4 * Copyright(c) 2010-2017 Intel Corporation. All rights reserved.
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
11 * * Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * * Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in
15 * the documentation and/or other materials provided with the
17 * * Neither the name of Intel Corporation nor the names of its
18 * contributors may be used to endorse or promote products derived
19 * from this software without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
22 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
23 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
24 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
25 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
26 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
27 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
28 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
29 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
30 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
31 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
34 #ifndef _RTE_ETHDEV_H_
35 #define _RTE_ETHDEV_H_
40 * RTE Ethernet Device API
42 * The Ethernet Device API is composed of two parts:
44 * - The application-oriented Ethernet API that includes functions to setup
45 * an Ethernet device (configure it, setup its RX and TX queues and start it),
46 * to get its MAC address, the speed and the status of its physical link,
47 * to receive and to transmit packets, and so on.
49 * - The driver-oriented Ethernet API that exports a function allowing
50 * an Ethernet Poll Mode Driver (PMD) to simultaneously register itself as
51 * an Ethernet device driver and as a PCI driver for a set of matching PCI
52 * [Ethernet] devices classes.
54 * By default, all the functions of the Ethernet Device API exported by a PMD
55 * are lock-free functions which assume to not be invoked in parallel on
56 * different logical cores to work on the same target object. For instance,
57 * the receive function of a PMD cannot be invoked in parallel on two logical
58 * cores to poll the same RX queue [of the same port]. Of course, this function
59 * can be invoked in parallel by different logical cores on different RX queues.
60 * It is the responsibility of the upper level application to enforce this rule.
62 * If needed, parallel accesses by multiple logical cores to shared queues
63 * shall be explicitly protected by dedicated inline lock-aware functions
64 * built on top of their corresponding lock-free functions of the PMD API.
66 * In all functions of the Ethernet API, the Ethernet device is
67 * designated by an integer >= 0 named the device port identifier.
69 * At the Ethernet driver level, Ethernet devices are represented by a generic
70 * data structure of type *rte_eth_dev*.
72 * Ethernet devices are dynamically registered during the PCI probing phase
73 * performed at EAL initialization time.
74 * When an Ethernet device is being probed, an *rte_eth_dev* structure and
75 * a new port identifier are allocated for that device. Then, the eth_dev_init()
76 * function supplied by the Ethernet driver matching the probed PCI
77 * device is invoked to properly initialize the device.
79 * The role of the device init function consists of resetting the hardware,
80 * checking access to Non-volatile Memory (NVM), reading the MAC address
83 * If the device init operation is successful, the correspondence between
84 * the port identifier assigned to the new device and its associated
85 * *rte_eth_dev* structure is effectively registered.
86 * Otherwise, both the *rte_eth_dev* structure and the port identifier are
89 * The functions exported by the application Ethernet API to setup a device
90 * designated by its port identifier must be invoked in the following order:
91 * - rte_eth_dev_configure()
92 * - rte_eth_tx_queue_setup()
93 * - rte_eth_rx_queue_setup()
94 * - rte_eth_dev_start()
96 * Then, the network application can invoke, in any order, the functions
97 * exported by the Ethernet API to get the MAC address of a given device, to
98 * get the speed and the status of a device physical link, to receive/transmit
99 * [burst of] packets, and so on.
101 * If the application wants to change the configuration (i.e. call
102 * rte_eth_dev_configure(), rte_eth_tx_queue_setup(), or
103 * rte_eth_rx_queue_setup()), it must call rte_eth_dev_stop() first to stop the
104 * device and then do the reconfiguration before calling rte_eth_dev_start()
105 * again. The transmit and receive functions should not be invoked when the
108 * Please note that some configuration is not stored between calls to
109 * rte_eth_dev_stop()/rte_eth_dev_start(). The following configuration will
112 * - flow control settings
113 * - receive mode configuration (promiscuous mode, hardware checksum mode,
114 * RSS/VMDQ settings etc.)
115 * - VLAN filtering configuration
116 * - MAC addresses supplied to MAC address array
117 * - flow director filtering mode (but not filtering rules)
118 * - NIC queue statistics mappings
120 * Any other configuration will not be stored and will need to be re-entered
121 * before a call to rte_eth_dev_start().
123 * Finally, a network application can close an Ethernet device by invoking the
124 * rte_eth_dev_close() function.
126 * Each function of the application Ethernet API invokes a specific function
127 * of the PMD that controls the target device designated by its port
129 * For this purpose, all device-specific functions of an Ethernet driver are
130 * supplied through a set of pointers contained in a generic structure of type
132 * The address of the *eth_dev_ops* structure is stored in the *rte_eth_dev*
133 * structure by the device init function of the Ethernet driver, which is
134 * invoked during the PCI probing phase, as explained earlier.
136 * In other words, each function of the Ethernet API simply retrieves the
137 * *rte_eth_dev* structure associated with the device port identifier and
138 * performs an indirect invocation of the corresponding driver function
139 * supplied in the *eth_dev_ops* structure of the *rte_eth_dev* structure.
141 * For performance reasons, the address of the burst-oriented RX and TX
142 * functions of the Ethernet driver are not contained in the *eth_dev_ops*
143 * structure. Instead, they are directly stored at the beginning of the
144 * *rte_eth_dev* structure to avoid an extra indirect memory access during
147 * RTE ethernet device drivers do not use interrupts for transmitting or
148 * receiving. Instead, Ethernet drivers export Poll-Mode receive and transmit
149 * functions to applications.
150 * Both receive and transmit functions are packet-burst oriented to minimize
151 * their cost per packet through the following optimizations:
153 * - Sharing among multiple packets the incompressible cost of the
154 * invocation of receive/transmit functions.
156 * - Enabling receive/transmit functions to take advantage of burst-oriented
157 * hardware features (L1 cache, prefetch instructions, NIC head/tail
158 * registers) to minimize the number of CPU cycles per packet, for instance,
159 * by avoiding useless read memory accesses to ring descriptors, or by
160 * systematically using arrays of pointers that exactly fit L1 cache line
161 * boundaries and sizes.
163 * The burst-oriented receive function does not provide any error notification,
164 * to avoid the corresponding overhead. As a hint, the upper-level application
165 * might check the status of the device link once being systematically returned
166 * a 0 value by the receive function of the driver for a given number of tries.
175 /* Use this macro to check if LRO API is supported */
176 #define RTE_ETHDEV_HAS_LRO_SUPPORT
179 #include <rte_interrupts.h>
181 #include <rte_devargs.h>
182 #include <rte_errno.h>
183 #include "rte_ether.h"
184 #include "rte_eth_ctrl.h"
185 #include "rte_dev_info.h"
190 * A structure used to retrieve statistics for an Ethernet port.
191 * Not all statistics fields in struct rte_eth_stats are supported
192 * by any type of network interface card (NIC). If any statistics
193 * field is not supported, its value is 0.
195 struct rte_eth_stats {
196 uint64_t ipackets; /**< Total number of successfully received packets. */
197 uint64_t opackets; /**< Total number of successfully transmitted packets.*/
198 uint64_t ibytes; /**< Total number of successfully received bytes. */
199 uint64_t obytes; /**< Total number of successfully transmitted bytes. */
201 /**< Total of RX packets dropped by the HW,
202 * because there are no available buffer (i.e. RX queues are full).
204 uint64_t ierrors; /**< Total number of erroneous received packets. */
205 uint64_t oerrors; /**< Total number of failed transmitted packets. */
206 uint64_t rx_nombuf; /**< Total number of RX mbuf allocation failures. */
207 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
208 /**< Total number of queue RX packets. */
209 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
210 /**< Total number of queue TX packets. */
211 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
212 /**< Total number of successfully received queue bytes. */
213 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
214 /**< Total number of successfully transmitted queue bytes. */
215 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS];
216 /**< Total number of queue packets received that are dropped. */
220 * Device supported speeds bitmap flags
222 #define ETH_LINK_SPEED_AUTONEG (0 << 0) /**< Autonegotiate (all speeds) */
223 #define ETH_LINK_SPEED_FIXED (1 << 0) /**< Disable autoneg (fixed speed) */
224 #define ETH_LINK_SPEED_10M_HD (1 << 1) /**< 10 Mbps half-duplex */
225 #define ETH_LINK_SPEED_10M (1 << 2) /**< 10 Mbps full-duplex */
226 #define ETH_LINK_SPEED_100M_HD (1 << 3) /**< 100 Mbps half-duplex */
227 #define ETH_LINK_SPEED_100M (1 << 4) /**< 100 Mbps full-duplex */
228 #define ETH_LINK_SPEED_1G (1 << 5) /**< 1 Gbps */
229 #define ETH_LINK_SPEED_2_5G (1 << 6) /**< 2.5 Gbps */
230 #define ETH_LINK_SPEED_5G (1 << 7) /**< 5 Gbps */
231 #define ETH_LINK_SPEED_10G (1 << 8) /**< 10 Gbps */
232 #define ETH_LINK_SPEED_20G (1 << 9) /**< 20 Gbps */
233 #define ETH_LINK_SPEED_25G (1 << 10) /**< 25 Gbps */
234 #define ETH_LINK_SPEED_40G (1 << 11) /**< 40 Gbps */
235 #define ETH_LINK_SPEED_50G (1 << 12) /**< 50 Gbps */
236 #define ETH_LINK_SPEED_56G (1 << 13) /**< 56 Gbps */
237 #define ETH_LINK_SPEED_100G (1 << 14) /**< 100 Gbps */
240 * Ethernet numeric link speeds in Mbps
242 #define ETH_SPEED_NUM_NONE 0 /**< Not defined */
243 #define ETH_SPEED_NUM_10M 10 /**< 10 Mbps */
244 #define ETH_SPEED_NUM_100M 100 /**< 100 Mbps */
245 #define ETH_SPEED_NUM_1G 1000 /**< 1 Gbps */
246 #define ETH_SPEED_NUM_2_5G 2500 /**< 2.5 Gbps */
247 #define ETH_SPEED_NUM_5G 5000 /**< 5 Gbps */
248 #define ETH_SPEED_NUM_10G 10000 /**< 10 Gbps */
249 #define ETH_SPEED_NUM_20G 20000 /**< 20 Gbps */
250 #define ETH_SPEED_NUM_25G 25000 /**< 25 Gbps */
251 #define ETH_SPEED_NUM_40G 40000 /**< 40 Gbps */
252 #define ETH_SPEED_NUM_50G 50000 /**< 50 Gbps */
253 #define ETH_SPEED_NUM_56G 56000 /**< 56 Gbps */
254 #define ETH_SPEED_NUM_100G 100000 /**< 100 Gbps */
257 * A structure used to retrieve link-level information of an Ethernet port.
260 struct rte_eth_link {
261 uint32_t link_speed; /**< ETH_SPEED_NUM_ */
262 uint16_t link_duplex : 1; /**< ETH_LINK_[HALF/FULL]_DUPLEX */
263 uint16_t link_autoneg : 1; /**< ETH_LINK_SPEED_[AUTONEG/FIXED] */
264 uint16_t link_status : 1; /**< ETH_LINK_[DOWN/UP] */
265 } __attribute__((aligned(8))); /**< aligned for atomic64 read/write */
267 /* Utility constants */
268 #define ETH_LINK_HALF_DUPLEX 0 /**< Half-duplex connection. */
269 #define ETH_LINK_FULL_DUPLEX 1 /**< Full-duplex connection. */
270 #define ETH_LINK_DOWN 0 /**< Link is down. */
271 #define ETH_LINK_UP 1 /**< Link is up. */
272 #define ETH_LINK_FIXED 0 /**< No autonegotiation. */
273 #define ETH_LINK_AUTONEG 1 /**< Autonegotiated. */
276 * A structure used to configure the ring threshold registers of an RX/TX
277 * queue for an Ethernet port.
279 struct rte_eth_thresh {
280 uint8_t pthresh; /**< Ring prefetch threshold. */
281 uint8_t hthresh; /**< Ring host threshold. */
282 uint8_t wthresh; /**< Ring writeback threshold. */
286 * Simple flags are used for rte_eth_conf.rxmode.mq_mode.
288 #define ETH_MQ_RX_RSS_FLAG 0x1
289 #define ETH_MQ_RX_DCB_FLAG 0x2
290 #define ETH_MQ_RX_VMDQ_FLAG 0x4
293 * A set of values to identify what method is to be used to route
294 * packets to multiple queues.
296 enum rte_eth_rx_mq_mode {
297 /** None of DCB,RSS or VMDQ mode */
300 /** For RX side, only RSS is on */
301 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG,
302 /** For RX side,only DCB is on. */
303 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG,
304 /** Both DCB and RSS enable */
305 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG,
307 /** Only VMDQ, no RSS nor DCB */
308 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG,
309 /** RSS mode with VMDQ */
310 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG,
311 /** Use VMDQ+DCB to route traffic to queues */
312 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG,
313 /** Enable both VMDQ and DCB in VMDq */
314 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG |
319 * for rx mq mode backward compatible
321 #define ETH_RSS ETH_MQ_RX_RSS
322 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB
323 #define ETH_DCB_RX ETH_MQ_RX_DCB
326 * A set of values to identify what method is to be used to transmit
327 * packets using multi-TCs.
329 enum rte_eth_tx_mq_mode {
330 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */
331 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */
332 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */
333 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */
337 * for tx mq mode backward compatible
339 #define ETH_DCB_NONE ETH_MQ_TX_NONE
340 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB
341 #define ETH_DCB_TX ETH_MQ_TX_DCB
344 * A structure used to configure the RX features of an Ethernet port.
346 struct rte_eth_rxmode {
347 /** The multi-queue packet distribution mode to be used, e.g. RSS. */
348 enum rte_eth_rx_mq_mode mq_mode;
349 uint32_t max_rx_pkt_len; /**< Only used if jumbo_frame enabled. */
350 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/
352 uint16_t header_split : 1, /**< Header Split enable. */
353 hw_ip_checksum : 1, /**< IP/UDP/TCP checksum offload enable. */
354 hw_vlan_filter : 1, /**< VLAN filter enable. */
355 hw_vlan_strip : 1, /**< VLAN strip enable. */
356 hw_vlan_extend : 1, /**< Extended VLAN enable. */
357 jumbo_frame : 1, /**< Jumbo Frame Receipt enable. */
358 hw_strip_crc : 1, /**< Enable CRC stripping by hardware. */
359 enable_scatter : 1, /**< Enable scatter packets rx handler */
360 enable_lro : 1; /**< Enable LRO */
364 * VLAN types to indicate if it is for single VLAN, inner VLAN or outer VLAN.
365 * Note that single VLAN is treated the same as inner VLAN.
368 ETH_VLAN_TYPE_UNKNOWN = 0,
369 ETH_VLAN_TYPE_INNER, /**< Inner VLAN. */
370 ETH_VLAN_TYPE_OUTER, /**< Single VLAN, or outer VLAN. */
375 * A structure used to describe a vlan filter.
376 * If the bit corresponding to a VID is set, such VID is on.
378 struct rte_vlan_filter_conf {
383 * A structure used to configure the Receive Side Scaling (RSS) feature
384 * of an Ethernet port.
385 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points
386 * to an array holding the RSS key to use for hashing specific header
387 * fields of received packets. The length of this array should be indicated
388 * by *rss_key_len* below. Otherwise, a default random hash key is used by
391 * The *rss_key_len* field of the *rss_conf* structure indicates the length
392 * in bytes of the array pointed by *rss_key*. To be compatible, this length
393 * will be checked in i40e only. Others assume 40 bytes to be used as before.
395 * The *rss_hf* field of the *rss_conf* structure indicates the different
396 * types of IPv4/IPv6 packets to which the RSS hashing must be applied.
397 * Supplying an *rss_hf* equal to zero disables the RSS feature.
399 struct rte_eth_rss_conf {
400 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */
401 uint8_t rss_key_len; /**< hash key length in bytes. */
402 uint64_t rss_hf; /**< Hash functions to apply - see below. */
406 * The RSS offload types are defined based on flow types which are defined
407 * in rte_eth_ctrl.h. Different NIC hardwares may support different RSS offload
408 * types. The supported flow types or RSS offload types can be queried by
409 * rte_eth_dev_info_get().
411 #define ETH_RSS_IPV4 (1ULL << RTE_ETH_FLOW_IPV4)
412 #define ETH_RSS_FRAG_IPV4 (1ULL << RTE_ETH_FLOW_FRAG_IPV4)
413 #define ETH_RSS_NONFRAG_IPV4_TCP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_TCP)
414 #define ETH_RSS_NONFRAG_IPV4_UDP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_UDP)
415 #define ETH_RSS_NONFRAG_IPV4_SCTP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_SCTP)
416 #define ETH_RSS_NONFRAG_IPV4_OTHER (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_OTHER)
417 #define ETH_RSS_IPV6 (1ULL << RTE_ETH_FLOW_IPV6)
418 #define ETH_RSS_FRAG_IPV6 (1ULL << RTE_ETH_FLOW_FRAG_IPV6)
419 #define ETH_RSS_NONFRAG_IPV6_TCP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_TCP)
420 #define ETH_RSS_NONFRAG_IPV6_UDP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_UDP)
421 #define ETH_RSS_NONFRAG_IPV6_SCTP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_SCTP)
422 #define ETH_RSS_NONFRAG_IPV6_OTHER (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_OTHER)
423 #define ETH_RSS_L2_PAYLOAD (1ULL << RTE_ETH_FLOW_L2_PAYLOAD)
424 #define ETH_RSS_IPV6_EX (1ULL << RTE_ETH_FLOW_IPV6_EX)
425 #define ETH_RSS_IPV6_TCP_EX (1ULL << RTE_ETH_FLOW_IPV6_TCP_EX)
426 #define ETH_RSS_IPV6_UDP_EX (1ULL << RTE_ETH_FLOW_IPV6_UDP_EX)
427 #define ETH_RSS_PORT (1ULL << RTE_ETH_FLOW_PORT)
428 #define ETH_RSS_VXLAN (1ULL << RTE_ETH_FLOW_VXLAN)
429 #define ETH_RSS_GENEVE (1ULL << RTE_ETH_FLOW_GENEVE)
430 #define ETH_RSS_NVGRE (1ULL << RTE_ETH_FLOW_NVGRE)
432 #define ETH_RSS_IP ( \
434 ETH_RSS_FRAG_IPV4 | \
435 ETH_RSS_NONFRAG_IPV4_OTHER | \
437 ETH_RSS_FRAG_IPV6 | \
438 ETH_RSS_NONFRAG_IPV6_OTHER | \
441 #define ETH_RSS_UDP ( \
442 ETH_RSS_NONFRAG_IPV4_UDP | \
443 ETH_RSS_NONFRAG_IPV6_UDP | \
446 #define ETH_RSS_TCP ( \
447 ETH_RSS_NONFRAG_IPV4_TCP | \
448 ETH_RSS_NONFRAG_IPV6_TCP | \
451 #define ETH_RSS_SCTP ( \
452 ETH_RSS_NONFRAG_IPV4_SCTP | \
453 ETH_RSS_NONFRAG_IPV6_SCTP)
455 #define ETH_RSS_TUNNEL ( \
461 /**< Mask of valid RSS hash protocols */
462 #define ETH_RSS_PROTO_MASK ( \
464 ETH_RSS_FRAG_IPV4 | \
465 ETH_RSS_NONFRAG_IPV4_TCP | \
466 ETH_RSS_NONFRAG_IPV4_UDP | \
467 ETH_RSS_NONFRAG_IPV4_SCTP | \
468 ETH_RSS_NONFRAG_IPV4_OTHER | \
470 ETH_RSS_FRAG_IPV6 | \
471 ETH_RSS_NONFRAG_IPV6_TCP | \
472 ETH_RSS_NONFRAG_IPV6_UDP | \
473 ETH_RSS_NONFRAG_IPV6_SCTP | \
474 ETH_RSS_NONFRAG_IPV6_OTHER | \
475 ETH_RSS_L2_PAYLOAD | \
477 ETH_RSS_IPV6_TCP_EX | \
478 ETH_RSS_IPV6_UDP_EX | \
485 * Definitions used for redirection table entry size.
