneling configuration.
3545 * - (0) if successful.
3546 * - (-ENODEV) if port identifier is invalid.
3547 * - (-EIO) if device is removed.
3548 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3551 rte_eth_dev_udp_tunnel_port_delete(uint16_t port_id,
3552 struct rte_eth_udp_tunnel *tunnel_udp);
3555 * Check whether the filter type is supported on an Ethernet device.
3556 * All the supported filter types are defined in 'rte_eth_ctrl.h'.
3559 * The port identifier of the Ethernet device.
3560 * @param filter_type
3563 * - (0) if successful.
3564 * - (-ENOTSUP) if hardware doesn't support this filter type.
3565 * - (-ENODEV) if *port_id* invalid.
3566 * - (-EIO) if device is removed.
3569 int rte_eth_dev_filter_supported(uint16_t port_id,
3570 enum rte_filter_type filter_type);
3573 * Take operations to assigned filter type on an Ethernet device.
3574 * All the supported operations and filter types are defined in 'rte_eth_ctrl.h'.
3577 * The port identifier of the Ethernet device.
3578 * @param filter_type
3581 * Type of operation.
3583 * A pointer to arguments defined specifically for the operation.
3585 * - (0) if successful.
3586 * - (-ENOTSUP) if hardware doesn't support.
3587 * - (-ENODEV) if *port_id* invalid.
3588 * - (-EIO) if device is removed.
3589 * - others depends on the specific operations implementation.
3592 int rte_eth_dev_filter_ctrl(uint16_t port_id, enum rte_filter_type filter_type,
3593 enum rte_filter_op filter_op, void *arg);
3596 * Get DCB information on an Ethernet device.
3599 * The port identifier of the Ethernet device.
3603 * - (0) if successful.
3604 * - (-ENODEV) if port identifier is invalid.
3605 * - (-EIO) if device is removed.
3606 * - (-ENOTSUP) if hardware doesn't support.
3608 int rte_eth_dev_get_dcb_info(uint16_t port_id,
3609 struct rte_eth_dcb_info *dcb_info);
3611 struct rte_eth_rxtx_callback;
3614 * Add a callback to be called on packet RX on a given port and queue.
3616 * This API configures a function to be called for each burst of
3617 * packets received on a given NIC port queue. The return value is a pointer
3618 * that can be used to later remove the callback using
3619 * rte_eth_remove_rx_callback().
3621 * Multiple functions are called in the order that they are added.
3624 * The port identifier of the Ethernet device.
3626 * The queue on the Ethernet device on which the callback is to be added.
3628 * The callback function
3630 * A generic pointer parameter which will be passed to each invocation of the
3631 * callback function on this port and queue.
3635 * On success, a pointer value which can later be used to remove the callback.
3637 const struct rte_eth_rxtx_callback *
3638 rte_eth_add_rx_callback(uint16_t port_id, uint16_t queue_id,
3639 rte_rx_callback_fn fn, void *user_param);
3642 * Add a callback that must be called first on packet RX on a given port
3645 * This API configures a first function to be called for each burst of
3646 * packets received on a given NIC port queue. The return value is a pointer
3647 * that can be used to later remove the callback using
3648 * rte_eth_remove_rx_callback().
3650 * Multiple functions are called in the order that they are added.
3653 * The port identifier of the Ethernet device.
3655 * The queue on the Ethernet device on which the callback is to be added.
3657 * The callback function
3659 * A generic pointer parameter which will be passed to each invocation of the
3660 * callback function on this port and queue.
3664 * On success, a pointer value which can later be used to remove the callback.
3666 const struct rte_eth_rxtx_callback *
3667 rte_eth_add_first_rx_callback(uint16_t port_id, uint16_t queue_id,
3668 rte_rx_callback_fn fn, void *user_param);
3671 * Add a callback to be called on packet TX on a given port and queue.
3673 * This API configures a function to be called for each burst of
3674 * packets sent on a given NIC port queue. The return value is a pointer
3675 * that can be used to later remove the callback using
3676 * rte_eth_remove_tx_callback().
3678 * Multiple functions are called in the order that they are added.
3681 * The port identifier of the Ethernet device.
3683 * The queue on the Ethernet device on which the callback is to be added.
3685 * The callback function
3687 * A generic pointer parameter which will be passed to each invocation of the
3688 * callback function on this port and queue.
3692 * On success, a pointer value which can later be used to remove the callback.
3694 const struct rte_eth_rxtx_callback *
3695 rte_eth_add_tx_callback(uint16_t port_id, uint16_t queue_id,
3696 rte_tx_callback_fn fn, void *user_param);
3699 * Remove an RX packet callback from a given port and queue.
3701 * This function is used to removed callbacks that were added to a NIC port
3702 * queue using rte_eth_add_rx_callback().
3704 * Note: the callback is removed from the callback list but it isn't freed
3705 * since the it may still be in use. The memory for the callback can be
3706 * subsequently freed back by the application by calling rte_free():
3708 * - Immediately - if the port is stopped, or the user knows that no
3709 * callbacks are in flight e.g. if called from the thread doing RX/TX
3712 * - After a short delay - where the delay is sufficient to allow any
3713 * in-flight callbacks to complete.
