4 * Copyright(c) 2010-2014 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 tramsit 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 * after 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>
182 #include <rte_devargs.h>
183 #include "rte_ether.h"
184 #include "rte_eth_ctrl.h"
189 * A structure used to retrieve statistics for an Ethernet port.
191 struct rte_eth_stats {
192 uint64_t ipackets; /**< Total number of successfully received packets. */
193 uint64_t opackets; /**< Total number of successfully transmitted packets.*/
194 uint64_t ibytes; /**< Total number of successfully received bytes. */
195 uint64_t obytes; /**< Total number of successfully transmitted bytes. */
196 uint64_t imissed; /**< Total of RX missed packets (e.g full FIFO). */
197 uint64_t ibadcrc; /**< Total of RX packets with CRC error. */
198 uint64_t ibadlen; /**< Total of RX packets with bad length. */
199 uint64_t ierrors; /**< Total number of erroneous received packets. */
200 uint64_t oerrors; /**< Total number of failed transmitted packets. */
201 uint64_t imcasts; /**< Total number of multicast received packets. */
202 uint64_t rx_nombuf; /**< Total number of RX mbuf allocation failures. */
203 uint64_t fdirmatch; /**< Total number of RX packets matching a filter. */
204 uint64_t fdirmiss; /**< Total number of RX packets not matching any filter. */
205 uint64_t tx_pause_xon; /**< Total nb. of XON pause frame sent. */
206 uint64_t rx_pause_xon; /**< Total nb. of XON pause frame received. */
207 uint64_t tx_pause_xoff; /**< Total nb. of XOFF pause frame sent. */
208 uint64_t rx_pause_xoff; /**< Total nb. of XOFF pause frame received. */
209 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
210 /**< Total number of queue RX packets. */
211 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
212 /**< Total number of queue TX packets. */
213 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
214 /**< Total number of successfully received queue bytes. */
215 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
216 /**< Total number of successfully transmitted queue bytes. */
217 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS];
218 /**< Total number of queue packets received that are dropped. */
220 /**< Total number of good packets received from loopback,VF Only */
222 /**< Total number of good packets transmitted to loopback,VF Only */
224 /**< Total number of good bytes received from loopback,VF Only */
226 /**< Total number of good bytes transmitted to loopback,VF Only */
230 * A structure used to retrieve link-level information of an Ethernet port.
232 struct rte_eth_link {
233 uint16_t link_speed; /**< ETH_LINK_SPEED_[10, 100, 1000, 10000] */
234 uint16_t link_duplex; /**< ETH_LINK_[HALF_DUPLEX, FULL_DUPLEX] */
235 uint8_t link_status : 1; /**< 1 -> link up, 0 -> link down */
236 }__attribute__((aligned(8))); /**< aligned for atomic64 read/write */
238 #define ETH_LINK_SPEED_AUTONEG 0 /**< Auto-negotiate link speed. */
239 #define ETH_LINK_SPEED_10 10 /**< 10 megabits/second. */
240 #define ETH_LINK_SPEED_100 100 /**< 100 megabits/second. */
241 #define ETH_LINK_SPEED_1000 1000 /**< 1 gigabits/second. */
242 #define ETH_LINK_SPEED_10000 10000 /**< 10 gigabits/second. */
243 #define ETH_LINK_SPEED_10G 10000 /**< alias of 10 gigabits/second. */
244 #define ETH_LINK_SPEED_20G 20000 /**< 20 gigabits/second. */
245 #define ETH_LINK_SPEED_40G 40000 /**< 40 gigabits/second. */
247 #define ETH_LINK_AUTONEG_DUPLEX 0 /**< Auto-negotiate duplex. */
248 #define ETH_LINK_HALF_DUPLEX 1 /**< Half-duplex connection. */
249 #define ETH_LINK_FULL_DUPLEX 2 /**< Full-duplex connection. */
252 * A structure used to configure the ring threshold registers of an RX/TX
253 * queue for an Ethernet port.
255 struct rte_eth_thresh {
256 uint8_t pthresh; /**< Ring prefetch threshold. */
257 uint8_t hthresh; /**< Ring host threshold. */
258 uint8_t wthresh; /**< Ring writeback threshold. */
262 * Simple flags are used for rte_eth_conf.rxmode.mq_mode.
264 #define ETH_MQ_RX_RSS_FLAG 0x1
265 #define ETH_MQ_RX_DCB_FLAG 0x2
266 #define ETH_MQ_RX_VMDQ_FLAG 0x4
269 * A set of values to identify what method is to be used to route
270 * packets to multiple queues.
272 enum rte_eth_rx_mq_mode {
273 /** None of DCB,RSS or VMDQ mode */
276 /** For RX side, only RSS is on */
277 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG,
278 /** For RX side,only DCB is on. */
279 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG,
280 /** Both DCB and RSS enable */
281 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG,
283 /** Only VMDQ, no RSS nor DCB */
284 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG,
285 /** RSS mode with VMDQ */
286 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG,
287 /** Use VMDQ+DCB to route traffic to queues */
288 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG,
289 /** Enable both VMDQ and DCB in VMDq */
290 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG |
295 * for rx mq mode backward compatible
297 #define ETH_RSS ETH_MQ_RX_RSS
298 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB
299 #define ETH_DCB_RX ETH_MQ_RX_DCB
302 * A set of values to identify what method is to be used to transmit
303 * packets using multi-TCs.
305 enum rte_eth_tx_mq_mode {
306 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */
307 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */
308 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */
309 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */
313 * for tx mq mode backward compatible
315 #define ETH_DCB_NONE ETH_MQ_TX_NONE
316 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB
317 #define ETH_DCB_TX ETH_MQ_TX_DCB
320 * A structure used to configure the RX features of an Ethernet port.
322 struct rte_eth_rxmode {
323 /** The multi-queue packet distribution mode to be used, e.g. RSS. */
324 enum rte_eth_rx_mq_mode mq_mode;
325 uint32_t max_rx_pkt_len; /**< Only used if jumbo_frame enabled. */
326 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/
327 uint16_t header_split : 1, /**< Header Split enable. */
328 hw_ip_checksum : 1, /**< IP/UDP/TCP checksum offload enable. */
329 hw_vlan_filter : 1, /**< VLAN filter enable. */
330 hw_vlan_strip : 1, /**< VLAN strip enable. */
331 hw_vlan_extend : 1, /**< Extended VLAN enable. */
332 jumbo_frame : 1, /**< Jumbo Frame Receipt enable. */
333 hw_strip_crc : 1, /**< Enable CRC stripping by hardware. */
334 enable_scatter : 1, /**< Enable scatter packets rx handler */
335 enable_lro : 1; /**< Enable LRO */
339 * A structure used to configure the Receive Side Scaling (RSS) feature
340 * of an Ethernet port.
341 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points
342 * to an array holding the RSS key to use for hashing specific header
343 * fields of received packets. The length of this array should be indicated
344 * by *rss_key_len* below. Otherwise, a default random hash key is used by
347 * The *rss_key_len* field of the *rss_conf* structure indicates the length
348 * in bytes of the array pointed by *rss_key*. To be compatible, this length
349 * will be checked in i40e only. Others assume 40 bytes to be used as before.
351 * The *rss_hf* field of the *rss_conf* structure indicates the different
352 * types of IPv4/IPv6 packets to which the RSS hashing must be applied.
353 * Supplying an *rss_hf* equal to zero disables the RSS feature.
355 struct rte_eth_rss_conf {
356 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */
357 uint8_t rss_key_len; /**< hash key length in bytes. */
358 uint64_t rss_hf; /**< Hash functions to apply - see below. */
362 * The RSS offload types are defined based on flow types which are defined
363 * in rte_eth_ctrl.h. Different NIC hardwares may support different RSS offload
364 * types. The supported flow types or RSS offload types can be queried by
365 * rte_eth_dev_info_get().
367 #define ETH_RSS_IPV4 (1ULL << RTE_ETH_FLOW_IPV4)
368 #define ETH_RSS_FRAG_IPV4 (1ULL << RTE_ETH_FLOW_FRAG_IPV4)
369 #define ETH_RSS_NONFRAG_IPV4_TCP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_TCP)
370 #define ETH_RSS_NONFRAG_IPV4_UDP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_UDP)
371 #define ETH_RSS_NONFRAG_IPV4_SCTP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_SCTP)
372 #define ETH_RSS_NONFRAG_IPV4_OTHER (1ULL << RTE_ETH_FLOW_NONFRAG_IPV4_OTHER)
373 #define ETH_RSS_IPV6 (1ULL << RTE_ETH_FLOW_IPV6)
374 #define ETH_RSS_FRAG_IPV6 (1ULL << RTE_ETH_FLOW_FRAG_IPV6)
375 #define ETH_RSS_NONFRAG_IPV6_TCP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_TCP)
376 #define ETH_RSS_NONFRAG_IPV6_UDP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_UDP)
377 #define ETH_RSS_NONFRAG_IPV6_SCTP (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_SCTP)
378 #define ETH_RSS_NONFRAG_IPV6_OTHER (1ULL << RTE_ETH_FLOW_NONFRAG_IPV6_OTHER)
379 #define ETH_RSS_L2_PAYLOAD (1ULL << RTE_ETH_FLOW_L2_PAYLOAD)
380 #define ETH_RSS_IPV6_EX (1ULL << RTE_ETH_FLOW_IPV6_EX)
381 #define ETH_RSS_IPV6_TCP_EX (1ULL << RTE_ETH_FLOW_IPV6_TCP_EX)
382 #define ETH_RSS_IPV6_UDP_EX (1ULL << RTE_ETH_FLOW_IPV6_UDP_EX)
384 #define ETH_RSS_IP ( \
386 ETH_RSS_FRAG_IPV4 | \
387 ETH_RSS_NONFRAG_IPV4_OTHER | \
389 ETH_RSS_FRAG_IPV6 | \
390 ETH_RSS_NONFRAG_IPV6_OTHER | \
393 #define ETH_RSS_UDP ( \
394 ETH_RSS_NONFRAG_IPV4_UDP | \
395 ETH_RSS_NONFRAG_IPV6_UDP | \
398 #define ETH_RSS_TCP ( \
399 ETH_RSS_NONFRAG_IPV4_TCP | \
400 ETH_RSS_NONFRAG_IPV6_TCP | \
403 #define ETH_RSS_SCTP ( \
404 ETH_RSS_NONFRAG_IPV4_SCTP | \
405 ETH_RSS_NONFRAG_IPV6_SCTP)
407 /**< Mask of valid RSS hash protocols */
408 #define ETH_RSS_PROTO_MASK ( \
410 ETH_RSS_FRAG_IPV4 | \
411 ETH_RSS_NONFRAG_IPV4_TCP | \
412 ETH_RSS_NONFRAG_IPV4_UDP | \
413 ETH_RSS_NONFRAG_IPV4_SCTP | \
414 ETH_RSS_NONFRAG_IPV4_OTHER | \
416 ETH_RSS_FRAG_IPV6 | \
417 ETH_RSS_NONFRAG_IPV6_TCP | \
418 ETH_RSS_NONFRAG_IPV6_UDP | \
419 ETH_RSS_NONFRAG_IPV6_SCTP | \
420 ETH_RSS_NONFRAG_IPV6_OTHER | \
421 ETH_RSS_L2_PAYLOAD | \
423 ETH_RSS_IPV6_TCP_EX | \
427 * Definitions used for redirection table entry size.
428 * Some RSS RETA sizes may not be supported by some drivers, check the
429 * documentation or the description of relevant functions for more details.
431 #define ETH_RSS_RETA_SIZE_64 64
432 #define ETH_RSS_RETA_SIZE_128 128
433 #define ETH_RSS_RETA_SIZE_512 512
434 #define RTE_RETA_GROUP_SIZE 64
436 /* Definitions used for VMDQ and DCB functionality */
437 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */
438 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */
439 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */
440 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */
442 /* DCB capability defines */
443 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */
444 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */
446 /* Definitions used for VLAN Offload functionality */
447 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */
448 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */
449 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */
451 /* Definitions used for mask VLAN setting */
452 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */
453 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/
454 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/
455 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/
457 /* Definitions used for receive MAC address */
458 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */
460 /* Definitions used for unicast hash */
461 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */
463 /* Definitions used for VMDQ pool rx mode setting */
464 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */
465 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */
466 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */
467 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */
468 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */
470 /** Maximum nb. of vlan per mirror rule */
471 #define ETH_MIRROR_MAX_VLANS 64
473 #define ETH_VMDQ_POOL_MIRROR 0x0001 /**< Virtual Pool Mirroring. */
474 #define ETH_VMDQ_UPLINK_MIRROR 0x0002 /**< Uplink Port Mirroring. */
475 #define ETH_VMDQ_DOWNLIN_MIRROR 0x0004 /**< Downlink Port Mirroring. */
476 #define ETH_VMDQ_VLAN_MIRROR 0x0008 /**< VLAN Mirroring. */
479 * A structure used to configure VLAN traffic mirror of an Ethernet port.
481 struct rte_eth_vlan_mirror {
482 uint64_t vlan_mask; /**< mask for valid VLAN ID. */
483 /** VLAN ID list for vlan mirroring. */
484 uint16_t vlan_id[ETH_MIRROR_MAX_VLANS];
488 * A structure used to configure traffic mirror of an Ethernet port.
490 struct rte_eth_mirror_conf {
491 uint8_t rule_type_mask; /**< Mirroring rule type mask we want to set */
492 uint8_t dst_pool; /**< Destination pool for this mirror rule. */
493 uint64_t pool_mask; /**< Bitmap of pool for pool mirroring */
494 /** VLAN ID setting for VLAN mirroring. */
495 struct rte_eth_vlan_mirror vlan;
499 * A structure used to configure 64 entries of Redirection Table of the
500 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure
501 * more than 64 entries supported by hardware, an array of this structure
504 struct rte_eth_rss_reta_entry64 {
506 /**< Mask bits indicate which entries need to be updated/queried. */
507 uint8_t reta[RTE_RETA_GROUP_SIZE];
508 /**< Group of 64 redirection table entries. */
512 * This enum indicates the possible number of traffic classes
513 * in DCB configratioins
515 enum rte_eth_nb_tcs {
516 ETH_4_TCS = 4, /**< 4 TCs with DCB. */
517 ETH_8_TCS = 8 /**< 8 TCs with DCB. */
521 * This enum indicates the possible number of queue pools
522 * in VMDQ configurations.
524 enum rte_eth_nb_pools {
525 ETH_8_POOLS = 8, /**< 8 VMDq pools. */
526 ETH_16_POOLS = 16, /**< 16 VMDq pools. */
527 ETH_32_POOLS = 32, /**< 32 VMDq pools. */
528 ETH_64_POOLS = 64 /**< 64 VMDq pools. */
531 /* This structure may be extended in future. */
532 struct rte_eth_dcb_rx_conf {
533 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */
534 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
535 /**< Possible DCB queue,4 or 8. */
538 struct rte_eth_vmdq_dcb_tx_conf {
539 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */
540 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
541 /**< Possible DCB queue,4 or 8. */
544 struct rte_eth_dcb_tx_conf {
545 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */
546 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
547 /**< Possible DCB queue,4 or 8. */
550 struct rte_eth_vmdq_tx_conf {
551 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */
555 * A structure used to configure the VMDQ+DCB feature
556 * of an Ethernet port.
558 * Using this feature, packets are routed to a pool of queues, based
559 * on the vlan id in the vlan tag, and then to a specific queue within
560 * that pool, using the user priority vlan tag field.
