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_bitops.h> 163 #include <rte_errno.h> 164 #include <rte_common.h> 165 #include <rte_config.h> 166 #include <rte_ether.h> 167 #include <rte_power_intrinsics.h> 168 169 #include "rte_ethdev_trace_fp.h" 170 #include "rte_dev_info.h" 171 172 extern int rte_eth_dev_logtype; 173 174 #define RTE_ETHDEV_LOG(level, ...) \ 175 rte_log(RTE_LOG_ ## level, rte_eth_dev_logtype, "" __VA_ARGS__) 176 177 struct rte_mbuf; 178 179 /** 180 * Initializes a device iterator. 181 * 182 * This iterator allows accessing a list of devices matching some devargs. 183 * 184 * @param iter 185 * Device iterator handle initialized by the function. 186 * The fields bus_str and cls_str might be dynamically allocated, 187 * and could be freed by calling rte_eth_iterator_cleanup(). 188 * 189 * @param devargs 190 * Device description string. 191 * 192 * @return 193 * 0 on successful initialization, negative otherwise. 194 */ 195 int rte_eth_iterator_init(struct rte_dev_iterator *iter, const char *devargs); 196 197 /** 198 * Iterates on devices with devargs filter. 199 * The ownership is not checked. 200 * 201 * The next port ID is returned, and the iterator is updated. 202 * 203 * @param iter 204 * Device iterator handle initialized by rte_eth_iterator_init(). 205 * Some fields bus_str and cls_str might be freed when no more port is found, 206 * by calling rte_eth_iterator_cleanup(). 207 * 208 * @return 209 * A port ID if found, RTE_MAX_ETHPORTS otherwise. 210 */ 211 uint16_t rte_eth_iterator_next(struct rte_dev_iterator *iter); 212 213 /** 214 * Free some allocated fields of the iterator. 215 * 216 * This function is automatically called by rte_eth_iterator_next() 217 * on the last iteration (i.e. when no more matching port is found). 218 * 219 * It is safe to call this function twice; it will do nothing more. 220 * 221 * @param iter 222 * Device iterator handle initialized by rte_eth_iterator_init(). 223 * The fields bus_str and cls_str are freed if needed. 224 */ 225 void rte_eth_iterator_cleanup(struct rte_dev_iterator *iter); 226 227 /** 228 * Macro to iterate over all ethdev ports matching some devargs. 229 * 230 * If a break is done before the end of the loop, 231 * the function rte_eth_iterator_cleanup() must be called. 232 * 233 * @param id 234 * Iterated port ID of type uint16_t. 235 * @param devargs 236 * Device parameters input as string of type char*. 237 * @param iter 238 * Iterator handle of type struct rte_dev_iterator, used internally. 239 */ 240 #define RTE_ETH_FOREACH_MATCHING_DEV(id, devargs, iter) \ 241 for (rte_eth_iterator_init(iter, devargs), \ 242 id = rte_eth_iterator_next(iter); \ 243 id != RTE_MAX_ETHPORTS; \ 244 id = rte_eth_iterator_next(iter)) 245 246 /** 247 * A structure used to retrieve statistics for an Ethernet port. 248 * Not all statistics fields in struct rte_eth_stats are supported 249 * by any type of network interface card (NIC). If any statistics 250 * field is not supported, its value is 0. 251 * All byte-related statistics do not include Ethernet FCS regardless 252 * of whether these bytes have been delivered to the application 253 * (see DEV_RX_OFFLOAD_KEEP_CRC). 254 */ 255 struct rte_eth_stats { 256 uint64_t ipackets; /**< Total number of successfully received packets. */ 257 uint64_t opackets; /**< Total number of successfully transmitted packets.*/ 258 uint64_t ibytes; /**< Total number of successfully received bytes. */ 259 uint64_t obytes; /**< Total number of successfully transmitted bytes. */ 260 /** 261 * Total of Rx packets dropped by the HW, 262 * because there are no available buffer (i.e. Rx queues are full). 263 */ 264 uint64_t imissed; 265 uint64_t ierrors; /**< Total number of erroneous received packets. */ 266 uint64_t oerrors; /**< Total number of failed transmitted packets. */ 267 uint64_t rx_nombuf; /**< Total number of Rx mbuf allocation failures. */ 268 /* Queue stats are limited to max 256 queues */ 269 /** Total number of queue Rx packets. */ 270 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 271 /** Total number of queue Tx packets. */ 272 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 273 /** Total number of successfully received queue bytes. */ 274 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 275 /** Total number of successfully transmitted queue bytes. */ 276 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 277 /** Total number of queue packets received that are dropped. */ 278 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS]; 279 }; 280 281 /**@{@name Link speed capabilities 282 * Device supported speeds bitmap flags 283 */ 284 #define ETH_LINK_SPEED_AUTONEG 0 /**< Autonegotiate (all speeds) */ 285 #define ETH_LINK_SPEED_FIXED RTE_BIT32(0) /**< Disable autoneg (fixed speed) */ 286 #define ETH_LINK_SPEED_10M_HD RTE_BIT32(1) /**< 10 Mbps half-duplex */ 287 #define ETH_LINK_SPEED_10M RTE_BIT32(2) /**< 10 Mbps full-duplex */ 288 #define ETH_LINK_SPEED_100M_HD RTE_BIT32(3) /**< 100 Mbps half-duplex */ 289 #define ETH_LINK_SPEED_100M RTE_BIT32(4) /**< 100 Mbps full-duplex */ 290 #define ETH_LINK_SPEED_1G RTE_BIT32(5) /**< 1 Gbps */ 291 #define ETH_LINK_SPEED_2_5G RTE_BIT32(6) /**< 2.5 Gbps */ 292 #define ETH_LINK_SPEED_5G RTE_BIT32(7) /**< 5 Gbps */ 293 #define ETH_LINK_SPEED_10G RTE_BIT32(8) /**< 10 Gbps */ 294 #define ETH_LINK_SPEED_20G RTE_BIT32(9) /**< 20 Gbps */ 295 #define ETH_LINK_SPEED_25G RTE_BIT32(10) /**< 25 Gbps */ 296 #define ETH_LINK_SPEED_40G RTE_BIT32(11) /**< 40 Gbps */ 297 #define ETH_LINK_SPEED_50G RTE_BIT32(12) /**< 50 Gbps */ 298 #define ETH_LINK_SPEED_56G RTE_BIT32(13) /**< 56 Gbps */ 299 #define ETH_LINK_SPEED_100G RTE_BIT32(14) /**< 100 Gbps */ 300 #define ETH_LINK_SPEED_200G RTE_BIT32(15) /**< 200 Gbps */ 301 /**@}*/ 302 303 /**@{@name Link speed 304 * Ethernet numeric link speeds in Mbps 305 */ 306 #define ETH_SPEED_NUM_NONE 0 /**< Not