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>
160 #include "rte_ether.h"
161 #include "rte_dev_info.h"
163 extern int rte_eth_dev_logtype;
165 #define RTE_ETHDEV_LOG(level, ...) \
166 rte_log(RTE_LOG_ ## level, rte_eth_dev_logtype, "" __VA_ARGS__)
171 * Initializes a device iterator.
173 * This iterator allows accessing a list of devices matching some devargs.
176 * Device iterator handle initialized by the function.
177 * The fields bus_str and cls_str might be dynamically allocated,
178 * and could be freed by calling rte_eth_iterator_cleanup().
181 * Device description string.
184 * 0 on successful initialization, negative otherwise.
186 int rte_eth_iterator_init(struct rte_dev_iterator *iter, const char *devargs);
189 * Iterates on devices with devargs filter.
190 * The ownership is not checked.
192 * The next port id is returned, and the iterator is updated.
195 * Device iterator handle initialized by rte_eth_iterator_init().
196 * Some fields bus_str and cls_str might be freed when no more port is found,
197 * by calling rte_eth_iterator_cleanup().
200 * A port id if found, RTE_MAX_ETHPORTS otherwise.
202 uint16_t rte_eth_iterator_next(struct rte_dev_iterator *iter);
205 * Free some allocated fields of the iterator.
207 * This function is automatically called by rte_eth_iterator_next()
208 * on the last iteration (i.e. when no more matching port is found).
210 * It is safe to call this function twice; it will do nothing more.
213 * Device iterator handle initialized by rte_eth_iterator_init().
214 * The fields bus_str and cls_str are freed if needed.
216 void rte_eth_iterator_cleanup(struct rte_dev_iterator *iter);
219 * Macro to iterate over all ethdev ports matching some devargs.
221 * If a break is done before the end of the loop,
222 * the function rte_eth_iterator_cleanup() must be called.
225 * Iterated port id of type uint16_t.
227 * Device parameters input as string of type char*.
229 * Iterator handle of type struct rte_dev_iterator, used internally.
231 #define RTE_ETH_FOREACH_MATCHING_DEV(id, devargs, iter) \
232 for (rte_eth_iterator_init(iter, devargs), \
233 id = rte_eth_iterator_next(iter); \
234 id != RTE_MAX_ETHPORTS; \
235 id = rte_eth_iterator_next(iter))
238 * A structure used to retrieve statistics for an Ethernet port.
239 * Not all statistics fields in struct rte_eth_stats are supported
240 * by any type of network interface card (NIC). If any statistics
241 * field is not supported, its value is 0.
243 struct rte_eth_stats {
244 uint64_t ipackets; /**< Total number of successfully received packets. */
245 uint64_t opackets; /**< Total number of successfully transmitted packets.*/
246 uint64_t ibytes; /**< Total number of successfully received bytes. */
247 uint64_t obytes; /**< Total number of successfully transmitted bytes. */
249 /**< Total of RX packets dropped by the HW,
250 * because there are no available buffer (i.e. RX queues are full).
252 uint64_t ierrors; /**< Total number of erroneous received packets. */
253 uint64_t oerrors; /**< Total number of failed transmitted packets. */
254 uint64_t rx_nombuf; /**< Total number of RX mbuf allocation failures. */
255 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
256 /**< Total number of queue RX packets. */
257 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
258 /**< Total number of queue TX packets. */
259 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
260 /**< Total number of successfully received queue bytes. */
261 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
262 /**< Total number of successfully transmitted queue bytes. */
263 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS];
264 /**< Total number of queue packets received that are dropped. */
268 * Device supported speeds bitmap flags
270 #define ETH_LINK_SPEED_AUTONEG (0 << 0) /**< Autonegotiate (all speeds) */
271 #define ETH_LINK_SPEED_FIXED (1 << 0) /**< Disable autoneg (fixed speed) */
272 #define ETH_LINK_SPEED_10M_HD (1 << 1) /**< 10 Mbps half-duplex */
273 #define ETH_LINK_SPEED_10M (1 << 2) /**< 10 Mbps full-duplex */
274 #define ETH_LINK_SPEED_100M_HD (1 << 3) /**< 100 Mbps half-duplex */
275 #define ETH_LINK_SPEED_100M (1 << 4) /**< 100 Mbps full-duplex */
276 #define ETH_LINK_SPEED_1G (1 << 5) /**< 1 Gbps */
277 #define ETH_LINK_SPEED_2_5G (1 << 6) /**< 2.5 Gbps */
278 #define ETH_LINK_SPEED_5G (1 << 7) /**< 5 Gbps */
279 #define ETH_LINK_SPEED_10G (1 << 8) /**< 10 Gbps */
280 #define ETH_LINK_SPEED_20G (1 << 9) /**< 20 Gbps */
281 #define ETH_LINK_SPEED_25G (1 << 10) /**< 25 Gbps */
282 #define ETH_LINK_SPEED_40G (1 << 11) /**< 40 Gbps */
283 #define ETH_LINK_SPEED_50G (1 << 12) /**< 50 Gbps */
284 #define ETH_LINK_SPEED_56G (1 << 13) /**< 56 Gbps */
285 #define ETH_LINK_SPEED_100G (1 << 14) /**< 100 Gbps */
288 * Ethernet numeric link speeds in Mbps
290 #define ETH_SPEED_NUM_NONE 0 /**< Not defined */
291 #define ETH_SPEED_NUM_10M 10 /**< 10 Mbps */
292 #define ETH_SPEED_NUM_100M 100 /**< 100 Mbps */
293 #define ETH_SPEED_NUM_1G 1000 /**< 1 Gbps */
294 #define ETH_SPEED_NUM_2_5G 2500 /**< 2.5 Gbps */
295 #define ETH_SPEED_NUM_5G 5000 /**< 5 Gbps */
296 #define ETH_SPEED_NUM_10G 10000 /**< 10 Gbps */
297 #define ETH_SPEED_NUM_20G 20000 /**< 20 Gbps */
298 #define ETH_SPEED_NUM_25G 25000 /**< 25 Gbps */
299 #define ETH_SPEED_NUM_40G 40000 /**< 40 Gbps */
300 #define ETH_SPEED_NUM_50G 50000 /**< 50 Gbps */
301 #define ETH_SPEED_NUM_56G 56000 /**< 56 Gbps */
302 #define ETH_SPEED_NUM_100G 100000 /**< 100 Gbps */
305 * A structure used to retrieve link-level information of an Ethernet port.
308 struct rte_eth_link {
309 uint32_t link_speed; /**< ETH_SPEED_NUM_ */
310 uint16_t link_duplex : 1; /**< ETH_LINK_[HALF/FULL]_DUPLEX */
311 uint16_t link_autoneg : 1; /**< ETH_LINK_[AUTONEG/FIXED] */
312 uint16_t link_status : 1; /**< ETH_LINK_[DOWN/UP] */
313 } __attribute__((aligned(8))); /**< aligned for atomic64 read/write */
315 /* Utility constants */
316 #define ETH_LINK_HALF_DUPLEX 0 /**< Half-duplex connection (see link_duplex). */
317 #define ETH_LINK_FULL_DUPLEX 1 /**< Full-duplex connection (see link_duplex). */
318 #define ETH_LINK_DOWN 0 /**< Link is down (see link_status). */
319 #define ETH_LINK_UP 1 /**< Link is up (see link_status). */
320 #define ETH_LINK_FIXED 0 /**< No autonegotiation (see link_autoneg). */
321 #define ETH_LINK_AUTONEG 1 /**< Autonegotiated (see link_autoneg). */
324 * A structure used to configure the ring threshold registers of an RX/TX
325 * queue for an Ethernet port.
327 struct rte_eth_thresh {
328 uint8_t pthresh; /**< Ring prefetch threshold. */
329 uint8_t hthresh; /**< Ring host threshold. */
330 uint8_t wthresh; /**< Ring writeback threshold. */
334 * Simple flags are used for rte_eth_conf.rxmode.mq_mode.
336 #define ETH_MQ_RX_RSS_FLAG 0x1
337 #define ETH_MQ_RX_DCB_FLAG 0x2
338 #define ETH_MQ_RX_VMDQ_FLAG 0x4
341 * A set of values to identify what method is to be used to route
342 * packets to multiple queues.
344 enum rte_eth_rx_mq_mode {
345 /** None of DCB,RSS or VMDQ mode */
348 /** For RX side, only RSS is on */
349 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG,
350 /** For RX side,only DCB is on. */
351 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG,
352 /** Both DCB and RSS enable */
353 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG,
355 /** Only VMDQ, no RSS nor DCB */
356 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG,
357 /** RSS mode with VMDQ */
358 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG,
359 /** Use VMDQ+DCB to route traffic to queues */
360 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG,
361 /** Enable both VMDQ and DCB in VMDq */
362 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG |
367 * for rx mq mode backward compatible
369 #define ETH_RSS ETH_MQ_RX_RSS
370 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB
371 #define ETH_DCB_RX ETH_MQ_RX_DCB
374 * A set of values to identify what method is to be used to transmit
375 * packets using multi-TCs.
377 enum rte_eth_tx_mq_mode {
378 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */
379 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */
380 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */
381 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */
385 * for tx mq mode backward compatible
387 #define ETH_DCB_NONE ETH_MQ_TX_NONE
388 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB
389 #define ETH_DCB_TX ETH_MQ_TX_DCB
392 * A structure used to configure the RX features of an Ethernet port.
394 struct rte_eth_rxmode {
395 /** The multi-queue packet distribution mode to be used, e.g. RSS. */
396 enum rte_eth_rx_mq_mode mq_mode;
397 uint32_t max_rx_pkt_len; /**< Only used if JUMBO_FRAME enabled. */
398 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/
400 * Per-port Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
401 * Only offloads set on rx_offload_capa field on rte_eth_dev_info
402 * structure are allowed to be set.
408 * VLAN types to indicate if it is for single VLAN, inner VLAN or outer VLAN.
409 * Note that single VLAN is treated the same as inner VLAN.
412 ETH_VLAN_TYPE_UNKNOWN = 0,
413 ETH_VLAN_TYPE_INNER, /**< Inner VLAN. */
414 ETH_VLAN_TYPE_OUTER, /**< Single VLAN, or outer VLAN. */
419 * A structure used to describe a vlan filter.
420 * If the bit corresponding to a VID is set, such VID is on.
422 struct rte_vlan_filter_conf {
427 * A structure used to configure the Receive Side Scaling (RSS) feature
428 * of an Ethernet port.
429 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points
430 * to an array holding the RSS key to use for hashing specific header
431 * fields of received packets. The length of this array should be indicated
432 * by *rss_key_len* below. Otherwise, a default random hash key is used by
435 * The *rss_key_len* field of the *rss_conf* structure indicates the length
436 * in bytes of the array pointed by *rss_key*. To be compatible, this length
437 * will be checked in i40e only. Others assume 40 bytes to be used as before.
439 * The *rss_hf* field of the *rss_conf* structure indicates the different
440 * types of IPv4/IPv6 packets to which the RSS hashing must be applied.
441 * Supplying an *rss_hf* equal to zero disables the RSS feature.
443 struct rte_eth_rss_conf {
444 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */
445 uint8_t rss_key_len; /**< hash key length in bytes. */
446 uint64_t rss_hf; /**< Hash functions to apply - see below. */
450 * A packet can be identified by hardware as different flow types. Different
451 * NIC hardware may support different flow types.
452 * Basically, the NIC hardware identifies the flow type as deep protocol as
453 * possible, and exclusively. For example, if a packet is identified as
454 * 'RTE_ETH_FLOW_NONFRAG_IPV4_TCP', it will not be any of other flow types,
455 * though it is an actual IPV4 packet.
456 * Note that the flow types are used to define RSS offload types.
458 #define RTE_ETH_FLOW_UNKNOWN 0
459 #define RTE_ETH_FLOW_RAW 1
460 #define RTE_ETH_FLOW_IPV4 2
461 #define RTE_ETH_FLOW_FRAG_IPV4 3
462 #define RTE_ETH_FLOW_NONFRAG_IPV4_TCP 4
463 #define RTE_ETH_FLOW_NONFRAG_IPV4_UDP 5
464 #define RTE_ETH_FLOW_NONFRAG_IPV4_SCTP 6
465 #define RTE_ETH_FLOW_NONFRAG_IPV4_OTHER 7
466 #define RTE_ETH_FLOW_IPV6 8
467 #define RTE_ETH_FLOW_FRAG_IPV6 9
468 #define RTE_ETH_FLOW_NONFRAG_IPV6_TCP 10
469 #define RTE_ETH_FLOW_NONFRAG_IPV6_UDP 11
470 #define RTE_ETH_FLOW_NONFRAG_IPV6_SCTP 12
471 #define RTE_ETH_FLOW_NONFRAG_IPV6_OTHER 13
472 #define RTE_ETH_FLOW_L2_PAYLOAD 14
473 #define RTE_ETH_FLOW_IPV6_EX 15
474 #define RTE_ETH_FLOW_IPV6_TCP_EX 16
475 #define RTE_ETH_FLOW_IPV6_UDP_EX 17
476 #define RTE_ETH_FLOW_PORT 18
477 /**< Consider device port number as a flow differentiator */
478 #define RTE_ETH_FLOW_VXLAN 19 /**< VXLAN protocol based flow */
479 #define RTE_ETH_FLOW_GENEVE 20 /**< GENEVE protocol based flow */
480 #define RTE_ETH_FLOW_NVGRE 21 /**< NVGRE protocol based flow */
481 #define RTE_ETH_FLOW_VXLAN_GPE 22 /**< VXLAN-GPE protocol based flow */
482 #define RTE_ETH_FLOW_MAX 23
485 * The RSS offload types are defined based on flow types.
486 * Different NIC hardware may support different RSS offload
487 * types. The supported flow types or RSS offload types can be queried by
488 * rte_eth_dev_info_get().
490 #define ETH_RSS_IPV4 (1ULL << RTE_ETH_FLOW_IPV4)
491 #define ETH_RSS_FRAG_IPV4 (1ULL << RTE_ETH_FLOW_FRAG_IPV4)
492 #define ETH_RSS_NONFRAG_IPV4_TCP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_TCP)
493 #define ETH_RSS_NONFRAG_IPV4_UDP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_UDP)
494 #define ETH_RSS_NONFRAG_IPV4_SCTP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_SCTP)
495 #define ETH_RSS_NONFRAG_IPV4_OTHER (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_OTHER)
496 #define ETH_RSS_IPV6 (1ULL << RTE_ETH_FLOW_IPV6)
497 #define ETH_RSS_FRAG_IPV6 (1ULL << RTE_ETH_FLOW_FRAG_IPV6)
498 #define ETH_RSS_NONFRAG_IPV6_TCP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_TCP)
499 #define ETH_RSS_NONFRAG_IPV6_UDP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_UDP)
500 #define ETH_RSS_NONFRAG_IPV6_SCTP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_SCTP)
501 #define ETH_RSS_NONFRAG_IPV6_OTHER (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_OTHER)
502 #define ETH_RSS_L2_PAYLOAD (1ULL << RTE_ETH_FLOW_L2_PAYLOAD)
503 #define ETH_RSS_IPV6_EX (1ULL << RTE_ETH_FLOW_IPV6_EX)
504 #define ETH_RSS_IPV6_TCP_EX (1ULL << RTE_ETH_FLOW_IPV6_TCP_EX)
505 #define ETH_RSS_IPV6_UDP_EX (1ULL << RTE_ETH_FLOW_IPV6_UDP_EX)
506 #define ETH_RSS_PORT (1ULL << RTE_ETH_FLOW_PORT)
507 #define ETH_RSS_VXLAN (1ULL << RTE_ETH_FLOW_VXLAN)
508 #define ETH_RSS_GENEVE (1ULL << RTE_ETH_FLOW_GENEVE)
509 #define ETH_RSS_NVGRE (1ULL << RTE_ETH_FLOW_NVGRE)
511 #define ETH_RSS_IP ( \
513 ETH_RSS_FRAG_IPV4 | \
514 ETH_RSS_NONFRAG_IPV4_OTHER | \
516 ETH_RSS_FRAG_IPV6 | \
517 ETH_RSS_NONFRAG_IPV6_OTHER | \
520 #define ETH_RSS_UDP ( \
521 ETH_RSS_NONFRAG_IPV4_UDP | \
522 ETH_RSS_NONFRAG_IPV6_UDP | \
525 #define ETH_RSS_TCP ( \
526 ETH_RSS_NONFRAG_IPV4_TCP | \
527 ETH_RSS_NONFRAG_IPV6_TCP | \
530 #define ETH_RSS_SCTP ( \
531 ETH_RSS_NONFRAG_IPV4_SCTP | \
532 ETH_RSS_NONFRAG_IPV6_SCTP)
534 #define ETH_RSS_TUNNEL ( \
539 /**< Mask of valid RSS hash protocols */
540 #define ETH_RSS_PROTO_MASK ( \
542 ETH_RSS_FRAG_IPV4 | \
543 ETH_RSS_NONFRAG_IPV4_TCP | \
544 ETH_RSS_NONFRAG_IPV4_UDP | \
545 ETH_RSS_NONFRAG_IPV4_SCTP | \
546 ETH_RSS_NONFRAG_IPV4_OTHER | \
548 ETH_RSS_FRAG_IPV6 | \
549 ETH_RSS_NONFRAG_IPV6_TCP | \
550 ETH_RSS_NONFRAG_IPV6_UDP | \
551 ETH_RSS_NONFRAG_IPV6_SCTP | \
552 ETH_RSS_NONFRAG_IPV6_OTHER | \
553 ETH_RSS_L2_PAYLOAD | \
555 ETH_RSS_IPV6_TCP_EX | \
556 ETH_RSS_IPV6_UDP_EX | \
563 * Definitions used for redirection table entry size.
564 * Some RSS RETA sizes may not be supported by some drivers, check the
565 * documentation or the description of relevant functions for more details.
