1 /* SPDX-License-Identifier: BSD-3-Clause 2 * Copyright(c) 2010-2017 Intel Corporation 3 */ 4 5 #ifndef _RTE_ETHDEV_H_ 6 #define _RTE_ETHDEV_H_ 7 8 /** 9 * @file 10 * 11 * RTE Ethernet Device API 12 * 13 * The Ethernet Device API is composed of two parts: 14 * 15 * - The application-oriented Ethernet API that includes functions to setup 16 * an Ethernet device (configure it, setup its RX and TX queues and start it), 17 * to get its MAC address, the speed and the status of its physical link, 18 * to receive and to transmit packets, and so on. 19 * 20 * - The driver-oriented Ethernet API that exports functions allowing 21 * an Ethernet Poll Mode Driver (PMD) to allocate an Ethernet device instance, 22 * create memzone for HW rings and process registered callbacks, and so on. 23 * PMDs should include ethdev_driver.h instead of this header. 24 * 25 * By default, all the functions of the Ethernet Device API exported by a PMD 26 * are lock-free functions which assume to not be invoked in parallel on 27 * different logical cores to work on the same target object. For instance, 28 * the receive function of a PMD cannot be invoked in parallel on two logical 29 * cores to poll the same RX queue [of the same port]. Of course, this function 30 * can be invoked in parallel by different logical cores on different RX queues. 31 * It is the responsibility of the upper level application to enforce this rule. 32 * 33 * If needed, parallel accesses by multiple logical cores to shared queues 34 * shall be explicitly protected by dedicated inline lock-aware functions 35 * built on top of their corresponding lock-free functions of the PMD API. 36 * 37 * In all functions of the Ethernet API, the Ethernet device is 38 * designated by an integer >= 0 named the device port identifier. 39 * 40 * At the Ethernet driver level, Ethernet devices are represented by a generic 41 * data structure of type *rte_eth_dev*. 42 * 43 * Ethernet devices are dynamically registered during the PCI probing phase 44 * performed at EAL initialization time. 45 * When an Ethernet device is being probed, an *rte_eth_dev* structure and 46 * a new port identifier are allocated for that device. Then, the eth_dev_init() 47 * function supplied by the Ethernet driver matching the probed PCI 48 * device is invoked to properly initialize the device. 49 * 50 * The role of the device init function consists of resetting the hardware, 51 * checking access to Non-volatile Memory (NVM), reading the MAC address 52 * from NVM etc. 53 * 54 * If the device init operation is successful, the correspondence between 55 * the port identifier assigned to the new device and its associated 56 * *rte_eth_dev* structure is effectively registered. 57 * Otherwise, both the *rte_eth_dev* structure and the port identifier are 58 * freed. 59 * 60 * The functions exported by the application Ethernet API to setup a device 61 * designated by its port identifier must be invoked in the following order: 62 * - rte_eth_dev_configure() 63 * - rte_eth_tx_queue_setup() 64 * - rte_eth_rx_queue_setup() 65 * - rte_eth_dev_start() 66 * 67 * Then, the network application can invoke, in any order, the functions 68 * exported by the Ethernet API to get the MAC address of a given device, to 69 * get the speed and the status of a device physical link, to receive/transmit 70 * [burst of] packets, and so on. 71 * 72 * If the application wants to change the configuration (i.e. call 73 * rte_eth_dev_configure(), rte_eth_tx_queue_setup(), or 74 * rte_eth_rx_queue_setup()), it must call rte_eth_dev_stop() first to stop the 75 * device and then do the reconfiguration before calling rte_eth_dev_start() 76 * again. The transmit and receive functions should not be invoked when the 77 * device is stopped. 78 * 79 * Please note that some configuration is not stored between calls to 80 * rte_eth_dev_stop()/rte_eth_dev_start(). The following configuration will 81 * be retained: 82 * 83 * - MTU 84 * - flow control settings 85 * - receive mode configuration (promiscuous mode, all-multicast mode, 86 * hardware checksum mode, RSS/VMDQ settings etc.) 87 * - VLAN filtering configuration 88 * - default MAC address 89 * - MAC addresses supplied to MAC address array 90 * - flow director filtering mode (but not filtering rules) 91 * - NIC queue statistics mappings 92 * 93 * Any other configuration will not be stored and will need to be re-entered 94 * before a call to rte_eth_dev_start(). 95 * 96 * Finally, a network application can close an Ethernet device by invoking the 97 * rte_eth_dev_close() function. 98 * 99 * Each function of the application Ethernet API invokes a specific function 100 * of the PMD that controls the target device designated by its port 101 * identifier. 102 * For this purpose, all device-specific functions of an Ethernet driver are 103 * supplied through a set of pointers contained in a generic structure of type 104 * *eth_dev_ops*. 105 * The address of the *eth_dev_ops* structure is stored in the *rte_eth_dev* 106 * structure by the device init function of the Ethernet driver, which is 107 * invoked during the PCI probing phase, as explained earlier. 108 * 109 * In other words, each function of the Ethernet API simply retrieves the 110 * *rte_eth_dev* structure associated with the device port identifier and 111 * performs an indirect invocation of the corresponding driver function 112 * supplied in the *eth_dev_ops* structure of the *rte_eth_dev* structure. 113 * 114 * For performance reasons, the address of the burst-oriented RX and TX 115 * functions of the Ethernet driver are not contained in the *eth_dev_ops* 116 * structure. Instead, they are directly stored at the beginning of the 117 * *rte_eth_dev* structure to avoid an extra indirect memory access during 118 * their invocation. 119 * 120 * RTE ethernet device drivers do not use interrupts for transmitting or 121 * receiving. Instead, Ethernet drivers export Poll-Mode receive and transmit 122 * functions to applications. 123 * Both receive and transmit functions are packet-burst oriented to minimize 124 * their cost per packet through the following optimizations: 125 * 126 * - Sharing among multiple packets the incompressible cost of the 127 * invocation of receive/transmit functions. 128 * 129 * - Enabling receive/transmit functions to take advantage of burst-oriented 130 * hardware features (L1 cache, prefetch instructions, NIC head/tail 131 * registers) to minimize the number of CPU cycles per packet, for instance, 132 * by avoiding useless read memory accesses to ring descriptors, or by 133 * systematically using arrays of pointers that exactly fit L1 cache line 134 * boundaries and sizes. 135 * 136 * The burst-oriented receive function does not provide any error notification, 137 * to avoid the corresponding overhead. As a hint, the upper-level application 138 * might check the status of the device link once being systematically returned 139 * a 0 value by the receive function of the driver for a given number of tries. 140 */ 141 142 #ifdef __cplusplus 143 extern "C" { 144 #endif 145 146 #include <stdint.h> 147 148 /* Use this macro to check if LRO API is supported */ 149 #define RTE_ETHDEV_HAS_LRO_SUPPORT 150 151 /* Alias RTE_LIBRTE_ETHDEV_DEBUG for backward compatibility. */ 152 #ifdef RTE_LIBRTE_ETHDEV_DEBUG 153 #define RTE_ETHDEV_DEBUG_RX 154 #define RTE_ETHDEV_DEBUG_TX 155 #endif 156 157 #include <rte_compat.h> 158 #include <rte_log.h> 159 #include <rte_interrupts.h> 160 #include <rte_dev.h> 161 #include <rte_devargs.h> 162 #include <rte_errno.h> 163 #include <rte_common.h> 164 #include <rte_config.h> 165 #include <rte_ether.h> 166 #include <rte_power_intrinsics.h> 167 168 #include "rte_ethdev_trace_fp.h" 169 #include "rte_dev_info.h" 170 171 extern int rte_eth_dev_logtype; 172 173 #define RTE_ETHDEV_LOG(level, ...) \ 174 rte_log(RTE_LOG_ ## level, rte_eth_dev_logtype, "" __VA_ARGS__) 175 176 struct rte_mbuf; 177 178 /** 179 * Initializes a device iterator. 180 * 181 * This iterator allows accessing a list of devices matching some devargs. 182 * 183 * @param iter 184 * Device iterator handle initialized by the function. 185 * The fields bus_str and cls_str might be dynamically allocated, 186 * and could be freed by calling rte_eth_iterator_cleanup(). 187 * 188 * @param devargs 189 * Device description string. 190 * 191 * @return 192 * 0 on successful initialization, negative otherwise. 193 */ 194 int rte_eth_iterator_init(struct rte_dev_iterator *iter, const char *devargs); 195 196 /** 197 * Iterates on devices with devargs filter. 198 * The ownership is not checked. 199 * 200 * The next port id is returned, and the iterator is updated. 201 * 202 * @param iter 203 * Device iterator handle initialized by rte_eth_iterator_init(). 204 * Some fields bus_str and cls_str might be freed when no more port is found, 205 * by calling rte_eth_iterator_cleanup(). 206 * 207 * @return 208 * A port id if found, RTE_MAX_ETHPORTS otherwise. 209 */ 210 uint16_t rte_eth_iterator_next(struct rte_dev_iterator *iter); 211 212 /** 213 * Free some allocated fields of the iterator. 214 * 215 * This function is automatically called by rte_eth_iterator_next() 216 * on the last iteration (i.e. when no more matching port is found). 217 * 218 * It is safe to call this function twice; it will do nothing more. 219 * 220 * @param iter 221 * Device iterator handle initialized by rte_eth_iterator_init(). 222 * The fields bus_str and cls_str are freed if needed. 223 */ 224 void rte_eth_iterator_cleanup(struct rte_dev_iterator *iter); 225 226 /** 227 * Macro to iterate over all ethdev ports matching some devargs. 228 * 229 * If a break is done before the end of the loop, 230 * the function rte_eth_iterator_cleanup() must be called. 231 * 232 * @param id 233 * Iterated port id of type uint16_t. 234 * @param devargs 235 * Device parameters input as string of type char*. 236 * @param iter 237 * Iterator handle of type struct rte_dev_iterator, used internally. 238 */ 239 #define RTE_ETH_FOREACH_MATCHING_DEV(id, devargs, iter) \ 240 for (rte_eth_iterator_init(iter, devargs), \ 241 id = rte_eth_iterator_next(iter); \ 242 id != RTE_MAX_ETHPORTS; \ 243 id = rte_eth_iterator_next(iter)) 244 245 /** 246 * A structure used to retrieve statistics for an Ethernet port. 247 * Not all statistics fields in struct rte_eth_stats are supported 248 * by any type of network interface card (NIC). If any statistics 249 * field is not supported, its value is 0. 250 * All byte-related statistics do not include Ethernet FCS regardless 251 * of whether these bytes have been delivered to the application 252 * (see DEV_RX_OFFLOAD_KEEP_CRC). 253 */ 254 struct rte_eth_stats { 255 uint64_t ipackets; /**< Total number of successfully received packets. */ 256 uint64_t opackets; /**< Total number of successfully transmitted packets.*/ 257 uint64_t ibytes; /**< Total number of successfully received bytes. */ 258 uint64_t obytes; /**< Total number of successfully transmitted bytes. */ 259 uint64_t imissed; 260 /**< Total of RX packets dropped by the HW, 261 * because there are no available buffer (i.e. RX queues are full). 262 */ 263 uint64_t ierrors; /**< Total number of erroneous received packets. */ 264 uint64_t oerrors; /**< Total number of failed transmitted packets. */ 265 uint64_t rx_nombuf; /**< Total number of RX mbuf allocation failures. */ 266 /* Queue stats are limited to max 256 queues */ 267 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 268 /**< Total number of queue RX packets. */ 269 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 270 /**< Total number of queue TX packets. */ 271 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 272 /**< Total number of successfully received queue bytes. */ 273 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 274 /**< Total number of successfully transmitted queue bytes. */ 275 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 276 /**< Total number of queue packets received that are dropped. */ 277 }; 278 279 /** 280 * Device supported speeds bitmap flags 281 */ 282 #define ETH_LINK_SPEED_AUTONEG (0 << 0) /**< Autonegotiate (all speeds) */ 283 #define ETH_LINK_SPEED_FIXED (1 << 0) /**< Disable autoneg (fixed speed) */ 284 #define ETH_LINK_SPEED_10M_HD (1 << 1) /**< 10 Mbps half-duplex */ 285 #define ETH_LINK_SPEED_10M (1 << 2) /**< 10 Mbps full-duplex */ 286 #define ETH_LINK_SPEED_100M_HD (1 << 3) /**< 100 Mbps half-duplex */ 287 #define ETH_LINK_SPEED_100M (1 << 4) /**< 100 Mbps full-duplex */ 288 #define ETH_LINK_SPEED_1G (1 << 5) /**< 1 Gbps */ 289 #define ETH_LINK_SPEED_2_5G (1 << 6) /**< 2.5 Gbps */ 290 #define ETH_LINK_SPEED_5G (1 << 7) /**< 5 Gbps */ 291 #define ETH_LINK_SPEED_10G (1 << 8) /**< 10 Gbps */ 292 #define ETH_LINK_SPEED_20G (1 << 9) /**< 20 Gbps */ 293 #define ETH_LINK_SPEED_25G (1 << 10) /**< 25 Gbps */ 294 #define ETH_LINK_SPEED_40G (1 << 11) /**< 40 Gbps */ 295 #define ETH_LINK_SPEED_50G (1 << 12) /**< 50 Gbps */ 296 #define ETH_LINK_SPEED_56G (1 << 13) /**< 56 Gbps */ 297 #define ETH_LINK_SPEED_100G (1 << 14) /**< 100 Gbps */ 298 #define ETH_LINK_SPEED_200G (1 << 15) /**< 200 Gbps */ 299 300 /** 301 * Ethernet numeric link speeds in Mbps 302 */ 303 #define ETH_SPEED_NUM_NONE 0 /**< Not defined */ 304 #define ETH_SPEED_NUM_10M 10 /**< 10 Mbps */ 305 #define ETH_SPEED_NUM_100M 100 /**< 100 Mbps */ 306 #define ETH_SPEED_NUM_1G 1000 /**< 1 Gbps */ 307 #define ETH_SPEED_NUM_2_5G 2500 /**< 2.5 Gbps */ 308 #define ETH_SPEED_NUM_5G 5000 /**< 5 Gbps */ 309 #define ETH_SPEED_NUM_10G 10000 /**< 10 Gbps */ 310 #define ETH_SPEED_NUM_20G 20000 /**< 20 Gbps */ 311 #define ETH_SPEED_NUM_25G 25000 /**< 25 Gbps */ 312 #define ETH_SPEED_NUM_40G 40000 /**< 40 Gbps */ 313 #define ETH_SPEED_NUM_50G 50000 /**< 50 Gbps */ 314 #define ETH_SPEED_NUM_56G 56000 /**< 56 Gbps */ 315 #define ETH_SPEED_NUM_100G 100000 /**< 100 Gbps */ 316 #define ETH_SPEED_NUM_200G 200000 /**< 200 Gbps */ 317 #define ETH_SPEED_NUM_UNKNOWN UINT32_MAX /**< Unknown */ 318 319 /** 320 * A structure used to retrieve link-level information of an Ethernet port. 321 */ 322 __extension__ 323 struct rte_eth_link { 324 uint32_t link_speed; /**< ETH_SPEED_NUM_ */ 325 uint16_t link_duplex : 1; /**< ETH_LINK_[HALF/FULL]_DUPLEX */ 326 uint16_t link_autoneg : 1; /**< ETH_LINK_[AUTONEG/FIXED] */ 327 uint16_t link_status : 1; /**< ETH_LINK_[DOWN/UP] */ 328 } __rte_aligned(8); /**< aligned for atomic64 read/write */ 329 330 /* Utility constants */ 331 #define ETH_LINK_HALF_DUPLEX 0 /**< Half-duplex connection (see link_duplex). */ 332 #define ETH_LINK_FULL_DUPLEX 1 /**< Full-duplex connection (see link_duplex). */ 333 #define ETH_LINK_DOWN 0 /**< Link is down (see link_status). */ 334 #define ETH_LINK_UP 1 /**< Link is up (see link_status). */ 335 #define ETH_LINK_FIXED 0 /**< No autonegotiation (see link_autoneg). */ 336 #define ETH_LINK_AUTONEG 1 /**< Autonegotiated (see link_autoneg). */ 337 #define RTE_ETH_LINK_MAX_STR_LEN 40 /**< Max length of default link string. */ 338 339 /** 340 * A structure used to configure the ring threshold registers of an RX/TX 341 * queue for an Ethernet port. 342 */ 343 struct rte_eth_thresh { 344 uint8_t pthresh; /**< Ring prefetch threshold. */ 345 uint8_t hthresh; /**< Ring host threshold. */ 346 uint8_t wthresh; /**< Ring writeback threshold. */ 347 }; 348 349 /** 350 * Simple flags are used for rte_eth_conf.rxmode.mq_mode. 351 */ 352 #define ETH_MQ_RX_RSS_FLAG 0x1 353 #define ETH_MQ_RX_DCB_FLAG 0x2 354 #define ETH_MQ_RX_VMDQ_FLAG 0x4 355 356 /** 357 * A set of values to identify what method is to be used to route 358 * packets to multiple queues. 359 */ 360 enum rte_eth_rx_mq_mode { 361 /** None of DCB,RSS or VMDQ mode */ 362 ETH_MQ_RX_NONE = 0, 363 364 /** For RX side, only RSS is on */ 365 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG, 366 /** For RX side,only DCB is on. */ 367 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG, 368 /** Both DCB and RSS enable */ 369 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG, 370 371 /** Only VMDQ, no RSS nor DCB */ 372 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG, 373 /** RSS mode with VMDQ */ 374 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG, 375 /** Use VMDQ+DCB to route traffic to queues */ 376 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG, 377 /** Enable both VMDQ and DCB in VMDq */ 378 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG | 379 ETH_MQ_RX_VMDQ_FLAG, 380 }; 381 382 /** 383 * for rx mq mode backward compatible 384 */ 385 #define ETH_RSS ETH_MQ_RX_RSS 386 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB 387 #define ETH_DCB_RX ETH_MQ_RX_DCB 388 389 /** 390 * A set of values to identify what method is to be used to transmit 391 * packets using multi-TCs. 392 */ 393 enum rte_eth_tx_mq_mode { 394 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */ 395 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */ 396 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */ 397 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */ 398 }; 399 400 /** 401 * for tx mq mode backward compatible 402 */ 403 #define ETH_DCB_NONE ETH_MQ_TX_NONE 404 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB 405 #define ETH_DCB_TX ETH_MQ_TX_DCB 406 407 /** 408 * A structure used to configure the RX features of an Ethernet port. 409 */ 410 struct rte_eth_rxmode { 411 /** The multi-queue packet distribution mode to be used, e.g. RSS. */ 412 enum rte_eth_rx_mq_mode mq_mode; 413 uint32_t max_rx_pkt_len; /**< Only used if JUMBO_FRAME enabled. */ 414 /** Maximum allowed size of LRO aggregated packet. */ 415 uint32_t max_lro_pkt_size; 416 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/ 417 /** 418 * Per-port Rx offloads to be set using DEV_RX_OFFLOAD_* flags. 419 * Only offloads set on rx_offload_capa field on rte_eth_dev_info 420 * structure are allowed to be set. 421 */ 422 uint64_t offloads; 423 424 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 425 void *reserved_ptrs[2]; /**< Reserved for future fields */ 426 }; 427 428 /** 429 * VLAN types to indicate if it is for single VLAN, inner VLAN or outer VLAN. 430 * Note that single VLAN is treated the same as inner VLAN. 431 */ 432 enum rte_vlan_type { 433 ETH_VLAN_TYPE_UNKNOWN = 0, 434 ETH_VLAN_TYPE_INNER, /**< Inner VLAN. */ 435 ETH_VLAN_TYPE_OUTER, /**< Single VLAN, or outer VLAN. */ 436 ETH_VLAN_TYPE_MAX, 437 }; 438 439 /** 440 * A structure used to describe a vlan filter. 441 * If the bit corresponding to a VID is set, such VID is on. 442 */ 443 struct rte_vlan_filter_conf { 444 uint64_t ids[64]; 445 }; 446 447 /** 448 * A structure used to configure the Receive Side Scaling (RSS) feature 449 * of an Ethernet port. 450 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points 451 * to an array holding the RSS key to use for hashing specific header 452 * fields of received packets. The length of this array should be indicated 453 * by *rss_key_len* below. Otherwise, a default random hash key is used by 454 * the device driver. 455 * 456 * The *rss_key_len* field of the *rss_conf* structure indicates the length 457 * in bytes of the array pointed by *rss_key*. To be compatible, this length 458 * will be checked in i40e only. Others assume 40 bytes to be used as before. 459 * 460 * The *rss_hf* field of the *rss_conf* structure indicates the different 461 * types of IPv4/IPv6 packets to which the RSS hashing must be applied. 462 * Supplying an *rss_hf* equal to zero disables the RSS feature. 463 */ 464 struct rte_eth_rss_conf { 465 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */ 466 uint8_t rss_key_len; /**< hash key length in bytes. */ 467 uint64_t rss_hf; /**< Hash functions to apply - see below. */ 468 }; 469 470 /* 471 * A packet can be identified by hardware as different flow types. Different 472 * NIC hardware may support different flow types. 473 * Basically, the NIC hardware identifies the flow type as deep protocol as 474 * possible, and exclusively. For example, if a packet is identified as 475 * 'RTE_ETH_FLOW_NONFRAG_IPV4_TCP', it will not be any of other flow types, 476 * though it is an actual IPV4 packet. 477 */ 478 #define RTE_ETH_FLOW_UNKNOWN 0 479 #define RTE_ETH_FLOW_RAW 1 480 #define RTE_ETH_FLOW_IPV4 2 481 #define RTE_ETH_FLOW_FRAG_IPV4 3 482 #define RTE_ETH_FLOW_NONFRAG_IPV4_TCP 4 483 #define RTE_ETH_FLOW_NONFRAG_IPV4_UDP 5 484 #define RTE_ETH_FLOW_NONFRAG_IPV4_SCTP 6 485 #define RTE_ETH_FLOW_NONFRAG_IPV4_OTHER 7 486 #define RTE_ETH_FLOW_IPV6 8 487 #define RTE_ETH_FLOW_FRAG_IPV6 9 488 #define RTE_ETH_FLOW_NONFRAG_IPV6_TCP 10 489 #define RTE_ETH_FLOW_NONFRAG_IPV6_UDP 11 490 #define RTE_ETH_FLOW_NONFRAG_IPV6_SCTP 12 491 #define RTE_ETH_FLOW_NONFRAG_IPV6_OTHER 13 492 #define RTE_ETH_FLOW_L2_PAYLOAD 14 493 #define RTE_ETH_FLOW_IPV6_EX 15 494 #define RTE_ETH_FLOW_IPV6_TCP_EX 16 495 #define RTE_ETH_FLOW_IPV6_UDP_EX 17 496 #define RTE_ETH_FLOW_PORT 18 497 /**< Consider device port number as a flow differentiator */ 498 #define RTE_ETH_FLOW_VXLAN 19 /**< VXLAN protocol based flow */ 499 #define RTE_ETH_FLOW_GENEVE 20 /**< GENEVE protocol based flow */ 500 #define RTE_ETH_FLOW_NVGRE 21 /**< NVGRE protocol based flow */ 501 #define RTE_ETH_FLOW_VXLAN_GPE 22 /**< VXLAN-GPE protocol based flow */ 502 #define RTE_ETH_FLOW_GTPU 23 /**< GTPU protocol based flow */ 503 #define RTE_ETH_FLOW_MAX 24 504 505 /* 506 * Below macros are defined for RSS offload types, they can be used to 507 * fill rte_eth_rss_conf.rss_hf or rte_flow_action_rss.types. 508 */ 509 #define ETH_RSS_IPV4 (1ULL << 2) 510 #define ETH_RSS_FRAG_IPV4 (1ULL << 3) 511 #define ETH_RSS_NONFRAG_IPV4_TCP (1ULL << 4) 512 #define ETH_RSS_NONFRAG_IPV4_UDP (1ULL << 5) 513 #define ETH_RSS_NONFRAG_IPV4_SCTP (1ULL << 6) 514 #define ETH_RSS_NONFRAG_IPV4_OTHER (1ULL << 7) 515 #define ETH_RSS_IPV6 (1ULL << 8) 516 #define ETH_RSS_FRAG_IPV6 (1ULL << 9) 517 #define ETH_RSS_NONFRAG_IPV6_TCP (1ULL << 10) 518 #define ETH_RSS_NONFRAG_IPV6_UDP (1ULL << 11) 519 #define ETH_RSS_NONFRAG_IPV6_SCTP (1ULL << 12) 520 #define ETH_RSS_NONFRAG_IPV6_OTHER (1ULL << 13) 521 #define ETH_RSS_L2_PAYLOAD (1ULL << 14) 522 #define ETH_RSS_IPV6_EX (1ULL << 15) 523 #define ETH_RSS_IPV6_TCP_EX (1ULL << 16) 524 #define ETH_RSS_IPV6_UDP_EX (1ULL << 17) 525 #define ETH_RSS_PORT (1ULL << 18) 526 #define ETH_RSS_VXLAN (1ULL << 19) 527 #define ETH_RSS_GENEVE (1ULL << 20) 528 #define ETH_RSS_NVGRE (1ULL << 21) 529 #define ETH_RSS_GTPU (1ULL << 23) 530 #define ETH_RSS_ETH (1ULL << 24) 531 #define ETH_RSS_S_VLAN (1ULL << 25) 532 #define ETH_RSS_C_VLAN (1ULL << 26) 533 #define ETH_RSS_ESP (1ULL << 27) 534 #define ETH_RSS_AH (1ULL << 28) 535 #define ETH_RSS_L2TPV3 (1ULL << 29) 536 #define ETH_RSS_PFCP (1ULL << 30) 537 #define ETH_RSS_PPPOE (1ULL << 31) 538 #define ETH_RSS_ECPRI (1ULL << 32) 539 #define ETH_RSS_MPLS (1ULL << 33) 540 541 /* 542 * We use the following macros to combine with above ETH_RSS_* for 543 * more specific input set selection. These bits are defined starting 544 * from the high end of the 64 bits. 545 * Note: If we use above ETH_RSS_* without SRC/DST_ONLY, it represents 546 * both SRC and DST are taken into account. If SRC_ONLY and DST_ONLY of 547 * the same level are used simultaneously, it is the same case as none of 548 * them are added. 549 */ 550 #define ETH_RSS_L3_SRC_ONLY (1ULL << 63) 551 #define ETH_RSS_L3_DST_ONLY (1ULL << 62) 552 #define ETH_RSS_L4_SRC_ONLY (1ULL << 61) 553 #define ETH_RSS_L4_DST_ONLY (1ULL << 60) 554 #define ETH_RSS_L2_SRC_ONLY (1ULL << 59) 555 #define ETH_RSS_L2_DST_ONLY (1ULL << 58) 556 557 /* 558 * Only select IPV6 address prefix as RSS input set according to 559 * https://tools.ietf.org/html/rfc6052 560 * Must be combined with ETH_RSS_IPV6, ETH_RSS_NONFRAG_IPV6_UDP, 561 * ETH_RSS_NONFRAG_IPV6_TCP, ETH_RSS_NONFRAG_IPV6_SCTP. 562 */ 563 #define RTE_ETH_RSS_L3_PRE32 (1ULL << 57) 564 #define RTE_ETH_RSS_L3_PRE40 (1ULL << 56) 565 #define RTE_ETH_RSS_L3_PRE48 (1ULL << 55) 566 #define RTE_ETH_RSS_L3_PRE56 (1ULL << 54) 567 #define RTE_ETH_RSS_L3_PRE64 (1ULL << 53) 568 #define RTE_ETH_RSS_L3_PRE96 (1ULL << 52) 569 570 /* 571 * Use the following macros to combine with the above layers 572 * to choose inner and outer layers or both for RSS computation. 573 * Bits 50 and 51 are reserved for this. 574 */ 575 576 /** 577 * level 0, requests the default behavior. 578 * Depending on the packet type, it can mean outermost, innermost, 579 * anything in between or even no RSS. 580 * It basically stands for the innermost encapsulation level RSS 581 * can be performed on according to PMD and device capabilities. 582 */ 583 #define ETH_RSS_LEVEL_PMD_DEFAULT (0ULL << 50) 584 585 /** 586 * level 1, requests RSS to be performed on the outermost packet 587 * encapsulation level. 588 */ 589 #define ETH_RSS_LEVEL_OUTERMOST (1ULL << 50) 590 591 /** 592 * level 2, requests RSS to be performed on the specified inner packet 593 * encapsulation level, from outermost to innermost (lower to higher values). 594 */ 595 #define ETH_RSS_LEVEL_INNERMOST (2ULL << 50) 596 #define ETH_RSS_LEVEL_MASK (3ULL << 50) 597 598 #define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50) 599 600 /** 601 * For input set change of hash filter, if SRC_ONLY and DST_ONLY of 602 * the same level are used simultaneously, it is the same case as 603 * none of them are added. 604 * 605 * @param rss_hf 606 * RSS types with SRC/DST_ONLY. 607 * @return 608 * RSS types. 