defined */ 307 #define ETH_SPEED_NUM_10M 10 /**< 10 Mbps */ 308 #define ETH_SPEED_NUM_100M 100 /**< 100 Mbps */ 309 #define ETH_SPEED_NUM_1G 1000 /**< 1 Gbps */ 310 #define ETH_SPEED_NUM_2_5G 2500 /**< 2.5 Gbps */ 311 #define ETH_SPEED_NUM_5G 5000 /**< 5 Gbps */ 312 #define ETH_SPEED_NUM_10G 10000 /**< 10 Gbps */ 313 #define ETH_SPEED_NUM_20G 20000 /**< 20 Gbps */ 314 #define ETH_SPEED_NUM_25G 25000 /**< 25 Gbps */ 315 #define ETH_SPEED_NUM_40G 40000 /**< 40 Gbps */ 316 #define ETH_SPEED_NUM_50G 50000 /**< 50 Gbps */ 317 #define ETH_SPEED_NUM_56G 56000 /**< 56 Gbps */ 318 #define ETH_SPEED_NUM_100G 100000 /**< 100 Gbps */ 319 #define ETH_SPEED_NUM_200G 200000 /**< 200 Gbps */ 320 #define ETH_SPEED_NUM_UNKNOWN UINT32_MAX /**< Unknown */ 321 /**@}*/ 322 323 /** 324 * A structure used to retrieve link-level information of an Ethernet port. 325 */ 326 __extension__ 327 struct rte_eth_link { 328 uint32_t link_speed; /**< ETH_SPEED_NUM_ */ 329 uint16_t link_duplex : 1; /**< ETH_LINK_[HALF/FULL]_DUPLEX */ 330 uint16_t link_autoneg : 1; /**< ETH_LINK_[AUTONEG/FIXED] */ 331 uint16_t link_status : 1; /**< ETH_LINK_[DOWN/UP] */ 332 } __rte_aligned(8); /**< aligned for atomic64 read/write */ 333 334 /**@{@name Link negotiation 335 * Constants used in link management. 336 */ 337 #define ETH_LINK_HALF_DUPLEX 0 /**< Half-duplex connection (see link_duplex). */ 338 #define ETH_LINK_FULL_DUPLEX 1 /**< Full-duplex connection (see link_duplex). */ 339 #define ETH_LINK_DOWN 0 /**< Link is down (see link_status). */ 340 #define ETH_LINK_UP 1 /**< Link is up (see link_status). */ 341 #define ETH_LINK_FIXED 0 /**< No autonegotiation (see link_autoneg). */ 342 #define ETH_LINK_AUTONEG 1 /**< Autonegotiated (see link_autoneg). */ 343 #define RTE_ETH_LINK_MAX_STR_LEN 40 /**< Max length of default link string. */ 344 /**@}*/ 345 346 /** 347 * A structure used to configure the ring threshold registers of an Rx/Tx 348 * queue for an Ethernet port. 349 */ 350 struct rte_eth_thresh { 351 uint8_t pthresh; /**< Ring prefetch threshold. */ 352 uint8_t hthresh; /**< Ring host threshold. */ 353 uint8_t wthresh; /**< Ring writeback threshold. */ 354 }; 355 356 /**@{@name Multi-queue mode 357 * @see rte_eth_conf.rxmode.mq_mode. 358 */ 359 #define ETH_MQ_RX_RSS_FLAG 0x1 /**< Enable RSS. @see rte_eth_rss_conf */ 360 #define ETH_MQ_RX_DCB_FLAG 0x2 /**< Enable DCB. */ 361 #define ETH_MQ_RX_VMDQ_FLAG 0x4 /**< Enable VMDq. */ 362 /**@}*/ 363 364 /** 365 * A set of values to identify what method is to be used to route 366 * packets to multiple queues. 367 */ 368 enum rte_eth_rx_mq_mode { 369 /** None of DCB, RSS or VMDq mode */ 370 ETH_MQ_RX_NONE = 0, 371 372 /** For Rx side, only RSS is on */ 373 ETH_MQ_RX_RSS = ETH_MQ_RX_RSS_FLAG, 374 /** For Rx side,only DCB is on. */ 375 ETH_MQ_RX_DCB = ETH_MQ_RX_DCB_FLAG, 376 /** Both DCB and RSS enable */ 377 ETH_MQ_RX_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG, 378 379 /** Only VMDq, no RSS nor DCB */ 380 ETH_MQ_RX_VMDQ_ONLY = ETH_MQ_RX_VMDQ_FLAG, 381 /** RSS mode with VMDq */ 382 ETH_MQ_RX_VMDQ_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_VMDQ_FLAG, 383 /** Use VMDq+DCB to route traffic to queues */ 384 ETH_MQ_RX_VMDQ_DCB = ETH_MQ_RX_VMDQ_FLAG | ETH_MQ_RX_DCB_FLAG, 385 /** Enable both VMDq and DCB in VMDq */ 386 ETH_MQ_RX_VMDQ_DCB_RSS = ETH_MQ_RX_RSS_FLAG | ETH_MQ_RX_DCB_FLAG | 387 ETH_MQ_RX_VMDQ_FLAG, 388 }; 389 390 /** 391 * for Rx mq mode backward compatible 392 */ 393 #define ETH_RSS ETH_MQ_RX_RSS 394 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB 395 #define ETH_DCB_RX ETH_MQ_RX_DCB 396 397 /** 398 * A set of values to identify what method is to be used to transmit 399 * packets using multi-TCs. 400 */ 401 enum rte_eth_tx_mq_mode { 402 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */ 403 ETH_MQ_TX_DCB, /**< For Tx side,only DCB is on. */ 404 ETH_MQ_TX_VMDQ_DCB, /**< For Tx side,both DCB and VT is on. */ 405 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */ 406 }; 407 408 /** 409 * for Tx mq mode backward compatible 410 */ 411 #define ETH_DCB_NONE ETH_MQ_TX_NONE 412 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB 413 #define ETH_DCB_TX ETH_MQ_TX_DCB 414 415 /** 416 * A structure used to configure the Rx features of an Ethernet port. 417 */ 418 struct rte_eth_rxmode { 419 /** The multi-queue packet distribution mode to be used, e.g. RSS. */ 420 enum rte_eth_rx_mq_mode mq_mode; 421 uint32_t mtu; /**< Requested MTU. */ 422 /** Maximum allowed size of LRO aggregated packet. */ 423 uint32_t max_lro_pkt_size; 424 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/ 425 /** 426 * Per-port Rx offloads to be set using DEV_RX_OFFLOAD_* flags. 427 * Only offloads set on rx_offload_capa field on rte_eth_dev_info 428 * structure are allowed to be set. 429 */ 430 uint64_t offloads; 431 432 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 433 void *reserved_ptrs[2]; /**< Reserved for future fields */ 434 }; 435 436 /** 437 * VLAN types to indicate if it is for single VLAN, inner VLAN or outer VLAN. 438 * Note that single VLAN is treated the same as inner VLAN. 439 */ 440 enum rte_vlan_type { 441 ETH_VLAN_TYPE_UNKNOWN = 0, 442 ETH_VLAN_TYPE_INNER, /**< Inner VLAN. */ 443 ETH_VLAN_TYPE_OUTER, /**< Single VLAN, or outer VLAN. */ 444 ETH_VLAN_TYPE_MAX, 445 }; 446 447 /** 448 * A structure used to describe a VLAN filter. 449 * If the bit corresponding to a VID is set, such VID is on. 450 */ 451 struct rte_vlan_filter_conf { 452 uint64_t ids[64]; 453 }; 454 455 /** 456 * A structure used to configure the Receive Side Scaling (RSS) feature 457 * of an Ethernet port. 458 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points 459 * to an array holding the RSS key to use for hashing specific header 460 * fields of received packets. The length of this array should be indicated 461 * by *rss_key_len* below. Otherwise, a default random hash key is used by 462 * the device driver. 