567 #define ETH_RSS_RETA_SIZE_64 64
568 #define ETH_RSS_RETA_SIZE_128 128
569 #define ETH_RSS_RETA_SIZE_256 256
570 #define ETH_RSS_RETA_SIZE_512 512
571 #define RTE_RETA_GROUP_SIZE 64
573 /* Definitions used for VMDQ and DCB functionality */
574 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */
575 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */
576 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */
577 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */
579 /* DCB capability defines */
580 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */
581 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */
583 /* Definitions used for VLAN Offload functionality */
584 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */
585 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */
586 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */
587 #define ETH_QINQ_STRIP_OFFLOAD 0x0008 /**< QINQ Strip On/Off */
589 /* Definitions used for mask VLAN setting */
590 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */
591 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/
592 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/
593 #define ETH_QINQ_STRIP_MASK 0x0008 /**< QINQ Strip setting mask */
594 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/
596 /* Definitions used for receive MAC address */
597 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */
599 /* Definitions used for unicast hash */
600 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */
602 /* Definitions used for VMDQ pool rx mode setting */
603 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */
604 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */
605 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */
606 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */
607 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */
609 /** Maximum nb. of vlan per mirror rule */
610 #define ETH_MIRROR_MAX_VLANS 64
612 #define ETH_MIRROR_VIRTUAL_POOL_UP 0x01 /**< Virtual Pool uplink Mirroring. */
613 #define ETH_MIRROR_UPLINK_PORT 0x02 /**< Uplink Port Mirroring. */
614 #define ETH_MIRROR_DOWNLINK_PORT 0x04 /**< Downlink Port Mirroring. */
615 #define ETH_MIRROR_VLAN 0x08 /**< VLAN Mirroring. */
616 #define ETH_MIRROR_VIRTUAL_POOL_DOWN 0x10 /**< Virtual Pool downlink Mirroring. */
619 * A structure used to configure VLAN traffic mirror of an Ethernet port.
621 struct rte_eth_vlan_mirror {
622 uint64_t vlan_mask; /**< mask for valid VLAN ID. */
623 /** VLAN ID list for vlan mirroring. */
624 uint16_t vlan_id[ETH_MIRROR_MAX_VLANS];
628 * A structure used to configure traffic mirror of an Ethernet port.
630 struct rte_eth_mirror_conf {
631 uint8_t rule_type; /**< Mirroring rule type */
632 uint8_t dst_pool; /**< Destination pool for this mirror rule. */
633 uint64_t pool_mask; /**< Bitmap of pool for pool mirroring */
634 /** VLAN ID setting for VLAN mirroring. */
635 struct rte_eth_vlan_mirror vlan;
639 * A structure used to configure 64 entries of Redirection Table of the
640 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure
641 * more than 64 entries supported by hardware, an array of this structure
644 struct rte_eth_rss_reta_entry64 {
646 /**< Mask bits indicate which entries need to be updated/queried. */
647 uint16_t reta[RTE_RETA_GROUP_SIZE];
648 /**< Group of 64 redirection table entries. */
652 * This enum indicates the possible number of traffic classes
653 * in DCB configurations
655 enum rte_eth_nb_tcs {
656 ETH_4_TCS = 4, /**< 4 TCs with DCB. */
657 ETH_8_TCS = 8 /**< 8 TCs with DCB. */
661 * This enum indicates the possible number of queue pools
662 * in VMDQ configurations.
664 enum rte_eth_nb_pools {
665 ETH_8_POOLS = 8, /**< 8 VMDq pools. */
666 ETH_16_POOLS = 16, /**< 16 VMDq pools. */
667 ETH_32_POOLS = 32, /**< 32 VMDq pools. */
668 ETH_64_POOLS = 64 /**< 64 VMDq pools. */
671 /* This structure may be extended in future. */
672 struct rte_eth_dcb_rx_conf {
673 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */
674 /** Traffic class each UP mapped to. */
675 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
678 struct rte_eth_vmdq_dcb_tx_conf {
679 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */
680 /** Traffic class each UP mapped to. */
681 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
684 struct rte_eth_dcb_tx_conf {
685 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */
686 /** Traffic class each UP mapped to. */
687 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
690 struct rte_eth_vmdq_tx_conf {
691 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */
695 * A structure used to configure the VMDQ+DCB feature
696 * of an Ethernet port.
698 * Using this feature, packets are routed to a pool of queues, based
699 * on the vlan id in the vlan tag, and then to a specific queue within
700 * that pool, using the user priority vlan tag field.
702 * A default pool may be used, if desired, to route all traffic which
703 * does not match the vlan filter rules.
705 struct rte_eth_vmdq_dcb_conf {
706 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */
707 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
708 uint8_t default_pool; /**< The default pool, if applicable */
709 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
711 uint16_t vlan_id; /**< The vlan id of the received frame */
712 uint64_t pools; /**< Bitmask of pools for packet rx */
713 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
714 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
715 /**< Selects a queue in a pool */
719 * A structure used to configure the VMDQ feature of an Ethernet port when
720 * not combined with the DCB feature.
722 * Using this feature, packets are routed to a pool of queues. By default,
723 * the pool selection is based on the MAC address, the vlan id in the
724 * vlan tag as specified in the pool_map array.
725 * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool
726 * selection using only the MAC address. MAC address to pool mapping is done
727 * using the rte_eth_dev_mac_addr_add function, with the pool parameter
728 * corresponding to the pool id.
730 * Queue selection within the selected pool will be done using RSS when
731 * it is enabled or revert to the first queue of the pool if not.
733 * A default pool may be used, if desired, to route all traffic which
734 * does not match the vlan filter rules or any pool MAC address.
736 struct rte_eth_vmdq_rx_conf {
737 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */
738 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
739 uint8_t default_pool; /**< The default pool, if applicable */
740 uint8_t enable_loop_back; /**< Enable VT loop back */
741 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
742 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */
744 uint16_t vlan_id; /**< The vlan id of the received frame */
745 uint64_t pools; /**< Bitmask of pools for packet rx */
746 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
750 * A structure used to configure the TX features of an Ethernet port.
752 struct rte_eth_txmode {
753 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */
755 * Per-port Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
756 * Only offloads set on tx_offload_capa field on rte_eth_dev_info
757 * structure are allowed to be set.
761 /* For i40e specifically */
764 uint8_t hw_vlan_reject_tagged : 1,
765 /**< If set, reject sending out tagged pkts */
766 hw_vlan_reject_untagged : 1,
767 /**< If set, reject sending out untagged pkts */
768 hw_vlan_insert_pvid : 1;
769 /**< If set, enable port based VLAN insertion */
773 * A structure used to configure an RX ring of an Ethernet port.
775 struct rte_eth_rxconf {
776 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */
777 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
778 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
779 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
781 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags.
782 * Only offloads set on rx_queue_offload_capa or rx_offload_capa
783 * fields on rte_eth_dev_info structure are allowed to be set.
789 * A structure used to configure a TX ring of an Ethernet port.
791 struct rte_eth_txconf {
792 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */
793 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */
794 uint16_t tx_free_thresh; /**< Start freeing TX buffers if there are
795 less free descriptors than this value. */
797 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
799 * Per-queue Tx offloads to be set using DEV_TX_OFFLOAD_* flags.
800 * Only offloads set on tx_queue_offload_capa or tx_offload_capa
801 * fields on rte_eth_dev_info structure are allowed to be set.
807 * A structure contains information about HW descriptor ring limitations.
809 struct rte_eth_desc_lim {
810 uint16_t nb_max; /**< Max allowed number of descriptors. */
811 uint16_t nb_min; /**< Min allowed number of descriptors. */
812 uint16_t nb_align; /**< Number of descriptors should be aligned to. */
815 * Max allowed number of segments per whole packet.
817 * - For TSO packet this is the total number of data descriptors allowed
820 * @see nb_mtu_seg_max
825 * Max number of segments per one MTU.
827 * - For non-TSO packet, this is the maximum allowed number of segments
828 * in a single transmit packet.
830 * - For TSO packet each segment within the TSO may span up to this
835 uint16_t nb_mtu_seg_max;
839 * This enum indicates the flow control mode
841 enum rte_eth_fc_mode {
842 RTE_FC_NONE = 0, /**< Disable flow control. */
843 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */
844 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */
845 RTE_FC_FULL /**< Enable flow control on both side. */
849 * A structure used to configure Ethernet flow control parameter.
850 * These parameters will be configured into the register of the NIC.
851 * Please refer to the corresponding data sheet for proper value.
853 struct rte_eth_fc_conf {
854 uint32_t high_water; /**< High threshold value to trigger XOFF */
855 uint32_t low_water; /**< Low threshold value to trigger XON */
856 uint16_t pause_time; /**< Pause quota in the Pause frame */
857 uint16_t send_xon; /**< Is XON frame need be sent */
858 enum rte_eth_fc_mode mode; /**< Link flow control mode */
859 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */
860 uint8_t autoneg; /**< Use Pause autoneg */
864 * A structure used to configure Ethernet priority flow control parameter.
865 * These parameters will be configured into the register of the NIC.
866 * Please refer to the corresponding data sheet for proper value.
868 struct rte_eth_pfc_conf {
869 struct rte_eth_fc_conf fc; /**< General flow control parameter. */
870 uint8_t priority; /**< VLAN User Priority. */
876 enum rte_eth_tunnel_type {
877 RTE_TUNNEL_TYPE_NONE = 0,
878 RTE_TUNNEL_TYPE_VXLAN,
879 RTE_TUNNEL_TYPE_GENEVE,
880 RTE_TUNNEL_TYPE_TEREDO,
881 RTE_TUNNEL_TYPE_NVGRE,
882 RTE_TUNNEL_TYPE_IP_IN_GRE,
883 RTE_L2_TUNNEL_TYPE_E_TAG,
884 RTE_TUNNEL_TYPE_VXLAN_GPE,
888 /* Deprecated API file for rte_eth_dev_filter_* functions */
889 #include "rte_eth_ctrl.h"
892 * Memory space that can be configured to store Flow Director filters
893 * in the board memory.
895 enum rte_fdir_pballoc_type {
896 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */
897 RTE_FDIR_PBALLOC_128K, /**< 128k. */
898 RTE_FDIR_PBALLOC_256K, /**< 256k. */
902 * Select report mode of FDIR hash information in RX descriptors.
904 enum rte_fdir_status_mode {
905 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */
906 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */
907 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */
911 * A structure used to configure the Flow Director (FDIR) feature
912 * of an Ethernet port.
914 * If mode is RTE_FDIR_MODE_NONE, the pballoc value is ignored.
916 struct rte_fdir_conf {
917 enum rte_fdir_mode mode; /**< Flow Director mode. */
918 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */
919 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */
920 /** RX queue of packets matching a "drop" filter in perfect mode. */
922 struct rte_eth_fdir_masks mask;
923 struct rte_eth_fdir_flex_conf flex_conf;
924 /**< Flex payload configuration. */
928 * UDP tunneling configuration.
929 * Used to config the UDP port for a type of tunnel.
930 * NICs need the UDP port to identify the tunnel type.
931 * Normally a type of tunnel has a default UDP port, this structure can be used
932 * in case if the users want to change or support more UDP port.
934 struct rte_eth_udp_tunnel {
935 uint16_t udp_port; /**< UDP port used for the tunnel. */
936 uint8_t prot_type; /**< Tunnel type. Defined in rte_eth_tunnel_type. */
940 * A structure used to enable/disable specific device interrupts.
942 struct rte_intr_conf {
943 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */
945 /** enable/disable rxq interrupt. 0 (default) - disable, 1 enable */
947 /** enable/disable rmv interrupt. 0 (default) - disable, 1 enable */
952 * A structure used to configure an Ethernet port.
953 * Depending upon the RX multi-queue mode, extra advanced
954 * configuration settings may be needed.
956 struct rte_eth_conf {
957 uint32_t link_speeds; /**< bitmap of ETH_LINK_SPEED_XXX of speeds to be
958 used. ETH_LINK_SPEED_FIXED disables link
959 autonegotiation, and a unique speed shall be
960 set. Otherwise, the bitmap defines the set of
961 speeds to be advertised. If the special value
962 ETH_LINK_SPEED_AUTONEG (0) is used, all speeds
963 supported are advertised. */
964 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */
965 struct rte_eth_txmode txmode; /**< Port TX configuration. */
966 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value
967 is 0, meaning the loopback mode is disabled.
968 Read the datasheet of given ethernet controller
969 for details. The possible values of this field
970 are defined in implementation of each driver. */
972 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */
973 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf;
974 /**< Port vmdq+dcb configuration. */
975 struct rte_eth_dcb_rx_conf dcb_rx_conf;
976 /**< Port dcb RX configuration. */
977 struct rte_eth_vmdq_rx_conf vmdq_rx_conf;
978 /**< Port vmdq RX configuration. */
979 } rx_adv_conf; /**< Port RX filtering configuration. */
981 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf;
982 /**< Port vmdq+dcb TX configuration. */
983 struct rte_eth_dcb_tx_conf dcb_tx_conf;
984 /**< Port dcb TX configuration. */
985 struct rte_eth_vmdq_tx_conf vmdq_tx_conf;
986 /**< Port vmdq TX configuration. */
987 } tx_adv_conf; /**< Port TX DCB configuration (union). */
988 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC
989 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */
990 uint32_t dcb_capability_en;
991 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. DEPRECATED */
992 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */
996 * RX offload capabilities of a device.
998 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001
999 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002
1000 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004
1001 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008
1002 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010
1003 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020
1004 #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040
1005 #define DEV_RX_OFFLOAD_MACSEC_STRIP 0x00000080
1006 #define DEV_RX_OFFLOAD_HEADER_SPLIT 0x00000100
1007 #define DEV_RX_OFFLOAD_VLAN_FILTER 0x00000200
1008 #define DEV_RX_OFFLOAD_VLAN_EXTEND 0x00000400
1009 #define DEV_RX_OFFLOAD_JUMBO_FRAME 0x00000800
1010 #define DEV_RX_OFFLOAD_SCATTER 0x00002000
1011 #define DEV_RX_OFFLOAD_TIMESTAMP 0x00004000
1012 #define DEV_RX_OFFLOAD_SECURITY 0x00008000
1013 #define DEV_RX_OFFLOAD_KEEP_CRC 0x00010000
1014 #define DEV_RX_OFFLOAD_SCTP_CKSUM 0x00020000
1015 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM 0x00040000
1017 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \
1018 DEV_RX_OFFLOAD_UDP_CKSUM | \
1019 DEV_RX_OFFLOAD_TCP_CKSUM)
1020 #define DEV_RX_OFFLOAD_VLAN (DEV_RX_OFFLOAD_VLAN_STRIP | \
1021 DEV_RX_OFFLOAD_VLAN_FILTER | \
1022 DEV_RX_OFFLOAD_VLAN_EXTEND | \
1023 DEV_RX_OFFLOAD_QINQ_STRIP)
1026 * If new Rx offload capabilities are defined, they also must be
1027 * mentioned in rte_rx_offload_names in rte_ethdev.c file.
1031 * TX offload capabilities of a device.
1033 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001
1034 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002
1035 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004
1036 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008
1037 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010
1038 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020
1039 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040
1040 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */
1041 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100
1042 #define DEV_TX_OFFLOAD_VXLAN_TNL_TSO 0x00000200 /**< Used for tunneling packet. */
1043 #define DEV_TX_OFFLOAD_GRE_TNL_TSO 0x00000400 /**< Used for tunneling packet. */
1044 #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */
1045 #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */
1046 #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000
1047 #define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000
1048 /**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same
1049 * tx queue without SW lock.
1051 #define DEV_TX_OFFLOAD_MULTI_SEGS 0x00008000
1052 /**< Device supports multi segment send. */
1053 #define DEV_TX_OFFLOAD_MBUF_FAST_FREE 0x00010000
1054 /**< Device supports optimization for fast release of mbufs.
1055 * When set application must guarantee that per-queue all mbufs comes from
1056 * the same mempool and has refcnt = 1.
1058 #define DEV_TX_OFFLOAD_SECURITY 0x00020000
1060 * Device supports generic UDP tunneled packet TSO.
1061 * Application must set PKT_TX_TUNNEL_UDP and other mbuf fields required
1064 #define DEV_TX_OFFLOAD_UDP_TNL_TSO 0x00040000
1066 * Device supports generic IP tunneled packet TSO.
1067 * Application must set PKT_TX_TUNNEL_IP and other mbuf fields required
1070 #define DEV_TX_OFFLOAD_IP_TNL_TSO 0x00080000
1071 /** Device supports outer UDP checksum */
1072 #define DEV_TX_OFFLOAD_OUTER_UDP_CKSUM 0x00100000
1074 * Device supports match on metadata Tx offload..
1075 * Application must set PKT_TX_METADATA and mbuf metadata field.
1077 #define DEV_TX_OFFLOAD_MATCH_METADATA 0x00200000
1079 #define RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP 0x00000001
1080 /**< Device supports Rx queue setup after device started*/
1081 #define RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP 0x00000002
1082 /**< Device supports Tx queue setup after device started*/
1085 * If new Tx offload capabilities are defined, they also must be
1086 * mentioned in rte_tx_offload_names in rte_ethdev.c file.
1090 * Fallback default preferred Rx/Tx port parameters.
1091 * These are used if an application requests default parameters
1092 * but the PMD does not provide preferred values.
1094 #define RTE_ETH_DEV_FALLBACK_RX_RINGSIZE 512
1095 #define RTE_ETH_DEV_FALLBACK_TX_RINGSIZE 512
1096 #define RTE_ETH_DEV_FALLBACK_RX_NBQUEUES 1
1097 #define RTE_ETH_DEV_FALLBACK_TX_NBQUEUES 1
1100 * Preferred Rx/Tx port parameters.
1101 * There are separate instances of this structure for transmission
1102 * and reception respectively.
1104 struct rte_eth_dev_portconf {
1105 uint16_t burst_size; /**< Device-preferred burst size */
1106 uint16_t ring_size; /**< Device-preferred size of queue rings */
1107 uint16_t nb_queues; /**< Device-preferred number of queues */
1111 * Default values for switch domain id when ethdev does not support switch
1112 * domain definitions.