3716 * The port identifier of the Ethernet device.
3718 * The queue on the Ethernet device from which the callback is to be removed.
3720 * User supplied callback created via rte_eth_add_rx_callback().
3723 * - 0: Success. Callback was removed.
3724 * - -ENOTSUP: Callback support is not available.
3725 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
3726 * is NULL or not found for the port/queue.
3728 int rte_eth_remove_rx_callback(uint16_t port_id, uint16_t queue_id,
3729 const struct rte_eth_rxtx_callback *user_cb);
3732 * Remove a TX packet callback from a given port and queue.
3734 * This function is used to removed callbacks that were added to a NIC port
3735 * queue using rte_eth_add_tx_callback().
3737 * Note: the callback is removed from the callback list but it isn't freed
3738 * since the it may still be in use. The memory for the callback can be
3739 * subsequently freed back by the application by calling rte_free():
3741 * - Immediately - if the port is stopped, or the user knows that no
3742 * callbacks are in flight e.g. if called from the thread doing RX/TX
3745 * - After a short delay - where the delay is sufficient to allow any
3746 * in-flight callbacks to complete.
3749 * The port identifier of the Ethernet device.
3751 * The queue on the Ethernet device from which the callback is to be removed.
3753 * User supplied callback created via rte_eth_add_tx_callback().
3756 * - 0: Success. Callback was removed.
3757 * - -ENOTSUP: Callback support is not available.
3758 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
3759 * is NULL or not found for the port/queue.
3761 int rte_eth_remove_tx_callback(uint16_t port_id, uint16_t queue_id,
3762 const struct rte_eth_rxtx_callback *user_cb);
3765 * Retrieve information about given port's RX queue.
3768 * The port identifier of the Ethernet device.
3770 * The RX queue on the Ethernet device for which information
3771 * will be retrieved.
3773 * A pointer to a structure of type *rte_eth_rxq_info_info* to be filled with
3774 * the information of the Ethernet device.
3778 * - -ENOTSUP: routine is not supported by the device PMD.
3779 * - -EINVAL: The port_id or the queue_id is out of range, or the queue
3782 int rte_eth_rx_queue_info_get(uint16_t port_id, uint16_t queue_id,
3783 struct rte_eth_rxq_info *qinfo);
3786 * Retrieve information about given port's TX queue.
3789 * The port identifier of the Ethernet device.
3791 * The TX queue on the Ethernet device for which information
3792 * will be retrieved.
3794 * A pointer to a structure of type *rte_eth_txq_info_info* to be filled with
3795 * the information of the Ethernet device.
3799 * - -ENOTSUP: routine is not supported by the device PMD.
3800 * - -EINVAL: The port_id or the queue_id is out of range, or the queue
3803 int rte_eth_tx_queue_info_get(uint16_t port_id, uint16_t queue_id,
3804 struct rte_eth_txq_info *qinfo);
3807 * Retrieve information about the Rx packet burst mode.
3810 * The port identifier of the Ethernet device.
3812 * The Rx queue on the Ethernet device for which information
3813 * will be retrieved.
3815 * A pointer to a structure of type *rte_eth_burst_mode* to be filled
3816 * with the information of the packet burst mode.
3820 * - -ENOTSUP: routine is not supported by the device PMD.
3821 * - -EINVAL: The port_id or the queue_id is out of range.
3824 int rte_eth_rx_burst_mode_get(uint16_t port_id, uint16_t queue_id,
3825 struct rte_eth_burst_mode *mode);
3828 * Retrieve information about the Tx packet burst mode.
3831 * The port identifier of the Ethernet device.
3833 * The Tx queue on the Ethernet device for which information
3834 * will be retrieved.
3836 * A pointer to a structure of type *rte_eth_burst_mode* to be filled
3837 * with the information of the packet burst mode.
3841 * - -ENOTSUP: routine is not supported by the device PMD.
3842 * - -EINVAL: The port_id or the queue_id is out of range.
3845 int rte_eth_tx_burst_mode_get(uint16_t port_id, uint16_t queue_id,
3846 struct rte_eth_burst_mode *mode);
3849 * Retrieve device registers and register attributes (number of registers and
3853 * The port identifier of the Ethernet device.
3855 * Pointer to rte_dev_reg_info structure to fill in. If info->data is
3856 * NULL the function fills in the width and length fields. If non-NULL
3857 * the registers are put into the buffer pointed at by the data field.
3859 * - (0) if successful.
3860 * - (-ENOTSUP) if hardware doesn't support.
3861 * - (-ENODEV) if *port_id* invalid.
3862 * - (-EIO) if device is removed.
3863 * - others depends on the specific operations implementation.
3865 int rte_eth_dev_get_reg_info(uint16_t port_id, struct rte_dev_reg_info *info);
3868 * Retrieve size of device EEPROM
3871 * The port identifier of the Ethernet device.
3873 * - (>=0) EEPROM size if successful.
3874 * - (-ENOTSUP) if hardware doesn't support.
3875 * - (-ENODEV) if *port_id* invalid.
3876 * - (-EIO) if device is removed.