463 * 464 * The *rss_key_len* field of the *rss_conf* structure indicates the length 465 * in bytes of the array pointed by *rss_key*. To be compatible, this length 466 * will be checked in i40e only. Others assume 40 bytes to be used as before. 467 * 468 * The *rss_hf* field of the *rss_conf* structure indicates the different 469 * types of IPv4/IPv6 packets to which the RSS hashing must be applied. 470 * Supplying an *rss_hf* equal to zero disables the RSS feature. 471 */ 472 struct rte_eth_rss_conf { 473 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */ 474 uint8_t rss_key_len; /**< hash key length in bytes. */ 475 uint64_t rss_hf; /**< Hash functions to apply - see below. */ 476 }; 477 478 /* 479 * A packet can be identified by hardware as different flow types. Different 480 * NIC hardware may support different flow types. 481 * Basically, the NIC hardware identifies the flow type as deep protocol as 482 * possible, and exclusively. For example, if a packet is identified as 483 * 'RTE_ETH_FLOW_NONFRAG_IPV4_TCP', it will not be any of other flow types, 484 * though it is an actual IPV4 packet. 485 */ 486 #define RTE_ETH_FLOW_UNKNOWN 0 487 #define RTE_ETH_FLOW_RAW 1 488 #define RTE_ETH_FLOW_IPV4 2 489 #define RTE_ETH_FLOW_FRAG_IPV4 3 490 #define RTE_ETH_FLOW_NONFRAG_IPV4_TCP 4 491 #define RTE_ETH_FLOW_NONFRAG_IPV4_UDP 5 492 #define RTE_ETH_FLOW_NONFRAG_IPV4_SCTP 6 493 #define RTE_ETH_FLOW_NONFRAG_IPV4_OTHER 7 494 #define RTE_ETH_FLOW_IPV6 8 495 #define RTE_ETH_FLOW_FRAG_IPV6 9 496 #define RTE_ETH_FLOW_NONFRAG_IPV6_TCP 10 497 #define RTE_ETH_FLOW_NONFRAG_IPV6_UDP 11 498 #define RTE_ETH_FLOW_NONFRAG_IPV6_SCTP 12 499 #define RTE_ETH_FLOW_NONFRAG_IPV6_OTHER 13 500 #define RTE_ETH_FLOW_L2_PAYLOAD 14 501 #define RTE_ETH_FLOW_IPV6_EX 15 502 #define RTE_ETH_FLOW_IPV6_TCP_EX 16 503 #define RTE_ETH_FLOW_IPV6_UDP_EX 17 504 /** Consider device port number as a flow differentiator */ 505 #define RTE_ETH_FLOW_PORT 18 506 #define RTE_ETH_FLOW_VXLAN 19 /**< VXLAN protocol based flow */ 507 #define RTE_ETH_FLOW_GENEVE 20 /**< GENEVE protocol based flow */ 508 #define RTE_ETH_FLOW_NVGRE 21 /**< NVGRE protocol based flow */ 509 #define RTE_ETH_FLOW_VXLAN_GPE 22 /**< VXLAN-GPE protocol based flow */ 510 #define RTE_ETH_FLOW_GTPU 23 /**< GTPU protocol based flow */ 511 #define RTE_ETH_FLOW_MAX 24 512 513 /* 514 * Below macros are defined for RSS offload types, they can be used to 515 * fill rte_eth_rss_conf.rss_hf or rte_flow_action_rss.types. 516 */ 517 #define ETH_RSS_IPV4 RTE_BIT64(2) 518 #define ETH_RSS_FRAG_IPV4 RTE_BIT64(3) 519 #define ETH_RSS_NONFRAG_IPV4_TCP RTE_BIT64(4) 520 #define ETH_RSS_NONFRAG_IPV4_UDP RTE_BIT64(5) 521 #define ETH_RSS_NONFRAG_IPV4_SCTP RTE_BIT64(6) 522 #define ETH_RSS_NONFRAG_IPV4_OTHER RTE_BIT64(7) 523 #define ETH_RSS_IPV6 RTE_BIT64(8) 524 #define ETH_RSS_FRAG_IPV6 RTE_BIT64(9) 525 #define ETH_RSS_NONFRAG_IPV6_TCP RTE_BIT64(10) 526 #define ETH_RSS_NONFRAG_IPV6_UDP RTE_BIT64(11) 527 #define ETH_RSS_NONFRAG_IPV6_SCTP RTE_BIT64(12) 528 #define ETH_RSS_NONFRAG_IPV6_OTHER RTE_BIT64(13) 529 #define ETH_RSS_L2_PAYLOAD RTE_BIT64(14) 530 #define ETH_RSS_IPV6_EX RTE_BIT64(15) 531 #define ETH_RSS_IPV6_TCP_EX RTE_BIT64(16) 532 #define ETH_RSS_IPV6_UDP_EX RTE_BIT64(17) 533 #define ETH_RSS_PORT RTE_BIT64(18) 534 #define ETH_RSS_VXLAN RTE_BIT64(19) 535 #define ETH_RSS_GENEVE RTE_BIT64(20) 536 #define ETH_RSS_NVGRE RTE_BIT64(21) 537 #define ETH_RSS_GTPU RTE_BIT64(23) 538 #define ETH_RSS_ETH RTE_BIT64(24) 539 #define ETH_RSS_S_VLAN RTE_BIT64(25) 540 #define ETH_RSS_C_VLAN RTE_BIT64(26) 541 #define ETH_RSS_ESP RTE_BIT64(27) 542 #define ETH_RSS_AH RTE_BIT64(28) 543 #define ETH_RSS_L2TPV3 RTE_BIT64(29) 544 #define ETH_RSS_PFCP RTE_BIT64(30) 545 #define ETH_RSS_PPPOE RTE_BIT64(31) 546 #define ETH_RSS_ECPRI RTE_BIT64(32) 547 #define ETH_RSS_MPLS RTE_BIT64(33) 548 #define ETH_RSS_IPV4_CHKSUM RTE_BIT64(34) 549 550 /** 551 * The ETH_RSS_L4_CHKSUM works on checksum field of any L4 header. 552 * It is similar to ETH_RSS_PORT that they don't specify the specific type of 553 * L4 header. This macro is defined to replace some specific L4 (TCP/UDP/SCTP) 554 * checksum type for constructing the use of RSS offload bits. 555 * 556 * Due to above reason, some old APIs (and configuration) don't support 557 * ETH_RSS_L4_CHKSUM. The rte_flow RSS API supports it. 558 * 559 * For the case that checksum is not used in an UDP header, 560 * it takes the reserved value 0 as input for the hash function. 561 */ 562 #define ETH_RSS_L4_CHKSUM RTE_BIT64(35) 563 564 /* 565 * We use the following macros to combine with above ETH_RSS_* for 566 * more specific input set selection. These bits are defined starting 567 * from the high end of the 64 bits. 568 * Note: If we use above ETH_RSS_* without SRC/DST_ONLY, it represents 569 * both SRC and DST are taken into account. If SRC_ONLY and DST_ONLY of 570 * the same level are used simultaneously, it is the same case as none of 571 * them are added. 572 */ 573 #define ETH_RSS_L3_SRC_ONLY RTE_BIT64(63) 574 #define ETH_RSS_L3_DST_ONLY RTE_BIT64(62) 575 #define ETH_RSS_L4_SRC_ONLY RTE_BIT64(61) 576 #define ETH_RSS_L4_DST_ONLY RTE_BIT64(60) 577 #define ETH_RSS_L2_SRC_ONLY RTE_BIT64(59) 578 #define ETH_RSS_L2_DST_ONLY RTE_BIT64(58) 579 580 /* 581 * Only select IPV6 address prefix as RSS input set according to 582 * https://tools.ietf.org/html/rfc6052 583 * Must be combined with ETH_RSS_IPV6, ETH_RSS_NONFRAG_IPV6_UDP, 584 * ETH_RSS_NONFRAG_IPV6_TCP, ETH_RSS_NONFRAG_IPV6_SCTP. 585 */ 586 #define RTE_ETH_RSS_L3_PRE32 RTE_BIT64(57) 587 #define RTE_ETH_RSS_L3_PRE40 RTE_BIT64(56) 588 #define RTE_ETH_RSS_L3_PRE48 RTE_BIT64(55) 589 #define RTE_ETH_RSS_L3_PRE56 RTE_BIT64(54) 590 #define RTE_ETH_RSS_L3_PRE64 RTE_BIT64(53) 591 #define RTE_ETH_RSS_L3_PRE96 RTE_BIT64(52) 592 593 /* 594 * Use the following macros to combine with the above layers 595 * to choose inner and outer layers or both for RSS computation. 