486 * Some RSS RETA sizes may not be supported by some drivers, check the
487 * documentation or the description of relevant functions for more details.
489 #define ETH_RSS_RETA_SIZE_64 64
490 #define ETH_RSS_RETA_SIZE_128 128
491 #define ETH_RSS_RETA_SIZE_256 256
492 #define ETH_RSS_RETA_SIZE_512 512
493 #define RTE_RETA_GROUP_SIZE 64
495 /* Definitions used for VMDQ and DCB functionality */
496 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */
497 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */
498 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */
499 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */
501 /* DCB capability defines */
502 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */
503 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */
505 /* Definitions used for VLAN Offload functionality */
506 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */
507 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */
508 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */
510 /* Definitions used for mask VLAN setting */
511 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */
512 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/
513 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/
514 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/
516 /* Definitions used for receive MAC address */
517 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */
519 /* Definitions used for unicast hash */
520 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */
522 /* Definitions used for VMDQ pool rx mode setting */
523 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */
524 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */
525 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */
526 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */
527 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */
529 /** Maximum nb. of vlan per mirror rule */
530 #define ETH_MIRROR_MAX_VLANS 64
532 #define ETH_MIRROR_VIRTUAL_POOL_UP 0x01 /**< Virtual Pool uplink Mirroring. */
533 #define ETH_MIRROR_UPLINK_PORT 0x02 /**< Uplink Port Mirroring. */
534 #define ETH_MIRROR_DOWNLINK_PORT 0x04 /**< Downlink Port Mirroring. */
535 #define ETH_MIRROR_VLAN 0x08 /**< VLAN Mirroring. */
536 #define ETH_MIRROR_VIRTUAL_POOL_DOWN 0x10 /**< Virtual Pool downlink Mirroring. */
539 * A structure used to configure VLAN traffic mirror of an Ethernet port.
541 struct rte_eth_vlan_mirror {
542 uint64_t vlan_mask; /**< mask for valid VLAN ID. */
543 /** VLAN ID list for vlan mirroring. */
544 uint16_t vlan_id[ETH_MIRROR_MAX_VLANS];
548 * A structure used to configure traffic mirror of an Ethernet port.
550 struct rte_eth_mirror_conf {
551 uint8_t rule_type; /**< Mirroring rule type */
552 uint8_t dst_pool; /**< Destination pool for this mirror rule. */
553 uint64_t pool_mask; /**< Bitmap of pool for pool mirroring */
554 /** VLAN ID setting for VLAN mirroring. */
555 struct rte_eth_vlan_mirror vlan;
559 * A structure used to configure 64 entries of Redirection Table of the
560 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure
561 * more than 64 entries supported by hardware, an array of this structure
564 struct rte_eth_rss_reta_entry64 {
566 /**< Mask bits indicate which entries need to be updated/queried. */
567 uint16_t reta[RTE_RETA_GROUP_SIZE];
568 /**< Group of 64 redirection table entries. */
572 * This enum indicates the possible number of traffic classes
573 * in DCB configurations
575 enum rte_eth_nb_tcs {
576 ETH_4_TCS = 4, /**< 4 TCs with DCB. */
577 ETH_8_TCS = 8 /**< 8 TCs with DCB. */
581 * This enum indicates the possible number of queue pools
582 * in VMDQ configurations.
584 enum rte_eth_nb_pools {
585 ETH_8_POOLS = 8, /**< 8 VMDq pools. */
586 ETH_16_POOLS = 16, /**< 16 VMDq pools. */
587 ETH_32_POOLS = 32, /**< 32 VMDq pools. */
588 ETH_64_POOLS = 64 /**< 64 VMDq pools. */
591 /* This structure may be extended in future. */
592 struct rte_eth_dcb_rx_conf {
593 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */
594 /** Traffic class each UP mapped to. */
595 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
598 struct rte_eth_vmdq_dcb_tx_conf {
599 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */
600 /** Traffic class each UP mapped to. */
601 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
604 struct rte_eth_dcb_tx_conf {
605 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */
606 /** Traffic class each UP mapped to. */
607 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
610 struct rte_eth_vmdq_tx_conf {
611 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */
615 * A structure used to configure the VMDQ+DCB feature
616 * of an Ethernet port.
618 * Using this feature, packets are routed to a pool of queues, based
619 * on the vlan id in the vlan tag, and then to a specific queue within
620 * that pool, using the user priority vlan tag field.
622 * A default pool may be used, if desired, to route all traffic which
623 * does not match the vlan filter rules.
625 struct rte_eth_vmdq_dcb_conf {
626 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */
627 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
628 uint8_t default_pool; /**< The default pool, if applicable */
629 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
631 uint16_t vlan_id; /**< The vlan id of the received frame */
632 uint64_t pools; /**< Bitmask of pools for packet rx */
633 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
634 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES];
635 /**< Selects a queue in a pool */
639 * A structure used to configure the VMDQ feature of an Ethernet port when
640 * not combined with the DCB feature.
642 * Using this feature, packets are routed to a pool of queues. By default,
643 * the pool selection is based on the MAC address, the vlan id in the
644 * vlan tag as specified in the pool_map array.
645 * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool
646 * selection using only the MAC address. MAC address to pool mapping is done
647 * using the rte_eth_dev_mac_addr_add function, with the pool parameter
648 * corresponding to the pool id.
650 * Queue selection within the selected pool will be done using RSS when
651 * it is enabled or revert to the first queue of the pool if not.
653 * A default pool may be used, if desired, to route all traffic which
654 * does not match the vlan filter rules or any pool MAC address.
656 struct rte_eth_vmdq_rx_conf {
657 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */
658 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
659 uint8_t default_pool; /**< The default pool, if applicable */
660 uint8_t enable_loop_back; /**< Enable VT loop back */
661 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
662 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */
664 uint16_t vlan_id; /**< The vlan id of the received frame */
665 uint64_t pools; /**< Bitmask of pools for packet rx */
666 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
670 * A structure used to configure the TX features of an Ethernet port.
672 struct rte_eth_txmode {
673 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */
675 /* For i40e specifically */
678 uint8_t hw_vlan_reject_tagged : 1,
679 /**< If set, reject sending out tagged pkts */
680 hw_vlan_reject_untagged : 1,
681 /**< If set, reject sending out untagged pkts */
682 hw_vlan_insert_pvid : 1;
683 /**< If set, enable port based VLAN insertion */
687 * A structure used to configure an RX ring of an Ethernet port.
689 struct rte_eth_rxconf {
690 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */
691 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
692 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
693 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
696 #define ETH_TXQ_FLAGS_NOMULTSEGS 0x0001 /**< nb_segs=1 for all mbufs */
697 #define ETH_TXQ_FLAGS_NOREFCOUNT 0x0002 /**< refcnt can be ignored */
698 #define ETH_TXQ_FLAGS_NOMULTMEMP 0x0004 /**< all bufs come from same mempool */
699 #define ETH_TXQ_FLAGS_NOVLANOFFL 0x0100 /**< disable VLAN offload */
700 #define ETH_TXQ_FLAGS_NOXSUMSCTP 0x0200 /**< disable SCTP checksum offload */
701 #define ETH_TXQ_FLAGS_NOXSUMUDP 0x0400 /**< disable UDP checksum offload */
702 #define ETH_TXQ_FLAGS_NOXSUMTCP 0x0800 /**< disable TCP checksum offload */
703 #define ETH_TXQ_FLAGS_NOOFFLOADS \
704 (ETH_TXQ_FLAGS_NOVLANOFFL | ETH_TXQ_FLAGS_NOXSUMSCTP | \
705 ETH_TXQ_FLAGS_NOXSUMUDP | ETH_TXQ_FLAGS_NOXSUMTCP)
706 #define ETH_TXQ_FLAGS_NOXSUMS \
707 (ETH_TXQ_FLAGS_NOXSUMSCTP | ETH_TXQ_FLAGS_NOXSUMUDP | \
708 ETH_TXQ_FLAGS_NOXSUMTCP)
710 * A structure used to configure a TX ring of an Ethernet port.
712 struct rte_eth_txconf {
713 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */
714 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */
715 uint16_t tx_free_thresh; /**< Start freeing TX buffers if there are
716 less free descriptors than this value. */
718 uint32_t txq_flags; /**< Set flags for the Tx queue */
719 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
723 * A structure contains information about HW descriptor ring limitations.
725 struct rte_eth_desc_lim {
726 uint16_t nb_max; /**< Max allowed number of descriptors. */
727 uint16_t nb_min; /**< Min allowed number of descriptors. */
728 uint16_t nb_align; /**< Number of descriptors should be aligned to. */
731 * Max allowed number of segments per whole packet.
733 * - For TSO packet this is the total number of data descriptors allowed
736 * @see nb_mtu_seg_max
741 * Max number of segments per one MTU.
743 * - For non-TSO packet, this is the maximum allowed number of segments
744 * in a single transmit packet.
746 * - For TSO packet each segment within the TSO may span up to this
751 uint16_t nb_mtu_seg_max;
755 * This enum indicates the flow control mode
757 enum rte_eth_fc_mode {
758 RTE_FC_NONE = 0, /**< Disable flow control. */
759 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */
760 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */
761 RTE_FC_FULL /**< Enable flow control on both side. */
765 * A structure used to configure Ethernet flow control parameter.
766 * These parameters will be configured into the register of the NIC.
767 * Please refer to the corresponding data sheet for proper value.
769 struct rte_eth_fc_conf {
770 uint32_t high_water; /**< High threshold value to trigger XOFF */
771 uint32_t low_water; /**< Low threshold value to trigger XON */
772 uint16_t pause_time; /**< Pause quota in the Pause frame */
773 uint16_t send_xon; /**< Is XON frame need be sent */
774 enum rte_eth_fc_mode mode; /**< Link flow control mode */
775 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */
776 uint8_t autoneg; /**< Use Pause autoneg */
780 * A structure used to configure Ethernet priority flow control parameter.
781 * These parameters will be configured into the register of the NIC.
782 * Please refer to the corresponding data sheet for proper value.
784 struct rte_eth_pfc_conf {
785 struct rte_eth_fc_conf fc; /**< General flow control parameter. */
786 uint8_t priority; /**< VLAN User Priority. */
790 * Memory space that can be configured to store Flow Director filters
791 * in the board memory.
793 enum rte_fdir_pballoc_type {
794 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */
795 RTE_FDIR_PBALLOC_128K, /**< 128k. */
796 RTE_FDIR_PBALLOC_256K, /**< 256k. */
800 * Select report mode of FDIR hash information in RX descriptors.
802 enum rte_fdir_status_mode {
803 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */
804 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */
805 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */
809 * A structure used to configure the Flow Director (FDIR) feature
810 * of an Ethernet port.
812 * If mode is RTE_FDIR_DISABLE, the pballoc value is ignored.
814 struct rte_fdir_conf {
815 enum rte_fdir_mode mode; /**< Flow Director mode. */
816 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */
817 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */
818 /** RX queue of packets matching a "drop" filter in perfect mode. */
820 struct rte_eth_fdir_masks mask;
821 struct rte_eth_fdir_flex_conf flex_conf;
822 /**< Flex payload configuration. */
826 * UDP tunneling configuration.
827 * Used to config the UDP port for a type of tunnel.
828 * NICs need the UDP port to identify the tunnel type.
829 * Normally a type of tunnel has a default UDP port, this structure can be used
830 * in case if the users want to change or support more UDP port.
832 struct rte_eth_udp_tunnel {
833 uint16_t udp_port; /**< UDP port used for the tunnel. */
834 uint8_t prot_type; /**< Tunnel type. Defined in rte_eth_tunnel_type. */
838 * A structure used to enable/disable specific device interrupts.
840 struct rte_intr_conf {
841 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */
843 /** enable/disable rxq interrupt. 0 (default) - disable, 1 enable */
845 /** enable/disable rmv interrupt. 0 (default) - disable, 1 enable */
850 * A structure used to configure an Ethernet port.
851 * Depending upon the RX multi-queue mode, extra advanced
852 * configuration settings may be needed.
854 struct rte_eth_conf {
855 uint32_t link_speeds; /**< bitmap of ETH_LINK_SPEED_XXX of speeds to be
856 used. ETH_LINK_SPEED_FIXED disables link
857 autonegotiation, and a unique speed shall be
858 set. Otherwise, the bitmap defines the set of
859 speeds to be advertised. If the special value
860 ETH_LINK_SPEED_AUTONEG (0) is used, all speeds
861 supported are advertised. */
862 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */
863 struct rte_eth_txmode txmode; /**< Port TX configuration. */
864 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value
865 is 0, meaning the loopback mode is disabled.
866 Read the datasheet of given ethernet controller
867 for details. The possible values of this field
868 are defined in implementation of each driver. */
870 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */
871 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf;
872 /**< Port vmdq+dcb configuration. */
873 struct rte_eth_dcb_rx_conf dcb_rx_conf;
874 /**< Port dcb RX configuration. */
875 struct rte_eth_vmdq_rx_conf vmdq_rx_conf;
876 /**< Port vmdq RX configuration. */
877 } rx_adv_conf; /**< Port RX filtering configuration (union). */
879 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf;
880 /**< Port vmdq+dcb TX configuration. */
881 struct rte_eth_dcb_tx_conf dcb_tx_conf;
882 /**< Port dcb TX configuration. */
883 struct rte_eth_vmdq_tx_conf vmdq_tx_conf;
884 /**< Port vmdq TX configuration. */
885 } tx_adv_conf; /**< Port TX DCB configuration (union). */
886 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC
887 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */
888 uint32_t dcb_capability_en;
889 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. */
890 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */
894 * A structure used to retrieve the contextual information of
895 * an Ethernet device, such as the controlling driver of the device,
896 * its PCI context, etc...
900 * RX offload capabilities of a device.
902 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001
903 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002
904 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004
905 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008
906 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010
907 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020
908 #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040
909 #define DEV_RX_OFFLOAD_MACSEC_STRIP 0x00000080
912 * TX offload capabilities of a device.
914 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001
915 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002
916 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004
917 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008
918 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010
919 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020
920 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040
921 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */
922 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100
923 #define DEV_TX_OFFLOAD_VXLAN_TNL_TSO 0x00000200 /**< Used for tunneling packet. */
924 #define DEV_TX_OFFLOAD_GRE_TNL_TSO 0x00000400 /**< Used for tunneling packet. */
925 #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */
926 #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */
927 #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000
928 #define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000
929 /**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same
930 * tx queue without SW lock.
933 struct rte_pci_device;
936 * Ethernet device information
938 struct rte_eth_dev_info {
939 struct rte_pci_device *pci_dev; /**< Device PCI information. */
940 const char *driver_name; /**< Device Driver name. */
941 unsigned int if_index; /**< Index to bound host interface, or 0 if none.
942 Use if_indextoname() to translate into an interface name. */
943 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */
944 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */
945 uint16_t max_rx_queues; /**< Maximum number of RX queues. */
946 uint16_t max_tx_queues; /**< Maximum number of TX queues. */
947 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */
948 uint32_t max_hash_mac_addrs;
949 /** Maximum number of hash MAC addresses for MTA and UTA. */
950 uint16_t max_vfs; /**< Maximum number of VFs. */
951 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
952 uint32_t rx_offload_capa; /**< Device RX offload capabilities. */
953 uint32_t tx_offload_capa; /**< Device TX offload capabilities. */
955 /**< Device redirection table size, the total number of entries. */
956 uint8_t hash_key_size; /**< Hash key size in bytes */
957 /** Bit mask of RSS offloads, the bit offset also means flow type */
958 uint64_t flow_type_rss_offloads;
959 struct rte_eth_rxconf default_rxconf; /**< Default RX configuration */
960 struct rte_eth_txconf default_txconf; /**< Default TX configuration */
961 uint16_t vmdq_queue_base; /**< First queue ID for VMDQ pools. */
962 uint16_t vmdq_queue_num; /**< Queue number for VMDQ pools. */
963 uint16_t vmdq_pool_base; /**< First ID of VMDQ pools. */
964 struct rte_eth_desc_lim rx_desc_lim; /**< RX descriptors limits */
965 struct rte_eth_desc_lim tx_desc_lim; /**< TX descriptors limits */
966 uint32_t speed_capa; /**< Supported speeds bitmap (ETH_LINK_SPEED_). */
967 /** Configured number of rx/tx queues */
968 uint16_t nb_rx_queues; /**< Number of RX queues. */
969 uint16_t nb_tx_queues; /**< Number of TX queues. */
973 * Ethernet device RX queue information structure.
974 * Used to retieve information about configured queue.
976 struct rte_eth_rxq_info {
977 struct rte_mempool *mp; /**< mempool used by that queue. */
978 struct rte_eth_rxconf conf; /**< queue config parameters. */
979 uint8_t scattered_rx; /**< scattered packets RX supported. */
980 uint16_t nb_desc; /**< configured number of RXDs. */
981 } __rte_cache_min_aligned;
984 * Ethernet device TX queue information structure.
985 * Used to retieve information about configured queue.
987 struct rte_eth_txq_info {
988 struct rte_eth_txconf conf; /**< queue config parameters. */
989 uint16_t nb_desc; /**< configured number of TXDs. */
990 } __rte_cache_min_aligned;
992 /** Maximum name length for extended statistics counters */
993 #define RTE_ETH_XSTATS_NAME_SIZE 64
996 * An Ethernet device extended statistic structure
998 * This structure is used by rte_eth_xstats_get() to provide
999 * statistics that are not provided in the generic *rte_eth_stats*
1001 * It maps a name id, corresponding to an index in the array returned
1002 * by rte_eth_xstats_get_names(), to a statistic value.
1004 struct rte_eth_xstat {
1005 uint64_t id; /**< The index in xstats name array. */
1006 uint64_t value; /**< The statistic counter value. */
1010 * A name element for extended statistics.
1012 * An array of this structure is returned by rte_eth_xstats_get_names().
1013 * It lists the names of extended statistics for a PMD. The *rte_eth_xstat*
1014 * structure references these names by their array index.
1016 struct rte_eth_xstat_name {
1017 char name[RTE_ETH_XSTATS_NAME_SIZE]; /**< The statistic name. */
1020 #define ETH_DCB_NUM_TCS 8
1021 #define ETH_MAX_VMDQ_POOL 64
1024 * A structure used to get the information of queue and
1025 * TC mapping on both TX and RX paths.
1027 struct rte_eth_dcb_tc_queue_mapping {
1028 /** rx queues assigned to tc per Pool */
1032 } tc_rxq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1033 /** rx queues assigned to tc per Pool */
1037 } tc_txq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS];
1041 * A structure used to get the information of DCB.
1042 * It includes TC UP mapping and queue TC mapping.