3877 * - others depends on the specific operations implementation.
3879 int rte_eth_dev_get_eeprom_length(uint16_t port_id);
3882 * Retrieve EEPROM and EEPROM attribute
3885 * The port identifier of the Ethernet device.
3887 * The template includes buffer for return EEPROM data and
3888 * EEPROM attributes to be filled.
3890 * - (0) if successful.
3891 * - (-ENOTSUP) if hardware doesn't support.
3892 * - (-ENODEV) if *port_id* invalid.
3893 * - (-EIO) if device is removed.
3894 * - others depends on the specific operations implementation.
3896 int rte_eth_dev_get_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
3899 * Program EEPROM with provided data
3902 * The port identifier of the Ethernet device.
3904 * The template includes EEPROM data for programming and
3905 * EEPROM attributes to be filled
3907 * - (0) if successful.
3908 * - (-ENOTSUP) if hardware doesn't support.
3909 * - (-ENODEV) if *port_id* invalid.
3910 * - (-EIO) if device is removed.
3911 * - others depends on the specific operations implementation.
3913 int rte_eth_dev_set_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
3917 * @b EXPERIMENTAL: this API may change without prior notice.
3919 * Retrieve the type and size of plugin module EEPROM
3922 * The port identifier of the Ethernet device.
3924 * The type and size of plugin module EEPROM.
3926 * - (0) if successful.
3927 * - (-ENOTSUP) if hardware doesn't support.
3928 * - (-ENODEV) if *port_id* invalid.
3929 * - (-EIO) if device is removed.
3930 * - others depends on the specific operations implementation.
3934 rte_eth_dev_get_module_info(uint16_t port_id,
3935 struct rte_eth_dev_module_info *modinfo);
3939 * @b EXPERIMENTAL: this API may change without prior notice.
3941 * Retrieve the data of plugin module EEPROM
3944 * The port identifier of the Ethernet device.
3946 * The template includes the plugin module EEPROM attributes, and the
3947 * buffer for return plugin module EEPROM data.
3949 * - (0) if successful.
3950 * - (-ENOTSUP) if hardware doesn't support.
3951 * - (-ENODEV) if *port_id* invalid.
3952 * - (-EIO) if device is removed.
3953 * - others depends on the specific operations implementation.
3957 rte_eth_dev_get_module_eeprom(uint16_t port_id,
3958 struct rte_dev_eeprom_info *info);
3961 * Set the list of multicast addresses to filter on an Ethernet device.
3964 * The port identifier of the Ethernet device.
3965 * @param mc_addr_set
3966 * The array of multicast addresses to set. Equal to NULL when the function
3967 * is invoked to flush the set of filtered addresses.
3969 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0
3970 * when the function is invoked to flush the set of filtered addresses.
3972 * - (0) if successful.
3973 * - (-ENODEV) if *port_id* invalid.
3974 * - (-EIO) if device is removed.
3975 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering.
3976 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources.
3978 int rte_eth_dev_set_mc_addr_list(uint16_t port_id,
3979 struct rte_ether_addr *mc_addr_set,
3980 uint32_t nb_mc_addr);
3983 * Enable IEEE1588/802.1AS timestamping for an Ethernet device.
3986 * The port identifier of the Ethernet device.
3990 * - -ENODEV: The port ID is invalid.
3991 * - -EIO: if device is removed.
3992 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3994 int rte_eth_timesync_enable(uint16_t port_id);
3997 * Disable IEEE1588/802.1AS timestamping for an Ethernet device.
4000 * The port identifier of the Ethernet device.
4004 * - -ENODEV: The port ID is invalid.
4005 * - -EIO: if device is removed.
4006 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4008 int rte_eth_timesync_disable(uint16_t port_id);
4011 * Read an IEEE1588/802.1AS RX timestamp from an Ethernet device.
4014 * The port identifier of the Ethernet device.
4016 * Pointer to the timestamp struct.
4018 * Device specific flags. Used to pass the RX timesync register index to
4019 * i40e. Unused in igb/ixgbe, pass 0 instead.
4023 * - -EINVAL: No timestamp is available.
4024 * - -ENODEV: The port ID is invalid.
4025 * - -EIO: if device is removed.
4026 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4028 int rte_eth_timesync_read_rx_timestamp(uint16_t port_id,
4029 struct timespec *timestamp, uint32_t flags);
4032 * Read an IEEE1588/802.1AS TX timestamp from an Ethernet device.
4035 * The port identifier of the Ethernet device.
4037 * Pointer to the timestamp struct.
4041 * - -EINVAL: No timestamp is available.
4042 * - -ENODEV: The port ID is invalid.
4043 * - -EIO: if device is removed.
4044 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4046 int rte_eth_timesync_read_tx_timestamp(uint16_t port_id,
4047 struct timespec *timestamp);
4050 * Adjust the timesync clock on an Ethernet device.
4052 * This is usually used in conjunction with other Ethdev timesync functions to
4053 * synchronize the device time using the IEEE1588/802.1AS protocol.
4056 * The port identifier of the Ethernet device.
4058 * The adjustment in nanoseconds.
4062 * - -ENODEV: The port ID is invalid.