596 * Bits 50 and 51 are reserved for this. 597 */ 598 599 /** 600 * level 0, requests the default behavior. 601 * Depending on the packet type, it can mean outermost, innermost, 602 * anything in between or even no RSS. 603 * It basically stands for the innermost encapsulation level RSS 604 * can be performed on according to PMD and device capabilities. 605 */ 606 #define ETH_RSS_LEVEL_PMD_DEFAULT (0ULL << 50) 607 608 /** 609 * level 1, requests RSS to be performed on the outermost packet 610 * encapsulation level. 611 */ 612 #define ETH_RSS_LEVEL_OUTERMOST (1ULL << 50) 613 614 /** 615 * level 2, requests RSS to be performed on the specified inner packet 616 * encapsulation level, from outermost to innermost (lower to higher values). 617 */ 618 #define ETH_RSS_LEVEL_INNERMOST (2ULL << 50) 619 #define ETH_RSS_LEVEL_MASK (3ULL << 50) 620 621 #define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50) 622 623 /** 624 * For input set change of hash filter, if SRC_ONLY and DST_ONLY of 625 * the same level are used simultaneously, it is the same case as 626 * none of them are added. 627 * 628 * @param rss_hf 629 * RSS types with SRC/DST_ONLY. 630 * @return 631 * RSS types. 632 */ 633 static inline uint64_t 634 rte_eth_rss_hf_refine(uint64_t rss_hf) 635 { 636 if ((rss_hf & ETH_RSS_L3_SRC_ONLY) && (rss_hf & ETH_RSS_L3_DST_ONLY)) 637 rss_hf &= ~(ETH_RSS_L3_SRC_ONLY | ETH_RSS_L3_DST_ONLY); 638 639 if ((rss_hf & ETH_RSS_L4_SRC_ONLY) && (rss_hf & ETH_RSS_L4_DST_ONLY)) 640 rss_hf &= ~(ETH_RSS_L4_SRC_ONLY | ETH_RSS_L4_DST_ONLY); 641 642 return rss_hf; 643 } 644 645 #define ETH_RSS_IPV6_PRE32 ( \ 646 ETH_RSS_IPV6 | \ 647 RTE_ETH_RSS_L3_PRE32) 648 649 #define ETH_RSS_IPV6_PRE40 ( \ 650 ETH_RSS_IPV6 | \ 651 RTE_ETH_RSS_L3_PRE40) 652 653 #define ETH_RSS_IPV6_PRE48 ( \ 654 ETH_RSS_IPV6 | \ 655 RTE_ETH_RSS_L3_PRE48) 656 657 #define ETH_RSS_IPV6_PRE56 ( \ 658 ETH_RSS_IPV6 | \ 659 RTE_ETH_RSS_L3_PRE56) 660 661 #define ETH_RSS_IPV6_PRE64 ( \ 662 ETH_RSS_IPV6 | \ 663 RTE_ETH_RSS_L3_PRE64) 664 665 #define ETH_RSS_IPV6_PRE96 ( \ 666 ETH_RSS_IPV6 | \ 667 RTE_ETH_RSS_L3_PRE96) 668 669 #define ETH_RSS_IPV6_PRE32_UDP ( \ 670 ETH_RSS_NONFRAG_IPV6_UDP | \ 671 RTE_ETH_RSS_L3_PRE32) 672 673 #define ETH_RSS_IPV6_PRE40_UDP ( \ 674 ETH_RSS_NONFRAG_IPV6_UDP | \ 675 RTE_ETH_RSS_L3_PRE40) 676 677 #define ETH_RSS_IPV6_PRE48_UDP ( \ 678 ETH_RSS_NONFRAG_IPV6_UDP | \ 679 RTE_ETH_RSS_L3_PRE48) 680 681 #define ETH_RSS_IPV6_PRE56_UDP ( \ 682 ETH_RSS_NONFRAG_IPV6_UDP | \ 683 RTE_ETH_RSS_L3_PRE56) 684 685 #define ETH_RSS_IPV6_PRE64_UDP ( \ 686 ETH_RSS_NONFRAG_IPV6_UDP | \ 687 RTE_ETH_RSS_L3_PRE64) 688 689 #define ETH_RSS_IPV6_PRE96_UDP ( \ 690 ETH_RSS_NONFRAG_IPV6_UDP | \ 691 RTE_ETH_RSS_L3_PRE96) 692 693 #define ETH_RSS_IPV6_PRE32_TCP ( \ 694 ETH_RSS_NONFRAG_IPV6_TCP | \ 695 RTE_ETH_RSS_L3_PRE32) 696 697 #define ETH_RSS_IPV6_PRE40_TCP ( \ 698 ETH_RSS_NONFRAG_IPV6_TCP | \ 699 RTE_ETH_RSS_L3_PRE40) 700 701 #define ETH_RSS_IPV6_PRE48_TCP ( \ 702 ETH_RSS_NONFRAG_IPV6_TCP | \ 703 RTE_ETH_RSS_L3_PRE48) 704 705 #define ETH_RSS_IPV6_PRE56_TCP ( \ 706 ETH_RSS_NONFRAG_IPV6_TCP | \ 707 RTE_ETH_RSS_L3_PRE56) 708 709 #define ETH_RSS_IPV6_PRE64_TCP ( \ 710 ETH_RSS_NONFRAG_IPV6_TCP | \ 711 RTE_ETH_RSS_L3_PRE64) 712 713 #define ETH_RSS_IPV6_PRE96_TCP ( \ 714 ETH_RSS_NONFRAG_IPV6_TCP | \ 715 RTE_ETH_RSS_L3_PRE96) 716 717 #define ETH_RSS_IPV6_PRE32_SCTP ( \ 718 ETH_RSS_NONFRAG_IPV6_SCTP | \ 719 RTE_ETH_RSS_L3_PRE32) 720 721 #define ETH_RSS_IPV6_PRE40_SCTP ( \ 722 ETH_RSS_NONFRAG_IPV6_SCTP | \ 723 RTE_ETH_RSS_L3_PRE40) 724 725 #define ETH_RSS_IPV6_PRE48_SCTP ( \ 726 ETH_RSS_NONFRAG_IPV6_SCTP | \ 727 RTE_ETH_RSS_L3_PRE48) 728 729 #define ETH_RSS_IPV6_PRE56_SCTP ( \ 730 ETH_RSS_NONFRAG_IPV6_SCTP | \ 731 RTE_ETH_RSS_L3_PRE56) 732 733 #define ETH_RSS_IPV6_PRE64_SCTP ( \ 734 ETH_RSS_NONFRAG_IPV6_SCTP | \ 735 RTE_ETH_RSS_L3_PRE64) 736 737 #define ETH_RSS_IPV6_PRE96_SCTP ( \ 738 ETH_RSS_NONFRAG_IPV6_SCTP | \ 739 RTE_ETH_RSS_L3_PRE96) 740 741 #define ETH_RSS_IP ( \ 742 ETH_RSS_IPV4 | \ 743 ETH_RSS_FRAG_IPV4 | \ 744 ETH_RSS_NONFRAG_IPV4_OTHER | \ 745 ETH_RSS_IPV6 | \ 746 ETH_RSS_FRAG_IPV6 | \ 747 ETH_RSS_NONFRAG_IPV6_OTHER | \ 748 ETH_RSS_IPV6_EX) 749 750 #define ETH_RSS_UDP ( \ 751 ETH_RSS_NONFRAG_IPV4_UDP | \ 752 ETH_RSS_NONFRAG_IPV6_UDP | \ 753 ETH_RSS_IPV6_UDP_EX) 754 755 #define ETH_RSS_TCP ( \ 756 ETH_RSS_NONFRAG_IPV4_TCP | \ 757 ETH_RSS_NONFRAG_IPV6_TCP | \ 758 ETH_RSS_IPV6_TCP_EX) 759 760 #define ETH_RSS_SCTP ( \ 761 ETH_RSS_NONFRAG_IPV4_SCTP | \ 762 ETH_RSS_NONFRAG_IPV6_SCTP) 763 764 #define ETH_RSS_TUNNEL ( \ 765 ETH_RSS_VXLAN | \ 766 ETH_RSS_GENEVE | \ 767 ETH_RSS_NVGRE) 768 769 #define ETH_RSS_VLAN ( \ 770 ETH_RSS_S_VLAN | \ 771 ETH_RSS_C_VLAN) 772 773 /** Mask of valid RSS hash protocols */ 774 #define ETH_RSS_PROTO_MASK ( \ 775 ETH_RSS_IPV4 | \ 776 ETH_RSS_FRAG_IPV4 | \ 777 ETH_RSS_NONFRAG_IPV4_TCP | \ 778 ETH_RSS_NONFRAG_IPV4_UDP | \ 779 ETH_RSS_NONFRAG_IPV4_SCTP | \ 780 ETH_RSS_NONFRAG_IPV4_OTHER | \ 781 ETH_RSS_IPV6 | \ 782 ETH_RSS_FRAG_IPV6 | \ 783 ETH_RSS_NONFRAG_IPV6_TCP | \ 784 ETH_RSS_NONFRAG_IPV6_UDP | \ 785 ETH_RSS_NONFRAG_IPV6_SCTP | \ 786 ETH_RSS_NONFRAG_IPV6_OTHER | \ 787 ETH_RSS_L2_PAYLOAD | \ 788 ETH_RSS_IPV6_EX | \ 789 ETH_RSS_IPV6_TCP_EX | \ 790 ETH_RSS_IPV6_UDP_EX | \ 791 ETH_RSS_PORT | \ 792 ETH_RSS_VXLAN | \ 793 ETH_RSS_GENEVE | \ 794 ETH_RSS_NVGRE | \ 795 ETH_RSS_MPLS) 796 797 /* 798 * Definitions used for redirection table entry size. 799 * Some RSS RETA sizes may not be supported by some drivers, check the 800 * documentation or the description of relevant functions for more details. 