1044 struct rte_eth_dcb_info {
1045 uint8_t nb_tcs; /**< number of TCs */
1046 uint8_t prio_tc[ETH_DCB_NUM_USER_PRIORITIES]; /**< Priority to tc */
1047 uint8_t tc_bws[ETH_DCB_NUM_TCS]; /**< TX BW percentage for each TC */
1048 /** rx queues assigned to tc */
1049 struct rte_eth_dcb_tc_queue_mapping tc_queue;
1053 * RX/TX queue states
1055 #define RTE_ETH_QUEUE_STATE_STOPPED 0
1056 #define RTE_ETH_QUEUE_STATE_STARTED 1
1060 struct rte_eth_dev_callback;
1061 /** @internal Structure to keep track of registered callbacks */
1062 TAILQ_HEAD(rte_eth_dev_cb_list, rte_eth_dev_callback);
1064 /* Macros to check for valid port */
1065 #define RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, retval) do { \
1066 if (!rte_eth_dev_is_valid_port(port_id)) { \
1067 RTE_PMD_DEBUG_TRACE("Invalid port_id=%d\n", port_id); \
1072 #define RTE_ETH_VALID_PORTID_OR_RET(port_id) do { \
1073 if (!rte_eth_dev_is_valid_port(port_id)) { \
1074 RTE_PMD_DEBUG_TRACE("Invalid port_id=%d\n", port_id); \
1079 #define RTE_ETH_DEV_TO_PCI(eth_dev) RTE_DEV_TO_PCI((eth_dev)->device)
1082 * l2 tunnel configuration.
1085 /**< l2 tunnel enable mask */
1086 #define ETH_L2_TUNNEL_ENABLE_MASK 0x00000001
1087 /**< l2 tunnel insertion mask */
1088 #define ETH_L2_TUNNEL_INSERTION_MASK 0x00000002
1089 /**< l2 tunnel stripping mask */
1090 #define ETH_L2_TUNNEL_STRIPPING_MASK 0x00000004
1091 /**< l2 tunnel forwarding mask */
1092 #define ETH_L2_TUNNEL_FORWARDING_MASK 0x00000008
1095 * Definitions of all functions exported by an Ethernet driver through the
1096 * the generic structure of type *eth_dev_ops* supplied in the *rte_eth_dev*
1097 * structure associated with an Ethernet device.
1100 typedef int (*eth_dev_configure_t)(struct rte_eth_dev *dev);
1101 /**< @internal Ethernet device configuration. */
1103 typedef int (*eth_dev_start_t)(struct rte_eth_dev *dev);
1104 /**< @internal Function used to start a configured Ethernet device. */
1106 typedef void (*eth_dev_stop_t)(struct rte_eth_dev *dev);
1107 /**< @internal Function used to stop a configured Ethernet device. */
1109 typedef int (*eth_dev_set_link_up_t)(struct rte_eth_dev *dev);
1110 /**< @internal Function used to link up a configured Ethernet device. */
1112 typedef int (*eth_dev_set_link_down_t)(struct rte_eth_dev *dev);
1113 /**< @internal Function used to link down a configured Ethernet device. */
1115 typedef void (*eth_dev_close_t)(struct rte_eth_dev *dev);
1116 /**< @internal Function used to close a configured Ethernet device. */
1118 typedef int (*eth_dev_reset_t)(struct rte_eth_dev *dev);
1119 /** <@internal Function used to reset a configured Ethernet device. */
1121 typedef void (*eth_promiscuous_enable_t)(struct rte_eth_dev *dev);
1122 /**< @internal Function used to enable the RX promiscuous mode of an Ethernet device. */
1124 typedef void (*eth_promiscuous_disable_t)(struct rte_eth_dev *dev);
1125 /**< @internal Function used to disable the RX promiscuous mode of an Ethernet device. */
1127 typedef void (*eth_allmulticast_enable_t)(struct rte_eth_dev *dev);
1128 /**< @internal Enable the receipt of all multicast packets by an Ethernet device. */
1130 typedef void (*eth_allmulticast_disable_t)(struct rte_eth_dev *dev);
1131 /**< @internal Disable the receipt of all multicast packets by an Ethernet device. */
1133 typedef int (*eth_link_update_t)(struct rte_eth_dev *dev,
1134 int wait_to_complete);
1135 /**< @internal Get link speed, duplex mode and state (up/down) of an Ethernet device. */
1137 typedef void (*eth_stats_get_t)(struct rte_eth_dev *dev,
1138 struct rte_eth_stats *igb_stats);
1139 /**< @internal Get global I/O statistics of an Ethernet device. */
1141 typedef void (*eth_stats_reset_t)(struct rte_eth_dev *dev);
1142 /**< @internal Reset global I/O statistics of an Ethernet device to 0. */
1144 typedef int (*eth_xstats_get_t)(struct rte_eth_dev *dev,
1145 struct rte_eth_xstat *stats, unsigned n);
1146 /**< @internal Get extended stats of an Ethernet device. */
1148 typedef int (*eth_xstats_get_by_id_t)(struct rte_eth_dev *dev,
1149 const uint64_t *ids,
1152 /**< @internal Get extended stats of an Ethernet device. */
1154 typedef void (*eth_xstats_reset_t)(struct rte_eth_dev *dev);
1155 /**< @internal Reset extended stats of an Ethernet device. */
1157 typedef int (*eth_xstats_get_names_t)(struct rte_eth_dev *dev,
1158 struct rte_eth_xstat_name *xstats_names, unsigned size);
1159 /**< @internal Get names of extended stats of an Ethernet device. */
1161 typedef int (*eth_xstats_get_names_by_id_t)(struct rte_eth_dev *dev,
1162 struct rte_eth_xstat_name *xstats_names, const uint64_t *ids,
1164 /**< @internal Get names of extended stats of an Ethernet device. */
1166 typedef int (*eth_queue_stats_mapping_set_t)(struct rte_eth_dev *dev,
1170 /**< @internal Set a queue statistics mapping for a tx/rx queue of an Ethernet device. */
1172 typedef void (*eth_dev_infos_get_t)(struct rte_eth_dev *dev,
1173 struct rte_eth_dev_info *dev_info);
1174 /**< @internal Get specific informations of an Ethernet device. */
1176 typedef const uint32_t *(*eth_dev_supported_ptypes_get_t)(struct rte_eth_dev *dev);
1177 /**< @internal Get supported ptypes of an Ethernet device. */
1179 typedef int (*eth_queue_start_t)(struct rte_eth_dev *dev,
1181 /**< @internal Start rx and tx of a queue of an Ethernet device. */
1183 typedef int (*eth_queue_stop_t)(struct rte_eth_dev *dev,
1185 /**< @internal Stop rx and tx of a queue of an Ethernet device. */
1187 typedef int (*eth_rx_queue_setup_t)(struct rte_eth_dev *dev,
1188 uint16_t rx_queue_id,
1189 uint16_t nb_rx_desc,
1190 unsigned int socket_id,
1191 const struct rte_eth_rxconf *rx_conf,
1192 struct rte_mempool *mb_pool);
1193 /**< @internal Set up a receive queue of an Ethernet device. */
1195 typedef int (*eth_tx_queue_setup_t)(struct rte_eth_dev *dev,
1196 uint16_t tx_queue_id,
1197 uint16_t nb_tx_desc,
1198 unsigned int socket_id,
1199 const struct rte_eth_txconf *tx_conf);
1200 /**< @internal Setup a transmit queue of an Ethernet device. */
1202 typedef int (*eth_rx_enable_intr_t)(struct rte_eth_dev *dev,
1203 uint16_t rx_queue_id);
1204 /**< @internal Enable interrupt of a receive queue of an Ethernet device. */
1206 typedef int (*eth_rx_disable_intr_t)(struct rte_eth_dev *dev,
1207 uint16_t rx_queue_id);
1208 /**< @internal Disable interrupt of a receive queue of an Ethernet device. */
1210 typedef void (*eth_queue_release_t)(void *queue);
1211 /**< @internal Release memory resources allocated by given RX/TX queue. */
1213 typedef uint32_t (*eth_rx_queue_count_t)(struct rte_eth_dev *dev,
1214 uint16_t rx_queue_id);
1215 /**< @internal Get number of used descriptors on a receive queue. */
1217 typedef int (*eth_rx_descriptor_done_t)(void *rxq, uint16_t offset);
1218 /**< @internal Check DD bit of specific RX descriptor */
1220 typedef int (*eth_rx_descriptor_status_t)(void *rxq, uint16_t offset);
1221 /**< @internal Check the status of a Rx descriptor */
1223 typedef int (*eth_tx_descriptor_status_t)(void *txq, uint16_t offset);
1224 /**< @internal Check the status of a Tx descriptor */
1226 typedef int (*eth_fw_version_get_t)(struct rte_eth_dev *dev,
1227 char *fw_version, size_t fw_size);
1228 /**< @internal Get firmware information of an Ethernet device. */
1230 typedef int (*eth_tx_done_cleanup_t)(void *txq, uint32_t free_cnt);
1231 /**< @internal Force mbufs to be from TX ring. */
1233 typedef void (*eth_rxq_info_get_t)(struct rte_eth_dev *dev,
1234 uint16_t rx_queue_id, struct rte_eth_rxq_info *qinfo);
1236 typedef void (*eth_txq_info_get_t)(struct rte_eth_dev *dev,
1237 uint16_t tx_queue_id, struct rte_eth_txq_info *qinfo);
1239 typedef int (*mtu_set_t)(struct rte_eth_dev *dev, uint16_t mtu);
1240 /**< @internal Set MTU. */
1242 typedef int (*vlan_filter_set_t)(struct rte_eth_dev *dev,
1245 /**< @internal filtering of a VLAN Tag Identifier by an Ethernet device. */
1247 typedef int (*vlan_tpid_set_t)(struct rte_eth_dev *dev,
1248 enum rte_vlan_type type, uint16_t tpid);
1249 /**< @internal set the outer/inner VLAN-TPID by an Ethernet device. */
1251 typedef void (*vlan_offload_set_t)(struct rte_eth_dev *dev, int mask);
1252 /**< @internal set VLAN offload function by an Ethernet device. */
1254 typedef int (*vlan_pvid_set_t)(struct rte_eth_dev *dev,
1257 /**< @internal set port based TX VLAN insertion by an Ethernet device. */
1259 typedef void (*vlan_strip_queue_set_t)(struct rte_eth_dev *dev,
1260 uint16_t rx_queue_id,
1262 /**< @internal VLAN stripping enable/disable by an queue of Ethernet device. */
1264 typedef uint16_t (*eth_rx_burst_t)(void *rxq,
1265 struct rte_mbuf **rx_pkts,
1267 /**< @internal Retrieve input packets from a receive queue of an Ethernet device. */
1269 typedef uint16_t (*eth_tx_burst_t)(void *txq,
1270 struct rte_mbuf **tx_pkts,
1272 /**< @internal Send output packets on a transmit queue of an Ethernet device. */
1274 typedef uint16_t (*eth_tx_prep_t)(void *txq,
1275 struct rte_mbuf **tx_pkts,
1277 /**< @internal Prepare output packets on a transmit queue of an Ethernet device. */
1279 typedef int (*flow_ctrl_get_t)(struct rte_eth_dev *dev,
1280 struct rte_eth_fc_conf *fc_conf);
1281 /**< @internal Get current flow control parameter on an Ethernet device */
1283 typedef int (*flow_ctrl_set_t)(struct rte_eth_dev *dev,
1284 struct rte_eth_fc_conf *fc_conf);
1285 /**< @internal Setup flow control parameter on an Ethernet device */
1287 typedef int (*priority_flow_ctrl_set_t)(struct rte_eth_dev *dev,
1288 struct rte_eth_pfc_conf *pfc_conf);
1289 /**< @internal Setup priority flow control parameter on an Ethernet device */
1291 typedef int (*reta_update_t)(struct rte_eth_dev *dev,
1292 struct rte_eth_rss_reta_entry64 *reta_conf,
1293 uint16_t reta_size);
1294 /**< @internal Update RSS redirection table on an Ethernet device */
1296 typedef int (*reta_query_t)(struct rte_eth_dev *dev,
1297 struct rte_eth_rss_reta_entry64 *reta_conf,
1298 uint16_t reta_size);
1299 /**< @internal Query RSS redirection table on an Ethernet device */
1301 typedef int (*rss_hash_update_t)(struct rte_eth_dev *dev,
1302 struct rte_eth_rss_conf *rss_conf);
1303 /**< @internal Update RSS hash configuration of an Ethernet device */
1305 typedef int (*rss_hash_conf_get_t)(struct rte_eth_dev *dev,
1306 struct rte_eth_rss_conf *rss_conf);
1307 /**< @internal Get current RSS hash configuration of an Ethernet device */
1309 typedef int (*eth_dev_led_on_t)(struct rte_eth_dev *dev);
1310 /**< @internal Turn on SW controllable LED on an Ethernet device */
1312 typedef int (*eth_dev_led_off_t)(struct rte_eth_dev *dev);
1313 /**< @internal Turn off SW controllable LED on an Ethernet device */
1315 typedef void (*eth_mac_addr_remove_t)(struct rte_eth_dev *dev, uint32_t index);
1316 /**< @internal Remove MAC address from receive address register */
1318 typedef int (*eth_mac_addr_add_t)(struct rte_eth_dev *dev,
1319 struct ether_addr *mac_addr,
1322 /**< @internal Set a MAC address into Receive Address Address Register */
1324 typedef void (*eth_mac_addr_set_t)(struct rte_eth_dev *dev,
1325 struct ether_addr *mac_addr);
1326 /**< @internal Set a MAC address into Receive Address Address Register */
1328 typedef int (*eth_uc_hash_table_set_t)(struct rte_eth_dev *dev,
1329 struct ether_addr *mac_addr,
1331 /**< @internal Set a Unicast Hash bitmap */
1333 typedef int (*eth_uc_all_hash_table_set_t)(struct rte_eth_dev *dev,
1335 /**< @internal Set all Unicast Hash bitmap */
1337 typedef int (*eth_set_queue_rate_limit_t)(struct rte_eth_dev *dev,
1340 /**< @internal Set queue TX rate */
1342 typedef int (*eth_mirror_rule_set_t)(struct rte_eth_dev *dev,
1343 struct rte_eth_mirror_conf *mirror_conf,
1346 /**< @internal Add a traffic mirroring rule on an Ethernet device */
1348 typedef int (*eth_mirror_rule_reset_t)(struct rte_eth_dev *dev,
1350 /**< @internal Remove a traffic mirroring rule on an Ethernet device */
1352 typedef int (*eth_udp_tunnel_port_add_t)(struct rte_eth_dev *dev,
1353 struct rte_eth_udp_tunnel *tunnel_udp);
1354 /**< @internal Add tunneling UDP port */
1356 typedef int (*eth_udp_tunnel_port_del_t)(struct rte_eth_dev *dev,
1357 struct rte_eth_udp_tunnel *tunnel_udp);
1358 /**< @internal Delete tunneling UDP port */
1360 typedef int (*eth_set_mc_addr_list_t)(struct rte_eth_dev *dev,
1361 struct ether_addr *mc_addr_set,
1362 uint32_t nb_mc_addr);
1363 /**< @internal set the list of multicast addresses on an Ethernet device */
1365 typedef int (*eth_timesync_enable_t)(struct rte_eth_dev *dev);
1366 /**< @internal Function used to enable IEEE1588/802.1AS timestamping. */
1368 typedef int (*eth_timesync_disable_t)(struct rte_eth_dev *dev);
1369 /**< @internal Function used to disable IEEE1588/802.1AS timestamping. */
1371 typedef int (*eth_timesync_read_rx_timestamp_t)(struct rte_eth_dev *dev,
1372 struct timespec *timestamp,
1374 /**< @internal Function used to read an RX IEEE1588/802.1AS timestamp. */
1376 typedef int (*eth_timesync_read_tx_timestamp_t)(struct rte_eth_dev *dev,
1377 struct timespec *timestamp);
1378 /**< @internal Function used to read a TX IEEE1588/802.1AS timestamp. */
1380 typedef int (*eth_timesync_adjust_time)(struct rte_eth_dev *dev, int64_t);
1381 /**< @internal Function used to adjust the device clock */
1383 typedef int (*eth_timesync_read_time)(struct rte_eth_dev *dev,
1384 struct timespec *timestamp);
1385 /**< @internal Function used to get time from the device clock. */
1387 typedef int (*eth_timesync_write_time)(struct rte_eth_dev *dev,
1388 const struct timespec *timestamp);
1389 /**< @internal Function used to get time from the device clock */
1391 typedef int (*eth_get_reg_t)(struct rte_eth_dev *dev,
1392 struct rte_dev_reg_info *info);
1393 /**< @internal Retrieve registers */
1395 typedef int (*eth_get_eeprom_length_t)(struct rte_eth_dev *dev);
1396 /**< @internal Retrieve eeprom size */
1398 typedef int (*eth_get_eeprom_t)(struct rte_eth_dev *dev,
1399 struct rte_dev_eeprom_info *info);
1400 /**< @internal Retrieve eeprom data */
1402 typedef int (*eth_set_eeprom_t)(struct rte_eth_dev *dev,
1403 struct rte_dev_eeprom_info *info);
1404 /**< @internal Program eeprom data */
1406 typedef int (*eth_l2_tunnel_eth_type_conf_t)
1407 (struct rte_eth_dev *dev, struct rte_eth_l2_tunnel_conf *l2_tunnel);
1408 /**< @internal config l2 tunnel ether type */
1410 typedef int (*eth_l2_tunnel_offload_set_t)
1411 (struct rte_eth_dev *dev,
1412 struct rte_eth_l2_tunnel_conf *l2_tunnel,
1415 /**< @internal enable/disable the l2 tunnel offload functions */
1418 typedef int (*eth_filter_ctrl_t)(struct rte_eth_dev *dev,
1419 enum rte_filter_type filter_type,
1420 enum rte_filter_op filter_op,
1422 /**< @internal Take operations to assigned filter type on an Ethernet device */
1424 typedef int (*eth_tm_ops_get_t)(struct rte_eth_dev *dev, void *ops);
1425 /**< @internal Get Traffic Management (TM) operations on an Ethernet device */
1427 typedef int (*eth_get_dcb_info)(struct rte_eth_dev *dev,
1428 struct rte_eth_dcb_info *dcb_info);
1429 /**< @internal Get dcb information on an Ethernet device */
1432 * @internal A structure containing the functions exported by an Ethernet driver.