609 */ 610 static inline uint64_t 611 rte_eth_rss_hf_refine(uint64_t rss_hf) 612 { 613 if ((rss_hf & ETH_RSS_L3_SRC_ONLY) && (rss_hf & ETH_RSS_L3_DST_ONLY)) 614 rss_hf &= ~(ETH_RSS_L3_SRC_ONLY | ETH_RSS_L3_DST_ONLY); 615 616 if ((rss_hf & ETH_RSS_L4_SRC_ONLY) && (rss_hf & ETH_RSS_L4_DST_ONLY)) 617 rss_hf &= ~(ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY); 618 619 return rss_hf; 620 } 621 622 #define ETH_RSS_IPV6_PRE32 ( \ 623 ETH_RSS_IPV6 | \ 624 RTE_ETH_RSS_L3_PRE32) 625 626 #define ETH_RSS_IPV6_PRE40 ( \ 627 ETH_RSS_IPV6 | \ 628 RTE_ETH_RSS_L3_PRE40) 629 630 #define ETH_RSS_IPV6_PRE48 ( \ 631 ETH_RSS_IPV6 | \ 632 RTE_ETH_RSS_L3_PRE48) 633 634 #define ETH_RSS_IPV6_PRE56 ( \ 635 ETH_RSS_IPV6 | \ 636 RTE_ETH_RSS_L3_PRE56) 637 638 #define ETH_RSS_IPV6_PRE64 ( \ 639 ETH_RSS_IPV6 | \ 640 RTE_ETH_RSS_L3_PRE64) 641 642 #define ETH_RSS_IPV6_PRE96 ( \ 643 ETH_RSS_IPV6 | \ 644 RTE_ETH_RSS_L3_PRE96) 645 646 #define ETH_RSS_IPV6_PRE32_UDP ( \ 647 ETH_RSS_NONFRAG_IPV6_UDP | \ 648 RTE_ETH_RSS_L3_PRE32) 649 650 #define ETH_RSS_IPV6_PRE40_UDP ( \ 651 ETH_RSS_NONFRAG_IPV6_UDP | \ 652 RTE_ETH_RSS_L3_PRE40) 653 654 #define ETH_RSS_IPV6_PRE48_UDP ( \ 655 ETH_RSS_NONFRAG_IPV6_UDP | \ 656 RTE_ETH_RSS_L3_PRE48) 657 658 #define ETH_RSS_IPV6_PRE56_UDP ( \ 659 ETH_RSS_NONFRAG_IPV6_UDP | \ 660 RTE_ETH_RSS_L3_PRE56) 661 662 #define ETH_RSS_IPV6_PRE64_UDP ( \ 663 ETH_RSS_NONFRAG_IPV6_UDP | \ 664 RTE_ETH_RSS_L3_PRE64) 665 666 #define ETH_RSS_IPV6_PRE96_UDP ( \ 667 ETH_RSS_NONFRAG_IPV6_UDP | \ 668 RTE_ETH_RSS_L3_PRE96) 669 670 #define ETH_RSS_IPV6_PRE32_TCP ( \ 671 ETH_RSS_NONFRAG_IPV6_TCP | \ 672 RTE_ETH_RSS_L3_PRE32) 673 674 #define ETH_RSS_IPV6_PRE40_TCP ( \ 675 ETH_RSS_NONFRAG_IPV6_TCP | \ 676 RTE_ETH_RSS_L3_PRE40) 677 678 #define ETH_RSS_IPV6_PRE48_TCP ( \ 679 ETH_RSS_NONFRAG_IPV6_TCP | \ 680 RTE_ETH_RSS_L3_PRE48) 681 682 #define ETH_RSS_IPV6_PRE56_TCP ( \ 683 ETH_RSS_NONFRAG_IPV6_TCP | \ 684 RTE_ETH_RSS_L3_PRE56) 685 686 #define ETH_RSS_IPV6_PRE64_TCP ( \ 687 ETH_RSS_NONFRAG_IPV6_TCP | \ 688 RTE_ETH_RSS_L3_PRE64) 689 690 #define ETH_RSS_IPV6_PRE96_TCP ( \ 691 ETH_RSS_NONFRAG_IPV6_TCP | \ 692 RTE_ETH_RSS_L3_PRE96) 693 694 #define ETH_RSS_IPV6_PRE32_SCTP ( \ 695 ETH_RSS_NONFRAG_IPV6_SCTP | \ 696 RTE_ETH_RSS_L3_PRE32) 697 698 #define ETH_RSS_IPV6_PRE40_SCTP ( \ 699 ETH_RSS_NONFRAG_IPV6_SCTP | \ 700 RTE_ETH_RSS_L3_PRE40) 701 702 #define ETH_RSS_IPV6_PRE48_SCTP ( \ 703 ETH_RSS_NONFRAG_IPV6_SCTP | \ 704 RTE_ETH_RSS_L3_PRE48) 705 706 #define ETH_RSS_IPV6_PRE56_SCTP ( \ 707 ETH_RSS_NONFRAG_IPV6_SCTP | \ 708 RTE_ETH_RSS_L3_PRE56) 709 710 #define ETH_RSS_IPV6_PRE64_SCTP ( \ 711 ETH_RSS_NONFRAG_IPV6_SCTP | \ 712 RTE_ETH_RSS_L3_PRE64) 713 714 #define ETH_RSS_IPV6_PRE96_SCTP ( \ 715 ETH_RSS_NONFRAG_IPV6_SCTP | \ 716 RTE_ETH_RSS_L3_PRE96) 717 718 #define ETH_RSS_IP ( \ 719 ETH_RSS_IPV4 | \ 720 ETH_RSS_FRAG_IPV4 | \ 721 ETH_RSS_NONFRAG_IPV4_OTHER | \ 722 ETH_RSS_IPV6 | \ 723 ETH_RSS_FRAG_IPV6 | \ 724 ETH_RSS_NONFRAG_IPV6_OTHER | \ 725 ETH_RSS_IPV6_EX) 726 727 #define ETH_RSS_UDP ( \ 728 ETH_RSS_NONFRAG_IPV4_UDP | \ 729 ETH_RSS_NONFRAG_IPV6_UDP | \ 730 ETH_RSS_IPV6_UDP_EX) 731 732 #define ETH_RSS_TCP ( \ 733 ETH_RSS_NONFRAG_IPV4_TCP | \ 734 ETH_RSS_NONFRAG_IPV6_TCP | \ 735 ETH_RSS_IPV6_TCP_EX) 736 737 #define ETH_RSS_SCTP ( \ 738 ETH_RSS_NONFRAG_IPV4_SCTP | \ 739 ETH_RSS_NONFRAG_IPV6_SCTP) 740 741 #define ETH_RSS_TUNNEL ( \ 742 ETH_RSS_VXLAN | \ 743 ETH_RSS_GENEVE | \ 744 ETH_RSS_NVGRE) 745 746 #define ETH_RSS_VLAN ( \ 747 ETH_RSS_S_VLAN | \ 748 ETH_RSS_C_VLAN) 749 750 /**< Mask of valid RSS hash protocols */ 751 #define ETH_RSS_PROTO_MASK ( \ 752 ETH_RSS_IPV4 | \ 753 ETH_RSS_FRAG_IPV4 | \ 754 ETH_RSS_NONFRAG_IPV4_TCP | \ 755 ETH_RSS_NONFRAG_IPV4_UDP | \ 756 ETH_RSS_NONFRAG_IPV4_SCTP | \ 757 ETH_RSS_NONFRAG_IPV4_OTHER | \ 758 ETH_RSS_IPV6 | \ 759 ETH_RSS_FRAG_IPV6 | \ 760 ETH_RSS_NONFRAG_IPV6_TCP | \ 761 ETH_RSS_NONFRAG_IPV6_UDP | \ 762 ETH_RSS_NONFRAG_IPV6_SCTP | \ 763 ETH_RSS_NONFRAG_IPV6_OTHER | \ 764 ETH_RSS_L2_PAYLOAD | \ 765 ETH_RSS_IPV6_EX | \ 766 ETH_RSS_IPV6_TCP_EX | \ 767 ETH_RSS_IPV6_UDP_EX | \ 768 ETH_RSS_PORT | \ 769 ETH_RSS_VXLAN | \ 770 ETH_RSS_GENEVE | \ 771 ETH_RSS_NVGRE | \ 772 ETH_RSS_MPLS) 773 774 /* 775 * Definitions used for redirection table entry size. 776 * Some RSS RETA sizes may not be supported by some drivers, check the 777 * documentation or the description of relevant functions for more details. 778 */ 779 #define ETH_RSS_RETA_SIZE_64 64 780 #define ETH_RSS_RETA_SIZE_128 128 781 #define ETH_RSS_RETA_SIZE_256 256 782 #define ETH_RSS_RETA_SIZE_512 512 783 #define RTE_RETA_GROUP_SIZE 64 784 785 /* Definitions used for VMDQ and DCB functionality */ 786 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */ 787 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */ 788 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */ 789 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */ 790 791 /* DCB capability defines */ 792 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */ 793 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */ 794 795 /* Definitions used for VLAN Offload functionality */ 796 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */ 797 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */ 798 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */ 799 #define ETH_QINQ_STRIP_OFFLOAD 0x0008 /**< QINQ Strip On/Off */ 800 801 /* Definitions used for mask VLAN setting */ 802 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */ 803 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/ 804 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/ 805 #define ETH_QINQ_STRIP_MASK 0x0008 /**< QINQ Strip setting mask */ 806 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/ 807 808 /* Definitions used for receive MAC address */ 809 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */ 810 811 /* Definitions used for unicast hash */ 812 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */ 813 814 /* Definitions used for VMDQ pool rx mode setting */ 815 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */ 816 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */ 817 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */ 818 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */ 819 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */ 820 821 /** Maximum nb. of vlan per mirror rule */ 822 #define ETH_MIRROR_MAX_VLANS 64 823 824 #define ETH_MIRROR_VIRTUAL_POOL_UP 0x01 /**< Virtual Pool uplink Mirroring. */ 825 #define ETH_MIRROR_UPLINK_PORT 0x02 /**< Uplink Port Mirroring. */ 826 #define ETH_MIRROR_DOWNLINK_PORT 0x04 /**< Downlink Port Mirroring. */ 827 #define ETH_MIRROR_VLAN 0x08 /**< VLAN Mirroring. */ 828 #define ETH_MIRROR_VIRTUAL_POOL_DOWN 0x10 /**< Virtual Pool downlink Mirroring. */ 829 830 /** 831 * A structure used to configure VLAN traffic mirror of an Ethernet port. 832 */ 833 struct rte_eth_vlan_mirror { 834 uint64_t vlan_mask; /**< mask for valid VLAN ID. */ 835 /** VLAN ID list for vlan mirroring. */ 836 uint16_t vlan_id[ETH_MIRROR_MAX_VLANS]; 837 }; 838 839 /** 840 * A structure used to configure traffic mirror of an Ethernet port. 841 */ 842 struct rte_eth_mirror_conf { 843 uint8_t rule_type; /**< Mirroring rule type */ 844 uint8_t dst_pool; /**< Destination pool for this mirror rule. */ 845 uint64_t pool_mask; /**< Bitmap of pool for pool mirroring */ 846 /** VLAN ID setting for VLAN mirroring. */ 847 struct rte_eth_vlan_mirror vlan; 848 }; 849 850 /** 851 * A structure used to configure 64 entries of Redirection Table of the 852 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure 853 * more than 64 entries supported by hardware, an array of this structure 854 * is needed. 855 */ 856 struct rte_eth_rss_reta_entry64 { 857 uint64_t mask; 858 /**< Mask bits indicate which entries need to be updated/queried. */ 859 uint16_t reta[RTE_RETA_GROUP_SIZE]; 860 /**< Group of 64 redirection table entries. */ 861 }; 862 863 /** 864 * This enum indicates the possible number of traffic classes 865 * in DCB configurations 866 */ 867 enum rte_eth_nb_tcs { 868 ETH_4_TCS = 4, /**< 4 TCs with DCB. */ 869 ETH_8_TCS = 8 /**< 8 TCs with DCB. */ 870 }; 871 872 /** 873 * This enum indicates the possible number of queue pools 874 * in VMDQ configurations. 875 */ 876 enum rte_eth_nb_pools { 877 ETH_8_POOLS = 8, /**< 8 VMDq pools. */ 878 ETH_16_POOLS = 16, /**< 16 VMDq pools. */ 879 ETH_32_POOLS = 32, /**< 32 VMDq pools. */ 880 ETH_64_POOLS = 64 /**< 64 VMDq pools. */ 881 }; 882 883 /* This structure may be extended in future. */ 884 struct rte_eth_dcb_rx_conf { 885 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */ 886 /** Traffic class each UP mapped to. */ 887 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 888 }; 889 890 struct rte_eth_vmdq_dcb_tx_conf { 891 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */ 892 /** Traffic class each UP mapped to. */ 893 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 894 }; 895 896 struct rte_eth_dcb_tx_conf { 897 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */ 898 /** Traffic class each UP mapped to. */ 899 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 900 }; 901 902 struct rte_eth_vmdq_tx_conf { 903 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */ 904 }; 905 906 /** 907 * A structure used to configure the VMDQ+DCB feature 908 * of an Ethernet port. 909 * 910 * Using this feature, packets are routed to a pool of queues, based 911 * on the vlan id in the vlan tag, and then to a specific queue within 912 * that pool, using the user priority vlan tag field. 913 * 914 * A default pool may be used, if desired, to route all traffic which 915 * does not match the vlan filter rules. 916 */ 917 struct rte_eth_vmdq_dcb_conf { 918 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */ 919 uint8_t enable_default_pool; /**< If non-zero, use a default pool */ 920 uint8_t default_pool; /**< The default pool, if applicable */ 921 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */ 922 struct { 923 uint16_t vlan_id; /**< The vlan id of the received frame */ 924 uint64_t pools; /**< Bitmask of pools for packet rx */ 925 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */ 926 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 927 /**< Selects a queue in a pool */ 928 }; 929 930 /** 931 * A structure used to configure the VMDQ feature of an Ethernet port when 932 * not combined with the DCB feature. 933 * 934 * Using this feature, packets are routed to a pool of queues. By default, 935 * the pool selection is based on the MAC address, the vlan id in the 936 * vlan tag as specified in the pool_map array. 937 * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool 938 * selection using only the MAC address. MAC address to pool mapping is done 939 * using the rte_eth_dev_mac_addr_add function, with the pool parameter 940 * corresponding to the pool id. 941 * 942 * Queue selection within the selected pool will be done using RSS when 943 * it is enabled or revert to the first queue of the pool if not. 944 * 945 * A default pool may be used, if desired, to route all traffic which 946 * does not match the vlan filter rules or any pool MAC address. 947 */ 948 struct rte_eth_vmdq_rx_conf { 949 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */ 950 uint8_t enable_default_pool; /**< If non-zero, use a default pool */ 951 uint8_t default_pool; /**< The default pool, if applicable */ 952 uint8_t enable_loop_back; /**< Enable VT loop back */ 953 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */ 954 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */ 955 struct { 956 uint16_t vlan_id; /**< The vlan id of the received frame */ 957 uint64_t pools; /**< Bitmask of pools for packet rx */ 958 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */ 959 }; 960 961 /** 962 * A structure used to configure the TX features of an Ethernet port. 963 */ 964 struct rte_eth_txmode { 965 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */ 966 /** 967 * Per-port Tx offloads to be set using DEV_TX_OFFLOAD_* flags. 968 * Only offloads set on tx_offload_capa field on rte_eth_dev_info 969 * structure are allowed to be set. 970 */ 971 uint64_t offloads; 972 973 uint16_t pvid; 974 __extension__ 975 uint8_t hw_vlan_reject_tagged : 1, 976 /**< If set, reject sending out tagged pkts */ 977 hw_vlan_reject_untagged : 1, 978 /**< If set, reject sending out untagged pkts */ 979 hw_vlan_insert_pvid : 1; 980 /**< If set, enable port based VLAN insertion */ 981 982 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 983 void *reserved_ptrs[2]; /**< Reserved for future fields */ 984 }; 985 986 /** 987 * @warning 988 * @b EXPERIMENTAL: this structure may change without prior notice. 989 * 990 * A structure used to configure an Rx packet segment to split. 991 * 992 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field, 993 * the PMD will split the received packets into multiple segments 994 * according to the specification in the description array: 995 * 996 * - The first network buffer will be allocated from the memory pool, 997 * specified in the first array element, the second buffer, from the 998 * pool in the second element, and so on. 999 * 1000 * - The offsets from the segment description elements specify 1001 * the data offset from the buffer beginning except the first mbuf. 1002 * The first segment offset is added with RTE_PKTMBUF_HEADROOM. 1003 * 1004 * - The lengths in the elements define the maximal data amount 1005 * being received to each segment. The receiving starts with filling 1006 * up the first mbuf data buffer up to specified length. If the 1007 * there are data remaining (packet is longer than buffer in the first 1008 * mbuf) the following data will be pushed to the next segment 1009 * up to its own length, and so on. 1010 * 1011 * - If the length in the segment description element is zero 1012 * the actual buffer size will be deduced from the appropriate 1013 * memory pool properties. 1014 * 1015 * - If there is not enough elements to describe the buffer for entire 1016 * packet of maximal length the following parameters will be used 1017 * for the all remaining segments: 1018 * - pool from the last valid element 1019 * - the buffer size from this pool 1020 * - zero offset 1021 */ 1022 struct rte_eth_rxseg_split { 1023 struct rte_mempool *mp; /**< Memory pool to allocate segment from. */ 1024 uint16_t length; /**< Segment data length, configures split point. */ 1025 uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */ 1026 uint32_t reserved; /**< Reserved field. */ 1027 }; 1028 1029 /** 1030 * @warning 1031 * @b EXPERIMENTAL: this structure may change without prior notice. 1032 * 1033 * A common structure used to describe Rx packet segment properties. 1034 */ 1035 union rte_eth_rxseg { 1036 /* The settings for buffer split offload. */ 1037 struct rte_eth_rxseg_split split; 1038 /* The other features settings should be added here. */ 1039 }; 1040 1041 /** 1042 * A structure used to configure an RX ring of an Ethernet port. 1043 */ 1044 struct rte_eth_rxconf { 1045 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */ 1046 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */ 1047 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */ 1048 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */ 1049 uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */ 1050 /** 1051 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags. 1052 * Only offloads set on rx_queue_offload_capa or rx_offload_capa 1053 * fields on rte_eth_dev_info structure are allowed to be set. 1054 */ 1055 uint64_t offloads; 1056 /** 1057 * Points to the array of segment descriptions for an entire packet. 1058 * Array elements are properties for consecutive Rx segments. 1059 * 1060 * The supported capabilities of receiving segmentation is reported 1061 * in rte_eth_dev_info.rx_seg_capa field. 1062 */ 1063 union rte_eth_rxseg *rx_seg; 1064 1065 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 1066 void *reserved_ptrs[2]; /**< Reserved for future fields */ 1067 }; 1068 1069 /** 1070 * A structure used to configure a TX ring of an Ethernet port. 1071 */ 1072 struct rte_eth_txconf { 1073 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */ 1074 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */ 1075 uint16_t tx_free_thresh; /**< Start freeing TX buffers if there are 1076 less free descriptors than this value. */ 1077 1078 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */ 1079 /** 1080 * Per-queue Tx offloads to be set using DEV_TX_OFFLOAD_* flags. 1081 * Only offloads set on tx_queue_offload_capa or tx_offload_capa 1082 * fields on rte_eth_dev_info structure are allowed to be set. 1083 */ 1084 uint64_t offloads; 1085 1086 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 1087 void *reserved_ptrs[2]; /**< Reserved for future fields */ 1088 }; 1089 1090 /** 1091 * @warning 1092 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 1093 * 1094 * A structure used to return the hairpin capabilities that are supported. 1095 */ 1096 struct rte_eth_hairpin_cap { 1097 /** The max number of hairpin queues (different bindings). */ 1098 uint16_t max_nb_queues; 1099 /** Max number of Rx queues to be connected to one Tx queue. */ 1100 uint16_t max_rx_2_tx; 1101 /** Max number of Tx queues to be connected to one Rx queue. */ 1102 uint16_t max_tx_2_rx; 1103 uint16_t max_nb_desc; /**< The max num of descriptors. */ 1104 }; 1105 1106 #define RTE_ETH_MAX_HAIRPIN_PEERS 32 1107 1108 /** 1109 * @warning 1110 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 1111 * 1112 * A structure used to hold hairpin peer data. 1113 */ 1114 struct rte_eth_hairpin_peer { 1115 uint16_t port; /**< Peer port. */ 1116 uint16_t queue; /**< Peer queue. */ 1117 }; 1118 1119 /** 1120 * @warning 1121 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 1122 * 1123 * A structure used to configure hairpin binding. 1124 */ 1125 struct rte_eth_hairpin_conf { 1126 uint32_t peer_count:16; /**< The number of peers. */ 1127 1128 /** 1129 * Explicit Tx flow rule mode. 1130 * One hairpin pair of queues should have the same attribute. 1131 * 1132 * - When set, the user should be responsible for inserting the hairpin 1133 * Tx part flows and removing them. 1134 * - When clear, the PMD will try to handle the Tx part of the flows, 1135 * e.g., by splitting one flow into two parts. 1136 */ 1137 uint32_t tx_explicit:1; 1138 1139 /** 1140 * Manually bind hairpin queues. 1141 * One hairpin pair of queues should have the same attribute. 1142 * 1143 * - When set, to enable hairpin, the user should call the hairpin bind 1144 * function after all the queues are set up properly and the ports are 1145 * started. Also, the hairpin unbind function should be called 1146 * accordingly before stopping a port that with hairpin configured. 1147 * - When clear, the PMD will try to enable the hairpin with the queues 1148 * configured automatically during port start. 1149 */ 1150 uint32_t manual_bind:1; 1151 uint32_t reserved:14; /**< Reserved bits. */ 1152 struct rte_eth_hairpin_peer peers[RTE_ETH_MAX_HAIRPIN_PEERS]; 1153 }; 1154 1155 /** 1156 * A structure contains information about HW descriptor ring limitations. 1157 */ 1158 struct rte_eth_desc_lim { 1159 uint16_t nb_max; /**< Max allowed number of descriptors. */ 1160 uint16_t nb_min; /**< Min allowed number of descriptors. */ 1161 uint16_t nb_align; /**< Number of descriptors should be aligned to. */ 1162 1163 /** 1164 * Max allowed number of segments per whole packet. 1165 * 1166 * - For TSO packet this is the total number of data descriptors allowed 1167 * by device. 1168 * 1169 * @see nb_mtu_seg_max 1170 */ 1171 uint16_t nb_seg_max; 1172 1173 /** 1174 * Max number of segments per one MTU. 1175 * 1176 * - For non-TSO packet, this is the maximum allowed number of segments 1177 * in a single transmit packet. 1178 * 1179 * - For TSO packet each segment within the TSO may span up to this 1180 * value. 1181 * 1182 * @see nb_seg_max 1183 */ 1184 uint16_t nb_mtu_seg_max; 1185 }; 1186 1187 /** 1188 * This enum indicates the flow control mode 1189 */ 1190 enum rte_eth_fc_mode { 1191 RTE_FC_NONE = 0, /**< Disable flow control. */ 1192 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */ 1193 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */ 1194 RTE_FC_FULL /**< Enable flow control on both side. */ 1195 }; 1196 1197 /** 1198 * A structure used to configure Ethernet flow control parameter. 1199 * These parameters will be configured into the register of the NIC. 1200 * Please refer to the corresponding data sheet for proper value. 1201 */ 1202 struct rte_eth_fc_conf { 1203 uint32_t high_water; /**< High threshold value to trigger XOFF */ 1204 uint32_t low_water; /**< Low threshold value to trigger XON */ 1205 uint16_t pause_time; /**< Pause quota in the Pause frame */ 1206 uint16_t send_xon; /**< Is XON frame need be sent */ 1207 enum rte_eth_fc_mode mode; /**< Link flow control mode */ 1208 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */ 1209 uint8_t autoneg; /**< Use Pause autoneg */ 1210 }; 1211 1212 /** 1213 * A structure used to configure Ethernet priority flow control parameter. 1214 * These parameters will be configured into the register of the NIC. 1215 * Please refer to the corresponding data sheet for proper value. 1216 */ 1217 struct rte_eth_pfc_conf { 1218 struct rte_eth_fc_conf fc; /**< General flow control parameter. */ 1219 uint8_t priority; /**< VLAN User Priority. */ 1220 }; 1221 1222 /** 1223 * Tunnel type for device-specific classifier configuration. 1224 * @see rte_eth_udp_tunnel 1225 */ 1226 enum rte_eth_tunnel_type { 1227 RTE_TUNNEL_TYPE_NONE = 0, 1228 RTE_TUNNEL_TYPE_VXLAN, 1229 RTE_TUNNEL_TYPE_GENEVE, 1230 RTE_TUNNEL_TYPE_TEREDO, 1231 RTE_TUNNEL_TYPE_NVGRE, 1232 RTE_TUNNEL_TYPE_IP_IN_GRE, 1233 RTE_L2_TUNNEL_TYPE_E_TAG, 1234 RTE_TUNNEL_TYPE_VXLAN_GPE, 1235 RTE_TUNNEL_TYPE_ECPRI, 1236 RTE_TUNNEL_TYPE_MAX, 1237 }; 1238 1239 /* Deprecated API file for rte_eth_dev_filter_* functions */ 1240 #include "rte_eth_ctrl.h" 1241 1242 /** 1243 * Memory space that can be configured to store Flow Director filters 1244 * in the board memory. 1245 */ 1246 enum rte_fdir_pballoc_type { 1247 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */ 1248 RTE_FDIR_PBALLOC_128K, /**< 128k. */ 1249 RTE_FDIR_PBALLOC_256K, /**< 256k. */ 1250 }; 1251 1252 /** 1253 * Select report mode of FDIR hash information in RX descriptors. 1254 */ 1255 enum rte_fdir_status_mode { 1256 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */ 1257 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */ 1258 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */ 1259 }; 1260 1261 /** 1262 * A structure used to configure the Flow Director (FDIR) feature 1263 * of an Ethernet port. 1264 * 1265 * If mode is RTE_FDIR_MODE_NONE, the pballoc value is ignored. 1266 */ 1267 struct rte_fdir_conf { 1268 enum rte_fdir_mode mode; /**< Flow Director mode. */ 1269 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */ 1270 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */ 1271 /** RX queue of packets matching a "drop" filter in perfect mode. */ 1272 uint8_t drop_queue; 1273 struct rte_eth_fdir_masks mask; 1274 struct rte_eth_fdir_flex_conf flex_conf; 1275 /**< Flex payload configuration. */ 1276 }; 1277 1278 /** 1279 * UDP tunneling configuration. 1280 * 1281 * Used to configure the classifier of a device, 1282 * associating an UDP port with a type of tunnel. 1283 * 1284 * Some NICs may need such configuration to properly parse a tunnel 1285 * with any standard or custom UDP port. 1286 */ 1287 struct rte_eth_udp_tunnel { 1288 uint16_t udp_port; /**< UDP port used for the tunnel. */ 1289 uint8_t prot_type; /**< Tunnel type. @see rte_eth_tunnel_type */ 1290 }; 1291 1292 /** 1293 * A structure used to enable/disable specific device interrupts. 1294 */ 1295 struct rte_intr_conf { 1296 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */ 1297 uint32_t lsc:1; 1298 /** enable/disable rxq interrupt. 0 (default) - disable, 1 enable */ 1299 uint32_t rxq:1; 1300 /** enable/disable rmv interrupt. 0 (default) - disable, 1 enable */ 1301 uint32_t rmv:1; 1302 }; 1303 1304 /** 1305 * A structure used to configure an Ethernet port. 1306 * Depending upon the RX multi-queue mode, extra advanced 1307 * configuration settings may be needed. 1308 */ 1309 struct rte_eth_conf { 1310 uint32_t link_speeds; /**< bitmap of ETH_LINK_SPEED_XXX of speeds to be 1311 used. ETH_LINK_SPEED_FIXED disables link 1312 autonegotiation, and a unique speed shall be 1313 set. Otherwise, the bitmap defines the set of 1314 speeds to be advertised. If the special value 1315 ETH_LINK_SPEED_AUTONEG (0) is used, all speeds 1316 supported are advertised. */ 1317 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */ 1318 struct rte_eth_txmode txmode; /**< Port TX configuration. */ 1319 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value 1320 is 0, meaning the loopback mode is disabled. 1321 Read the datasheet of given ethernet controller 1322 for details. The possible values of this field 1323 are defined in implementation of each driver. */ 1324 struct { 1325 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */ 1326 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf; 1327 /**< Port vmdq+dcb configuration. */ 1328 struct rte_eth_dcb_rx_conf dcb_rx_conf; 1329 /**< Port dcb RX configuration. */ 1330 struct rte_eth_vmdq_rx_conf vmdq_rx_conf; 1331 /**< Port vmdq RX configuration. */ 1332 } rx_adv_conf; /**< Port RX filtering configuration. */ 1333 union { 1334 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf; 1335 /**< Port vmdq+dcb TX configuration. */ 1336 struct rte_eth_dcb_tx_conf dcb_tx_conf; 1337 /**< Port dcb TX configuration. */ 1338 struct rte_eth_vmdq_tx_conf vmdq_tx_conf; 1339 /**< Port vmdq TX configuration. */ 1340 } tx_adv_conf; /**< Port TX DCB configuration (union). */ 1341 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC 1342 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */ 1343 uint32_t dcb_capability_en; 1344 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. DEPRECATED */ 1345 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */ 1346 }; 1347 1348 /** 1349 * RX offload capabilities of a device. 