801 */ 802 #define ETH_RSS_RETA_SIZE_64 64 803 #define ETH_RSS_RETA_SIZE_128 128 804 #define ETH_RSS_RETA_SIZE_256 256 805 #define ETH_RSS_RETA_SIZE_512 512 806 #define RTE_RETA_GROUP_SIZE 64 807 808 /**@{@name VMDq and DCB maximums */ 809 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDq VLAN filters. */ 810 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */ 811 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDq DCB queues. */ 812 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */ 813 /**@}*/ 814 815 /**@{@name DCB capabilities */ 816 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */ 817 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */ 818 /**@}*/ 819 820 /**@{@name VLAN offload bits */ 821 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */ 822 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */ 823 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */ 824 #define ETH_QINQ_STRIP_OFFLOAD 0x0008 /**< QINQ Strip On/Off */ 825 826 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */ 827 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/ 828 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/ 829 #define ETH_QINQ_STRIP_MASK 0x0008 /**< QINQ Strip setting mask */ 830 #define ETH_VLAN_ID_MAX 0x0FFF /**< VLAN ID is in lower 12 bits*/ 831 /**@}*/ 832 833 /* Definitions used for receive MAC address */ 834 #define ETH_NUM_RECEIVE_MAC_ADDR 128 /**< Maximum nb. of receive mac addr. */ 835 836 /* Definitions used for unicast hash */ 837 #define ETH_VMDQ_NUM_UC_HASH_ARRAY 128 /**< Maximum nb. of UC hash array. */ 838 839 /**@{@name VMDq Rx mode 840 * @see rte_eth_vmdq_rx_conf.rx_mode 841 */ 842 #define ETH_VMDQ_ACCEPT_UNTAG 0x0001 /**< accept untagged packets. */ 843 #define ETH_VMDQ_ACCEPT_HASH_MC 0x0002 /**< accept packets in multicast table . */ 844 #define ETH_VMDQ_ACCEPT_HASH_UC 0x0004 /**< accept packets in unicast table. */ 845 #define ETH_VMDQ_ACCEPT_BROADCAST 0x0008 /**< accept broadcast packets. */ 846 #define ETH_VMDQ_ACCEPT_MULTICAST 0x0010 /**< multicast promiscuous. */ 847 /**@}*/ 848 849 /** 850 * A structure used to configure 64 entries of Redirection Table of the 851 * Receive Side Scaling (RSS) feature of an Ethernet port. To configure 852 * more than 64 entries supported by hardware, an array of this structure 853 * is needed. 854 */ 855 struct rte_eth_rss_reta_entry64 { 856 /** Mask bits indicate which entries need to be updated/queried. */ 857 uint64_t mask; 858 /** Group of 64 redirection table entries. */ 859 uint16_t reta[RTE_RETA_GROUP_SIZE]; 860 }; 861 862 /** 863 * This enum indicates the possible number of traffic classes 864 * in DCB configurations 865 */ 866 enum rte_eth_nb_tcs { 867 ETH_4_TCS = 4, /**< 4 TCs with DCB. */ 868 ETH_8_TCS = 8 /**< 8 TCs with DCB. */ 869 }; 870 871 /** 872 * This enum indicates the possible number of queue pools 873 * in VMDq configurations. 874 */ 875 enum rte_eth_nb_pools { 876 ETH_8_POOLS = 8, /**< 8 VMDq pools. */ 877 ETH_16_POOLS = 16, /**< 16 VMDq pools. */ 878 ETH_32_POOLS = 32, /**< 32 VMDq pools. */ 879 ETH_64_POOLS = 64 /**< 64 VMDq pools. */ 880 }; 881 882 /* This structure may be extended in future. */ 883 struct rte_eth_dcb_rx_conf { 884 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */ 885 /** Traffic class each UP mapped to. */ 886 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 887 }; 888 889 struct rte_eth_vmdq_dcb_tx_conf { 890 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */ 891 /** Traffic class each UP mapped to. */ 892 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 893 }; 894 895 struct rte_eth_dcb_tx_conf { 896 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs. */ 897 /** Traffic class each UP mapped to. */ 898 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 899 }; 900 901 struct rte_eth_vmdq_tx_conf { 902 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq mode, 64 pools. */ 903 }; 904 905 /** 906 * A structure used to configure the VMDq+DCB feature 907 * of an Ethernet port. 908 * 909 * Using this feature, packets are routed to a pool of queues, based 910 * on the VLAN ID in the VLAN tag, and then to a specific queue within 911 * that pool, using the user priority VLAN tag field. 912 * 913 * A default pool may be used, if desired, to route all traffic which 914 * does not match the VLAN filter rules. 915 */ 916 struct rte_eth_vmdq_dcb_conf { 917 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */ 918 uint8_t enable_default_pool; /**< If non-zero, use a default pool */ 919 uint8_t default_pool; /**< The default pool, if applicable */ 920 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */ 921 struct { 922 uint16_t vlan_id; /**< The VLAN ID of the received frame */ 923 uint64_t pools; /**< Bitmask of pools for packet Rx */ 924 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq VLAN pool maps. */ 925 /** Selects a queue in a pool */ 926 uint8_t dcb_tc[ETH_DCB_NUM_USER_PRIORITIES]; 927 }; 928 929 /** 930 * A structure used to configure the VMDq feature of an Ethernet port when 931 * not combined with the DCB feature. 932 * 933 * Using this feature, packets are routed to a pool of queues. By default, 934 * the pool selection is based on the MAC address, the VLAN ID in the 935 * VLAN tag as specified in the pool_map array. 936 * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool 937 * selection using only the MAC address. MAC address to pool mapping is done 938 * using the rte_eth_dev_mac_addr_add function, with the pool parameter 939 * corresponding to the pool ID. 940 * 941 * Queue selection within the selected pool will be done using RSS when 942 * it is enabled or revert to the first queue of the pool if not. 943 * 944 * A default pool may be used, if desired, to route all traffic which 945 * does not match the VLAN filter rules or any pool MAC address. 