1114 #define RTE_ETH_DEV_SWITCH_DOMAIN_ID_INVALID (0)
1117 * Ethernet device associated switch information
1119 struct rte_eth_switch_info {
1120 const char *name; /**< switch name */
1121 uint16_t domain_id; /**< switch domain id */
1124 * mapping to the devices physical switch port as enumerated from the
1125 * perspective of the embedded interconnect/switch. For SR-IOV enabled
1126 * device this may correspond to the VF_ID of each virtual function,
1127 * but each driver should explicitly define the mapping of switch
1128 * port identifier to that physical interconnect/switch
1133 * Ethernet device information
1137 * A structure used to retrieve the contextual information of
1138 * an Ethernet device, such as the controlling driver of the
1141 struct rte_eth_dev_info {
1142 struct rte_device *device; /** Generic device information */
1143 const char *driver_name; /**< Device Driver name. */
1144 unsigned int if_index; /**< Index to bound host interface, or 0 if none.
1145 Use if_indextoname() to translate into an interface name. */
1146 uint16_t min_mtu; /**< Minimum MTU allowed */
1147 uint16_t max_mtu; /**< Maximum MTU allowed */
1148 const uint32_t *dev_flags; /**< Device flags */
1149 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */
1150 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */
1151 uint16_t max_rx_queues; /**< Maximum number of RX queues. */
1152 uint16_t max_tx_queues; /**< Maximum number of TX queues. */
1153 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */
1154 uint32_t max_hash_mac_addrs;
1155 /** Maximum number of hash MAC addresses for MTA and UTA. */
1156 uint16_t max_vfs; /**< Maximum number of VFs. */
1157 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
1158 uint64_t rx_offload_capa;
1159 /**< All RX offload capabilities including all per-queue ones */
1160 uint64_t tx_offload_capa;
1161 /**< All TX offload capabilities including all per-queue ones */
1162 uint64_t rx_queue_offload_capa;
1163 /**< Device per-queue RX offload capabilities. */
1164 uint64_t tx_queue_offload_capa;
1165 /**< Device per-queue TX offload capabilities. */
1167 /**< Device redirection table size, the total number of entries. */
1168 uint8_t hash_key_size; /**< Hash key size in bytes */
1169 /** Bit mask of RSS offloads, the bit offset also means flow type */
1170 uint64_t flow_type_rss_offloads;
1171 struct rte_eth_rxconf default_rxconf; /**< Default RX configuration */
1172 struct rte_eth_txconf default_txconf; /**< Default TX configuration */
1173 uint16_t vmdq_queue_base; /**< First queue ID for VMDQ pools. */
1174 uint16_t vmdq_queue_num; /**< Queue number for VMDQ pools. */
1175 uint16_t vmdq_pool_base; /**< First ID of VMDQ pools. */
1176 struct rte_eth_desc_lim rx_desc_lim; /**< RX descriptors limits */
1177 struct rte_eth_desc_lim tx_desc_lim; /**< TX descriptors limits */
1178 uint32_t speed_capa; /**< Supported speeds bitmap (ETH_LINK_SPEED_). */
1179 /** Configured number of rx/tx queues */
1180 uint16_t nb_rx_queues; /**< Number of RX queues. */
1181 uint16_t nb_tx_queues; /**< Number of TX queues. */
1182 /** Rx parameter recommendations */
1183 struct rte_eth_dev_portconf default_rxportconf;
1184 /** Tx parameter recommendations */
1185 struct rte_eth_dev_portconf default_txportconf;
1186 /** Generic device capabilities (RTE_ETH_DEV_CAPA_). */
1189 * Switching information for ports on a device with a
1190 * embedded managed interconnect/switch.
1192 struct rte_eth_switch_info switch_info;
1196 * Ethernet device RX queue information structure.
1197 * Used to retrieve information about configured queue.
1199 struct rte_eth_rxq_info {
1200 struct rte_mempool *mp; /**< mempool used by that queue. */
1201 struct rte_eth_rxconf conf; /**< queue config parameters. */
1202 uint8_t scattered_rx; /**< scattered packets RX supported. */
1203 uint16_t nb_desc; /**< configured number of RXDs. */
1204 } __rte_cache_min_aligned;
1207 * Ethernet device TX queue information structure.
1208 * Used to retrieve information about configured queue.
1210 struct rte_eth_txq_info {
1211 struct rte_eth_txconf conf; /**< queue config parameters. */
1212 uint16_t nb_desc; /**< configured number of TXDs. */
1213 } __rte_cache_min_aligned;
1215 /** Maximum name length for extended statistics counters */
1216 #define RTE_ETH_XSTATS_NAME_SIZE 64
1219 * An Ethernet device extended statistic structure
1221 * This structure is used by rte_eth_xstats_get() to provide
1222 * statistics that are not provided in the generic *rte_eth_stats*
1224 * It maps a name id, corresponding to an index in the array returned
1225 * by rte_eth_xstats_get_names(), to a statistic value.
1227 struct rte_eth_xstat {
1228 uint64_t id; /**< The index in xstats name array. */
1229 uint64_t value; /**< The statistic counter value. */
1233 * A name element for extended statistics.
1235 * An array of this structure is returned by rte_eth_xstats_get_names().
1236 * It lists the names of extended statistics for a PMD. The *rte_eth_xstat*
1237 * structure references these names by their array index.
1239 struct rte_eth_xstat_name {
1240 char name[RTE_ETH_XSTATS_NAME_SIZE]; /**< The statistic name. */
1243 #define ETH_DCB_NUM_TCS 8
1244 #define ETH_MAX_VMDQ_POOL 64
1247 * A structure used to get the information of queue and
1248 * TC mapping on both TX and RX paths.
1250 struct rte_eth_dcb_tc_queue_mapping {
1251 /** rx queues assigned to tc per Pool */
1255 } tc_rxq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1256 /** rx queues assigned to tc per Pool */
1260 } tc_txq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1264 * A structure used to get the information of DCB.
1265 * It includes TC UP mapping and queue TC mapping.
1267 struct rte_eth_dcb_info {
1268 uint8_t nb_tcs; /**< number of TCs */
1269 uint8_t prio_tc[ETH_DCB_NUM_USER_PRIORITIES]; /**< Priority to tc */
1270 uint8_t tc_bws[ETH_DCB_NUM_TCS]; /**< TX BW percentage for each TC */
1271 /** rx queues assigned to tc */
1272 struct rte_eth_dcb_tc_queue_mapping tc_queue;
1276 * RX/TX queue states
1278 #define RTE_ETH_QUEUE_STATE_STOPPED 0
1279 #define RTE_ETH_QUEUE_STATE_STARTED 1
1281 #define RTE_ETH_ALL RTE_MAX_ETHPORTS
1283 /* Macros to check for valid port */
1284 #define RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, retval) do { \
1285 if (!rte_eth_dev_is_valid_port(port_id)) { \
1286 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1291 #define RTE_ETH_VALID_PORTID_OR_RET(port_id) do { \
1292 if (!rte_eth_dev_is_valid_port(port_id)) { \
1293 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \
1299 * l2 tunnel configuration.
1302 /**< l2 tunnel enable mask */
1303 #define ETH_L2_TUNNEL_ENABLE_MASK 0x00000001
1304 /**< l2 tunnel insertion mask */
1305 #define ETH_L2_TUNNEL_INSERTION_MASK 0x00000002
1306 /**< l2 tunnel stripping mask */
1307 #define ETH_L2_TUNNEL_STRIPPING_MASK 0x00000004
1308 /**< l2 tunnel forwarding mask */
1309 #define ETH_L2_TUNNEL_FORWARDING_MASK 0x00000008
1312 * Function type used for RX packet processing packet callbacks.
1314 * The callback function is called on RX with a burst of packets that have
1315 * been received on the given port and queue.
1318 * The Ethernet port on which RX is being performed.
1320 * The queue on the Ethernet port which is being used to receive the packets.
1322 * The burst of packets that have just been received.
1324 * The number of packets in the burst pointed to by "pkts".
1326 * The max number of packets that can be stored in the "pkts" array.
1328 * The arbitrary user parameter passed in by the application when the callback
1329 * was originally configured.
1331 * The number of packets returned to the user.
1333 typedef uint16_t (*rte_rx_callback_fn)(uint16_t port_id, uint16_t queue,
1334 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts,
1338 * Function type used for TX packet processing packet callbacks.
1340 * The callback function is called on TX with a burst of packets immediately
1341 * before the packets are put onto the hardware queue for transmission.
1344 * The Ethernet port on which TX is being performed.
1346 * The queue on the Ethernet port which is being used to transmit the packets.
1348 * The burst of packets that are about to be transmitted.
1350 * The number of packets in the burst pointed to by "pkts".
1352 * The arbitrary user parameter passed in by the application when the callback
1353 * was originally configured.
1355 * The number of packets to be written to the NIC.
1357 typedef uint16_t (*rte_tx_callback_fn)(uint16_t port_id, uint16_t queue,
1358 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param);
1361 * Possible states of an ethdev port.
1363 enum rte_eth_dev_state {
1364 /** Device is unused before being probed. */
1365 RTE_ETH_DEV_UNUSED = 0,
1366 /** Device is attached when allocated in probing. */
1367 RTE_ETH_DEV_ATTACHED,
1368 /** Device is in removed state when plug-out is detected. */
1369 RTE_ETH_DEV_REMOVED,
1372 struct rte_eth_dev_sriov {
1373 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */
1374 uint8_t nb_q_per_pool; /**< rx queue number per pool */
1375 uint16_t def_vmdq_idx; /**< Default pool num used for PF */
1376 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */
1378 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
1380 #define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN
1382 #define RTE_ETH_DEV_NO_OWNER 0
1384 #define RTE_ETH_MAX_OWNER_NAME_LEN 64
1386 struct rte_eth_dev_owner {
1387 uint64_t id; /**< The owner unique identifier. */
1388 char name[RTE_ETH_MAX_OWNER_NAME_LEN]; /**< The owner name. */
1392 * Port is released (i.e. totally freed and data erased) on close.
1393 * Temporary flag for PMD migration to new rte_eth_dev_close() behaviour.
1395 #define RTE_ETH_DEV_CLOSE_REMOVE 0x0001
1396 /** Device supports link state interrupt */
1397 #define RTE_ETH_DEV_INTR_LSC 0x0002
1398 /** Device is a bonded slave */
1399 #define RTE_ETH_DEV_BONDED_SLAVE 0x0004
1400 /** Device supports device removal interrupt */
1401 #define RTE_ETH_DEV_INTR_RMV 0x0008
1402 /** Device is port representor */
1403 #define RTE_ETH_DEV_REPRESENTOR 0x0010
1404 /** Device does not support MAC change after started */
1405 #define RTE_ETH_DEV_NOLIVE_MAC_ADDR 0x0020
1408 * Iterates over valid ethdev ports owned by a specific owner.
1411 * The id of the next possible valid owned port.
1413 * The owner identifier.
1414 * RTE_ETH_DEV_NO_OWNER means iterate over all valid ownerless ports.
1416 * Next valid port id owned by owner_id, RTE_MAX_ETHPORTS if there is none.
1418 uint64_t rte_eth_find_next_owned_by(uint16_t port_id,
1419 const uint64_t owner_id);
1422 * Macro to iterate over all enabled ethdev ports owned by a specific owner.
1424 #define RTE_ETH_FOREACH_DEV_OWNED_BY(p, o) \
1425 for (p = rte_eth_find_next_owned_by(0, o); \
1426 (unsigned int)p < (unsigned int)RTE_MAX_ETHPORTS; \
1427 p = rte_eth_find_next_owned_by(p + 1, o))
1430 * Iterates over valid ethdev ports.
1433 * The id of the next possible valid port.
1435 * Next valid port id, RTE_MAX_ETHPORTS if there is none.
1437 uint16_t rte_eth_find_next(uint16_t port_id);
1440 * Macro to iterate over all enabled and ownerless ethdev ports.
1442 #define RTE_ETH_FOREACH_DEV(p) \
1443 RTE_ETH_FOREACH_DEV_OWNED_BY(p, RTE_ETH_DEV_NO_OWNER)
1447 * @b EXPERIMENTAL: this API may change without prior notice.
1449 * Iterates over ethdev ports of a specified device.
1451 * @param port_id_start
1452 * The id of the next possible valid port.
1454 * The generic device behind the ports to iterate.
1456 * Next port id of the device, possibly port_id_start,
1457 * RTE_MAX_ETHPORTS if there is none.
1461 rte_eth_find_next_of(uint16_t port_id_start,
1462 const struct rte_device *parent);
1465 * Macro to iterate over all ethdev ports of a specified device.
1468 * The id of the matching port being iterated.
1470 * The rte_device pointer matching the iterated ports.
1472 #define RTE_ETH_FOREACH_DEV_OF(port_id, parent) \
1473 for (port_id = rte_eth_find_next_of(0, parent); \
1474 port_id < RTE_MAX_ETHPORTS; \
1475 port_id = rte_eth_find_next_of(port_id + 1, parent))
1479 * @b EXPERIMENTAL: this API may change without prior notice.
1481 * Iterates over sibling ethdev ports (i.e. sharing the same rte_device).
1483 * @param port_id_start
1484 * The id of the next possible valid sibling port.
1485 * @param ref_port_id
1486 * The id of a reference port to compare rte_device with.
1488 * Next sibling port id, possibly port_id_start or ref_port_id itself,
1489 * RTE_MAX_ETHPORTS if there is none.
1493 rte_eth_find_next_sibling(uint16_t port_id_start,
1494 uint16_t ref_port_id);
1497 * Macro to iterate over all ethdev ports sharing the same rte_device
1498 * as the specified port.
1499 * Note: the specified reference port is part of the loop iterations.
1502 * The id of the matching port being iterated.
1503 * @param ref_port_id
1504 * The id of the port being compared.
1506 #define RTE_ETH_FOREACH_DEV_SIBLING(port_id, ref_port_id) \
1507 for (port_id = rte_eth_find_next_sibling(0, ref_port_id); \
1508 port_id < RTE_MAX_ETHPORTS; \
1509 port_id = rte_eth_find_next_sibling(port_id + 1, ref_port_id))
1513 * @b EXPERIMENTAL: this API may change without prior notice.
1515 * Get a new unique owner identifier.
1516 * An owner identifier is used to owns Ethernet devices by only one DPDK entity
1517 * to avoid multiple management of device by different entities.
1520 * Owner identifier pointer.
1522 * Negative errno value on error, 0 on success.
1525 int rte_eth_dev_owner_new(uint64_t *owner_id);
1529 * @b EXPERIMENTAL: this API may change without prior notice.
1531 * Set an Ethernet device owner.
1534 * The identifier of the port to own.
1536 * The owner pointer.
1538 * Negative errno value on error, 0 on success.
1541 int rte_eth_dev_owner_set(const uint16_t port_id,
1542 const struct rte_eth_dev_owner *owner);
1546 * @b EXPERIMENTAL: this API may change without prior notice.
1548 * Unset Ethernet device owner to make the device ownerless.
1551 * The identifier of port to make ownerless.
1553 * The owner identifier.
1555 * 0 on success, negative errno value on error.
1558 int rte_eth_dev_owner_unset(const uint16_t port_id,
1559 const uint64_t owner_id);
1563 * @b EXPERIMENTAL: this API may change without prior notice.
1565 * Remove owner from all Ethernet devices owned by a specific owner.
1568 * The owner identifier.
1570 * 0 on success, negative errno value on error.
1573 int rte_eth_dev_owner_delete(const uint64_t owner_id);
1577 * @b EXPERIMENTAL: this API may change without prior notice.
1579 * Get the owner of an Ethernet device.
1582 * The port identifier.
1584 * The owner structure pointer to fill.
1586 * 0 on success, negative errno value on error..
1589 int rte_eth_dev_owner_get(const uint16_t port_id,
1590 struct rte_eth_dev_owner *owner);
1593 * Get the total number of Ethernet devices that have been successfully
1594 * initialized by the matching Ethernet driver during the PCI probing phase
1595 * and that are available for applications to use. These devices must be
1596 * accessed by using the ``RTE_ETH_FOREACH_DEV()`` macro to deal with
1597 * non-contiguous ranges of devices.
1598 * These non-contiguous ranges can be created by calls to hotplug functions or
1602 * - The total number of usable Ethernet devices.
1605 uint16_t rte_eth_dev_count(void);
1608 * Get the number of ports which are usable for the application.
1610 * These devices must be iterated by using the macro
1611 * ``RTE_ETH_FOREACH_DEV`` or ``RTE_ETH_FOREACH_DEV_OWNED_BY``
1612 * to deal with non-contiguous ranges of devices.
1615 * The count of available Ethernet devices.
1617 uint16_t rte_eth_dev_count_avail(void);
1620 * Get the total number of ports which are allocated.
1622 * Some devices may not be available for the application.
1625 * The total count of Ethernet devices.
1627 uint16_t rte_eth_dev_count_total(void);
1630 * Convert a numerical speed in Mbps to a bitmap flag that can be used in
1631 * the bitmap link_speeds of the struct rte_eth_conf
1634 * Numerical speed value in Mbps
1636 * ETH_LINK_[HALF/FULL]_DUPLEX (only for 10/100M speeds)
1638 * 0 if the speed cannot be mapped
1640 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex);
1643 * Get DEV_RX_OFFLOAD_* flag name.
1648 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
1650 const char *rte_eth_dev_rx_offload_name(uint64_t offload);
1653 * Get DEV_TX_OFFLOAD_* flag name.
1658 * Offload name or 'UNKNOWN' if the flag cannot be recognised.
1660 const char *rte_eth_dev_tx_offload_name(uint64_t offload);
1663 * Configure an Ethernet device.
1664 * This function must be invoked first before any other function in the
1665 * Ethernet API. This function can also be re-invoked when a device is in the
1669 * The port identifier of the Ethernet device to configure.
1670 * @param nb_rx_queue
1671 * The number of receive queues to set up for the Ethernet device.
1672 * @param nb_tx_queue
1673 * The number of transmit queues to set up for the Ethernet device.
1675 * The pointer to the configuration data to be used for the Ethernet device.
1676 * The *rte_eth_conf* structure includes:
1677 * - the hardware offload features to activate, with dedicated fields for
1678 * each statically configurable offload hardware feature provided by
1679 * Ethernet devices, such as IP checksum or VLAN tag stripping for
1681 * The Rx offload bitfield API is obsolete and will be deprecated.
1682 * Applications should set the ignore_bitfield_offloads bit on *rxmode*
1683 * structure and use offloads field to set per-port offloads instead.
1684 * - Any offloading set in eth_conf->[rt]xmode.offloads must be within
1685 * the [rt]x_offload_capa returned from rte_eth_dev_info_get().