1350 */ 1351 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001 1352 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002 1353 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004 1354 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008 1355 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010 1356 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020 1357 #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040 1358 #define DEV_RX_OFFLOAD_MACSEC_STRIP 0x00000080 1359 #define DEV_RX_OFFLOAD_HEADER_SPLIT 0x00000100 1360 #define DEV_RX_OFFLOAD_VLAN_FILTER 0x00000200 1361 #define DEV_RX_OFFLOAD_VLAN_EXTEND 0x00000400 1362 #define DEV_RX_OFFLOAD_JUMBO_FRAME 0x00000800 1363 #define DEV_RX_OFFLOAD_SCATTER 0x00002000 1364 /** 1365 * Timestamp is set by the driver in RTE_MBUF_DYNFIELD_TIMESTAMP_NAME 1366 * and RTE_MBUF_DYNFLAG_RX_TIMESTAMP_NAME is set in ol_flags. 1367 * The mbuf field and flag are registered when the offload is configured. 1368 */ 1369 #define DEV_RX_OFFLOAD_TIMESTAMP 0x00004000 1370 #define DEV_RX_OFFLOAD_SECURITY 0x00008000 1371 #define DEV_RX_OFFLOAD_KEEP_CRC 0x00010000 1372 #define DEV_RX_OFFLOAD_SCTP_CKSUM 0x00020000 1373 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM 0x00040000 1374 #define DEV_RX_OFFLOAD_RSS_HASH 0x00080000 1375 #define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000 1376 1377 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \ 1378 DEV_RX_OFFLOAD_UDP_CKSUM | \ 1379 DEV_RX_OFFLOAD_TCP_CKSUM) 1380 #define DEV_RX_OFFLOAD_VLAN (DEV_RX_OFFLOAD_VLAN_STRIP | \ 1381 DEV_RX_OFFLOAD_VLAN_FILTER | \ 1382 DEV_RX_OFFLOAD_VLAN_EXTEND | \ 1383 DEV_RX_OFFLOAD_QINQ_STRIP) 1384 1385 /* 1386 * If new Rx offload capabilities are defined, they also must be 1387 * mentioned in rte_rx_offload_names in rte_ethdev.c file. 1388 */ 1389 1390 /** 1391 * TX offload capabilities of a device. 1392 */ 1393 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001 1394 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002 1395 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004 1396 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008 1397 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010 1398 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020 1399 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040 1400 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */ 1401 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100 1402 #define DEV_TX_OFFLOAD_VXLAN_TNL_TSO 0x00000200 /**< Used for tunneling packet. */ 1403 #define DEV_TX_OFFLOAD_GRE_TNL_TSO 0x00000400 /**< Used for tunneling packet. */ 1404 #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */ 1405 #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */ 1406 #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000 1407 #define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000 1408 /**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same 1409 * tx queue without SW lock. 1410 */ 1411 #define DEV_TX_OFFLOAD_MULTI_SEGS 0x00008000 1412 /**< Device supports multi segment send. */ 1413 #define DEV_TX_OFFLOAD_MBUF_FAST_FREE 0x00010000 1414 /**< Device supports optimization for fast release of mbufs. 1415 * When set application must guarantee that per-queue all mbufs comes from 1416 * the same mempool and has refcnt = 1. 1417 */ 1418 #define DEV_TX_OFFLOAD_SECURITY 0x00020000 1419 /** 1420 * Device supports generic UDP tunneled packet TSO. 1421 * Application must set PKT_TX_TUNNEL_UDP and other mbuf fields required 1422 * for tunnel TSO. 1423 */ 1424 #define DEV_TX_OFFLOAD_UDP_TNL_TSO 0x00040000 1425 /** 1426 * Device supports generic IP tunneled packet TSO. 1427 * Application must set PKT_TX_TUNNEL_IP and other mbuf fields required 1428 * for tunnel TSO. 1429 */ 1430 #define DEV_TX_OFFLOAD_IP_TNL_TSO 0x00080000 1431 /** Device supports outer UDP checksum */ 1432 #define DEV_TX_OFFLOAD_OUTER_UDP_CKSUM 0x00100000 1433 /** 1434 * Device sends on time read from RTE_MBUF_DYNFIELD_TIMESTAMP_NAME 1435 * if RTE_MBUF_DYNFLAG_TX_TIMESTAMP_NAME is set in ol_flags. 1436 * The mbuf field and flag are registered when the offload is configured. 1437 */ 1438 #define DEV_TX_OFFLOAD_SEND_ON_TIMESTAMP 0x00200000 1439 /* 1440 * If new Tx offload capabilities are defined, they also must be 1441 * mentioned in rte_tx_offload_names in rte_ethdev.c file. 1442 */ 1443 1444 /**@{@name Device capabilities 1445 * Non-offload capabilities reported in rte_eth_dev_info.dev_capa. 1446 */ 1447 /** Device supports Rx queue setup after device started. */ 1448 #define RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP 0x00000001 1449 /** Device supports Tx queue setup after device started. */ 1450 #define RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP 0x00000002 1451 /**@}*/ 1452 1453 /* 1454 * Fallback default preferred Rx/Tx port parameters. 1455 * These are used if an application requests default parameters 1456 * but the PMD does not provide preferred values. 1457 */ 1458 #define RTE_ETH_DEV_FALLBACK_RX_RINGSIZE 512 1459 #define RTE_ETH_DEV_FALLBACK_TX_RINGSIZE 512 1460 #define RTE_ETH_DEV_FALLBACK_RX_NBQUEUES 1 1461 #define RTE_ETH_DEV_FALLBACK_TX_NBQUEUES 1 1462 1463 /** 1464 * Preferred Rx/Tx port parameters. 1465 * There are separate instances of this structure for transmission 1466 * and reception respectively. 1467 */ 1468 struct rte_eth_dev_portconf { 1469 uint16_t burst_size; /**< Device-preferred burst size */ 1470 uint16_t ring_size; /**< Device-preferred size of queue rings */ 1471 uint16_t nb_queues; /**< Device-preferred number of queues */ 1472 }; 1473 1474 /** 1475 * Default values for switch domain id when ethdev does not support switch 1476 * domain definitions. 1477 */ 1478 #define RTE_ETH_DEV_SWITCH_DOMAIN_ID_INVALID (UINT16_MAX) 1479 1480 /** 1481 * Ethernet device associated switch information 1482 */ 1483 struct rte_eth_switch_info { 1484 const char *name; /**< switch name */ 1485 uint16_t domain_id; /**< switch domain id */ 1486 uint16_t port_id; 1487 /**< 1488 * mapping to the devices physical switch port as enumerated from the 1489 * perspective of the embedded interconnect/switch. For SR-IOV enabled 1490 * device this may correspond to the VF_ID of each virtual function, 1491 * but each driver should explicitly define the mapping of switch 1492 * port identifier to that physical interconnect/switch 1493 */ 1494 }; 1495 1496 /** 1497 * @warning 1498 * @b EXPERIMENTAL: this structure may change without prior notice. 1499 * 1500 * Ethernet device Rx buffer segmentation capabilities. 1501 */ 1502 struct rte_eth_rxseg_capa { 1503 __extension__ 1504 uint32_t multi_pools:1; /**< Supports receiving to multiple pools.*/ 1505 uint32_t offset_allowed:1; /**< Supports buffer offsets. */ 1506 uint32_t offset_align_log2:4; /**< Required offset alignment. */ 1507 uint16_t max_nseg; /**< Maximum amount of segments to split. */ 1508 uint16_t reserved; /**< Reserved field. */ 1509 }; 1510 1511 /** 1512 * Ethernet device information 1513 */ 1514 1515 /** 1516 * Ethernet device representor port type. 1517 */ 1518 enum rte_eth_representor_type { 1519 RTE_ETH_REPRESENTOR_NONE, /**< not a representor. */ 1520 RTE_ETH_REPRESENTOR_VF, /**< representor of Virtual Function. */ 1521 RTE_ETH_REPRESENTOR_SF, /**< representor of Sub Function. */ 1522 RTE_ETH_REPRESENTOR_PF, /**< representor of Physical Function. */ 1523 }; 1524 1525 /** 1526 * A structure used to retrieve the contextual information of 1527 * an Ethernet device, such as the controlling driver of the 1528 * device, etc... 1529 */ 1530 struct rte_eth_dev_info { 1531 struct rte_device *device; /** Generic device information */ 1532 const char *driver_name; /**< Device Driver name. */ 1533 unsigned int if_index; /**< Index to bound host interface, or 0 if none. 1534 Use if_indextoname() to translate into an interface name. */ 1535 uint16_t min_mtu; /**< Minimum MTU allowed */ 1536 uint16_t max_mtu; /**< Maximum MTU allowed */ 1537 const uint32_t *dev_flags; /**< Device flags */ 1538 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */ 1539 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */ 1540 /** Maximum configurable size of LRO aggregated packet. */ 1541 uint32_t max_lro_pkt_size; 1542 uint16_t max_rx_queues; /**< Maximum number of RX queues. */ 1543 uint16_t max_tx_queues; /**< Maximum number of TX queues. */ 1544 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */ 1545 uint32_t max_hash_mac_addrs; 1546 /** Maximum number of hash MAC addresses for MTA and UTA. */ 1547 uint16_t max_vfs; /**< Maximum number of VFs. */ 1548 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */ 1549 struct rte_eth_rxseg_capa rx_seg_capa; /**< Segmentation capability.*/ 1550 uint64_t rx_offload_capa; 1551 /**< All RX offload capabilities including all per-queue ones */ 1552 uint64_t tx_offload_capa; 1553 /**< All TX offload capabilities including all per-queue ones */ 1554 uint64_t rx_queue_offload_capa; 1555 /**< Device per-queue RX offload capabilities. */ 1556 uint64_t tx_queue_offload_capa; 1557 /**< Device per-queue TX offload capabilities. */ 1558 uint16_t reta_size; 1559 /**< Device redirection table size, the total number of entries. */ 1560 uint8_t hash_key_size; /**< Hash key size in bytes */ 1561 /** Bit mask of RSS offloads, the bit offset also means flow type */ 1562 uint64_t flow_type_rss_offloads; 1563 struct rte_eth_rxconf default_rxconf; /**< Default RX configuration */ 1564 struct rte_eth_txconf default_txconf; /**< Default TX configuration */ 1565 uint16_t vmdq_queue_base; /**< First queue ID for VMDQ pools. */ 1566 uint16_t vmdq_queue_num; /**< Queue number for VMDQ pools. */ 1567 uint16_t vmdq_pool_base; /**< First ID of VMDQ pools. */ 1568 struct rte_eth_desc_lim rx_desc_lim; /**< RX descriptors limits */ 1569 struct rte_eth_desc_lim tx_desc_lim; /**< TX descriptors limits */ 1570 uint32_t speed_capa; /**< Supported speeds bitmap (ETH_LINK_SPEED_). */ 1571 /** Configured number of rx/tx queues */ 1572 uint16_t nb_rx_queues; /**< Number of RX queues. */ 1573 uint16_t nb_tx_queues; /**< Number of TX queues. */ 1574 /** Rx parameter recommendations */ 1575 struct rte_eth_dev_portconf default_rxportconf; 1576 /** Tx parameter recommendations */ 1577 struct rte_eth_dev_portconf default_txportconf; 1578 /** Generic device capabilities (RTE_ETH_DEV_CAPA_). */ 1579 uint64_t dev_capa; 1580 /** 1581 * Switching information for ports on a device with a 1582 * embedded managed interconnect/switch. 1583 */ 1584 struct rte_eth_switch_info switch_info; 1585 1586 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 1587 void *reserved_ptrs[2]; /**< Reserved for future fields */ 1588 }; 1589 1590 /** 1591 * Ethernet device RX queue information structure. 1592 * Used to retrieve information about configured queue. 1593 */ 1594 struct rte_eth_rxq_info { 1595 struct rte_mempool *mp; /**< mempool used by that queue. */ 1596 struct rte_eth_rxconf conf; /**< queue config parameters. */ 1597 uint8_t scattered_rx; /**< scattered packets RX supported. */ 1598 uint16_t nb_desc; /**< configured number of RXDs. */ 1599 uint16_t rx_buf_size; /**< hardware receive buffer size. */ 1600 } __rte_cache_min_aligned; 1601 1602 /** 1603 * Ethernet device TX queue information structure. 1604 * Used to retrieve information about configured queue. 1605 */ 1606 struct rte_eth_txq_info { 1607 struct rte_eth_txconf conf; /**< queue config parameters. */ 1608 uint16_t nb_desc; /**< configured number of TXDs. */ 1609 } __rte_cache_min_aligned; 1610 1611 /* Generic Burst mode flag definition, values can be ORed. */ 1612 1613 /** 1614 * If the queues have different burst mode description, this bit will be set 1615 * by PMD, then the application can iterate to retrieve burst description for 1616 * all other queues. 1617 */ 1618 #define RTE_ETH_BURST_FLAG_PER_QUEUE (1ULL << 0) 1619 1620 /** 1621 * Ethernet device RX/TX queue packet burst mode information structure. 1622 * Used to retrieve information about packet burst mode setting. 1623 */ 1624 struct rte_eth_burst_mode { 1625 uint64_t flags; /**< The ORed values of RTE_ETH_BURST_FLAG_xxx */ 1626 1627 #define RTE_ETH_BURST_MODE_INFO_SIZE 1024 /**< Maximum size for information */ 1628 char info[RTE_ETH_BURST_MODE_INFO_SIZE]; /**< burst mode information */ 1629 }; 1630 1631 /** Maximum name length for extended statistics counters */ 1632 #define RTE_ETH_XSTATS_NAME_SIZE 64 1633 1634 /** 1635 * An Ethernet device extended statistic structure 1636 * 1637 * This structure is used by rte_eth_xstats_get() to provide 1638 * statistics that are not provided in the generic *rte_eth_stats* 1639 * structure. 1640 * It maps a name id, corresponding to an index in the array returned 1641 * by rte_eth_xstats_get_names(), to a statistic value. 1642 */ 1643 struct rte_eth_xstat { 1644 uint64_t id; /**< The index in xstats name array. */ 1645 uint64_t value; /**< The statistic counter value. */ 1646 }; 1647 1648 /** 1649 * A name element for extended statistics. 1650 * 1651 * An array of this structure is returned by rte_eth_xstats_get_names(). 1652 * It lists the names of extended statistics for a PMD. The *rte_eth_xstat* 1653 * structure references these names by their array index. 1654 * 1655 * The xstats should follow a common naming scheme. 1656 * Some names are standardized in rte_stats_strings. 1657 * Examples: 1658 * - rx_missed_errors 1659 * - tx_q3_bytes 1660 * - tx_size_128_to_255_packets 1661 */ 1662 struct rte_eth_xstat_name { 1663 char name[RTE_ETH_XSTATS_NAME_SIZE]; /**< The statistic name. */ 1664 }; 1665 1666 #define ETH_DCB_NUM_TCS 8 1667 #define ETH_MAX_VMDQ_POOL 64 1668 1669 /** 1670 * A structure used to get the information of queue and 1671 * TC mapping on both TX and RX paths. 1672 */ 1673 struct rte_eth_dcb_tc_queue_mapping { 1674 /** rx queues assigned to tc per Pool */ 1675 struct { 1676 uint16_t base; 1677 uint16_t nb_queue; 1678 } tc_rxq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS]; 1679 /** rx queues assigned to tc per Pool */ 1680 struct { 1681 uint16_t base; 1682 uint16_t nb_queue; 1683 } tc_txq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS]; 1684 }; 1685 1686 /** 1687 * A structure used to get the information of DCB. 1688 * It includes TC UP mapping and queue TC mapping. 1689 */ 1690 struct rte_eth_dcb_info { 1691 uint8_t nb_tcs; /**< number of TCs */ 1692 uint8_t prio_tc[ETH_DCB_NUM_USER_PRIORITIES]; /**< Priority to tc */ 1693 uint8_t tc_bws[ETH_DCB_NUM_TCS]; /**< TX BW percentage for each TC */ 1694 /** rx queues assigned to tc */ 1695 struct rte_eth_dcb_tc_queue_mapping tc_queue; 1696 }; 1697 1698 /** 1699 * This enum indicates the possible Forward Error Correction (FEC) modes 1700 * of an ethdev port. 1701 */ 1702 enum rte_eth_fec_mode { 1703 RTE_ETH_FEC_NOFEC = 0, /**< FEC is off */ 1704 RTE_ETH_FEC_AUTO, /**< FEC autonegotiation modes */ 1705 RTE_ETH_FEC_BASER, /**< FEC using common algorithm */ 1706 RTE_ETH_FEC_RS, /**< FEC using RS algorithm */ 1707 }; 1708 1709 /* Translate from FEC mode to FEC capa */ 1710 #define RTE_ETH_FEC_MODE_TO_CAPA(x) (1U << (x)) 1711 1712 /* This macro indicates FEC capa mask */ 1713 #define RTE_ETH_FEC_MODE_CAPA_MASK(x) (1U << (RTE_ETH_FEC_ ## x)) 1714 1715 /* A structure used to get capabilities per link speed */ 1716 struct rte_eth_fec_capa { 1717 uint32_t speed; /**< Link speed (see ETH_SPEED_NUM_*) */ 1718 uint32_t capa; /**< FEC capabilities bitmask */ 1719 }; 1720 1721 #define RTE_ETH_ALL RTE_MAX_ETHPORTS 1722 1723 /* Macros to check for valid port */ 1724 #define RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, retval) do { \ 1725 if (!rte_eth_dev_is_valid_port(port_id)) { \ 1726 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \ 1727 return retval; \ 1728 } \ 1729 } while (0) 1730 1731 #define RTE_ETH_VALID_PORTID_OR_RET(port_id) do { \ 1732 if (!rte_eth_dev_is_valid_port(port_id)) { \ 1733 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \ 1734 return; \ 1735 } \ 1736 } while (0) 1737 1738 /** 1739 * l2 tunnel configuration. 1740 */ 1741 1742 /**< l2 tunnel enable mask */ 1743 #define ETH_L2_TUNNEL_ENABLE_MASK 0x00000001 1744 /**< l2 tunnel insertion mask */ 1745 #define ETH_L2_TUNNEL_INSERTION_MASK 0x00000002 1746 /**< l2 tunnel stripping mask */ 1747 #define ETH_L2_TUNNEL_STRIPPING_MASK 0x00000004 1748 /**< l2 tunnel forwarding mask */ 1749 #define ETH_L2_TUNNEL_FORWARDING_MASK 0x00000008 1750 1751 /** 1752 * Function type used for RX packet processing packet callbacks. 1753 * 1754 * The callback function is called on RX with a burst of packets that have 1755 * been received on the given port and queue. 1756 * 1757 * @param port_id 1758 * The Ethernet port on which RX is being performed. 1759 * @param queue 1760 * The queue on the Ethernet port which is being used to receive the packets. 1761 * @param pkts 1762 * The burst of packets that have just been received. 1763 * @param nb_pkts 1764 * The number of packets in the burst pointed to by "pkts". 1765 * @param max_pkts 1766 * The max number of packets that can be stored in the "pkts" array. 1767 * @param user_param 1768 * The arbitrary user parameter passed in by the application when the callback 1769 * was originally configured. 1770 * @return 1771 * The number of packets returned to the user. 1772 */ 1773 typedef uint16_t (*rte_rx_callback_fn)(uint16_t port_id, uint16_t queue, 1774 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts, 1775 void *user_param); 1776 1777 /** 1778 * Function type used for TX packet processing packet callbacks. 1779 * 1780 * The callback function is called on TX with a burst of packets immediately 1781 * before the packets are put onto the hardware queue for transmission. 1782 * 1783 * @param port_id 1784 * The Ethernet port on which TX is being performed. 1785 * @param queue 1786 * The queue on the Ethernet port which is being used to transmit the packets. 1787 * @param pkts 1788 * The burst of packets that are about to be transmitted. 1789 * @param nb_pkts 1790 * The number of packets in the burst pointed to by "pkts". 1791 * @param user_param 1792 * The arbitrary user parameter passed in by the application when the callback 1793 * was originally configured. 1794 * @return 1795 * The number of packets to be written to the NIC. 1796 */ 1797 typedef uint16_t (*rte_tx_callback_fn)(uint16_t port_id, uint16_t queue, 1798 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param); 1799 1800 /** 1801 * Possible states of an ethdev port. 1802 */ 1803 enum rte_eth_dev_state { 1804 /** Device is unused before being probed. */ 1805 RTE_ETH_DEV_UNUSED = 0, 1806 /** Device is attached when allocated in probing. */ 1807 RTE_ETH_DEV_ATTACHED, 1808 /** Device is in removed state when plug-out is detected. */ 1809 RTE_ETH_DEV_REMOVED, 1810 }; 1811 1812 struct rte_eth_dev_sriov { 1813 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */ 1814 uint8_t nb_q_per_pool; /**< rx queue number per pool */ 1815 uint16_t def_vmdq_idx; /**< Default pool num used for PF */ 1816 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */ 1817 }; 1818 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov) 1819 1820 #define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN 1821 1822 #define RTE_ETH_DEV_NO_OWNER 0 1823 1824 #define RTE_ETH_MAX_OWNER_NAME_LEN 64 1825 1826 struct rte_eth_dev_owner { 1827 uint64_t id; /**< The owner unique identifier. */ 1828 char name[RTE_ETH_MAX_OWNER_NAME_LEN]; /**< The owner name. */ 1829 }; 1830 1831 /** PMD supports thread-safe flow operations */ 1832 #define RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE 0x0001 1833 /** Device supports link state interrupt */ 1834 #define RTE_ETH_DEV_INTR_LSC 0x0002 1835 /** Device is a bonded slave */ 1836 #define RTE_ETH_DEV_BONDED_SLAVE 0x0004 1837 /** Device supports device removal interrupt */ 1838 #define RTE_ETH_DEV_INTR_RMV 0x0008 1839 /** Device is port representor */ 1840 #define RTE_ETH_DEV_REPRESENTOR 0x0010 1841 /** Device does not support MAC change after started */ 1842 #define RTE_ETH_DEV_NOLIVE_MAC_ADDR 0x0020 1843 /** 1844 * Queue xstats filled automatically by ethdev layer. 1845 * PMDs filling the queue xstats themselves should not set this flag 1846 */ 1847 #define RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS 0x0040 1848 1849 /** 1850 * Iterates over valid ethdev ports owned by a specific owner. 1851 * 1852 * @param port_id 1853 * The id of the next possible valid owned port. 1854 * @param owner_id 1855 * The owner identifier. 1856 * RTE_ETH_DEV_NO_OWNER means iterate over all valid ownerless ports. 1857 * @return 1858 * Next valid port id owned by owner_id, RTE_MAX_ETHPORTS if there is none. 1859 */ 1860 uint64_t rte_eth_find_next_owned_by(uint16_t port_id, 1861 const uint64_t owner_id); 1862 1863 /** 1864 * Macro to iterate over all enabled ethdev ports owned by a specific owner. 1865 */ 1866 #define RTE_ETH_FOREACH_DEV_OWNED_BY(p, o) \ 1867 for (p = rte_eth_find_next_owned_by(0, o); \ 1868 (unsigned int)p < (unsigned int)RTE_MAX_ETHPORTS; \ 1869 p = rte_eth_find_next_owned_by(p + 1, o)) 1870 1871 /** 1872 * Iterates over valid ethdev ports. 1873 * 1874 * @param port_id 1875 * The id of the next possible valid port. 1876 * @return 1877 * Next valid port id, RTE_MAX_ETHPORTS if there is none. 1878 */ 1879 uint16_t rte_eth_find_next(uint16_t port_id); 1880 1881 /** 1882 * Macro to iterate over all enabled and ownerless ethdev ports. 1883 */ 1884 #define RTE_ETH_FOREACH_DEV(p) \ 1885 RTE_ETH_FOREACH_DEV_OWNED_BY(p, RTE_ETH_DEV_NO_OWNER) 1886 1887 /** 1888 * @warning 1889 * @b EXPERIMENTAL: this API may change without prior notice. 1890 * 1891 * Iterates over ethdev ports of a specified device. 1892 * 1893 * @param port_id_start 1894 * The id of the next possible valid port. 1895 * @param parent 1896 * The generic device behind the ports to iterate. 1897 * @return 1898 * Next port id of the device, possibly port_id_start, 1899 * RTE_MAX_ETHPORTS if there is none. 1900 */ 1901 __rte_experimental 1902 uint16_t 1903 rte_eth_find_next_of(uint16_t port_id_start, 1904 const struct rte_device *parent); 1905 1906 /** 1907 * Macro to iterate over all ethdev ports of a specified device. 1908 * 1909 * @param port_id 1910 * The id of the matching port being iterated. 1911 * @param parent 1912 * The rte_device pointer matching the iterated ports. 1913 */ 1914 #define RTE_ETH_FOREACH_DEV_OF(port_id, parent) \ 1915 for (port_id = rte_eth_find_next_of(0, parent); \ 1916 port_id < RTE_MAX_ETHPORTS; \ 1917 port_id = rte_eth_find_next_of(port_id + 1, parent)) 1918 1919 /** 1920 * @warning 1921 * @b EXPERIMENTAL: this API may change without prior notice. 1922 * 1923 * Iterates over sibling ethdev ports (i.e. sharing the same rte_device). 1924 * 1925 * @param port_id_start 1926 * The id of the next possible valid sibling port. 1927 * @param ref_port_id 1928 * The id of a reference port to compare rte_device with. 1929 * @return 1930 * Next sibling port id, possibly port_id_start or ref_port_id itself, 1931 * RTE_MAX_ETHPORTS if there is none. 1932 */ 1933 __rte_experimental 1934 uint16_t 1935 rte_eth_find_next_sibling(uint16_t port_id_start, uint16_t ref_port_id); 1936 1937 /** 1938 * Macro to iterate over all ethdev ports sharing the same rte_device 1939 * as the specified port. 1940 * Note: the specified reference port is part of the loop iterations. 1941 * 1942 * @param port_id 1943 * The id of the matching port being iterated. 1944 * @param ref_port_id 1945 * The id of the port being compared. 1946 */ 1947 #define RTE_ETH_FOREACH_DEV_SIBLING(port_id, ref_port_id) \ 1948 for (port_id = rte_eth_find_next_sibling(0, ref_port_id); \ 1949 port_id < RTE_MAX_ETHPORTS; \ 1950 port_id = rte_eth_find_next_sibling(port_id + 1, ref_port_id)) 1951 1952 /** 1953 * @warning 1954 * @b EXPERIMENTAL: this API may change without prior notice. 1955 * 1956 * Get a new unique owner identifier. 1957 * An owner identifier is used to owns Ethernet devices by only one DPDK entity 1958 * to avoid multiple management of device by different entities. 1959 * 1960 * @param owner_id 1961 * Owner identifier pointer. 1962 * @return 1963 * Negative errno value on error, 0 on success. 1964 */ 1965 __rte_experimental 1966 int rte_eth_dev_owner_new(uint64_t *owner_id); 1967 1968 /** 1969 * @warning 1970 * @b EXPERIMENTAL: this API may change without prior notice. 1971 * 1972 * Set an Ethernet device owner. 1973 * 1974 * @param port_id 1975 * The identifier of the port to own. 1976 * @param owner 1977 * The owner pointer. 1978 * @return 1979 * Negative errno value on error, 0 on success. 1980 */ 1981 __rte_experimental 1982 int rte_eth_dev_owner_set(const uint16_t port_id, 1983 const struct rte_eth_dev_owner *owner); 1984 1985 /** 1986 * @warning 1987 * @b EXPERIMENTAL: this API may change without prior notice. 1988 * 1989 * Unset Ethernet device owner to make the device ownerless. 1990 * 1991 * @param port_id 1992 * The identifier of port to make ownerless. 1993 * @param owner_id 1994 * The owner identifier. 1995 * @return 1996 * 0 on success, negative errno value on error. 1997 */ 1998 __rte_experimental 1999 int rte_eth_dev_owner_unset(const uint16_t port_id, 2000 const uint64_t owner_id); 2001 2002 /** 2003 * @warning 2004 * @b EXPERIMENTAL: this API may change without prior notice. 2005 * 2006 * Remove owner from all Ethernet devices owned by a specific owner. 2007 * 2008 * @param owner_id 2009 * The owner identifier. 2010 * @return 2011 * 0 on success, negative errno value on error. 2012 */ 2013 __rte_experimental 2014 int rte_eth_dev_owner_delete(const uint64_t owner_id); 2015 2016 /** 2017 * @warning 2018 * @b EXPERIMENTAL: this API may change without prior notice. 2019 * 2020 * Get the owner of an Ethernet device. 2021 * 2022 * @param port_id 2023 * The port identifier. 