1434 struct eth_dev_ops {
1435 eth_dev_configure_t dev_configure; /**< Configure device. */
1436 eth_dev_start_t dev_start; /**< Start device. */
1437 eth_dev_stop_t dev_stop; /**< Stop device. */
1438 eth_dev_set_link_up_t dev_set_link_up; /**< Device link up. */
1439 eth_dev_set_link_down_t dev_set_link_down; /**< Device link down. */
1440 eth_dev_close_t dev_close; /**< Close device. */
1441 eth_dev_reset_t dev_reset; /**< Reset device. */
1442 eth_link_update_t link_update; /**< Get device link state. */
1444 eth_promiscuous_enable_t promiscuous_enable; /**< Promiscuous ON. */
1445 eth_promiscuous_disable_t promiscuous_disable;/**< Promiscuous OFF. */
1446 eth_allmulticast_enable_t allmulticast_enable;/**< RX multicast ON. */
1447 eth_allmulticast_disable_t allmulticast_disable;/**< RX multicast OFF. */
1448 eth_mac_addr_remove_t mac_addr_remove; /**< Remove MAC address. */
1449 eth_mac_addr_add_t mac_addr_add; /**< Add a MAC address. */
1450 eth_mac_addr_set_t mac_addr_set; /**< Set a MAC address. */
1451 eth_set_mc_addr_list_t set_mc_addr_list; /**< set list of mcast addrs. */
1452 mtu_set_t mtu_set; /**< Set MTU. */
1454 eth_stats_get_t stats_get; /**< Get generic device statistics. */
1455 eth_stats_reset_t stats_reset; /**< Reset generic device statistics. */
1456 eth_xstats_get_t xstats_get; /**< Get extended device statistics. */
1457 eth_xstats_reset_t xstats_reset; /**< Reset extended device statistics. */
1458 eth_xstats_get_names_t xstats_get_names;
1459 /**< Get names of extended statistics. */
1460 eth_queue_stats_mapping_set_t queue_stats_mapping_set;
1461 /**< Configure per queue stat counter mapping. */
1463 eth_dev_infos_get_t dev_infos_get; /**< Get device info. */
1464 eth_rxq_info_get_t rxq_info_get; /**< retrieve RX queue information. */
1465 eth_txq_info_get_t txq_info_get; /**< retrieve TX queue information. */
1466 eth_fw_version_get_t fw_version_get; /**< Get firmware version. */
1467 eth_dev_supported_ptypes_get_t dev_supported_ptypes_get;
1468 /**< Get packet types supported and identified by device. */
1470 vlan_filter_set_t vlan_filter_set; /**< Filter VLAN Setup. */
1471 vlan_tpid_set_t vlan_tpid_set; /**< Outer/Inner VLAN TPID Setup. */
1472 vlan_strip_queue_set_t vlan_strip_queue_set; /**< VLAN Stripping on queue. */
1473 vlan_offload_set_t vlan_offload_set; /**< Set VLAN Offload. */
1474 vlan_pvid_set_t vlan_pvid_set; /**< Set port based TX VLAN insertion. */
1476 eth_queue_start_t rx_queue_start;/**< Start RX for a queue. */
1477 eth_queue_stop_t rx_queue_stop; /**< Stop RX for a queue. */
1478 eth_queue_start_t tx_queue_start;/**< Start TX for a queue. */
1479 eth_queue_stop_t tx_queue_stop; /**< Stop TX for a queue. */
1480 eth_rx_queue_setup_t rx_queue_setup;/**< Set up device RX queue. */
1481 eth_queue_release_t rx_queue_release; /**< Release RX queue. */
1482 eth_rx_queue_count_t rx_queue_count;
1483 /**< Get the number of used RX descriptors. */
1484 eth_rx_descriptor_done_t rx_descriptor_done; /**< Check rxd DD bit. */
1485 eth_rx_descriptor_status_t rx_descriptor_status;
1486 /**< Check the status of a Rx descriptor. */
1487 eth_tx_descriptor_status_t tx_descriptor_status;
1488 /**< Check the status of a Tx descriptor. */
1489 eth_rx_enable_intr_t rx_queue_intr_enable; /**< Enable Rx queue interrupt. */
1490 eth_rx_disable_intr_t rx_queue_intr_disable; /**< Disable Rx queue interrupt. */
1491 eth_tx_queue_setup_t tx_queue_setup;/**< Set up device TX queue. */
1492 eth_queue_release_t tx_queue_release; /**< Release TX queue. */
1493 eth_tx_done_cleanup_t tx_done_cleanup;/**< Free tx ring mbufs */
1495 eth_dev_led_on_t dev_led_on; /**< Turn on LED. */
1496 eth_dev_led_off_t dev_led_off; /**< Turn off LED. */
1498 flow_ctrl_get_t flow_ctrl_get; /**< Get flow control. */
1499 flow_ctrl_set_t flow_ctrl_set; /**< Setup flow control. */
1500 priority_flow_ctrl_set_t priority_flow_ctrl_set; /**< Setup priority flow control. */
1502 eth_uc_hash_table_set_t uc_hash_table_set; /**< Set Unicast Table Array. */
1503 eth_uc_all_hash_table_set_t uc_all_hash_table_set; /**< Set Unicast hash bitmap. */
1505 eth_mirror_rule_set_t mirror_rule_set; /**< Add a traffic mirror rule. */
1506 eth_mirror_rule_reset_t mirror_rule_reset; /**< reset a traffic mirror rule. */
1508 eth_udp_tunnel_port_add_t udp_tunnel_port_add; /** Add UDP tunnel port. */
1509 eth_udp_tunnel_port_del_t udp_tunnel_port_del; /** Del UDP tunnel port. */
1510 eth_l2_tunnel_eth_type_conf_t l2_tunnel_eth_type_conf;
1511 /** Config ether type of l2 tunnel. */
1512 eth_l2_tunnel_offload_set_t l2_tunnel_offload_set;
1513 /** Enable/disable l2 tunnel offload functions. */
1515 eth_set_queue_rate_limit_t set_queue_rate_limit; /**< Set queue rate limit. */
1517 rss_hash_update_t rss_hash_update; /** Configure RSS hash protocols. */
1518 rss_hash_conf_get_t rss_hash_conf_get; /** Get current RSS hash configuration. */
1519 reta_update_t reta_update; /** Update redirection table. */
1520 reta_query_t reta_query; /** Query redirection table. */
1522 eth_get_reg_t get_reg; /**< Get registers. */
1523 eth_get_eeprom_length_t get_eeprom_length; /**< Get eeprom length. */
1524 eth_get_eeprom_t get_eeprom; /**< Get eeprom data. */
1525 eth_set_eeprom_t set_eeprom; /**< Set eeprom. */
1528 eth_filter_ctrl_t filter_ctrl; /**< common filter control. */
1530 eth_get_dcb_info get_dcb_info; /** Get DCB information. */
1532 eth_timesync_enable_t timesync_enable;
1533 /** Turn IEEE1588/802.1AS timestamping on. */
1534 eth_timesync_disable_t timesync_disable;
1535 /** Turn IEEE1588/802.1AS timestamping off. */
1536 eth_timesync_read_rx_timestamp_t timesync_read_rx_timestamp;
1537 /** Read the IEEE1588/802.1AS RX timestamp. */
1538 eth_timesync_read_tx_timestamp_t timesync_read_tx_timestamp;
1539 /** Read the IEEE1588/802.1AS TX timestamp. */
1540 eth_timesync_adjust_time timesync_adjust_time; /** Adjust the device clock. */
1541 eth_timesync_read_time timesync_read_time; /** Get the device clock time. */
1542 eth_timesync_write_time timesync_write_time; /** Set the device clock time. */
1544 eth_xstats_get_by_id_t xstats_get_by_id;
1545 /**< Get extended device statistic values by ID. */
1546 eth_xstats_get_names_by_id_t xstats_get_names_by_id;
1547 /**< Get name of extended device statistics by ID. */
1549 eth_tm_ops_get_t tm_ops_get;
1550 /**< Get Traffic Management (TM) operations. */
1554 * Function type used for RX packet processing packet callbacks.
1556 * The callback function is called on RX with a burst of packets that have
1557 * been received on the given port and queue.
1560 * The Ethernet port on which RX is being performed.
1562 * The queue on the Ethernet port which is being used to receive the packets.
1564 * The burst of packets that have just been received.
1566 * The number of packets in the burst pointed to by "pkts".
1568 * The max number of packets that can be stored in the "pkts" array.
1570 * The arbitrary user parameter passed in by the application when the callback
1571 * was originally configured.
1573 * The number of packets returned to the user.
1575 typedef uint16_t (*rte_rx_callback_fn)(uint8_t port, uint16_t queue,
1576 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts,
1580 * Function type used for TX packet processing packet callbacks.
1582 * The callback function is called on TX with a burst of packets immediately
1583 * before the packets are put onto the hardware queue for transmission.
1586 * The Ethernet port on which TX is being performed.
1588 * The queue on the Ethernet port which is being used to transmit the packets.
1590 * The burst of packets that are about to be transmitted.
1592 * The number of packets in the burst pointed to by "pkts".
1594 * The arbitrary user parameter passed in by the application when the callback
1595 * was originally configured.
1597 * The number of packets to be written to the NIC.
1599 typedef uint16_t (*rte_tx_callback_fn)(uint8_t port, uint16_t queue,
1600 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param);
1604 * Structure used to hold information about the callbacks to be called for a
1605 * queue on RX and TX.
1607 struct rte_eth_rxtx_callback {
1608 struct rte_eth_rxtx_callback *next;
1610 rte_rx_callback_fn rx;
1611 rte_tx_callback_fn tx;
1617 * A set of values to describe the possible states of an eth device.
1619 enum rte_eth_dev_state {
1620 RTE_ETH_DEV_UNUSED = 0,
1621 RTE_ETH_DEV_ATTACHED,
1622 RTE_ETH_DEV_DEFERRED,
1627 * The generic data structure associated with each ethernet device.
1629 * Pointers to burst-oriented packet receive and transmit functions are
1630 * located at the beginning of the structure, along with the pointer to
1631 * where all the data elements for the particular device are stored in shared
1632 * memory. This split allows the function pointer and driver data to be per-
1633 * process, while the actual configuration data for the device is shared.
1635 struct rte_eth_dev {
1636 eth_rx_burst_t rx_pkt_burst; /**< Pointer to PMD receive function. */
1637 eth_tx_burst_t tx_pkt_burst; /**< Pointer to PMD transmit function. */
1638 eth_tx_prep_t tx_pkt_prepare; /**< Pointer to PMD transmit prepare function. */
1639 struct rte_eth_dev_data *data; /**< Pointer to device data */
1640 const struct eth_dev_ops *dev_ops; /**< Functions exported by PMD */
1641 struct rte_device *device; /**< Backing device */
1642 struct rte_intr_handle *intr_handle; /**< Device interrupt handle */
1643 /** User application callbacks for NIC interrupts */
1644 struct rte_eth_dev_cb_list link_intr_cbs;
1646 * User-supplied functions called from rx_burst to post-process
1647 * received packets before passing them to the user
1649 struct rte_eth_rxtx_callback *post_rx_burst_cbs[RTE_MAX_QUEUES_PER_PORT];
1651 * User-supplied functions called from tx_burst to pre-process
1652 * received packets before passing them to the driver for transmission.
1654 struct rte_eth_rxtx_callback *pre_tx_burst_cbs[RTE_MAX_QUEUES_PER_PORT];
1655 enum rte_eth_dev_state state; /**< Flag indicating the port state */
1656 } __rte_cache_aligned;
1658 struct rte_eth_dev_sriov {
1659 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */
1660 uint8_t nb_q_per_pool; /**< rx queue number per pool */
1661 uint16_t def_vmdq_idx; /**< Default pool num used for PF */
1662 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */
1664 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
1666 #define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN
1670 * The data part, with no function pointers, associated with each ethernet device.
1672 * This structure is safe to place in shared memory to be common among different
1673 * processes in a multi-process configuration.
1675 struct rte_eth_dev_data {
1676 char name[RTE_ETH_NAME_MAX_LEN]; /**< Unique identifier name */
1678 void **rx_queues; /**< Array of pointers to RX queues. */
1679 void **tx_queues; /**< Array of pointers to TX queues. */
1680 uint16_t nb_rx_queues; /**< Number of RX queues. */
1681 uint16_t nb_tx_queues; /**< Number of TX queues. */
1683 struct rte_eth_dev_sriov sriov; /**< SRIOV data */
1685 void *dev_private; /**< PMD-specific private data */
1687 struct rte_eth_link dev_link;
1688 /**< Link-level information & status */
1690 struct rte_eth_conf dev_conf; /**< Configuration applied to device. */
1691 uint16_t mtu; /**< Maximum Transmission Unit. */
1693 uint32_t min_rx_buf_size;
1694 /**< Common rx buffer size handled by all queues */
1696 uint64_t rx_mbuf_alloc_failed; /**< RX ring mbuf allocation failures. */
1697 struct ether_addr* mac_addrs;/**< Device Ethernet Link address. */
1698 uint64_t mac_pool_sel[ETH_NUM_RECEIVE_MAC_ADDR];
1699 /** bitmap array of associating Ethernet MAC addresses to pools */
1700 struct ether_addr* hash_mac_addrs;
1701 /** Device Ethernet MAC addresses of hash filtering. */
1702 uint8_t port_id; /**< Device [external] port identifier. */
1704 uint8_t promiscuous : 1, /**< RX promiscuous mode ON(1) / OFF(0). */
1705 scattered_rx : 1, /**< RX of scattered packets is ON(1) / OFF(0) */
1706 all_multicast : 1, /**< RX all multicast mode ON(1) / OFF(0). */
1707 dev_started : 1, /**< Device state: STARTED(1) / STOPPED(0). */
1708 lro : 1; /**< RX LRO is ON(1) / OFF(0) */
1709 uint8_t rx_queue_state[RTE_MAX_QUEUES_PER_PORT];
1710 /** Queues state: STARTED(1) / STOPPED(0) */
1711 uint8_t tx_queue_state[RTE_MAX_QUEUES_PER_PORT];
1712 /** Queues state: STARTED(1) / STOPPED(0) */
1713 uint32_t dev_flags; /**< Capabilities */
1714 enum rte_kernel_driver kdrv; /**< Kernel driver passthrough */
1715 int numa_node; /**< NUMA node connection */
1716 struct rte_vlan_filter_conf vlan_filter_conf;
1717 /**< VLAN filter configuration. */
1720 /** Device supports hotplug detach */
1721 #define RTE_ETH_DEV_DETACHABLE 0x0001
1722 /** Device supports link state interrupt */
1723 #define RTE_ETH_DEV_INTR_LSC 0x0002
1724 /** Device is a bonded slave */
1725 #define RTE_ETH_DEV_BONDED_SLAVE 0x0004
1726 /** Device supports device removal interrupt */
1727 #define RTE_ETH_DEV_INTR_RMV 0x0008
1731 * The pool of *rte_eth_dev* structures. The size of the pool
1732 * is configured at compile-time in the <rte_ethdev.c> file.
1734 extern struct rte_eth_dev rte_eth_devices[];
1737 * Iterates over valid ethdev ports.
1740 * The id of the next possible valid port.
1742 * Next valid port id, RTE_MAX_ETHPORTS if there is none.
1744 uint8_t rte_eth_find_next(uint8_t port_id);
1747 * Macro to iterate over all enabled ethdev ports.
1749 #define RTE_ETH_FOREACH_DEV(p) \
1750 for (p = rte_eth_find_next(0); \
1751 (unsigned int)p < (unsigned int)RTE_MAX_ETHPORTS; \
1752 p = rte_eth_find_next(p + 1))
1756 * Get the total number of Ethernet devices that have been successfully
1757 * initialized by the matching Ethernet driver during the PCI probing phase
1758 * and that are available for applications to use. These devices must be
1759 * accessed by using the ``RTE_ETH_FOREACH_DEV()`` macro to deal with
1760 * non-contiguous ranges of devices.
1761 * These non-contiguous ranges can be created by calls to hotplug functions or
1765 * - The total number of usable Ethernet devices.
1767 uint8_t rte_eth_dev_count(void);
1771 * Returns a ethdev slot specified by the unique identifier name.
1774 * The pointer to the Unique identifier name for each Ethernet device
1776 * - The pointer to the ethdev slot, on success. NULL on error
1778 struct rte_eth_dev *rte_eth_dev_allocated(const char *name);
1782 * Allocates a new ethdev slot for an ethernet device and returns the pointer
1783 * to that slot for the driver to use.
1785 * @param name Unique identifier name for each Ethernet device
1786 * @param type Device type of this Ethernet device
1788 * - Slot in the rte_dev_devices array for a new device;
1790 struct rte_eth_dev *rte_eth_dev_allocate(const char *name);
1794 * Attach to the ethdev already initialized by the primary
1797 * @param name Ethernet device's name.
1799 * - Success: Slot in the rte_dev_devices array for attached
1801 * - Error: Null pointer.
1803 struct rte_eth_dev *rte_eth_dev_attach_secondary(const char *name);
1807 * Release the specified ethdev port.
1810 * The *eth_dev* pointer is the address of the *rte_eth_dev* structure.
1812 * - 0 on success, negative on error
1814 int rte_eth_dev_release_port(struct rte_eth_dev *eth_dev);
1817 * Attach a new Ethernet device specified by arguments.
1820 * A pointer to a strings array describing the new device
1821 * to be attached. The strings should be a pci address like
1822 * '0000:01:00.0' or virtual device name like 'net_pcap0'.
1824 * A pointer to a port identifier actually attached.
1826 * 0 on success and port_id is filled, negative on error
1828 int rte_eth_dev_attach(const char *devargs, uint8_t *port_id);
1831 * Detach a Ethernet device specified by port identifier.
1832 * This function must be called when the device is in the
1836 * The port identifier of the device to detach.
1838 * A pointer to a buffer that will be filled with the device name.
1839 * This buffer must be at least RTE_DEV_NAME_MAX_LEN long.
1841 * 0 on success and devname is filled, negative on error
1843 int rte_eth_dev_detach(uint8_t port_id, char *devname);
1846 * Convert a numerical speed in Mbps to a bitmap flag that can be used in
1847 * the bitmap link_speeds of the struct rte_eth_conf
1850 * Numerical speed value in Mbps
1852 * ETH_LINK_[HALF/FULL]_DUPLEX (only for 10/100M speeds)
1854 * 0 if the speed cannot be mapped
1856 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex);
1859 * Configure an Ethernet device.
1860 * This function must be invoked first before any other function in the
1861 * Ethernet API. This function can also be re-invoked when a device is in the
1865 * The port identifier of the Ethernet device to configure.
1866 * @param nb_rx_queue
1867 * The number of receive queues to set up for the Ethernet device.
1868 * @param nb_tx_queue
1869 * The number of transmit queues to set up for the Ethernet device.
1871 * The pointer to the configuration data to be used for the Ethernet device.
1872 * The *rte_eth_conf* structure includes:
1873 * - the hardware offload features to activate, with dedicated fields for
1874 * each statically configurable offload hardware feature provided by
1875 * Ethernet devices, such as IP checksum or VLAN tag stripping for
1877 * - the Receive Side Scaling (RSS) configuration when using multiple RX
1880 * Embedding all configuration information in a single data structure
1881 * is the more flexible method that allows the addition of new features
1882 * without changing the syntax of the API.
1884 * - 0: Success, device configured.
1885 * - <0: Error code returned by the driver configuration function.
1887 int rte_eth_dev_configure(uint8_t port_id, uint16_t nb_rx_queue,
1888 uint16_t nb_tx_queue, const struct rte_eth_conf *eth_conf);
1892 * Release device queues and clear its configuration to force the user
1893 * application to reconfigure it. It is for internal use only.
1896 * Pointer to struct rte_eth_dev.
1901 void _rte_eth_dev_reset(struct rte_eth_dev *dev);
1904 * Allocate and set up a receive queue for an Ethernet device.
1906 * The function allocates a contiguous block of memory for *nb_rx_desc*
1907 * receive descriptors from a memory zone associated with *socket_id*
1908 * and initializes each receive descriptor with a network buffer allocated
1909 * from the memory pool *mb_pool*.
1912 * The port identifier of the Ethernet device.
1913 * @param rx_queue_id
1914 * The index of the receive queue to set up.
1915 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1916 * to rte_eth_dev_configure().
1918 * The number of receive descriptors to allocate for the receive ring.
1920 * The *socket_id* argument is the socket identifier in case of NUMA.
1921 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1922 * the DMA memory allocated for the receive descriptors of the ring.
1924 * The pointer to the configuration data to be used for the receive queue.
1925 * NULL value is allowed, in which case default RX configuration
1927 * The *rx_conf* structure contains an *rx_thresh* structure with the values
1928 * of the Prefetch, Host, and Write-Back threshold registers of the receive
1931 * The pointer to the memory pool from which to allocate *rte_mbuf* network
1932 * memory buffers to populate each descriptor of the receive ring.