946 */ 947 struct rte_eth_vmdq_rx_conf { 948 enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */ 949 uint8_t enable_default_pool; /**< If non-zero, use a default pool */ 950 uint8_t default_pool; /**< The default pool, if applicable */ 951 uint8_t enable_loop_back; /**< Enable VT loop back */ 952 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */ 953 uint32_t rx_mode; /**< Flags from ETH_VMDQ_ACCEPT_* */ 954 struct { 955 uint16_t vlan_id; /**< The VLAN ID of the received frame */ 956 uint64_t pools; /**< Bitmask of pools for packet Rx */ 957 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq VLAN pool maps. */ 958 }; 959 960 /** 961 * A structure used to configure the Tx features of an Ethernet port. 962 */ 963 struct rte_eth_txmode { 964 enum rte_eth_tx_mq_mode mq_mode; /**< Tx multi-queues mode. */ 965 /** 966 * Per-port Tx offloads to be set using DEV_TX_OFFLOAD_* flags. 967 * Only offloads set on tx_offload_capa field on rte_eth_dev_info 968 * structure are allowed to be set. 969 */ 970 uint64_t offloads; 971 972 uint16_t pvid; 973 __extension__ 974 uint8_t /** If set, reject sending out tagged pkts */ 975 hw_vlan_reject_tagged : 1, 976 /** If set, reject sending out untagged pkts */ 977 hw_vlan_reject_untagged : 1, 978 /** If set, enable port based VLAN insertion */ 979 hw_vlan_insert_pvid : 1; 980 981 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 982 void *reserved_ptrs[2]; /**< Reserved for future fields */ 983 }; 984 985 /** 986 * @warning 987 * @b EXPERIMENTAL: this structure may change without prior notice. 988 * 989 * A structure used to configure an Rx packet segment to split. 990 * 991 * If RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT flag is set in offloads field, 992 * the PMD will split the received packets into multiple segments 993 * according to the specification in the description array: 994 * 995 * - The first network buffer will be allocated from the memory pool, 996 * specified in the first array element, the second buffer, from the 997 * pool in the second element, and so on. 998 * 999 * - The offsets from the segment description elements specify 1000 * the data offset from the buffer beginning except the first mbuf. 1001 * The first segment offset is added with RTE_PKTMBUF_HEADROOM. 1002 * 1003 * - The lengths in the elements define the maximal data amount 1004 * being received to each segment. The receiving starts with filling 1005 * up the first mbuf data buffer up to specified length. If the 1006 * there are data remaining (packet is longer than buffer in the first 1007 * mbuf) the following data will be pushed to the next segment 1008 * up to its own length, and so on. 1009 * 1010 * - If the length in the segment description element is zero 1011 * the actual buffer size will be deduced from the appropriate 1012 * memory pool properties. 1013 * 1014 * - If there is not enough elements to describe the buffer for entire 1015 * packet of maximal length the following parameters will be used 1016 * for the all remaining segments: 1017 * - pool from the last valid element 1018 * - the buffer size from this pool 1019 * - zero offset 1020 */ 1021 struct rte_eth_rxseg_split { 1022 struct rte_mempool *mp; /**< Memory pool to allocate segment from. */ 1023 uint16_t length; /**< Segment data length, configures split point. */ 1024 uint16_t offset; /**< Data offset from beginning of mbuf data buffer. */ 1025 uint32_t reserved; /**< Reserved field. */ 1026 }; 1027 1028 /** 1029 * @warning 1030 * @b EXPERIMENTAL: this structure may change without prior notice. 1031 * 1032 * A common structure used to describe Rx packet segment properties. 1033 */ 1034 union rte_eth_rxseg { 1035 /* The settings for buffer split offload. */ 1036 struct rte_eth_rxseg_split split; 1037 /* The other features settings should be added here. */ 1038 }; 1039 1040 /** 1041 * A structure used to configure an Rx ring of an Ethernet port. 1042 */ 1043 struct rte_eth_rxconf { 1044 struct rte_eth_thresh rx_thresh; /**< Rx ring threshold registers. */ 1045 uint16_t rx_free_thresh; /**< Drives the freeing of Rx descriptors. */ 1046 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */ 1047 uint8_t rx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */ 1048 uint16_t rx_nseg; /**< Number of descriptions in rx_seg array. */ 1049 /** 1050 * Share group index in Rx domain and switch domain. 1051 * Non-zero value to enable Rx queue share, zero value disable share. 1052 * PMD is responsible for Rx queue consistency checks to avoid member 1053 * port's configuration contradict to each other. 1054 */ 1055 uint16_t share_group; 1056 uint16_t share_qid; /**< Shared Rx queue ID in group */ 1057 /** 1058 * Per-queue Rx offloads to be set using DEV_RX_OFFLOAD_* flags. 1059 * Only offloads set on rx_queue_offload_capa or rx_offload_capa 1060 * fields on rte_eth_dev_info structure are allowed to be set. 1061 */ 1062 uint64_t offloads; 1063 /** 1064 * Points to the array of segment descriptions for an entire packet. 1065 * Array elements are properties for consecutive Rx segments. 1066 * 1067 * The supported capabilities of receiving segmentation is reported 1068 * in rte_eth_dev_info.rx_seg_capa field. 1069 */ 1070 union rte_eth_rxseg *rx_seg; 1071 1072 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 1073 void *reserved_ptrs[2]; /**< Reserved for future fields */ 1074 }; 1075 1076 /** 1077 * A structure used to configure a Tx ring of an Ethernet port. 1078 */ 1079 struct rte_eth_txconf { 1080 struct rte_eth_thresh tx_thresh; /**< Tx ring threshold registers. */ 1081 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */ 1082 uint16_t tx_free_thresh; /**< Start freeing Tx buffers if there are 1083 less free descriptors than this value. */ 1084 1085 uint8_t tx_deferred_start; /**< Do not start queue with rte_eth_dev_start(). */ 1086 /** 1087 * Per-queue Tx offloads to be set using DEV_TX_OFFLOAD_* flags. 1088 * Only offloads set on tx_queue_offload_capa or tx_offload_capa 1089 * fields on rte_eth_dev_info structure are allowed to be set. 1090 */ 1091 uint64_t offloads; 1092 1093 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 1094 void *reserved_ptrs[2]; /**< Reserved for future fields */ 1095 }; 1096 1097 /** 1098 * @warning 1099 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 1100 * 1101 * A structure used to return the hairpin capabilities that are supported. 