2024 * @param owner 2025 * The owner structure pointer to fill. 2026 * @return 2027 * 0 on success, negative errno value on error.. 2028 */ 2029 __rte_experimental 2030 int rte_eth_dev_owner_get(const uint16_t port_id, 2031 struct rte_eth_dev_owner *owner); 2032 2033 /** 2034 * Get the number of ports which are usable for the application. 2035 * 2036 * These devices must be iterated by using the macro 2037 * ``RTE_ETH_FOREACH_DEV`` or ``RTE_ETH_FOREACH_DEV_OWNED_BY`` 2038 * to deal with non-contiguous ranges of devices. 2039 * 2040 * @return 2041 * The count of available Ethernet devices. 2042 */ 2043 uint16_t rte_eth_dev_count_avail(void); 2044 2045 /** 2046 * Get the total number of ports which are allocated. 2047 * 2048 * Some devices may not be available for the application. 2049 * 2050 * @return 2051 * The total count of Ethernet devices. 2052 */ 2053 uint16_t rte_eth_dev_count_total(void); 2054 2055 /** 2056 * Convert a numerical speed in Mbps to a bitmap flag that can be used in 2057 * the bitmap link_speeds of the struct rte_eth_conf 2058 * 2059 * @param speed 2060 * Numerical speed value in Mbps 2061 * @param duplex 2062 * ETH_LINK_[HALF/FULL]_DUPLEX (only for 10/100M speeds) 2063 * @return 2064 * 0 if the speed cannot be mapped 2065 */ 2066 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex); 2067 2068 /** 2069 * Get DEV_RX_OFFLOAD_* flag name. 2070 * 2071 * @param offload 2072 * Offload flag. 2073 * @return 2074 * Offload name or 'UNKNOWN' if the flag cannot be recognised. 2075 */ 2076 const char *rte_eth_dev_rx_offload_name(uint64_t offload); 2077 2078 /** 2079 * Get DEV_TX_OFFLOAD_* flag name. 2080 * 2081 * @param offload 2082 * Offload flag. 2083 * @return 2084 * Offload name or 'UNKNOWN' if the flag cannot be recognised. 2085 */ 2086 const char *rte_eth_dev_tx_offload_name(uint64_t offload); 2087 2088 /** 2089 * Configure an Ethernet device. 2090 * This function must be invoked first before any other function in the 2091 * Ethernet API. This function can also be re-invoked when a device is in the 2092 * stopped state. 2093 * 2094 * @param port_id 2095 * The port identifier of the Ethernet device to configure. 2096 * @param nb_rx_queue 2097 * The number of receive queues to set up for the Ethernet device. 2098 * @param nb_tx_queue 2099 * The number of transmit queues to set up for the Ethernet device. 2100 * @param eth_conf 2101 * The pointer to the configuration data to be used for the Ethernet device. 2102 * The *rte_eth_conf* structure includes: 2103 * - the hardware offload features to activate, with dedicated fields for 2104 * each statically configurable offload hardware feature provided by 2105 * Ethernet devices, such as IP checksum or VLAN tag stripping for 2106 * example. 2107 * The Rx offload bitfield API is obsolete and will be deprecated. 2108 * Applications should set the ignore_bitfield_offloads bit on *rxmode* 2109 * structure and use offloads field to set per-port offloads instead. 2110 * - Any offloading set in eth_conf->[rt]xmode.offloads must be within 2111 * the [rt]x_offload_capa returned from rte_eth_dev_info_get(). 2112 * Any type of device supported offloading set in the input argument 2113 * eth_conf->[rt]xmode.offloads to rte_eth_dev_configure() is enabled 2114 * on all queues and it can't be disabled in rte_eth_[rt]x_queue_setup() 2115 * - the Receive Side Scaling (RSS) configuration when using multiple RX 2116 * queues per port. Any RSS hash function set in eth_conf->rss_conf.rss_hf 2117 * must be within the flow_type_rss_offloads provided by drivers via 2118 * rte_eth_dev_info_get() API. 2119 * 2120 * Embedding all configuration information in a single data structure 2121 * is the more flexible method that allows the addition of new features 2122 * without changing the syntax of the API. 2123 * @return 2124 * - 0: Success, device configured. 2125 * - <0: Error code returned by the driver configuration function. 2126 */ 2127 int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue, 2128 uint16_t nb_tx_queue, const struct rte_eth_conf *eth_conf); 2129 2130 /** 2131 * @warning 2132 * @b EXPERIMENTAL: this API may change without prior notice. 2133 * 2134 * Check if an Ethernet device was physically removed. 2135 * 2136 * @param port_id 2137 * The port identifier of the Ethernet device. 2138 * @return 2139 * 1 when the Ethernet device is removed, otherwise 0. 2140 */ 2141 __rte_experimental 2142 int 2143 rte_eth_dev_is_removed(uint16_t port_id); 2144 2145 /** 2146 * Allocate and set up a receive queue for an Ethernet device. 2147 * 2148 * The function allocates a contiguous block of memory for *nb_rx_desc* 2149 * receive descriptors from a memory zone associated with *socket_id* 2150 * and initializes each receive descriptor with a network buffer allocated 2151 * from the memory pool *mb_pool*. 2152 * 2153 * @param port_id 2154 * The port identifier of the Ethernet device. 2155 * @param rx_queue_id 2156 * The index of the receive queue to set up. 2157 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2158 * to rte_eth_dev_configure(). 2159 * @param nb_rx_desc 2160 * The number of receive descriptors to allocate for the receive ring. 2161 * @param socket_id 2162 * The *socket_id* argument is the socket identifier in case of NUMA. 2163 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for 2164 * the DMA memory allocated for the receive descriptors of the ring. 2165 * @param rx_conf 2166 * The pointer to the configuration data to be used for the receive queue. 2167 * NULL value is allowed, in which case default RX configuration 2168 * will be used. 2169 * The *rx_conf* structure contains an *rx_thresh* structure with the values 2170 * of the Prefetch, Host, and Write-Back threshold registers of the receive 2171 * ring. 2172 * In addition it contains the hardware offloads features to activate using 2173 * the DEV_RX_OFFLOAD_* flags. 2174 * If an offloading set in rx_conf->offloads 2175 * hasn't been set in the input argument eth_conf->rxmode.offloads 2176 * to rte_eth_dev_configure(), it is a new added offloading, it must be 2177 * per-queue type and it is enabled for the queue. 2178 * No need to repeat any bit in rx_conf->offloads which has already been 2179 * enabled in rte_eth_dev_configure() at port level. An offloading enabled 2180 * at port level can't be disabled at queue level. 2181 * The configuration structure also contains the pointer to the array 2182 * of the receiving buffer segment descriptions, see rx_seg and rx_nseg 2183 * fields, this extended configuration might be used by split offloads like 2184 * RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mp_pool is not NULL, 2185 * the extended configuration fields must be set to NULL and zero. 2186 * @param mb_pool 2187 * The pointer to the memory pool from which to allocate *rte_mbuf* network 2188 * memory buffers to populate each descriptor of the receive ring. There are 2189 * two options to provide Rx buffer configuration: 2190 * - single pool: 2191 * mb_pool is not NULL, rx_conf.rx_nseg is 0. 2192 * - multiple segments description: 2193 * mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0. 2194 * Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads. 2195 * 2196 * @return 2197 * - 0: Success, receive queue correctly set up. 2198 * - -EIO: if device is removed. 2199 * - -ENODEV: if *port_id* is invalid. 2200 * - -EINVAL: The memory pool pointer is null or the size of network buffers 2201 * which can be allocated from this memory pool does not fit the various 2202 * buffer sizes allowed by the device controller. 2203 * - -ENOMEM: Unable to allocate the receive ring descriptors or to 2204 * allocate network memory buffers from the memory pool when 2205 * initializing receive descriptors. 2206 */ 2207 int rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id, 2208 uint16_t nb_rx_desc, unsigned int socket_id, 2209 const struct rte_eth_rxconf *rx_conf, 2210 struct rte_mempool *mb_pool); 2211 2212 /** 2213 * @warning 2214 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2215 * 2216 * Allocate and set up a hairpin receive queue for an Ethernet device. 2217 * 2218 * The function set up the selected queue to be used in hairpin. 2219 * 2220 * @param port_id 2221 * The port identifier of the Ethernet device. 2222 * @param rx_queue_id 2223 * The index of the receive queue to set up. 2224 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2225 * to rte_eth_dev_configure(). 2226 * @param nb_rx_desc 2227 * The number of receive descriptors to allocate for the receive ring. 2228 * 0 means the PMD will use default value. 2229 * @param conf 2230 * The pointer to the hairpin configuration. 2231 * 2232 * @return 2233 * - (0) if successful. 2234 * - (-ENODEV) if *port_id* is invalid. 2235 * - (-ENOTSUP) if hardware doesn't support. 2236 * - (-EINVAL) if bad parameter. 2237 * - (-ENOMEM) if unable to allocate the resources. 2238 */ 2239 __rte_experimental 2240 int rte_eth_rx_hairpin_queue_setup 2241 (uint16_t port_id, uint16_t rx_queue_id, uint16_t nb_rx_desc, 2242 const struct rte_eth_hairpin_conf *conf); 2243 2244 /** 2245 * Allocate and set up a transmit queue for an Ethernet device. 2246 * 2247 * @param port_id 2248 * The port identifier of the Ethernet device. 2249 * @param tx_queue_id 2250 * The index of the transmit queue to set up. 2251 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2252 * to rte_eth_dev_configure(). 2253 * @param nb_tx_desc 2254 * The number of transmit descriptors to allocate for the transmit ring. 2255 * @param socket_id 2256 * The *socket_id* argument is the socket identifier in case of NUMA. 2257 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for 2258 * the DMA memory allocated for the transmit descriptors of the ring. 2259 * @param tx_conf 2260 * The pointer to the configuration data to be used for the transmit queue. 2261 * NULL value is allowed, in which case default TX configuration 2262 * will be used. 2263 * The *tx_conf* structure contains the following data: 2264 * - The *tx_thresh* structure with the values of the Prefetch, Host, and 2265 * Write-Back threshold registers of the transmit ring. 2266 * When setting Write-Back threshold to the value greater then zero, 2267 * *tx_rs_thresh* value should be explicitly set to one. 2268 * - The *tx_free_thresh* value indicates the [minimum] number of network 2269 * buffers that must be pending in the transmit ring to trigger their 2270 * [implicit] freeing by the driver transmit function. 2271 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit 2272 * descriptors that must be pending in the transmit ring before setting the 2273 * RS bit on a descriptor by the driver transmit function. 2274 * The *tx_rs_thresh* value should be less or equal then 2275 * *tx_free_thresh* value, and both of them should be less then 2276 * *nb_tx_desc* - 3. 2277 * - The *offloads* member contains Tx offloads to be enabled. 2278 * If an offloading set in tx_conf->offloads 2279 * hasn't been set in the input argument eth_conf->txmode.offloads 2280 * to rte_eth_dev_configure(), it is a new added offloading, it must be 2281 * per-queue type and it is enabled for the queue. 2282 * No need to repeat any bit in tx_conf->offloads which has already been 2283 * enabled in rte_eth_dev_configure() at port level. An offloading enabled 2284 * at port level can't be disabled at queue level. 2285 * 2286 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces 2287 * the transmit function to use default values. 2288 * @return 2289 * - 0: Success, the transmit queue is correctly set up. 2290 * - -ENOMEM: Unable to allocate the transmit ring descriptors. 2291 */ 2292 int rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id, 2293 uint16_t nb_tx_desc, unsigned int socket_id, 2294 const struct rte_eth_txconf *tx_conf); 2295 2296 /** 2297 * @warning 2298 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2299 * 2300 * Allocate and set up a transmit hairpin queue for an Ethernet device. 2301 * 2302 * @param port_id 2303 * The port identifier of the Ethernet device. 2304 * @param tx_queue_id 2305 * The index of the transmit queue to set up. 2306 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2307 * to rte_eth_dev_configure(). 2308 * @param nb_tx_desc 2309 * The number of transmit descriptors to allocate for the transmit ring. 2310 * 0 to set default PMD value. 2311 * @param conf 2312 * The hairpin configuration. 2313 * 2314 * @return 2315 * - (0) if successful. 2316 * - (-ENODEV) if *port_id* is invalid. 2317 * - (-ENOTSUP) if hardware doesn't support. 2318 * - (-EINVAL) if bad parameter. 2319 * - (-ENOMEM) if unable to allocate the resources. 2320 */ 2321 __rte_experimental 2322 int rte_eth_tx_hairpin_queue_setup 2323 (uint16_t port_id, uint16_t tx_queue_id, uint16_t nb_tx_desc, 2324 const struct rte_eth_hairpin_conf *conf); 2325 2326 /** 2327 * @warning 2328 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2329 * 2330 * Get all the hairpin peer Rx / Tx ports of the current port. 2331 * The caller should ensure that the array is large enough to save the ports 2332 * list. 2333 * 2334 * @param port_id 2335 * The port identifier of the Ethernet device. 2336 * @param peer_ports 2337 * Pointer to the array to store the peer ports list. 2338 * @param len 2339 * Length of the array to store the port identifiers. 2340 * @param direction 2341 * Current port to peer port direction 2342 * positive - current used as Tx to get all peer Rx ports. 2343 * zero - current used as Rx to get all peer Tx ports. 2344 * 2345 * @return 2346 * - (0 or positive) actual peer ports number. 2347 * - (-EINVAL) if bad parameter. 2348 * - (-ENODEV) if *port_id* invalid 2349 * - (-ENOTSUP) if hardware doesn't support. 2350 * - Others detailed errors from PMD drivers. 2351 */ 2352 __rte_experimental 2353 int rte_eth_hairpin_get_peer_ports(uint16_t port_id, uint16_t *peer_ports, 2354 size_t len, uint32_t direction); 2355 2356 /** 2357 * @warning 2358 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2359 * 2360 * Bind all hairpin Tx queues of one port to the Rx queues of the peer port. 2361 * It is only allowed to call this function after all hairpin queues are 2362 * configured properly and the devices are in started state. 2363 * 2364 * @param tx_port 2365 * The identifier of the Tx port. 2366 * @param rx_port 2367 * The identifier of peer Rx port. 2368 * RTE_MAX_ETHPORTS is allowed for the traversal of all devices. 2369 * Rx port ID could have the same value as Tx port ID. 2370 * 2371 * @return 2372 * - (0) if successful. 2373 * - (-ENODEV) if Tx port ID is invalid. 2374 * - (-EBUSY) if device is not in started state. 2375 * - (-ENOTSUP) if hardware doesn't support. 2376 * - Others detailed errors from PMD drivers. 2377 */ 2378 __rte_experimental 2379 int rte_eth_hairpin_bind(uint16_t tx_port, uint16_t rx_port); 2380 2381 /** 2382 * @warning 2383 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2384 * 2385 * Unbind all hairpin Tx queues of one port from the Rx queues of the peer port. 2386 * This should be called before closing the Tx or Rx devices, if the bind 2387 * function is called before. 2388 * After unbinding the hairpin ports pair, it is allowed to bind them again. 2389 * Changing queues configuration should be after stopping the device(s). 2390 * 2391 * @param tx_port 2392 * The identifier of the Tx port. 2393 * @param rx_port 2394 * The identifier of peer Rx port. 2395 * RTE_MAX_ETHPORTS is allowed for traversal of all devices. 2396 * Rx port ID could have the same value as Tx port ID. 2397 * 2398 * @return 2399 * - (0) if successful. 2400 * - (-ENODEV) if Tx port ID is invalid. 2401 * - (-EBUSY) if device is in stopped state. 2402 * - (-ENOTSUP) if hardware doesn't support. 2403 * - Others detailed errors from PMD drivers. 2404 */ 2405 __rte_experimental 2406 int rte_eth_hairpin_unbind(uint16_t tx_port, uint16_t rx_port); 2407 2408 /** 2409 * Return the NUMA socket to which an Ethernet device is connected 2410 * 2411 * @param port_id 2412 * The port identifier of the Ethernet device 2413 * @return 2414 * The NUMA socket id to which the Ethernet device is connected or 2415 * a default of zero if the socket could not be determined. 2416 * -1 is returned is the port_id value is out of range. 2417 */ 2418 int rte_eth_dev_socket_id(uint16_t port_id); 2419 2420 /** 2421 * Check if port_id of device is attached 2422 * 2423 * @param port_id 2424 * The port identifier of the Ethernet device 2425 * @return 2426 * - 0 if port is out of range or not attached 2427 * - 1 if device is attached 2428 */ 2429 int rte_eth_dev_is_valid_port(uint16_t port_id); 2430 2431 /** 2432 * Start specified RX queue of a port. It is used when rx_deferred_start 2433 * flag of the specified queue is true. 2434 * 2435 * @param port_id 2436 * The port identifier of the Ethernet device 2437 * @param rx_queue_id 2438 * The index of the rx queue to update the ring. 2439 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2440 * to rte_eth_dev_configure(). 2441 * @return 2442 * - 0: Success, the receive queue is started. 2443 * - -ENODEV: if *port_id* is invalid. 2444 * - -EINVAL: The queue_id out of range or belong to hairpin. 2445 * - -EIO: if device is removed. 2446 * - -ENOTSUP: The function not supported in PMD driver. 2447 */ 2448 int rte_eth_dev_rx_queue_start(uint16_t port_id, uint16_t rx_queue_id); 2449 2450 /** 2451 * Stop specified RX queue of a port 2452 * 2453 * @param port_id 2454 * The port identifier of the Ethernet device 2455 * @param rx_queue_id 2456 * The index of the rx queue to update the ring. 2457 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2458 * to rte_eth_dev_configure(). 2459 * @return 2460 * - 0: Success, the receive queue is stopped. 2461 * - -ENODEV: if *port_id* is invalid. 2462 * - -EINVAL: The queue_id out of range or belong to hairpin. 2463 * - -EIO: if device is removed. 2464 * - -ENOTSUP: The function not supported in PMD driver. 2465 */ 2466 int rte_eth_dev_rx_queue_stop(uint16_t port_id, uint16_t rx_queue_id); 2467 2468 /** 2469 * Start TX for specified queue of a port. It is used when tx_deferred_start 2470 * flag of the specified queue is true. 2471 * 2472 * @param port_id 2473 * The port identifier of the Ethernet device 2474 * @param tx_queue_id 2475 * The index of the tx queue to update the ring. 2476 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2477 * to rte_eth_dev_configure(). 2478 * @return 2479 * - 0: Success, the transmit queue is started. 2480 * - -ENODEV: if *port_id* is invalid. 2481 * - -EINVAL: The queue_id out of range or belong to hairpin. 2482 * - -EIO: if device is removed. 2483 * - -ENOTSUP: The function not supported in PMD driver. 2484 */ 2485 int rte_eth_dev_tx_queue_start(uint16_t port_id, uint16_t tx_queue_id); 2486 2487 /** 2488 * Stop specified TX queue of a port 2489 * 2490 * @param port_id 2491 * The port identifier of the Ethernet device 2492 * @param tx_queue_id 2493 * The index of the tx queue to update the ring. 2494 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2495 * to rte_eth_dev_configure(). 2496 * @return 2497 * - 0: Success, the transmit queue is stopped. 2498 * - -ENODEV: if *port_id* is invalid. 2499 * - -EINVAL: The queue_id out of range or belong to hairpin. 2500 * - -EIO: if device is removed. 2501 * - -ENOTSUP: The function not supported in PMD driver. 2502 */ 2503 int rte_eth_dev_tx_queue_stop(uint16_t port_id, uint16_t tx_queue_id); 2504 2505 /** 2506 * Start an Ethernet device. 2507 * 2508 * The device start step is the last one and consists of setting the configured 2509 * offload features and in starting the transmit and the receive units of the 2510 * device. 2511 * 2512 * Device RTE_ETH_DEV_NOLIVE_MAC_ADDR flag causes MAC address to be set before 2513 * PMD port start callback function is invoked. 2514 * 2515 * On success, all basic functions exported by the Ethernet API (link status, 2516 * receive/transmit, and so on) can be invoked. 2517 * 2518 * @param port_id 2519 * The port identifier of the Ethernet device. 2520 * @return 2521 * - 0: Success, Ethernet device started. 2522 * - <0: Error code of the driver device start function. 2523 */ 2524 int rte_eth_dev_start(uint16_t port_id); 2525 2526 /** 2527 * Stop an Ethernet device. The device can be restarted with a call to 2528 * rte_eth_dev_start() 2529 * 2530 * @param port_id 2531 * The port identifier of the Ethernet device. 2532 * @return 2533 * - 0: Success, Ethernet device stopped. 2534 * - <0: Error code of the driver device stop function. 2535 */ 2536 int rte_eth_dev_stop(uint16_t port_id); 2537 2538 /** 2539 * Link up an Ethernet device. 2540 * 2541 * Set device link up will re-enable the device rx/tx 2542 * functionality after it is previously set device linked down. 2543 * 2544 * @param port_id 2545 * The port identifier of the Ethernet device. 2546 * @return 2547 * - 0: Success, Ethernet device linked up. 2548 * - <0: Error code of the driver device link up function. 2549 */ 2550 int rte_eth_dev_set_link_up(uint16_t port_id); 2551 2552 /** 2553 * Link down an Ethernet device. 2554 * The device rx/tx functionality will be disabled if success, 2555 * and it can be re-enabled with a call to 2556 * rte_eth_dev_set_link_up() 2557 * 2558 * @param port_id 2559 * The port identifier of the Ethernet device. 2560 */ 2561 int rte_eth_dev_set_link_down(uint16_t port_id); 2562 2563 /** 2564 * Close a stopped Ethernet device. The device cannot be restarted! 2565 * The function frees all port resources. 2566 * 2567 * @param port_id 2568 * The port identifier of the Ethernet device. 2569 * @return 2570 * - Zero if the port is closed successfully. 2571 * - Negative if something went wrong. 2572 */ 2573 int rte_eth_dev_close(uint16_t port_id); 2574 2575 /** 2576 * Reset a Ethernet device and keep its port id. 2577 * 2578 * When a port has to be reset passively, the DPDK application can invoke 2579 * this function. For example when a PF is reset, all its VFs should also 2580 * be reset. Normally a DPDK application can invoke this function when 2581 * RTE_ETH_EVENT_INTR_RESET event is detected, but can also use it to start 2582 * a port reset in other circumstances. 2583 * 2584 * When this function is called, it first stops the port and then calls the 2585 * PMD specific dev_uninit( ) and dev_init( ) to return the port to initial 2586 * state, in which no Tx and Rx queues are setup, as if the port has been 2587 * reset and not started. The port keeps the port id it had before the 2588 * function call. 2589 * 2590 * After calling rte_eth_dev_reset( ), the application should use 2591 * rte_eth_dev_configure( ), rte_eth_rx_queue_setup( ), 2592 * rte_eth_tx_queue_setup( ), and rte_eth_dev_start( ) 2593 * to reconfigure the device as appropriate. 2594 * 2595 * Note: To avoid unexpected behavior, the application should stop calling 2596 * Tx and Rx functions before calling rte_eth_dev_reset( ). For thread 2597 * safety, all these controlling functions should be called from the same 2598 * thread. 2599 * 2600 * @param port_id 2601 * The port identifier of the Ethernet device. 2602 * 2603 * @return 2604 * - (0) if successful. 2605 * - (-ENODEV) if *port_id* is invalid. 2606 * - (-ENOTSUP) if hardware doesn't support this function. 2607 * - (-EPERM) if not ran from the primary process. 2608 * - (-EIO) if re-initialisation failed or device is removed. 2609 * - (-ENOMEM) if the reset failed due to OOM. 2610 * - (-EAGAIN) if the reset temporarily failed and should be retried later. 2611 */ 2612 int rte_eth_dev_reset(uint16_t port_id); 2613 2614 /** 2615 * Enable receipt in promiscuous mode for an Ethernet device. 2616 * 2617 * @param port_id 2618 * The port identifier of the Ethernet device. 2619 * @return 2620 * - (0) if successful. 2621 * - (-ENOTSUP) if support for promiscuous_enable() does not exist 2622 * for the device. 2623 * - (-ENODEV) if *port_id* invalid. 2624 */ 2625 int rte_eth_promiscuous_enable(uint16_t port_id); 2626 2627 /** 2628 * Disable receipt in promiscuous mode for an Ethernet device. 2629 * 2630 * @param port_id 2631 * The port identifier of the Ethernet device. 2632 * @return 2633 * - (0) if successful. 2634 * - (-ENOTSUP) if support for promiscuous_disable() does not exist 2635 * for the device. 2636 * - (-ENODEV) if *port_id* invalid. 2637 */ 2638 int rte_eth_promiscuous_disable(uint16_t port_id); 2639 2640 /** 2641 * Return the value of promiscuous mode for an Ethernet device. 2642 * 2643 * @param port_id 2644 * The port identifier of the Ethernet device. 2645 * @return 2646 * - (1) if promiscuous is enabled 2647 * - (0) if promiscuous is disabled. 2648 * - (-1) on error 2649 */ 2650 int rte_eth_promiscuous_get(uint16_t port_id); 2651 2652 /** 2653 * Enable the receipt of any multicast frame by an Ethernet device. 2654 * 2655 * @param port_id 2656 * The port identifier of the Ethernet device. 2657 * @return 2658 * - (0) if successful. 2659 * - (-ENOTSUP) if support for allmulticast_enable() does not exist 2660 * for the device. 2661 * - (-ENODEV) if *port_id* invalid. 2662 */ 2663 int rte_eth_allmulticast_enable(uint16_t port_id); 2664 2665 /** 2666 * Disable the receipt of all multicast frames by an Ethernet device. 2667 * 2668 * @param port_id 2669 * The port identifier of the Ethernet device. 2670 * @return 2671 * - (0) if successful. 2672 * - (-ENOTSUP) if support for allmulticast_disable() does not exist 2673 * for the device. 2674 * - (-ENODEV) if *port_id* invalid. 2675 */ 2676 int rte_eth_allmulticast_disable(uint16_t port_id); 2677 2678 /** 2679 * Return the value of allmulticast mode for an Ethernet device. 2680 * 2681 * @param port_id 2682 * The port identifier of the Ethernet device. 2683 * @return 2684 * - (1) if allmulticast is enabled 2685 * - (0) if allmulticast is disabled. 