4063 * - -EIO: if device is removed.
4064 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4066 int rte_eth_timesync_adjust_time(uint16_t port_id, int64_t delta);
4069 * Read the time from the timesync clock on an Ethernet device.
4071 * This is usually used in conjunction with other Ethdev timesync functions to
4072 * synchronize the device time using the IEEE1588/802.1AS protocol.
4075 * The port identifier of the Ethernet device.
4077 * Pointer to the timespec struct that holds the time.
4082 int rte_eth_timesync_read_time(uint16_t port_id, struct timespec *time);
4085 * Set the time of the timesync clock on an Ethernet device.
4087 * This is usually used in conjunction with other Ethdev timesync functions to
4088 * synchronize the device time using the IEEE1588/802.1AS protocol.
4091 * The port identifier of the Ethernet device.
4093 * Pointer to the timespec struct that holds the time.
4097 * - -EINVAL: No timestamp is available.
4098 * - -ENODEV: The port ID is invalid.
4099 * - -EIO: if device is removed.
4100 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4102 int rte_eth_timesync_write_time(uint16_t port_id, const struct timespec *time);
4106 * @b EXPERIMENTAL: this API may change without prior notice.
4108 * Read the current clock counter of an Ethernet device
4110 * This returns the current raw clock value of an Ethernet device. It is
4111 * a raw amount of ticks, with no given time reference.
4112 * The value returned here is from the same clock than the one
4113 * filling timestamp field of Rx packets when using hardware timestamp
4114 * offload. Therefore it can be used to compute a precise conversion of
4115 * the device clock to the real time.
4117 * E.g, a simple heuristic to derivate the frequency would be:
4118 * uint64_t start, end;
4119 * rte_eth_read_clock(port, start);
4120 * rte_delay_ms(100);
4121 * rte_eth_read_clock(port, end);
4122 * double freq = (end - start) * 10;
4124 * Compute a common reference with:
4125 * uint64_t base_time_sec = current_time();
4126 * uint64_t base_clock;
4127 * rte_eth_read_clock(port, base_clock);
4129 * Then, convert the raw mbuf timestamp with:
4130 * base_time_sec + (double)(mbuf->timestamp - base_clock) / freq;
4132 * This simple example will not provide a very good accuracy. One must
4133 * at least measure multiple times the frequency and do a regression.
4134 * To avoid deviation from the system time, the common reference can
4135 * be repeated from time to time. The integer division can also be
4136 * converted by a multiplication and a shift for better performance.
4139 * The port identifier of the Ethernet device.
4141 * Pointer to the uint64_t that holds the raw clock value.
4145 * - -ENODEV: The port ID is invalid.
4146 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4150 rte_eth_read_clock(uint16_t port_id, uint64_t *clock);
4153 * Config l2 tunnel ether type of an Ethernet device for filtering specific
4154 * tunnel packets by ether type.
4157 * The port identifier of the Ethernet device.
4159 * l2 tunnel configuration.
4162 * - (0) if successful.
4163 * - (-ENODEV) if port identifier is invalid.
4164 * - (-EIO) if device is removed.
4165 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4168 rte_eth_dev_l2_tunnel_eth_type_conf(uint16_t port_id,
4169 struct rte_eth_l2_tunnel_conf *l2_tunnel);
4172 * Enable/disable l2 tunnel offload functions. Include,
4173 * 1, The ability of parsing a type of l2 tunnel of an Ethernet device.
4174 * Filtering, forwarding and offloading this type of tunnel packets depend on
4176 * 2, Stripping the l2 tunnel tag.
4177 * 3, Insertion of the l2 tunnel tag.
4178 * 4, Forwarding the packets based on the l2 tunnel tag.
4181 * The port identifier of the Ethernet device.
4183 * l2 tunnel parameters.
4185 * Indicate the offload function.
4187 * Enable or disable this function.
4190 * - (0) if successful.
4191 * - (-ENODEV) if port identifier is invalid.
4192 * - (-EIO) if device is removed.
4193 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4196 rte_eth_dev_l2_tunnel_offload_set(uint16_t port_id,
4197 struct rte_eth_l2_tunnel_conf *l2_tunnel,
4202 * Get the port id from device name. The device name should be specified
4204 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:2:00.0
4205 * - SoC device name, for example- fsl-gmac0
4206 * - vdev dpdk name, for example- net_[pcap0|null0|tap0]
4209 * pci address or name of the device
4211 * pointer to port identifier of the device
4213 * - (0) if successful and port_id is filled.
4214 * - (-ENODEV or -EINVAL) on failure.
4217 rte_eth_dev_get_port_by_name(const char *name, uint16_t *port_id);
4220 * Get the device name from port id. The device name is specified as below:
4221 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:02:00.0
4222 * - SoC device name, for example- fsl-gmac0
4223 * - vdev dpdk name, for example- net_[pcap0|null0|tun0|tap0]
4226 * Port identifier of the device.
4228 * Buffer of size RTE_ETH_NAME_MAX_LEN to store the name.
4230 * - (0) if successful.
4231 * - (-EINVAL) on failure.