1102 */ 1103 struct rte_eth_hairpin_cap { 1104 /** The max number of hairpin queues (different bindings). */ 1105 uint16_t max_nb_queues; 1106 /** Max number of Rx queues to be connected to one Tx queue. */ 1107 uint16_t max_rx_2_tx; 1108 /** Max number of Tx queues to be connected to one Rx queue. */ 1109 uint16_t max_tx_2_rx; 1110 uint16_t max_nb_desc; /**< The max num of descriptors. */ 1111 }; 1112 1113 #define RTE_ETH_MAX_HAIRPIN_PEERS 32 1114 1115 /** 1116 * @warning 1117 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 1118 * 1119 * A structure used to hold hairpin peer data. 1120 */ 1121 struct rte_eth_hairpin_peer { 1122 uint16_t port; /**< Peer port. */ 1123 uint16_t queue; /**< Peer queue. */ 1124 }; 1125 1126 /** 1127 * @warning 1128 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 1129 * 1130 * A structure used to configure hairpin binding. 1131 */ 1132 struct rte_eth_hairpin_conf { 1133 uint32_t peer_count:16; /**< The number of peers. */ 1134 1135 /** 1136 * Explicit Tx flow rule mode. 1137 * One hairpin pair of queues should have the same attribute. 1138 * 1139 * - When set, the user should be responsible for inserting the hairpin 1140 * Tx part flows and removing them. 1141 * - When clear, the PMD will try to handle the Tx part of the flows, 1142 * e.g., by splitting one flow into two parts. 1143 */ 1144 uint32_t tx_explicit:1; 1145 1146 /** 1147 * Manually bind hairpin queues. 1148 * One hairpin pair of queues should have the same attribute. 1149 * 1150 * - When set, to enable hairpin, the user should call the hairpin bind 1151 * function after all the queues are set up properly and the ports are 1152 * started. Also, the hairpin unbind function should be called 1153 * accordingly before stopping a port that with hairpin configured. 1154 * - When clear, the PMD will try to enable the hairpin with the queues 1155 * configured automatically during port start. 1156 */ 1157 uint32_t manual_bind:1; 1158 uint32_t reserved:14; /**< Reserved bits. */ 1159 struct rte_eth_hairpin_peer peers[RTE_ETH_MAX_HAIRPIN_PEERS]; 1160 }; 1161 1162 /** 1163 * A structure contains information about HW descriptor ring limitations. 1164 */ 1165 struct rte_eth_desc_lim { 1166 uint16_t nb_max; /**< Max allowed number of descriptors. */ 1167 uint16_t nb_min; /**< Min allowed number of descriptors. */ 1168 uint16_t nb_align; /**< Number of descriptors should be aligned to. */ 1169 1170 /** 1171 * Max allowed number of segments per whole packet. 1172 * 1173 * - For TSO packet this is the total number of data descriptors allowed 1174 * by device. 1175 * 1176 * @see nb_mtu_seg_max 1177 */ 1178 uint16_t nb_seg_max; 1179 1180 /** 1181 * Max number of segments per one MTU. 1182 * 1183 * - For non-TSO packet, this is the maximum allowed number of segments 1184 * in a single transmit packet. 1185 * 1186 * - For TSO packet each segment within the TSO may span up to this 1187 * value. 1188 * 1189 * @see nb_seg_max 1190 */ 1191 uint16_t nb_mtu_seg_max; 1192 }; 1193 1194 /** 1195 * This enum indicates the flow control mode 1196 */ 1197 enum rte_eth_fc_mode { 1198 RTE_FC_NONE = 0, /**< Disable flow control. */ 1199 RTE_FC_RX_PAUSE, /**< Rx pause frame, enable flowctrl on Tx side. */ 1200 RTE_FC_TX_PAUSE, /**< Tx pause frame, enable flowctrl on Rx side. */ 1201 RTE_FC_FULL /**< Enable flow control on both side. */ 1202 }; 1203 1204 /** 1205 * A structure used to configure Ethernet flow control parameter. 1206 * These parameters will be configured into the register of the NIC. 1207 * Please refer to the corresponding data sheet for proper value. 1208 */ 1209 struct rte_eth_fc_conf { 1210 uint32_t high_water; /**< High threshold value to trigger XOFF */ 1211 uint32_t low_water; /**< Low threshold value to trigger XON */ 1212 uint16_t pause_time; /**< Pause quota in the Pause frame */ 1213 uint16_t send_xon; /**< Is XON frame need be sent */ 1214 enum rte_eth_fc_mode mode; /**< Link flow control mode */ 1215 uint8_t mac_ctrl_frame_fwd; /**< Forward MAC control frames */ 1216 uint8_t autoneg; /**< Use Pause autoneg */ 1217 }; 1218 1219 /** 1220 * A structure used to configure Ethernet priority flow control parameter. 1221 * These parameters will be configured into the register of the NIC. 1222 * Please refer to the corresponding data sheet for proper value. 1223 */ 1224 struct rte_eth_pfc_conf { 1225 struct rte_eth_fc_conf fc; /**< General flow control parameter. */ 1226 uint8_t priority; /**< VLAN User Priority. */ 1227 }; 1228 1229 /** 1230 * Tunnel type for device-specific classifier configuration. 1231 * @see rte_eth_udp_tunnel 1232 */ 1233 enum rte_eth_tunnel_type { 1234 RTE_TUNNEL_TYPE_NONE = 0, 1235 RTE_TUNNEL_TYPE_VXLAN, 1236 RTE_TUNNEL_TYPE_GENEVE, 1237 RTE_TUNNEL_TYPE_TEREDO, 1238 RTE_TUNNEL_TYPE_NVGRE, 1239 RTE_TUNNEL_TYPE_IP_IN_GRE, 1240 RTE_L2_TUNNEL_TYPE_E_TAG, 1241 RTE_TUNNEL_TYPE_VXLAN_GPE, 1242 RTE_TUNNEL_TYPE_ECPRI, 1243 RTE_TUNNEL_TYPE_MAX, 1244 }; 1245 1246 /* Deprecated API file for rte_eth_dev_filter_* functions */ 1247 #include "rte_eth_ctrl.h" 1248 1249 /** 1250 * Memory space that can be configured to store Flow Director filters 1251 * in the board memory. 1252 */ 1253 enum rte_fdir_pballoc_type { 1254 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */ 1255 RTE_FDIR_PBALLOC_128K, /**< 128k. */ 1256 RTE_FDIR_PBALLOC_256K, /**< 256k. */ 1257 }; 1258 1259 /** 1260 * Select report mode of FDIR hash information in Rx descriptors. 1261 */ 1262 enum rte_fdir_status_mode { 1263 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */ 1264 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */ 1265 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */ 1266 }; 1267 1268 /** 1269 * A structure used to configure the Flow Director (FDIR) feature 1270 * of an Ethernet port. 1271 * 1272 * If mode is RTE_FDIR_MODE_NONE, the pballoc value is ignored. 