562 * A default pool may be used, if desired, to route all traffic which
563 * does not match the vlan filter rules.
565 struct rte_eth_vmdq_dcb_conf {
566 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */
567 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
568 uint8_t default_pool; /**< The default pool, if applicable */
569 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
571 uint16_t vlan_id; /**< The vlan id of the received frame */
572 uint64_t pools; /**< Bitmask of pools for packet rx */
573 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
574 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
575 /**< Selects a queue in a pool */
578 struct rte_eth_vmdq_rx_conf {
579 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */
580 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
581 uint8_t default_pool; /**< The default pool, if applicable */
582 uint8_t enable_loop_back; /**< Enable VT loop back */
583 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
584 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */
586 uint16_t vlan_id; /**< The vlan id of the received frame */
587 uint64_t pools; /**< Bitmask of pools for packet rx */
588 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
592 * A structure used to configure the TX features of an Ethernet port.
594 struct rte_eth_txmode {
595 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */
597 /* For i40e specifically */
599 uint8_t hw_vlan_reject_tagged : 1,
600 /**< If set, reject sending out tagged pkts */
601 hw_vlan_reject_untagged : 1,
602 /**< If set, reject sending out untagged pkts */
603 hw_vlan_insert_pvid : 1;
604 /**< If set, enable port based VLAN insertion */
608 * A structure used to configure an RX ring of an Ethernet port.
610 struct rte_eth_rxconf {
611 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */
612 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
613 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
614 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
617 #define ETH_TXQ_FLAGS_NOMULTSEGS 0x0001 /**< nb_segs=1 for all mbufs */
618 #define ETH_TXQ_FLAGS_NOREFCOUNT 0x0002 /**< refcnt can be ignored */
619 #define ETH_TXQ_FLAGS_NOMULTMEMP 0x0004 /**< all bufs come from same mempool */
620 #define ETH_TXQ_FLAGS_NOVLANOFFL 0x0100 /**< disable VLAN offload */
621 #define ETH_TXQ_FLAGS_NOXSUMSCTP 0x0200 /**< disable SCTP checksum offload */
622 #define ETH_TXQ_FLAGS_NOXSUMUDP 0x0400 /**< disable UDP checksum offload */
623 #define ETH_TXQ_FLAGS_NOXSUMTCP 0x0800 /**< disable TCP checksum offload */
624 #define ETH_TXQ_FLAGS_NOOFFLOADS \
625 (ETH_TXQ_FLAGS_NOVLANOFFL | ETH_TXQ_FLAGS_NOXSUMSCTP | \
626 ETH_TXQ_FLAGS_NOXSUMUDP | ETH_TXQ_FLAGS_NOXSUMTCP)
627 #define ETH_TXQ_FLAGS_NOXSUMS \
628 (ETH_TXQ_FLAGS_NOXSUMSCTP | ETH_TXQ_FLAGS_NOXSUMUDP | \
629 ETH_TXQ_FLAGS_NOXSUMTCP)
631 * A structure used to configure a TX ring of an Ethernet port.
633 struct rte_eth_txconf {
634 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */
635 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */
636 uint16_t tx_free_thresh; /**< Start freeing TX buffers if there are
637 less free descriptors than this value. */
639 uint32_t txq_flags; /**< Set flags for the Tx queue */
640 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */
644 * This enum indicates the flow control mode
646 enum rte_eth_fc_mode {
647 RTE_FC_NONE = 0, /**< Disable flow control. */
648 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */
649 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */
650 RTE_FC_FULL /**< Enable flow control on both side. */
654 * A structure used to configure Ethernet flow control parameter.
655 * These parameters will be configured into the register of the NIC.
656 * Please refer to the corresponding data sheet for proper value.
658 struct rte_eth_fc_conf {
659 uint32_t high_water; /**< High threshold value to trigger XOFF */
660 uint32_t low_water; /**< Low threshold value to trigger XON */
661 uint16_t pause_time; /**< Pause quota in the Pause frame */
662 uint16_t send_xon; /**< Is XON frame need be sent */
663 enum rte_eth_fc_mode mode; /**< Link flow control mode */
664 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */
665 uint8_t autoneg; /**< Use Pause autoneg */
669 * A structure used to configure Ethernet priority flow control parameter.
670 * These parameters will be configured into the register of the NIC.
671 * Please refer to the corresponding data sheet for proper value.
673 struct rte_eth_pfc_conf {
674 struct rte_eth_fc_conf fc; /**< General flow control parameter. */
675 uint8_t priority; /**< VLAN User Priority. */
679 * Memory space that can be configured to store Flow Director filters
680 * in the board memory.
682 enum rte_fdir_pballoc_type {
683 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */
684 RTE_FDIR_PBALLOC_128K, /**< 128k. */
685 RTE_FDIR_PBALLOC_256K, /**< 256k. */
689 * Select report mode of FDIR hash information in RX descriptors.
691 enum rte_fdir_status_mode {
692 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */
693 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */
694 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */
698 * A structure used to configure the Flow Director (FDIR) feature
699 * of an Ethernet port.
701 * If mode is RTE_FDIR_DISABLE, the pballoc value is ignored.
703 struct rte_fdir_conf {
704 enum rte_fdir_mode mode; /**< Flow Director mode. */
705 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */
706 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */
707 /** RX queue of packets matching a "drop" filter in perfect mode. */
709 struct rte_eth_fdir_masks mask;
710 struct rte_eth_fdir_flex_conf flex_conf;
711 /**< Flex payload configuration. */
715 * UDP tunneling configuration.
717 struct rte_eth_udp_tunnel {
723 * Possible l4type of FDIR filters.
726 RTE_FDIR_L4TYPE_NONE = 0, /**< None. */
727 RTE_FDIR_L4TYPE_UDP, /**< UDP. */
728 RTE_FDIR_L4TYPE_TCP, /**< TCP. */
729 RTE_FDIR_L4TYPE_SCTP, /**< SCTP. */
733 * Select IPv4 or IPv6 FDIR filters.
736 RTE_FDIR_IPTYPE_IPV4 = 0, /**< IPv4. */
737 RTE_FDIR_IPTYPE_IPV6 , /**< IPv6. */
741 * A structure used to define a FDIR packet filter.
743 struct rte_fdir_filter {
744 uint16_t flex_bytes; /**< Flex bytes value to match. */
745 uint16_t vlan_id; /**< VLAN ID value to match, 0 otherwise. */
746 uint16_t port_src; /**< Source port to match, 0 otherwise. */
747 uint16_t port_dst; /**< Destination port to match, 0 otherwise. */
749 uint32_t ipv4_addr; /**< IPv4 source address to match. */
750 uint32_t ipv6_addr[4]; /**< IPv6 source address to match. */
751 } ip_src; /**< IPv4/IPv6 source address to match (union of above). */
753 uint32_t ipv4_addr; /**< IPv4 destination address to match. */
754 uint32_t ipv6_addr[4]; /**< IPv6 destination address to match */
755 } ip_dst; /**< IPv4/IPv6 destination address to match (union of above). */
756 enum rte_l4type l4type; /**< l4type to match: NONE/UDP/TCP/SCTP. */
757 enum rte_iptype iptype; /**< IP packet type to match: IPv4 or IPv6. */
761 * A structure used to configure FDIR masks that are used by the device
762 * to match the various fields of RX packet headers.
763 * @note The only_ip_flow field has the opposite meaning compared to other
766 struct rte_fdir_masks {
767 /** When set to 1, packet l4type is \b NOT relevant in filters, and
768 source and destination port masks must be set to zero. */
769 uint8_t only_ip_flow;
770 /** If set to 1, vlan_id is relevant in filters. */
772 /** If set to 1, vlan_prio is relevant in filters. */
774 /** If set to 1, flexbytes is relevant in filters. */
776 /** If set to 1, set the IPv6 masks. Otherwise set the IPv4 masks. */
777 uint8_t set_ipv6_mask;
778 /** When set to 1, comparison of destination IPv6 address with IP6AT
779 registers is meaningful. */
780 uint8_t comp_ipv6_dst;
781 /** Mask of Destination IPv4 Address. All bits set to 1 define the
782 relevant bits to use in the destination address of an IPv4 packet
783 when matching it against FDIR filters. */
784 uint32_t dst_ipv4_mask;
785 /** Mask of Source IPv4 Address. All bits set to 1 define
786 the relevant bits to use in the source address of an IPv4 packet
787 when matching it against FDIR filters. */
788 uint32_t src_ipv4_mask;
789 /** Mask of Source IPv6 Address. All bits set to 1 define the
790 relevant BYTES to use in the source address of an IPv6 packet
791 when matching it against FDIR filters. */
792 uint16_t dst_ipv6_mask;
793 /** Mask of Destination IPv6 Address. All bits set to 1 define the
794 relevant BYTES to use in the destination address of an IPv6 packet
795 when matching it against FDIR filters. */
796 uint16_t src_ipv6_mask;
797 /** Mask of Source Port. All bits set to 1 define the relevant
798 bits to use in the source port of an IP packets when matching it
799 against FDIR filters. */
800 uint16_t src_port_mask;
801 /** Mask of Destination Port. All bits set to 1 define the relevant
802 bits to use in the destination port of an IP packet when matching it
803 against FDIR filters. */
804 uint16_t dst_port_mask;
808 * A structure used to report the status of the flow director filters in use.
810 struct rte_eth_fdir {
811 /** Number of filters with collision indication. */
813 /** Number of free (non programmed) filters. */
815 /** The Lookup hash value of the added filter that updated the value
816 of the MAXLEN field */
818 /** Longest linked list of filters in the table. */
820 /** Number of added filters. */
822 /** Number of removed filters. */
824 /** Number of failed added filters (no more space in device). */
826 /** Number of failed removed filters. */
831 * A structure used to enable/disable specific device interrupts.
833 struct rte_intr_conf {
834 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */
839 * A structure used to configure an Ethernet port.
840 * Depending upon the RX multi-queue mode, extra advanced
841 * configuration settings may be needed.
843 struct rte_eth_conf {
845 /**< ETH_LINK_SPEED_10[0|00|000], or 0 for autonegotation */
846 uint16_t link_duplex;
847 /**< ETH_LINK_[HALF_DUPLEX|FULL_DUPLEX], or 0 for autonegotation */
848 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */
849 struct rte_eth_txmode txmode; /**< Port TX configuration. */
850 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value
851 is 0, meaning the loopback mode is disabled.
852 Read the datasheet of given ethernet controller
853 for details. The possible values of this field
854 are defined in implementation of each driver. */
856 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */
857 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf;
858 /**< Port vmdq+dcb configuration. */
859 struct rte_eth_dcb_rx_conf dcb_rx_conf;
860 /**< Port dcb RX configuration. */
861 struct rte_eth_vmdq_rx_conf vmdq_rx_conf;
862 /**< Port vmdq RX configuration. */
863 } rx_adv_conf; /**< Port RX filtering configuration (union). */
865 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf;
866 /**< Port vmdq+dcb TX configuration. */
867 struct rte_eth_dcb_tx_conf dcb_tx_conf;
868 /**< Port dcb TX configuration. */
869 struct rte_eth_vmdq_tx_conf vmdq_tx_conf;
870 /**< Port vmdq TX configuration. */
871 } tx_adv_conf; /**< Port TX DCB configuration (union). */
872 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC
873 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */
874 uint32_t dcb_capability_en;
875 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. */
876 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */
880 * A structure used to retrieve the contextual information of
881 * an Ethernet device, such as the controlling driver of the device,
882 * its PCI context, etc...
886 * RX offload capabilities of a device.
888 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001
889 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002
890 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004
891 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008
892 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010
893 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020
896 * TX offload capabilities of a device.
898 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001
899 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002
900 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004
901 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008
902 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010
903 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020
904 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040
905 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */
906 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100
908 struct rte_eth_dev_info {
909 struct rte_pci_device *pci_dev; /**< Device PCI information. */
910 const char *driver_name; /**< Device Driver name. */
911 unsigned int if_index; /**< Index to bound host interface, or 0 if none.
912 Use if_indextoname() to translate into an interface name. */
913 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */
914 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */
915 uint16_t max_rx_queues; /**< Maximum number of RX queues. */
916 uint16_t max_tx_queues; /**< Maximum number of TX queues. */
917 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */
918 uint32_t max_hash_mac_addrs;
919 /** Maximum number of hash MAC addresses for MTA and UTA. */
920 uint16_t max_vfs; /**< Maximum number of VFs. */
921 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */
922 uint32_t rx_offload_capa; /**< Device RX offload capabilities. */
923 uint32_t tx_offload_capa; /**< Device TX offload capabilities. */
925 /**< Device redirection table size, the total number of entries. */
926 /** Bit mask of RSS offloads, the bit offset also means flow type */
927 uint64_t flow_type_rss_offloads;
928 struct rte_eth_rxconf default_rxconf; /**< Default RX configuration */
929 struct rte_eth_txconf default_txconf; /**< Default TX configuration */
930 uint16_t vmdq_queue_base; /**< First queue ID for VMDQ pools. */
931 uint16_t vmdq_queue_num; /**< Queue number for VMDQ pools. */
932 uint16_t vmdq_pool_base; /**< First ID of VMDQ pools. */
935 /** Maximum name length for extended statistics counters */
936 #define RTE_ETH_XSTATS_NAME_SIZE 64
939 * An Ethernet device extended statistic structure
941 * This structure is used by ethdev->eth_xstats_get() to provide
942 * statistics that are not provided in the generic rte_eth_stats
945 struct rte_eth_xstats {
946 char name[RTE_ETH_XSTATS_NAME_SIZE];
952 struct rte_eth_dev_callback;
953 /** @internal Structure to keep track of registered callbacks */
954 TAILQ_HEAD(rte_eth_dev_cb_list, rte_eth_dev_callback);
957 * Definitions of all functions exported by an Ethernet driver through the
958 * the generic structure of type *eth_dev_ops* supplied in the *rte_eth_dev*
959 * structure associated with an Ethernet device.