4234 rte_eth_dev_get_name_by_port(uint16_t port_id, char *name);
4237 * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from
4238 * the ethernet device information, otherwise adjust them to boundaries.
4241 * The port identifier of the Ethernet device.
4243 * A pointer to a uint16_t where the number of receive
4244 * descriptors stored.
4246 * A pointer to a uint16_t where the number of transmit
4247 * descriptors stored.
4249 * - (0) if successful.
4250 * - (-ENOTSUP, -ENODEV or -EINVAL) on failure.
4252 int rte_eth_dev_adjust_nb_rx_tx_desc(uint16_t port_id,
4253 uint16_t *nb_rx_desc,
4254 uint16_t *nb_tx_desc);
4257 * Test if a port supports specific mempool ops.
4260 * Port identifier of the Ethernet device.
4262 * The name of the pool operations to test.
4264 * - 0: best mempool ops choice for this port.
4265 * - 1: mempool ops are supported for this port.
4266 * - -ENOTSUP: mempool ops not supported for this port.
4267 * - -ENODEV: Invalid port Identifier.
4268 * - -EINVAL: Pool param is null.
4271 rte_eth_dev_pool_ops_supported(uint16_t port_id, const char *pool);
4274 * Get the security context for the Ethernet device.
4277 * Port identifier of the Ethernet device
4280 * - pointer to security context on success.
4283 rte_eth_dev_get_sec_ctx(uint16_t port_id);
4287 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice
4289 * Query the device hairpin capabilities.
4292 * The port identifier of the Ethernet device.
4294 * Pointer to a structure that will hold the hairpin capabilities.
4296 * - (0) if successful.
4297 * - (-ENOTSUP) if hardware doesn't support.
4300 int rte_eth_dev_hairpin_capability_get(uint16_t port_id,
4301 struct rte_eth_hairpin_cap *cap);
4303 #include <rte_ethdev_core.h>
4307 * Retrieve a burst of input packets from a receive queue of an Ethernet
4308 * device. The retrieved packets are stored in *rte_mbuf* structures whose
4309 * pointers are supplied in the *rx_pkts* array.
4311 * The rte_eth_rx_burst() function loops, parsing the RX ring of the
4312 * receive queue, up to *nb_pkts* packets, and for each completed RX
4313 * descriptor in the ring, it performs the following operations:
4315 * - Initialize the *rte_mbuf* data structure associated with the
4316 * RX descriptor according to the information provided by the NIC into
4317 * that RX descriptor.
4319 * - Store the *rte_mbuf* data structure into the next entry of the
4322 * - Replenish the RX descriptor with a new *rte_mbuf* buffer
4323 * allocated from the memory pool associated with the receive queue at
4324 * initialization time.
4326 * When retrieving an input packet that was scattered by the controller
4327 * into multiple receive descriptors, the rte_eth_rx_burst() function
4328 * appends the associated *rte_mbuf* buffers to the first buffer of the
4331 * The rte_eth_rx_burst() function returns the number of packets
4332 * actually retrieved, which is the number of *rte_mbuf* data structures
4333 * effectively supplied into the *rx_pkts* array.
4334 * A return value equal to *nb_pkts* indicates that the RX queue contained
4335 * at least *rx_pkts* packets, and this is likely to signify that other
4336 * received packets remain in the input queue. Applications implementing
4337 * a "retrieve as much received packets as possible" policy can check this
4338 * specific case and keep invoking the rte_eth_rx_burst() function until
4339 * a value less than *nb_pkts* is returned.
4341 * This receive method has the following advantages:
4343 * - It allows a run-to-completion network stack engine to retrieve and
4344 * to immediately process received packets in a fast burst-oriented
4345 * approach, avoiding the overhead of unnecessary intermediate packet
4346 * queue/dequeue operations.
4348 * - Conversely, it also allows an asynchronous-oriented processing
4349 * method to retrieve bursts of received packets and to immediately
4350 * queue them for further parallel processing by another logical core,
4351 * for instance. However, instead of having received packets being
4352 * individually queued by the driver, this approach allows the caller
4353 * of the rte_eth_rx_burst() function to queue a burst of retrieved
4354 * packets at a time and therefore dramatically reduce the cost of
4355 * enqueue/dequeue operations per packet.
4357 * - It allows the rte_eth_rx_burst() function of the driver to take
4358 * advantage of burst-oriented hardware features (CPU cache,
4359 * prefetch instructions, and so on) to minimize the number of CPU
4360 * cycles per packet.
4362 * To summarize, the proposed receive API enables many
4363 * burst-oriented optimizations in both synchronous and asynchronous
4364 * packet processing environments with no overhead in both cases.
4366 * The rte_eth_rx_burst() function does not provide any error
4367 * notification to avoid the corresponding overhead. As a hint, the
4368 * upper-level application might check the status of the device link once
4369 * being systematically returned a 0 value for a given number of tries.
4372 * The port identifier of the Ethernet device.
4374 * The index of the receive queue from which to retrieve input packets.
4375 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
4376 * to rte_eth_dev_configure().
4378 * The address of an array of pointers to *rte_mbuf* structures that
4379 * must be large enough to store *nb_pkts* pointers in it.
4381 * The maximum number of packets to retrieve.