1934 * - 0: Success, receive queue correctly set up.
1935 * - -EINVAL: The size of network buffers which can be allocated from the
1936 * memory pool does not fit the various buffer sizes allowed by the
1937 * device controller.
1938 * - -ENOMEM: Unable to allocate the receive ring descriptors or to
1939 * allocate network memory buffers from the memory pool when
1940 * initializing receive descriptors.
1942 int rte_eth_rx_queue_setup(uint8_t port_id, uint16_t rx_queue_id,
1943 uint16_t nb_rx_desc, unsigned int socket_id,
1944 const struct rte_eth_rxconf *rx_conf,
1945 struct rte_mempool *mb_pool);
1948 * Allocate and set up a transmit queue for an Ethernet device.
1951 * The port identifier of the Ethernet device.
1952 * @param tx_queue_id
1953 * The index of the transmit queue to set up.
1954 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1955 * to rte_eth_dev_configure().
1957 * The number of transmit descriptors to allocate for the transmit ring.
1959 * The *socket_id* argument is the socket identifier in case of NUMA.
1960 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1961 * the DMA memory allocated for the transmit descriptors of the ring.
1963 * The pointer to the configuration data to be used for the transmit queue.
1964 * NULL value is allowed, in which case default RX configuration
1966 * The *tx_conf* structure contains the following data:
1967 * - The *tx_thresh* structure with the values of the Prefetch, Host, and
1968 * Write-Back threshold registers of the transmit ring.
1969 * When setting Write-Back threshold to the value greater then zero,
1970 * *tx_rs_thresh* value should be explicitly set to one.
1971 * - The *tx_free_thresh* value indicates the [minimum] number of network
1972 * buffers that must be pending in the transmit ring to trigger their
1973 * [implicit] freeing by the driver transmit function.
1974 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit
1975 * descriptors that must be pending in the transmit ring before setting the
1976 * RS bit on a descriptor by the driver transmit function.
1977 * The *tx_rs_thresh* value should be less or equal then
1978 * *tx_free_thresh* value, and both of them should be less then
1980 * - The *txq_flags* member contains flags to pass to the TX queue setup
1981 * function to configure the behavior of the TX queue. This should be set
1982 * to 0 if no special configuration is required.
1984 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces
1985 * the transmit function to use default values.
1987 * - 0: Success, the transmit queue is correctly set up.
1988 * - -ENOMEM: Unable to allocate the transmit ring descriptors.
1990 int rte_eth_tx_queue_setup(uint8_t port_id, uint16_t tx_queue_id,
1991 uint16_t nb_tx_desc, unsigned int socket_id,
1992 const struct rte_eth_txconf *tx_conf);
1995 * Return the NUMA socket to which an Ethernet device is connected
1998 * The port identifier of the Ethernet device
2000 * The NUMA socket id to which the Ethernet device is connected or
2001 * a default of zero if the socket could not be determined.
2002 * -1 is returned is the port_id value is out of range.
2004 int rte_eth_dev_socket_id(uint8_t port_id);
2007 * Check if port_id of device is attached
2010 * The port identifier of the Ethernet device
2012 * - 0 if port is out of range or not attached
2013 * - 1 if device is attached
2015 int rte_eth_dev_is_valid_port(uint8_t port_id);
2018 * Start specified RX queue of a port. It is used when rx_deferred_start
2019 * flag of the specified queue is true.
2022 * The port identifier of the Ethernet device
2023 * @param rx_queue_id
2024 * The index of the rx queue to update the ring.
2025 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2026 * to rte_eth_dev_configure().
2028 * - 0: Success, the receive queue is started.
2029 * - -EINVAL: The port_id or the queue_id out of range.
2030 * - -ENOTSUP: The function not supported in PMD driver.
2032 int rte_eth_dev_rx_queue_start(uint8_t port_id, uint16_t rx_queue_id);
2035 * Stop specified RX queue of a port
2038 * The port identifier of the Ethernet device
2039 * @param rx_queue_id
2040 * The index of the rx queue to update the ring.
2041 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2042 * to rte_eth_dev_configure().
2044 * - 0: Success, the receive queue is stopped.
2045 * - -EINVAL: The port_id or the queue_id out of range.
2046 * - -ENOTSUP: The function not supported in PMD driver.
2048 int rte_eth_dev_rx_queue_stop(uint8_t port_id, uint16_t rx_queue_id);
2051 * Start TX for specified queue of a port. It is used when tx_deferred_start
2052 * flag of the specified queue is true.
2055 * The port identifier of the Ethernet device
2056 * @param tx_queue_id
2057 * The index of the tx queue to update the ring.
2058 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2059 * to rte_eth_dev_configure().
2061 * - 0: Success, the transmit queue is started.
2062 * - -EINVAL: The port_id or the queue_id out of range.
2063 * - -ENOTSUP: The function not supported in PMD driver.
2065 int rte_eth_dev_tx_queue_start(uint8_t port_id, uint16_t tx_queue_id);
2068 * Stop specified TX queue of a port
2071 * The port identifier of the Ethernet device
2072 * @param tx_queue_id
2073 * The index of the tx queue to update the ring.
2074 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2075 * to rte_eth_dev_configure().
2077 * - 0: Success, the transmit queue is stopped.
2078 * - -EINVAL: The port_id or the queue_id out of range.
2079 * - -ENOTSUP: The function not supported in PMD driver.
2081 int rte_eth_dev_tx_queue_stop(uint8_t port_id, uint16_t tx_queue_id);
2086 * Start an Ethernet device.
2088 * The device start step is the last one and consists of setting the configured
2089 * offload features and in starting the transmit and the receive units of the
2091 * On success, all basic functions exported by the Ethernet API (link status,
2092 * receive/transmit, and so on) can be invoked.
2095 * The port identifier of the Ethernet device.
2097 * - 0: Success, Ethernet device started.
2098 * - <0: Error code of the driver device start function.
2100 int rte_eth_dev_start(uint8_t port_id);
2103 * Stop an Ethernet device. The device can be restarted with a call to
2104 * rte_eth_dev_start()
2107 * The port identifier of the Ethernet device.
2109 void rte_eth_dev_stop(uint8_t port_id);
2113 * Link up an Ethernet device.
2115 * Set device link up will re-enable the device rx/tx
2116 * functionality after it is previously set device linked down.
2119 * The port identifier of the Ethernet device.
2121 * - 0: Success, Ethernet device linked up.
2122 * - <0: Error code of the driver device link up function.
2124 int rte_eth_dev_set_link_up(uint8_t port_id);
2127 * Link down an Ethernet device.
2128 * The device rx/tx functionality will be disabled if success,
2129 * and it can be re-enabled with a call to
2130 * rte_eth_dev_set_link_up()
2133 * The port identifier of the Ethernet device.
2135 int rte_eth_dev_set_link_down(uint8_t port_id);
2138 * Close a stopped Ethernet device. The device cannot be restarted!
2139 * The function frees all resources except for needed by the
2140 * closed state. To free these resources, call rte_eth_dev_detach().
2143 * The port identifier of the Ethernet device.
2145 void rte_eth_dev_close(uint8_t port_id);
2148 * Reset a Ethernet device and keep its port id.
2150 * When a port has to be reset passively, the DPDK application can invoke
2151 * this function. For example when a PF is reset, all its VFs should also
2152 * be reset. Normally a DPDK application can invoke this function when
2153 * RTE_ETH_EVENT_INTR_RESET event is detected, but can also use it to start
2154 * a port reset in other circumstances.
2156 * When this function is called, it first stops the port and then calls the
2157 * PMD specific dev_uninit( ) and dev_init( ) to return the port to initial
2158 * state, in which no Tx and Rx queues are setup, as if the port has been
2159 * reset and not started. The port keeps the port id it had before the
2162 * After calling rte_eth_dev_reset( ), the application should use
2163 * rte_eth_dev_configure( ), rte_eth_rx_queue_setup( ),
2164 * rte_eth_tx_queue_setup( ), and rte_eth_dev_start( )
2165 * to reconfigure the device as appropriate.
2167 * Note: To avoid unexpected behavior, the application should stop calling
2168 * Tx and Rx functions before calling rte_eth_dev_reset( ). For thread
2169 * safety, all these controlling functions should be called from the same
2173 * The port identifier of the Ethernet device.
2175 int rte_eth_dev_reset(uint8_t port_id);
2178 * Enable receipt in promiscuous mode for an Ethernet device.
2181 * The port identifier of the Ethernet device.
2183 void rte_eth_promiscuous_enable(uint8_t port_id);
2186 * Disable receipt in promiscuous mode for an Ethernet device.
2189 * The port identifier of the Ethernet device.
2191 void rte_eth_promiscuous_disable(uint8_t port_id);
2194 * Return the value of promiscuous mode for an Ethernet device.
2197 * The port identifier of the Ethernet device.
2199 * - (1) if promiscuous is enabled
2200 * - (0) if promiscuous is disabled.
2203 int rte_eth_promiscuous_get(uint8_t port_id);
2206 * Enable the receipt of any multicast frame by an Ethernet device.
2209 * The port identifier of the Ethernet device.
2211 void rte_eth_allmulticast_enable(uint8_t port_id);
2214 * Disable the receipt of all multicast frames by an Ethernet device.
2217 * The port identifier of the Ethernet device.
2219 void rte_eth_allmulticast_disable(uint8_t port_id);
2222 * Return the value of allmulticast mode for an Ethernet device.
2225 * The port identifier of the Ethernet device.
2227 * - (1) if allmulticast is enabled
2228 * - (0) if allmulticast is disabled.
2231 int rte_eth_allmulticast_get(uint8_t port_id);
2234 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2235 * or FULL-DUPLEX) of the physical link of an Ethernet device. It might need
2236 * to wait up to 9 seconds in it.
2239 * The port identifier of the Ethernet device.
2241 * A pointer to an *rte_eth_link* structure to be filled with
2242 * the status, the speed and the mode of the Ethernet device link.
2244 void rte_eth_link_get(uint8_t port_id, struct rte_eth_link *link);
2247 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2248 * or FULL-DUPLEX) of the physical link of an Ethernet device. It is a no-wait
2249 * version of rte_eth_link_get().
2252 * The port identifier of the Ethernet device.
2254 * A pointer to an *rte_eth_link* structure to be filled with
2255 * the status, the speed and the mode of the Ethernet device link.
2257 void rte_eth_link_get_nowait(uint8_t port_id, struct rte_eth_link *link);
2260 * Retrieve the general I/O statistics of an Ethernet device.
2263 * The port identifier of the Ethernet device.
2265 * A pointer to a structure of type *rte_eth_stats* to be filled with
2266 * the values of device counters for the following set of statistics:
2267 * - *ipackets* with the total of successfully received packets.
2268 * - *opackets* with the total of successfully transmitted packets.
2269 * - *ibytes* with the total of successfully received bytes.
2270 * - *obytes* with the total of successfully transmitted bytes.
2271 * - *ierrors* with the total of erroneous received packets.
2272 * - *oerrors* with the total of failed transmitted packets.
2274 * Zero if successful. Non-zero otherwise.
2276 int rte_eth_stats_get(uint8_t port_id, struct rte_eth_stats *stats);
2279 * Reset the general I/O statistics of an Ethernet device.
2282 * The port identifier of the Ethernet device.
2284 void rte_eth_stats_reset(uint8_t port_id);
2287 * Retrieve names of extended statistics of an Ethernet device.
2290 * The port identifier of the Ethernet device.
2291 * @param xstats_names
2292 * An rte_eth_xstat_name array of at least *size* elements to
2293 * be filled. If set to NULL, the function returns the required number
2296 * The size of the xstats_names array (number of elements).
2298 * - A positive value lower or equal to size: success. The return value
2299 * is the number of entries filled in the stats table.
2300 * - A positive value higher than size: error, the given statistics table
2301 * is too small. The return value corresponds to the size that should
2302 * be given to succeed. The entries in the table are not valid and
2303 * shall not be used by the caller.
2304 * - A negative value on error (invalid port id).
2306 int rte_eth_xstats_get_names(uint8_t port_id,
2307 struct rte_eth_xstat_name *xstats_names,
2311 * Retrieve extended statistics of an Ethernet device.
2314 * The port identifier of the Ethernet device.
2316 * A pointer to a table of structure of type *rte_eth_xstat*
2317 * to be filled with device statistics ids and values: id is the
2318 * index of the name string in xstats_names (see rte_eth_xstats_get_names()),
2319 * and value is the statistic counter.
2320 * This parameter can be set to NULL if n is 0.
2322 * The size of the xstats array (number of elements).
2324 * - A positive value lower or equal to n: success. The return value
2325 * is the number of entries filled in the stats table.
2326 * - A positive value higher than n: error, the given statistics table
2327 * is too small. The return value corresponds to the size that should
2328 * be given to succeed. The entries in the table are not valid and
2329 * shall not be used by the caller.
2330 * - A negative value on error (invalid port id).
2332 int rte_eth_xstats_get(uint8_t port_id, struct rte_eth_xstat *xstats,
2336 * Retrieve names of extended statistics of an Ethernet device.
2339 * The port identifier of the Ethernet device.
2340 * @param xstats_names
2341 * An rte_eth_xstat_name array of at least *size* elements to
2342 * be filled. If set to NULL, the function returns the required number
2345 * IDs array given by app to retrieve specific statistics
2347 * The size of the xstats_names array (number of elements).
2349 * - A positive value lower or equal to size: success. The return value
2350 * is the number of entries filled in the stats table.
2351 * - A positive value higher than size: error, the given statistics table
2352 * is too small. The return value corresponds to the size that should
2353 * be given to succeed. The entries in the table are not valid and
2354 * shall not be used by the caller.
2355 * - A negative value on error (invalid port id).
2358 rte_eth_xstats_get_names_by_id(uint8_t port_id,
2359 struct rte_eth_xstat_name *xstats_names, unsigned int size,
2363 * Retrieve extended statistics of an Ethernet device.
2366 * The port identifier of the Ethernet device.
2368 * A pointer to an ids array passed by application. This tells which
2369 * statistics values function should retrieve. This parameter
2370 * can be set to NULL if n is 0. In this case function will retrieve
2371 * all avalible statistics.
2373 * A pointer to a table to be filled with device statistics values.
2375 * The size of the ids array (number of elements).
2377 * - A positive value lower or equal to n: success. The return value
2378 * is the number of entries filled in the stats table.
2379 * - A positive value higher than n: error, the given statistics table
2380 * is too small. The return value corresponds to the size that should
2381 * be given to succeed. The entries in the table are not valid and
2382 * shall not be used by the caller.
2383 * - A negative value on error (invalid port id).
2385 int rte_eth_xstats_get_by_id(uint8_t port_id, const uint64_t *ids,
2386 uint64_t *values, unsigned int n);
2389 * Gets the ID of a statistic from its name.
2391 * This function searches for the statistics using string compares, and
2392 * as such should not be used on the fast-path. For fast-path retrieval of
2393 * specific statistics, store the ID as provided in *id* from this function,
2394 * and pass the ID to rte_eth_xstats_get()
2396 * @param port_id The port to look up statistics from
2397 * @param xstat_name The name of the statistic to return
2398 * @param[out] id A pointer to an app-supplied uint64_t which should be
2399 * set to the ID of the stat if the stat exists.
2402 * -ENODEV for invalid port_id,
2403 * -EINVAL if the xstat_name doesn't exist in port_id
2405 int rte_eth_xstats_get_id_by_name(uint8_t port_id, const char *xstat_name,
2409 * Reset extended statistics of an Ethernet device.
2412 * The port identifier of the Ethernet device.
2414 void rte_eth_xstats_reset(uint8_t port_id);
2417 * Set a mapping for the specified transmit queue to the specified per-queue
2418 * statistics counter.
2421 * The port identifier of the Ethernet device.
2422 * @param tx_queue_id
2423 * The index of the transmit queue for which a queue stats mapping is required.
2424 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2425 * to rte_eth_dev_configure().
2427 * The per-queue packet statistics functionality number that the transmit
2428 * queue is to be assigned.
2429 * The value must be in the range [0, RTE_MAX_ETHPORT_QUEUE_STATS_MAPS - 1].
2431 * Zero if successful. Non-zero otherwise.
2433 int rte_eth_dev_set_tx_queue_stats_mapping(uint8_t port_id,
2434 uint16_t tx_queue_id, uint8_t stat_idx);
2437 * Set a mapping for the specified receive queue to the specified per-queue
2438 * statistics counter.
2441 * The port identifier of the Ethernet device.
2442 * @param rx_queue_id
2443 * The index of the receive queue for which a queue stats mapping is required.
2444 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2445 * to rte_eth_dev_configure().
2447 * The per-queue packet statistics functionality number that the receive
2448 * queue is to be assigned.
2449 * The value must be in the range [0, RTE_MAX_ETHPORT_QUEUE_STATS_MAPS - 1].
2451 * Zero if successful. Non-zero otherwise.
2453 int rte_eth_dev_set_rx_queue_stats_mapping(uint8_t port_id,
2454 uint16_t rx_queue_id,
2458 * Retrieve the Ethernet address of an Ethernet device.
2461 * The port identifier of the Ethernet device.
2463 * A pointer to a structure of type *ether_addr* to be filled with
2464 * the Ethernet address of the Ethernet device.
2466 void rte_eth_macaddr_get(uint8_t port_id, struct ether_addr *mac_addr);
2469 * Retrieve the contextual information of an Ethernet device.
2472 * The port identifier of the Ethernet device.
2474 * A pointer to a structure of type *rte_eth_dev_info* to be filled with
2475 * the contextual information of the Ethernet device.
2477 void rte_eth_dev_info_get(uint8_t port_id, struct rte_eth_dev_info *dev_info);
2480 * Retrieve the firmware version of a device.
2483 * The port identifier of the device.
2485 * A pointer to a string array storing the firmware version of a device,
2486 * the string includes terminating null. This pointer is allocated by caller.
2488 * The size of the string array pointed by fw_version, which should be
2489 * large enough to store firmware version of the device.
2491 * - (0) if successful.
2492 * - (-ENOTSUP) if operation is not supported.
2493 * - (-ENODEV) if *port_id* invalid.
2494 * - (>0) if *fw_size* is not enough to store firmware version, return
2495 * the size of the non truncated string.
2497 int rte_eth_dev_fw_version_get(uint8_t port_id,
2498 char *fw_version, size_t fw_size);
2501 * Retrieve the supported packet types of an Ethernet device.
2503 * When a packet type is announced as supported, it *must* be recognized by
2504 * the PMD. For instance, if RTE_PTYPE_L2_ETHER, RTE_PTYPE_L2_ETHER_VLAN
2505 * and RTE_PTYPE_L3_IPV4 are announced, the PMD must return the following
2506 * packet types for these packets:
2507 * - Ether/IPv4 -> RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4
2508 * - Ether/Vlan/IPv4 -> RTE_PTYPE_L2_ETHER_VLAN | RTE_PTYPE_L3_IPV4
2509 * - Ether/[anything else] -> RTE_PTYPE_L2_ETHER
2510 * - Ether/Vlan/[anything else] -> RTE_PTYPE_L2_ETHER_VLAN
2512 * When a packet is received by a PMD, the most precise type must be
2513 * returned among the ones supported. However a PMD is allowed to set
2514 * packet type that is not in the supported list, at the condition that it
2515 * is more precise. Therefore, a PMD announcing no supported packet types
2516 * can still set a matching packet type in a received packet.
2519 * Better to invoke this API after the device is already started or rx burst
2520 * function is decided, to obtain correct supported ptypes.
2522 * if a given PMD does not report what ptypes it supports, then the supported
2523 * ptype count is reported as 0.