2686 * - (-1) on error 2687 */ 2688 int rte_eth_allmulticast_get(uint16_t port_id); 2689 2690 /** 2691 * Retrieve the link status (up/down), the duplex mode (half/full), 2692 * the negotiation (auto/fixed), and if available, the speed (Mbps). 2693 * 2694 * It might need to wait up to 9 seconds. 2695 * @see rte_eth_link_get_nowait. 2696 * 2697 * @param port_id 2698 * The port identifier of the Ethernet device. 2699 * @param link 2700 * Link information written back. 2701 * @return 2702 * - (0) if successful. 2703 * - (-ENOTSUP) if the function is not supported in PMD driver. 2704 * - (-ENODEV) if *port_id* invalid. 2705 */ 2706 int rte_eth_link_get(uint16_t port_id, struct rte_eth_link *link); 2707 2708 /** 2709 * Retrieve the link status (up/down), the duplex mode (half/full), 2710 * the negotiation (auto/fixed), and if available, the speed (Mbps). 2711 * 2712 * @param port_id 2713 * The port identifier of the Ethernet device. 2714 * @param link 2715 * Link information written back. 2716 * @return 2717 * - (0) if successful. 2718 * - (-ENOTSUP) if the function is not supported in PMD driver. 2719 * - (-ENODEV) if *port_id* invalid. 2720 */ 2721 int rte_eth_link_get_nowait(uint16_t port_id, struct rte_eth_link *link); 2722 2723 /** 2724 * @warning 2725 * @b EXPERIMENTAL: this API may change without prior notice. 2726 * 2727 * The function converts a link_speed to a string. It handles all special 2728 * values like unknown or none speed. 2729 * 2730 * @param link_speed 2731 * link_speed of rte_eth_link struct 2732 * @return 2733 * Link speed in textual format. It's pointer to immutable memory. 2734 * No free is required. 2735 */ 2736 __rte_experimental 2737 const char *rte_eth_link_speed_to_str(uint32_t link_speed); 2738 2739 /** 2740 * @warning 2741 * @b EXPERIMENTAL: this API may change without prior notice. 2742 * 2743 * The function converts a rte_eth_link struct representing a link status to 2744 * a string. 2745 * 2746 * @param str 2747 * A pointer to a string to be filled with textual representation of 2748 * device status. At least ETH_LINK_MAX_STR_LEN bytes should be allocated to 2749 * store default link status text. 2750 * @param len 2751 * Length of available memory at 'str' string. 2752 * @param eth_link 2753 * Link status returned by rte_eth_link_get function 2754 * @return 2755 * Number of bytes written to str array. 2756 */ 2757 __rte_experimental 2758 int rte_eth_link_to_str(char *str, size_t len, 2759 const struct rte_eth_link *eth_link); 2760 2761 /** 2762 * Retrieve the general I/O statistics of an Ethernet device. 2763 * 2764 * @param port_id 2765 * The port identifier of the Ethernet device. 2766 * @param stats 2767 * A pointer to a structure of type *rte_eth_stats* to be filled with 2768 * the values of device counters for the following set of statistics: 2769 * - *ipackets* with the total of successfully received packets. 2770 * - *opackets* with the total of successfully transmitted packets. 2771 * - *ibytes* with the total of successfully received bytes. 2772 * - *obytes* with the total of successfully transmitted bytes. 2773 * - *ierrors* with the total of erroneous received packets. 2774 * - *oerrors* with the total of failed transmitted packets. 2775 * @return 2776 * Zero if successful. Non-zero otherwise. 2777 */ 2778 int rte_eth_stats_get(uint16_t port_id, struct rte_eth_stats *stats); 2779 2780 /** 2781 * Reset the general I/O statistics of an Ethernet device. 2782 * 2783 * @param port_id 2784 * The port identifier of the Ethernet device. 2785 * @return 2786 * - (0) if device notified to reset stats. 2787 * - (-ENOTSUP) if hardware doesn't support. 2788 * - (-ENODEV) if *port_id* invalid. 2789 * - (<0): Error code of the driver stats reset function. 2790 */ 2791 int rte_eth_stats_reset(uint16_t port_id); 2792 2793 /** 2794 * Retrieve names of extended statistics of an Ethernet device. 2795 * 2796 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched 2797 * by array index: 2798 * xstats_names[i].name => xstats[i].value 2799 * 2800 * And the array index is same with id field of 'struct rte_eth_xstat': 2801 * xstats[i].id == i 2802 * 2803 * This assumption makes key-value pair matching less flexible but simpler. 2804 * 2805 * @param port_id 2806 * The port identifier of the Ethernet device. 2807 * @param xstats_names 2808 * An rte_eth_xstat_name array of at least *size* elements to 2809 * be filled. If set to NULL, the function returns the required number 2810 * of elements. 2811 * @param size 2812 * The size of the xstats_names array (number of elements). 2813 * @return 2814 * - A positive value lower or equal to size: success. The return value 2815 * is the number of entries filled in the stats table. 2816 * - A positive value higher than size: error, the given statistics table 2817 * is too small. The return value corresponds to the size that should 2818 * be given to succeed. The entries in the table are not valid and 2819 * shall not be used by the caller. 2820 * - A negative value on error (invalid port id). 2821 */ 2822 int rte_eth_xstats_get_names(uint16_t port_id, 2823 struct rte_eth_xstat_name *xstats_names, 2824 unsigned int size); 2825 2826 /** 2827 * Retrieve extended statistics of an Ethernet device. 2828 * 2829 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched 2830 * by array index: 2831 * xstats_names[i].name => xstats[i].value 2832 * 2833 * And the array index is same with id field of 'struct rte_eth_xstat': 2834 * xstats[i].id == i 2835 * 2836 * This assumption makes key-value pair matching less flexible but simpler. 2837 * 2838 * @param port_id 2839 * The port identifier of the Ethernet device. 2840 * @param xstats 2841 * A pointer to a table of structure of type *rte_eth_xstat* 2842 * to be filled with device statistics ids and values. 2843 * This parameter can be set to NULL if n is 0. 2844 * @param n 2845 * The size of the xstats array (number of elements). 2846 * @return 2847 * - A positive value lower or equal to n: success. The return value 2848 * is the number of entries filled in the stats table. 2849 * - A positive value higher than n: error, the given statistics table 2850 * is too small. The return value corresponds to the size that should 2851 * be given to succeed. The entries in the table are not valid and 2852 * shall not be used by the caller. 2853 * - A negative value on error (invalid port id). 2854 */ 2855 int rte_eth_xstats_get(uint16_t port_id, struct rte_eth_xstat *xstats, 2856 unsigned int n); 2857 2858 /** 2859 * Retrieve names of extended statistics of an Ethernet device. 2860 * 2861 * @param port_id 2862 * The port identifier of the Ethernet device. 2863 * @param xstats_names 2864 * An rte_eth_xstat_name array of at least *size* elements to 2865 * be filled. If set to NULL, the function returns the required number 2866 * of elements. 2867 * @param ids 2868 * IDs array given by app to retrieve specific statistics 2869 * @param size 2870 * The size of the xstats_names array (number of elements). 2871 * @return 2872 * - A positive value lower or equal to size: success. The return value 2873 * is the number of entries filled in the stats table. 2874 * - A positive value higher than size: error, the given statistics table 2875 * is too small. The return value corresponds to the size that should 2876 * be given to succeed. The entries in the table are not valid and 2877 * shall not be used by the caller. 2878 * - A negative value on error (invalid port id). 2879 */ 2880 int 2881 rte_eth_xstats_get_names_by_id(uint16_t port_id, 2882 struct rte_eth_xstat_name *xstats_names, unsigned int size, 2883 uint64_t *ids); 2884 2885 /** 2886 * Retrieve extended statistics of an Ethernet device. 2887 * 2888 * @param port_id 2889 * The port identifier of the Ethernet device. 2890 * @param ids 2891 * A pointer to an ids array passed by application. This tells which 2892 * statistics values function should retrieve. This parameter 2893 * can be set to NULL if size is 0. In this case function will retrieve 2894 * all available statistics. 2895 * @param values 2896 * A pointer to a table to be filled with device statistics values. 2897 * @param size 2898 * The size of the ids array (number of elements). 2899 * @return 2900 * - A positive value lower or equal to size: success. The return value 2901 * is the number of entries filled in the stats table. 2902 * - A positive value higher than size: error, the given statistics table 2903 * is too small. The return value corresponds to the size that should 2904 * be given to succeed. The entries in the table are not valid and 2905 * shall not be used by the caller. 2906 * - A negative value on error (invalid port id). 2907 */ 2908 int rte_eth_xstats_get_by_id(uint16_t port_id, const uint64_t *ids, 2909 uint64_t *values, unsigned int size); 2910 2911 /** 2912 * Gets the ID of a statistic from its name. 2913 * 2914 * This function searches for the statistics using string compares, and 2915 * as such should not be used on the fast-path. For fast-path retrieval of 2916 * specific statistics, store the ID as provided in *id* from this function, 2917 * and pass the ID to rte_eth_xstats_get() 2918 * 2919 * @param port_id The port to look up statistics from 2920 * @param xstat_name The name of the statistic to return 2921 * @param[out] id A pointer to an app-supplied uint64_t which should be 2922 * set to the ID of the stat if the stat exists. 2923 * @return 2924 * 0 on success 2925 * -ENODEV for invalid port_id, 2926 * -EIO if device is removed, 2927 * -EINVAL if the xstat_name doesn't exist in port_id 2928 */ 2929 int rte_eth_xstats_get_id_by_name(uint16_t port_id, const char *xstat_name, 2930 uint64_t *id); 2931 2932 /** 2933 * Reset extended statistics of an Ethernet device. 2934 * 2935 * @param port_id 2936 * The port identifier of the Ethernet device. 2937 * @return 2938 * - (0) if device notified to reset extended stats. 2939 * - (-ENOTSUP) if pmd doesn't support both 2940 * extended stats and basic stats reset. 2941 * - (-ENODEV) if *port_id* invalid. 2942 * - (<0): Error code of the driver xstats reset function. 2943 */ 2944 int rte_eth_xstats_reset(uint16_t port_id); 2945 2946 /** 2947 * Set a mapping for the specified transmit queue to the specified per-queue 2948 * statistics counter. 2949 * 2950 * @param port_id 2951 * The port identifier of the Ethernet device. 2952 * @param tx_queue_id 2953 * The index of the transmit queue for which a queue stats mapping is required. 2954 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2955 * to rte_eth_dev_configure(). 2956 * @param stat_idx 2957 * The per-queue packet statistics functionality number that the transmit 2958 * queue is to be assigned. 2959 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1]. 2960 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256. 2961 * @return 2962 * Zero if successful. Non-zero otherwise. 2963 */ 2964 int rte_eth_dev_set_tx_queue_stats_mapping(uint16_t port_id, 2965 uint16_t tx_queue_id, uint8_t stat_idx); 2966 2967 /** 2968 * Set a mapping for the specified receive queue to the specified per-queue 2969 * statistics counter. 2970 * 2971 * @param port_id 2972 * The port identifier of the Ethernet device. 2973 * @param rx_queue_id 2974 * The index of the receive queue for which a queue stats mapping is required. 2975 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2976 * to rte_eth_dev_configure(). 2977 * @param stat_idx 2978 * The per-queue packet statistics functionality number that the receive 2979 * queue is to be assigned. 2980 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1]. 2981 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256. 2982 * @return 2983 * Zero if successful. Non-zero otherwise. 2984 */ 2985 int rte_eth_dev_set_rx_queue_stats_mapping(uint16_t port_id, 2986 uint16_t rx_queue_id, 2987 uint8_t stat_idx); 2988 2989 /** 2990 * Retrieve the Ethernet address of an Ethernet device. 2991 * 2992 * @param port_id 2993 * The port identifier of the Ethernet device. 2994 * @param mac_addr 2995 * A pointer to a structure of type *ether_addr* to be filled with 2996 * the Ethernet address of the Ethernet device. 2997 * @return 2998 * - (0) if successful 2999 * - (-ENODEV) if *port_id* invalid. 3000 */ 3001 int rte_eth_macaddr_get(uint16_t port_id, struct rte_ether_addr *mac_addr); 3002 3003 /** 3004 * Retrieve the contextual information of an Ethernet device. 3005 * 3006 * As part of this function, a number of of fields in dev_info will be 3007 * initialized as follows: 3008 * 3009 * rx_desc_lim = lim 3010 * tx_desc_lim = lim 3011 * 3012 * Where lim is defined within the rte_eth_dev_info_get as 3013 * 3014 * const struct rte_eth_desc_lim lim = { 3015 * .nb_max = UINT16_MAX, 3016 * .nb_min = 0, 3017 * .nb_align = 1, 3018 * .nb_seg_max = UINT16_MAX, 3019 * .nb_mtu_seg_max = UINT16_MAX, 3020 * }; 3021 * 3022 * device = dev->device 3023 * min_mtu = RTE_ETHER_MIN_MTU 3024 * max_mtu = UINT16_MAX 3025 * 3026 * The following fields will be populated if support for dev_infos_get() 3027 * exists for the device and the rte_eth_dev 'dev' has been populated 3028 * successfully with a call to it: 3029 * 3030 * driver_name = dev->device->driver->name 3031 * nb_rx_queues = dev->data->nb_rx_queues 3032 * nb_tx_queues = dev->data->nb_tx_queues 3033 * dev_flags = &dev->data->dev_flags 3034 * 3035 * @param port_id 3036 * The port identifier of the Ethernet device. 3037 * @param dev_info 3038 * A pointer to a structure of type *rte_eth_dev_info* to be filled with 3039 * the contextual information of the Ethernet device. 3040 * @return 3041 * - (0) if successful. 3042 * - (-ENOTSUP) if support for dev_infos_get() does not exist for the device. 3043 * - (-ENODEV) if *port_id* invalid. 3044 */ 3045 int rte_eth_dev_info_get(uint16_t port_id, struct rte_eth_dev_info *dev_info); 3046 3047 /** 3048 * Retrieve the firmware version of a device. 3049 * 3050 * @param port_id 3051 * The port identifier of the device. 3052 * @param fw_version 3053 * A pointer to a string array storing the firmware version of a device, 3054 * the string includes terminating null. This pointer is allocated by caller. 3055 * @param fw_size 3056 * The size of the string array pointed by fw_version, which should be 3057 * large enough to store firmware version of the device. 3058 * @return 3059 * - (0) if successful. 3060 * - (-ENOTSUP) if operation is not supported. 3061 * - (-ENODEV) if *port_id* invalid. 3062 * - (-EIO) if device is removed. 3063 * - (>0) if *fw_size* is not enough to store firmware version, return 3064 * the size of the non truncated string. 3065 */ 3066 int rte_eth_dev_fw_version_get(uint16_t port_id, 3067 char *fw_version, size_t fw_size); 3068 3069 /** 3070 * Retrieve the supported packet types of an Ethernet device. 3071 * 3072 * When a packet type is announced as supported, it *must* be recognized by 3073 * the PMD. For instance, if RTE_PTYPE_L2_ETHER, RTE_PTYPE_L2_ETHER_VLAN 3074 * and RTE_PTYPE_L3_IPV4 are announced, the PMD must return the following 3075 * packet types for these packets: 3076 * - Ether/IPv4 -> RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4 3077 * - Ether/Vlan/IPv4 -> RTE_PTYPE_L2_ETHER_VLAN | RTE_PTYPE_L3_IPV4 3078 * - Ether/[anything else] -> RTE_PTYPE_L2_ETHER 3079 * - Ether/Vlan/[anything else] -> RTE_PTYPE_L2_ETHER_VLAN 3080 * 3081 * When a packet is received by a PMD, the most precise type must be 3082 * returned among the ones supported. However a PMD is allowed to set 3083 * packet type that is not in the supported list, at the condition that it 3084 * is more precise. Therefore, a PMD announcing no supported packet types 3085 * can still set a matching packet type in a received packet. 3086 * 3087 * @note 3088 * Better to invoke this API after the device is already started or rx burst 3089 * function is decided, to obtain correct supported ptypes. 3090 * @note 3091 * if a given PMD does not report what ptypes it supports, then the supported 3092 * ptype count is reported as 0. 3093 * @param port_id 3094 * The port identifier of the Ethernet device. 3095 * @param ptype_mask 3096 * A hint of what kind of packet type which the caller is interested in. 3097 * @param ptypes 3098 * An array pointer to store adequate packet types, allocated by caller. 3099 * @param num 3100 * Size of the array pointed by param ptypes. 3101 * @return 3102 * - (>=0) Number of supported ptypes. If the number of types exceeds num, 3103 * only num entries will be filled into the ptypes array, but the full 3104 * count of supported ptypes will be returned. 3105 * - (-ENODEV) if *port_id* invalid. 3106 */ 3107 int rte_eth_dev_get_supported_ptypes(uint16_t port_id, uint32_t ptype_mask, 3108 uint32_t *ptypes, int num); 3109 /** 3110 * @warning 3111 * @b EXPERIMENTAL: this API may change without prior notice. 3112 * 3113 * Inform Ethernet device about reduced range of packet types to handle. 3114 * 3115 * Application can use this function to set only specific ptypes that it's 3116 * interested. This information can be used by the PMD to optimize Rx path. 3117 * 3118 * The function accepts an array `set_ptypes` allocated by the caller to 3119 * store the packet types set by the driver, the last element of the array 3120 * is set to RTE_PTYPE_UNKNOWN. The size of the `set_ptype` array should be 3121 * `rte_eth_dev_get_supported_ptypes() + 1` else it might only be filled 3122 * partially. 3123 * 3124 * @param port_id 3125 * The port identifier of the Ethernet device. 3126 * @param ptype_mask 3127 * The ptype family that application is interested in should be bitwise OR of 3128 * RTE_PTYPE_*_MASK or 0. 3129 * @param set_ptypes 3130 * An array pointer to store set packet types, allocated by caller. The 3131 * function marks the end of array with RTE_PTYPE_UNKNOWN. 3132 * @param num 3133 * Size of the array pointed by param ptypes. 3134 * Should be rte_eth_dev_get_supported_ptypes() + 1 to accommodate the 3135 * set ptypes. 3136 * @return 3137 * - (0) if Success. 3138 * - (-ENODEV) if *port_id* invalid. 3139 * - (-EINVAL) if *ptype_mask* is invalid (or) set_ptypes is NULL and 3140 * num > 0. 3141 */ 3142 __rte_experimental 3143 int rte_eth_dev_set_ptypes(uint16_t port_id, uint32_t ptype_mask, 3144 uint32_t *set_ptypes, unsigned int num); 3145 3146 /** 3147 * Retrieve the MTU of an Ethernet device. 3148 * 3149 * @param port_id 3150 * The port identifier of the Ethernet device. 3151 * @param mtu 3152 * A pointer to a uint16_t where the retrieved MTU is to be stored. 3153 * @return 3154 * - (0) if successful. 3155 * - (-ENODEV) if *port_id* invalid. 3156 */ 3157 int rte_eth_dev_get_mtu(uint16_t port_id, uint16_t *mtu); 3158 3159 /** 3160 * Change the MTU of an Ethernet device. 3161 * 3162 * @param port_id 3163 * The port identifier of the Ethernet device. 3164 * @param mtu 3165 * A uint16_t for the MTU to be applied. 3166 * @return 3167 * - (0) if successful. 3168 * - (-ENOTSUP) if operation is not supported. 3169 * - (-ENODEV) if *port_id* invalid. 3170 * - (-EIO) if device is removed. 3171 * - (-EINVAL) if *mtu* invalid, validation of mtu can occur within 3172 * rte_eth_dev_set_mtu if dev_infos_get is supported by the device or 3173 * when the mtu is set using dev->dev_ops->mtu_set. 3174 * - (-EBUSY) if operation is not allowed when the port is running 3175 */ 3176 int rte_eth_dev_set_mtu(uint16_t port_id, uint16_t mtu); 3177 3178 /** 3179 * Enable/Disable hardware filtering by an Ethernet device of received 3180 * VLAN packets tagged with a given VLAN Tag Identifier. 3181 * 3182 * @param port_id 3183 * The port identifier of the Ethernet device. 3184 * @param vlan_id 3185 * The VLAN Tag Identifier whose filtering must be enabled or disabled. 3186 * @param on 3187 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*. 3188 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*. 3189 * @return 3190 * - (0) if successful. 3191 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured. 3192 * - (-ENODEV) if *port_id* invalid. 3193 * - (-EIO) if device is removed. 3194 * - (-ENOSYS) if VLAN filtering on *port_id* disabled. 3195 * - (-EINVAL) if *vlan_id* > 4095. 3196 */ 3197 int rte_eth_dev_vlan_filter(uint16_t port_id, uint16_t vlan_id, int on); 3198 3199 /** 3200 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device. 3201 * 3202 * @param port_id 3203 * The port identifier of the Ethernet device. 3204 * @param rx_queue_id 3205 * The index of the receive queue for which a queue stats mapping is required. 3206 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3207 * to rte_eth_dev_configure(). 3208 * @param on 3209 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port. 3210 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port. 3211 * @return 3212 * - (0) if successful. 3213 * - (-ENOTSUP) if hardware-assisted VLAN stripping not configured. 3214 * - (-ENODEV) if *port_id* invalid. 3215 * - (-EINVAL) if *rx_queue_id* invalid. 3216 */ 3217 int rte_eth_dev_set_vlan_strip_on_queue(uint16_t port_id, uint16_t rx_queue_id, 3218 int on); 3219 3220 /** 3221 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to 3222 * the VLAN header. 3223 * 3224 * @param port_id 3225 * The port identifier of the Ethernet device. 3226 * @param vlan_type 3227 * The vlan type. 3228 * @param tag_type 3229 * The Tag Protocol ID 3230 * @return 3231 * - (0) if successful. 3232 * - (-ENOTSUP) if hardware-assisted VLAN TPID setup is not supported. 3233 * - (-ENODEV) if *port_id* invalid. 3234 * - (-EIO) if device is removed. 3235 */ 3236 int rte_eth_dev_set_vlan_ether_type(uint16_t port_id, 3237 enum rte_vlan_type vlan_type, 3238 uint16_t tag_type); 3239 3240 /** 3241 * Set VLAN offload configuration on an Ethernet device. 3242 * 3243 * @param port_id 3244 * The port identifier of the Ethernet device. 3245 * @param offload_mask 3246 * The VLAN Offload bit mask can be mixed use with "OR" 3247 * ETH_VLAN_STRIP_OFFLOAD 3248 * ETH_VLAN_FILTER_OFFLOAD 3249 * ETH_VLAN_EXTEND_OFFLOAD 3250 * ETH_QINQ_STRIP_OFFLOAD 3251 * @return 3252 * - (0) if successful. 3253 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured. 3254 * - (-ENODEV) if *port_id* invalid. 3255 * - (-EIO) if device is removed. 3256 */ 3257 int rte_eth_dev_set_vlan_offload(uint16_t port_id, int offload_mask); 3258 3259 /** 3260 * Read VLAN Offload configuration from an Ethernet device 3261 * 3262 * @param port_id 3263 * The port identifier of the Ethernet device. 3264 * @return 3265 * - (>0) if successful. Bit mask to indicate 3266 * ETH_VLAN_STRIP_OFFLOAD 3267 * ETH_VLAN_FILTER_OFFLOAD 3268 * ETH_VLAN_EXTEND_OFFLOAD 3269 * ETH_QINQ_STRIP_OFFLOAD 3270 * - (-ENODEV) if *port_id* invalid. 3271 */ 3272 int rte_eth_dev_get_vlan_offload(uint16_t port_id); 3273 3274 /** 3275 * Set port based TX VLAN insertion on or off. 3276 * 3277 * @param port_id 3278 * The port identifier of the Ethernet device. 3279 * @param pvid 3280 * Port based TX VLAN identifier together with user priority. 3281 * @param on 3282 * Turn on or off the port based TX VLAN insertion. 3283 * 3284 * @return 3285 * - (0) if successful. 3286 * - negative if failed. 3287 */ 3288 int rte_eth_dev_set_vlan_pvid(uint16_t port_id, uint16_t pvid, int on); 3289 3290 typedef void (*buffer_tx_error_fn)(struct rte_mbuf **unsent, uint16_t count, 3291 void *userdata); 3292 3293 /** 3294 * Structure used to buffer packets for future TX 3295 * Used by APIs rte_eth_tx_buffer and rte_eth_tx_buffer_flush 3296 */ 3297 struct rte_eth_dev_tx_buffer { 3298 buffer_tx_error_fn error_callback; 3299 void *error_userdata; 3300 uint16_t size; /**< Size of buffer for buffered tx */ 3301 uint16_t length; /**< Number of packets in the array */ 3302 struct rte_mbuf *pkts[]; 3303 /**< Pending packets to be sent on explicit flush or when full */ 3304 }; 3305 3306 /** 3307 * Calculate the size of the tx buffer. 3308 * 3309 * @param sz 3310 * Number of stored packets. 3311 */ 3312 #define RTE_ETH_TX_BUFFER_SIZE(sz) \ 3313 (sizeof(struct rte_eth_dev_tx_buffer) + (sz) * sizeof(struct rte_mbuf *)) 3314 3315 /** 3316 * Initialize default values for buffered transmitting 3317 * 3318 * @param buffer 3319 * Tx buffer to be initialized. 3320 * @param size 3321 * Buffer size 3322 * @return 3323 * 0 if no error 3324 */ 3325 int 3326 rte_eth_tx_buffer_init(struct rte_eth_dev_tx_buffer *buffer, uint16_t size); 3327 3328 /** 3329 * Configure a callback for buffered packets which cannot be sent 3330 * 3331 * Register a specific callback to be called when an attempt is made to send 3332 * all packets buffered on an ethernet port, but not all packets can 3333 * successfully be sent. The callback registered here will be called only 3334 * from calls to rte_eth_tx_buffer() and rte_eth_tx_buffer_flush() APIs. 3335 * The default callback configured for each queue by default just frees the 3336 * packets back to the calling mempool. If additional behaviour is required, 3337 * for example, to count dropped packets, or to retry transmission of packets 3338 * which cannot be sent, this function should be used to register a suitable 3339 * callback function to implement the desired behaviour. 3340 * The example callback "rte_eth_count_unsent_packet_callback()" is also 3341 * provided as reference. 3342 * 3343 * @param buffer 3344 * The port identifier of the Ethernet device. 3345 * @param callback 3346 * The function to be used as the callback. 3347 * @param userdata 3348 * Arbitrary parameter to be passed to the callback function 3349 * @return 3350 * 0 on success, or -1 on error with rte_errno set appropriately 3351 */ 3352 int 3353 rte_eth_tx_buffer_set_err_callback(struct rte_eth_dev_tx_buffer *buffer, 3354 buffer_tx_error_fn callback, void *userdata); 3355 3356 /** 3357 * Callback function for silently dropping unsent buffered packets. 3358 * 3359 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to 3360 * adjust the default behavior when buffered packets cannot be sent. This 3361 * function drops any unsent packets silently and is used by tx buffered 3362 * operations as default behavior. 