1686 * Any type of device supported offloading set in the input argument
1687 * eth_conf->[rt]xmode.offloads to rte_eth_dev_configure() is enabled
1688 * on all queues and it can't be disabled in rte_eth_[rt]x_queue_setup()
1689 * - the Receive Side Scaling (RSS) configuration when using multiple RX
1690 * queues per port. Any RSS hash function set in eth_conf->rss_conf.rss_hf
1691 * must be within the flow_type_rss_offloads provided by drivers via
1692 * rte_eth_dev_info_get() API.
1694 * Embedding all configuration information in a single data structure
1695 * is the more flexible method that allows the addition of new features
1696 * without changing the syntax of the API.
1698 * - 0: Success, device configured.
1699 * - <0: Error code returned by the driver configuration function.
1701 int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue,
1702 uint16_t nb_tx_queue, const struct rte_eth_conf *eth_conf);
1706 * @b EXPERIMENTAL: this API may change without prior notice.
1708 * Check if an Ethernet device was physically removed.
1711 * The port identifier of the Ethernet device.
1713 * 1 when the Ethernet device is removed, otherwise 0.
1717 rte_eth_dev_is_removed(uint16_t port_id);
1720 * Allocate and set up a receive queue for an Ethernet device.
1722 * The function allocates a contiguous block of memory for *nb_rx_desc*
1723 * receive descriptors from a memory zone associated with *socket_id*
1724 * and initializes each receive descriptor with a network buffer allocated
1725 * from the memory pool *mb_pool*.
1728 * The port identifier of the Ethernet device.
1729 * @param rx_queue_id
1730 * The index of the receive queue to set up.
1731 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1732 * to rte_eth_dev_configure().
1734 * The number of receive descriptors to allocate for the receive ring.
1736 * The *socket_id* argument is the socket identifier in case of NUMA.
1737 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1738 * the DMA memory allocated for the receive descriptors of the ring.
1740 * The pointer to the configuration data to be used for the receive queue.
1741 * NULL value is allowed, in which case default RX configuration
1743 * The *rx_conf* structure contains an *rx_thresh* structure with the values
1744 * of the Prefetch, Host, and Write-Back threshold registers of the receive
1746 * In addition it contains the hardware offloads features to activate using
1747 * the DEV_RX_OFFLOAD_* flags.
1748 * If an offloading set in rx_conf->offloads
1749 * hasn't been set in the input argument eth_conf->rxmode.offloads
1750 * to rte_eth_dev_configure(), it is a new added offloading, it must be
1751 * per-queue type and it is enabled for the queue.
1752 * No need to repeat any bit in rx_conf->offloads which has already been
1753 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
1754 * at port level can't be disabled at queue level.
1756 * The pointer to the memory pool from which to allocate *rte_mbuf* network
1757 * memory buffers to populate each descriptor of the receive ring.
1759 * - 0: Success, receive queue correctly set up.
1760 * - -EIO: if device is removed.
1761 * - -EINVAL: The memory pool pointer is null or the size of network buffers
1762 * which can be allocated from this memory pool does not fit the various
1763 * buffer sizes allowed by the device controller.
1764 * - -ENOMEM: Unable to allocate the receive ring descriptors or to
1765 * allocate network memory buffers from the memory pool when
1766 * initializing receive descriptors.
1768 int rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
1769 uint16_t nb_rx_desc, unsigned int socket_id,
1770 const struct rte_eth_rxconf *rx_conf,
1771 struct rte_mempool *mb_pool);
1774 * Allocate and set up a transmit queue for an Ethernet device.
1777 * The port identifier of the Ethernet device.
1778 * @param tx_queue_id
1779 * The index of the transmit queue to set up.
1780 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1781 * to rte_eth_dev_configure().
1783 * The number of transmit descriptors to allocate for the transmit ring.
1785 * The *socket_id* argument is the socket identifier in case of NUMA.
1786 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1787 * the DMA memory allocated for the transmit descriptors of the ring.
1789 * The pointer to the configuration data to be used for the transmit queue.
1790 * NULL value is allowed, in which case default TX configuration
1792 * The *tx_conf* structure contains the following data:
1793 * - The *tx_thresh* structure with the values of the Prefetch, Host, and
1794 * Write-Back threshold registers of the transmit ring.
1795 * When setting Write-Back threshold to the value greater then zero,
1796 * *tx_rs_thresh* value should be explicitly set to one.
1797 * - The *tx_free_thresh* value indicates the [minimum] number of network
1798 * buffers that must be pending in the transmit ring to trigger their
1799 * [implicit] freeing by the driver transmit function.
1800 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit
1801 * descriptors that must be pending in the transmit ring before setting the
1802 * RS bit on a descriptor by the driver transmit function.
1803 * The *tx_rs_thresh* value should be less or equal then
1804 * *tx_free_thresh* value, and both of them should be less then
1806 * - The *offloads* member contains Tx offloads to be enabled.
1807 * If an offloading set in tx_conf->offloads
1808 * hasn't been set in the input argument eth_conf->txmode.offloads
1809 * to rte_eth_dev_configure(), it is a new added offloading, it must be
1810 * per-queue type and it is enabled for the queue.
1811 * No need to repeat any bit in tx_conf->offloads which has already been
1812 * enabled in rte_eth_dev_configure() at port level. An offloading enabled
1813 * at port level can't be disabled at queue level.
1815 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces
1816 * the transmit function to use default values.
1818 * - 0: Success, the transmit queue is correctly set up.
1819 * - -ENOMEM: Unable to allocate the transmit ring descriptors.
1821 int rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,
1822 uint16_t nb_tx_desc, unsigned int socket_id,
1823 const struct rte_eth_txconf *tx_conf);
1826 * Return the NUMA socket to which an Ethernet device is connected
1829 * The port identifier of the Ethernet device
1831 * The NUMA socket id to which the Ethernet device is connected or
1832 * a default of zero if the socket could not be determined.
1833 * -1 is returned is the port_id value is out of range.
1835 int rte_eth_dev_socket_id(uint16_t port_id);
1838 * Check if port_id of device is attached
1841 * The port identifier of the Ethernet device
1843 * - 0 if port is out of range or not attached
1844 * - 1 if device is attached
1846 int rte_eth_dev_is_valid_port(uint16_t port_id);
1849 * Start specified RX queue of a port. It is used when rx_deferred_start
1850 * flag of the specified queue is true.
1853 * The port identifier of the Ethernet device
1854 * @param rx_queue_id
1855 * The index of the rx queue to update the ring.
1856 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1857 * to rte_eth_dev_configure().
1859 * - 0: Success, the receive queue is started.
1860 * - -EINVAL: The port_id or the queue_id out of range.
1861 * - -EIO: if device is removed.
1862 * - -ENOTSUP: The function not supported in PMD driver.
1864 int rte_eth_dev_rx_queue_start(uint16_t port_id, uint16_t rx_queue_id);
1867 * Stop specified RX queue of a port
1870 * The port identifier of the Ethernet device
1871 * @param rx_queue_id
1872 * The index of the rx queue to update the ring.
1873 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1874 * to rte_eth_dev_configure().
1876 * - 0: Success, the receive queue is stopped.
1877 * - -EINVAL: The port_id or the queue_id out of range.
1878 * - -EIO: if device is removed.
1879 * - -ENOTSUP: The function not supported in PMD driver.
1881 int rte_eth_dev_rx_queue_stop(uint16_t port_id, uint16_t rx_queue_id);
1884 * Start TX for specified queue of a port. It is used when tx_deferred_start
1885 * flag of the specified queue is true.
1888 * The port identifier of the Ethernet device
1889 * @param tx_queue_id
1890 * The index of the tx queue to update the ring.
1891 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1892 * to rte_eth_dev_configure().
1894 * - 0: Success, the transmit queue is started.
1895 * - -EINVAL: The port_id or the queue_id out of range.
1896 * - -EIO: if device is removed.
1897 * - -ENOTSUP: The function not supported in PMD driver.
1899 int rte_eth_dev_tx_queue_start(uint16_t port_id, uint16_t tx_queue_id);
1902 * Stop specified TX queue of a port
1905 * The port identifier of the Ethernet device
1906 * @param tx_queue_id
1907 * The index of the tx queue to update the ring.
1908 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1909 * to rte_eth_dev_configure().
1911 * - 0: Success, the transmit queue is stopped.
1912 * - -EINVAL: The port_id or the queue_id out of range.
1913 * - -EIO: if device is removed.
1914 * - -ENOTSUP: The function not supported in PMD driver.
1916 int rte_eth_dev_tx_queue_stop(uint16_t port_id, uint16_t tx_queue_id);
1919 * Start an Ethernet device.
1921 * The device start step is the last one and consists of setting the configured
1922 * offload features and in starting the transmit and the receive units of the
1925 * Device RTE_ETH_DEV_NOLIVE_MAC_ADDR flag causes MAC address to be set before
1926 * PMD port start callback function is invoked.
1928 * On success, all basic functions exported by the Ethernet API (link status,
1929 * receive/transmit, and so on) can be invoked.
1932 * The port identifier of the Ethernet device.
1934 * - 0: Success, Ethernet device started.
1935 * - <0: Error code of the driver device start function.
1937 int rte_eth_dev_start(uint16_t port_id);
1940 * Stop an Ethernet device. The device can be restarted with a call to
1941 * rte_eth_dev_start()
1944 * The port identifier of the Ethernet device.
1946 void rte_eth_dev_stop(uint16_t port_id);
1949 * Link up an Ethernet device.
1951 * Set device link up will re-enable the device rx/tx
1952 * functionality after it is previously set device linked down.
1955 * The port identifier of the Ethernet device.
1957 * - 0: Success, Ethernet device linked up.
1958 * - <0: Error code of the driver device link up function.
1960 int rte_eth_dev_set_link_up(uint16_t port_id);
1963 * Link down an Ethernet device.
1964 * The device rx/tx functionality will be disabled if success,
1965 * and it can be re-enabled with a call to
1966 * rte_eth_dev_set_link_up()
1969 * The port identifier of the Ethernet device.
1971 int rte_eth_dev_set_link_down(uint16_t port_id);
1974 * Close a stopped Ethernet device. The device cannot be restarted!
1975 * The function frees all port resources if the driver supports
1976 * the flag RTE_ETH_DEV_CLOSE_REMOVE.
1979 * The port identifier of the Ethernet device.
1981 void rte_eth_dev_close(uint16_t port_id);
1984 * Reset a Ethernet device and keep its port id.
1986 * When a port has to be reset passively, the DPDK application can invoke
1987 * this function. For example when a PF is reset, all its VFs should also
1988 * be reset. Normally a DPDK application can invoke this function when
1989 * RTE_ETH_EVENT_INTR_RESET event is detected, but can also use it to start
1990 * a port reset in other circumstances.
1992 * When this function is called, it first stops the port and then calls the
1993 * PMD specific dev_uninit( ) and dev_init( ) to return the port to initial
1994 * state, in which no Tx and Rx queues are setup, as if the port has been
1995 * reset and not started. The port keeps the port id it had before the
1998 * After calling rte_eth_dev_reset( ), the application should use
1999 * rte_eth_dev_configure( ), rte_eth_rx_queue_setup( ),
2000 * rte_eth_tx_queue_setup( ), and rte_eth_dev_start( )
2001 * to reconfigure the device as appropriate.
2003 * Note: To avoid unexpected behavior, the application should stop calling
2004 * Tx and Rx functions before calling rte_eth_dev_reset( ). For thread
2005 * safety, all these controlling functions should be called from the same
2009 * The port identifier of the Ethernet device.
2012 * - (0) if successful.
2013 * - (-EINVAL) if port identifier is invalid.
2014 * - (-ENOTSUP) if hardware doesn't support this function.
2015 * - (-EPERM) if not ran from the primary process.
2016 * - (-EIO) if re-initialisation failed or device is removed.
2017 * - (-ENOMEM) if the reset failed due to OOM.
2018 * - (-EAGAIN) if the reset temporarily failed and should be retried later.
2020 int rte_eth_dev_reset(uint16_t port_id);
2023 * Enable receipt in promiscuous mode for an Ethernet device.
2026 * The port identifier of the Ethernet device.
2028 * - (0) if successful.
2029 * - (-ENOTSUP) if support for promiscuous_enable() does not exist
2031 * - (-ENODEV) if *port_id* invalid.
2033 int rte_eth_promiscuous_enable(uint16_t port_id);
2036 * Disable receipt in promiscuous mode for an Ethernet device.
2039 * The port identifier of the Ethernet device.
2041 * - (0) if successful.
2042 * - (-ENOTSUP) if support for promiscuous_disable() does not exist
2044 * - (-ENODEV) if *port_id* invalid.
2046 int rte_eth_promiscuous_disable(uint16_t port_id);
2049 * Return the value of promiscuous mode for an Ethernet device.
2052 * The port identifier of the Ethernet device.
2054 * - (1) if promiscuous is enabled
2055 * - (0) if promiscuous is disabled.
2058 int rte_eth_promiscuous_get(uint16_t port_id);
2061 * Enable the receipt of any multicast frame by an Ethernet device.
2064 * The port identifier of the Ethernet device.
2066 * - (0) if successful.
2067 * - (-ENOTSUP) if support for allmulticast_enable() does not exist
2069 * - (-ENODEV) if *port_id* invalid.
2071 int rte_eth_allmulticast_enable(uint16_t port_id);
2074 * Disable the receipt of all multicast frames by an Ethernet device.
2077 * The port identifier of the Ethernet device.
2079 * - (0) if successful.
2080 * - (-ENOTSUP) if support for allmulticast_disable() does not exist
2082 * - (-ENODEV) if *port_id* invalid.
2084 int rte_eth_allmulticast_disable(uint16_t port_id);
2087 * Return the value of allmulticast mode for an Ethernet device.
2090 * The port identifier of the Ethernet device.
2092 * - (1) if allmulticast is enabled
2093 * - (0) if allmulticast is disabled.
2096 int rte_eth_allmulticast_get(uint16_t port_id);
2099 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2100 * or FULL-DUPLEX) of the physical link of an Ethernet device. It might need
2101 * to wait up to 9 seconds in it.
2104 * The port identifier of the Ethernet device.
2106 * A pointer to an *rte_eth_link* structure to be filled with
2107 * the status, the speed and the mode of the Ethernet device link.
2109 * - (0) if successful.
2110 * - (-ENOTSUP) if the function is not supported in PMD driver.
2111 * - (-ENODEV) if *port_id* invalid.
2113 int rte_eth_link_get(uint16_t port_id, struct rte_eth_link *link);
2116 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2117 * or FULL-DUPLEX) of the physical link of an Ethernet device. It is a no-wait
2118 * version of rte_eth_link_get().
2121 * The port identifier of the Ethernet device.
2123 * A pointer to an *rte_eth_link* structure to be filled with
2124 * the status, the speed and the mode of the Ethernet device link.
2126 * - (0) if successful.
2127 * - (-ENOTSUP) if the function is not supported in PMD driver.
2128 * - (-ENODEV) if *port_id* invalid.
2130 int rte_eth_link_get_nowait(uint16_t port_id, struct rte_eth_link *link);
2133 * Retrieve the general I/O statistics of an Ethernet device.
2136 * The port identifier of the Ethernet device.
2138 * A pointer to a structure of type *rte_eth_stats* to be filled with
2139 * the values of device counters for the following set of statistics:
2140 * - *ipackets* with the total of successfully received packets.
2141 * - *opackets* with the total of successfully transmitted packets.
2142 * - *ibytes* with the total of successfully received bytes.
2143 * - *obytes* with the total of successfully transmitted bytes.
2144 * - *ierrors* with the total of erroneous received packets.
2145 * - *oerrors* with the total of failed transmitted packets.
2147 * Zero if successful. Non-zero otherwise.
2149 int rte_eth_stats_get(uint16_t port_id, struct rte_eth_stats *stats);
2152 * Reset the general I/O statistics of an Ethernet device.
2155 * The port identifier of the Ethernet device.
2157 * - (0) if device notified to reset stats.
2158 * - (-ENOTSUP) if hardware doesn't support.
2159 * - (-ENODEV) if *port_id* invalid.
2160 * - (<0): Error code of the driver stats reset function.
2162 int rte_eth_stats_reset(uint16_t port_id);
2165 * Retrieve names of extended statistics of an Ethernet device.
2167 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2169 * xstats_names[i].name => xstats[i].value
2171 * And the array index is same with id field of 'struct rte_eth_xstat':
2174 * This assumption makes key-value pair matching less flexible but simpler.
2177 * The port identifier of the Ethernet device.
2178 * @param xstats_names
2179 * An rte_eth_xstat_name array of at least *size* elements to
2180 * be filled. If set to NULL, the function returns the required number
2183 * The size of the xstats_names array (number of elements).
2185 * - A positive value lower or equal to size: success. The return value
2186 * is the number of entries filled in the stats table.
2187 * - A positive value higher than size: error, the given statistics table
2188 * is too small. The return value corresponds to the size that should
2189 * be given to succeed. The entries in the table are not valid and
2190 * shall not be used by the caller.
2191 * - A negative value on error (invalid port id).
2193 int rte_eth_xstats_get_names(uint16_t port_id,
2194 struct rte_eth_xstat_name *xstats_names,
2198 * Retrieve extended statistics of an Ethernet device.
2200 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched
2202 * xstats_names[i].name => xstats[i].value
2204 * And the array index is same with id field of 'struct rte_eth_xstat':
2207 * This assumption makes key-value pair matching less flexible but simpler.
2210 * The port identifier of the Ethernet device.
2212 * A pointer to a table of structure of type *rte_eth_xstat*
2213 * to be filled with device statistics ids and values.
2214 * This parameter can be set to NULL if n is 0.
2216 * The size of the xstats array (number of elements).
2218 * - A positive value lower or equal to n: success. The return value
2219 * is the number of entries filled in the stats table.
2220 * - A positive value higher than n: error, the given statistics table
2221 * is too small. The return value corresponds to the size that should
2222 * be given to succeed. The entries in the table are not valid and
2223 * shall not be used by the caller.
2224 * - A negative value on error (invalid port id).
2226 int rte_eth_xstats_get(uint16_t port_id, struct rte_eth_xstat *xstats,
2230 * Retrieve names of extended statistics of an Ethernet device.