4383 * The number of packets actually retrieved, which is the number
4384 * of pointers to *rte_mbuf* structures effectively supplied to the
4387 static inline uint16_t
4388 rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id,
4389 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
4391 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
4394 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4395 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
4396 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_pkt_burst, 0);
4398 if (queue_id >= dev->data->nb_rx_queues) {
4399 RTE_ETHDEV_LOG(ERR, "Invalid RX queue_id=%u\n", queue_id);
4403 nb_rx = (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id],
4406 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
4407 if (unlikely(dev->post_rx_burst_cbs[queue_id] != NULL)) {
4408 struct rte_eth_rxtx_callback *cb =
4409 dev->post_rx_burst_cbs[queue_id];
4412 nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
4413 nb_pkts, cb->param);
4415 } while (cb != NULL);
4419 rte_ethdev_trace_rx_burst(port_id, queue_id, (void **)rx_pkts, nb_rx);
4424 * Get the number of used descriptors of a rx queue
4427 * The port identifier of the Ethernet device.
4429 * The queue id on the specific port.
4431 * The number of used descriptors in the specific queue, or:
4432 * (-EINVAL) if *port_id* or *queue_id* is invalid
4433 * (-ENOTSUP) if the device does not support this function
4436 rte_eth_rx_queue_count(uint16_t port_id, uint16_t queue_id)
4438 struct rte_eth_dev *dev;
4440 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -EINVAL);
4441 dev = &rte_eth_devices[port_id];
4442 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_count, -ENOTSUP);
4443 if (queue_id >= dev->data->nb_rx_queues)
4446 return (int)(*dev->dev_ops->rx_queue_count)(dev, queue_id);
4450 * Check if the DD bit of the specific RX descriptor in the queue has been set
4453 * The port identifier of the Ethernet device.
4455 * The queue id on the specific port.
4457 * The offset of the descriptor ID from tail.
4459 * - (1) if the specific DD bit is set.
4460 * - (0) if the specific DD bit is not set.
4461 * - (-ENODEV) if *port_id* invalid.
4462 * - (-ENOTSUP) if the device does not support this function
4465 rte_eth_rx_descriptor_done(uint16_t port_id, uint16_t queue_id, uint16_t offset)
4467 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
4468 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
4469 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_descriptor_done, -ENOTSUP);
4470 return (*dev->dev_ops->rx_descriptor_done)( \
4471 dev->data->rx_queues[queue_id], offset);
4474 #define RTE_ETH_RX_DESC_AVAIL 0 /**< Desc available for hw. */
4475 #define RTE_ETH_RX_DESC_DONE 1 /**< Desc done, filled by hw. */
4476 #define RTE_ETH_RX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
4479 * Check the status of a Rx descriptor in the queue
4481 * It should be called in a similar context than the Rx function:
4482 * - on a dataplane core
4483 * - not concurrently on the same queue
4485 * Since it's a dataplane function, no check is performed on port_id and
4486 * queue_id. The caller must therefore ensure that the port is enabled
4487 * and the queue is configured and running.
4489 * Note: accessing to a random descriptor in the ring may trigger cache
4490 * misses and have a performance impact.
4493 * A valid port identifier of the Ethernet device which.
4495 * A valid Rx queue identifier on this port.
4497 * The offset of the descriptor starting from tail (0 is the next
4498 * packet to be received by the driver).
4501 * - (RTE_ETH_RX_DESC_AVAIL): Descriptor is available for the hardware to
4503 * - (RTE_ETH_RX_DESC_DONE): Descriptor is done, it is filled by hw, but
4504 * not yet processed by the driver (i.e. in the receive queue).
4505 * - (RTE_ETH_RX_DESC_UNAVAIL): Descriptor is unavailable, either hold by
4506 * the driver and not yet returned to hw, or reserved by the hw.
4507 * - (-EINVAL) bad descriptor offset.
4508 * - (-ENOTSUP) if the device does not support this function.
4509 * - (-ENODEV) bad port or queue (only if compiled with debug).
4512 rte_eth_rx_descriptor_status(uint16_t port_id, uint16_t queue_id,
4515 struct rte_eth_dev *dev;
4518 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4519 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
4521 dev = &rte_eth_devices[port_id];
4522 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4523 if (queue_id >= dev->data->nb_rx_queues)
4526 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_descriptor_status, -ENOTSUP);
4527 rxq = dev->data->rx_queues[queue_id];
4529 return (*dev->dev_ops->rx_descriptor_status)(rxq, offset);
4532 #define RTE_ETH_TX_DESC_FULL 0 /**< Desc filled for hw, waiting xmit. */
4533 #define RTE_ETH_TX_DESC_DONE 1 /**< Desc done, packet is transmitted. */
4534 #define RTE_ETH_TX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
4537 * Check the status of a Tx descriptor in the queue.
4539 * It should be called in a similar context than the Tx function:
4540 * - on a dataplane core
4541 * - not concurrently on the same queue
4543 * Since it's a dataplane function, no check is performed on port_id and
4544 * queue_id. The caller must therefore ensure that the port is enabled
4545 * and the queue is configured and running.