3363 * 3364 * NOTE: this function should not be called directly, instead it should be used 3365 * as a callback for packet buffering. 3366 * 3367 * NOTE: when configuring this function as a callback with 3368 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter 3369 * should point to an uint64_t value. 3370 * 3371 * @param pkts 3372 * The previously buffered packets which could not be sent 3373 * @param unsent 3374 * The number of unsent packets in the pkts array 3375 * @param userdata 3376 * Not used 3377 */ 3378 void 3379 rte_eth_tx_buffer_drop_callback(struct rte_mbuf **pkts, uint16_t unsent, 3380 void *userdata); 3381 3382 /** 3383 * Callback function for tracking unsent buffered packets. 3384 * 3385 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to 3386 * adjust the default behavior when buffered packets cannot be sent. This 3387 * function drops any unsent packets, but also updates a user-supplied counter 3388 * to track the overall number of packets dropped. The counter should be an 3389 * uint64_t variable. 3390 * 3391 * NOTE: this function should not be called directly, instead it should be used 3392 * as a callback for packet buffering. 3393 * 3394 * NOTE: when configuring this function as a callback with 3395 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter 3396 * should point to an uint64_t value. 3397 * 3398 * @param pkts 3399 * The previously buffered packets which could not be sent 3400 * @param unsent 3401 * The number of unsent packets in the pkts array 3402 * @param userdata 3403 * Pointer to an uint64_t value, which will be incremented by unsent 3404 */ 3405 void 3406 rte_eth_tx_buffer_count_callback(struct rte_mbuf **pkts, uint16_t unsent, 3407 void *userdata); 3408 3409 /** 3410 * Request the driver to free mbufs currently cached by the driver. The 3411 * driver will only free the mbuf if it is no longer in use. It is the 3412 * application's responsibility to ensure rte_eth_tx_buffer_flush(..) is 3413 * called if needed. 3414 * 3415 * @param port_id 3416 * The port identifier of the Ethernet device. 3417 * @param queue_id 3418 * The index of the transmit queue through which output packets must be 3419 * sent. 3420 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 3421 * to rte_eth_dev_configure(). 3422 * @param free_cnt 3423 * Maximum number of packets to free. Use 0 to indicate all possible packets 3424 * should be freed. Note that a packet may be using multiple mbufs. 3425 * @return 3426 * Failure: < 0 3427 * -ENODEV: Invalid interface 3428 * -EIO: device is removed 3429 * -ENOTSUP: Driver does not support function 3430 * Success: >= 0 3431 * 0-n: Number of packets freed. More packets may still remain in ring that 3432 * are in use. 3433 */ 3434 int 3435 rte_eth_tx_done_cleanup(uint16_t port_id, uint16_t queue_id, uint32_t free_cnt); 3436 3437 /** 3438 * Subtypes for IPsec offload event(@ref RTE_ETH_EVENT_IPSEC) raised by 3439 * eth device. 3440 */ 3441 enum rte_eth_event_ipsec_subtype { 3442 RTE_ETH_EVENT_IPSEC_UNKNOWN = 0, 3443 /**< Unknown event type */ 3444 RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW, 3445 /**< Sequence number overflow */ 3446 RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY, 3447 /**< Soft time expiry of SA */ 3448 RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY, 3449 /**< Soft byte expiry of SA */ 3450 RTE_ETH_EVENT_IPSEC_MAX 3451 /**< Max value of this enum */ 3452 }; 3453 3454 /** 3455 * Descriptor for @ref RTE_ETH_EVENT_IPSEC event. Used by eth dev to send extra 3456 * information of the IPsec offload event. 3457 */ 3458 struct rte_eth_event_ipsec_desc { 3459 enum rte_eth_event_ipsec_subtype subtype; 3460 /**< Type of RTE_ETH_EVENT_IPSEC_* event */ 3461 uint64_t metadata; 3462 /**< Event specific metadata 3463 * 3464 * For the following events, *userdata* registered 3465 * with the *rte_security_session* would be returned 3466 * as metadata, 3467 * 3468 * - @ref RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW 3469 * - @ref RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY 3470 * - @ref RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY 3471 * 3472 * @see struct rte_security_session_conf 3473 * 3474 */ 3475 }; 3476 3477 /** 3478 * The eth device event type for interrupt, and maybe others in the future. 3479 */ 3480 enum rte_eth_event_type { 3481 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */ 3482 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */ 3483 RTE_ETH_EVENT_QUEUE_STATE, 3484 /**< queue state event (enabled/disabled) */ 3485 RTE_ETH_EVENT_INTR_RESET, 3486 /**< reset interrupt event, sent to VF on PF reset */ 3487 RTE_ETH_EVENT_VF_MBOX, /**< message from the VF received by PF */ 3488 RTE_ETH_EVENT_MACSEC, /**< MACsec offload related event */ 3489 RTE_ETH_EVENT_INTR_RMV, /**< device removal event */ 3490 RTE_ETH_EVENT_NEW, /**< port is probed */ 3491 RTE_ETH_EVENT_DESTROY, /**< port is released */ 3492 RTE_ETH_EVENT_IPSEC, /**< IPsec offload related event */ 3493 RTE_ETH_EVENT_FLOW_AGED,/**< New aged-out flows is detected */ 3494 RTE_ETH_EVENT_MAX /**< max value of this enum */ 3495 }; 3496 3497 typedef int (*rte_eth_dev_cb_fn)(uint16_t port_id, 3498 enum rte_eth_event_type event, void *cb_arg, void *ret_param); 3499 /**< user application callback to be registered for interrupts */ 3500 3501 /** 3502 * Register a callback function for port event. 3503 * 3504 * @param port_id 3505 * Port id. 3506 * RTE_ETH_ALL means register the event for all port ids. 3507 * @param event 3508 * Event interested. 3509 * @param cb_fn 3510 * User supplied callback function to be called. 3511 * @param cb_arg 3512 * Pointer to the parameters for the registered callback. 3513 * 3514 * @return 3515 * - On success, zero. 3516 * - On failure, a negative value. 3517 */ 3518 int rte_eth_dev_callback_register(uint16_t port_id, 3519 enum rte_eth_event_type event, 3520 rte_eth_dev_cb_fn cb_fn, void *cb_arg); 3521 3522 /** 3523 * Unregister a callback function for port event. 3524 * 3525 * @param port_id 3526 * Port id. 3527 * RTE_ETH_ALL means unregister the event for all port ids. 3528 * @param event 3529 * Event interested. 3530 * @param cb_fn 3531 * User supplied callback function to be called. 3532 * @param cb_arg 3533 * Pointer to the parameters for the registered callback. -1 means to 3534 * remove all for the same callback address and same event. 3535 * 3536 * @return 3537 * - On success, zero. 3538 * - On failure, a negative value. 3539 */ 3540 int rte_eth_dev_callback_unregister(uint16_t port_id, 3541 enum rte_eth_event_type event, 3542 rte_eth_dev_cb_fn cb_fn, void *cb_arg); 3543 3544 /** 3545 * When there is no rx packet coming in Rx Queue for a long time, we can 3546 * sleep lcore related to RX Queue for power saving, and enable rx interrupt 3547 * to be triggered when Rx packet arrives. 3548 * 3549 * The rte_eth_dev_rx_intr_enable() function enables rx queue 3550 * interrupt on specific rx queue of a port. 3551 * 3552 * @param port_id 3553 * The port identifier of the Ethernet device. 3554 * @param queue_id 3555 * The index of the receive queue from which to retrieve input packets. 3556 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3557 * to rte_eth_dev_configure(). 3558 * @return 3559 * - (0) if successful. 3560 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3561 * that operation. 3562 * - (-ENODEV) if *port_id* invalid. 3563 * - (-EIO) if device is removed. 3564 */ 3565 int rte_eth_dev_rx_intr_enable(uint16_t port_id, uint16_t queue_id); 3566 3567 /** 3568 * When lcore wakes up from rx interrupt indicating packet coming, disable rx 3569 * interrupt and returns to polling mode. 3570 * 3571 * The rte_eth_dev_rx_intr_disable() function disables rx queue 3572 * interrupt on specific rx queue of a port. 3573 * 3574 * @param port_id 3575 * The port identifier of the Ethernet device. 3576 * @param queue_id 3577 * The index of the receive queue from which to retrieve input packets. 3578 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3579 * to rte_eth_dev_configure(). 3580 * @return 3581 * - (0) if successful. 3582 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3583 * that operation. 3584 * - (-ENODEV) if *port_id* invalid. 3585 * - (-EIO) if device is removed. 3586 */ 3587 int rte_eth_dev_rx_intr_disable(uint16_t port_id, uint16_t queue_id); 3588 3589 /** 3590 * RX Interrupt control per port. 3591 * 3592 * @param port_id 3593 * The port identifier of the Ethernet device. 3594 * @param epfd 3595 * Epoll instance fd which the intr vector associated to. 3596 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance. 3597 * @param op 3598 * The operation be performed for the vector. 3599 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}. 3600 * @param data 3601 * User raw data. 3602 * @return 3603 * - On success, zero. 3604 * - On failure, a negative value. 3605 */ 3606 int rte_eth_dev_rx_intr_ctl(uint16_t port_id, int epfd, int op, void *data); 3607 3608 /** 3609 * RX Interrupt control per queue. 3610 * 3611 * @param port_id 3612 * The port identifier of the Ethernet device. 3613 * @param queue_id 3614 * The index of the receive queue from which to retrieve input packets. 3615 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3616 * to rte_eth_dev_configure(). 3617 * @param epfd 3618 * Epoll instance fd which the intr vector associated to. 3619 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance. 3620 * @param op 3621 * The operation be performed for the vector. 3622 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}. 3623 * @param data 3624 * User raw data. 3625 * @return 3626 * - On success, zero. 3627 * - On failure, a negative value. 3628 */ 3629 int rte_eth_dev_rx_intr_ctl_q(uint16_t port_id, uint16_t queue_id, 3630 int epfd, int op, void *data); 3631 3632 /** 3633 * @warning 3634 * @b EXPERIMENTAL: this API may change without prior notice. 3635 * 3636 * Get interrupt fd per Rx queue. 3637 * 3638 * @param port_id 3639 * The port identifier of the Ethernet device. 3640 * @param queue_id 3641 * The index of the receive queue from which to retrieve input packets. 3642 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3643 * to rte_eth_dev_configure(). 3644 * @return 3645 * - (>=0) the interrupt fd associated to the requested Rx queue if 3646 * successful. 3647 * - (-1) on error. 3648 */ 3649 __rte_experimental 3650 int 3651 rte_eth_dev_rx_intr_ctl_q_get_fd(uint16_t port_id, uint16_t queue_id); 3652 3653 /** 3654 * Turn on the LED on the Ethernet device. 3655 * This function turns on the LED on the Ethernet device. 3656 * 3657 * @param port_id 3658 * The port identifier of the Ethernet device. 3659 * @return 3660 * - (0) if successful. 3661 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3662 * that operation. 3663 * - (-ENODEV) if *port_id* invalid. 3664 * - (-EIO) if device is removed. 3665 */ 3666 int rte_eth_led_on(uint16_t port_id); 3667 3668 /** 3669 * Turn off the LED on the Ethernet device. 3670 * This function turns off the LED on the Ethernet device. 3671 * 3672 * @param port_id 3673 * The port identifier of the Ethernet device. 3674 * @return 3675 * - (0) if successful. 3676 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3677 * that operation. 3678 * - (-ENODEV) if *port_id* invalid. 3679 * - (-EIO) if device is removed. 3680 */ 3681 int rte_eth_led_off(uint16_t port_id); 3682 3683 /** 3684 * @warning 3685 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 3686 * 3687 * Get Forward Error Correction(FEC) capability. 3688 * 3689 * @param port_id 3690 * The port identifier of the Ethernet device. 3691 * @param speed_fec_capa 3692 * speed_fec_capa is out only with per-speed capabilities. 3693 * If set to NULL, the function returns the required number 3694 * of required array entries. 3695 * @param num 3696 * a number of elements in an speed_fec_capa array. 3697 * 3698 * @return 3699 * - A non-negative value lower or equal to num: success. The return value 3700 * is the number of entries filled in the fec capa array. 3701 * - A non-negative value higher than num: error, the given fec capa array 3702 * is too small. The return value corresponds to the num that should 3703 * be given to succeed. The entries in fec capa array are not valid and 3704 * shall not be used by the caller. 3705 * - (-ENOTSUP) if underlying hardware OR driver doesn't support. 3706 * that operation. 3707 * - (-EIO) if device is removed. 3708 * - (-ENODEV) if *port_id* invalid. 3709 * - (-EINVAL) if *num* or *speed_fec_capa* invalid 3710 */ 3711 __rte_experimental 3712 int rte_eth_fec_get_capability(uint16_t port_id, 3713 struct rte_eth_fec_capa *speed_fec_capa, 3714 unsigned int num); 3715 3716 /** 3717 * @warning 3718 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 3719 * 3720 * Get current Forward Error Correction(FEC) mode. 3721 * If link is down and AUTO is enabled, AUTO is returned, otherwise, 3722 * configured FEC mode is returned. 3723 * If link is up, current FEC mode is returned. 3724 * 3725 * @param port_id 3726 * The port identifier of the Ethernet device. 3727 * @param fec_capa 3728 * A bitmask of enabled FEC modes. If AUTO bit is set, other 3729 * bits specify FEC modes which may be negotiated. If AUTO 3730 * bit is clear, specify FEC modes to be used (only one valid 3731 * mode per speed may be set). 3732 * @return 3733 * - (0) if successful. 3734 * - (-ENOTSUP) if underlying hardware OR driver doesn't support. 3735 * that operation. 3736 * - (-EIO) if device is removed. 3737 * - (-ENODEV) if *port_id* invalid. 3738 */ 3739 __rte_experimental 3740 int rte_eth_fec_get(uint16_t port_id, uint32_t *fec_capa); 3741 3742 /** 3743 * @warning 3744 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 3745 * 3746 * Set Forward Error Correction(FEC) mode. 3747 * 3748 * @param port_id 3749 * The port identifier of the Ethernet device. 3750 * @param fec_capa 3751 * A bitmask of allowed FEC modes. If AUTO bit is set, other 3752 * bits specify FEC modes which may be negotiated. If AUTO 3753 * bit is clear, specify FEC modes to be used (only one valid 3754 * mode per speed may be set). 3755 * @return 3756 * - (0) if successful. 3757 * - (-EINVAL) if the FEC mode is not valid. 3758 * - (-ENOTSUP) if underlying hardware OR driver doesn't support. 3759 * - (-EIO) if device is removed. 3760 * - (-ENODEV) if *port_id* invalid. 3761 */ 3762 __rte_experimental 3763 int rte_eth_fec_set(uint16_t port_id, uint32_t fec_capa); 3764 3765 /** 3766 * Get current status of the Ethernet link flow control for Ethernet device 3767 * 3768 * @param port_id 3769 * The port identifier of the Ethernet device. 3770 * @param fc_conf 3771 * The pointer to the structure where to store the flow control parameters. 3772 * @return 3773 * - (0) if successful. 3774 * - (-ENOTSUP) if hardware doesn't support flow control. 3775 * - (-ENODEV) if *port_id* invalid. 3776 * - (-EIO) if device is removed. 3777 */ 3778 int rte_eth_dev_flow_ctrl_get(uint16_t port_id, 3779 struct rte_eth_fc_conf *fc_conf); 3780 3781 /** 3782 * Configure the Ethernet link flow control for Ethernet device 3783 * 3784 * @param port_id 3785 * The port identifier of the Ethernet device. 3786 * @param fc_conf 3787 * The pointer to the structure of the flow control parameters. 3788 * @return 3789 * - (0) if successful. 3790 * - (-ENOTSUP) if hardware doesn't support flow control mode. 3791 * - (-ENODEV) if *port_id* invalid. 3792 * - (-EINVAL) if bad parameter 3793 * - (-EIO) if flow control setup failure or device is removed. 3794 */ 3795 int rte_eth_dev_flow_ctrl_set(uint16_t port_id, 3796 struct rte_eth_fc_conf *fc_conf); 3797 3798 /** 3799 * Configure the Ethernet priority flow control under DCB environment 3800 * for Ethernet device. 3801 * 3802 * @param port_id 3803 * The port identifier of the Ethernet device. 3804 * @param pfc_conf 3805 * The pointer to the structure of the priority flow control parameters. 3806 * @return 3807 * - (0) if successful. 3808 * - (-ENOTSUP) if hardware doesn't support priority flow control mode. 3809 * - (-ENODEV) if *port_id* invalid. 3810 * - (-EINVAL) if bad parameter 3811 * - (-EIO) if flow control setup failure or device is removed. 3812 */ 3813 int rte_eth_dev_priority_flow_ctrl_set(uint16_t port_id, 3814 struct rte_eth_pfc_conf *pfc_conf); 3815 3816 /** 3817 * Add a MAC address to the set used for filtering incoming packets. 3818 * 3819 * @param port_id 3820 * The port identifier of the Ethernet device. 3821 * @param mac_addr 3822 * The MAC address to add. 3823 * @param pool 3824 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is 3825 * not enabled, this should be set to 0. 3826 * @return 3827 * - (0) if successfully added or *mac_addr* was already added. 3828 * - (-ENOTSUP) if hardware doesn't support this feature. 3829 * - (-ENODEV) if *port* is invalid. 3830 * - (-EIO) if device is removed. 3831 * - (-ENOSPC) if no more MAC addresses can be added. 3832 * - (-EINVAL) if MAC address is invalid. 3833 */ 3834 int rte_eth_dev_mac_addr_add(uint16_t port_id, struct rte_ether_addr *mac_addr, 3835 uint32_t pool); 3836 3837 /** 3838 * Remove a MAC address from the internal array of addresses. 3839 * 3840 * @param port_id 3841 * The port identifier of the Ethernet device. 3842 * @param mac_addr 3843 * MAC address to remove. 3844 * @return 3845 * - (0) if successful, or *mac_addr* didn't exist. 3846 * - (-ENOTSUP) if hardware doesn't support. 3847 * - (-ENODEV) if *port* invalid. 3848 * - (-EADDRINUSE) if attempting to remove the default MAC address 3849 */ 3850 int rte_eth_dev_mac_addr_remove(uint16_t port_id, 3851 struct rte_ether_addr *mac_addr); 3852 3853 /** 3854 * Set the default MAC address. 3855 * 3856 * @param port_id 3857 * The port identifier of the Ethernet device. 3858 * @param mac_addr 3859 * New default MAC address. 3860 * @return 3861 * - (0) if successful, or *mac_addr* didn't exist. 3862 * - (-ENOTSUP) if hardware doesn't support. 3863 * - (-ENODEV) if *port* invalid. 3864 * - (-EINVAL) if MAC address is invalid. 3865 */ 3866 int rte_eth_dev_default_mac_addr_set(uint16_t port_id, 3867 struct rte_ether_addr *mac_addr); 3868 3869 /** 3870 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device. 3871 * 3872 * @param port_id 3873 * The port identifier of the Ethernet device. 3874 * @param reta_conf 3875 * RETA to update. 3876 * @param reta_size 3877 * Redirection table size. The table size can be queried by 3878 * rte_eth_dev_info_get(). 3879 * @return 3880 * - (0) if successful. 3881 * - (-ENODEV) if *port_id* is invalid. 3882 * - (-ENOTSUP) if hardware doesn't support. 3883 * - (-EINVAL) if bad parameter. 3884 * - (-EIO) if device is removed. 3885 */ 3886 int rte_eth_dev_rss_reta_update(uint16_t port_id, 3887 struct rte_eth_rss_reta_entry64 *reta_conf, 3888 uint16_t reta_size); 3889 3890 /** 3891 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device. 3892 * 3893 * @param port_id 3894 * The port identifier of the Ethernet device. 3895 * @param reta_conf 3896 * RETA to query. For each requested reta entry, corresponding bit 3897 * in mask must be set. 3898 * @param reta_size 3899 * Redirection table size. The table size can be queried by 3900 * rte_eth_dev_info_get(). 3901 * @return 3902 * - (0) if successful. 3903 * - (-ENODEV) if *port_id* is invalid. 3904 * - (-ENOTSUP) if hardware doesn't support. 3905 * - (-EINVAL) if bad parameter. 3906 * - (-EIO) if device is removed. 3907 */ 3908 int rte_eth_dev_rss_reta_query(uint16_t port_id, 3909 struct rte_eth_rss_reta_entry64 *reta_conf, 3910 uint16_t reta_size); 3911 3912 /** 3913 * Updates unicast hash table for receiving packet with the given destination 3914 * MAC address, and the packet is routed to all VFs for which the RX mode is 3915 * accept packets that match the unicast hash table. 3916 * 3917 * @param port_id 3918 * The port identifier of the Ethernet device. 3919 * @param addr 3920 * Unicast MAC address. 3921 * @param on 3922 * 1 - Set an unicast hash bit for receiving packets with the MAC address. 3923 * 0 - Clear an unicast hash bit. 3924 * @return 3925 * - (0) if successful. 3926 * - (-ENOTSUP) if hardware doesn't support. 3927 * - (-ENODEV) if *port_id* invalid. 3928 * - (-EIO) if device is removed. 3929 * - (-EINVAL) if bad parameter. 3930 */ 3931 int rte_eth_dev_uc_hash_table_set(uint16_t port_id, struct rte_ether_addr *addr, 3932 uint8_t on); 3933 3934 /** 3935 * Updates all unicast hash bitmaps for receiving packet with any Unicast 3936 * Ethernet MAC addresses,the packet is routed to all VFs for which the RX 3937 * mode is accept packets that match the unicast hash table. 3938 * 3939 * @param port_id 3940 * The port identifier of the Ethernet device. 3941 * @param on 3942 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet 3943 * MAC addresses 3944 * 0 - Clear all unicast hash bitmaps 3945 * @return 3946 * - (0) if successful. 3947 * - (-ENOTSUP) if hardware doesn't support. 3948 * - (-ENODEV) if *port_id* invalid. 3949 * - (-EIO) if device is removed. 3950 * - (-EINVAL) if bad parameter. 3951 */ 3952 int rte_eth_dev_uc_all_hash_table_set(uint16_t port_id, uint8_t on); 3953 3954 /** 3955 * Set a traffic mirroring rule on an Ethernet device 3956 * 3957 * @param port_id 3958 * The port identifier of the Ethernet device. 3959 * @param mirror_conf 3960 * The pointer to the traffic mirroring structure describing the mirroring rule. 3961 * The *rte_eth_vm_mirror_conf* structure includes the type of mirroring rule, 3962 * destination pool and the value of rule if enable vlan or pool mirroring. 3963 * 3964 * @param rule_id 3965 * The index of traffic mirroring rule, we support four separated rules. 3966 * @param on 3967 * 1 - Enable a mirroring rule. 3968 * 0 - Disable a mirroring rule. 3969 * @return 3970 * - (0) if successful. 3971 * - (-ENOTSUP) if hardware doesn't support this feature. 3972 * - (-ENODEV) if *port_id* invalid. 3973 * - (-EIO) if device is removed. 3974 * - (-EINVAL) if the mr_conf information is not correct. 3975 */ 3976 int rte_eth_mirror_rule_set(uint16_t port_id, 3977 struct rte_eth_mirror_conf *mirror_conf, 3978 uint8_t rule_id, 3979 uint8_t on); 3980 3981 /** 3982 * Reset a traffic mirroring rule on an Ethernet device. 3983 * 3984 * @param port_id 3985 * The port identifier of the Ethernet device. 3986 * @param rule_id 3987 * The index of traffic mirroring rule, we support four separated rules. 3988 * @return 3989 * - (0) if successful. 3990 * - (-ENOTSUP) if hardware doesn't support this feature. 3991 * - (-ENODEV) if *port_id* invalid. 3992 * - (-EIO) if device is removed. 3993 * - (-EINVAL) if bad parameter. 3994 */ 3995 int rte_eth_mirror_rule_reset(uint16_t port_id, 3996 uint8_t rule_id); 3997 3998 /** 3999 * Set the rate limitation for a queue on an Ethernet device. 4000 * 4001 * @param port_id 4002 * The port identifier of the Ethernet device. 4003 * @param queue_idx 4004 * The queue id. 4005 * @param tx_rate 4006 * The tx rate in Mbps. Allocated from the total port link speed. 4007 * @return 4008 * - (0) if successful. 4009 * - (-ENOTSUP) if hardware doesn't support this feature. 4010 * - (-ENODEV) if *port_id* invalid. 4011 * - (-EIO) if device is removed. 4012 * - (-EINVAL) if bad parameter. 4013 */ 4014 int rte_eth_set_queue_rate_limit(uint16_t port_id, uint16_t queue_idx, 4015 uint16_t tx_rate); 4016 4017 /** 4018 * Configuration of Receive Side Scaling hash computation of Ethernet device. 4019 * 4020 * @param port_id 4021 * The port identifier of the Ethernet device. 4022 * @param rss_conf 4023 * The new configuration to use for RSS hash computation on the port. 4024 * @return 4025 * - (0) if successful. 4026 * - (-ENODEV) if port identifier is invalid. 4027 * - (-EIO) if device is removed. 4028 * - (-ENOTSUP) if hardware doesn't support. 4029 * - (-EINVAL) if bad parameter. 4030 */ 4031 int rte_eth_dev_rss_hash_update(uint16_t port_id, 4032 struct rte_eth_rss_conf *rss_conf); 4033 4034 /** 4035 * Retrieve current configuration of Receive Side Scaling hash computation 4036 * of Ethernet device. 4037 * 4038 * @param port_id 4039 * The port identifier of the Ethernet device. 4040 * @param rss_conf 4041 * Where to store the current RSS hash configuration of the Ethernet device. 4042 * @return 4043 * - (0) if successful. 4044 * - (-ENODEV) if port identifier is invalid. 4045 * - (-EIO) if device is removed. 4046 * - (-ENOTSUP) if hardware doesn't support RSS. 4047 */ 4048 int 4049 rte_eth_dev_rss_hash_conf_get(uint16_t port_id, 4050 struct rte_eth_rss_conf *rss_conf); 4051 4052 /** 4053 * Add UDP tunneling port for a type of tunnel. 4054 * 4055 * Some NICs may require such configuration to properly parse a tunnel 4056 * with any standard or custom UDP port. 4057 * The packets with this UDP port will be parsed for this type of tunnel. 4058 * The device parser will also check the rest of the tunnel headers 4059 * before classifying the packet. 4060 * 4061 * With some devices, this API will affect packet classification, i.e.: 4062 * - mbuf.packet_type reported on Rx 4063 * - rte_flow rules with tunnel items 4064 * 4065 * @param port_id 4066 * The port identifier of the Ethernet device. 4067 * @param tunnel_udp 4068 * UDP tunneling configuration. 4069 * 4070 * @return 4071 * - (0) if successful. 4072 * - (-ENODEV) if port identifier is invalid. 4073 * - (-EIO) if device is removed. 4074 * - (-ENOTSUP) if hardware doesn't support tunnel type. 4075 */ 4076 int 4077 rte_eth_dev_udp_tunnel_port_add(uint16_t port_id, 4078 struct rte_eth_udp_tunnel *tunnel_udp); 4079 4080 /** 4081 * Delete UDP tunneling port for a type of tunnel. 4082 * 4083 * The packets with this UDP port will not be classified as this type of tunnel 4084 * anymore if the device use such mapping for tunnel packet classification. 