2233 * The port identifier of the Ethernet device.
2234 * @param xstats_names
2235 * An rte_eth_xstat_name array of at least *size* elements to
2236 * be filled. If set to NULL, the function returns the required number
2239 * IDs array given by app to retrieve specific statistics
2241 * The size of the xstats_names array (number of elements).
2243 * - A positive value lower or equal to size: success. The return value
2244 * is the number of entries filled in the stats table.
2245 * - A positive value higher than size: error, the given statistics table
2246 * is too small. The return value corresponds to the size that should
2247 * be given to succeed. The entries in the table are not valid and
2248 * shall not be used by the caller.
2249 * - A negative value on error (invalid port id).
2252 rte_eth_xstats_get_names_by_id(uint16_t port_id,
2253 struct rte_eth_xstat_name *xstats_names, unsigned int size,
2257 * Retrieve extended statistics of an Ethernet device.
2260 * The port identifier of the Ethernet device.
2262 * A pointer to an ids array passed by application. This tells which
2263 * statistics values function should retrieve. This parameter
2264 * can be set to NULL if size is 0. In this case function will retrieve
2265 * all available statistics.
2267 * A pointer to a table to be filled with device statistics values.
2269 * The size of the ids array (number of elements).
2271 * - A positive value lower or equal to size: success. The return value
2272 * is the number of entries filled in the stats table.
2273 * - A positive value higher than size: error, the given statistics table
2274 * is too small. The return value corresponds to the size that should
2275 * be given to succeed. The entries in the table are not valid and
2276 * shall not be used by the caller.
2277 * - A negative value on error (invalid port id).
2279 int rte_eth_xstats_get_by_id(uint16_t port_id, const uint64_t *ids,
2280 uint64_t *values, unsigned int size);
2283 * Gets the ID of a statistic from its name.
2285 * This function searches for the statistics using string compares, and
2286 * as such should not be used on the fast-path. For fast-path retrieval of
2287 * specific statistics, store the ID as provided in *id* from this function,
2288 * and pass the ID to rte_eth_xstats_get()
2290 * @param port_id The port to look up statistics from
2291 * @param xstat_name The name of the statistic to return
2292 * @param[out] id A pointer to an app-supplied uint64_t which should be
2293 * set to the ID of the stat if the stat exists.
2296 * -ENODEV for invalid port_id,
2297 * -EIO if device is removed,
2298 * -EINVAL if the xstat_name doesn't exist in port_id
2300 int rte_eth_xstats_get_id_by_name(uint16_t port_id, const char *xstat_name,
2304 * Reset extended statistics of an Ethernet device.
2307 * The port identifier of the Ethernet device.
2309 * - (0) if device notified to reset extended stats.
2310 * - (-ENOTSUP) if pmd doesn't support both
2311 * extended stats and basic stats reset.
2312 * - (-ENODEV) if *port_id* invalid.
2313 * - (<0): Error code of the driver xstats reset function.
2315 int rte_eth_xstats_reset(uint16_t port_id);
2318 * Set a mapping for the specified transmit queue to the specified per-queue
2319 * statistics counter.
2322 * The port identifier of the Ethernet device.
2323 * @param tx_queue_id
2324 * The index of the transmit queue for which a queue stats mapping is required.
2325 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2326 * to rte_eth_dev_configure().
2328 * The per-queue packet statistics functionality number that the transmit
2329 * queue is to be assigned.
2330 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
2332 * Zero if successful. Non-zero otherwise.
2334 int rte_eth_dev_set_tx_queue_stats_mapping(uint16_t port_id,
2335 uint16_t tx_queue_id, uint8_t stat_idx);
2338 * Set a mapping for the specified receive queue to the specified per-queue
2339 * statistics counter.
2342 * The port identifier of the Ethernet device.
2343 * @param rx_queue_id
2344 * The index of the receive queue for which a queue stats mapping is required.
2345 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2346 * to rte_eth_dev_configure().
2348 * The per-queue packet statistics functionality number that the receive
2349 * queue is to be assigned.
2350 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1].
2352 * Zero if successful. Non-zero otherwise.
2354 int rte_eth_dev_set_rx_queue_stats_mapping(uint16_t port_id,
2355 uint16_t rx_queue_id,
2359 * Retrieve the Ethernet address of an Ethernet device.
2362 * The port identifier of the Ethernet device.
2364 * A pointer to a structure of type *ether_addr* to be filled with
2365 * the Ethernet address of the Ethernet device.
2367 * - (0) if successful
2368 * - (-ENODEV) if *port_id* invalid.
2370 int rte_eth_macaddr_get(uint16_t port_id, struct rte_ether_addr *mac_addr);
2373 * Retrieve the contextual information of an Ethernet device.
2375 * As part of this function, a number of of fields in dev_info will be
2376 * initialized as follows:
2381 * Where lim is defined within the rte_eth_dev_info_get as
2383 * const struct rte_eth_desc_lim lim = {
2384 * .nb_max = UINT16_MAX,
2387 * .nb_seg_max = UINT16_MAX,
2388 * .nb_mtu_seg_max = UINT16_MAX,
2391 * device = dev->device
2392 * min_mtu = RTE_ETHER_MIN_MTU
2393 * max_mtu = UINT16_MAX
2395 * The following fields will be populated if support for dev_infos_get()
2396 * exists for the device and the rte_eth_dev 'dev' has been populated
2397 * successfully with a call to it:
2399 * driver_name = dev->device->driver->name
2400 * nb_rx_queues = dev->data->nb_rx_queues
2401 * nb_tx_queues = dev->data->nb_tx_queues
2402 * dev_flags = &dev->data->dev_flags
2405 * The port identifier of the Ethernet device.
2407 * A pointer to a structure of type *rte_eth_dev_info* to be filled with
2408 * the contextual information of the Ethernet device.
2410 * - (0) if successful.
2411 * - (-ENOTSUP) if support for dev_infos_get() does not exist for the device.
2412 * - (-ENODEV) if *port_id* invalid.
2414 int rte_eth_dev_info_get(uint16_t port_id, struct rte_eth_dev_info *dev_info);
2417 * Retrieve the firmware version of a device.
2420 * The port identifier of the device.
2422 * A pointer to a string array storing the firmware version of a device,
2423 * the string includes terminating null. This pointer is allocated by caller.
2425 * The size of the string array pointed by fw_version, which should be
2426 * large enough to store firmware version of the device.
2428 * - (0) if successful.
2429 * - (-ENOTSUP) if operation is not supported.
2430 * - (-ENODEV) if *port_id* invalid.
2431 * - (-EIO) if device is removed.
2432 * - (>0) if *fw_size* is not enough to store firmware version, return
2433 * the size of the non truncated string.
2435 int rte_eth_dev_fw_version_get(uint16_t port_id,
2436 char *fw_version, size_t fw_size);
2439 * Retrieve the supported packet types of an Ethernet device.
2441 * When a packet type is announced as supported, it *must* be recognized by
2442 * the PMD. For instance, if RTE_PTYPE_L2_ETHER, RTE_PTYPE_L2_ETHER_VLAN
2443 * and RTE_PTYPE_L3_IPV4 are announced, the PMD must return the following
2444 * packet types for these packets:
2445 * - Ether/IPv4 -> RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4
2446 * - Ether/Vlan/IPv4 -> RTE_PTYPE_L2_ETHER_VLAN | RTE_PTYPE_L3_IPV4
2447 * - Ether/[anything else] -> RTE_PTYPE_L2_ETHER
2448 * - Ether/Vlan/[anything else] -> RTE_PTYPE_L2_ETHER_VLAN
2450 * When a packet is received by a PMD, the most precise type must be
2451 * returned among the ones supported. However a PMD is allowed to set
2452 * packet type that is not in the supported list, at the condition that it
2453 * is more precise. Therefore, a PMD announcing no supported packet types
2454 * can still set a matching packet type in a received packet.
2457 * Better to invoke this API after the device is already started or rx burst
2458 * function is decided, to obtain correct supported ptypes.
2460 * if a given PMD does not report what ptypes it supports, then the supported
2461 * ptype count is reported as 0.
2463 * The port identifier of the Ethernet device.
2465 * A hint of what kind of packet type which the caller is interested in.
2467 * An array pointer to store adequate packet types, allocated by caller.
2469 * Size of the array pointed by param ptypes.
2471 * - (>=0) Number of supported ptypes. If the number of types exceeds num,
2472 * only num entries will be filled into the ptypes array, but the full
2473 * count of supported ptypes will be returned.
2474 * - (-ENODEV) if *port_id* invalid.
2476 int rte_eth_dev_get_supported_ptypes(uint16_t port_id, uint32_t ptype_mask,
2477 uint32_t *ptypes, int num);
2480 * Retrieve the MTU of an Ethernet device.
2483 * The port identifier of the Ethernet device.
2485 * A pointer to a uint16_t where the retrieved MTU is to be stored.
2487 * - (0) if successful.
2488 * - (-ENODEV) if *port_id* invalid.
2490 int rte_eth_dev_get_mtu(uint16_t port_id, uint16_t *mtu);
2493 * Change the MTU of an Ethernet device.
2496 * The port identifier of the Ethernet device.
2498 * A uint16_t for the MTU to be applied.
2500 * - (0) if successful.
2501 * - (-ENOTSUP) if operation is not supported.
2502 * - (-ENODEV) if *port_id* invalid.
2503 * - (-EIO) if device is removed.
2504 * - (-EINVAL) if *mtu* invalid, validation of mtu can occur within
2505 * rte_eth_dev_set_mtu if dev_infos_get is supported by the device or
2506 * when the mtu is set using dev->dev_ops->mtu_set.
2507 * - (-EBUSY) if operation is not allowed when the port is running
2509 int rte_eth_dev_set_mtu(uint16_t port_id, uint16_t mtu);
2512 * Enable/Disable hardware filtering by an Ethernet device of received
2513 * VLAN packets tagged with a given VLAN Tag Identifier.
2516 * The port identifier of the Ethernet device.
2518 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
2520 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*.
2521 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*.
2523 * - (0) if successful.
2524 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
2525 * - (-ENODEV) if *port_id* invalid.
2526 * - (-EIO) if device is removed.
2527 * - (-ENOSYS) if VLAN filtering on *port_id* disabled.
2528 * - (-EINVAL) if *vlan_id* > 4095.
2530 int rte_eth_dev_vlan_filter(uint16_t port_id, uint16_t vlan_id, int on);
2533 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device.
2534 * 82599/X540/X550 can support VLAN stripping at the rx queue level
2537 * The port identifier of the Ethernet device.
2538 * @param rx_queue_id
2539 * The index of the receive queue for which a queue stats mapping is required.
2540 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2541 * to rte_eth_dev_configure().
2543 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port.
2544 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port.
2546 * - (0) if successful.
2547 * - (-ENOTSUP) if hardware-assisted VLAN stripping not configured.
2548 * - (-ENODEV) if *port_id* invalid.
2549 * - (-EINVAL) if *rx_queue_id* invalid.
2551 int rte_eth_dev_set_vlan_strip_on_queue(uint16_t port_id, uint16_t rx_queue_id,
2555 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to
2556 * the VLAN Header. This is a register setup available on some Intel NIC, not
2557 * but all, please check the data sheet for availability.
2560 * The port identifier of the Ethernet device.
2564 * The Tag Protocol ID
2566 * - (0) if successful.
2567 * - (-ENOTSUP) if hardware-assisted VLAN TPID setup is not supported.
2568 * - (-ENODEV) if *port_id* invalid.
2569 * - (-EIO) if device is removed.
2571 int rte_eth_dev_set_vlan_ether_type(uint16_t port_id,
2572 enum rte_vlan_type vlan_type,
2576 * Set VLAN offload configuration on an Ethernet device
2577 * Enable/Disable Extended VLAN by an Ethernet device, This is a register setup
2578 * available on some Intel NIC, not but all, please check the data sheet for
2580 * Enable/Disable VLAN Strip can be done on rx queue for certain NIC, but here
2581 * the configuration is applied on the port level.
2584 * The port identifier of the Ethernet device.
2585 * @param offload_mask
2586 * The VLAN Offload bit mask can be mixed use with "OR"
2587 * ETH_VLAN_STRIP_OFFLOAD
2588 * ETH_VLAN_FILTER_OFFLOAD
2589 * ETH_VLAN_EXTEND_OFFLOAD
2590 * ETH_QINQ_STRIP_OFFLOAD
2592 * - (0) if successful.
2593 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured.
2594 * - (-ENODEV) if *port_id* invalid.
2595 * - (-EIO) if device is removed.
2597 int rte_eth_dev_set_vlan_offload(uint16_t port_id, int offload_mask);
2600 * Read VLAN Offload configuration from an Ethernet device
2603 * The port identifier of the Ethernet device.
2605 * - (>0) if successful. Bit mask to indicate
2606 * ETH_VLAN_STRIP_OFFLOAD
2607 * ETH_VLAN_FILTER_OFFLOAD
2608 * ETH_VLAN_EXTEND_OFFLOAD
2609 * ETH_QINQ_STRIP_OFFLOAD
2610 * - (-ENODEV) if *port_id* invalid.
2612 int rte_eth_dev_get_vlan_offload(uint16_t port_id);
2615 * Set port based TX VLAN insertion on or off.
2618 * The port identifier of the Ethernet device.
2620 * Port based TX VLAN identifier together with user priority.
2622 * Turn on or off the port based TX VLAN insertion.
2625 * - (0) if successful.
2626 * - negative if failed.
2628 int rte_eth_dev_set_vlan_pvid(uint16_t port_id, uint16_t pvid, int on);
2630 typedef void (*buffer_tx_error_fn)(struct rte_mbuf **unsent, uint16_t count,
2634 * Structure used to buffer packets for future TX
2635 * Used by APIs rte_eth_tx_buffer and rte_eth_tx_buffer_flush
2637 struct rte_eth_dev_tx_buffer {
2638 buffer_tx_error_fn error_callback;
2639 void *error_userdata;
2640 uint16_t size; /**< Size of buffer for buffered tx */
2641 uint16_t length; /**< Number of packets in the array */
2642 struct rte_mbuf *pkts[];
2643 /**< Pending packets to be sent on explicit flush or when full */
2647 * Calculate the size of the tx buffer.
2650 * Number of stored packets.
2652 #define RTE_ETH_TX_BUFFER_SIZE(sz) \
2653 (sizeof(struct rte_eth_dev_tx_buffer) + (sz) * sizeof(struct rte_mbuf *))
2656 * Initialize default values for buffered transmitting
2659 * Tx buffer to be initialized.
2666 rte_eth_tx_buffer_init(struct rte_eth_dev_tx_buffer *buffer, uint16_t size);
2669 * Configure a callback for buffered packets which cannot be sent
2671 * Register a specific callback to be called when an attempt is made to send
2672 * all packets buffered on an ethernet port, but not all packets can
2673 * successfully be sent. The callback registered here will be called only
2674 * from calls to rte_eth_tx_buffer() and rte_eth_tx_buffer_flush() APIs.
2675 * The default callback configured for each queue by default just frees the
2676 * packets back to the calling mempool. If additional behaviour is required,
2677 * for example, to count dropped packets, or to retry transmission of packets
2678 * which cannot be sent, this function should be used to register a suitable
2679 * callback function to implement the desired behaviour.
2680 * The example callback "rte_eth_count_unsent_packet_callback()" is also
2681 * provided as reference.
2684 * The port identifier of the Ethernet device.
2686 * The function to be used as the callback.
2688 * Arbitrary parameter to be passed to the callback function
2690 * 0 on success, or -1 on error with rte_errno set appropriately
2693 rte_eth_tx_buffer_set_err_callback(struct rte_eth_dev_tx_buffer *buffer,
2694 buffer_tx_error_fn callback, void *userdata);
2697 * Callback function for silently dropping unsent buffered packets.
2699 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
2700 * adjust the default behavior when buffered packets cannot be sent. This
2701 * function drops any unsent packets silently and is used by tx buffered
2702 * operations as default behavior.
2704 * NOTE: this function should not be called directly, instead it should be used
2705 * as a callback for packet buffering.
2707 * NOTE: when configuring this function as a callback with
2708 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
2709 * should point to an uint64_t value.
2712 * The previously buffered packets which could not be sent
2714 * The number of unsent packets in the pkts array
2719 rte_eth_tx_buffer_drop_callback(struct rte_mbuf **pkts, uint16_t unsent,
2723 * Callback function for tracking unsent buffered packets.
2725 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
2726 * adjust the default behavior when buffered packets cannot be sent. This
2727 * function drops any unsent packets, but also updates a user-supplied counter
2728 * to track the overall number of packets dropped. The counter should be an
2729 * uint64_t variable.
2731 * NOTE: this function should not be called directly, instead it should be used
2732 * as a callback for packet buffering.
2734 * NOTE: when configuring this function as a callback with
2735 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
2736 * should point to an uint64_t value.
2739 * The previously buffered packets which could not be sent
2741 * The number of unsent packets in the pkts array
2743 * Pointer to an uint64_t value, which will be incremented by unsent
2746 rte_eth_tx_buffer_count_callback(struct rte_mbuf **pkts, uint16_t unsent,
2750 * Request the driver to free mbufs currently cached by the driver. The
2751 * driver will only free the mbuf if it is no longer in use. It is the
2752 * application's responsibility to ensure rte_eth_tx_buffer_flush(..) is
2756 * The port identifier of the Ethernet device.
2758 * The index of the transmit queue through which output packets must be
2760 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2761 * to rte_eth_dev_configure().
2763 * Maximum number of packets to free. Use 0 to indicate all possible packets
2764 * should be freed. Note that a packet may be using multiple mbufs.
2767 * -ENODEV: Invalid interface
2768 * -EIO: device is removed
2769 * -ENOTSUP: Driver does not support function
2771 * 0-n: Number of packets freed. More packets may still remain in ring that
2775 rte_eth_tx_done_cleanup(uint16_t port_id, uint16_t queue_id, uint32_t free_cnt);
2778 * Subtypes for IPsec offload event(@ref RTE_ETH_EVENT_IPSEC) raised by
2781 enum rte_eth_event_ipsec_subtype {
2782 RTE_ETH_EVENT_IPSEC_UNKNOWN = 0,
2783 /**< Unknown event type */
2784 RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW,
2785 /**< Sequence number overflow */
2786 RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY,
2787 /**< Soft time expiry of SA */
2788 RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY,
2789 /**< Soft byte expiry of SA */
2790 RTE_ETH_EVENT_IPSEC_MAX
2791 /**< Max value of this enum */
2795 * Descriptor for @ref RTE_ETH_EVENT_IPSEC event. Used by eth dev to send extra
2796 * information of the IPsec offload event.