962 typedef int (*eth_dev_configure_t)(struct rte_eth_dev *dev);
963 /**< @internal Ethernet device configuration. */
965 typedef int (*eth_dev_start_t)(struct rte_eth_dev *dev);
966 /**< @internal Function used to start a configured Ethernet device. */
968 typedef void (*eth_dev_stop_t)(struct rte_eth_dev *dev);
969 /**< @internal Function used to stop a configured Ethernet device. */
971 typedef int (*eth_dev_set_link_up_t)(struct rte_eth_dev *dev);
972 /**< @internal Function used to link up a configured Ethernet device. */
974 typedef int (*eth_dev_set_link_down_t)(struct rte_eth_dev *dev);
975 /**< @internal Function used to link down a configured Ethernet device. */
977 typedef void (*eth_dev_close_t)(struct rte_eth_dev *dev);
978 /**< @internal Function used to close a configured Ethernet device. */
980 typedef void (*eth_promiscuous_enable_t)(struct rte_eth_dev *dev);
981 /**< @internal Function used to enable the RX promiscuous mode of an Ethernet device. */
983 typedef void (*eth_promiscuous_disable_t)(struct rte_eth_dev *dev);
984 /**< @internal Function used to disable the RX promiscuous mode of an Ethernet device. */
986 typedef void (*eth_allmulticast_enable_t)(struct rte_eth_dev *dev);
987 /**< @internal Enable the receipt of all multicast packets by an Ethernet device. */
989 typedef void (*eth_allmulticast_disable_t)(struct rte_eth_dev *dev);
990 /**< @internal Disable the receipt of all multicast packets by an Ethernet device. */
992 typedef int (*eth_link_update_t)(struct rte_eth_dev *dev,
993 int wait_to_complete);
994 /**< @internal Get link speed, duplex mode and state (up/down) of an Ethernet device. */
996 typedef void (*eth_stats_get_t)(struct rte_eth_dev *dev,
997 struct rte_eth_stats *igb_stats);
998 /**< @internal Get global I/O statistics of an Ethernet device. */
1000 typedef void (*eth_stats_reset_t)(struct rte_eth_dev *dev);
1001 /**< @internal Reset global I/O statistics of an Ethernet device to 0. */
1003 typedef int (*eth_xstats_get_t)(struct rte_eth_dev *dev,
1004 struct rte_eth_xstats *stats, unsigned n);
1005 /**< @internal Get extended stats of an Ethernet device. */
1007 typedef void (*eth_xstats_reset_t)(struct rte_eth_dev *dev);
1008 /**< @internal Reset extended stats of an Ethernet device. */
1010 typedef int (*eth_queue_stats_mapping_set_t)(struct rte_eth_dev *dev,
1014 /**< @internal Set a queue statistics mapping for a tx/rx queue of an Ethernet device. */
1016 typedef void (*eth_dev_infos_get_t)(struct rte_eth_dev *dev,
1017 struct rte_eth_dev_info *dev_info);
1018 /**< @internal Get specific informations of an Ethernet device. */
1020 typedef int (*eth_queue_start_t)(struct rte_eth_dev *dev,
1022 /**< @internal Start rx and tx of a queue of an Ethernet device. */
1024 typedef int (*eth_queue_stop_t)(struct rte_eth_dev *dev,
1026 /**< @internal Stop rx and tx of a queue of an Ethernet device. */
1028 typedef int (*eth_rx_queue_setup_t)(struct rte_eth_dev *dev,
1029 uint16_t rx_queue_id,
1030 uint16_t nb_rx_desc,
1031 unsigned int socket_id,
1032 const struct rte_eth_rxconf *rx_conf,
1033 struct rte_mempool *mb_pool);
1034 /**< @internal Set up a receive queue of an Ethernet device. */
1036 typedef int (*eth_tx_queue_setup_t)(struct rte_eth_dev *dev,
1037 uint16_t tx_queue_id,
1038 uint16_t nb_tx_desc,
1039 unsigned int socket_id,
1040 const struct rte_eth_txconf *tx_conf);
1041 /**< @internal Setup a transmit queue of an Ethernet device. */
1043 typedef void (*eth_queue_release_t)(void *queue);
1044 /**< @internal Release memory resources allocated by given RX/TX queue. */
1046 typedef uint32_t (*eth_rx_queue_count_t)(struct rte_eth_dev *dev,
1047 uint16_t rx_queue_id);
1048 /**< @internal Get number of available descriptors on a receive queue of an Ethernet device. */
1050 typedef int (*eth_rx_descriptor_done_t)(void *rxq, uint16_t offset);
1051 /**< @internal Check DD bit of specific RX descriptor */
1053 typedef int (*mtu_set_t)(struct rte_eth_dev *dev, uint16_t mtu);
1054 /**< @internal Set MTU. */
1056 typedef int (*vlan_filter_set_t)(struct rte_eth_dev *dev,
1059 /**< @internal filtering of a VLAN Tag Identifier by an Ethernet device. */
1061 typedef void (*vlan_tpid_set_t)(struct rte_eth_dev *dev,
1063 /**< @internal set the outer VLAN-TPID by an Ethernet device. */
1065 typedef void (*vlan_offload_set_t)(struct rte_eth_dev *dev, int mask);
1066 /**< @internal set VLAN offload function by an Ethernet device. */
1068 typedef int (*vlan_pvid_set_t)(struct rte_eth_dev *dev,
1071 /**< @internal set port based TX VLAN insertion by an Ethernet device. */
1073 typedef void (*vlan_strip_queue_set_t)(struct rte_eth_dev *dev,
1074 uint16_t rx_queue_id,
1076 /**< @internal VLAN stripping enable/disable by an queue of Ethernet device. */
1078 typedef uint16_t (*eth_rx_burst_t)(void *rxq,
1079 struct rte_mbuf **rx_pkts,
1081 /**< @internal Retrieve input packets from a receive queue of an Ethernet device. */
1083 typedef uint16_t (*eth_tx_burst_t)(void *txq,
1084 struct rte_mbuf **tx_pkts,
1086 /**< @internal Send output packets on a transmit queue of an Ethernet device. */
1088 typedef int (*fdir_add_signature_filter_t)(struct rte_eth_dev *dev,
1089 struct rte_fdir_filter *fdir_ftr,
1091 /**< @internal Setup a new signature filter rule on an Ethernet device */
1093 typedef int (*fdir_update_signature_filter_t)(struct rte_eth_dev *dev,
1094 struct rte_fdir_filter *fdir_ftr,
1096 /**< @internal Update a signature filter rule on an Ethernet device */
1098 typedef int (*fdir_remove_signature_filter_t)(struct rte_eth_dev *dev,
1099 struct rte_fdir_filter *fdir_ftr);
1100 /**< @internal Remove a signature filter rule on an Ethernet device */
1102 typedef void (*fdir_infos_get_t)(struct rte_eth_dev *dev,
1103 struct rte_eth_fdir *fdir);
1104 /**< @internal Get information about fdir status */
1106 typedef int (*fdir_add_perfect_filter_t)(struct rte_eth_dev *dev,
1107 struct rte_fdir_filter *fdir_ftr,
1108 uint16_t soft_id, uint8_t rx_queue,
1110 /**< @internal Setup a new perfect filter rule on an Ethernet device */
1112 typedef int (*fdir_update_perfect_filter_t)(struct rte_eth_dev *dev,
1113 struct rte_fdir_filter *fdir_ftr,
1114 uint16_t soft_id, uint8_t rx_queue,
1116 /**< @internal Update a perfect filter rule on an Ethernet device */
1118 typedef int (*fdir_remove_perfect_filter_t)(struct rte_eth_dev *dev,
1119 struct rte_fdir_filter *fdir_ftr,
1121 /**< @internal Remove a perfect filter rule on an Ethernet device */
1123 typedef int (*fdir_set_masks_t)(struct rte_eth_dev *dev,
1124 struct rte_fdir_masks *fdir_masks);
1125 /**< @internal Setup flow director masks on an Ethernet device */
1127 typedef int (*flow_ctrl_get_t)(struct rte_eth_dev *dev,
1128 struct rte_eth_fc_conf *fc_conf);
1129 /**< @internal Get current flow control parameter on an Ethernet device */
1131 typedef int (*flow_ctrl_set_t)(struct rte_eth_dev *dev,
1132 struct rte_eth_fc_conf *fc_conf);
1133 /**< @internal Setup flow control parameter on an Ethernet device */
1135 typedef int (*priority_flow_ctrl_set_t)(struct rte_eth_dev *dev,
1136 struct rte_eth_pfc_conf *pfc_conf);
1137 /**< @internal Setup priority flow control parameter on an Ethernet device */
1139 typedef int (*reta_update_t)(struct rte_eth_dev *dev,
1140 struct rte_eth_rss_reta_entry64 *reta_conf,
1141 uint16_t reta_size);
1142 /**< @internal Update RSS redirection table on an Ethernet device */
1144 typedef int (*reta_query_t)(struct rte_eth_dev *dev,
1145 struct rte_eth_rss_reta_entry64 *reta_conf,
1146 uint16_t reta_size);
1147 /**< @internal Query RSS redirection table on an Ethernet device */
1149 typedef int (*rss_hash_update_t)(struct rte_eth_dev *dev,
1150 struct rte_eth_rss_conf *rss_conf);
1151 /**< @internal Update RSS hash configuration of an Ethernet device */
1153 typedef int (*rss_hash_conf_get_t)(struct rte_eth_dev *dev,
1154 struct rte_eth_rss_conf *rss_conf);
1155 /**< @internal Get current RSS hash configuration of an Ethernet device */
1157 typedef int (*eth_dev_led_on_t)(struct rte_eth_dev *dev);
1158 /**< @internal Turn on SW controllable LED on an Ethernet device */
1160 typedef int (*eth_dev_led_off_t)(struct rte_eth_dev *dev);
1161 /**< @internal Turn off SW controllable LED on an Ethernet device */
1163 typedef void (*eth_mac_addr_remove_t)(struct rte_eth_dev *dev, uint32_t index);
1164 /**< @internal Remove MAC address from receive address register */
1166 typedef void (*eth_mac_addr_add_t)(struct rte_eth_dev *dev,
1167 struct ether_addr *mac_addr,
1170 /**< @internal Set a MAC address into Receive Address Address Register */
1172 typedef void (*eth_mac_addr_set_t)(struct rte_eth_dev *dev,
1173 struct ether_addr *mac_addr);
1174 /**< @internal Set a MAC address into Receive Address Address Register */
1176 typedef int (*eth_uc_hash_table_set_t)(struct rte_eth_dev *dev,
1177 struct ether_addr *mac_addr,
1179 /**< @internal Set a Unicast Hash bitmap */
1181 typedef int (*eth_uc_all_hash_table_set_t)(struct rte_eth_dev *dev,
1183 /**< @internal Set all Unicast Hash bitmap */
1185 typedef int (*eth_set_vf_rx_mode_t)(struct rte_eth_dev *dev,
1189 /**< @internal Set a VF receive mode */
1191 typedef int (*eth_set_vf_rx_t)(struct rte_eth_dev *dev,
1194 /**< @internal Set a VF receive mode */
1196 typedef int (*eth_set_vf_tx_t)(struct rte_eth_dev *dev,
1199 /**< @internal Enable or disable a VF transmit */
1201 typedef int (*eth_set_vf_vlan_filter_t)(struct rte_eth_dev *dev,
1205 /**< @internal Set VF VLAN pool filter */
1207 typedef int (*eth_set_queue_rate_limit_t)(struct rte_eth_dev *dev,
1210 /**< @internal Set queue TX rate */
1212 typedef int (*eth_set_vf_rate_limit_t)(struct rte_eth_dev *dev,
1216 /**< @internal Set VF TX rate */
1218 typedef int (*eth_mirror_rule_set_t)(struct rte_eth_dev *dev,
1219 struct rte_eth_mirror_conf *mirror_conf,
1222 /**< @internal Add a traffic mirroring rule on an Ethernet device */
1224 typedef int (*eth_mirror_rule_reset_t)(struct rte_eth_dev *dev,
1226 /**< @internal Remove a traffic mirroring rule on an Ethernet device */
1228 typedef int (*eth_udp_tunnel_add_t)(struct rte_eth_dev *dev,
1229 struct rte_eth_udp_tunnel *tunnel_udp);
1230 /**< @internal Add tunneling UDP info */
1232 typedef int (*eth_udp_tunnel_del_t)(struct rte_eth_dev *dev,
1233 struct rte_eth_udp_tunnel *tunnel_udp);
1234 /**< @internal Delete tunneling UDP info */
1236 typedef int (*eth_set_mc_addr_list_t)(struct rte_eth_dev *dev,
1237 struct ether_addr *mc_addr_set,
1238 uint32_t nb_mc_addr);
1239 /**< @internal set the list of multicast addresses on an Ethernet device */
1241 #ifdef RTE_NIC_BYPASS
1244 RTE_BYPASS_MODE_NONE,
1245 RTE_BYPASS_MODE_NORMAL,
1246 RTE_BYPASS_MODE_BYPASS,
1247 RTE_BYPASS_MODE_ISOLATE,
1248 RTE_BYPASS_MODE_NUM,
1251 #define RTE_BYPASS_MODE_VALID(x) \
1252 ((x) > RTE_BYPASS_MODE_NONE && (x) < RTE_BYPASS_MODE_NUM)
1255 RTE_BYPASS_EVENT_NONE,
1256 RTE_BYPASS_EVENT_START,
1257 RTE_BYPASS_EVENT_OS_ON = RTE_BYPASS_EVENT_START,
1258 RTE_BYPASS_EVENT_POWER_ON,
1259 RTE_BYPASS_EVENT_OS_OFF,
1260 RTE_BYPASS_EVENT_POWER_OFF,
1261 RTE_BYPASS_EVENT_TIMEOUT,
1262 RTE_BYPASS_EVENT_NUM
1265 #define RTE_BYPASS_EVENT_VALID(x) \
1266 ((x) > RTE_BYPASS_EVENT_NONE && (x) < RTE_BYPASS_MODE_NUM)
1269 RTE_BYPASS_TMT_OFF, /* timeout disabled. */
1270 RTE_BYPASS_TMT_1_5_SEC, /* timeout for 1.5 seconds */
1271 RTE_BYPASS_TMT_2_SEC, /* timeout for 2 seconds */
1272 RTE_BYPASS_TMT_3_SEC, /* timeout for 3 seconds */
1273 RTE_BYPASS_TMT_4_SEC, /* timeout for 4 seconds */
1274 RTE_BYPASS_TMT_8_SEC, /* timeout for 8 seconds */
1275 RTE_BYPASS_TMT_16_SEC, /* timeout for 16 seconds */
1276 RTE_BYPASS_TMT_32_SEC, /* timeout for 32 seconds */
1280 #define RTE_BYPASS_TMT_VALID(x) \
1281 ((x) == RTE_BYPASS_TMT_OFF || \
1282 ((x) > RTE_BYPASS_TMT_OFF && (x) < RTE_BYPASS_TMT_NUM))
1284 typedef void (*bypass_init_t)(struct rte_eth_dev *dev);
1285 typedef int32_t (*bypass_state_set_t)(struct rte_eth_dev *dev, uint32_t *new_state);
1286 typedef int32_t (*bypass_state_show_t)(struct rte_eth_dev *dev, uint32_t *state);
1287 typedef int32_t (*bypass_event_set_t)(struct rte_eth_dev *dev, uint32_t state, uint32_t event);
1288 typedef int32_t (*bypass_event_show_t)(struct rte_eth_dev *dev, uint32_t event_shift, uint32_t *event);
1289 typedef int32_t (*bypass_wd_timeout_set_t)(struct rte_eth_dev *dev, uint32_t timeout);
1290 typedef int32_t (*bypass_wd_timeout_show_t)(struct rte_eth_dev *dev, uint32_t *wd_timeout);
1291 typedef int32_t (*bypass_ver_show_t)(struct rte_eth_dev *dev, uint32_t *ver);
1292 typedef int32_t (*bypass_wd_reset_t)(struct rte_eth_dev *dev);
1295 typedef int (*eth_filter_ctrl_t)(struct rte_eth_dev *dev,
1296 enum rte_filter_type filter_type,
1297 enum rte_filter_op filter_op,
1299 /**< @internal Take operations to assigned filter type on an Ethernet device */
1302 * @internal A structure containing the functions exported by an Ethernet driver.