4085 * 4086 * @see rte_eth_dev_udp_tunnel_port_add 4087 * 4088 * @param port_id 4089 * The port identifier of the Ethernet device. 4090 * @param tunnel_udp 4091 * UDP tunneling configuration. 4092 * 4093 * @return 4094 * - (0) if successful. 4095 * - (-ENODEV) if port identifier is invalid. 4096 * - (-EIO) if device is removed. 4097 * - (-ENOTSUP) if hardware doesn't support tunnel type. 4098 */ 4099 int 4100 rte_eth_dev_udp_tunnel_port_delete(uint16_t port_id, 4101 struct rte_eth_udp_tunnel *tunnel_udp); 4102 4103 /** 4104 * Get DCB information on an Ethernet device. 4105 * 4106 * @param port_id 4107 * The port identifier of the Ethernet device. 4108 * @param dcb_info 4109 * dcb information. 4110 * @return 4111 * - (0) if successful. 4112 * - (-ENODEV) if port identifier is invalid. 4113 * - (-EIO) if device is removed. 4114 * - (-ENOTSUP) if hardware doesn't support. 4115 */ 4116 int rte_eth_dev_get_dcb_info(uint16_t port_id, 4117 struct rte_eth_dcb_info *dcb_info); 4118 4119 struct rte_eth_rxtx_callback; 4120 4121 /** 4122 * Add a callback to be called on packet RX on a given port and queue. 4123 * 4124 * This API configures a function to be called for each burst of 4125 * packets received on a given NIC port queue. The return value is a pointer 4126 * that can be used to later remove the callback using 4127 * rte_eth_remove_rx_callback(). 4128 * 4129 * Multiple functions are called in the order that they are added. 4130 * 4131 * @param port_id 4132 * The port identifier of the Ethernet device. 4133 * @param queue_id 4134 * The queue on the Ethernet device on which the callback is to be added. 4135 * @param fn 4136 * The callback function 4137 * @param user_param 4138 * A generic pointer parameter which will be passed to each invocation of the 4139 * callback function on this port and queue. Inter-thread synchronization 4140 * of any user data changes is the responsibility of the user. 4141 * 4142 * @return 4143 * NULL on error. 4144 * On success, a pointer value which can later be used to remove the callback. 4145 */ 4146 const struct rte_eth_rxtx_callback * 4147 rte_eth_add_rx_callback(uint16_t port_id, uint16_t queue_id, 4148 rte_rx_callback_fn fn, void *user_param); 4149 4150 /** 4151 * Add a callback that must be called first on packet RX on a given port 4152 * and queue. 4153 * 4154 * This API configures a first function to be called for each burst of 4155 * packets received on a given NIC port queue. The return value is a pointer 4156 * that can be used to later remove the callback using 4157 * rte_eth_remove_rx_callback(). 4158 * 4159 * Multiple functions are called in the order that they are added. 4160 * 4161 * @param port_id 4162 * The port identifier of the Ethernet device. 4163 * @param queue_id 4164 * The queue on the Ethernet device on which the callback is to be added. 4165 * @param fn 4166 * The callback function 4167 * @param user_param 4168 * A generic pointer parameter which will be passed to each invocation of the 4169 * callback function on this port and queue. Inter-thread synchronization 4170 * of any user data changes is the responsibility of the user. 4171 * 4172 * @return 4173 * NULL on error. 4174 * On success, a pointer value which can later be used to remove the callback. 4175 */ 4176 const struct rte_eth_rxtx_callback * 4177 rte_eth_add_first_rx_callback(uint16_t port_id, uint16_t queue_id, 4178 rte_rx_callback_fn fn, void *user_param); 4179 4180 /** 4181 * Add a callback to be called on packet TX on a given port and queue. 4182 * 4183 * This API configures a function to be called for each burst of 4184 * packets sent on a given NIC port queue. The return value is a pointer 4185 * that can be used to later remove the callback using 4186 * rte_eth_remove_tx_callback(). 4187 * 4188 * Multiple functions are called in the order that they are added. 4189 * 4190 * @param port_id 4191 * The port identifier of the Ethernet device. 4192 * @param queue_id 4193 * The queue on the Ethernet device on which the callback is to be added. 4194 * @param fn 4195 * The callback function 4196 * @param user_param 4197 * A generic pointer parameter which will be passed to each invocation of the 4198 * callback function on this port and queue. Inter-thread synchronization 4199 * of any user data changes is the responsibility of the user. 4200 * 4201 * @return 4202 * NULL on error. 4203 * On success, a pointer value which can later be used to remove the callback. 4204 */ 4205 const struct rte_eth_rxtx_callback * 4206 rte_eth_add_tx_callback(uint16_t port_id, uint16_t queue_id, 4207 rte_tx_callback_fn fn, void *user_param); 4208 4209 /** 4210 * Remove an RX packet callback from a given port and queue. 4211 * 4212 * This function is used to removed callbacks that were added to a NIC port 4213 * queue using rte_eth_add_rx_callback(). 4214 * 4215 * Note: the callback is removed from the callback list but it isn't freed 4216 * since the it may still be in use. The memory for the callback can be 4217 * subsequently freed back by the application by calling rte_free(): 4218 * 4219 * - Immediately - if the port is stopped, or the user knows that no 4220 * callbacks are in flight e.g. if called from the thread doing RX/TX 4221 * on that queue. 4222 * 4223 * - After a short delay - where the delay is sufficient to allow any 4224 * in-flight callbacks to complete. Alternately, the RCU mechanism can be 4225 * used to detect when data plane threads have ceased referencing the 4226 * callback memory. 4227 * 4228 * @param port_id 4229 * The port identifier of the Ethernet device. 4230 * @param queue_id 4231 * The queue on the Ethernet device from which the callback is to be removed. 4232 * @param user_cb 4233 * User supplied callback created via rte_eth_add_rx_callback(). 4234 * 4235 * @return 4236 * - 0: Success. Callback was removed. 4237 * - -ENODEV: If *port_id* is invalid. 4238 * - -ENOTSUP: Callback support is not available. 4239 * - -EINVAL: The queue_id is out of range, or the callback 4240 * is NULL or not found for the port/queue. 4241 */ 4242 int rte_eth_remove_rx_callback(uint16_t port_id, uint16_t queue_id, 4243 const struct rte_eth_rxtx_callback *user_cb); 4244 4245 /** 4246 * Remove a TX packet callback from a given port and queue. 4247 * 4248 * This function is used to removed callbacks that were added to a NIC port 4249 * queue using rte_eth_add_tx_callback(). 4250 * 4251 * Note: the callback is removed from the callback list but it isn't freed 4252 * since the it may still be in use. The memory for the callback can be 4253 * subsequently freed back by the application by calling rte_free(): 4254 * 4255 * - Immediately - if the port is stopped, or the user knows that no 4256 * callbacks are in flight e.g. if called from the thread doing RX/TX 4257 * on that queue. 4258 * 4259 * - After a short delay - where the delay is sufficient to allow any 4260 * in-flight callbacks to complete. Alternately, the RCU mechanism can be 4261 * used to detect when data plane threads have ceased referencing the 4262 * callback memory. 4263 * 4264 * @param port_id 4265 * The port identifier of the Ethernet device. 4266 * @param queue_id 4267 * The queue on the Ethernet device from which the callback is to be removed. 4268 * @param user_cb 4269 * User supplied callback created via rte_eth_add_tx_callback(). 4270 * 4271 * @return 4272 * - 0: Success. Callback was removed. 4273 * - -ENODEV: If *port_id* is invalid. 4274 * - -ENOTSUP: Callback support is not available. 4275 * - -EINVAL: The queue_id is out of range, or the callback 4276 * is NULL or not found for the port/queue. 4277 */ 4278 int rte_eth_remove_tx_callback(uint16_t port_id, uint16_t queue_id, 4279 const struct rte_eth_rxtx_callback *user_cb); 4280 4281 /** 4282 * Retrieve information about given port's RX queue. 4283 * 4284 * @param port_id 4285 * The port identifier of the Ethernet device. 4286 * @param queue_id 4287 * The RX queue on the Ethernet device for which information 4288 * will be retrieved. 4289 * @param qinfo 4290 * A pointer to a structure of type *rte_eth_rxq_info_info* to be filled with 4291 * the information of the Ethernet device. 4292 * 4293 * @return 4294 * - 0: Success 4295 * - -ENODEV: If *port_id* is invalid. 4296 * - -ENOTSUP: routine is not supported by the device PMD. 4297 * - -EINVAL: The queue_id is out of range, or the queue 4298 * is hairpin queue. 4299 */ 4300 int rte_eth_rx_queue_info_get(uint16_t port_id, uint16_t queue_id, 4301 struct rte_eth_rxq_info *qinfo); 4302 4303 /** 4304 * Retrieve information about given port's TX queue. 4305 * 4306 * @param port_id 4307 * The port identifier of the Ethernet device. 4308 * @param queue_id 4309 * The TX queue on the Ethernet device for which information 4310 * will be retrieved. 4311 * @param qinfo 4312 * A pointer to a structure of type *rte_eth_txq_info_info* to be filled with 4313 * the information of the Ethernet device. 4314 * 4315 * @return 4316 * - 0: Success 4317 * - -ENODEV: If *port_id* is invalid. 4318 * - -ENOTSUP: routine is not supported by the device PMD. 4319 * - -EINVAL: The queue_id is out of range, or the queue 4320 * is hairpin queue. 4321 */ 4322 int rte_eth_tx_queue_info_get(uint16_t port_id, uint16_t queue_id, 4323 struct rte_eth_txq_info *qinfo); 4324 4325 /** 4326 * Retrieve information about the Rx packet burst mode. 4327 * 4328 * @param port_id 4329 * The port identifier of the Ethernet device. 4330 * @param queue_id 4331 * The Rx queue on the Ethernet device for which information 4332 * will be retrieved. 4333 * @param mode 4334 * A pointer to a structure of type *rte_eth_burst_mode* to be filled 4335 * with the information of the packet burst mode. 4336 * 4337 * @return 4338 * - 0: Success 4339 * - -ENODEV: If *port_id* is invalid. 4340 * - -ENOTSUP: routine is not supported by the device PMD. 4341 * - -EINVAL: The queue_id is out of range. 4342 */ 4343 __rte_experimental 4344 int rte_eth_rx_burst_mode_get(uint16_t port_id, uint16_t queue_id, 4345 struct rte_eth_burst_mode *mode); 4346 4347 /** 4348 * Retrieve information about the Tx packet burst mode. 4349 * 4350 * @param port_id 4351 * The port identifier of the Ethernet device. 4352 * @param queue_id 4353 * The Tx queue on the Ethernet device for which information 4354 * will be retrieved. 4355 * @param mode 4356 * A pointer to a structure of type *rte_eth_burst_mode* to be filled 4357 * with the information of the packet burst mode. 4358 * 4359 * @return 4360 * - 0: Success 4361 * - -ENODEV: If *port_id* is invalid. 4362 * - -ENOTSUP: routine is not supported by the device PMD. 4363 * - -EINVAL: The queue_id is out of range. 4364 */ 4365 __rte_experimental 4366 int rte_eth_tx_burst_mode_get(uint16_t port_id, uint16_t queue_id, 4367 struct rte_eth_burst_mode *mode); 4368 4369 /** 4370 * @warning 4371 * @b EXPERIMENTAL: this API may change without prior notice. 4372 * 4373 * Retrieve the monitor condition for a given receive queue. 4374 * 4375 * @param port_id 4376 * The port identifier of the Ethernet device. 4377 * @param queue_id 4378 * The Rx queue on the Ethernet device for which information 4379 * will be retrieved. 4380 * @param pmc 4381 * The pointer to power-optimized monitoring condition structure. 4382 * 4383 * @return 4384 * - 0: Success. 4385 * -ENOTSUP: Operation not supported. 4386 * -EINVAL: Invalid parameters. 4387 * -ENODEV: Invalid port ID. 4388 */ 4389 __rte_experimental 4390 int rte_eth_get_monitor_addr(uint16_t port_id, uint16_t queue_id, 4391 struct rte_power_monitor_cond *pmc); 4392 4393 /** 4394 * Retrieve device registers and register attributes (number of registers and 4395 * register size) 4396 * 4397 * @param port_id 4398 * The port identifier of the Ethernet device. 4399 * @param info 4400 * Pointer to rte_dev_reg_info structure to fill in. If info->data is 4401 * NULL the function fills in the width and length fields. If non-NULL 4402 * the registers are put into the buffer pointed at by the data field. 4403 * @return 4404 * - (0) if successful. 4405 * - (-ENOTSUP) if hardware doesn't support. 4406 * - (-EINVAL) if bad parameter. 4407 * - (-ENODEV) if *port_id* invalid. 4408 * - (-EIO) if device is removed. 4409 * - others depends on the specific operations implementation. 4410 */ 4411 int rte_eth_dev_get_reg_info(uint16_t port_id, struct rte_dev_reg_info *info); 4412 4413 /** 4414 * Retrieve size of device EEPROM 4415 * 4416 * @param port_id 4417 * The port identifier of the Ethernet device. 4418 * @return 4419 * - (>=0) EEPROM size if successful. 4420 * - (-ENOTSUP) if hardware doesn't support. 4421 * - (-ENODEV) if *port_id* invalid. 4422 * - (-EIO) if device is removed. 4423 * - others depends on the specific operations implementation. 4424 */ 4425 int rte_eth_dev_get_eeprom_length(uint16_t port_id); 4426 4427 /** 4428 * Retrieve EEPROM and EEPROM attribute 4429 * 4430 * @param port_id 4431 * The port identifier of the Ethernet device. 4432 * @param info 4433 * The template includes buffer for return EEPROM data and 4434 * EEPROM attributes to be filled. 4435 * @return 4436 * - (0) if successful. 4437 * - (-ENOTSUP) if hardware doesn't support. 4438 * - (-EINVAL) if bad parameter. 4439 * - (-ENODEV) if *port_id* invalid. 4440 * - (-EIO) if device is removed. 4441 * - others depends on the specific operations implementation. 4442 */ 4443 int rte_eth_dev_get_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info); 4444 4445 /** 4446 * Program EEPROM with provided data 4447 * 4448 * @param port_id 4449 * The port identifier of the Ethernet device. 4450 * @param info 4451 * The template includes EEPROM data for programming and 4452 * EEPROM attributes to be filled 4453 * @return 4454 * - (0) if successful. 4455 * - (-ENOTSUP) if hardware doesn't support. 4456 * - (-ENODEV) if *port_id* invalid. 4457 * - (-EINVAL) if bad parameter. 4458 * - (-EIO) if device is removed. 4459 * - others depends on the specific operations implementation. 4460 */ 4461 int rte_eth_dev_set_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info); 4462 4463 /** 4464 * @warning 4465 * @b EXPERIMENTAL: this API may change without prior notice. 4466 * 4467 * Retrieve the type and size of plugin module EEPROM 4468 * 4469 * @param port_id 4470 * The port identifier of the Ethernet device. 4471 * @param modinfo 4472 * The type and size of plugin module EEPROM. 4473 * @return 4474 * - (0) if successful. 4475 * - (-ENOTSUP) if hardware doesn't support. 4476 * - (-ENODEV) if *port_id* invalid. 4477 * - (-EINVAL) if bad parameter. 4478 * - (-EIO) if device is removed. 4479 * - others depends on the specific operations implementation. 4480 */ 4481 __rte_experimental 4482 int 4483 rte_eth_dev_get_module_info(uint16_t port_id, 4484 struct rte_eth_dev_module_info *modinfo); 4485 4486 /** 4487 * @warning 4488 * @b EXPERIMENTAL: this API may change without prior notice. 4489 * 4490 * Retrieve the data of plugin module EEPROM 4491 * 4492 * @param port_id 4493 * The port identifier of the Ethernet device. 4494 * @param info 4495 * The template includes the plugin module EEPROM attributes, and the 4496 * buffer for return plugin module EEPROM data. 4497 * @return 4498 * - (0) if successful. 4499 * - (-ENOTSUP) if hardware doesn't support. 4500 * - (-EINVAL) if bad parameter. 4501 * - (-ENODEV) if *port_id* invalid. 4502 * - (-EIO) if device is removed. 4503 * - others depends on the specific operations implementation. 4504 */ 4505 __rte_experimental 4506 int 4507 rte_eth_dev_get_module_eeprom(uint16_t port_id, 4508 struct rte_dev_eeprom_info *info); 4509 4510 /** 4511 * Set the list of multicast addresses to filter on an Ethernet device. 4512 * 4513 * @param port_id 4514 * The port identifier of the Ethernet device. 4515 * @param mc_addr_set 4516 * The array of multicast addresses to set. Equal to NULL when the function 4517 * is invoked to flush the set of filtered addresses. 4518 * @param nb_mc_addr 4519 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0 4520 * when the function is invoked to flush the set of filtered addresses. 4521 * @return 4522 * - (0) if successful. 4523 * - (-ENODEV) if *port_id* invalid. 4524 * - (-EIO) if device is removed. 4525 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering. 4526 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources. 4527 */ 4528 int rte_eth_dev_set_mc_addr_list(uint16_t port_id, 4529 struct rte_ether_addr *mc_addr_set, 4530 uint32_t nb_mc_addr); 4531 4532 /** 4533 * Enable IEEE1588/802.1AS timestamping for an Ethernet device. 4534 * 4535 * @param port_id 4536 * The port identifier of the Ethernet device. 4537 * 4538 * @return 4539 * - 0: Success. 4540 * - -ENODEV: The port ID is invalid. 4541 * - -EIO: if device is removed. 4542 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4543 */ 4544 int rte_eth_timesync_enable(uint16_t port_id); 4545 4546 /** 4547 * Disable IEEE1588/802.1AS timestamping for an Ethernet device. 4548 * 4549 * @param port_id 4550 * The port identifier of the Ethernet device. 4551 * 4552 * @return 4553 * - 0: Success. 4554 * - -ENODEV: The port ID is invalid. 4555 * - -EIO: if device is removed. 4556 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4557 */ 4558 int rte_eth_timesync_disable(uint16_t port_id); 4559 4560 /** 4561 * Read an IEEE1588/802.1AS RX timestamp from an Ethernet device. 4562 * 4563 * @param port_id 4564 * The port identifier of the Ethernet device. 4565 * @param timestamp 4566 * Pointer to the timestamp struct. 4567 * @param flags 4568 * Device specific flags. Used to pass the RX timesync register index to 4569 * i40e. Unused in igb/ixgbe, pass 0 instead. 4570 * 4571 * @return 4572 * - 0: Success. 4573 * - -EINVAL: No timestamp is available. 4574 * - -ENODEV: The port ID is invalid. 4575 * - -EIO: if device is removed. 4576 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4577 */ 4578 int rte_eth_timesync_read_rx_timestamp(uint16_t port_id, 4579 struct timespec *timestamp, uint32_t flags); 4580 4581 /** 4582 * Read an IEEE1588/802.1AS TX timestamp from an Ethernet device. 4583 * 4584 * @param port_id 4585 * The port identifier of the Ethernet device. 4586 * @param timestamp 4587 * Pointer to the timestamp struct. 4588 * 4589 * @return 4590 * - 0: Success. 4591 * - -EINVAL: No timestamp is available. 4592 * - -ENODEV: The port ID is invalid. 4593 * - -EIO: if device is removed. 4594 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4595 */ 4596 int rte_eth_timesync_read_tx_timestamp(uint16_t port_id, 4597 struct timespec *timestamp); 4598 4599 /** 4600 * Adjust the timesync clock on an Ethernet device. 4601 * 4602 * This is usually used in conjunction with other Ethdev timesync functions to 4603 * synchronize the device time using the IEEE1588/802.1AS protocol. 4604 * 4605 * @param port_id 4606 * The port identifier of the Ethernet device. 4607 * @param delta 4608 * The adjustment in nanoseconds. 4609 * 4610 * @return 4611 * - 0: Success. 4612 * - -ENODEV: The port ID is invalid. 4613 * - -EIO: if device is removed. 4614 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4615 */ 4616 int rte_eth_timesync_adjust_time(uint16_t port_id, int64_t delta); 4617 4618 /** 4619 * Read the time from the timesync clock on an Ethernet device. 4620 * 4621 * This is usually used in conjunction with other Ethdev timesync functions to 4622 * synchronize the device time using the IEEE1588/802.1AS protocol. 4623 * 4624 * @param port_id 4625 * The port identifier of the Ethernet device. 4626 * @param time 4627 * Pointer to the timespec struct that holds the time. 4628 * 4629 * @return 4630 * - 0: Success. 4631 */ 4632 int rte_eth_timesync_read_time(uint16_t port_id, struct timespec *time); 4633 4634 /** 4635 * Set the time of the timesync clock on an Ethernet device. 4636 * 4637 * This is usually used in conjunction with other Ethdev timesync functions to 4638 * synchronize the device time using the IEEE1588/802.1AS protocol. 4639 * 4640 * @param port_id 4641 * The port identifier of the Ethernet device. 4642 * @param time 4643 * Pointer to the timespec struct that holds the time. 4644 * 4645 * @return 4646 * - 0: Success. 4647 * - -EINVAL: No timestamp is available. 4648 * - -ENODEV: The port ID is invalid. 4649 * - -EIO: if device is removed. 4650 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4651 */ 4652 int rte_eth_timesync_write_time(uint16_t port_id, const struct timespec *time); 4653 4654 /** 4655 * @warning 4656 * @b EXPERIMENTAL: this API may change without prior notice. 4657 * 4658 * Read the current clock counter of an Ethernet device 4659 * 4660 * This returns the current raw clock value of an Ethernet device. It is 4661 * a raw amount of ticks, with no given time reference. 4662 * The value returned here is from the same clock than the one 4663 * filling timestamp field of Rx packets when using hardware timestamp 4664 * offload. Therefore it can be used to compute a precise conversion of 4665 * the device clock to the real time. 4666 * 4667 * E.g, a simple heuristic to derivate the frequency would be: 4668 * uint64_t start, end; 4669 * rte_eth_read_clock(port, start); 4670 * rte_delay_ms(100); 4671 * rte_eth_read_clock(port, end); 4672 * double freq = (end - start) * 10; 4673 * 4674 * Compute a common reference with: 4675 * uint64_t base_time_sec = current_time(); 4676 * uint64_t base_clock; 4677 * rte_eth_read_clock(port, base_clock); 4678 * 4679 * Then, convert the raw mbuf timestamp with: 4680 * base_time_sec + (double)(*timestamp_dynfield(mbuf) - base_clock) / freq; 4681 * 4682 * This simple example will not provide a very good accuracy. One must 4683 * at least measure multiple times the frequency and do a regression. 4684 * To avoid deviation from the system time, the common reference can 4685 * be repeated from time to time. The integer division can also be 4686 * converted by a multiplication and a shift for better performance. 4687 * 4688 * @param port_id 4689 * The port identifier of the Ethernet device. 4690 * @param clock 4691 * Pointer to the uint64_t that holds the raw clock value. 4692 * 4693 * @return 4694 * - 0: Success. 4695 * - -ENODEV: The port ID is invalid. 4696 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4697 */ 4698 __rte_experimental 4699 int 4700 rte_eth_read_clock(uint16_t port_id, uint64_t *clock); 4701 4702 /** 4703 * Get the port id from device name. The device name should be specified 4704 * as below: 4705 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:2:00.0 4706 * - SoC device name, for example- fsl-gmac0 4707 * - vdev dpdk name, for example- net_[pcap0|null0|tap0] 4708 * 4709 * @param name 4710 * pci address or name of the device 4711 * @param port_id 4712 * pointer to port identifier of the device 4713 * @return 4714 * - (0) if successful and port_id is filled. 4715 * - (-ENODEV or -EINVAL) on failure. 4716 */ 4717 int 4718 rte_eth_dev_get_port_by_name(const char *name, uint16_t *port_id); 4719 4720 /** 4721 * Get the device name from port id. The device name is specified as below: 4722 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:02:00.0 4723 * - SoC device name, for example- fsl-gmac0 4724 * - vdev dpdk name, for example- net_[pcap0|null0|tun0|tap0] 4725 * 4726 * @param port_id 4727 * Port identifier of the device. 4728 * @param name 4729 * Buffer of size RTE_ETH_NAME_MAX_LEN to store the name. 4730 * @return 4731 * - (0) if successful. 4732 * - (-ENODEV) if *port_id* is invalid. 4733 * - (-EINVAL) on failure. 4734 */ 4735 int 4736 rte_eth_dev_get_name_by_port(uint16_t port_id, char *name); 4737 4738 /** 4739 * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from 4740 * the ethernet device information, otherwise adjust them to boundaries. 4741 * 4742 * @param port_id 4743 * The port identifier of the Ethernet device. 4744 * @param nb_rx_desc 4745 * A pointer to a uint16_t where the number of receive 4746 * descriptors stored. 4747 * @param nb_tx_desc 4748 * A pointer to a uint16_t where the number of transmit 4749 * descriptors stored. 4750 * @return 4751 * - (0) if successful. 4752 * - (-ENOTSUP, -ENODEV or -EINVAL) on failure. 4753 */ 4754 int rte_eth_dev_adjust_nb_rx_tx_desc(uint16_t port_id, 4755 uint16_t *nb_rx_desc, 4756 uint16_t *nb_tx_desc); 4757 4758 /** 4759 * Test if a port supports specific mempool ops. 4760 * 4761 * @param port_id 4762 * Port identifier of the Ethernet device. 4763 * @param [in] pool 4764 * The name of the pool operations to test. 4765 * @return 4766 * - 0: best mempool ops choice for this port. 4767 * - 1: mempool ops are supported for this port. 4768 * - -ENOTSUP: mempool ops not supported for this port. 4769 * - -ENODEV: Invalid port Identifier. 4770 * - -EINVAL: Pool param is null. 4771 */ 4772 int 4773 rte_eth_dev_pool_ops_supported(uint16_t port_id, const char *pool); 4774 4775 /** 4776 * Get the security context for the Ethernet device. 4777 * 4778 * @param port_id 4779 * Port identifier of the Ethernet device 4780 * @return 4781 * - NULL on error. 4782 * - pointer to security context on success. 4783 */ 4784 void * 4785 rte_eth_dev_get_sec_ctx(uint16_t port_id); 4786 4787 /** 4788 * @warning 4789 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 4790 * 4791 * Query the device hairpin capabilities. 4792 * 4793 * @param port_id 4794 * The port identifier of the Ethernet device. 4795 * @param cap 4796 * Pointer to a structure that will hold the hairpin capabilities. 4797 * @return 4798 * - (0) if successful. 4799 * - (-ENOTSUP) if hardware doesn't support. 4800 */ 4801 __rte_experimental 4802 int rte_eth_dev_hairpin_capability_get(uint16_t port_id, 4803 struct rte_eth_hairpin_cap *cap); 4804 4805 /** 4806 * @warning 4807 * @b EXPERIMENTAL: this structure may change without prior notice. 