2798 struct rte_eth_event_ipsec_desc {
2799 enum rte_eth_event_ipsec_subtype subtype;
2800 /**< Type of RTE_ETH_EVENT_IPSEC_* event */
2802 /**< Event specific metadata
2804 * For the following events, *userdata* registered
2805 * with the *rte_security_session* would be returned
2808 * - @ref RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW
2809 * - @ref RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY
2810 * - @ref RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY
2812 * @see struct rte_security_session_conf
2818 * The eth device event type for interrupt, and maybe others in the future.
2820 enum rte_eth_event_type {
2821 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */
2822 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */
2823 RTE_ETH_EVENT_QUEUE_STATE,
2824 /**< queue state event (enabled/disabled) */
2825 RTE_ETH_EVENT_INTR_RESET,
2826 /**< reset interrupt event, sent to VF on PF reset */
2827 RTE_ETH_EVENT_VF_MBOX, /**< message from the VF received by PF */
2828 RTE_ETH_EVENT_MACSEC, /**< MACsec offload related event */
2829 RTE_ETH_EVENT_INTR_RMV, /**< device removal event */
2830 RTE_ETH_EVENT_NEW, /**< port is probed */
2831 RTE_ETH_EVENT_DESTROY, /**< port is released */
2832 RTE_ETH_EVENT_IPSEC, /**< IPsec offload related event */
2833 RTE_ETH_EVENT_MAX /**< max value of this enum */
2836 typedef int (*rte_eth_dev_cb_fn)(uint16_t port_id,
2837 enum rte_eth_event_type event, void *cb_arg, void *ret_param);
2838 /**< user application callback to be registered for interrupts */
2841 * Register a callback function for port event.
2845 * RTE_ETH_ALL means register the event for all port ids.
2849 * User supplied callback function to be called.
2851 * Pointer to the parameters for the registered callback.
2854 * - On success, zero.
2855 * - On failure, a negative value.
2857 int rte_eth_dev_callback_register(uint16_t port_id,
2858 enum rte_eth_event_type event,
2859 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
2862 * Unregister a callback function for port event.
2866 * RTE_ETH_ALL means unregister the event for all port ids.
2870 * User supplied callback function to be called.
2872 * Pointer to the parameters for the registered callback. -1 means to
2873 * remove all for the same callback address and same event.
2876 * - On success, zero.
2877 * - On failure, a negative value.
2879 int rte_eth_dev_callback_unregister(uint16_t port_id,
2880 enum rte_eth_event_type event,
2881 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
2884 * When there is no rx packet coming in Rx Queue for a long time, we can
2885 * sleep lcore related to RX Queue for power saving, and enable rx interrupt
2886 * to be triggered when Rx packet arrives.
2888 * The rte_eth_dev_rx_intr_enable() function enables rx queue
2889 * interrupt on specific rx queue of a port.
2892 * The port identifier of the Ethernet device.
2894 * The index of the receive queue from which to retrieve input packets.
2895 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2896 * to rte_eth_dev_configure().
2898 * - (0) if successful.
2899 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
2901 * - (-ENODEV) if *port_id* invalid.
2902 * - (-EIO) if device is removed.
2904 int rte_eth_dev_rx_intr_enable(uint16_t port_id, uint16_t queue_id);
2907 * When lcore wakes up from rx interrupt indicating packet coming, disable rx
2908 * interrupt and returns to polling mode.
2910 * The rte_eth_dev_rx_intr_disable() function disables rx queue
2911 * interrupt on specific rx queue of a port.
2914 * The port identifier of the Ethernet device.
2916 * The index of the receive queue from which to retrieve input packets.
2917 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2918 * to rte_eth_dev_configure().
2920 * - (0) if successful.
2921 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
2923 * - (-ENODEV) if *port_id* invalid.
2924 * - (-EIO) if device is removed.
2926 int rte_eth_dev_rx_intr_disable(uint16_t port_id, uint16_t queue_id);
2929 * RX Interrupt control per port.
2932 * The port identifier of the Ethernet device.
2934 * Epoll instance fd which the intr vector associated to.
2935 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
2937 * The operation be performed for the vector.
2938 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
2942 * - On success, zero.
2943 * - On failure, a negative value.
2945 int rte_eth_dev_rx_intr_ctl(uint16_t port_id, int epfd, int op, void *data);
2948 * RX Interrupt control per queue.
2951 * The port identifier of the Ethernet device.
2953 * The index of the receive queue from which to retrieve input packets.
2954 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2955 * to rte_eth_dev_configure().
2957 * Epoll instance fd which the intr vector associated to.
2958 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
2960 * The operation be performed for the vector.
2961 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
2965 * - On success, zero.
2966 * - On failure, a negative value.
2968 int rte_eth_dev_rx_intr_ctl_q(uint16_t port_id, uint16_t queue_id,
2969 int epfd, int op, void *data);
2973 * @b EXPERIMENTAL: this API may change without prior notice.
2975 * Get interrupt fd per Rx queue.
2978 * The port identifier of the Ethernet device.
2980 * The index of the receive queue from which to retrieve input packets.
2981 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2982 * to rte_eth_dev_configure().
2984 * - (>=0) the interrupt fd associated to the requested Rx queue if
2990 rte_eth_dev_rx_intr_ctl_q_get_fd(uint16_t port_id, uint16_t queue_id);
2993 * Turn on the LED on the Ethernet device.
2994 * This function turns on the LED on the Ethernet device.
2997 * The port identifier of the Ethernet device.
2999 * - (0) if successful.
3000 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3002 * - (-ENODEV) if *port_id* invalid.
3003 * - (-EIO) if device is removed.
3005 int rte_eth_led_on(uint16_t port_id);
3008 * Turn off the LED on the Ethernet device.
3009 * This function turns off the LED on the Ethernet device.
3012 * The port identifier of the Ethernet device.
3014 * - (0) if successful.
3015 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3017 * - (-ENODEV) if *port_id* invalid.
3018 * - (-EIO) if device is removed.
3020 int rte_eth_led_off(uint16_t port_id);
3023 * Get current status of the Ethernet link flow control for Ethernet device
3026 * The port identifier of the Ethernet device.
3028 * The pointer to the structure where to store the flow control parameters.
3030 * - (0) if successful.
3031 * - (-ENOTSUP) if hardware doesn't support flow control.
3032 * - (-ENODEV) if *port_id* invalid.
3033 * - (-EIO) if device is removed.
3035 int rte_eth_dev_flow_ctrl_get(uint16_t port_id,
3036 struct rte_eth_fc_conf *fc_conf);
3039 * Configure the Ethernet link flow control for Ethernet device
3042 * The port identifier of the Ethernet device.
3044 * The pointer to the structure of the flow control parameters.
3046 * - (0) if successful.
3047 * - (-ENOTSUP) if hardware doesn't support flow control mode.
3048 * - (-ENODEV) if *port_id* invalid.
3049 * - (-EINVAL) if bad parameter
3050 * - (-EIO) if flow control setup failure or device is removed.
3052 int rte_eth_dev_flow_ctrl_set(uint16_t port_id,
3053 struct rte_eth_fc_conf *fc_conf);
3056 * Configure the Ethernet priority flow control under DCB environment
3057 * for Ethernet device.
3060 * The port identifier of the Ethernet device.
3062 * The pointer to the structure of the priority flow control parameters.
3064 * - (0) if successful.
3065 * - (-ENOTSUP) if hardware doesn't support priority flow control mode.
3066 * - (-ENODEV) if *port_id* invalid.
3067 * - (-EINVAL) if bad parameter
3068 * - (-EIO) if flow control setup failure or device is removed.
3070 int rte_eth_dev_priority_flow_ctrl_set(uint16_t port_id,
3071 struct rte_eth_pfc_conf *pfc_conf);
3074 * Add a MAC address to an internal array of addresses used to enable whitelist
3075 * filtering to accept packets only if the destination MAC address matches.
3078 * The port identifier of the Ethernet device.
3080 * The MAC address to add.
3082 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is
3083 * not enabled, this should be set to 0.
3085 * - (0) if successfully added or *mac_addr* was already added.
3086 * - (-ENOTSUP) if hardware doesn't support this feature.
3087 * - (-ENODEV) if *port* is invalid.
3088 * - (-EIO) if device is removed.
3089 * - (-ENOSPC) if no more MAC addresses can be added.
3090 * - (-EINVAL) if MAC address is invalid.
3092 int rte_eth_dev_mac_addr_add(uint16_t port_id, struct rte_ether_addr *mac_addr,
3096 * Remove a MAC address from the internal array of addresses.
3099 * The port identifier of the Ethernet device.
3101 * MAC address to remove.
3103 * - (0) if successful, or *mac_addr* didn't exist.
3104 * - (-ENOTSUP) if hardware doesn't support.
3105 * - (-ENODEV) if *port* invalid.
3106 * - (-EADDRINUSE) if attempting to remove the default MAC address
3108 int rte_eth_dev_mac_addr_remove(uint16_t port_id,
3109 struct rte_ether_addr *mac_addr);
3112 * Set the default MAC address.
3115 * The port identifier of the Ethernet device.
3117 * New default MAC address.
3119 * - (0) if successful, or *mac_addr* didn't exist.
3120 * - (-ENOTSUP) if hardware doesn't support.
3121 * - (-ENODEV) if *port* invalid.
3122 * - (-EINVAL) if MAC address is invalid.
3124 int rte_eth_dev_default_mac_addr_set(uint16_t port_id,
3125 struct rte_ether_addr *mac_addr);
3128 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3131 * The port identifier of the Ethernet device.
3135 * Redirection table size. The table size can be queried by
3136 * rte_eth_dev_info_get().
3138 * - (0) if successful.
3139 * - (-ENOTSUP) if hardware doesn't support.
3140 * - (-EINVAL) if bad parameter.
3141 * - (-EIO) if device is removed.
3143 int rte_eth_dev_rss_reta_update(uint16_t port_id,
3144 struct rte_eth_rss_reta_entry64 *reta_conf,
3145 uint16_t reta_size);
3148 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3151 * The port identifier of the Ethernet device.
3153 * RETA to query. For each requested reta entry, corresponding bit
3154 * in mask must be set.
3156 * Redirection table size. The table size can be queried by
3157 * rte_eth_dev_info_get().
3159 * - (0) if successful.
3160 * - (-ENOTSUP) if hardware doesn't support.
3161 * - (-EINVAL) if bad parameter.
3162 * - (-EIO) if device is removed.
3164 int rte_eth_dev_rss_reta_query(uint16_t port_id,
3165 struct rte_eth_rss_reta_entry64 *reta_conf,
3166 uint16_t reta_size);
3169 * Updates unicast hash table for receiving packet with the given destination
3170 * MAC address, and the packet is routed to all VFs for which the RX mode is
3171 * accept packets that match the unicast hash table.
3174 * The port identifier of the Ethernet device.
3176 * Unicast MAC address.
3178 * 1 - Set an unicast hash bit for receiving packets with the MAC address.
3179 * 0 - Clear an unicast hash bit.
3181 * - (0) if successful.
3182 * - (-ENOTSUP) if hardware doesn't support.
3183 * - (-ENODEV) if *port_id* invalid.
3184 * - (-EIO) if device is removed.
3185 * - (-EINVAL) if bad parameter.
3187 int rte_eth_dev_uc_hash_table_set(uint16_t port_id, struct rte_ether_addr *addr,
3191 * Updates all unicast hash bitmaps for receiving packet with any Unicast
3192 * Ethernet MAC addresses,the packet is routed to all VFs for which the RX
3193 * mode is accept packets that match the unicast hash table.
3196 * The port identifier of the Ethernet device.
3198 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet
3200 * 0 - Clear all unicast hash bitmaps
3202 * - (0) if successful.
3203 * - (-ENOTSUP) if hardware doesn't support.
3204 * - (-ENODEV) if *port_id* invalid.
3205 * - (-EIO) if device is removed.
3206 * - (-EINVAL) if bad parameter.
3208 int rte_eth_dev_uc_all_hash_table_set(uint16_t port_id, uint8_t on);
3211 * Set a traffic mirroring rule on an Ethernet device
3214 * The port identifier of the Ethernet device.
3215 * @param mirror_conf
3216 * The pointer to the traffic mirroring structure describing the mirroring rule.
3217 * The *rte_eth_vm_mirror_conf* structure includes the type of mirroring rule,
3218 * destination pool and the value of rule if enable vlan or pool mirroring.
3221 * The index of traffic mirroring rule, we support four separated rules.
3223 * 1 - Enable a mirroring rule.
3224 * 0 - Disable a mirroring rule.
3226 * - (0) if successful.
3227 * - (-ENOTSUP) if hardware doesn't support this feature.
3228 * - (-ENODEV) if *port_id* invalid.
3229 * - (-EIO) if device is removed.
3230 * - (-EINVAL) if the mr_conf information is not correct.
3232 int rte_eth_mirror_rule_set(uint16_t port_id,
3233 struct rte_eth_mirror_conf *mirror_conf,
3238 * Reset a traffic mirroring rule on an Ethernet device.
3241 * The port identifier of the Ethernet device.
3243 * The index of traffic mirroring rule, we support four separated rules.
3245 * - (0) if successful.
3246 * - (-ENOTSUP) if hardware doesn't support this feature.
3247 * - (-ENODEV) if *port_id* invalid.
3248 * - (-EIO) if device is removed.
3249 * - (-EINVAL) if bad parameter.
3251 int rte_eth_mirror_rule_reset(uint16_t port_id,
3255 * Set the rate limitation for a queue on an Ethernet device.
3258 * The port identifier of the Ethernet device.
3262 * The tx rate in Mbps. Allocated from the total port link speed.
3264 * - (0) if successful.
3265 * - (-ENOTSUP) if hardware doesn't support this feature.
3266 * - (-ENODEV) if *port_id* invalid.
3267 * - (-EIO) if device is removed.
3268 * - (-EINVAL) if bad parameter.
3270 int rte_eth_set_queue_rate_limit(uint16_t port_id, uint16_t queue_idx,
3274 * Configuration of Receive Side Scaling hash computation of Ethernet device.
3277 * The port identifier of the Ethernet device.
3279 * The new configuration to use for RSS hash computation on the port.
3281 * - (0) if successful.
3282 * - (-ENODEV) if port identifier is invalid.
3283 * - (-EIO) if device is removed.
3284 * - (-ENOTSUP) if hardware doesn't support.
3285 * - (-EINVAL) if bad parameter.
3287 int rte_eth_dev_rss_hash_update(uint16_t port_id,
3288 struct rte_eth_rss_conf *rss_conf);
3291 * Retrieve current configuration of Receive Side Scaling hash computation
3292 * of Ethernet device.
3295 * The port identifier of the Ethernet device.
3297 * Where to store the current RSS hash configuration of the Ethernet device.
3299 * - (0) if successful.
3300 * - (-ENODEV) if port identifier is invalid.
3301 * - (-EIO) if device is removed.
3302 * - (-ENOTSUP) if hardware doesn't support RSS.
3305 rte_eth_dev_rss_hash_conf_get(uint16_t port_id,
3306 struct rte_eth_rss_conf *rss_conf);
3309 * Add UDP tunneling port for a specific type of tunnel.
3310 * The packets with this UDP port will be identified as this type of tunnel.
3311 * Before enabling any offloading function for a tunnel, users can call this API
3312 * to change or add more UDP port for the tunnel. So the offloading function
3313 * can take effect on the packets with the specific UDP port.
3316 * The port identifier of the Ethernet device.
3318 * UDP tunneling configuration.
3321 * - (0) if successful.
3322 * - (-ENODEV) if port identifier is invalid.
3323 * - (-EIO) if device is removed.
3324 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3327 rte_eth_dev_udp_tunnel_port_add(uint16_t port_id,
3328 struct rte_eth_udp_tunnel *tunnel_udp);
3331 * Delete UDP tunneling port a specific type of tunnel.
3332 * The packets with this UDP port will not be identified as this type of tunnel
3334 * Before enabling any offloading function for a tunnel, users can call this API
3335 * to delete a UDP port for the tunnel. So the offloading function will not take
3336 * effect on the packets with the specific UDP port.
3339 * The port identifier of the Ethernet device.
3341 * UDP tunneling configuration.
3344 * - (0) if successful.
3345 * - (-ENODEV) if port identifier is invalid.
3346 * - (-EIO) if device is removed.
3347 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3350 rte_eth_dev_udp_tunnel_port_delete(uint16_t port_id,
3351 struct rte_eth_udp_tunnel *tunnel_udp);
3354 * Check whether the filter type is supported on an Ethernet device.
3355 * All the supported filter types are defined in 'rte_eth_ctrl.h'.
3358 * The port identifier of the Ethernet device.
3359 * @param filter_type
3362 * - (0) if successful.
3363 * - (-ENOTSUP) if hardware doesn't support this filter type.
3364 * - (-ENODEV) if *port_id* invalid.
3365 * - (-EIO) if device is removed.
3368 int rte_eth_dev_filter_supported(uint16_t port_id,
3369 enum rte_filter_type filter_type);
3372 * Take operations to assigned filter type on an Ethernet device.
3373 * All the supported operations and filter types are defined in 'rte_eth_ctrl.h'.
3376 * The port identifier of the Ethernet device.
3377 * @param filter_type
3380 * Type of operation.
3382 * A pointer to arguments defined specifically for the operation.
3384 * - (0) if successful.
3385 * - (-ENOTSUP) if hardware doesn't support.
3386 * - (-ENODEV) if *port_id* invalid.
3387 * - (-EIO) if device is removed.
3388 * - others depends on the specific operations implementation.
3391 int rte_eth_dev_filter_ctrl(uint16_t port_id, enum rte_filter_type filter_type,
3392 enum rte_filter_op filter_op, void *arg);
3395 * Get DCB information on an Ethernet device.
3398 * The port identifier of the Ethernet device.
3402 * - (0) if successful.
3403 * - (-ENODEV) if port identifier is invalid.