2525 * The port identifier of the Ethernet device.
2527 * A hint of what kind of packet type which the caller is interested in.
2529 * An array pointer to store adequate packet types, allocated by caller.
2531 * Size of the array pointed by param ptypes.
2533 * - (>=0) Number of supported ptypes. If the number of types exceeds num,
2534 * only num entries will be filled into the ptypes array, but the full
2535 * count of supported ptypes will be returned.
2536 * - (-ENODEV) if *port_id* invalid.
2538 int rte_eth_dev_get_supported_ptypes(uint8_t port_id, uint32_t ptype_mask,
2539 uint32_t *ptypes, int num);
2542 * Retrieve the MTU of an Ethernet device.
2545 * The port identifier of the Ethernet device.
2547 * A pointer to a uint16_t where the retrieved MTU is to be stored.
2549 * - (0) if successful.
2550 * - (-ENODEV) if *port_id* invalid.
2552 int rte_eth_dev_get_mtu(uint8_t port_id, uint16_t *mtu);
2555 * Change the MTU of an Ethernet device.
2558 * The port identifier of the Ethernet device.
2560 * A uint16_t for the MTU to be applied.
2562 * - (0) if successful.
2563 * - (-ENOTSUP) if operation is not supported.
2564 * - (-ENODEV) if *port_id* invalid.
2565 * - (-EINVAL) if *mtu* invalid.
2566 * - (-EBUSY) if operation is not allowed when the port is running
2568 int rte_eth_dev_set_mtu(uint8_t port_id, uint16_t mtu);
2571 * Enable/Disable hardware filtering by an Ethernet device of received
2572 * VLAN packets tagged with a given VLAN Tag Identifier.
2575 * The port identifier of the Ethernet device.
2577 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
2579 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*.
2580 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*.
2582 * - (0) if successful.
2583 * - (-ENOSUP) if hardware-assisted VLAN filtering not configured.
2584 * - (-ENODEV) if *port_id* invalid.
2585 * - (-ENOSYS) if VLAN filtering on *port_id* disabled.
2586 * - (-EINVAL) if *vlan_id* > 4095.
2588 int rte_eth_dev_vlan_filter(uint8_t port_id, uint16_t vlan_id, int on);
2591 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device.
2592 * 82599/X540/X550 can support VLAN stripping at the rx queue level
2595 * The port identifier of the Ethernet device.
2596 * @param rx_queue_id
2597 * The index of the receive queue for which a queue stats mapping is required.
2598 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2599 * to rte_eth_dev_configure().
2601 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port.
2602 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port.
2604 * - (0) if successful.
2605 * - (-ENOSUP) if hardware-assisted VLAN stripping not configured.
2606 * - (-ENODEV) if *port_id* invalid.
2607 * - (-EINVAL) if *rx_queue_id* invalid.
2609 int rte_eth_dev_set_vlan_strip_on_queue(uint8_t port_id, uint16_t rx_queue_id,
2613 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to
2614 * the VLAN Header. This is a register setup available on some Intel NIC, not
2615 * but all, please check the data sheet for availability.
2618 * The port identifier of the Ethernet device.
2622 * The Tag Protocol ID
2624 * - (0) if successful.
2625 * - (-ENOSUP) if hardware-assisted VLAN TPID setup is not supported.
2626 * - (-ENODEV) if *port_id* invalid.
2628 int rte_eth_dev_set_vlan_ether_type(uint8_t port_id,
2629 enum rte_vlan_type vlan_type,
2633 * Set VLAN offload configuration on an Ethernet device
2634 * Enable/Disable Extended VLAN by an Ethernet device, This is a register setup
2635 * available on some Intel NIC, not but all, please check the data sheet for
2637 * Enable/Disable VLAN Strip can be done on rx queue for certain NIC, but here
2638 * the configuration is applied on the port level.
2641 * The port identifier of the Ethernet device.
2642 * @param offload_mask
2643 * The VLAN Offload bit mask can be mixed use with "OR"
2644 * ETH_VLAN_STRIP_OFFLOAD
2645 * ETH_VLAN_FILTER_OFFLOAD
2646 * ETH_VLAN_EXTEND_OFFLOAD
2648 * - (0) if successful.
2649 * - (-ENOSUP) if hardware-assisted VLAN filtering not configured.
2650 * - (-ENODEV) if *port_id* invalid.
2652 int rte_eth_dev_set_vlan_offload(uint8_t port_id, int offload_mask);
2655 * Read VLAN Offload configuration from an Ethernet device
2658 * The port identifier of the Ethernet device.
2660 * - (>0) if successful. Bit mask to indicate
2661 * ETH_VLAN_STRIP_OFFLOAD
2662 * ETH_VLAN_FILTER_OFFLOAD
2663 * ETH_VLAN_EXTEND_OFFLOAD
2664 * - (-ENODEV) if *port_id* invalid.
2666 int rte_eth_dev_get_vlan_offload(uint8_t port_id);
2669 * Set port based TX VLAN insertion on or off.
2672 * The port identifier of the Ethernet device.
2674 * Port based TX VLAN identifier together with user priority.
2676 * Turn on or off the port based TX VLAN insertion.
2679 * - (0) if successful.
2680 * - negative if failed.
2682 int rte_eth_dev_set_vlan_pvid(uint8_t port_id, uint16_t pvid, int on);
2686 * Retrieve a burst of input packets from a receive queue of an Ethernet
2687 * device. The retrieved packets are stored in *rte_mbuf* structures whose
2688 * pointers are supplied in the *rx_pkts* array.
2690 * The rte_eth_rx_burst() function loops, parsing the RX ring of the
2691 * receive queue, up to *nb_pkts* packets, and for each completed RX
2692 * descriptor in the ring, it performs the following operations:
2694 * - Initialize the *rte_mbuf* data structure associated with the
2695 * RX descriptor according to the information provided by the NIC into
2696 * that RX descriptor.
2698 * - Store the *rte_mbuf* data structure into the next entry of the
2701 * - Replenish the RX descriptor with a new *rte_mbuf* buffer
2702 * allocated from the memory pool associated with the receive queue at
2703 * initialization time.
2705 * When retrieving an input packet that was scattered by the controller
2706 * into multiple receive descriptors, the rte_eth_rx_burst() function
2707 * appends the associated *rte_mbuf* buffers to the first buffer of the
2710 * The rte_eth_rx_burst() function returns the number of packets
2711 * actually retrieved, which is the number of *rte_mbuf* data structures
2712 * effectively supplied into the *rx_pkts* array.
2713 * A return value equal to *nb_pkts* indicates that the RX queue contained
2714 * at least *rx_pkts* packets, and this is likely to signify that other
2715 * received packets remain in the input queue. Applications implementing
2716 * a "retrieve as much received packets as possible" policy can check this
2717 * specific case and keep invoking the rte_eth_rx_burst() function until
2718 * a value less than *nb_pkts* is returned.
2720 * This receive method has the following advantages:
2722 * - It allows a run-to-completion network stack engine to retrieve and
2723 * to immediately process received packets in a fast burst-oriented
2724 * approach, avoiding the overhead of unnecessary intermediate packet
2725 * queue/dequeue operations.
2727 * - Conversely, it also allows an asynchronous-oriented processing
2728 * method to retrieve bursts of received packets and to immediately
2729 * queue them for further parallel processing by another logical core,
2730 * for instance. However, instead of having received packets being
2731 * individually queued by the driver, this approach allows the caller
2732 * of the rte_eth_rx_burst() function to queue a burst of retrieved
2733 * packets at a time and therefore dramatically reduce the cost of
2734 * enqueue/dequeue operations per packet.
2736 * - It allows the rte_eth_rx_burst() function of the driver to take
2737 * advantage of burst-oriented hardware features (CPU cache,
2738 * prefetch instructions, and so on) to minimize the number of CPU
2739 * cycles per packet.
2741 * To summarize, the proposed receive API enables many
2742 * burst-oriented optimizations in both synchronous and asynchronous
2743 * packet processing environments with no overhead in both cases.
2745 * The rte_eth_rx_burst() function does not provide any error
2746 * notification to avoid the corresponding overhead. As a hint, the
2747 * upper-level application might check the status of the device link once
2748 * being systematically returned a 0 value for a given number of tries.
2751 * The port identifier of the Ethernet device.
2753 * The index of the receive queue from which to retrieve input packets.
2754 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2755 * to rte_eth_dev_configure().
2757 * The address of an array of pointers to *rte_mbuf* structures that
2758 * must be large enough to store *nb_pkts* pointers in it.
2760 * The maximum number of packets to retrieve.
2762 * The number of packets actually retrieved, which is the number
2763 * of pointers to *rte_mbuf* structures effectively supplied to the
2766 static inline uint16_t
2767 rte_eth_rx_burst(uint8_t port_id, uint16_t queue_id,
2768 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
2770 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
2772 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2773 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
2774 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_pkt_burst, 0);
2776 if (queue_id >= dev->data->nb_rx_queues) {
2777 RTE_PMD_DEBUG_TRACE("Invalid RX queue_id=%d\n", queue_id);
2781 int16_t nb_rx = (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id],
2784 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
2785 struct rte_eth_rxtx_callback *cb = dev->post_rx_burst_cbs[queue_id];
2787 if (unlikely(cb != NULL)) {
2789 nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
2790 nb_pkts, cb->param);
2792 } while (cb != NULL);
2800 * Get the number of used descriptors of a rx queue
2803 * The port identifier of the Ethernet device.
2805 * The queue id on the specific port.
2807 * The number of used descriptors in the specific queue, or:
2808 * (-EINVAL) if *port_id* or *queue_id* is invalid
2809 * (-ENOTSUP) if the device does not support this function
2812 rte_eth_rx_queue_count(uint8_t port_id, uint16_t queue_id)
2814 struct rte_eth_dev *dev;
2816 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -EINVAL);
2817 dev = &rte_eth_devices[port_id];
2818 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_queue_count, -ENOTSUP);
2819 if (queue_id >= dev->data->nb_rx_queues)
2822 return (*dev->dev_ops->rx_queue_count)(dev, queue_id);
2826 * Check if the DD bit of the specific RX descriptor in the queue has been set
2829 * The port identifier of the Ethernet device.
2831 * The queue id on the specific port.
2833 * The offset of the descriptor ID from tail.
2835 * - (1) if the specific DD bit is set.
2836 * - (0) if the specific DD bit is not set.
2837 * - (-ENODEV) if *port_id* invalid.
2838 * - (-ENOTSUP) if the device does not support this function
2841 rte_eth_rx_descriptor_done(uint8_t port_id, uint16_t queue_id, uint16_t offset)
2843 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
2844 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
2845 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_descriptor_done, -ENOTSUP);
2846 return (*dev->dev_ops->rx_descriptor_done)( \
2847 dev->data->rx_queues[queue_id], offset);
2850 #define RTE_ETH_RX_DESC_AVAIL 0 /**< Desc available for hw. */
2851 #define RTE_ETH_RX_DESC_DONE 1 /**< Desc done, filled by hw. */
2852 #define RTE_ETH_RX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
2855 * Check the status of a Rx descriptor in the queue
2857 * It should be called in a similar context than the Rx function:
2858 * - on a dataplane core
2859 * - not concurrently on the same queue
2861 * Since it's a dataplane function, no check is performed on port_id and
2862 * queue_id. The caller must therefore ensure that the port is enabled
2863 * and the queue is configured and running.
2865 * Note: accessing to a random descriptor in the ring may trigger cache
2866 * misses and have a performance impact.
2869 * A valid port identifier of the Ethernet device which.
2871 * A valid Rx queue identifier on this port.
2873 * The offset of the descriptor starting from tail (0 is the next
2874 * packet to be received by the driver).
2877 * - (RTE_ETH_RX_DESC_AVAIL): Descriptor is available for the hardware to
2879 * - (RTE_ETH_RX_DESC_DONE): Descriptor is done, it is filled by hw, but
2880 * not yet processed by the driver (i.e. in the receive queue).
2881 * - (RTE_ETH_RX_DESC_UNAVAIL): Descriptor is unavailable, either hold by
2882 * the driver and not yet returned to hw, or reserved by the hw.
2883 * - (-EINVAL) bad descriptor offset.
2884 * - (-ENOTSUP) if the device does not support this function.
2885 * - (-ENODEV) bad port or queue (only if compiled with debug).
2888 rte_eth_rx_descriptor_status(uint8_t port_id, uint16_t queue_id,
2891 struct rte_eth_dev *dev;
2894 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2895 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
2897 dev = &rte_eth_devices[port_id];
2898 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2899 if (queue_id >= dev->data->nb_rx_queues)
2902 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->rx_descriptor_status, -ENOTSUP);
2903 rxq = dev->data->rx_queues[queue_id];
2905 return (*dev->dev_ops->rx_descriptor_status)(rxq, offset);
2908 #define RTE_ETH_TX_DESC_FULL 0 /**< Desc filled for hw, waiting xmit. */
2909 #define RTE_ETH_TX_DESC_DONE 1 /**< Desc done, packet is transmitted. */
2910 #define RTE_ETH_TX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */
2913 * Check the status of a Tx descriptor in the queue.
2915 * It should be called in a similar context than the Tx function:
2916 * - on a dataplane core
2917 * - not concurrently on the same queue
2919 * Since it's a dataplane function, no check is performed on port_id and
2920 * queue_id. The caller must therefore ensure that the port is enabled
2921 * and the queue is configured and running.
2923 * Note: accessing to a random descriptor in the ring may trigger cache
2924 * misses and have a performance impact.
2927 * A valid port identifier of the Ethernet device which.
2929 * A valid Tx queue identifier on this port.
2931 * The offset of the descriptor starting from tail (0 is the place where
2932 * the next packet will be send).
2935 * - (RTE_ETH_TX_DESC_FULL) Descriptor is being processed by the hw, i.e.
2936 * in the transmit queue.
2937 * - (RTE_ETH_TX_DESC_DONE) Hardware is done with this descriptor, it can
2938 * be reused by the driver.
2939 * - (RTE_ETH_TX_DESC_UNAVAIL): Descriptor is unavailable, reserved by the
2940 * driver or the hardware.
2941 * - (-EINVAL) bad descriptor offset.
2942 * - (-ENOTSUP) if the device does not support this function.
2943 * - (-ENODEV) bad port or queue (only if compiled with debug).
2945 static inline int rte_eth_tx_descriptor_status(uint8_t port_id,
2946 uint16_t queue_id, uint16_t offset)
2948 struct rte_eth_dev *dev;
2951 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2952 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
2954 dev = &rte_eth_devices[port_id];
2955 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2956 if (queue_id >= dev->data->nb_tx_queues)
2959 RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->tx_descriptor_status, -ENOTSUP);
2960 txq = dev->data->tx_queues[queue_id];
2962 return (*dev->dev_ops->tx_descriptor_status)(txq, offset);
2966 * Send a burst of output packets on a transmit queue of an Ethernet device.
2968 * The rte_eth_tx_burst() function is invoked to transmit output packets
2969 * on the output queue *queue_id* of the Ethernet device designated by its
2971 * The *nb_pkts* parameter is the number of packets to send which are
2972 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
2973 * allocated from a pool created with rte_pktmbuf_pool_create().
2974 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets,
2975 * up to the number of transmit descriptors available in the TX ring of the
2977 * For each packet to send, the rte_eth_tx_burst() function performs
2978 * the following operations:
2980 * - Pick up the next available descriptor in the transmit ring.
2982 * - Free the network buffer previously sent with that descriptor, if any.
2984 * - Initialize the transmit descriptor with the information provided
2985 * in the *rte_mbuf data structure.
2987 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers,
2988 * the rte_eth_tx_burst() function uses several transmit descriptors
2991 * The rte_eth_tx_burst() function returns the number of packets it
2992 * actually sent. A return value equal to *nb_pkts* means that all packets
2993 * have been sent, and this is likely to signify that other output packets
2994 * could be immediately transmitted again. Applications that implement a
2995 * "send as many packets to transmit as possible" policy can check this
2996 * specific case and keep invoking the rte_eth_tx_burst() function until
2997 * a value less than *nb_pkts* is returned.
2999 * It is the responsibility of the rte_eth_tx_burst() function to
3000 * transparently free the memory buffers of packets previously sent.
3001 * This feature is driven by the *tx_free_thresh* value supplied to the
3002 * rte_eth_dev_configure() function at device configuration time.
3003 * When the number of free TX descriptors drops below this threshold, the
3004 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers
3005 * of those packets whose transmission was effectively completed.
3007 * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can
3008 * invoke this function concurrently on the same tx queue without SW lock.
3009 * @see rte_eth_dev_info_get, struct rte_eth_txconf::txq_flags
3012 * The port identifier of the Ethernet device.
3014 * The index of the transmit queue through which output packets must be
3016 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
3017 * to rte_eth_dev_configure().
3019 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
3020 * which contain the output packets.
3022 * The maximum number of packets to transmit.
3024 * The number of output packets actually stored in transmit descriptors of
3025 * the transmit ring. The return value can be less than the value of the
3026 * *tx_pkts* parameter when the transmit ring is full or has been filled up.
3028 static inline uint16_t
3029 rte_eth_tx_burst(uint8_t port_id, uint16_t queue_id,
3030 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
3032 struct rte_eth_dev *dev = &rte_eth_devices[port_id];
3034 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
3035 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0);
3036 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_pkt_burst, 0);
3038 if (queue_id >= dev->data->nb_tx_queues) {
3039 RTE_PMD_DEBUG_TRACE("Invalid TX queue_id=%d\n", queue_id);
3044 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
3045 struct rte_eth_rxtx_callback *cb = dev->pre_tx_burst_cbs[queue_id];
3047 if (unlikely(cb != NULL)) {
3049 nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
3052 } while (cb != NULL);
3056 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts);
3061 * @b EXPERIMENTAL: this API may change without prior notice
3063 * Process a burst of output packets on a transmit queue of an Ethernet device.
3065 * The rte_eth_tx_prepare() function is invoked to prepare output packets to be
3066 * transmitted on the output queue *queue_id* of the Ethernet device designated
3068 * The *nb_pkts* parameter is the number of packets to be prepared which are
3069 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them
3070 * allocated from a pool created with rte_pktmbuf_pool_create().
3071 * For each packet to send, the rte_eth_tx_prepare() function performs
3072 * the following operations:
3074 * - Check if packet meets devices requirements for tx offloads.
3076 * - Check limitations about number of segments.
3078 * - Check additional requirements when debug is enabled.
3080 * - Update and/or reset required checksums when tx offload is set for packet.
3082 * Since this function can modify packet data, provided mbufs must be safely
3083 * writable (e.g. modified data cannot be in shared segment).
3085 * The rte_eth_tx_prepare() function returns the number of packets ready to be
3086 * sent. A return value equal to *nb_pkts* means that all packets are valid and
3087 * ready to be sent, otherwise stops processing on the first invalid packet and
3088 * leaves the rest packets untouched.
3090 * When this functionality is not implemented in the driver, all packets are
3091 * are returned untouched.
3094 * The port identifier of the Ethernet device.
3095 * The value must be a valid port id.
3097 * The index of the transmit queue through which output packets must be
3099 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
3100 * to rte_eth_dev_configure().
3102 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
3103 * which contain the output packets.
3105 * The maximum number of packets to process.