1304 struct eth_dev_ops {
1305 eth_dev_configure_t dev_configure; /**< Configure device. */
1306 eth_dev_start_t dev_start; /**< Start device. */
1307 eth_dev_stop_t dev_stop; /**< Stop device. */
1308 eth_dev_set_link_up_t dev_set_link_up; /**< Device link up. */
1309 eth_dev_set_link_down_t dev_set_link_down; /**< Device link down. */
1310 eth_dev_close_t dev_close; /**< Close device. */
1311 eth_promiscuous_enable_t promiscuous_enable; /**< Promiscuous ON. */
1312 eth_promiscuous_disable_t promiscuous_disable;/**< Promiscuous OFF. */
1313 eth_allmulticast_enable_t allmulticast_enable;/**< RX multicast ON. */
1314 eth_allmulticast_disable_t allmulticast_disable;/**< RX multicast OF. */
1315 eth_link_update_t link_update; /**< Get device link state. */
1316 eth_stats_get_t stats_get; /**< Get generic device statistics. */
1317 eth_stats_reset_t stats_reset; /**< Reset generic device statistics. */
1318 eth_xstats_get_t xstats_get; /**< Get extended device statistics. */
1319 eth_xstats_reset_t xstats_reset; /**< Reset extended device statistics. */
1320 eth_queue_stats_mapping_set_t queue_stats_mapping_set;
1321 /**< Configure per queue stat counter mapping. */
1322 eth_dev_infos_get_t dev_infos_get; /**< Get device info. */
1323 mtu_set_t mtu_set; /**< Set MTU. */
1324 vlan_filter_set_t vlan_filter_set; /**< Filter VLAN Setup. */
1325 vlan_tpid_set_t vlan_tpid_set; /**< Outer VLAN TPID Setup. */
1326 vlan_strip_queue_set_t vlan_strip_queue_set; /**< VLAN Stripping on queue. */
1327 vlan_offload_set_t vlan_offload_set; /**< Set VLAN Offload. */
1328 vlan_pvid_set_t vlan_pvid_set; /**< Set port based TX VLAN insertion */
1329 eth_queue_start_t rx_queue_start;/**< Start RX for a queue.*/
1330 eth_queue_stop_t rx_queue_stop;/**< Stop RX for a queue.*/
1331 eth_queue_start_t tx_queue_start;/**< Start TX for a queue.*/
1332 eth_queue_stop_t tx_queue_stop;/**< Stop TX for a queue.*/
1333 eth_rx_queue_setup_t rx_queue_setup;/**< Set up device RX queue.*/
1334 eth_queue_release_t rx_queue_release;/**< Release RX queue.*/
1335 eth_rx_queue_count_t rx_queue_count; /**< Get Rx queue count. */
1336 eth_rx_descriptor_done_t rx_descriptor_done; /**< Check rxd DD bit */
1337 eth_tx_queue_setup_t tx_queue_setup;/**< Set up device TX queue.*/
1338 eth_queue_release_t tx_queue_release;/**< Release TX queue.*/
1339 eth_dev_led_on_t dev_led_on; /**< Turn on LED. */
1340 eth_dev_led_off_t dev_led_off; /**< Turn off LED. */
1341 flow_ctrl_get_t flow_ctrl_get; /**< Get flow control. */
1342 flow_ctrl_set_t flow_ctrl_set; /**< Setup flow control. */
1343 priority_flow_ctrl_set_t priority_flow_ctrl_set; /**< Setup priority flow control.*/
1344 eth_mac_addr_remove_t mac_addr_remove; /**< Remove MAC address */
1345 eth_mac_addr_add_t mac_addr_add; /**< Add a MAC address */
1346 eth_mac_addr_set_t mac_addr_set; /**< Set a MAC address */
1347 eth_uc_hash_table_set_t uc_hash_table_set; /**< Set Unicast Table Array */
1348 eth_uc_all_hash_table_set_t uc_all_hash_table_set; /**< Set Unicast hash bitmap */
1349 eth_mirror_rule_set_t mirror_rule_set; /**< Add a traffic mirror rule.*/
1350 eth_mirror_rule_reset_t mirror_rule_reset; /**< reset a traffic mirror rule.*/
1351 eth_set_vf_rx_mode_t set_vf_rx_mode; /**< Set VF RX mode */
1352 eth_set_vf_rx_t set_vf_rx; /**< enable/disable a VF receive */
1353 eth_set_vf_tx_t set_vf_tx; /**< enable/disable a VF transmit */
1354 eth_set_vf_vlan_filter_t set_vf_vlan_filter; /**< Set VF VLAN filter */
1355 eth_udp_tunnel_add_t udp_tunnel_add;
1356 eth_udp_tunnel_del_t udp_tunnel_del;
1357 eth_set_queue_rate_limit_t set_queue_rate_limit; /**< Set queue rate limit */
1358 eth_set_vf_rate_limit_t set_vf_rate_limit; /**< Set VF rate limit */
1360 /** Add a signature filter. */
1361 fdir_add_signature_filter_t fdir_add_signature_filter;
1362 /** Update a signature filter. */
1363 fdir_update_signature_filter_t fdir_update_signature_filter;
1364 /** Remove a signature filter. */
1365 fdir_remove_signature_filter_t fdir_remove_signature_filter;
1366 /** Get information about FDIR status. */
1367 fdir_infos_get_t fdir_infos_get;
1368 /** Add a perfect filter. */
1369 fdir_add_perfect_filter_t fdir_add_perfect_filter;
1370 /** Update a perfect filter. */
1371 fdir_update_perfect_filter_t fdir_update_perfect_filter;
1372 /** Remove a perfect filter. */
1373 fdir_remove_perfect_filter_t fdir_remove_perfect_filter;
1374 /** Setup masks for FDIR filtering. */
1375 fdir_set_masks_t fdir_set_masks;
1376 /** Update redirection table. */
1377 reta_update_t reta_update;
1378 /** Query redirection table. */
1379 reta_query_t reta_query;
1380 /* bypass control */
1381 #ifdef RTE_NIC_BYPASS
1382 bypass_init_t bypass_init;
1383 bypass_state_set_t bypass_state_set;
1384 bypass_state_show_t bypass_state_show;
1385 bypass_event_set_t bypass_event_set;
1386 bypass_event_show_t bypass_event_show;
1387 bypass_wd_timeout_set_t bypass_wd_timeout_set;
1388 bypass_wd_timeout_show_t bypass_wd_timeout_show;
1389 bypass_ver_show_t bypass_ver_show;
1390 bypass_wd_reset_t bypass_wd_reset;
1393 /** Configure RSS hash protocols. */
1394 rss_hash_update_t rss_hash_update;
1395 /** Get current RSS hash configuration. */
1396 rss_hash_conf_get_t rss_hash_conf_get;
1397 eth_filter_ctrl_t filter_ctrl; /**< common filter control*/
1398 eth_set_mc_addr_list_t set_mc_addr_list; /**< set list of mcast addrs */
1402 * Function type used for RX packet processing packet callbacks.
1404 * The callback function is called on RX with a burst of packets that have
1405 * been received on the given port and queue.
1408 * The Ethernet port on which RX is being performed.
1410 * The queue on the Ethernet port which is being used to receive the packets.
1412 * The burst of packets that have just been received.
1414 * The number of packets in the burst pointed to by "pkts".
1416 * The max number of packets that can be stored in the "pkts" array.
1418 * The arbitrary user parameter passed in by the application when the callback
1419 * was originally configured.
1421 * The number of packets returned to the user.
1423 typedef uint16_t (*rte_rx_callback_fn)(uint8_t port, uint16_t queue,
1424 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts,
1428 * Function type used for TX packet processing packet callbacks.
1430 * The callback function is called on TX with a burst of packets immediately
1431 * before the packets are put onto the hardware queue for transmission.
1434 * The Ethernet port on which TX is being performed.
1436 * The queue on the Ethernet port which is being used to transmit the packets.
1438 * The burst of packets that are about to be transmitted.
1440 * The number of packets in the burst pointed to by "pkts".
1442 * The arbitrary user parameter passed in by the application when the callback
1443 * was originally configured.
1445 * The number of packets to be written to the NIC.
1447 typedef uint16_t (*rte_tx_callback_fn)(uint8_t port, uint16_t queue,
1448 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param);
1452 * Structure used to hold information about the callbacks to be called for a
1453 * queue on RX and TX.
1455 struct rte_eth_rxtx_callback {
1456 struct rte_eth_rxtx_callback *next;
1458 rte_rx_callback_fn rx;
1459 rte_tx_callback_fn tx;
1465 * The eth device type.
1467 enum rte_eth_dev_type {
1468 RTE_ETH_DEV_UNKNOWN, /**< unknown device type */
1470 /**< Physical function and Virtual function of PCI devices */
1471 RTE_ETH_DEV_VIRTUAL, /**< non hardware device */
1472 RTE_ETH_DEV_MAX /**< max value of this enum */
1477 * The generic data structure associated with each ethernet device.
1479 * Pointers to burst-oriented packet receive and transmit functions are
1480 * located at the beginning of the structure, along with the pointer to
1481 * where all the data elements for the particular device are stored in shared
1482 * memory. This split allows the function pointer and driver data to be per-
1483 * process, while the actual configuration data for the device is shared.
1485 struct rte_eth_dev {
1486 eth_rx_burst_t rx_pkt_burst; /**< Pointer to PMD receive function. */
1487 eth_tx_burst_t tx_pkt_burst; /**< Pointer to PMD transmit function. */
1488 struct rte_eth_dev_data *data; /**< Pointer to device data */
1489 const struct eth_driver *driver;/**< Driver for this device */
1490 const struct eth_dev_ops *dev_ops; /**< Functions exported by PMD */
1491 struct rte_pci_device *pci_dev; /**< PCI info. supplied by probing */
1492 /** User application callbacks for NIC interrupts */
1493 struct rte_eth_dev_cb_list link_intr_cbs;
1495 * User-supplied functions called from rx_burst to post-process
1496 * received packets before passing them to the user
1498 struct rte_eth_rxtx_callback *post_rx_burst_cbs[RTE_MAX_QUEUES_PER_PORT];
1500 * User-supplied functions called from tx_burst to pre-process
1501 * received packets before passing them to the driver for transmission.
1503 struct rte_eth_rxtx_callback *pre_tx_burst_cbs[RTE_MAX_QUEUES_PER_PORT];
1504 uint8_t attached; /**< Flag indicating the port is attached */
1505 enum rte_eth_dev_type dev_type; /**< Flag indicating the device type */
1508 struct rte_eth_dev_sriov {
1509 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */
1510 uint8_t nb_q_per_pool; /**< rx queue number per pool */
1511 uint16_t def_vmdq_idx; /**< Default pool num used for PF */
1512 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */
1514 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
1516 #define RTE_ETH_NAME_MAX_LEN (32)
1520 * The data part, with no function pointers, associated with each ethernet device.
1522 * This structure is safe to place in shared memory to be common among different
1523 * processes in a multi-process configuration.
1525 struct rte_eth_dev_data {
1526 char name[RTE_ETH_NAME_MAX_LEN]; /**< Unique identifier name */
1528 void **rx_queues; /**< Array of pointers to RX queues. */
1529 void **tx_queues; /**< Array of pointers to TX queues. */
1530 uint16_t nb_rx_queues; /**< Number of RX queues. */
1531 uint16_t nb_tx_queues; /**< Number of TX queues. */
1533 struct rte_eth_dev_sriov sriov; /**< SRIOV data */
1535 void *dev_private; /**< PMD-specific private data */
1537 struct rte_eth_link dev_link;
1538 /**< Link-level information & status */
1540 struct rte_eth_conf dev_conf; /**< Configuration applied to device. */
1541 uint16_t mtu; /**< Maximum Transmission Unit. */
1543 uint32_t min_rx_buf_size;
1544 /**< Common rx buffer size handled by all queues */
1546 uint64_t rx_mbuf_alloc_failed; /**< RX ring mbuf allocation failures. */
1547 struct ether_addr* mac_addrs;/**< Device Ethernet Link address. */
1548 uint64_t mac_pool_sel[ETH_NUM_RECEIVE_MAC_ADDR];
1549 /** bitmap array of associating Ethernet MAC addresses to pools */
1550 struct ether_addr* hash_mac_addrs;
1551 /** Device Ethernet MAC addresses of hash filtering. */
1552 uint8_t port_id; /**< Device [external] port identifier. */
1553 uint8_t promiscuous : 1, /**< RX promiscuous mode ON(1) / OFF(0). */
1554 scattered_rx : 1, /**< RX of scattered packets is ON(1) / OFF(0) */
1555 lro : 1, /**< RX LRO is ON(1) / OFF(0) */
1556 all_multicast : 1, /**< RX all multicast mode ON(1) / OFF(0). */
1557 dev_started : 1; /**< Device state: STARTED(1) / STOPPED(0). */
1562 * The pool of *rte_eth_dev* structures. The size of the pool
1563 * is configured at compile-time in the <rte_ethdev.c> file.
1565 extern struct rte_eth_dev rte_eth_devices[];
1568 * Get the total number of Ethernet devices that have been successfully
1569 * initialized by the [matching] Ethernet driver during the PCI probing phase.
1570 * All devices whose port identifier is in the range
1571 * [0, rte_eth_dev_count() - 1] can be operated on by network applications
1572 * immediately after invoking rte_eal_init().
1573 * If the application unplugs a port using hotplug function, The enabled port
1574 * numbers may be noncontiguous. In the case, the applications need to manage
1575 * enabled port by themselves.
1578 * - The total number of usable Ethernet devices.
1580 extern uint8_t rte_eth_dev_count(void);
1583 * Function for internal use by port hotplug functions.
1584 * Returns a ethdev slot specified by the unique identifier name.
1586 * The pointer to the Unique identifier name for each Ethernet device
1588 * - The pointer to the ethdev slot, on success. NULL on error
1590 extern struct rte_eth_dev *rte_eth_dev_allocated(const char *name);
1593 * Function for internal use by dummy drivers primarily, e.g. ring-based
1595 * Allocates a new ethdev slot for an ethernet device and returns the pointer
1596 * to that slot for the driver to use.
1598 * @param name Unique identifier name for each Ethernet device
1599 * @param type Device type of this Ethernet device
1601 * - Slot in the rte_dev_devices array for a new device;
1603 struct rte_eth_dev *rte_eth_dev_allocate(const char *name,
1604 enum rte_eth_dev_type type);
1607 * Function for internal use by dummy drivers primarily, e.g. ring-based
1609 * Release the specified ethdev port.
1612 * The *eth_dev* pointer is the address of the *rte_eth_dev* structure.
1614 * - 0 on success, negative on error
1616 int rte_eth_dev_release_port(struct rte_eth_dev *eth_dev);
1619 * Attach a new Ethernet device specified by aruguments.
1622 * A pointer to a strings array describing the new device
1623 * to be attached. The strings should be a pci address like
1624 * '0000:01:00.0' or virtual device name like 'eth_pcap0'.
1626 * A pointer to a port identifier actually attached.
1628 * 0 on success and port_id is filled, negative on error
1630 int rte_eth_dev_attach(const char *devargs, uint8_t *port_id);
1633 * Detach a Ethernet device specified by port identifier.
1636 * The port identifier of the device to detach.
1638 * A pointer to a device name actually detached.
1640 * 0 on success and devname is filled, negative on error
1642 int rte_eth_dev_detach(uint8_t port_id, char *devname);
1647 * Initialization function of an Ethernet driver invoked for each matching
1648 * Ethernet PCI device detected during the PCI probing phase.
1651 * The *eth_dev* pointer is the address of the *rte_eth_dev* structure
1652 * associated with the matching device and which have been [automatically]
1653 * allocated in the *rte_eth_devices* array.
1654 * The *eth_dev* structure is supplied to the driver initialization function
1655 * with the following fields already initialized:
1657 * - *pci_dev*: Holds the pointers to the *rte_pci_device* structure which
1658 * contains the generic PCI information of the matching device.
1660 * - *driver*: Holds the pointer to the *eth_driver* structure.
1662 * - *dev_private*: Holds a pointer to the device private data structure.
1664 * - *mtu*: Contains the default Ethernet maximum frame length (1500).