3404 * - (-EIO) if device is removed.
3405 * - (-ENOTSUP) if hardware doesn't support.
3407 int rte_eth_dev_get_dcb_info(uint16_t port_id,
3408 struct rte_eth_dcb_info *dcb_info);
3410 struct rte_eth_rxtx_callback;
3413 * Add a callback to be called on packet RX on a given port and queue.
3415 * This API configures a function to be called for each burst of
3416 * packets received on a given NIC port queue. The return value is a pointer
3417 * that can be used to later remove the callback using
3418 * rte_eth_remove_rx_callback().
3420 * Multiple functions are called in the order that they are added.
3423 * The port identifier of the Ethernet device.
3425 * The queue on the Ethernet device on which the callback is to be added.
3427 * The callback function
3429 * A generic pointer parameter which will be passed to each invocation of the
3430 * callback function on this port and queue.
3434 * On success, a pointer value which can later be used to remove the callback.
3436 const struct rte_eth_rxtx_callback *
3437 rte_eth_add_rx_callback(uint16_t port_id, uint16_t queue_id,
3438 rte_rx_callback_fn fn, void *user_param);
3441 * Add a callback that must be called first on packet RX on a given port
3444 * This API configures a first function to be called for each burst of
3445 * packets received on a given NIC port queue. The return value is a pointer
3446 * that can be used to later remove the callback using
3447 * rte_eth_remove_rx_callback().
3449 * Multiple functions are called in the order that they are added.
3452 * The port identifier of the Ethernet device.
3454 * The queue on the Ethernet device on which the callback is to be added.
3456 * The callback function
3458 * A generic pointer parameter which will be passed to each invocation of the
3459 * callback function on this port and queue.
3463 * On success, a pointer value which can later be used to remove the callback.
3465 const struct rte_eth_rxtx_callback *
3466 rte_eth_add_first_rx_callback(uint16_t port_id, uint16_t queue_id,
3467 rte_rx_callback_fn fn, void *user_param);
3470 * Add a callback to be called on packet TX on a given port and queue.
3472 * This API configures a function to be called for each burst of
3473 * packets sent on a given NIC port queue. The return value is a pointer
3474 * that can be used to later remove the callback using
3475 * rte_eth_remove_tx_callback().
3477 * Multiple functions are called in the order that they are added.
3480 * The port identifier of the Ethernet device.
3482 * The queue on the Ethernet device on which the callback is to be added.
3484 * The callback function
3486 * A generic pointer parameter which will be passed to each invocation of the
3487 * callback function on this port and queue.
3491 * On success, a pointer value which can later be used to remove the callback.
3493 const struct rte_eth_rxtx_callback *
3494 rte_eth_add_tx_callback(uint16_t port_id, uint16_t queue_id,
3495 rte_tx_callback_fn fn, void *user_param);
3498 * Remove an RX packet callback from a given port and queue.
3500 * This function is used to removed callbacks that were added to a NIC port
3501 * queue using rte_eth_add_rx_callback().
3503 * Note: the callback is removed from the callback list but it isn't freed
3504 * since the it may still be in use. The memory for the callback can be
3505 * subsequently freed back by the application by calling rte_free():
3507 * - Immediately - if the port is stopped, or the user knows that no
3508 * callbacks are in flight e.g. if called from the thread doing RX/TX
3511 * - After a short delay - where the delay is sufficient to allow any
3512 * in-flight callbacks to complete.
3515 * The port identifier of the Ethernet device.
3517 * The queue on the Ethernet device from which the callback is to be removed.
3519 * User supplied callback created via rte_eth_add_rx_callback().
3522 * - 0: Success. Callback was removed.
3523 * - -ENOTSUP: Callback support is not available.
3524 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
3525 * is NULL or not found for the port/queue.
3527 int rte_eth_remove_rx_callback(uint16_t port_id, uint16_t queue_id,
3528 const struct rte_eth_rxtx_callback *user_cb);
3531 * Remove a TX packet callback from a given port and queue.
3533 * This function is used to removed callbacks that were added to a NIC port
3534 * queue using rte_eth_add_tx_callback().
3536 * Note: the callback is removed from the callback list but it isn't freed
3537 * since the it may still be in use. The memory for the callback can be
3538 * subsequently freed back by the application by calling rte_free():
3540 * - Immediately - if the port is stopped, or the user knows that no
3541 * callbacks are in flight e.g. if called from the thread doing RX/TX
3544 * - After a short delay - where the delay is sufficient to allow any
3545 * in-flight callbacks to complete.
3548 * The port identifier of the Ethernet device.
3550 * The queue on the Ethernet device from which the callback is to be removed.
3552 * User supplied callback created via rte_eth_add_tx_callback().
3555 * - 0: Success. Callback was removed.
3556 * - -ENOTSUP: Callback support is not available.
3557 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
3558 * is NULL or not found for the port/queue.
3560 int rte_eth_remove_tx_callback(uint16_t port_id, uint16_t queue_id,
3561 const struct rte_eth_rxtx_callback *user_cb);
3564 * Retrieve information about given port's RX queue.
3567 * The port identifier of the Ethernet device.
3569 * The RX queue on the Ethernet device for which information
3570 * will be retrieved.
3572 * A pointer to a structure of type *rte_eth_rxq_info_info* to be filled with
3573 * the information of the Ethernet device.
3577 * - -ENOTSUP: routine is not supported by the device PMD.
3578 * - -EINVAL: The port_id or the queue_id is out of range.
3580 int rte_eth_rx_queue_info_get(uint16_t port_id, uint16_t queue_id,
3581 struct rte_eth_rxq_info *qinfo);
3584 * Retrieve information about given port's TX queue.
3587 * The port identifier of the Ethernet device.
3589 * The TX queue on the Ethernet device for which information
3590 * will be retrieved.
3592 * A pointer to a structure of type *rte_eth_txq_info_info* to be filled with
3593 * the information of the Ethernet device.
3597 * - -ENOTSUP: routine is not supported by the device PMD.
3598 * - -EINVAL: The port_id or the queue_id is out of range.
3600 int rte_eth_tx_queue_info_get(uint16_t port_id, uint16_t queue_id,
3601 struct rte_eth_txq_info *qinfo);
3604 * Retrieve device registers and register attributes (number of registers and
3608 * The port identifier of the Ethernet device.
3610 * Pointer to rte_dev_reg_info structure to fill in. If info->data is
3611 * NULL the function fills in the width and length fields. If non-NULL
3612 * the registers are put into the buffer pointed at by the data field.
3614 * - (0) if successful.
3615 * - (-ENOTSUP) if hardware doesn't support.
3616 * - (-ENODEV) if *port_id* invalid.
3617 * - (-EIO) if device is removed.
3618 * - others depends on the specific operations implementation.
3620 int rte_eth_dev_get_reg_info(uint16_t port_id, struct rte_dev_reg_info *info);
3623 * Retrieve size of device EEPROM
3626 * The port identifier of the Ethernet device.
3628 * - (>=0) EEPROM size if successful.
3629 * - (-ENOTSUP) if hardware doesn't support.
3630 * - (-ENODEV) if *port_id* invalid.
3631 * - (-EIO) if device is removed.
3632 * - others depends on the specific operations implementation.
3634 int rte_eth_dev_get_eeprom_length(uint16_t port_id);
3637 * Retrieve EEPROM and EEPROM attribute
3640 * The port identifier of the Ethernet device.
3642 * The template includes buffer for return EEPROM data and
3643 * EEPROM attributes to be filled.
3645 * - (0) if successful.
3646 * - (-ENOTSUP) if hardware doesn't support.
3647 * - (-ENODEV) if *port_id* invalid.
3648 * - (-EIO) if device is removed.
3649 * - others depends on the specific operations implementation.
3651 int rte_eth_dev_get_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
3654 * Program EEPROM with provided data
3657 * The port identifier of the Ethernet device.
3659 * The template includes EEPROM data for programming and
3660 * EEPROM attributes to be filled
3662 * - (0) if successful.
3663 * - (-ENOTSUP) if hardware doesn't support.
3664 * - (-ENODEV) if *port_id* invalid.
3665 * - (-EIO) if device is removed.
3666 * - others depends on the specific operations implementation.
3668 int rte_eth_dev_set_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info);
3672 * @b EXPERIMENTAL: this API may change without prior notice.
3674 * Retrieve the type and size of plugin module EEPROM
3677 * The port identifier of the Ethernet device.
3679 * The type and size of plugin module EEPROM.
3681 * - (0) if successful.
3682 * - (-ENOTSUP) if hardware doesn't support.
3683 * - (-ENODEV) if *port_id* invalid.
3684 * - (-EIO) if device is removed.
3685 * - others depends on the specific operations implementation.
3689 rte_eth_dev_get_module_info(uint16_t port_id,
3690 struct rte_eth_dev_module_info *modinfo);
3694 * @b EXPERIMENTAL: this API may change without prior notice.
3696 * Retrieve the data of plugin module EEPROM
3699 * The port identifier of the Ethernet device.
3701 * The template includes the plugin module EEPROM attributes, and the
3702 * buffer for return plugin module EEPROM data.
3704 * - (0) if successful.
3705 * - (-ENOTSUP) if hardware doesn't support.
3706 * - (-ENODEV) if *port_id* invalid.
3707 * - (-EIO) if device is removed.
3708 * - others depends on the specific operations implementation.
3712 rte_eth_dev_get_module_eeprom(uint16_t port_id,
3713 struct rte_dev_eeprom_info *info);
3716 * Set the list of multicast addresses to filter on an Ethernet device.
3719 * The port identifier of the Ethernet device.
3720 * @param mc_addr_set
3721 * The array of multicast addresses to set. Equal to NULL when the function
3722 * is invoked to flush the set of filtered addresses.
3724 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0
3725 * when the function is invoked to flush the set of filtered addresses.
3727 * - (0) if successful.
3728 * - (-ENODEV) if *port_id* invalid.
3729 * - (-EIO) if device is removed.
3730 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering.
3731 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources.
3733 int rte_eth_dev_set_mc_addr_list(uint16_t port_id,
3734 struct rte_ether_addr *mc_addr_set,
3735 uint32_t nb_mc_addr);
3738 * Enable IEEE1588/802.1AS timestamping for an Ethernet device.
3741 * The port identifier of the Ethernet device.
3745 * - -ENODEV: The port ID is invalid.
3746 * - -EIO: if device is removed.
3747 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3749 int rte_eth_timesync_enable(uint16_t port_id);
3752 * Disable IEEE1588/802.1AS timestamping for an Ethernet device.
3755 * The port identifier of the Ethernet device.
3759 * - -ENODEV: The port ID is invalid.
3760 * - -EIO: if device is removed.
3761 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3763 int rte_eth_timesync_disable(uint16_t port_id);
3766 * Read an IEEE1588/802.1AS RX timestamp from an Ethernet device.
3769 * The port identifier of the Ethernet device.
3771 * Pointer to the timestamp struct.
3773 * Device specific flags. Used to pass the RX timesync register index to
3774 * i40e. Unused in igb/ixgbe, pass 0 instead.
3778 * - -EINVAL: No timestamp is available.
3779 * - -ENODEV: The port ID is invalid.
3780 * - -EIO: if device is removed.
3781 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3783 int rte_eth_timesync_read_rx_timestamp(uint16_t port_id,
3784 struct timespec *timestamp, uint32_t flags);
3787 * Read an IEEE1588/802.1AS TX timestamp from an Ethernet device.
3790 * The port identifier of the Ethernet device.
3792 * Pointer to the timestamp struct.
3796 * - -EINVAL: No timestamp is available.
3797 * - -ENODEV: The port ID is invalid.
3798 * - -EIO: if device is removed.
3799 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3801 int rte_eth_timesync_read_tx_timestamp(uint16_t port_id,
3802 struct timespec *timestamp);
3805 * Adjust the timesync clock on an Ethernet device.
3807 * This is usually used in conjunction with other Ethdev timesync functions to
3808 * synchronize the device time using the IEEE1588/802.1AS protocol.
3811 * The port identifier of the Ethernet device.
3813 * The adjustment in nanoseconds.
3817 * - -ENODEV: The port ID is invalid.
3818 * - -EIO: if device is removed.
3819 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3821 int rte_eth_timesync_adjust_time(uint16_t port_id, int64_t delta);
3824 * Read the time from the timesync clock on an Ethernet device.
3826 * This is usually used in conjunction with other Ethdev timesync functions to
3827 * synchronize the device time using the IEEE1588/802.1AS protocol.
3830 * The port identifier of the Ethernet device.
3832 * Pointer to the timespec struct that holds the time.
3837 int rte_eth_timesync_read_time(uint16_t port_id, struct timespec *time);
3840 * Set the time of the timesync clock on an Ethernet device.
3842 * This is usually used in conjunction with other Ethdev timesync functions to
3843 * synchronize the device time using the IEEE1588/802.1AS protocol.
3846 * The port identifier of the Ethernet device.
3848 * Pointer to the timespec struct that holds the time.
3852 * - -EINVAL: No timestamp is available.
3853 * - -ENODEV: The port ID is invalid.
3854 * - -EIO: if device is removed.
3855 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3857 int rte_eth_timesync_write_time(uint16_t port_id, const struct timespec *time);
3861 * @b EXPERIMENTAL: this API may change without prior notice.
3863 * Read the current clock counter of an Ethernet device
3865 * This returns the current raw clock value of an Ethernet device. It is
3866 * a raw amount of ticks, with no given time reference.
3867 * The value returned here is from the same clock than the one
3868 * filling timestamp field of Rx packets when using hardware timestamp
3869 * offload. Therefore it can be used to compute a precise conversion of
3870 * the device clock to the real time.
3872 * E.g, a simple heuristic to derivate the frequency would be:
3873 * uint64_t start, end;
3874 * rte_eth_read_clock(port, start);
3875 * rte_delay_ms(100);
3876 * rte_eth_read_clock(port, end);
3877 * double freq = (end - start) * 10;
3879 * Compute a common reference with:
3880 * uint64_t base_time_sec = current_time();
3881 * uint64_t base_clock;
3882 * rte_eth_read_clock(port, base_clock);
3884 * Then, convert the raw mbuf timestamp with:
3885 * base_time_sec + (double)(mbuf->timestamp - base_clock) / freq;
3887 * This simple example will not provide a very good accuracy. One must
3888 * at least measure multiple times the frequency and do a regression.
3889 * To avoid deviation from the system time, the common reference can
3890 * be repeated from time to time. The integer division can also be
3891 * converted by a multiplication and a shift for better performance.
3894 * The port identifier of the Ethernet device.
3896 * Pointer to the uint64_t that holds the raw clock value.
3900 * - -ENODEV: The port ID is invalid.
3901 * - -ENOTSUP: The function is not supported by the Ethernet driver.
3905 rte_eth_read_clock(uint16_t port_id, uint64_t *clock);
3908 * Config l2 tunnel ether type of an Ethernet device for filtering specific
3909 * tunnel packets by ether type.
3912 * The port identifier of the Ethernet device.
3914 * l2 tunnel configuration.
3917 * - (0) if successful.
3918 * - (-ENODEV) if port identifier is invalid.
3919 * - (-EIO) if device is removed.
3920 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3923 rte_eth_dev_l2_tunnel_eth_type_conf(uint16_t port_id,
3924 struct rte_eth_l2_tunnel_conf *l2_tunnel);
3927 * Enable/disable l2 tunnel offload functions. Include,
3928 * 1, The ability of parsing a type of l2 tunnel of an Ethernet device.
3929 * Filtering, forwarding and offloading this type of tunnel packets depend on
3931 * 2, Stripping the l2 tunnel tag.
3932 * 3, Insertion of the l2 tunnel tag.
3933 * 4, Forwarding the packets based on the l2 tunnel tag.
3936 * The port identifier of the Ethernet device.
3938 * l2 tunnel parameters.
3940 * Indicate the offload function.
3942 * Enable or disable this function.
3945 * - (0) if successful.
3946 * - (-ENODEV) if port identifier is invalid.
3947 * - (-EIO) if device is removed.
3948 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3951 rte_eth_dev_l2_tunnel_offload_set(uint16_t port_id,
3952 struct rte_eth_l2_tunnel_conf *l2_tunnel,
3957 * Get the port id from device name. The device name should be specified
3959 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:2:00.0
3960 * - SoC device name, for example- fsl-gmac0
3961 * - vdev dpdk name, for example- net_[pcap0|null0|tap0]
3964 * pci address or name of the device
3966 * pointer to port identifier of the device
3968 * - (0) if successful and port_id is filled.
3969 * - (-ENODEV or -EINVAL) on failure.
3972 rte_eth_dev_get_port_by_name(const char *name, uint16_t *port_id);
3975 * Get the device name from port id. The device name is specified as below:
3976 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:02:00.0
3977 * - SoC device name, for example- fsl-gmac0
3978 * - vdev dpdk name, for example- net_[pcap0|null0|tun0|tap0]
3981 * Port identifier of the device.
3983 * Buffer of size RTE_ETH_NAME_MAX_LEN to store the name.
3985 * - (0) if successful.
3986 * - (-EINVAL) on failure.
3989 rte_eth_dev_get_name_by_port(uint16_t port_id, char *name);
3992 * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from
3993 * the ethernet device information, otherwise adjust them to boundaries.
3996 * The port identifier of the Ethernet device.
3998 * A pointer to a uint16_t where the number of receive
3999 * descriptors stored.
4001 * A pointer to a uint16_t where the number of transmit
4002 * descriptors stored.
4004 * - (0) if successful.
4005 * - (-ENOTSUP, -ENODEV or -EINVAL) on failure.
4007 int rte_eth_dev_adjust_nb_rx_tx_desc(uint16_t port_id,
4008 uint16_t *nb_rx_desc,
4009 uint16_t *nb_tx_desc);
4012 * Test if a port supports specific mempool ops.
4015 * Port identifier of the Ethernet device.
4017 * The name of the pool operations to test.
4019 * - 0: best mempool ops choice for this port.
4020 * - 1: mempool ops are supported for this port.
4021 * - -ENOTSUP: mempool ops not supported for this port.
4022 * - -ENODEV: Invalid port Identifier.
4023 * - -EINVAL: Pool param is null.
4026 rte_eth_dev_pool_ops_supported(uint16_t port_id, const char *pool);
4029 * Get the security context for the Ethernet device.