4547 * Note: accessing to a random descriptor in the ring may trigger cache
4548 * misses and have a performance impact.
4551 * A valid port identifier of the Ethernet device which.
4553 * A valid Tx queue identifier on this port.
4555 * The offset of the descriptor starting from tail (0 is the place where
4556 * the next packet will be send).
4559 * - (RTE_ETH_TX_DESC_FULL) Descriptor is being processed by the hw, i.e.
4560 * in the transmit queue.
4561 * - (RTE_ETH_TX_DESC_DONE) Hardware is done with this descriptor, it can
4562 * be reused by the driver.
4563 * - (RTE_ETH_TX_DESC_UNAVAIL): Descriptor is unavailable, reserved by the
4564 * driver or the hardware.
4565 * - (-EINVAL) bad descriptor offset.
4566 * - (-ENOTSUP) if the device does not support this function.
4567 * - (-ENODEV) bad port or queue (only if compiled with debug).
4569 static inline int rte_eth_tx_descriptor_status(uint16_t port_id,
4570 uint16_t queue_id, uint16_t offset)
4572 struct rte_eth_dev *dev;
4575 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4576 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
4578 dev = &rte_eth_devices[port_id];
4579 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4580 if (queue_id >= dev->data->nb_tx_queues)
4583 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->tx_descriptor_status, -ENOTSUP);
4584 txq = dev->data->tx_queues[queue_id];
4586 return (*dev->dev_ops->tx_descriptor_status)(txq, offset);
4590 * Send a burst of output packets on a transmit queue of an Ethernet device.
4592 * The rte_eth_tx_burst() function is invoked to transmit output packets
4593 * on the output queue *queue_id* of the Ethernet device designated by its
4595 * The *nb_pkts* parameter is the number of packets to send which are
4596 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
4597 * allocated from a pool created with rte_pktmbuf_pool_create().
4598 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets,
4599 * up to the number of transmit descriptors available in the TX ring of the
4601 * For each packet to send, the rte_eth_tx_burst() function performs
4602 * the following operations:
4604 * - Pick up the next available descriptor in the transmit ring.
4606 * - Free the network buffer previously sent with that descriptor, if any.
4608 * - Initialize the transmit descriptor with the information provided
4609 * in the *rte_mbuf data structure.
4611 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers,
4612 * the rte_eth_tx_burst() function uses several transmit descriptors
4615 * The rte_eth_tx_burst() function returns the number of packets it
4616 * actually sent. A return value equal to *nb_pkts* means that all packets
4617 * have been sent, and this is likely to signify that other output packets
4618 * could be immediately transmitted again. Applications that implement a
4619 * "send as many packets to transmit as possible" policy can check this
4620 * specific case and keep invoking the rte_eth_tx_burst() function until
4621 * a value less than *nb_pkts* is returned.
4623 * It is the responsibility of the rte_eth_tx_burst() function to
4624 * transparently free the memory buffers of packets previously sent.
4625 * This feature is driven by the *tx_free_thresh* value supplied to the
4626 * rte_eth_dev_configure() function at device configuration time.
4627 * When the number of free TX descriptors drops below this threshold, the
4628 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers
4629 * of those packets whose transmission was effectively completed.
4631 * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can
4632 * invoke this function concurrently on the same tx queue without SW lock.
4633 * @see rte_eth_dev_info_get, struct rte_eth_txconf::offloads
4635 * @see rte_eth_tx_prepare to perform some prior checks or adjustments
4639 * The port identifier of the Ethernet device.
4641 * The index of the transmit queue through which output packets must be
4643 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4644 * to rte_eth_dev_configure().
4646 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
4647 * which contain the output packets.
4649 * The maximum number of packets to transmit.
4651 * The number of output packets actually stored in transmit descriptors of
4652 * the transmit ring. The return value can be less than the value of the
4653 * *tx_pkts* parameter when the transmit ring is full or has been filled up.
4655 static inline uint16_t
4656 rte_eth_tx_burst(uint16_t port_id, uint16_t queue_id,
4657 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
4659 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
4661 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4662 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
4663 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_pkt_burst, 0);
4665 if (queue_id >= dev->data->nb_tx_queues) {
4666 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
4671 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
4672 struct rte_eth_rxtx_callback *cb = dev->pre_tx_burst_cbs[queue_id];
4674 if (unlikely(cb != NULL)) {
4676 nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
4679 } while (cb != NULL);
4683 rte_ethdev_trace_tx_burst(port_id, queue_id, (void **)tx_pkts,
4685 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts);
4689 * Process a burst of output packets on a transmit queue of an Ethernet device.
4691 * The rte_eth_tx_prepare() function is invoked to prepare output packets to be
4692 * transmitted on the output queue *queue_id* of the Ethernet device designated
4694 * The *nb_pkts* parameter is the number of packets to be prepared which are
4695 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
4696 * allocated from a pool created with rte_pktmbuf_pool_create().
4697 * For each packet to send, the rte_eth_tx_prepare() function performs
4698 * the following operations:
4700 * - Check if packet meets devices requirements for tx offloads.
4702 * - Check limitations about number of segments.
4704 * - Check additional requirements when debug is enabled.