1666 * - *port_id*: Contains the port index of the device (actually the index
1667 * of the *eth_dev* structure in the *rte_eth_devices* array).
1670 * - 0: Success, the device is properly initialized by the driver.
1671 * In particular, the driver MUST have set up the *dev_ops* pointer
1672 * of the *eth_dev* structure.
1673 * - <0: Error code of the device initialization failure.
1675 typedef int (*eth_dev_init_t)(struct rte_eth_dev *eth_dev);
1679 * Finalization function of an Ethernet driver invoked for each matching
1680 * Ethernet PCI device detected during the PCI closing phase.
1683 * The *eth_dev* pointer is the address of the *rte_eth_dev* structure
1684 * associated with the matching device and which have been [automatically]
1685 * allocated in the *rte_eth_devices* array.
1687 * - 0: Success, the device is properly finalized by the driver.
1688 * In particular, the driver MUST free the *dev_ops* pointer
1689 * of the *eth_dev* structure.
1690 * - <0: Error code of the device initialization failure.
1692 typedef int (*eth_dev_uninit_t)(struct rte_eth_dev *eth_dev);
1696 * The structure associated with a PMD Ethernet driver.
1698 * Each Ethernet driver acts as a PCI driver and is represented by a generic
1699 * *eth_driver* structure that holds:
1701 * - An *rte_pci_driver* structure (which must be the first field).
1703 * - The *eth_dev_init* function invoked for each matching PCI device.
1705 * - The *eth_dev_uninit* function invoked for each matching PCI device.
1707 * - The size of the private data to allocate for each matching device.
1710 struct rte_pci_driver pci_drv; /**< The PMD is also a PCI driver. */
1711 eth_dev_init_t eth_dev_init; /**< Device init function. */
1712 eth_dev_uninit_t eth_dev_uninit; /**< Device uninit function. */
1713 unsigned int dev_private_size; /**< Size of device private data. */
1718 * A function invoked by the initialization function of an Ethernet driver
1719 * to simultaneously register itself as a PCI driver and as an Ethernet
1720 * Poll Mode Driver (PMD).
1723 * The pointer to the *eth_driver* structure associated with
1724 * the Ethernet driver.
1726 extern void rte_eth_driver_register(struct eth_driver *eth_drv);
1729 * Configure an Ethernet device.
1730 * This function must be invoked first before any other function in the
1731 * Ethernet API. This function can also be re-invoked when a device is in the
1735 * The port identifier of the Ethernet device to configure.
1736 * @param nb_rx_queue
1737 * The number of receive queues to set up for the Ethernet device.
1738 * @param nb_tx_queue
1739 * The number of transmit queues to set up for the Ethernet device.
1741 * The pointer to the configuration data to be used for the Ethernet device.
1742 * The *rte_eth_conf* structure includes:
1743 * - the hardware offload features to activate, with dedicated fields for
1744 * each statically configurable offload hardware feature provided by
1745 * Ethernet devices, such as IP checksum or VLAN tag stripping for
1747 * - the Receive Side Scaling (RSS) configuration when using multiple RX
1750 * Embedding all configuration information in a single data structure
1751 * is the more flexible method that allows the addition of new features
1752 * without changing the syntax of the API.
1754 * - 0: Success, device configured.
1755 * - <0: Error code returned by the driver configuration function.
1757 extern int rte_eth_dev_configure(uint8_t port_id,
1758 uint16_t nb_rx_queue,
1759 uint16_t nb_tx_queue,
1760 const struct rte_eth_conf *eth_conf);
1763 * Allocate and set up a receive queue for an Ethernet device.
1765 * The function allocates a contiguous block of memory for *nb_rx_desc*
1766 * receive descriptors from a memory zone associated with *socket_id*
1767 * and initializes each receive descriptor with a network buffer allocated
1768 * from the memory pool *mb_pool*.
1771 * The port identifier of the Ethernet device.
1772 * @param rx_queue_id
1773 * The index of the receive queue to set up.
1774 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1775 * to rte_eth_dev_configure().
1777 * The number of receive descriptors to allocate for the receive ring.
1779 * The *socket_id* argument is the socket identifier in case of NUMA.
1780 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1781 * the DMA memory allocated for the receive descriptors of the ring.
1783 * The pointer to the configuration data to be used for the receive queue.
1784 * NULL value is allowed, in which case default RX configuration
1786 * The *rx_conf* structure contains an *rx_thresh* structure with the values
1787 * of the Prefetch, Host, and Write-Back threshold registers of the receive
1790 * The pointer to the memory pool from which to allocate *rte_mbuf* network
1791 * memory buffers to populate each descriptor of the receive ring.
1793 * - 0: Success, receive queue correctly set up.
1794 * - -EINVAL: The size of network buffers which can be allocated from the
1795 * memory pool does not fit the various buffer sizes allowed by the
1796 * device controller.
1797 * - -ENOMEM: Unable to allocate the receive ring descriptors or to
1798 * allocate network memory buffers from the memory pool when
1799 * initializing receive descriptors.
1801 extern int rte_eth_rx_queue_setup(uint8_t port_id, uint16_t rx_queue_id,
1802 uint16_t nb_rx_desc, unsigned int socket_id,
1803 const struct rte_eth_rxconf *rx_conf,
1804 struct rte_mempool *mb_pool);
1807 * Allocate and set up a transmit queue for an Ethernet device.
1810 * The port identifier of the Ethernet device.
1811 * @param tx_queue_id
1812 * The index of the transmit queue to set up.
1813 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1814 * to rte_eth_dev_configure().
1816 * The number of transmit descriptors to allocate for the transmit ring.
1818 * The *socket_id* argument is the socket identifier in case of NUMA.
1819 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1820 * the DMA memory allocated for the transmit descriptors of the ring.
1822 * The pointer to the configuration data to be used for the transmit queue.
1823 * NULL value is allowed, in which case default RX configuration
1825 * The *tx_conf* structure contains the following data:
1826 * - The *tx_thresh* structure with the values of the Prefetch, Host, and
1827 * Write-Back threshold registers of the transmit ring.
1828 * When setting Write-Back threshold to the value greater then zero,
1829 * *tx_rs_thresh* value should be explicitly set to one.
1830 * - The *tx_free_thresh* value indicates the [minimum] number of network
1831 * buffers that must be pending in the transmit ring to trigger their
1832 * [implicit] freeing by the driver transmit function.
1833 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit
1834 * descriptors that must be pending in the transmit ring before setting the
1835 * RS bit on a descriptor by the driver transmit function.
1836 * The *tx_rs_thresh* value should be less or equal then
1837 * *tx_free_thresh* value, and both of them should be less then
1839 * - The *txq_flags* member contains flags to pass to the TX queue setup
1840 * function to configure the behavior of the TX queue. This should be set
1841 * to 0 if no special configuration is required.
1843 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces
1844 * the transmit function to use default values.
1846 * - 0: Success, the transmit queue is correctly set up.
1847 * - -ENOMEM: Unable to allocate the transmit ring descriptors.
1849 extern int rte_eth_tx_queue_setup(uint8_t port_id, uint16_t tx_queue_id,
1850 uint16_t nb_tx_desc, unsigned int socket_id,
1851 const struct rte_eth_txconf *tx_conf);
1854 * Return the NUMA socket to which an Ethernet device is connected
1857 * The port identifier of the Ethernet device
1859 * The NUMA socket id to which the Ethernet device is connected or
1860 * a default of zero if the socket could not be determined.
1861 * -1 is returned is the port_id value is out of range.
1863 extern int rte_eth_dev_socket_id(uint8_t port_id);
1866 * Allocate mbuf from mempool, setup the DMA physical address
1867 * and then start RX for specified queue of a port. It is used
1868 * when rx_deferred_start flag of the specified queue is true.
1871 * The port identifier of the Ethernet device
1872 * @param rx_queue_id
1873 * The index of the rx queue to update the ring.
1874 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1875 * to rte_eth_dev_configure().
1877 * - 0: Success, the transmit queue is correctly set up.
1878 * - -EINVAL: The port_id or the queue_id out of range.
1879 * - -ENOTSUP: The function not supported in PMD driver.
1881 extern int rte_eth_dev_rx_queue_start(uint8_t port_id, uint16_t rx_queue_id);
1884 * Stop specified RX queue of a port
1887 * The port identifier of the Ethernet device
1888 * @param rx_queue_id
1889 * The index of the rx queue to update the ring.
1890 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1891 * to rte_eth_dev_configure().
1893 * - 0: Success, the transmit queue is correctly set up.
1894 * - -EINVAL: The port_id or the queue_id out of range.
1895 * - -ENOTSUP: The function not supported in PMD driver.
1897 extern int rte_eth_dev_rx_queue_stop(uint8_t port_id, uint16_t rx_queue_id);
1900 * Start TX for specified queue of a port. It is used when tx_deferred_start
1901 * flag of the specified queue is true.
1904 * The port identifier of the Ethernet device
1905 * @param tx_queue_id
1906 * The index of the tx queue to update the ring.
1907 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1908 * to rte_eth_dev_configure().
1910 * - 0: Success, the transmit queue is correctly set up.
1911 * - -EINVAL: The port_id or the queue_id out of range.
1912 * - -ENOTSUP: The function not supported in PMD driver.
1914 extern int rte_eth_dev_tx_queue_start(uint8_t port_id, uint16_t tx_queue_id);
1917 * Stop specified TX queue of a port
1920 * The port identifier of the Ethernet device
1921 * @param tx_queue_id
1922 * The index of the tx queue to update the ring.
1923 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1924 * to rte_eth_dev_configure().
1926 * - 0: Success, the transmit queue is correctly set up.
1927 * - -EINVAL: The port_id or the queue_id out of range.
1928 * - -ENOTSUP: The function not supported in PMD driver.
1930 extern int rte_eth_dev_tx_queue_stop(uint8_t port_id, uint16_t tx_queue_id);
1935 * Start an Ethernet device.
1937 * The device start step is the last one and consists of setting the configured
1938 * offload features and in starting the transmit and the receive units of the
1940 * On success, all basic functions exported by the Ethernet API (link status,
1941 * receive/transmit, and so on) can be invoked.
1944 * The port identifier of the Ethernet device.
1946 * - 0: Success, Ethernet device started.
1947 * - <0: Error code of the driver device start function.
1949 extern int rte_eth_dev_start(uint8_t port_id);
1952 * Stop an Ethernet device. The device can be restarted with a call to
1953 * rte_eth_dev_start()
1956 * The port identifier of the Ethernet device.
1958 extern void rte_eth_dev_stop(uint8_t port_id);
1962 * Link up an Ethernet device.
1964 * Set device link up will re-enable the device rx/tx
1965 * functionality after it is previously set device linked down.
1968 * The port identifier of the Ethernet device.
1970 * - 0: Success, Ethernet device linked up.
1971 * - <0: Error code of the driver device link up function.
1973 extern int rte_eth_dev_set_link_up(uint8_t port_id);
1976 * Link down an Ethernet device.
1977 * The device rx/tx functionality will be disabled if success,
1978 * and it can be re-enabled with a call to
1979 * rte_eth_dev_set_link_up()
1982 * The port identifier of the Ethernet device.
1984 extern int rte_eth_dev_set_link_down(uint8_t port_id);
1987 * Close an Ethernet device. The device cannot be restarted!
1990 * The port identifier of the Ethernet device.
1992 extern void rte_eth_dev_close(uint8_t port_id);
1995 * Enable receipt in promiscuous mode for an Ethernet device.
1998 * The port identifier of the Ethernet device.
2000 extern void rte_eth_promiscuous_enable(uint8_t port_id);
2003 * Disable receipt in promiscuous mode for an Ethernet device.
2006 * The port identifier of the Ethernet device.
2008 extern void rte_eth_promiscuous_disable(uint8_t port_id);
2011 * Return the value of promiscuous mode for an Ethernet device.
2014 * The port identifier of the Ethernet device.
2016 * - (1) if promiscuous is enabled
2017 * - (0) if promiscuous is disabled.
2020 extern int rte_eth_promiscuous_get(uint8_t port_id);
2023 * Enable the receipt of any multicast frame by an Ethernet device.
2026 * The port identifier of the Ethernet device.
2028 extern void rte_eth_allmulticast_enable(uint8_t port_id);
2031 * Disable the receipt of all multicast frames by an Ethernet device.
2034 * The port identifier of the Ethernet device.
2036 extern void rte_eth_allmulticast_disable(uint8_t port_id);
2039 * Return the value of allmulticast mode for an Ethernet device.
2042 * The port identifier of the Ethernet device.
2044 * - (1) if allmulticast is enabled
2045 * - (0) if allmulticast is disabled.
2048 extern int rte_eth_allmulticast_get(uint8_t port_id);
2051 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2052 * or FULL-DUPLEX) of the physical link of an Ethernet device. It might need
2053 * to wait up to 9 seconds in it.
2056 * The port identifier of the Ethernet device.
2058 * A pointer to an *rte_eth_link* structure to be filled with
2059 * the status, the speed and the mode of the Ethernet device link.
2061 extern void rte_eth_link_get(uint8_t port_id, struct rte_eth_link *link);
2064 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
2065 * or FULL-DUPLEX) of the physical link of an Ethernet device. It is a no-wait
2066 * version of rte_eth_link_get().
2069 * The port identifier of the Ethernet device.
2071 * A pointer to an *rte_eth_link* structure to be filled with
2072 * the status, the speed and the mode of the Ethernet device link.
2074 extern void rte_eth_link_get_nowait(uint8_t port_id,
2075 struct rte_eth_link *link);
2078 * Retrieve the general I/O statistics of an Ethernet device.
2081 * The port identifier of the Ethernet device.
2083 * A pointer to a structure of type *rte_eth_stats* to be filled with
2084 * the values of device counters for the following set of statistics:
2085 * - *ipackets* with the total of successfully received packets.
2086 * - *opackets* with the total of successfully transmitted packets.
2087 * - *ibytes* with the total of successfully received bytes.
2088 * - *obytes* with the total of successfully transmitted bytes.
2089 * - *ierrors* with the total of erroneous received packets.
2090 * - *oerrors* with the total of failed transmitted packets.
2092 * Zero if successful. Non-zero otherwise.
2094 extern int rte_eth_stats_get(uint8_t port_id, struct rte_eth_stats *stats);
2097 * Reset the general I/O statistics of an Ethernet device.
2100 * The port identifier of the Ethernet device.
2102 extern void rte_eth_stats_reset(uint8_t port_id);
2105 * Retrieve extended statistics of an Ethernet device.
2108 * The port identifier of the Ethernet device.
2110 * A pointer to a table of structure of type *rte_eth_xstats*
2111 * to be filled with device statistics names and values.
2112 * This parameter can be set to NULL if n is 0.
2114 * The size of the stats table, which should be large enough to store
2115 * all the statistics of the device.
2117 * - positive value lower or equal to n: success. The return value
2118 * is the number of entries filled in the stats table.
2119 * - positive value higher than n: error, the given statistics table
2120 * is too small. The return value corresponds to the size that should
2121 * be given to succeed. The entries in the table are not valid and
2122 * shall not be used by the caller.
2123 * - negative value on error (invalid port id)
2125 extern int rte_eth_xstats_get(uint8_t port_id,
2126 struct rte_eth_xstats *xstats, unsigned n);
2129 * Reset extended statistics of an Ethernet device.
2132 * The port identifier of the Ethernet device.
2134 extern void rte_eth_xstats_reset(uint8_t port_id);
2137 * Set a mapping for the specified transmit queue to the specified per-queue
2138 * statistics counter.