3107 * The number of packets correct and ready to be sent. The return value can be
3108 * less than the value of the *tx_pkts* parameter when some packet doesn't
3109 * meet devices requirements with rte_errno set appropriately:
3110 * - -EINVAL: offload flags are not correctly set
3111 * - -ENOTSUP: the offload feature is not supported by the hardware
3115 #ifndef RTE_ETHDEV_TX_PREPARE_NOOP
3117 static inline uint16_t
3118 rte_eth_tx_prepare(uint8_t port_id, uint16_t queue_id,
3119 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
3121 struct rte_eth_dev *dev;
3123 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
3124 if (!rte_eth_dev_is_valid_port(port_id)) {
3125 RTE_PMD_DEBUG_TRACE("Invalid TX port_id=%d\n", port_id);
3126 rte_errno = -EINVAL;
3131 dev = &rte_eth_devices[port_id];
3133 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
3134 if (queue_id >= dev->data->nb_tx_queues) {
3135 RTE_PMD_DEBUG_TRACE("Invalid TX queue_id=%d\n", queue_id);
3136 rte_errno = -EINVAL;
3141 if (!dev->tx_pkt_prepare)
3144 return (*dev->tx_pkt_prepare)(dev->data->tx_queues[queue_id],
3151 * Native NOOP operation for compilation targets which doesn't require any
3152 * preparations steps, and functional NOOP may introduce unnecessary performance
3155 * Generally this is not a good idea to turn it on globally and didn't should
3156 * be used if behavior of tx_preparation can change.
3159 static inline uint16_t
3160 rte_eth_tx_prepare(__rte_unused uint8_t port_id, __rte_unused uint16_t queue_id,
3161 __rte_unused struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
3168 typedef void (*buffer_tx_error_fn)(struct rte_mbuf **unsent, uint16_t count,
3172 * Structure used to buffer packets for future TX
3173 * Used by APIs rte_eth_tx_buffer and rte_eth_tx_buffer_flush
3175 struct rte_eth_dev_tx_buffer {
3176 buffer_tx_error_fn error_callback;
3177 void *error_userdata;
3178 uint16_t size; /**< Size of buffer for buffered tx */
3179 uint16_t length; /**< Number of packets in the array */
3180 struct rte_mbuf *pkts[];
3181 /**< Pending packets to be sent on explicit flush or when full */
3185 * Calculate the size of the tx buffer.
3188 * Number of stored packets.
3190 #define RTE_ETH_TX_BUFFER_SIZE(sz) \
3191 (sizeof(struct rte_eth_dev_tx_buffer) + (sz) * sizeof(struct rte_mbuf *))
3194 * Initialize default values for buffered transmitting
3197 * Tx buffer to be initialized.
3204 rte_eth_tx_buffer_init(struct rte_eth_dev_tx_buffer *buffer, uint16_t size);
3207 * Send any packets queued up for transmission on a port and HW queue
3209 * This causes an explicit flush of packets previously buffered via the
3210 * rte_eth_tx_buffer() function. It returns the number of packets successfully
3211 * sent to the NIC, and calls the error callback for any unsent packets. Unless
3212 * explicitly set up otherwise, the default callback simply frees the unsent
3213 * packets back to the owning mempool.
3216 * The port identifier of the Ethernet device.
3218 * The index of the transmit queue through which output packets must be
3220 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
3221 * to rte_eth_dev_configure().
3223 * Buffer of packets to be transmit.
3225 * The number of packets successfully sent to the Ethernet device. The error
3226 * callback is called for any packets which could not be sent.
3228 static inline uint16_t
3229 rte_eth_tx_buffer_flush(uint8_t port_id, uint16_t queue_id,
3230 struct rte_eth_dev_tx_buffer *buffer)
3233 uint16_t to_send = buffer->length;
3238 sent = rte_eth_tx_burst(port_id, queue_id, buffer->pkts, to_send);
3242 /* All packets sent, or to be dealt with by callback below */
3243 if (unlikely(sent != to_send))
3244 buffer->error_callback(&buffer->pkts[sent], to_send - sent,
3245 buffer->error_userdata);
3251 * Buffer a single packet for future transmission on a port and queue
3253 * This function takes a single mbuf/packet and buffers it for later
3254 * transmission on the particular port and queue specified. Once the buffer is
3255 * full of packets, an attempt will be made to transmit all the buffered
3256 * packets. In case of error, where not all packets can be transmitted, a
3257 * callback is called with the unsent packets as a parameter. If no callback
3258 * is explicitly set up, the unsent packets are just freed back to the owning
3259 * mempool. The function returns the number of packets actually sent i.e.
3260 * 0 if no buffer flush occurred, otherwise the number of packets successfully
3264 * The port identifier of the Ethernet device.
3266 * The index of the transmit queue through which output packets must be
3268 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
3269 * to rte_eth_dev_configure().
3271 * Buffer used to collect packets to be sent.
3273 * Pointer to the packet mbuf to be sent.
3275 * 0 = packet has been buffered for later transmission
3276 * N > 0 = packet has been buffered, and the buffer was subsequently flushed,
3277 * causing N packets to be sent, and the error callback to be called for
3280 static __rte_always_inline uint16_t
3281 rte_eth_tx_buffer(uint8_t port_id, uint16_t queue_id,
3282 struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt)
3284 buffer->pkts[buffer->length++] = tx_pkt;
3285 if (buffer->length < buffer->size)
3288 return rte_eth_tx_buffer_flush(port_id, queue_id, buffer);
3292 * Configure a callback for buffered packets which cannot be sent
3294 * Register a specific callback to be called when an attempt is made to send
3295 * all packets buffered on an ethernet port, but not all packets can
3296 * successfully be sent. The callback registered here will be called only
3297 * from calls to rte_eth_tx_buffer() and rte_eth_tx_buffer_flush() APIs.
3298 * The default callback configured for each queue by default just frees the
3299 * packets back to the calling mempool. If additional behaviour is required,
3300 * for example, to count dropped packets, or to retry transmission of packets
3301 * which cannot be sent, this function should be used to register a suitable
3302 * callback function to implement the desired behaviour.
3303 * The example callback "rte_eth_count_unsent_packet_callback()" is also
3304 * provided as reference.
3307 * The port identifier of the Ethernet device.
3309 * The function to be used as the callback.
3311 * Arbitrary parameter to be passed to the callback function
3313 * 0 on success, or -1 on error with rte_errno set appropriately
3316 rte_eth_tx_buffer_set_err_callback(struct rte_eth_dev_tx_buffer *buffer,
3317 buffer_tx_error_fn callback, void *userdata);
3320 * Callback function for silently dropping unsent buffered packets.
3322 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
3323 * adjust the default behavior when buffered packets cannot be sent. This
3324 * function drops any unsent packets silently and is used by tx buffered
3325 * operations as default behavior.
3327 * NOTE: this function should not be called directly, instead it should be used
3328 * as a callback for packet buffering.
3330 * NOTE: when configuring this function as a callback with
3331 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
3332 * should point to an uint64_t value.
3335 * The previously buffered packets which could not be sent
3337 * The number of unsent packets in the pkts array
3342 rte_eth_tx_buffer_drop_callback(struct rte_mbuf **pkts, uint16_t unsent,
3346 * Callback function for tracking unsent buffered packets.
3348 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to
3349 * adjust the default behavior when buffered packets cannot be sent. This
3350 * function drops any unsent packets, but also updates a user-supplied counter
3351 * to track the overall number of packets dropped. The counter should be an
3352 * uint64_t variable.
3354 * NOTE: this function should not be called directly, instead it should be used
3355 * as a callback for packet buffering.
3357 * NOTE: when configuring this function as a callback with
3358 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter
3359 * should point to an uint64_t value.
3362 * The previously buffered packets which could not be sent
3364 * The number of unsent packets in the pkts array
3366 * Pointer to an uint64_t value, which will be incremented by unsent
3369 rte_eth_tx_buffer_count_callback(struct rte_mbuf **pkts, uint16_t unsent,
3373 * Request the driver to free mbufs currently cached by the driver. The
3374 * driver will only free the mbuf if it is no longer in use. It is the
3375 * application's responsibity to ensure rte_eth_tx_buffer_flush(..) is
3379 * The port identifier of the Ethernet device.
3381 * The index of the transmit queue through which output packets must be
3383 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
3384 * to rte_eth_dev_configure().
3386 * Maximum number of packets to free. Use 0 to indicate all possible packets
3387 * should be freed. Note that a packet may be using multiple mbufs.
3390 * -ENODEV: Invalid interface
3391 * -ENOTSUP: Driver does not support function
3393 * 0-n: Number of packets freed. More packets may still remain in ring that
3397 rte_eth_tx_done_cleanup(uint8_t port_id, uint16_t queue_id, uint32_t free_cnt);
3400 * The eth device event type for interrupt, and maybe others in the future.
3402 enum rte_eth_event_type {
3403 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */
3404 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */
3405 RTE_ETH_EVENT_QUEUE_STATE,
3406 /**< queue state event (enabled/disabled) */
3407 RTE_ETH_EVENT_INTR_RESET,
3408 /**< reset interrupt event, sent to VF on PF reset */
3409 RTE_ETH_EVENT_VF_MBOX, /**< message from the VF received by PF */
3410 RTE_ETH_EVENT_MACSEC, /**< MACsec offload related event */
3411 RTE_ETH_EVENT_INTR_RMV, /**< device removal event */
3412 RTE_ETH_EVENT_MAX /**< max value of this enum */
3415 typedef int (*rte_eth_dev_cb_fn)(uint8_t port_id,
3416 enum rte_eth_event_type event, void *cb_arg, void *ret_param);
3417 /**< user application callback to be registered for interrupts */
3422 * Register a callback function for specific port id.
3429 * User supplied callback function to be called.
3431 * Pointer to the parameters for the registered callback.
3434 * - On success, zero.
3435 * - On failure, a negative value.
3437 int rte_eth_dev_callback_register(uint8_t port_id,
3438 enum rte_eth_event_type event,
3439 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3442 * Unregister a callback function for specific port id.
3449 * User supplied callback function to be called.
3451 * Pointer to the parameters for the registered callback. -1 means to
3452 * remove all for the same callback address and same event.
3455 * - On success, zero.
3456 * - On failure, a negative value.
3458 int rte_eth_dev_callback_unregister(uint8_t port_id,
3459 enum rte_eth_event_type event,
3460 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
3463 * @internal Executes all the user application registered callbacks for
3464 * the specific device. It is for DPDK internal user only. User
3465 * application should not call it directly.
3468 * Pointer to struct rte_eth_dev.
3470 * Eth device interrupt event type.
3472 * callback parameter.
3474 * To pass data back to user application.
3475 * This allows the user application to decide if a particular function
3476 * is permitted or not.
3481 int _rte_eth_dev_callback_process(struct rte_eth_dev *dev,
3482 enum rte_eth_event_type event, void *cb_arg, void *ret_param);
3485 * When there is no rx packet coming in Rx Queue for a long time, we can
3486 * sleep lcore related to RX Queue for power saving, and enable rx interrupt
3487 * to be triggered when Rx packet arrives.
3489 * The rte_eth_dev_rx_intr_enable() function enables rx queue
3490 * interrupt on specific rx queue of a port.
3493 * The port identifier of the Ethernet device.
3495 * The index of the receive queue from which to retrieve input packets.
3496 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3497 * to rte_eth_dev_configure().
3499 * - (0) if successful.
3500 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3502 * - (-ENODEV) if *port_id* invalid.
3504 int rte_eth_dev_rx_intr_enable(uint8_t port_id, uint16_t queue_id);
3507 * When lcore wakes up from rx interrupt indicating packet coming, disable rx
3508 * interrupt and returns to polling mode.
3510 * The rte_eth_dev_rx_intr_disable() function disables rx queue
3511 * interrupt on specific rx queue of a port.
3514 * The port identifier of the Ethernet device.
3516 * The index of the receive queue from which to retrieve input packets.
3517 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3518 * to rte_eth_dev_configure().
3520 * - (0) if successful.
3521 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3523 * - (-ENODEV) if *port_id* invalid.
3525 int rte_eth_dev_rx_intr_disable(uint8_t port_id, uint16_t queue_id);
3528 * RX Interrupt control per port.
3531 * The port identifier of the Ethernet device.
3533 * Epoll instance fd which the intr vector associated to.
3534 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3536 * The operation be performed for the vector.
3537 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3541 * - On success, zero.
3542 * - On failure, a negative value.
3544 int rte_eth_dev_rx_intr_ctl(uint8_t port_id, int epfd, int op, void *data);
3547 * RX Interrupt control per queue.
3550 * The port identifier of the Ethernet device.
3552 * The index of the receive queue from which to retrieve input packets.
3553 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
3554 * to rte_eth_dev_configure().
3556 * Epoll instance fd which the intr vector associated to.
3557 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance.
3559 * The operation be performed for the vector.
3560 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}.
3564 * - On success, zero.
3565 * - On failure, a negative value.
3567 int rte_eth_dev_rx_intr_ctl_q(uint8_t port_id, uint16_t queue_id,
3568 int epfd, int op, void *data);
3571 * Turn on the LED on the Ethernet device.
3572 * This function turns on the LED on the Ethernet device.
3575 * The port identifier of the Ethernet device.
3577 * - (0) if successful.
3578 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3580 * - (-ENODEV) if *port_id* invalid.
3582 int rte_eth_led_on(uint8_t port_id);
3585 * Turn off the LED on the Ethernet device.
3586 * This function turns off the LED on the Ethernet device.
3589 * The port identifier of the Ethernet device.
3591 * - (0) if successful.
3592 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
3594 * - (-ENODEV) if *port_id* invalid.
3596 int rte_eth_led_off(uint8_t port_id);
3599 * Get current status of the Ethernet link flow control for Ethernet device
3602 * The port identifier of the Ethernet device.
3604 * The pointer to the structure where to store the flow control parameters.
3606 * - (0) if successful.
3607 * - (-ENOTSUP) if hardware doesn't support flow control.
3608 * - (-ENODEV) if *port_id* invalid.
3610 int rte_eth_dev_flow_ctrl_get(uint8_t port_id,
3611 struct rte_eth_fc_conf *fc_conf);
3614 * Configure the Ethernet link flow control for Ethernet device
3617 * The port identifier of the Ethernet device.
3619 * The pointer to the structure of the flow control parameters.
3621 * - (0) if successful.
3622 * - (-ENOTSUP) if hardware doesn't support flow control mode.
3623 * - (-ENODEV) if *port_id* invalid.
3624 * - (-EINVAL) if bad parameter
3625 * - (-EIO) if flow control setup failure
3627 int rte_eth_dev_flow_ctrl_set(uint8_t port_id,
3628 struct rte_eth_fc_conf *fc_conf);
3631 * Configure the Ethernet priority flow control under DCB environment
3632 * for Ethernet device.
3635 * The port identifier of the Ethernet device.
3637 * The pointer to the structure of the priority flow control parameters.
3639 * - (0) if successful.
3640 * - (-ENOTSUP) if hardware doesn't support priority flow control mode.
3641 * - (-ENODEV) if *port_id* invalid.
3642 * - (-EINVAL) if bad parameter
3643 * - (-EIO) if flow control setup failure
3645 int rte_eth_dev_priority_flow_ctrl_set(uint8_t port_id,
3646 struct rte_eth_pfc_conf *pfc_conf);
3649 * Add a MAC address to an internal array of addresses used to enable whitelist
3650 * filtering to accept packets only if the destination MAC address matches.
3653 * The port identifier of the Ethernet device.
3655 * The MAC address to add.
3657 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is
3658 * not enabled, this should be set to 0.
3660 * - (0) if successfully added or *mac_addr" was already added.
3661 * - (-ENOTSUP) if hardware doesn't support this feature.
3662 * - (-ENODEV) if *port* is invalid.
3663 * - (-ENOSPC) if no more MAC addresses can be added.
3664 * - (-EINVAL) if MAC address is invalid.
3666 int rte_eth_dev_mac_addr_add(uint8_t port, struct ether_addr *mac_addr,
3670 * Remove a MAC address from the internal array of addresses.
3673 * The port identifier of the Ethernet device.
3675 * MAC address to remove.
3677 * - (0) if successful, or *mac_addr* didn't exist.
3678 * - (-ENOTSUP) if hardware doesn't support.
3679 * - (-ENODEV) if *port* invalid.
3680 * - (-EADDRINUSE) if attempting to remove the default MAC address
3682 int rte_eth_dev_mac_addr_remove(uint8_t port, struct ether_addr *mac_addr);
3685 * Set the default MAC address.
3688 * The port identifier of the Ethernet device.
3690 * New default MAC address.
3692 * - (0) if successful, or *mac_addr* didn't exist.
3693 * - (-ENOTSUP) if hardware doesn't support.
3694 * - (-ENODEV) if *port* invalid.
3695 * - (-EINVAL) if MAC address is invalid.
3697 int rte_eth_dev_default_mac_addr_set(uint8_t port, struct ether_addr *mac_addr);
3701 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3704 * The port identifier of the Ethernet device.
3708 * Redirection table size. The table size can be queried by
3709 * rte_eth_dev_info_get().
3711 * - (0) if successful.
3712 * - (-ENOTSUP) if hardware doesn't support.
3713 * - (-EINVAL) if bad parameter.
3715 int rte_eth_dev_rss_reta_update(uint8_t port,
3716 struct rte_eth_rss_reta_entry64 *reta_conf,
3717 uint16_t reta_size);
3720 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3723 * The port identifier of the Ethernet device.
3727 * Redirection table size. The table size can be queried by
3728 * rte_eth_dev_info_get().
3730 * - (0) if successful.
3731 * - (-ENOTSUP) if hardware doesn't support.
3732 * - (-EINVAL) if bad parameter.
3734 int rte_eth_dev_rss_reta_query(uint8_t port,
3735 struct rte_eth_rss_reta_entry64 *reta_conf,
3736 uint16_t reta_size);
3739 * Updates unicast hash table for receiving packet with the given destination
3740 * MAC address, and the packet is routed to all VFs for which the RX mode is
3741 * accept packets that match the unicast hash table.
3744 * The port identifier of the Ethernet device.
3746 * Unicast MAC address.
3748 * 1 - Set an unicast hash bit for receiving packets with the MAC address.
3749 * 0 - Clear an unicast hash bit.
3751 * - (0) if successful.
3752 * - (-ENOTSUP) if hardware doesn't support.
3753 * - (-ENODEV) if *port_id* invalid.
3754 * - (-EINVAL) if bad parameter.
3756 int rte_eth_dev_uc_hash_table_set(uint8_t port,struct ether_addr *addr,
3760 * Updates all unicast hash bitmaps for receiving packet with any Unicast
3761 * Ethernet MAC addresses,the packet is routed to all VFs for which the RX
3762 * mode is accept packets that match the unicast hash table.
3765 * The port identifier of the Ethernet device.
3767 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet
3769 * 0 - Clear all unicast hash bitmaps
3771 * - (0) if successful.
3772 * - (-ENOTSUP) if hardware doesn't support.
3773 * - (-ENODEV) if *port_id* invalid.
3774 * - (-EINVAL) if bad parameter.
3776 int rte_eth_dev_uc_all_hash_table_set(uint8_t port,uint8_t on);
3779 * Set a traffic mirroring rule on an Ethernet device
3782 * The port identifier of the Ethernet device.
3783 * @param mirror_conf
3784 * The pointer to the traffic mirroring structure describing the mirroring rule.
3785 * The *rte_eth_vm_mirror_conf* structure includes the type of mirroring rule,
3786 * destination pool and the value of rule if enable vlan or pool mirroring.
3789 * The index of traffic mirroring rule, we support four separated rules.
3791 * 1 - Enable a mirroring rule.
3792 * 0 - Disable a mirroring rule.
3794 * - (0) if successful.
3795 * - (-ENOTSUP) if hardware doesn't support this feature.
3796 * - (-ENODEV) if *port_id* invalid.