4706 * - Update and/or reset required checksums when tx offload is set for packet.
4708 * Since this function can modify packet data, provided mbufs must be safely
4709 * writable (e.g. modified data cannot be in shared segment).
4711 * The rte_eth_tx_prepare() function returns the number of packets ready to be
4712 * sent. A return value equal to *nb_pkts* means that all packets are valid and
4713 * ready to be sent, otherwise stops processing on the first invalid packet and
4714 * leaves the rest packets untouched.
4716 * When this functionality is not implemented in the driver, all packets are
4717 * are returned untouched.
4720 * The port identifier of the Ethernet device.
4721 * The value must be a valid port id.
4723 * The index of the transmit queue through which output packets must be
4725 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4726 * to rte_eth_dev_configure().
4728 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
4729 * which contain the output packets.
4731 * The maximum number of packets to process.
4733 * The number of packets correct and ready to be sent. The return value can be
4734 * less than the value of the *tx_pkts* parameter when some packet doesn't
4735 * meet devices requirements with rte_errno set appropriately:
4736 * - EINVAL: offload flags are not correctly set
4737 * - ENOTSUP: the offload feature is not supported by the hardware
4741 #ifndef RTE_ETHDEV_TX_PREPARE_NOOP
4743 static inline uint16_t
4744 rte_eth_tx_prepare(uint16_t port_id, uint16_t queue_id,
4745 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
4747 struct rte_eth_dev *dev;
4749 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4750 if (!rte_eth_dev_is_valid_port(port_id)) {
4751 RTE_ETHDEV_LOG(ERR, "Invalid TX port_id=%u\n", port_id);
4757 dev = &rte_eth_devices[port_id];
4759 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4760 if (queue_id >= dev->data->nb_tx_queues) {
4761 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
4767 if (!dev->tx_pkt_prepare)
4770 return (*dev->tx_pkt_prepare)(dev->data->tx_queues[queue_id],
4777 * Native NOOP operation for compilation targets which doesn't require any
4778 * preparations steps, and functional NOOP may introduce unnecessary performance
4781 * Generally this is not a good idea to turn it on globally and didn't should
4782 * be used if behavior of tx_preparation can change.
4785 static inline uint16_t
4786 rte_eth_tx_prepare(__rte_unused uint16_t port_id,
4787 __rte_unused uint16_t queue_id,
4788 __rte_unused struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
4796 * Send any packets queued up for transmission on a port and HW queue
4798 * This causes an explicit flush of packets previously buffered via the
4799 * rte_eth_tx_buffer() function. It returns the number of packets successfully
4800 * sent to the NIC, and calls the error callback for any unsent packets. Unless
4801 * explicitly set up otherwise, the default callback simply frees the unsent
4802 * packets back to the owning mempool.
4805 * The port identifier of the Ethernet device.
4807 * The index of the transmit queue through which output packets must be
4809 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4810 * to rte_eth_dev_configure().
4812 * Buffer of packets to be transmit.
4814 * The number of packets successfully sent to the Ethernet device. The error
4815 * callback is called for any packets which could not be sent.
4817 static inline uint16_t
4818 rte_eth_tx_buffer_flush(uint16_t port_id, uint16_t queue_id,
4819 struct rte_eth_dev_tx_buffer *buffer)
4822 uint16_t to_send = buffer->length;
4827 sent = rte_eth_tx_burst(port_id, queue_id, buffer->pkts, to_send);
4831 /* All packets sent, or to be dealt with by callback below */
4832 if (unlikely(sent != to_send))
4833 buffer->error_callback(&buffer->pkts[sent],
4834 (uint16_t)(to_send - sent),
4835 buffer->error_userdata);
4841 * Buffer a single packet for future transmission on a port and queue
4843 * This function takes a single mbuf/packet and buffers it for later
4844 * transmission on the particular port and queue specified. Once the buffer is
4845 * full of packets, an attempt will be made to transmit all the buffered
4846 * packets. In case of error, where not all packets can be transmitted, a
4847 * callback is called with the unsent packets as a parameter. If no callback
4848 * is explicitly set up, the unsent packets are just freed back to the owning
4849 * mempool. The function returns the number of packets actually sent i.e.
4850 * 0 if no buffer flush occurred, otherwise the number of packets successfully
4854 * The port identifier of the Ethernet device.
4856 * The index of the transmit queue through which output packets must be
4858 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4859 * to rte_eth_dev_configure().
4861 * Buffer used to collect packets to be sent.
4863 * Pointer to the packet mbuf to be sent.
4865 * 0 = packet has been buffered for later transmission
4866 * N > 0 = packet has been buffered, and the buffer was subsequently flushed,
4867 * causing N packets to be sent, and the error callback to be called for
4870 static __rte_always_inline uint16_t
4871 rte_eth_tx_buffer(uint16_t port_id, uint16_t queue_id,
4872 struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt)
4874 buffer->pkts[buffer->length++] = tx_pkt;
4875 if (buffer->length < buffer->size)
4878 return rte_eth_tx_buffer_flush(port_id, queue_id, buffer);
4885 #endif /* _RTE_ETHDEV_H_ */