4032 * Port identifier of the Ethernet device
4035 * - pointer to security context on success.
4038 rte_eth_dev_get_sec_ctx(uint16_t port_id);
4041 #include <rte_ethdev_core.h>
4045 * Retrieve a burst of input packets from a receive queue of an Ethernet
4046 * device. The retrieved packets are stored in *rte_mbuf* structures whose
4047 * pointers are supplied in the *rx_pkts* array.
4049 * The rte_eth_rx_burst() function loops, parsing the RX ring of the
4050 * receive queue, up to *nb_pkts* packets, and for each completed RX
4051 * descriptor in the ring, it performs the following operations:
4053 * - Initialize the *rte_mbuf* data structure associated with the
4054 * RX descriptor according to the information provided by the NIC into
4055 * that RX descriptor.
4057 * - Store the *rte_mbuf* data structure into the next entry of the
4060 * - Replenish the RX descriptor with a new *rte_mbuf* buffer
4061 * allocated from the memory pool associated with the receive queue at
4062 * initialization time.
4064 * When retrieving an input packet that was scattered by the controller
4065 * into multiple receive descriptors, the rte_eth_rx_burst() function
4066 * appends the associated *rte_mbuf* buffers to the first buffer of the
4069 * The rte_eth_rx_burst() function returns the number of packets
4070 * actually retrieved, which is the number of *rte_mbuf* data structures
4071 * effectively supplied into the *rx_pkts* array.
4072 * A return value equal to *nb_pkts* indicates that the RX queue contained
4073 * at least *rx_pkts* packets, and this is likely to signify that other
4074 * received packets remain in the input queue. Applications implementing
4075 * a "retrieve as much received packets as possible" policy can check this
4076 * specific case and keep invoking the rte_eth_rx_burst() function until
4077 * a value less than *nb_pkts* is returned.
4079 * This receive method has the following advantages:
4081 * - It allows a run-to-completion network stack engine to retrieve and
4082 * to immediately process received packets in a fast burst-oriented
4083 * approach, avoiding the overhead of unnecessary intermediate packet
4084 * queue/dequeue operations.
4086 * - Conversely, it also allows an asynchronous-oriented processing
4087 * method to retrieve bursts of received packets and to immediately
4088 * queue them for further parallel processing by another logical core,
4089 * for instance. However, instead of having received packets being
4090 * individually queued by the driver, this approach allows the caller
4091 * of the rte_eth_rx_burst() function to queue a burst of retrieved
4092 * packets at a time and therefore dramatically reduce the cost of
4093 * enqueue/dequeue operations per packet.
4095 * - It allows the rte_eth_rx_burst() function of the driver to take
4096 * advantage of burst-oriented hardware features (CPU cache,
4097 * prefetch instructions, and so on) to minimize the number of CPU
4098 * cycles per packet.
4100 * To summarize, the proposed receive API enables many
4101 * burst-oriented optimizations in both synchronous and asynchronous
4102 * packet processing environments with no overhead in both cases.
4104 * The rte_eth_rx_burst() function does not provide any error
4105 * notification to avoid the corresponding overhead. As a hint, the
4106 * upper-level application might check the status of the device link once
4107 * being systematically returned a 0 value for a given number of tries.
4110 * The port identifier of the Ethernet device.
4112 * The index of the receive queue from which to retrieve input packets.
4113 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
4114 * to rte_eth_dev_configure().
4116 * The address of an array of pointers to *rte_mbuf* structures that
4117 * must be large enough to store *nb_pkts* pointers in it.
4119 * The maximum number of packets to retrieve.
4121 * The number of packets actually retrieved, which is the number
4122 * of pointers to *rte_mbuf* structures effectively supplied to the
4125 static inline uint16_t
4126 rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id,
4127 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
4129 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
4132 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4133 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
4134 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_pkt_burst, 0);
4136 if (queue_id >= dev->data->nb_rx_queues) {
4137 RTE_ETHDEV_LOG(ERR, "Invalid RX queue_id=%u\n", queue_id);
4141 nb_rx = (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id],
4144 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
4145 if (unlikely(dev->post_rx_burst_cbs[queue_id] != NULL)) {
4146 struct rte_eth_rxtx_callback *cb =
4147 dev->post_rx_burst_cbs[queue_id];
4150 nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
4151 nb_pkts, cb->param);
4153 } while (cb != NULL);
4161 * Get the number of used descriptors of a rx queue
4164 * The port identifier of the Ethernet device.
4166 * The queue id on the specific port.
4168 * The number of used descriptors in the specific queue, or:
4169 * (-EINVAL) if *port_id* or *queue_id* is invalid
4170 * (-ENOTSUP) if the device does not support this function
4173 rte_eth_rx_queue_count(uint16_t port_id, uint16_t queue_id)
4175 struct rte_eth_dev *dev;
4177 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -EINVAL);
4178 dev = &rte_eth_devices[port_id];
4179 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_count, -ENOTSUP);
4180 if (queue_id >= dev->data->nb_rx_queues)
4183 return (int)(*dev->dev_ops->rx_queue_count)(dev, queue_id);
4187 * Check if the DD bit of the specific RX descriptor in the queue has been set
4190 * The port identifier of the Ethernet device.
4192 * The queue id on the specific port.
4194 * The offset of the descriptor ID from tail.
4196 * - (1) if the specific DD bit is set.
4197 * - (0) if the specific DD bit is not set.
4198 * - (-ENODEV) if *port_id* invalid.
4199 * - (-ENOTSUP) if the device does not support this function
4202 rte_eth_rx_descriptor_done(uint16_t port_id, uint16_t queue_id, uint16_t offset)
4204 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
4205 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
4206 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_descriptor_done, -ENOTSUP);
4207 return (*dev->dev_ops->rx_descriptor_done)( \
4208 dev->data->rx_queues[queue_id], offset);
4211 #define RTE_ETH_RX_DESC_AVAIL 0 /**< Desc available for hw. */
4212 #define RTE_ETH_RX_DESC_DONE 1 /**< Desc done, filled by hw. */
4213 #define RTE_ETH_RX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
4216 * Check the status of a Rx descriptor in the queue
4218 * It should be called in a similar context than the Rx function:
4219 * - on a dataplane core
4220 * - not concurrently on the same queue
4222 * Since it's a dataplane function, no check is performed on port_id and
4223 * queue_id. The caller must therefore ensure that the port is enabled
4224 * and the queue is configured and running.
4226 * Note: accessing to a random descriptor in the ring may trigger cache
4227 * misses and have a performance impact.
4230 * A valid port identifier of the Ethernet device which.
4232 * A valid Rx queue identifier on this port.
4234 * The offset of the descriptor starting from tail (0 is the next
4235 * packet to be received by the driver).
4238 * - (RTE_ETH_RX_DESC_AVAIL): Descriptor is available for the hardware to
4240 * - (RTE_ETH_RX_DESC_DONE): Descriptor is done, it is filled by hw, but
4241 * not yet processed by the driver (i.e. in the receive queue).
4242 * - (RTE_ETH_RX_DESC_UNAVAIL): Descriptor is unavailable, either hold by
4243 * the driver and not yet returned to hw, or reserved by the hw.
4244 * - (-EINVAL) bad descriptor offset.
4245 * - (-ENOTSUP) if the device does not support this function.
4246 * - (-ENODEV) bad port or queue (only if compiled with debug).
4249 rte_eth_rx_descriptor_status(uint16_t port_id, uint16_t queue_id,
4252 struct rte_eth_dev *dev;
4255 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4256 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
4258 dev = &rte_eth_devices[port_id];
4259 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4260 if (queue_id >= dev->data->nb_rx_queues)
4263 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_descriptor_status, -ENOTSUP);
4264 rxq = dev->data->rx_queues[queue_id];
4266 return (*dev->dev_ops->rx_descriptor_status)(rxq, offset);
4269 #define RTE_ETH_TX_DESC_FULL 0 /**< Desc filled for hw, waiting xmit. */
4270 #define RTE_ETH_TX_DESC_DONE 1 /**< Desc done, packet is transmitted. */
4271 #define RTE_ETH_TX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
4274 * Check the status of a Tx descriptor in the queue.
4276 * It should be called in a similar context than the Tx function:
4277 * - on a dataplane core
4278 * - not concurrently on the same queue
4280 * Since it's a dataplane function, no check is performed on port_id and
4281 * queue_id. The caller must therefore ensure that the port is enabled
4282 * and the queue is configured and running.
4284 * Note: accessing to a random descriptor in the ring may trigger cache
4285 * misses and have a performance impact.
4288 * A valid port identifier of the Ethernet device which.
4290 * A valid Tx queue identifier on this port.
4292 * The offset of the descriptor starting from tail (0 is the place where
4293 * the next packet will be send).
4296 * - (RTE_ETH_TX_DESC_FULL) Descriptor is being processed by the hw, i.e.
4297 * in the transmit queue.
4298 * - (RTE_ETH_TX_DESC_DONE) Hardware is done with this descriptor, it can
4299 * be reused by the driver.
4300 * - (RTE_ETH_TX_DESC_UNAVAIL): Descriptor is unavailable, reserved by the
4301 * driver or the hardware.
4302 * - (-EINVAL) bad descriptor offset.
4303 * - (-ENOTSUP) if the device does not support this function.
4304 * - (-ENODEV) bad port or queue (only if compiled with debug).
4306 static inline int rte_eth_tx_descriptor_status(uint16_t port_id,
4307 uint16_t queue_id, uint16_t offset)
4309 struct rte_eth_dev *dev;
4312 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4313 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
4315 dev = &rte_eth_devices[port_id];
4316 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4317 if (queue_id >= dev->data->nb_tx_queues)
4320 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->tx_descriptor_status, -ENOTSUP);
4321 txq = dev->data->tx_queues[queue_id];
4323 return (*dev->dev_ops->tx_descriptor_status)(txq, offset);
4327 * Send a burst of output packets on a transmit queue of an Ethernet device.
4329 * The rte_eth_tx_burst() function is invoked to transmit output packets
4330 * on the output queue *queue_id* of the Ethernet device designated by its
4332 * The *nb_pkts* parameter is the number of packets to send which are
4333 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
4334 * allocated from a pool created with rte_pktmbuf_pool_create().
4335 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets,
4336 * up to the number of transmit descriptors available in the TX ring of the
4338 * For each packet to send, the rte_eth_tx_burst() function performs
4339 * the following operations:
4341 * - Pick up the next available descriptor in the transmit ring.
4343 * - Free the network buffer previously sent with that descriptor, if any.
4345 * - Initialize the transmit descriptor with the information provided
4346 * in the *rte_mbuf data structure.
4348 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers,
4349 * the rte_eth_tx_burst() function uses several transmit descriptors
4352 * The rte_eth_tx_burst() function returns the number of packets it
4353 * actually sent. A return value equal to *nb_pkts* means that all packets
4354 * have been sent, and this is likely to signify that other output packets
4355 * could be immediately transmitted again. Applications that implement a
4356 * "send as many packets to transmit as possible" policy can check this
4357 * specific case and keep invoking the rte_eth_tx_burst() function until
4358 * a value less than *nb_pkts* is returned.
4360 * It is the responsibility of the rte_eth_tx_burst() function to
4361 * transparently free the memory buffers of packets previously sent.
4362 * This feature is driven by the *tx_free_thresh* value supplied to the
4363 * rte_eth_dev_configure() function at device configuration time.
4364 * When the number of free TX descriptors drops below this threshold, the
4365 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers
4366 * of those packets whose transmission was effectively completed.
4368 * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can
4369 * invoke this function concurrently on the same tx queue without SW lock.
4370 * @see rte_eth_dev_info_get, struct rte_eth_txconf::offloads
4372 * @see rte_eth_tx_prepare to perform some prior checks or adjustments
4376 * The port identifier of the Ethernet device.
4378 * The index of the transmit queue through which output packets must be
4380 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4381 * to rte_eth_dev_configure().
4383 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
4384 * which contain the output packets.
4386 * The maximum number of packets to transmit.
4388 * The number of output packets actually stored in transmit descriptors of
4389 * the transmit ring. The return value can be less than the value of the
4390 * *tx_pkts* parameter when the transmit ring is full or has been filled up.
4392 static inline uint16_t
4393 rte_eth_tx_burst(uint16_t port_id, uint16_t queue_id,
4394 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
4396 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
4398 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4399 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
4400 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_pkt_burst, 0);
4402 if (queue_id >= dev->data->nb_tx_queues) {
4403 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
4408 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
4409 struct rte_eth_rxtx_callback *cb = dev->pre_tx_burst_cbs[queue_id];
4411 if (unlikely(cb != NULL)) {
4413 nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
4416 } while (cb != NULL);
4420 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts);
4424 * Process a burst of output packets on a transmit queue of an Ethernet device.
4426 * The rte_eth_tx_prepare() function is invoked to prepare output packets to be
4427 * transmitted on the output queue *queue_id* of the Ethernet device designated
4429 * The *nb_pkts* parameter is the number of packets to be prepared which are
4430 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
4431 * allocated from a pool created with rte_pktmbuf_pool_create().
4432 * For each packet to send, the rte_eth_tx_prepare() function performs
4433 * the following operations:
4435 * - Check if packet meets devices requirements for tx offloads.
4437 * - Check limitations about number of segments.
4439 * - Check additional requirements when debug is enabled.
4441 * - Update and/or reset required checksums when tx offload is set for packet.
4443 * Since this function can modify packet data, provided mbufs must be safely
4444 * writable (e.g. modified data cannot be in shared segment).
4446 * The rte_eth_tx_prepare() function returns the number of packets ready to be
4447 * sent. A return value equal to *nb_pkts* means that all packets are valid and
4448 * ready to be sent, otherwise stops processing on the first invalid packet and
4449 * leaves the rest packets untouched.
4451 * When this functionality is not implemented in the driver, all packets are
4452 * are returned untouched.
4455 * The port identifier of the Ethernet device.
4456 * The value must be a valid port id.
4458 * The index of the transmit queue through which output packets must be
4460 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4461 * to rte_eth_dev_configure().
4463 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
4464 * which contain the output packets.
4466 * The maximum number of packets to process.
4468 * The number of packets correct and ready to be sent. The return value can be
4469 * less than the value of the *tx_pkts* parameter when some packet doesn't
4470 * meet devices requirements with rte_errno set appropriately:
4471 * - EINVAL: offload flags are not correctly set
4472 * - ENOTSUP: the offload feature is not supported by the hardware
4476 #ifndef RTE_ETHDEV_TX_PREPARE_NOOP
4478 static inline uint16_t
4479 rte_eth_tx_prepare(uint16_t port_id, uint16_t queue_id,
4480 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
4482 struct rte_eth_dev *dev;
4484 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4485 if (!rte_eth_dev_is_valid_port(port_id)) {
4486 RTE_ETHDEV_LOG(ERR, "Invalid TX port_id=%u\n", port_id);
4492 dev = &rte_eth_devices[port_id];
4494 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
4495 if (queue_id >= dev->data->nb_tx_queues) {
4496 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id);
4502 if (!dev->tx_pkt_prepare)
4505 return (*dev->tx_pkt_prepare)(dev->data->tx_queues[queue_id],
4512 * Native NOOP operation for compilation targets which doesn't require any
4513 * preparations steps, and functional NOOP may introduce unnecessary performance
4516 * Generally this is not a good idea to turn it on globally and didn't should
4517 * be used if behavior of tx_preparation can change.
4520 static inline uint16_t
4521 rte_eth_tx_prepare(__rte_unused uint16_t port_id,
4522 __rte_unused uint16_t queue_id,
4523 __rte_unused struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
4531 * Send any packets queued up for transmission on a port and HW queue
4533 * This causes an explicit flush of packets previously buffered via the
4534 * rte_eth_tx_buffer() function. It returns the number of packets successfully
4535 * sent to the NIC, and calls the error callback for any unsent packets. Unless
4536 * explicitly set up otherwise, the default callback simply frees the unsent
4537 * packets back to the owning mempool.
4540 * The port identifier of the Ethernet device.
4542 * The index of the transmit queue through which output packets must be
4544 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4545 * to rte_eth_dev_configure().
4547 * Buffer of packets to be transmit.
4549 * The number of packets successfully sent to the Ethernet device. The error
4550 * callback is called for any packets which could not be sent.
4552 static inline uint16_t
4553 rte_eth_tx_buffer_flush(uint16_t port_id, uint16_t queue_id,
4554 struct rte_eth_dev_tx_buffer *buffer)
4557 uint16_t to_send = buffer->length;
4562 sent = rte_eth_tx_burst(port_id, queue_id, buffer->pkts, to_send);
4566 /* All packets sent, or to be dealt with by callback below */
4567 if (unlikely(sent != to_send))
4568 buffer->error_callback(&buffer->pkts[sent],
4569 (uint16_t)(to_send - sent),
4570 buffer->error_userdata);
4576 * Buffer a single packet for future transmission on a port and queue
4578 * This function takes a single mbuf/packet and buffers it for later
4579 * transmission on the particular port and queue specified. Once the buffer is
4580 * full of packets, an attempt will be made to transmit all the buffered
4581 * packets. In case of error, where not all packets can be transmitted, a
4582 * callback is called with the unsent packets as a parameter. If no callback
4583 * is explicitly set up, the unsent packets are just freed back to the owning
4584 * mempool. The function returns the number of packets actually sent i.e.
4585 * 0 if no buffer flush occurred, otherwise the number of packets successfully
4589 * The port identifier of the Ethernet device.
4591 * The index of the transmit queue through which output packets must be
4593 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
4594 * to rte_eth_dev_configure().
4596 * Buffer used to collect packets to be sent.
4598 * Pointer to the packet mbuf to be sent.
4600 * 0 = packet has been buffered for later transmission
4601 * N > 0 = packet has been buffered, and the buffer was subsequently flushed,
4602 * causing N packets to be sent, and the error callback to be called for
4605 static __rte_always_inline uint16_t
4606 rte_eth_tx_buffer(uint16_t port_id, uint16_t queue_id,
4607 struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt)
4609 buffer->pkts[buffer->length++] = tx_pkt;
4610 if (buffer->length < buffer->size)
4613 return rte_eth_tx_buffer_flush(port_id, queue_id, buffer);
4620 #endif /* _RTE_ETHDEV_H_ */