2141 * The port identifier of the Ethernet device.
2142 * @param tx_queue_id
2143 * The index of the transmit queue for which a queue stats mapping is required.
2144 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2145 * to rte_eth_dev_configure().
2147 * The per-queue packet statistics functionality number that the transmit
2148 * queue is to be assigned.
2149 * The value must be in the range [0, RTE_MAX_ETHPORT_QUEUE_STATS_MAPS - 1].
2151 * Zero if successful. Non-zero otherwise.
2153 extern int rte_eth_dev_set_tx_queue_stats_mapping(uint8_t port_id,
2154 uint16_t tx_queue_id,
2158 * Set a mapping for the specified receive queue to the specified per-queue
2159 * statistics counter.
2162 * The port identifier of the Ethernet device.
2163 * @param rx_queue_id
2164 * The index of the receive queue for which a queue stats mapping is required.
2165 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2166 * to rte_eth_dev_configure().
2168 * The per-queue packet statistics functionality number that the receive
2169 * queue is to be assigned.
2170 * The value must be in the range [0, RTE_MAX_ETHPORT_QUEUE_STATS_MAPS - 1].
2172 * Zero if successful. Non-zero otherwise.
2174 extern int rte_eth_dev_set_rx_queue_stats_mapping(uint8_t port_id,
2175 uint16_t rx_queue_id,
2179 * Retrieve the Ethernet address of an Ethernet device.
2182 * The port identifier of the Ethernet device.
2184 * A pointer to a structure of type *ether_addr* to be filled with
2185 * the Ethernet address of the Ethernet device.
2187 extern void rte_eth_macaddr_get(uint8_t port_id, struct ether_addr *mac_addr);
2190 * Retrieve the contextual information of an Ethernet device.
2193 * The port identifier of the Ethernet device.
2195 * A pointer to a structure of type *rte_eth_dev_info* to be filled with
2196 * the contextual information of the Ethernet device.
2198 extern void rte_eth_dev_info_get(uint8_t port_id,
2199 struct rte_eth_dev_info *dev_info);
2202 * Retrieve the MTU of an Ethernet device.
2205 * The port identifier of the Ethernet device.
2207 * A pointer to a uint16_t where the retrieved MTU is to be stored.
2209 * - (0) if successful.
2210 * - (-ENODEV) if *port_id* invalid.
2212 extern int rte_eth_dev_get_mtu(uint8_t port_id, uint16_t *mtu);
2215 * Change the MTU of an Ethernet device.
2218 * The port identifier of the Ethernet device.
2220 * A uint16_t for the MTU to be applied.
2222 * - (0) if successful.
2223 * - (-ENOTSUP) if operation is not supported.
2224 * - (-ENODEV) if *port_id* invalid.
2225 * - (-EINVAL) if *mtu* invalid.
2227 extern int rte_eth_dev_set_mtu(uint8_t port_id, uint16_t mtu);
2230 * Enable/Disable hardware filtering by an Ethernet device of received
2231 * VLAN packets tagged with a given VLAN Tag Identifier.
2234 * The port identifier of the Ethernet device.
2236 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
2238 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*.
2239 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*.
2241 * - (0) if successful.
2242 * - (-ENOSUP) if hardware-assisted VLAN filtering not configured.
2243 * - (-ENODEV) if *port_id* invalid.
2244 * - (-ENOSYS) if VLAN filtering on *port_id* disabled.
2245 * - (-EINVAL) if *vlan_id* > 4095.
2247 extern int rte_eth_dev_vlan_filter(uint8_t port_id, uint16_t vlan_id , int on);
2250 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device.
2251 * 82599/X540/X550 can support VLAN stripping at the rx queue level
2254 * The port identifier of the Ethernet device.
2255 * @param rx_queue_id
2256 * The index of the receive queue for which a queue stats mapping is required.
2257 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2258 * to rte_eth_dev_configure().
2260 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port.
2261 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port.
2263 * - (0) if successful.
2264 * - (-ENOSUP) if hardware-assisted VLAN stripping not configured.
2265 * - (-ENODEV) if *port_id* invalid.
2266 * - (-EINVAL) if *rx_queue_id* invalid.
2268 extern int rte_eth_dev_set_vlan_strip_on_queue(uint8_t port_id,
2269 uint16_t rx_queue_id, int on);
2272 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to
2273 * the VLAN Header. This is a register setup available on some Intel NIC, not
2274 * but all, please check the data sheet for availability.
2277 * The port identifier of the Ethernet device.
2279 * The Tag Protocol ID
2281 * - (0) if successful.
2282 * - (-ENOSUP) if hardware-assisted VLAN TPID setup is not supported.
2283 * - (-ENODEV) if *port_id* invalid.
2285 extern int rte_eth_dev_set_vlan_ether_type(uint8_t port_id, uint16_t tag_type);
2288 * Set VLAN offload configuration on an Ethernet device
2289 * Enable/Disable Extended VLAN by an Ethernet device, This is a register setup
2290 * available on some Intel NIC, not but all, please check the data sheet for
2292 * Enable/Disable VLAN Strip can be done on rx queue for certain NIC, but here
2293 * the configuration is applied on the port level.
2296 * The port identifier of the Ethernet device.
2297 * @param offload_mask
2298 * The VLAN Offload bit mask can be mixed use with "OR"
2299 * ETH_VLAN_STRIP_OFFLOAD
2300 * ETH_VLAN_FILTER_OFFLOAD
2301 * ETH_VLAN_EXTEND_OFFLOAD
2303 * - (0) if successful.
2304 * - (-ENOSUP) if hardware-assisted VLAN filtering not configured.
2305 * - (-ENODEV) if *port_id* invalid.
2307 extern int rte_eth_dev_set_vlan_offload(uint8_t port_id, int offload_mask);
2310 * Read VLAN Offload configuration from an Ethernet device
2313 * The port identifier of the Ethernet device.
2315 * - (>0) if successful. Bit mask to indicate
2316 * ETH_VLAN_STRIP_OFFLOAD
2317 * ETH_VLAN_FILTER_OFFLOAD
2318 * ETH_VLAN_EXTEND_OFFLOAD
2319 * - (-ENODEV) if *port_id* invalid.
2321 extern int rte_eth_dev_get_vlan_offload(uint8_t port_id);
2324 * Set port based TX VLAN insersion on or off.
2327 * The port identifier of the Ethernet device.
2329 * Port based TX VLAN identifier togeth with user priority.
2331 * Turn on or off the port based TX VLAN insertion.
2334 * - (0) if successful.
2335 * - negative if failed.
2337 extern int rte_eth_dev_set_vlan_pvid(uint8_t port_id, uint16_t pvid, int on);
2341 * Retrieve a burst of input packets from a receive queue of an Ethernet
2342 * device. The retrieved packets are stored in *rte_mbuf* structures whose
2343 * pointers are supplied in the *rx_pkts* array.
2345 * The rte_eth_rx_burst() function loops, parsing the RX ring of the
2346 * receive queue, up to *nb_pkts* packets, and for each completed RX
2347 * descriptor in the ring, it performs the following operations:
2349 * - Initialize the *rte_mbuf* data structure associated with the
2350 * RX descriptor according to the information provided by the NIC into
2351 * that RX descriptor.
2353 * - Store the *rte_mbuf* data structure into the next entry of the
2356 * - Replenish the RX descriptor with a new *rte_mbuf* buffer
2357 * allocated from the memory pool associated with the receive queue at
2358 * initialization time.
2360 * When retrieving an input packet that was scattered by the controller
2361 * into multiple receive descriptors, the rte_eth_rx_burst() function
2362 * appends the associated *rte_mbuf* buffers to the first buffer of the
2365 * The rte_eth_rx_burst() function returns the number of packets
2366 * actually retrieved, which is the number of *rte_mbuf* data structures
2367 * effectively supplied into the *rx_pkts* array.
2368 * A return value equal to *nb_pkts* indicates that the RX queue contained
2369 * at least *rx_pkts* packets, and this is likely to signify that other
2370 * received packets remain in the input queue. Applications implementing
2371 * a "retrieve as much received packets as possible" policy can check this
2372 * specific case and keep invoking the rte_eth_rx_burst() function until
2373 * a value less than *nb_pkts* is returned.
2375 * This receive method has the following advantages:
2377 * - It allows a run-to-completion network stack engine to retrieve and
2378 * to immediately process received packets in a fast burst-oriented
2379 * approach, avoiding the overhead of unnecessary intermediate packet
2380 * queue/dequeue operations.
2382 * - Conversely, it also allows an asynchronous-oriented processing
2383 * method to retrieve bursts of received packets and to immediately
2384 * queue them for further parallel processing by another logical core,
2385 * for instance. However, instead of having received packets being
2386 * individually queued by the driver, this approach allows the invoker
2387 * of the rte_eth_rx_burst() function to queue a burst of retrieved
2388 * packets at a time and therefore dramatically reduce the cost of
2389 * enqueue/dequeue operations per packet.
2391 * - It allows the rte_eth_rx_burst() function of the driver to take
2392 * advantage of burst-oriented hardware features (CPU cache,
2393 * prefetch instructions, and so on) to minimize the number of CPU
2394 * cycles per packet.
2396 * To summarize, the proposed receive API enables many
2397 * burst-oriented optimizations in both synchronous and asynchronous
2398 * packet processing environments with no overhead in both cases.
2400 * The rte_eth_rx_burst() function does not provide any error
2401 * notification to avoid the corresponding overhead. As a hint, the
2402 * upper-level application might check the status of the device link once
2403 * being systematically returned a 0 value for a given number of tries.
2406 * The port identifier of the Ethernet device.
2408 * The index of the receive queue from which to retrieve input packets.
2409 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
2410 * to rte_eth_dev_configure().
2412 * The address of an array of pointers to *rte_mbuf* structures that
2413 * must be large enough to store *nb_pkts* pointers in it.
2415 * The maximum number of packets to retrieve.
2417 * The number of packets actually retrieved, which is the number
2418 * of pointers to *rte_mbuf* structures effectively supplied to the
2421 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2422 extern uint16_t rte_eth_rx_burst(uint8_t port_id, uint16_t queue_id,
2423 struct rte_mbuf **rx_pkts, uint16_t nb_pkts);
2425 static inline uint16_t
2426 rte_eth_rx_burst(uint8_t port_id, uint16_t queue_id,
2427 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
2429 struct rte_eth_dev *dev;
2431 dev = &rte_eth_devices[port_id];
2433 int16_t nb_rx = (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id],
2436 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
2437 struct rte_eth_rxtx_callback *cb = dev->post_rx_burst_cbs[queue_id];
2439 if (unlikely(cb != NULL)) {
2441 nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
2442 nb_pkts, cb->param);
2444 } while (cb != NULL);
2453 * Get the number of used descriptors in a specific queue
2456 * The port identifier of the Ethernet device.
2458 * The queue id on the specific port.
2460 * The number of used descriptors in the specific queue.
2462 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2463 extern uint32_t rte_eth_rx_queue_count(uint8_t port_id, uint16_t queue_id);
2465 static inline uint32_t
2466 rte_eth_rx_queue_count(uint8_t port_id, uint16_t queue_id)
2468 struct rte_eth_dev *dev;
2470 dev = &rte_eth_devices[port_id];
2471 return (*dev->dev_ops->rx_queue_count)(dev, queue_id);
2476 * Check if the DD bit of the specific RX descriptor in the queue has been set
2479 * The port identifier of the Ethernet device.
2481 * The queue id on the specific port.
2483 * The offset of the descriptor ID from tail.
2485 * - (1) if the specific DD bit is set.
2486 * - (0) if the specific DD bit is not set.
2487 * - (-ENODEV) if *port_id* invalid.
2489 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2490 extern int rte_eth_rx_descriptor_done(uint8_t port_id,
2495 rte_eth_rx_descriptor_done(uint8_t port_id, uint16_t queue_id, uint16_t offset)
2497 struct rte_eth_dev *dev;
2499 dev = &rte_eth_devices[port_id];
2500 return (*dev->dev_ops->rx_descriptor_done)( \
2501 dev->data->rx_queues[queue_id], offset);
2506 * Send a burst of output packets on a transmit queue of an Ethernet device.
2508 * The rte_eth_tx_burst() function is invoked to transmit output packets
2509 * on the output queue *queue_id* of the Ethernet device designated by its
2511 * The *nb_pkts* parameter is the number of packets to send which are
2512 * supplied in the *tx_pkts* array of *rte_mbuf* structures.
2513 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets,
2514 * up to the number of transmit descriptors available in the TX ring of the
2516 * For each packet to send, the rte_eth_tx_burst() function performs
2517 * the following operations:
2519 * - Pick up the next available descriptor in the transmit ring.
2521 * - Free the network buffer previously sent with that descriptor, if any.
2523 * - Initialize the transmit descriptor with the information provided
2524 * in the *rte_mbuf data structure.
2526 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers,
2527 * the rte_eth_tx_burst() function uses several transmit descriptors
2530 * The rte_eth_tx_burst() function returns the number of packets it
2531 * actually sent. A return value equal to *nb_pkts* means that all packets
2532 * have been sent, and this is likely to signify that other output packets
2533 * could be immediately transmitted again. Applications that implement a
2534 * "send as many packets to transmit as possible" policy can check this
2535 * specific case and keep invoking the rte_eth_tx_burst() function until
2536 * a value less than *nb_pkts* is returned.
2538 * It is the responsibility of the rte_eth_tx_burst() function to
2539 * transparently free the memory buffers of packets previously sent.
2540 * This feature is driven by the *tx_free_thresh* value supplied to the
2541 * rte_eth_dev_configure() function at device configuration time.
2542 * When the number of free TX descriptors drops below this threshold, the
2543 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers
2544 * of those packets whose transmission was effectively completed.
2547 * The port identifier of the Ethernet device.
2549 * The index of the transmit queue through which output packets must be
2551 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
2552 * to rte_eth_dev_configure().
2554 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
2555 * which contain the output packets.
2557 * The maximum number of packets to transmit.
2559 * The number of output packets actually stored in transmit descriptors of
2560 * the transmit ring. The return value can be less than the value of the
2561 * *tx_pkts* parameter when the transmit ring is full or has been filled up.
2563 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
2564 extern uint16_t rte_eth_tx_burst(uint8_t port_id, uint16_t queue_id,
2565 struct rte_mbuf **tx_pkts, uint16_t nb_pkts);
2567 static inline uint16_t
2568 rte_eth_tx_burst(uint8_t port_id, uint16_t queue_id,
2569 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
2571 struct rte_eth_dev *dev;
2573 dev = &rte_eth_devices[port_id];
2575 #ifdef RTE_ETHDEV_RXTX_CALLBACKS
2576 struct rte_eth_rxtx_callback *cb = dev->pre_tx_burst_cbs[queue_id];
2578 if (unlikely(cb != NULL)) {
2580 nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
2583 } while (cb != NULL);
2587 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts);
2592 * Setup a new signature filter rule on an Ethernet device
2595 * The port identifier of the Ethernet device.
2596 * @param fdir_filter
2597 * The pointer to the fdir filter structure describing the signature filter
2599 * The *rte_fdir_filter* structure includes the values of the different fields
2600 * to match: source and destination IP addresses, vlan id, flexbytes, source
2601 * and destination ports, and so on.
2603 * The index of the RX queue where to store RX packets matching the added
2604 * signature filter defined in fdir_filter.
2606 * - (0) if successful.
2607 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2608 * - (-ENODEV) if *port_id* invalid.