3797 * - (-EINVAL) if the mr_conf information is not correct.
3799 int rte_eth_mirror_rule_set(uint8_t port_id,
3800 struct rte_eth_mirror_conf *mirror_conf,
3805 * Reset a traffic mirroring rule on an Ethernet device.
3808 * The port identifier of the Ethernet device.
3810 * The index of traffic mirroring rule, we support four separated rules.
3812 * - (0) if successful.
3813 * - (-ENOTSUP) if hardware doesn't support this feature.
3814 * - (-ENODEV) if *port_id* invalid.
3815 * - (-EINVAL) if bad parameter.
3817 int rte_eth_mirror_rule_reset(uint8_t port_id,
3821 * Set the rate limitation for a queue on an Ethernet device.
3824 * The port identifier of the Ethernet device.
3828 * The tx rate in Mbps. Allocated from the total port link speed.
3830 * - (0) if successful.
3831 * - (-ENOTSUP) if hardware doesn't support this feature.
3832 * - (-ENODEV) if *port_id* invalid.
3833 * - (-EINVAL) if bad parameter.
3835 int rte_eth_set_queue_rate_limit(uint8_t port_id, uint16_t queue_idx,
3839 * Configuration of Receive Side Scaling hash computation of Ethernet device.
3842 * The port identifier of the Ethernet device.
3844 * The new configuration to use for RSS hash computation on the port.
3846 * - (0) if successful.
3847 * - (-ENODEV) if port identifier is invalid.
3848 * - (-ENOTSUP) if hardware doesn't support.
3849 * - (-EINVAL) if bad parameter.
3851 int rte_eth_dev_rss_hash_update(uint8_t port_id,
3852 struct rte_eth_rss_conf *rss_conf);
3855 * Retrieve current configuration of Receive Side Scaling hash computation
3856 * of Ethernet device.
3859 * The port identifier of the Ethernet device.
3861 * Where to store the current RSS hash configuration of the Ethernet device.
3863 * - (0) if successful.
3864 * - (-ENODEV) if port identifier is invalid.
3865 * - (-ENOTSUP) if hardware doesn't support RSS.
3868 rte_eth_dev_rss_hash_conf_get(uint8_t port_id,
3869 struct rte_eth_rss_conf *rss_conf);
3872 * Add UDP tunneling port for a specific type of tunnel.
3873 * The packets with this UDP port will be identified as this type of tunnel.
3874 * Before enabling any offloading function for a tunnel, users can call this API
3875 * to change or add more UDP port for the tunnel. So the offloading function
3876 * can take effect on the packets with the specific UDP port.
3879 * The port identifier of the Ethernet device.
3881 * UDP tunneling configuration.
3884 * - (0) if successful.
3885 * - (-ENODEV) if port identifier is invalid.
3886 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3889 rte_eth_dev_udp_tunnel_port_add(uint8_t port_id,
3890 struct rte_eth_udp_tunnel *tunnel_udp);
3893 * Delete UDP tunneling port a specific type of tunnel.
3894 * The packets with this UDP port will not be identified as this type of tunnel
3896 * Before enabling any offloading function for a tunnel, users can call this API
3897 * to delete a UDP port for the tunnel. So the offloading function will not take
3898 * effect on the packets with the specific UDP port.
3901 * The port identifier of the Ethernet device.
3903 * UDP tunneling configuration.
3906 * - (0) if successful.
3907 * - (-ENODEV) if port identifier is invalid.
3908 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3911 rte_eth_dev_udp_tunnel_port_delete(uint8_t port_id,
3912 struct rte_eth_udp_tunnel *tunnel_udp);
3915 * Check whether the filter type is supported on an Ethernet device.
3916 * All the supported filter types are defined in 'rte_eth_ctrl.h'.
3919 * The port identifier of the Ethernet device.
3920 * @param filter_type
3923 * - (0) if successful.
3924 * - (-ENOTSUP) if hardware doesn't support this filter type.
3925 * - (-ENODEV) if *port_id* invalid.
3927 int rte_eth_dev_filter_supported(uint8_t port_id, enum rte_filter_type filter_type);
3930 * Take operations to assigned filter type on an Ethernet device.
3931 * All the supported operations and filter types are defined in 'rte_eth_ctrl.h'.
3934 * The port identifier of the Ethernet device.
3935 * @param filter_type
3938 * Type of operation.
3940 * A pointer to arguments defined specifically for the operation.
3942 * - (0) if successful.
3943 * - (-ENOTSUP) if hardware doesn't support.
3944 * - (-ENODEV) if *port_id* invalid.
3945 * - others depends on the specific operations implementation.
3947 int rte_eth_dev_filter_ctrl(uint8_t port_id, enum rte_filter_type filter_type,
3948 enum rte_filter_op filter_op, void *arg);
3951 * Get DCB information on an Ethernet device.
3954 * The port identifier of the Ethernet device.
3958 * - (0) if successful.
3959 * - (-ENODEV) if port identifier is invalid.
3960 * - (-ENOTSUP) if hardware doesn't support.
3962 int rte_eth_dev_get_dcb_info(uint8_t port_id,
3963 struct rte_eth_dcb_info *dcb_info);
3966 * Add a callback to be called on packet RX on a given port and queue.
3968 * This API configures a function to be called for each burst of
3969 * packets received on a given NIC port queue. The return value is a pointer
3970 * that can be used to later remove the callback using
3971 * rte_eth_remove_rx_callback().
3973 * Multiple functions are called in the order that they are added.
3976 * The port identifier of the Ethernet device.
3978 * The queue on the Ethernet device on which the callback is to be added.
3980 * The callback function
3982 * A generic pointer parameter which will be passed to each invocation of the
3983 * callback function on this port and queue.
3987 * On success, a pointer value which can later be used to remove the callback.
3989 void *rte_eth_add_rx_callback(uint8_t port_id, uint16_t queue_id,
3990 rte_rx_callback_fn fn, void *user_param);
3993 * Add a callback that must be called first on packet RX on a given port
3996 * This API configures a first function to be called for each burst of
3997 * packets received on a given NIC port queue. The return value is a pointer
3998 * that can be used to later remove the callback using
3999 * rte_eth_remove_rx_callback().
4001 * Multiple functions are called in the order that they are added.
4004 * The port identifier of the Ethernet device.
4006 * The queue on the Ethernet device on which the callback is to be added.
4008 * The callback function
4010 * A generic pointer parameter which will be passed to each invocation of the
4011 * callback function on this port and queue.
4015 * On success, a pointer value which can later be used to remove the callback.
4017 void *rte_eth_add_first_rx_callback(uint8_t port_id, uint16_t queue_id,
4018 rte_rx_callback_fn fn, void *user_param);
4021 * Add a callback to be called on packet TX on a given port and queue.
4023 * This API configures a function to be called for each burst of
4024 * packets sent on a given NIC port queue. The return value is a pointer
4025 * that can be used to later remove the callback using
4026 * rte_eth_remove_tx_callback().
4028 * Multiple functions are called in the order that they are added.
4031 * The port identifier of the Ethernet device.
4033 * The queue on the Ethernet device on which the callback is to be added.
4035 * The callback function
4037 * A generic pointer parameter which will be passed to each invocation of the
4038 * callback function on this port and queue.
4042 * On success, a pointer value which can later be used to remove the callback.
4044 void *rte_eth_add_tx_callback(uint8_t port_id, uint16_t queue_id,
4045 rte_tx_callback_fn fn, void *user_param);
4048 * Remove an RX packet callback from a given port and queue.
4050 * This function is used to removed callbacks that were added to a NIC port
4051 * queue using rte_eth_add_rx_callback().
4053 * Note: the callback is removed from the callback list but it isn't freed
4054 * since the it may still be in use. The memory for the callback can be
4055 * subsequently freed back by the application by calling rte_free():
4057 * - Immediately - if the port is stopped, or the user knows that no
4058 * callbacks are in flight e.g. if called from the thread doing RX/TX
4061 * - After a short delay - where the delay is sufficient to allow any
4062 * in-flight callbacks to complete.
4065 * The port identifier of the Ethernet device.
4067 * The queue on the Ethernet device from which the callback is to be removed.
4069 * User supplied callback created via rte_eth_add_rx_callback().
4072 * - 0: Success. Callback was removed.
4073 * - -ENOTSUP: Callback support is not available.
4074 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
4075 * is NULL or not found for the port/queue.
4077 int rte_eth_remove_rx_callback(uint8_t port_id, uint16_t queue_id,
4078 struct rte_eth_rxtx_callback *user_cb);
4081 * Remove a TX packet callback from a given port and queue.
4083 * This function is used to removed callbacks that were added to a NIC port
4084 * queue using rte_eth_add_tx_callback().
4086 * Note: the callback is removed from the callback list but it isn't freed
4087 * since the it may still be in use. The memory for the callback can be
4088 * subsequently freed back by the application by calling rte_free():
4090 * - Immediately - if the port is stopped, or the user knows that no
4091 * callbacks are in flight e.g. if called from the thread doing RX/TX
4094 * - After a short delay - where the delay is sufficient to allow any
4095 * in-flight callbacks to complete.
4098 * The port identifier of the Ethernet device.
4100 * The queue on the Ethernet device from which the callback is to be removed.
4102 * User supplied callback created via rte_eth_add_tx_callback().
4105 * - 0: Success. Callback was removed.
4106 * - -ENOTSUP: Callback support is not available.
4107 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
4108 * is NULL or not found for the port/queue.
4110 int rte_eth_remove_tx_callback(uint8_t port_id, uint16_t queue_id,
4111 struct rte_eth_rxtx_callback *user_cb);
4114 * Retrieve information about given port's RX queue.
4117 * The port identifier of the Ethernet device.
4119 * The RX queue on the Ethernet device for which information
4120 * will be retrieved.
4122 * A pointer to a structure of type *rte_eth_rxq_info_info* to be filled with
4123 * the information of the Ethernet device.
4127 * - -ENOTSUP: routine is not supported by the device PMD.
4128 * - -EINVAL: The port_id or the queue_id is out of range.
4130 int rte_eth_rx_queue_info_get(uint8_t port_id, uint16_t queue_id,
4131 struct rte_eth_rxq_info *qinfo);
4134 * Retrieve information about given port's TX queue.
4137 * The port identifier of the Ethernet device.
4139 * The TX queue on the Ethernet device for which information
4140 * will be retrieved.
4142 * A pointer to a structure of type *rte_eth_txq_info_info* to be filled with
4143 * the information of the Ethernet device.
4147 * - -ENOTSUP: routine is not supported by the device PMD.
4148 * - -EINVAL: The port_id or the queue_id is out of range.
4150 int rte_eth_tx_queue_info_get(uint8_t port_id, uint16_t queue_id,
4151 struct rte_eth_txq_info *qinfo);
4154 * Retrieve device registers and register attributes (number of registers and
4158 * The port identifier of the Ethernet device.
4160 * Pointer to rte_dev_reg_info structure to fill in. If info->data is
4161 * NULL the function fills in the width and length fields. If non-NULL
4162 * the registers are put into the buffer pointed at by the data field.
4164 * - (0) if successful.
4165 * - (-ENOTSUP) if hardware doesn't support.
4166 * - (-ENODEV) if *port_id* invalid.
4167 * - others depends on the specific operations implementation.
4169 int rte_eth_dev_get_reg_info(uint8_t port_id, struct rte_dev_reg_info *info);
4172 * Retrieve size of device EEPROM
4175 * The port identifier of the Ethernet device.
4177 * - (>=0) EEPROM size if successful.
4178 * - (-ENOTSUP) if hardware doesn't support.
4179 * - (-ENODEV) if *port_id* invalid.
4180 * - others depends on the specific operations implementation.
4182 int rte_eth_dev_get_eeprom_length(uint8_t port_id);
4185 * Retrieve EEPROM and EEPROM attribute
4188 * The port identifier of the Ethernet device.
4190 * The template includes buffer for return EEPROM data and
4191 * EEPROM attributes to be filled.
4193 * - (0) if successful.
4194 * - (-ENOTSUP) if hardware doesn't support.
4195 * - (-ENODEV) if *port_id* invalid.
4196 * - others depends on the specific operations implementation.
4198 int rte_eth_dev_get_eeprom(uint8_t port_id, struct rte_dev_eeprom_info *info);
4201 * Program EEPROM with provided data
4204 * The port identifier of the Ethernet device.
4206 * The template includes EEPROM data for programming and
4207 * EEPROM attributes to be filled
4209 * - (0) if successful.
4210 * - (-ENOTSUP) if hardware doesn't support.
4211 * - (-ENODEV) if *port_id* invalid.
4212 * - others depends on the specific operations implementation.
4214 int rte_eth_dev_set_eeprom(uint8_t port_id, struct rte_dev_eeprom_info *info);
4217 * Set the list of multicast addresses to filter on an Ethernet device.
4220 * The port identifier of the Ethernet device.
4221 * @param mc_addr_set
4222 * The array of multicast addresses to set. Equal to NULL when the function
4223 * is invoked to flush the set of filtered addresses.
4225 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0
4226 * when the function is invoked to flush the set of filtered addresses.
4228 * - (0) if successful.
4229 * - (-ENODEV) if *port_id* invalid.
4230 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering.
4231 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources.
4233 int rte_eth_dev_set_mc_addr_list(uint8_t port_id,
4234 struct ether_addr *mc_addr_set,
4235 uint32_t nb_mc_addr);
4238 * Enable IEEE1588/802.1AS timestamping for an Ethernet device.
4241 * The port identifier of the Ethernet device.
4245 * - -ENODEV: The port ID is invalid.
4246 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4248 int rte_eth_timesync_enable(uint8_t port_id);
4251 * Disable IEEE1588/802.1AS timestamping for an Ethernet device.
4254 * The port identifier of the Ethernet device.
4258 * - -ENODEV: The port ID is invalid.
4259 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4261 int rte_eth_timesync_disable(uint8_t port_id);
4264 * Read an IEEE1588/802.1AS RX timestamp from an Ethernet device.
4267 * The port identifier of the Ethernet device.
4269 * Pointer to the timestamp struct.
4271 * Device specific flags. Used to pass the RX timesync register index to
4272 * i40e. Unused in igb/ixgbe, pass 0 instead.
4276 * - -EINVAL: No timestamp is available.
4277 * - -ENODEV: The port ID is invalid.
4278 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4280 int rte_eth_timesync_read_rx_timestamp(uint8_t port_id,
4281 struct timespec *timestamp, uint32_t flags);
4284 * Read an IEEE1588/802.1AS TX timestamp from an Ethernet device.
4287 * The port identifier of the Ethernet device.
4289 * Pointer to the timestamp struct.
4293 * - -EINVAL: No timestamp is available.
4294 * - -ENODEV: The port ID is invalid.
4295 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4297 int rte_eth_timesync_read_tx_timestamp(uint8_t port_id,
4298 struct timespec *timestamp);
4301 * Adjust the timesync clock on an Ethernet device.
4303 * This is usually used in conjunction with other Ethdev timesync functions to
4304 * synchronize the device time using the IEEE1588/802.1AS protocol.
4307 * The port identifier of the Ethernet device.
4309 * The adjustment in nanoseconds.
4313 * - -ENODEV: The port ID is invalid.
4314 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4316 int rte_eth_timesync_adjust_time(uint8_t port_id, int64_t delta);
4319 * Read the time from the timesync clock on an Ethernet device.
4321 * This is usually used in conjunction with other Ethdev timesync functions to
4322 * synchronize the device time using the IEEE1588/802.1AS protocol.
4325 * The port identifier of the Ethernet device.
4327 * Pointer to the timespec struct that holds the time.
4332 int rte_eth_timesync_read_time(uint8_t port_id, struct timespec *time);
4335 * Set the time of the timesync clock on an Ethernet device.
4337 * This is usually used in conjunction with other Ethdev timesync functions to
4338 * synchronize the device time using the IEEE1588/802.1AS protocol.
4341 * The port identifier of the Ethernet device.
4343 * Pointer to the timespec struct that holds the time.
4347 * - -EINVAL: No timestamp is available.
4348 * - -ENODEV: The port ID is invalid.
4349 * - -ENOTSUP: The function is not supported by the Ethernet driver.
4351 int rte_eth_timesync_write_time(uint8_t port_id, const struct timespec *time);
4354 * Create memzone for HW rings.
4355 * malloc can't be used as the physical address is needed.
4356 * If the memzone is already created, then this function returns a ptr
4360 * The *eth_dev* pointer is the address of the *rte_eth_dev* structure
4362 * The name of the memory zone
4364 * The index of the queue to add to name
4366 * The sizeof of the memory area
4368 * Alignment for resulting memzone. Must be a power of 2.
4370 * The *socket_id* argument is the socket identifier in case of NUMA.
4372 const struct rte_memzone *
4373 rte_eth_dma_zone_reserve(const struct rte_eth_dev *eth_dev, const char *name,
4374 uint16_t queue_id, size_t size,
4375 unsigned align, int socket_id);
4378 * Config l2 tunnel ether type of an Ethernet device for filtering specific
4379 * tunnel packets by ether type.
4382 * The port identifier of the Ethernet device.
4384 * l2 tunnel configuration.
4387 * - (0) if successful.
4388 * - (-ENODEV) if port identifier is invalid.
4389 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4392 rte_eth_dev_l2_tunnel_eth_type_conf(uint8_t port_id,
4393 struct rte_eth_l2_tunnel_conf *l2_tunnel);
4396 * Enable/disable l2 tunnel offload functions. Include,
4397 * 1, The ability of parsing a type of l2 tunnel of an Ethernet device.
4398 * Filtering, forwarding and offloading this type of tunnel packets depend on
4400 * 2, Stripping the l2 tunnel tag.
4401 * 3, Insertion of the l2 tunnel tag.
4402 * 4, Forwarding the packets based on the l2 tunnel tag.
4405 * The port identifier of the Ethernet device.
4407 * l2 tunnel parameters.
4409 * Indicate the offload function.
4411 * Enable or disable this function.
4414 * - (0) if successful.
4415 * - (-ENODEV) if port identifier is invalid.
4416 * - (-ENOTSUP) if hardware doesn't support tunnel type.
4419 rte_eth_dev_l2_tunnel_offload_set(uint8_t port_id,
4420 struct rte_eth_l2_tunnel_conf *l2_tunnel,
4425 * Get the port id from pci address or device name
4426 * Ex: 0000:2:00.0 or vdev name net_pcap0
4429 * pci address or name of the device
4431 * pointer to port identifier of the device
4433 * - (0) if successful and port_id is filled.
4434 * - (-ENODEV or -EINVAL) on failure.
4437 rte_eth_dev_get_port_by_name(const char *name, uint8_t *port_id);
4440 * Get the device name from port id
4443 * pointer to port identifier of the device
4445 * pci address or name of the device
4447 * - (0) if successful.
4448 * - (-EINVAL) on failure.
4451 rte_eth_dev_get_name_by_port(uint8_t port_id, char *name);
4454 * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from
4455 * the ethernet device information, otherwise adjust them to boundaries.
4458 * The port identifier of the Ethernet device.
4460 * A pointer to a uint16_t where the number of receive
4461 * descriptors stored.
4463 * A pointer to a uint16_t where the number of transmit
4464 * descriptors stored.
4466 * - (0) if successful.
4467 * - (-ENOTSUP, -ENODEV or -EINVAL) on failure.
4469 int rte_eth_dev_adjust_nb_rx_tx_desc(uint8_t port_id,
4470 uint16_t *nb_rx_desc,
4471 uint16_t *nb_tx_desc);
4477 #endif /* _RTE_ETHDEV_H_ */