4808 * 4809 * ethernet device representor ID range entry 4810 */ 4811 struct rte_eth_representor_range { 4812 enum rte_eth_representor_type type; /**< Representor type */ 4813 int controller; /**< Controller index */ 4814 int pf; /**< Physical function index */ 4815 __extension__ 4816 union { 4817 int vf; /**< VF start index */ 4818 int sf; /**< SF start index */ 4819 }; 4820 uint32_t id_base; /**< Representor ID start index */ 4821 uint32_t id_end; /**< Representor ID end index */ 4822 char name[RTE_DEV_NAME_MAX_LEN]; /**< Representor name */ 4823 }; 4824 4825 /** 4826 * @warning 4827 * @b EXPERIMENTAL: this structure may change without prior notice. 4828 * 4829 * Ethernet device representor information 4830 */ 4831 struct rte_eth_representor_info { 4832 uint16_t controller; /**< Controller ID of caller device. */ 4833 uint16_t pf; /**< Physical function ID of caller device. */ 4834 struct rte_eth_representor_range ranges[];/**< Representor ID range. */ 4835 }; 4836 4837 /** 4838 * Retrieve the representor info of the device. 4839 * 4840 * Get device representor info to be able to calculate a unique 4841 * representor ID. @see rte_eth_representor_id_get helper. 4842 * 4843 * @param port_id 4844 * The port identifier of the device. 4845 * @param info 4846 * A pointer to a representor info structure. 4847 * NULL to return number of range entries and allocate memory 4848 * for next call to store detail. 4849 * @return 4850 * - (-ENOTSUP) if operation is not supported. 4851 * - (-ENODEV) if *port_id* invalid. 4852 * - (-EIO) if device is removed. 4853 * - (>=0) number of representor range entries supported by device. 4854 */ 4855 __rte_experimental 4856 int rte_eth_representor_info_get(uint16_t port_id, 4857 struct rte_eth_representor_info *info); 4858 4859 #include <rte_ethdev_core.h> 4860 4861 /** 4862 * 4863 * Retrieve a burst of input packets from a receive queue of an Ethernet 4864 * device. The retrieved packets are stored in *rte_mbuf* structures whose 4865 * pointers are supplied in the *rx_pkts* array. 4866 * 4867 * The rte_eth_rx_burst() function loops, parsing the RX ring of the 4868 * receive queue, up to *nb_pkts* packets, and for each completed RX 4869 * descriptor in the ring, it performs the following operations: 4870 * 4871 * - Initialize the *rte_mbuf* data structure associated with the 4872 * RX descriptor according to the information provided by the NIC into 4873 * that RX descriptor. 4874 * 4875 * - Store the *rte_mbuf* data structure into the next entry of the 4876 * *rx_pkts* array. 4877 * 4878 * - Replenish the RX descriptor with a new *rte_mbuf* buffer 4879 * allocated from the memory pool associated with the receive queue at 4880 * initialization time. 4881 * 4882 * When retrieving an input packet that was scattered by the controller 4883 * into multiple receive descriptors, the rte_eth_rx_burst() function 4884 * appends the associated *rte_mbuf* buffers to the first buffer of the 4885 * packet. 4886 * 4887 * The rte_eth_rx_burst() function returns the number of packets 4888 * actually retrieved, which is the number of *rte_mbuf* data structures 4889 * effectively supplied into the *rx_pkts* array. 4890 * A return value equal to *nb_pkts* indicates that the RX queue contained 4891 * at least *rx_pkts* packets, and this is likely to signify that other 4892 * received packets remain in the input queue. Applications implementing 4893 * a "retrieve as much received packets as possible" policy can check this 4894 * specific case and keep invoking the rte_eth_rx_burst() function until 4895 * a value less than *nb_pkts* is returned. 4896 * 4897 * This receive method has the following advantages: 4898 * 4899 * - It allows a run-to-completion network stack engine to retrieve and 4900 * to immediately process received packets in a fast burst-oriented 4901 * approach, avoiding the overhead of unnecessary intermediate packet 4902 * queue/dequeue operations. 4903 * 4904 * - Conversely, it also allows an asynchronous-oriented processing 4905 * method to retrieve bursts of received packets and to immediately 4906 * queue them for further parallel processing by another logical core, 4907 * for instance. However, instead of having received packets being 4908 * individually queued by the driver, this approach allows the caller 4909 * of the rte_eth_rx_burst() function to queue a burst of retrieved 4910 * packets at a time and therefore dramatically reduce the cost of 4911 * enqueue/dequeue operations per packet. 4912 * 4913 * - It allows the rte_eth_rx_burst() function of the driver to take 4914 * advantage of burst-oriented hardware features (CPU cache, 4915 * prefetch instructions, and so on) to minimize the number of CPU 4916 * cycles per packet. 4917 * 4918 * To summarize, the proposed receive API enables many 4919 * burst-oriented optimizations in both synchronous and asynchronous 4920 * packet processing environments with no overhead in both cases. 4921 * 4922 * @note 4923 * Some drivers using vector instructions require that *nb_pkts* is 4924 * divisible by 4 or 8, depending on the driver implementation. 4925 * 4926 * The rte_eth_rx_burst() function does not provide any error 4927 * notification to avoid the corresponding overhead. As a hint, the 4928 * upper-level application might check the status of the device link once 4929 * being systematically returned a 0 value for a given number of tries. 4930 * 4931 * @param port_id 4932 * The port identifier of the Ethernet device. 4933 * @param queue_id 4934 * The index of the receive queue from which to retrieve input packets. 4935 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 4936 * to rte_eth_dev_configure(). 4937 * @param rx_pkts 4938 * The address of an array of pointers to *rte_mbuf* structures that 4939 * must be large enough to store *nb_pkts* pointers in it. 4940 * @param nb_pkts 4941 * The maximum number of packets to retrieve. 4942 * The value must be divisible by 8 in order to work with any driver. 4943 * @return 4944 * The number of packets actually retrieved, which is the number 4945 * of pointers to *rte_mbuf* structures effectively supplied to the 4946 * *rx_pkts* array. 4947 */ 4948 static inline uint16_t 4949 rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id, 4950 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts) 4951 { 4952 struct rte_eth_dev *dev = &rte_eth_devices[port_id]; 4953 uint16_t nb_rx; 4954 4955 #ifdef RTE_ETHDEV_DEBUG_RX 4956 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0); 4957 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_pkt_burst, 0); 4958 4959 if (queue_id >= dev->data->nb_rx_queues) { 4960 RTE_ETHDEV_LOG(ERR, "Invalid RX queue_id=%u\n", queue_id); 4961 return 0; 4962 } 4963 #endif 4964 nb_rx = (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id], 4965 rx_pkts, nb_pkts); 4966 4967 #ifdef RTE_ETHDEV_RXTX_CALLBACKS 4968 struct rte_eth_rxtx_callback *cb; 4969 4970 /* __ATOMIC_RELEASE memory order was used when the 4971 * call back was inserted into the list. 4972 * Since there is a clear dependency between loading 4973 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is 4974 * not required. 4975 */ 4976 cb = __atomic_load_n(&dev->post_rx_burst_cbs[queue_id], 4977 __ATOMIC_RELAXED); 4978 4979 if (unlikely(cb != NULL)) { 4980 do { 4981 nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx, 4982 nb_pkts, cb->param); 4983 cb = cb->next; 4984 } while (cb != NULL); 4985 } 4986 #endif 4987 4988 rte_ethdev_trace_rx_burst(port_id, queue_id, (void **)rx_pkts, nb_rx); 4989 return nb_rx; 4990 } 4991 4992 /** 4993 * Get the number of used descriptors of a rx queue 4994 * 4995 * @param port_id 4996 * The port identifier of the Ethernet device. 4997 * @param queue_id 4998 * The queue id on the specific port. 4999 * @return 5000 * The number of used descriptors in the specific queue, or: 5001 * - (-ENODEV) if *port_id* is invalid. 5002 * (-EINVAL) if *queue_id* is invalid 5003 * (-ENOTSUP) if the device does not support this function 5004 */ 5005 static inline int 5006 rte_eth_rx_queue_count(uint16_t port_id, uint16_t queue_id) 5007 { 5008 struct rte_eth_dev *dev; 5009 5010 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); 5011 dev = &rte_eth_devices[port_id]; 5012 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_queue_count, -ENOTSUP); 5013 if (queue_id >= dev->data->nb_rx_queues || 5014 dev->data->rx_queues[queue_id] == NULL) 5015 return -EINVAL; 5016 5017 return (int)(*dev->rx_queue_count)(dev, queue_id); 5018 } 5019 5020 /** 5021 * Check if the DD bit of the specific RX descriptor in the queue has been set 5022 * 5023 * @param port_id 5024 * The port identifier of the Ethernet device. 5025 * @param queue_id 5026 * The queue id on the specific port. 5027 * @param offset 5028 * The offset of the descriptor ID from tail. 5029 * @return 5030 * - (1) if the specific DD bit is set. 5031 * - (0) if the specific DD bit is not set. 5032 * - (-ENODEV) if *port_id* invalid. 5033 * - (-ENOTSUP) if the device does not support this function 5034 */ 5035 __rte_deprecated 5036 static inline int 5037 rte_eth_rx_descriptor_done(uint16_t port_id, uint16_t queue_id, uint16_t offset) 5038 { 5039 struct rte_eth_dev *dev = &rte_eth_devices[port_id]; 5040 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); 5041 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_descriptor_done, -ENOTSUP); 5042 return (*dev->rx_descriptor_done)(dev->data->rx_queues[queue_id], offset); 5043 } 5044 5045 #define RTE_ETH_RX_DESC_AVAIL 0 /**< Desc available for hw. */ 5046 #define RTE_ETH_RX_DESC_DONE 1 /**< Desc done, filled by hw. */ 5047 #define RTE_ETH_RX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */ 5048 5049 /** 5050 * Check the status of a Rx descriptor in the queue 5051 * 5052 * It should be called in a similar context than the Rx function: 5053 * - on a dataplane core 5054 * - not concurrently on the same queue 5055 * 5056 * Since it's a dataplane function, no check is performed on port_id and 5057 * queue_id. The caller must therefore ensure that the port is enabled 5058 * and the queue is configured and running. 5059 * 5060 * Note: accessing to a random descriptor in the ring may trigger cache 5061 * misses and have a performance impact. 5062 * 5063 * @param port_id 5064 * A valid port identifier of the Ethernet device which. 5065 * @param queue_id 5066 * A valid Rx queue identifier on this port. 5067 * @param offset 5068 * The offset of the descriptor starting from tail (0 is the next 5069 * packet to be received by the driver). 5070 * 5071 * @return 5072 * - (RTE_ETH_RX_DESC_AVAIL): Descriptor is available for the hardware to 5073 * receive a packet. 5074 * - (RTE_ETH_RX_DESC_DONE): Descriptor is done, it is filled by hw, but 5075 * not yet processed by the driver (i.e. in the receive queue). 5076 * - (RTE_ETH_RX_DESC_UNAVAIL): Descriptor is unavailable, either hold by 5077 * the driver and not yet returned to hw, or reserved by the hw. 5078 * - (-EINVAL) bad descriptor offset. 5079 * - (-ENOTSUP) if the device does not support this function. 5080 * - (-ENODEV) bad port or queue (only if compiled with debug). 5081 */ 5082 static inline int 5083 rte_eth_rx_descriptor_status(uint16_t port_id, uint16_t queue_id, 5084 uint16_t offset) 5085 { 5086 struct rte_eth_dev *dev; 5087 void *rxq; 5088 5089 #ifdef RTE_ETHDEV_DEBUG_RX 5090 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); 5091 #endif 5092 dev = &rte_eth_devices[port_id]; 5093 #ifdef RTE_ETHDEV_DEBUG_RX 5094 if (queue_id >= dev->data->nb_rx_queues) 5095 return -ENODEV; 5096 #endif 5097 RTE_FUNC_PTR_OR_ERR_RET(*dev->rx_descriptor_status, -ENOTSUP); 5098 rxq = dev->data->rx_queues[queue_id]; 5099 5100 return (*dev->rx_descriptor_status)(rxq, offset); 5101 } 5102 5103 #define RTE_ETH_TX_DESC_FULL 0 /**< Desc filled for hw, waiting xmit. */ 5104 #define RTE_ETH_TX_DESC_DONE 1 /**< Desc done, packet is transmitted. */ 5105 #define RTE_ETH_TX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */ 5106 5107 /** 5108 * Check the status of a Tx descriptor in the queue. 5109 * 5110 * It should be called in a similar context than the Tx function: 5111 * - on a dataplane core 5112 * - not concurrently on the same queue 5113 * 5114 * Since it's a dataplane function, no check is performed on port_id and 5115 * queue_id. The caller must therefore ensure that the port is enabled 5116 * and the queue is configured and running. 5117 * 5118 * Note: accessing to a random descriptor in the ring may trigger cache 5119 * misses and have a performance impact. 5120 * 5121 * @param port_id 5122 * A valid port identifier of the Ethernet device which. 5123 * @param queue_id 5124 * A valid Tx queue identifier on this port. 5125 * @param offset 5126 * The offset of the descriptor starting from tail (0 is the place where 5127 * the next packet will be send). 5128 * 5129 * @return 5130 * - (RTE_ETH_TX_DESC_FULL) Descriptor is being processed by the hw, i.e. 5131 * in the transmit queue. 5132 * - (RTE_ETH_TX_DESC_DONE) Hardware is done with this descriptor, it can 5133 * be reused by the driver. 5134 * - (RTE_ETH_TX_DESC_UNAVAIL): Descriptor is unavailable, reserved by the 5135 * driver or the hardware. 5136 * - (-EINVAL) bad descriptor offset. 5137 * - (-ENOTSUP) if the device does not support this function. 5138 * - (-ENODEV) bad port or queue (only if compiled with debug). 5139 */ 5140 static inline int rte_eth_tx_descriptor_status(uint16_t port_id, 5141 uint16_t queue_id, uint16_t offset) 5142 { 5143 struct rte_eth_dev *dev; 5144 void *txq; 5145 5146 #ifdef RTE_ETHDEV_DEBUG_TX 5147 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); 5148 #endif 5149 dev = &rte_eth_devices[port_id]; 5150 #ifdef RTE_ETHDEV_DEBUG_TX 5151 if (queue_id >= dev->data->nb_tx_queues) 5152 return -ENODEV; 5153 #endif 5154 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_descriptor_status, -ENOTSUP); 5155 txq = dev->data->tx_queues[queue_id]; 5156 5157 return (*dev->tx_descriptor_status)(txq, offset); 5158 } 5159 5160 /** 5161 * Send a burst of output packets on a transmit queue of an Ethernet device. 5162 * 5163 * The rte_eth_tx_burst() function is invoked to transmit output packets 5164 * on the output queue *queue_id* of the Ethernet device designated by its 5165 * *port_id*. 5166 * The *nb_pkts* parameter is the number of packets to send which are 5167 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them 5168 * allocated from a pool created with rte_pktmbuf_pool_create(). 5169 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets, 5170 * up to the number of transmit descriptors available in the TX ring of the 5171 * transmit queue. 5172 * For each packet to send, the rte_eth_tx_burst() function performs 5173 * the following operations: 5174 * 5175 * - Pick up the next available descriptor in the transmit ring. 5176 * 5177 * - Free the network buffer previously sent with that descriptor, if any. 5178 * 5179 * - Initialize the transmit descriptor with the information provided 5180 * in the *rte_mbuf data structure. 5181 * 5182 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers, 5183 * the rte_eth_tx_burst() function uses several transmit descriptors 5184 * of the ring. 5185 * 5186 * The rte_eth_tx_burst() function returns the number of packets it 5187 * actually sent. A return value equal to *nb_pkts* means that all packets 5188 * have been sent, and this is likely to signify that other output packets 5189 * could be immediately transmitted again. Applications that implement a 5190 * "send as many packets to transmit as possible" policy can check this 5191 * specific case and keep invoking the rte_eth_tx_burst() function until 5192 * a value less than *nb_pkts* is returned. 5193 * 5194 * It is the responsibility of the rte_eth_tx_burst() function to 5195 * transparently free the memory buffers of packets previously sent. 5196 * This feature is driven by the *tx_free_thresh* value supplied to the 5197 * rte_eth_dev_configure() function at device configuration time. 5198 * When the number of free TX descriptors drops below this threshold, the 5199 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers 5200 * of those packets whose transmission was effectively completed. 5201 * 5202 * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can 5203 * invoke this function concurrently on the same tx queue without SW lock. 5204 * @see rte_eth_dev_info_get, struct rte_eth_txconf::offloads 5205 * 5206 * @see rte_eth_tx_prepare to perform some prior checks or adjustments 5207 * for offloads. 5208 * 5209 * @param port_id 5210 * The port identifier of the Ethernet device. 5211 * @param queue_id 5212 * The index of the transmit queue through which output packets must be 5213 * sent. 5214 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5215 * to rte_eth_dev_configure(). 5216 * @param tx_pkts 5217 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures 5218 * which contain the output packets. 5219 * @param nb_pkts 5220 * The maximum number of packets to transmit. 5221 * @return 5222 * The number of output packets actually stored in transmit descriptors of 5223 * the transmit ring. The return value can be less than the value of the 5224 * *tx_pkts* parameter when the transmit ring is full or has been filled up. 5225 */ 5226 static inline uint16_t 5227 rte_eth_tx_burst(uint16_t port_id, uint16_t queue_id, 5228 struct rte_mbuf **tx_pkts, uint16_t nb_pkts) 5229 { 5230 struct rte_eth_dev *dev = &rte_eth_devices[port_id]; 5231 5232 #ifdef RTE_ETHDEV_DEBUG_TX 5233 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0); 5234 RTE_FUNC_PTR_OR_ERR_RET(*dev->tx_pkt_burst, 0); 5235 5236 if (queue_id >= dev->data->nb_tx_queues) { 5237 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id); 5238 return 0; 5239 } 5240 #endif 5241 5242 #ifdef RTE_ETHDEV_RXTX_CALLBACKS 5243 struct rte_eth_rxtx_callback *cb; 5244 5245 /* __ATOMIC_RELEASE memory order was used when the 5246 * call back was inserted into the list. 5247 * Since there is a clear dependency between loading 5248 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is 5249 * not required. 5250 */ 5251 cb = __atomic_load_n(&dev->pre_tx_burst_cbs[queue_id], 5252 __ATOMIC_RELAXED); 5253 5254 if (unlikely(cb != NULL)) { 5255 do { 5256 nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts, 5257 cb->param); 5258 cb = cb->next; 5259 } while (cb != NULL); 5260 } 5261 #endif 5262 5263 rte_ethdev_trace_tx_burst(port_id, queue_id, (void **)tx_pkts, 5264 nb_pkts); 5265 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts); 5266 } 5267 5268 /** 5269 * Process a burst of output packets on a transmit queue of an Ethernet device. 5270 * 5271 * The rte_eth_tx_prepare() function is invoked to prepare output packets to be 5272 * transmitted on the output queue *queue_id* of the Ethernet device designated 5273 * by its *port_id*. 5274 * The *nb_pkts* parameter is the number of packets to be prepared which are 5275 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them 5276 * allocated from a pool created with rte_pktmbuf_pool_create(). 5277 * For each packet to send, the rte_eth_tx_prepare() function performs 5278 * the following operations: 5279 * 5280 * - Check if packet meets devices requirements for tx offloads. 5281 * 5282 * - Check limitations about number of segments. 5283 * 5284 * - Check additional requirements when debug is enabled. 5285 * 5286 * - Update and/or reset required checksums when tx offload is set for packet. 5287 * 5288 * Since this function can modify packet data, provided mbufs must be safely 5289 * writable (e.g. modified data cannot be in shared segment). 5290 * 5291 * The rte_eth_tx_prepare() function returns the number of packets ready to be 5292 * sent. A return value equal to *nb_pkts* means that all packets are valid and 5293 * ready to be sent, otherwise stops processing on the first invalid packet and 5294 * leaves the rest packets untouched. 5295 * 5296 * When this functionality is not implemented in the driver, all packets are 5297 * are returned untouched. 5298 * 5299 * @param port_id 5300 * The port identifier of the Ethernet device. 5301 * The value must be a valid port id. 5302 * @param queue_id 5303 * The index of the transmit queue through which output packets must be 5304 * sent. 5305 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5306 * to rte_eth_dev_configure(). 5307 * @param tx_pkts 5308 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures 5309 * which contain the output packets. 5310 * @param nb_pkts 5311 * The maximum number of packets to process. 5312 * @return 5313 * The number of packets correct and ready to be sent. The return value can be 5314 * less than the value of the *tx_pkts* parameter when some packet doesn't 5315 * meet devices requirements with rte_errno set appropriately: 5316 * - EINVAL: offload flags are not correctly set 5317 * - ENOTSUP: the offload feature is not supported by the hardware 5318 * - ENODEV: if *port_id* is invalid (with debug enabled only) 5319 * 5320 */ 5321 5322 #ifndef RTE_ETHDEV_TX_PREPARE_NOOP 5323 5324 static inline uint16_t 5325 rte_eth_tx_prepare(uint16_t port_id, uint16_t queue_id, 5326 struct rte_mbuf **tx_pkts, uint16_t nb_pkts) 5327 { 5328 struct rte_eth_dev *dev; 5329 5330 #ifdef RTE_ETHDEV_DEBUG_TX 5331 if (!rte_eth_dev_is_valid_port(port_id)) { 5332 RTE_ETHDEV_LOG(ERR, "Invalid TX port_id=%u\n", port_id); 5333 rte_errno = ENODEV; 5334 return 0; 5335 } 5336 #endif 5337 5338 dev = &rte_eth_devices[port_id]; 5339 5340 #ifdef RTE_ETHDEV_DEBUG_TX 5341 if (queue_id >= dev->data->nb_tx_queues) { 5342 RTE_ETHDEV_LOG(ERR, "Invalid TX queue_id=%u\n", queue_id); 5343 rte_errno = EINVAL; 5344 return 0; 5345 } 5346 #endif 5347 5348 if (!dev->tx_pkt_prepare) 5349 return nb_pkts; 5350 5351 return (*dev->tx_pkt_prepare)(dev->data->tx_queues[queue_id], 5352 tx_pkts, nb_pkts); 5353 } 5354 5355 #else 5356 5357 /* 5358 * Native NOOP operation for compilation targets which doesn't require any 5359 * preparations steps, and functional NOOP may introduce unnecessary performance 5360 * drop. 5361 * 5362 * Generally this is not a good idea to turn it on globally and didn't should 5363 * be used if behavior of tx_preparation can change. 5364 */ 5365 5366 static inline uint16_t 5367 rte_eth_tx_prepare(__rte_unused uint16_t port_id, 5368 __rte_unused uint16_t queue_id, 5369 __rte_unused struct rte_mbuf **tx_pkts, uint16_t nb_pkts) 5370 { 5371 return nb_pkts; 5372 } 5373 5374 #endif 5375 5376 /** 5377 * Send any packets queued up for transmission on a port and HW queue 5378 * 5379 * This causes an explicit flush of packets previously buffered via the 5380 * rte_eth_tx_buffer() function. It returns the number of packets successfully 5381 * sent to the NIC, and calls the error callback for any unsent packets. Unless 5382 * explicitly set up otherwise, the default callback simply frees the unsent 5383 * packets back to the owning mempool. 5384 * 5385 * @param port_id 5386 * The port identifier of the Ethernet device. 5387 * @param queue_id 5388 * The index of the transmit queue through which output packets must be 5389 * sent. 5390 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5391 * to rte_eth_dev_configure(). 5392 * @param buffer 5393 * Buffer of packets to be transmit. 5394 * @return 5395 * The number of packets successfully sent to the Ethernet device. The error 5396 * callback is called for any packets which could not be sent. 5397 */ 5398 static inline uint16_t 5399 rte_eth_tx_buffer_flush(uint16_t port_id, uint16_t queue_id, 5400 struct rte_eth_dev_tx_buffer *buffer) 5401 { 5402 uint16_t sent; 5403 uint16_t to_send = buffer->length; 5404 5405 if (to_send == 0) 5406 return 0; 5407 5408 sent = rte_eth_tx_burst(port_id, queue_id, buffer->pkts, to_send); 5409 5410 buffer->length = 0; 5411 5412 /* All packets sent, or to be dealt with by callback below */ 5413 if (unlikely(sent != to_send)) 5414 buffer->error_callback(&buffer->pkts[sent], 5415 (uint16_t)(to_send - sent), 5416 buffer->error_userdata); 5417 5418 return sent; 5419 } 5420 5421 /** 5422 * Buffer a single packet for future transmission on a port and queue 5423 * 5424 * This function takes a single mbuf/packet and buffers it for later 5425 * transmission on the particular port and queue specified. Once the buffer is 5426 * full of packets, an attempt will be made to transmit all the buffered 5427 * packets. In case of error, where not all packets can be transmitted, a 5428 * callback is called with the unsent packets as a parameter. If no callback 5429 * is explicitly set up, the unsent packets are just freed back to the owning 5430 * mempool. The function returns the number of packets actually sent i.e. 5431 * 0 if no buffer flush occurred, otherwise the number of packets successfully 5432 * flushed 5433 * 5434 * @param port_id 5435 * The port identifier of the Ethernet device. 5436 * @param queue_id 5437 * The index of the transmit queue through which output packets must be 5438 * sent. 5439 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5440 * to rte_eth_dev_configure(). 5441 * @param buffer 5442 * Buffer used to collect packets to be sent. 5443 * @param tx_pkt 5444 * Pointer to the packet mbuf to be sent. 5445 * @return 5446 * 0 = packet has been buffered for later transmission 5447 * N > 0 = packet has been buffered, and the buffer was subsequently flushed, 5448 * causing N packets to be sent, and the error callback to be called for 5449 * the rest. 5450 */ 5451 static __rte_always_inline uint16_t 5452 rte_eth_tx_buffer(uint16_t port_id, uint16_t queue_id, 5453 struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt) 5454 { 5455 buffer->pkts[buffer->length++] = tx_pkt; 5456 if (buffer->length < buffer->size) 5457 return 0; 5458 5459 return rte_eth_tx_buffer_flush(port_id, queue_id, buffer); 5460 } 5461 5462 #ifdef __cplusplus 5463 } 5464 #endif 5465 5466 #endif /* _RTE_ETHDEV_H_ */ 5467