2609 * - (-ENOSYS) if the FDIR mode is not configured in signature mode
2611 * - (-EINVAL) if the fdir_filter information is not correct.
2613 int rte_eth_dev_fdir_add_signature_filter(uint8_t port_id,
2614 struct rte_fdir_filter *fdir_filter,
2618 * Update a signature filter rule on an Ethernet device.
2619 * If the rule doesn't exits, it is created.
2622 * The port identifier of the Ethernet device.
2624 * The pointer to the structure describing the signature filter rule.
2625 * The *rte_fdir_filter* structure includes the values of the different fields
2626 * to match: source and destination IP addresses, vlan id, flexbytes, source
2627 * and destination ports, and so on.
2629 * The index of the RX queue where to store RX packets matching the added
2630 * signature filter defined in fdir_ftr.
2632 * - (0) if successful.
2633 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2634 * - (-ENODEV) if *port_id* invalid.
2635 * - (-ENOSYS) if the flow director mode is not configured in signature mode
2637 * - (-EINVAL) if the fdir_filter information is not correct.
2639 int rte_eth_dev_fdir_update_signature_filter(uint8_t port_id,
2640 struct rte_fdir_filter *fdir_ftr,
2644 * Remove a signature filter rule on an Ethernet device.
2647 * The port identifier of the Ethernet device.
2649 * The pointer to the structure describing the signature filter rule.
2650 * The *rte_fdir_filter* structure includes the values of the different fields
2651 * to match: source and destination IP addresses, vlan id, flexbytes, source
2652 * and destination ports, and so on.
2654 * - (0) if successful.
2655 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2656 * - (-ENODEV) if *port_id* invalid.
2657 * - (-ENOSYS) if the flow director mode is not configured in signature mode
2659 * - (-EINVAL) if the fdir_filter information is not correct.
2661 int rte_eth_dev_fdir_remove_signature_filter(uint8_t port_id,
2662 struct rte_fdir_filter *fdir_ftr);
2665 * Retrieve the flow director information of an Ethernet device.
2668 * The port identifier of the Ethernet device.
2670 * A pointer to a structure of type *rte_eth_dev_fdir* to be filled with
2671 * the flow director information of the Ethernet device.
2673 * - (0) if successful.
2674 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2675 * - (-ENODEV) if *port_id* invalid.
2676 * - (-ENOSYS) if the flow director mode is not configured on *port_id*.
2678 int rte_eth_dev_fdir_get_infos(uint8_t port_id, struct rte_eth_fdir *fdir);
2681 * Add a new perfect filter rule on an Ethernet device.
2684 * The port identifier of the Ethernet device.
2685 * @param fdir_filter
2686 * The pointer to the structure describing the perfect filter rule.
2687 * The *rte_fdir_filter* structure includes the values of the different fields
2688 * to match: source and destination IP addresses, vlan id, flexbytes, source
2689 * and destination ports, and so on.
2690 * IPv6 are not supported.
2692 * The 16-bit value supplied in the field hash.fdir.id of mbuf for RX
2693 * packets matching the perfect filter.
2695 * The index of the RX queue where to store RX packets matching the added
2696 * perfect filter defined in fdir_filter.
2698 * If drop is set to 1, matching RX packets are stored into the RX drop
2699 * queue defined in the rte_fdir_conf.
2701 * - (0) if successful.
2702 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2703 * - (-ENODEV) if *port_id* invalid.
2704 * - (-ENOSYS) if the flow director mode is not configured in perfect mode
2706 * - (-EINVAL) if the fdir_filter information is not correct.
2708 int rte_eth_dev_fdir_add_perfect_filter(uint8_t port_id,
2709 struct rte_fdir_filter *fdir_filter,
2710 uint16_t soft_id, uint8_t rx_queue,
2714 * Update a perfect filter rule on an Ethernet device.
2715 * If the rule doesn't exits, it is created.
2718 * The port identifier of the Ethernet device.
2719 * @param fdir_filter
2720 * The pointer to the structure describing the perfect filter rule.
2721 * The *rte_fdir_filter* structure includes the values of the different fields
2722 * to match: source and destination IP addresses, vlan id, flexbytes, source
2723 * and destination ports, and so on.
2724 * IPv6 are not supported.
2726 * The 16-bit value supplied in the field hash.fdir.id of mbuf for RX
2727 * packets matching the perfect filter.
2729 * The index of the RX queue where to store RX packets matching the added
2730 * perfect filter defined in fdir_filter.
2732 * If drop is set to 1, matching RX packets are stored into the RX drop
2733 * queue defined in the rte_fdir_conf.
2735 * - (0) if successful.
2736 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2737 * - (-ENODEV) if *port_id* invalid.
2738 * - (-ENOSYS) if the flow director mode is not configured in perfect mode
2740 * - (-EINVAL) if the fdir_filter information is not correct.
2742 int rte_eth_dev_fdir_update_perfect_filter(uint8_t port_id,
2743 struct rte_fdir_filter *fdir_filter,
2744 uint16_t soft_id, uint8_t rx_queue,
2748 * Remove a perfect filter rule on an Ethernet device.
2751 * The port identifier of the Ethernet device.
2752 * @param fdir_filter
2753 * The pointer to the structure describing the perfect filter rule.
2754 * The *rte_fdir_filter* structure includes the values of the different fields
2755 * to match: source and destination IP addresses, vlan id, flexbytes, source
2756 * and destination ports, and so on.
2757 * IPv6 are not supported.
2759 * The soft_id value provided when adding/updating the removed filter.
2761 * - (0) if successful.
2762 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2763 * - (-ENODEV) if *port_id* invalid.
2764 * - (-ENOSYS) if the flow director mode is not configured in perfect mode
2766 * - (-EINVAL) if the fdir_filter information is not correct.
2768 int rte_eth_dev_fdir_remove_perfect_filter(uint8_t port_id,
2769 struct rte_fdir_filter *fdir_filter,
2772 * Configure globally the masks for flow director mode for an Ethernet device.
2773 * For example, the device can match packets with only the first 24 bits of
2774 * the IPv4 source address.
2776 * The following fields can be masked: IPv4 addresses and L4 port numbers.
2777 * The following fields can be either enabled or disabled completely for the
2778 * matching functionality: VLAN ID tag; VLAN Priority + CFI bit; Flexible 2-byte
2780 * IPv6 masks are not supported.
2782 * All filters must comply with the masks previously configured.
2783 * For example, with a mask equal to 255.255.255.0 for the source IPv4 address,
2784 * all IPv4 filters must be created with a source IPv4 address that fits the
2787 * This function flushes all filters that have been previously added in
2791 * The port identifier of the Ethernet device.
2793 * The pointer to the fdir mask structure describing relevant headers fields
2794 * and relevant bits to use when matching packets addresses and ports.
2795 * IPv6 masks are not supported.
2797 * - (0) if successful.
2798 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2799 * - (-ENODEV) if *port_id* invalid.
2800 * - (-ENOSYS) if the flow director mode is not configured in perfect
2801 * mode on *port_id*.
2802 * - (-EINVAL) if the fdir_filter information is not correct
2804 int rte_eth_dev_fdir_set_masks(uint8_t port_id,
2805 struct rte_fdir_masks *fdir_mask);
2808 * The eth device event type for interrupt, and maybe others in the future.
2810 enum rte_eth_event_type {
2811 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */
2812 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */
2813 RTE_ETH_EVENT_MAX /**< max value of this enum */
2816 typedef void (*rte_eth_dev_cb_fn)(uint8_t port_id, \
2817 enum rte_eth_event_type event, void *cb_arg);
2818 /**< user application callback to be registered for interrupts */
2823 * Register a callback function for specific port id.
2830 * User supplied callback function to be called.
2832 * Pointer to the parameters for the registered callback.
2835 * - On success, zero.
2836 * - On failure, a negative value.
2838 int rte_eth_dev_callback_register(uint8_t port_id,
2839 enum rte_eth_event_type event,
2840 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
2843 * Unregister a callback function for specific port id.
2850 * User supplied callback function to be called.
2852 * Pointer to the parameters for the registered callback. -1 means to
2853 * remove all for the same callback address and same event.
2856 * - On success, zero.
2857 * - On failure, a negative value.
2859 int rte_eth_dev_callback_unregister(uint8_t port_id,
2860 enum rte_eth_event_type event,
2861 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
2864 * @internal Executes all the user application registered callbacks for
2865 * the specific device. It is for DPDK internal user only. User
2866 * application should not call it directly.
2869 * Pointer to struct rte_eth_dev.
2871 * Eth device interrupt event type.
2876 void _rte_eth_dev_callback_process(struct rte_eth_dev *dev,
2877 enum rte_eth_event_type event);
2880 * Turn on the LED on the Ethernet device.
2881 * This function turns on the LED on the Ethernet device.
2884 * The port identifier of the Ethernet device.
2886 * - (0) if successful.
2887 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
2889 * - (-ENODEV) if *port_id* invalid.
2891 int rte_eth_led_on(uint8_t port_id);
2894 * Turn off the LED on the Ethernet device.
2895 * This function turns off the LED on the Ethernet device.
2898 * The port identifier of the Ethernet device.
2900 * - (0) if successful.
2901 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
2903 * - (-ENODEV) if *port_id* invalid.
2905 int rte_eth_led_off(uint8_t port_id);
2908 * Get current status of the Ethernet link flow control for Ethernet device
2911 * The port identifier of the Ethernet device.
2913 * The pointer to the structure where to store the flow control parameters.
2915 * - (0) if successful.
2916 * - (-ENOTSUP) if hardware doesn't support flow control.
2917 * - (-ENODEV) if *port_id* invalid.
2919 int rte_eth_dev_flow_ctrl_get(uint8_t port_id,
2920 struct rte_eth_fc_conf *fc_conf);
2923 * Configure the Ethernet link flow control for Ethernet device
2926 * The port identifier of the Ethernet device.
2928 * The pointer to the structure of the flow control parameters.
2930 * - (0) if successful.
2931 * - (-ENOTSUP) if hardware doesn't support flow control mode.
2932 * - (-ENODEV) if *port_id* invalid.
2933 * - (-EINVAL) if bad parameter
2934 * - (-EIO) if flow control setup failure
2936 int rte_eth_dev_flow_ctrl_set(uint8_t port_id,
2937 struct rte_eth_fc_conf *fc_conf);
2940 * Configure the Ethernet priority flow control under DCB environment
2941 * for Ethernet device.
2944 * The port identifier of the Ethernet device.
2946 * The pointer to the structure of the priority flow control parameters.
2948 * - (0) if successful.
2949 * - (-ENOTSUP) if hardware doesn't support priority flow control mode.
2950 * - (-ENODEV) if *port_id* invalid.
2951 * - (-EINVAL) if bad parameter
2952 * - (-EIO) if flow control setup failure
2954 int rte_eth_dev_priority_flow_ctrl_set(uint8_t port_id,
2955 struct rte_eth_pfc_conf *pfc_conf);
2958 * Add a MAC address to an internal array of addresses used to enable whitelist
2959 * filtering to accept packets only if the destination MAC address matches.
2962 * The port identifier of the Ethernet device.
2964 * The MAC address to add.
2966 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is
2967 * not enabled, this should be set to 0.
2969 * - (0) if successfully added or *mac_addr" was already added.
2970 * - (-ENOTSUP) if hardware doesn't support this feature.
2971 * - (-ENODEV) if *port* is invalid.
2972 * - (-ENOSPC) if no more MAC addresses can be added.
2973 * - (-EINVAL) if MAC address is invalid.
2975 int rte_eth_dev_mac_addr_add(uint8_t port, struct ether_addr *mac_addr,
2979 * Remove a MAC address from the internal array of addresses.
2982 * The port identifier of the Ethernet device.
2984 * MAC address to remove.
2986 * - (0) if successful, or *mac_addr* didn't exist.
2987 * - (-ENOTSUP) if hardware doesn't support.
2988 * - (-ENODEV) if *port* invalid.
2989 * - (-EADDRINUSE) if attempting to remove the default MAC address
2991 int rte_eth_dev_mac_addr_remove(uint8_t port, struct ether_addr *mac_addr);
2994 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
2997 * The port identifier of the Ethernet device.
3001 * Redirection table size. The table size can be queried by
3002 * rte_eth_dev_info_get().
3004 * - (0) if successful.
3005 * - (-ENOTSUP) if hardware doesn't support.
3006 * - (-EINVAL) if bad parameter.
3008 int rte_eth_dev_rss_reta_update(uint8_t port,
3009 struct rte_eth_rss_reta_entry64 *reta_conf,
3010 uint16_t reta_size);
3013 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
3016 * The port identifier of the Ethernet device.
3020 * Redirection table size. The table size can be queried by
3021 * rte_eth_dev_info_get().
3023 * - (0) if successful.
3024 * - (-ENOTSUP) if hardware doesn't support.
3025 * - (-EINVAL) if bad parameter.
3027 int rte_eth_dev_rss_reta_query(uint8_t port,
3028 struct rte_eth_rss_reta_entry64 *reta_conf,
3029 uint16_t reta_size);
3032 * Updates unicast hash table for receiving packet with the given destination
3033 * MAC address, and the packet is routed to all VFs for which the RX mode is
3034 * accept packets that match the unicast hash table.
3037 * The port identifier of the Ethernet device.
3039 * Unicast MAC address.
3041 * 1 - Set an unicast hash bit for receiving packets with the MAC address.
3042 * 0 - Clear an unicast hash bit.
3044 * - (0) if successful.
3045 * - (-ENOTSUP) if hardware doesn't support.
3046 * - (-ENODEV) if *port_id* invalid.
3047 * - (-EINVAL) if bad parameter.
3049 int rte_eth_dev_uc_hash_table_set(uint8_t port,struct ether_addr *addr,
3053 * Updates all unicast hash bitmaps for receiving packet with any Unicast
3054 * Ethernet MAC addresses,the packet is routed to all VFs for which the RX
3055 * mode is accept packets that match the unicast hash table.
3058 * The port identifier of the Ethernet device.
3060 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet
3062 * 0 - Clear all unicast hash bitmaps
3064 * - (0) if successful.
3065 * - (-ENOTSUP) if hardware doesn't support.
3066 * - (-ENODEV) if *port_id* invalid.
3067 * - (-EINVAL) if bad parameter.
3069 int rte_eth_dev_uc_all_hash_table_set(uint8_t port,uint8_t on);
3072 * Set RX L2 Filtering mode of a VF of an Ethernet device.
3075 * The port identifier of the Ethernet device.
3079 * The RX mode mask, which is one or more of accepting Untagged Packets,
3080 * packets that match the PFUTA table, Broadcast and Multicast Promiscuous.
3081 * ETH_VMDQ_ACCEPT_UNTAG,ETH_VMDQ_ACCEPT_HASH_UC,
3082 * ETH_VMDQ_ACCEPT_BROADCAST and ETH_VMDQ_ACCEPT_MULTICAST will be used
3085 * 1 - Enable a VF RX mode.
3086 * 0 - Disable a VF RX mode.
3088 * - (0) if successful.
3089 * - (-ENOTSUP) if hardware doesn't support.
3090 * - (-ENOTSUP) if hardware doesn't support.
3091 * - (-EINVAL) if bad parameter.
3093 int rte_eth_dev_set_vf_rxmode(uint8_t port, uint16_t vf, uint16_t rx_mode,
3097 * Enable or disable a VF traffic transmit of the Ethernet device.
3100 * The port identifier of the Ethernet device.
3104 * 1 - Enable a VF traffic transmit.
3105 * 0 - Disable a VF traffic transmit.
3107 * - (0) if successful.
3108 * - (-ENODEV) if *port_id* invalid.
3109 * - (-ENOTSUP) if hardware doesn't support.
3110 * - (-EINVAL) if bad parameter.
3113 rte_eth_dev_set_vf_tx(uint8_t port,uint16_t vf, uint8_t on);
3116 * Enable or disable a VF traffic receive of an Ethernet device.
3119 * The port identifier of the Ethernet device.
3123 * 1 - Enable a VF traffic receive.
3124 * 0 - Disable a VF traffic receive.
3126 * - (0) if successful.
3127 * - (-ENOTSUP) if hardware doesn't support.
3128 * - (-ENODEV) if *port_id* invalid.
3129 * - (-EINVAL) if bad parameter.
3132 rte_eth_dev_set_vf_rx(uint8_t port,uint16_t vf, uint8_t on);
3135 * Enable/Disable hardware VF VLAN filtering by an Ethernet device of
3136 * received VLAN packets tagged with a given VLAN Tag Identifier.
3139 * The port identifier of the Ethernet device.
3141 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
3143 * Bitmap listing which VFs participate in the VLAN filtering.
3145 * 1 - Enable VFs VLAN filtering.
3146 * 0 - Disable VFs VLAN filtering.
3148 * - (0) if successful.
3149 * - (-ENOTSUP) if hardware doesn't support.
3150 * - (-ENODEV) if *port_id* invalid.
3151 * - (-EINVAL) if bad parameter.
3154 rte_eth_dev_set_vf_vlan_filter(uint8_t port, uint16_t vlan_id,
3159 * Set a traffic mirroring rule on an Ethernet device
3162 * The port identifier of the Ethernet device.
3163 * @param mirror_conf
3164 * The pointer to the traffic mirroring structure describing the mirroring rule.
3165 * The *rte_eth_vm_mirror_conf* structure includes the type of mirroring rule,
3166 * destination pool and the value of rule if enable vlan or pool mirroring.
3169 * The index of traffic mirroring rule, we support four separated rules.
3171 * 1 - Enable a mirroring rule.
3172 * 0 - Disable a mirroring rule.
3174 * - (0) if successful.
3175 * - (-ENOTSUP) if hardware doesn't support this feature.
3176 * - (-ENODEV) if *port_id* invalid.
3177 * - (-EINVAL) if the mr_conf information is not correct.
3179 int rte_eth_mirror_rule_set(uint8_t port_id,
3180 struct rte_eth_mirror_conf *mirror_conf,
3185 * Reset a traffic mirroring rule on an Ethernet device.
3188 * The port identifier of the Ethernet device.
3190 * The index of traffic mirroring rule, we support four separated rules.
3192 * - (0) if successful.
3193 * - (-ENOTSUP) if hardware doesn't support this feature.
3194 * - (-ENODEV) if *port_id* invalid.
3195 * - (-EINVAL) if bad parameter.
3197 int rte_eth_mirror_rule_reset(uint8_t port_id,
3201 * Set the rate limitation for a queue on an Ethernet device.
3204 * The port identifier of the Ethernet device.
3208 * The tx rate allocated from the total link speed for this queue.
3210 * - (0) if successful.
3211 * - (-ENOTSUP) if hardware doesn't support this feature.
3212 * - (-ENODEV) if *port_id* invalid.
3213 * - (-EINVAL) if bad parameter.
3215 int rte_eth_set_queue_rate_limit(uint8_t port_id, uint16_t queue_idx,
3219 * Set the rate limitation for a vf on an Ethernet device.
3222 * The port identifier of the Ethernet device.
3226 * The tx rate allocated from the total link speed for this VF id.
3228 * The queue mask which need to set the rate.
3230 * - (0) if successful.
3231 * - (-ENOTSUP) if hardware doesn't support this feature.
3232 * - (-ENODEV) if *port_id* invalid.
3233 * - (-EINVAL) if bad parameter.
3235 int rte_eth_set_vf_rate_limit(uint8_t port_id, uint16_t vf,
3236 uint16_t tx_rate, uint64_t q_msk);
3239 * Initialize bypass logic. This function needs to be called before
3240 * executing any other bypass API.
3243 * The port identifier of the Ethernet device.
3245 * - (0) if successful.
3246 * - (-ENOTSUP) if hardware doesn't support.
3247 * - (-EINVAL) if bad parameter.
3249 int rte_eth_dev_bypass_init(uint8_t port);
3252 * Return bypass state.
3255 * The port identifier of the Ethernet device.
3257 * The return bypass state.
3260 * - (3) Isolate mode
3262 * - (0) if successful.
3263 * - (-ENOTSUP) if hardware doesn't support.
3264 * - (-EINVAL) if bad parameter.
3266 int rte_eth_dev_bypass_state_show(uint8_t port, uint32_t *state);
3272 * The port identifier of the Ethernet device.
3274 * The current bypass state.
3277 * - (3) Isolate mode
3279 * - (0) if successful.
3280 * - (-ENOTSUP) if hardware doesn't support.
3281 * - (-EINVAL) if bad parameter.
3283 int rte_eth_dev_bypass_state_set(uint8_t port, uint32_t *new_state);
3286 * Return bypass state when given event occurs.
3289 * The port identifier of the Ethernet device.
3292 * - (1) Main power on (power button is pushed)
3293 * - (2) Auxiliary power on (power supply is being plugged)
3294 * - (3) Main power off (system shutdown and power supply is left plugged in)
3295 * - (4) Auxiliary power off (power supply is being unplugged)
3296 * - (5) Display or set the watchdog timer
3298 * The bypass state when given event occurred.
3301 * - (3) Isolate mode
3303 * - (0) if successful.
3304 * - (-ENOTSUP) if hardware doesn't support.
3305 * - (-EINVAL) if bad parameter.
3307 int rte_eth_dev_bypass_event_show(uint8_t port, uint32_t event, uint32_t *state);
3310 * Set bypass state when given event occurs.
3313 * The port identifier of the Ethernet device.
3316 * - (1) Main power on (power button is pushed)
3317 * - (2) Auxiliary power on (power supply is being plugged)
3318 * - (3) Main power off (system shutdown and power supply is left plugged in)
3319 * - (4) Auxiliary power off (power supply is being unplugged)
3320 * - (5) Display or set the watchdog timer
3322 * The assigned state when given event occurs.
3325 * - (3) Isolate mode
3327 * - (0) if successful.
3328 * - (-ENOTSUP) if hardware doesn't support.
3329 * - (-EINVAL) if bad parameter.
3331 int rte_eth_dev_bypass_event_store(uint8_t port, uint32_t event, uint32_t state);
3334 * Set bypass watchdog timeout count.
3337 * The port identifier of the Ethernet device.
3339 * The timeout to be set.
3340 * - (0) 0 seconds (timer is off)
3349 * - (0) if successful.
3350 * - (-ENOTSUP) if hardware doesn't support.
3351 * - (-EINVAL) if bad parameter.
3353 int rte_eth_dev_wd_timeout_store(uint8_t port, uint32_t timeout);
3356 * Get bypass firmware version.
3359 * The port identifier of the Ethernet device.
3361 * The firmware version
3363 * - (0) if successful.
3364 * - (-ENOTSUP) if hardware doesn't support.
3365 * - (-EINVAL) if bad parameter.
3367 int rte_eth_dev_bypass_ver_show(uint8_t port, uint32_t *ver);
3370 * Return bypass watchdog timeout in seconds
3373 * The port identifier of the Ethernet device.
3375 * The return watchdog timeout. "0" represents timer expired
3376 * - (0) 0 seconds (timer is off)
3385 * - (0) if successful.
3386 * - (-ENOTSUP) if hardware doesn't support.
3387 * - (-EINVAL) if bad parameter.
3389 int rte_eth_dev_bypass_wd_timeout_show(uint8_t port, uint32_t *wd_timeout);
3392 * Reset bypass watchdog timer
3395 * The port identifier of the Ethernet device.
3397 * - (0) if successful.
3398 * - (-ENOTSUP) if hardware doesn't support.
3399 * - (-EINVAL) if bad parameter.
3401 int rte_eth_dev_bypass_wd_reset(uint8_t port);
3404 * Configuration of Receive Side Scaling hash computation of Ethernet device.
3407 * The port identifier of the Ethernet device.
3409 * The new configuration to use for RSS hash computation on the port.
3411 * - (0) if successful.
3412 * - (-ENODEV) if port identifier is invalid.
3413 * - (-ENOTSUP) if hardware doesn't support.
3414 * - (-EINVAL) if bad parameter.
3416 int rte_eth_dev_rss_hash_update(uint8_t port_id,
3417 struct rte_eth_rss_conf *rss_conf);
3420 * Retrieve current configuration of Receive Side Scaling hash computation
3421 * of Ethernet device.
3424 * The port identifier of the Ethernet device.
3426 * Where to store the current RSS hash configuration of the Ethernet device.
3428 * - (0) if successful.
3429 * - (-ENODEV) if port identifier is invalid.
3430 * - (-ENOTSUP) if hardware doesn't support RSS.
3433 rte_eth_dev_rss_hash_conf_get(uint8_t port_id,
3434 struct rte_eth_rss_conf *rss_conf);
3437 * Add UDP tunneling port of an Ethernet device for filtering a specific
3438 * tunneling packet by UDP port number.
3441 * The port identifier of the Ethernet device.
3443 * UDP tunneling configuration.
3446 * - (0) if successful.
3447 * - (-ENODEV) if port identifier is invalid.
3448 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3451 rte_eth_dev_udp_tunnel_add(uint8_t port_id,
3452 struct rte_eth_udp_tunnel *tunnel_udp);
3455 * Detete UDP tunneling port configuration of Ethernet device
3458 * The port identifier of the Ethernet device.
3460 * UDP tunneling configuration.
3463 * - (0) if successful.
3464 * - (-ENODEV) if port identifier is invalid.
3465 * - (-ENOTSUP) if hardware doesn't support tunnel type.
3468 rte_eth_dev_udp_tunnel_delete(uint8_t port_id,
3469 struct rte_eth_udp_tunnel *tunnel_udp);
3472 * Check whether the filter type is supported on an Ethernet device.
3473 * All the supported filter types are defined in 'rte_eth_ctrl.h'.
3476 * The port identifier of the Ethernet device.
3477 * @param filter_type
3480 * - (0) if successful.
3481 * - (-ENOTSUP) if hardware doesn't support this filter type.
3482 * - (-ENODEV) if *port_id* invalid.
3484 int rte_eth_dev_filter_supported(uint8_t port_id, enum rte_filter_type filter_type);
3487 * Take operations to assigned filter type on an Ethernet device.
3488 * All the supported operations and filter types are defined in 'rte_eth_ctrl.h'.
3491 * The port identifier of the Ethernet device.
3492 * @param filter_type
3495 * Type of operation.
3497 * A pointer to arguments defined specifically for the operation.
3499 * - (0) if successful.
3500 * - (-ENOTSUP) if hardware doesn't support.
3501 * - (-ENODEV) if *port_id* invalid.
3502 * - others depends on the specific operations implementation.
3504 int rte_eth_dev_filter_ctrl(uint8_t port_id, enum rte_filter_type filter_type,
3505 enum rte_filter_op filter_op, void *arg);
3508 * Add a callback to be called on packet RX on a given port and queue.
3510 * This API configures a function to be called for each burst of
3511 * packets received on a given NIC port queue. The return value is a pointer
3512 * that can be used to later remove the callback using
3513 * rte_eth_remove_rx_callback().
3516 * The port identifier of the Ethernet device.
3518 * The queue on the Ethernet device on which the callback is to be added.
3520 * The callback function
3522 * A generic pointer parameter which will be passed to each invocation of the
3523 * callback function on this port and queue.
3527 * On success, a pointer value which can later be used to remove the callback.
3529 void *rte_eth_add_rx_callback(uint8_t port_id, uint16_t queue_id,
3530 rte_rx_callback_fn fn, void *user_param);
3533 * Add a callback to be called on packet TX on a given port and queue.
3535 * This API configures a function to be called for each burst of
3536 * packets sent on a given NIC port queue. The return value is a pointer
3537 * that can be used to later remove the callback using
3538 * rte_eth_remove_tx_callback().
3541 * The port identifier of the Ethernet device.
3543 * The queue on the Ethernet device on which the callback is to be added.
3545 * The callback function
3547 * A generic pointer parameter which will be passed to each invocation of the
3548 * callback function on this port and queue.
3552 * On success, a pointer value which can later be used to remove the callback.
3554 void *rte_eth_add_tx_callback(uint8_t port_id, uint16_t queue_id,
3555 rte_tx_callback_fn fn, void *user_param);
3558 * Remove an RX packet callback from a given port and queue.
3560 * This function is used to removed callbacks that were added to a NIC port
3561 * queue using rte_eth_add_rx_callback().
3563 * Note: the callback is removed from the callback list but it isn't freed
3564 * since the it may still be in use. The memory for the callback can be
3565 * subsequently freed back by the application by calling rte_free():
3567 * - Immediately - if the port is stopped, or the user knows that no
3568 * callbacks are in flight e.g. if called from the thread doing RX/TX
3571 * - After a short delay - where the delay is sufficient to allow any
3572 * in-flight callbacks to complete.
3575 * The port identifier of the Ethernet device.
3577 * The queue on the Ethernet device from which the callback is to be removed.
3579 * User supplied callback created via rte_eth_add_rx_callback().
3582 * - 0: Success. Callback was removed.
3583 * - -ENOTSUP: Callback support is not available.
3584 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
3585 * is NULL or not found for the port/queue.
3587 int rte_eth_remove_rx_callback(uint8_t port_id, uint16_t queue_id,
3588 struct rte_eth_rxtx_callback *user_cb);
3591 * Remove a TX packet callback from a given port and queue.
3593 * This function is used to removed callbacks that were added to a NIC port
3594 * queue using rte_eth_add_tx_callback().
3596 * Note: the callback is removed from the callback list but it isn't freed
3597 * since the it may still be in use. The memory for the callback can be
3598 * subsequently freed back by the application by calling rte_free():
3600 * - Immediately - if the port is stopped, or the user knows that no
3601 * callbacks are in flight e.g. if called from the thread doing RX/TX
3604 * - After a short delay - where the delay is sufficient to allow any
3605 * in-flight callbacks to complete.
3608 * The port identifier of the Ethernet device.
3610 * The queue on the Ethernet device from which the callback is to be removed.
3612 * User supplied callback created via rte_eth_add_tx_callback().
3615 * - 0: Success. Callback was removed.
3616 * - -ENOTSUP: Callback support is not available.
3617 * - -EINVAL: The port_id or the queue_id is out of range, or the callback
3618 * is NULL or not found for the port/queue.
3620 int rte_eth_remove_tx_callback(uint8_t port_id, uint16_t queue_id,
3621 struct rte_eth_rxtx_callback *user_cb);
3628 * Set the list of multicast addresses to filter on an Ethernet device.
3631 * The port identifier of the Ethernet device.
3632 * @param mc_addr_set
3633 * The array of multicast addresses to set. Equal to NULL when the function
3634 * is invoked to flush the set of filtered addresses.
3636 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0
3637 * when the function is invoked to flush the set of filtered addresses.
3639 * - (0) if successful.
3640 * - (-ENODEV) if *port_id* invalid.
3641 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering.
3642 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources.
3644 int rte_eth_dev_set_mc_addr_list(uint8_t port_id,
3645 struct ether_addr *mc_addr_set,
3646 uint32_t nb_mc_addr);
3648 #endif /* _RTE_ETHDEV_H_ */