1273 */ 1274 struct rte_fdir_conf { 1275 enum rte_fdir_mode mode; /**< Flow Director mode. */ 1276 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */ 1277 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */ 1278 /** Rx queue of packets matching a "drop" filter in perfect mode. */ 1279 uint8_t drop_queue; 1280 struct rte_eth_fdir_masks mask; 1281 /** Flex payload configuration. */ 1282 struct rte_eth_fdir_flex_conf flex_conf; 1283 }; 1284 1285 /** 1286 * UDP tunneling configuration. 1287 * 1288 * Used to configure the classifier of a device, 1289 * associating an UDP port with a type of tunnel. 1290 * 1291 * Some NICs may need such configuration to properly parse a tunnel 1292 * with any standard or custom UDP port. 1293 */ 1294 struct rte_eth_udp_tunnel { 1295 uint16_t udp_port; /**< UDP port used for the tunnel. */ 1296 uint8_t prot_type; /**< Tunnel type. @see rte_eth_tunnel_type */ 1297 }; 1298 1299 /** 1300 * A structure used to enable/disable specific device interrupts. 1301 */ 1302 struct rte_intr_conf { 1303 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */ 1304 uint32_t lsc:1; 1305 /** enable/disable rxq interrupt. 0 (default) - disable, 1 enable */ 1306 uint32_t rxq:1; 1307 /** enable/disable rmv interrupt. 0 (default) - disable, 1 enable */ 1308 uint32_t rmv:1; 1309 }; 1310 1311 /** 1312 * A structure used to configure an Ethernet port. 1313 * Depending upon the Rx multi-queue mode, extra advanced 1314 * configuration settings may be needed. 1315 */ 1316 struct rte_eth_conf { 1317 uint32_t link_speeds; /**< bitmap of ETH_LINK_SPEED_XXX of speeds to be 1318 used. ETH_LINK_SPEED_FIXED disables link 1319 autonegotiation, and a unique speed shall be 1320 set. Otherwise, the bitmap defines the set of 1321 speeds to be advertised. If the special value 1322 ETH_LINK_SPEED_AUTONEG (0) is used, all speeds 1323 supported are advertised. */ 1324 struct rte_eth_rxmode rxmode; /**< Port Rx configuration. */ 1325 struct rte_eth_txmode txmode; /**< Port Tx configuration. */ 1326 uint32_t lpbk_mode; /**< Loopback operation mode. By default the value 1327 is 0, meaning the loopback mode is disabled. 1328 Read the datasheet of given Ethernet controller 1329 for details. The possible values of this field 1330 are defined in implementation of each driver. */ 1331 struct { 1332 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */ 1333 /** Port VMDq+DCB configuration. */ 1334 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf; 1335 /** Port DCB Rx configuration. */ 1336 struct rte_eth_dcb_rx_conf dcb_rx_conf; 1337 /** Port VMDq Rx configuration. */ 1338 struct rte_eth_vmdq_rx_conf vmdq_rx_conf; 1339 } rx_adv_conf; /**< Port Rx filtering configuration. */ 1340 union { 1341 /** Port VMDq+DCB Tx configuration. */ 1342 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf; 1343 /** Port DCB Tx configuration. */ 1344 struct rte_eth_dcb_tx_conf dcb_tx_conf; 1345 /** Port VMDq Tx configuration. */ 1346 struct rte_eth_vmdq_tx_conf vmdq_tx_conf; 1347 } tx_adv_conf; /**< Port Tx DCB configuration (union). */ 1348 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC 1349 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */ 1350 uint32_t dcb_capability_en; 1351 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. DEPRECATED */ 1352 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */ 1353 }; 1354 1355 /** 1356 * Rx offload capabilities of a device. 1357 */ 1358 #define DEV_RX_OFFLOAD_VLAN_STRIP 0x00000001 1359 #define DEV_RX_OFFLOAD_IPV4_CKSUM 0x00000002 1360 #define DEV_RX_OFFLOAD_UDP_CKSUM 0x00000004 1361 #define DEV_RX_OFFLOAD_TCP_CKSUM 0x00000008 1362 #define DEV_RX_OFFLOAD_TCP_LRO 0x00000010 1363 #define DEV_RX_OFFLOAD_QINQ_STRIP 0x00000020 1364 #define DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000040 1365 #define DEV_RX_OFFLOAD_MACSEC_STRIP 0x00000080 1366 #define DEV_RX_OFFLOAD_HEADER_SPLIT 0x00000100 1367 #define DEV_RX_OFFLOAD_VLAN_FILTER 0x00000200 1368 #define DEV_RX_OFFLOAD_VLAN_EXTEND 0x00000400 1369 #define DEV_RX_OFFLOAD_SCATTER 0x00002000 1370 /** 1371 * Timestamp is set by the driver in RTE_MBUF_DYNFIELD_TIMESTAMP_NAME 1372 * and RTE_MBUF_DYNFLAG_RX_TIMESTAMP_NAME is set in ol_flags. 1373 * The mbuf field and flag are registered when the offload is configured. 1374 */ 1375 #define DEV_RX_OFFLOAD_TIMESTAMP 0x00004000 1376 #define DEV_RX_OFFLOAD_SECURITY 0x00008000 1377 #define DEV_RX_OFFLOAD_KEEP_CRC 0x00010000 1378 #define DEV_RX_OFFLOAD_SCTP_CKSUM 0x00020000 1379 #define DEV_RX_OFFLOAD_OUTER_UDP_CKSUM 0x00040000 1380 #define DEV_RX_OFFLOAD_RSS_HASH 0x00080000 1381 #define RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT 0x00100000 1382 1383 #define DEV_RX_OFFLOAD_CHECKSUM (DEV_RX_OFFLOAD_IPV4_CKSUM | \ 1384 DEV_RX_OFFLOAD_UDP_CKSUM | \ 1385 DEV_RX_OFFLOAD_TCP_CKSUM) 1386 #define DEV_RX_OFFLOAD_VLAN (DEV_RX_OFFLOAD_VLAN_STRIP | \ 1387 DEV_RX_OFFLOAD_VLAN_FILTER | \ 1388 DEV_RX_OFFLOAD_VLAN_EXTEND | \ 1389 DEV_RX_OFFLOAD_QINQ_STRIP) 1390 1391 /* 1392 * If new Rx offload capabilities are defined, they also must be 1393 * mentioned in rte_rx_offload_names in rte_ethdev.c file. 1394 */ 1395 1396 /** 1397 * Tx offload capabilities of a device. 1398 */ 1399 #define DEV_TX_OFFLOAD_VLAN_INSERT 0x00000001 1400 #define DEV_TX_OFFLOAD_IPV4_CKSUM 0x00000002 1401 #define DEV_TX_OFFLOAD_UDP_CKSUM 0x00000004 1402 #define DEV_TX_OFFLOAD_TCP_CKSUM 0x00000008 1403 #define DEV_TX_OFFLOAD_SCTP_CKSUM 0x00000010 1404 #define DEV_TX_OFFLOAD_TCP_TSO 0x00000020 1405 #define DEV_TX_OFFLOAD_UDP_TSO 0x00000040 1406 #define DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM 0x00000080 /**< Used for tunneling packet. */ 1407 #define DEV_TX_OFFLOAD_QINQ_INSERT 0x00000100 1408 #define DEV_TX_OFFLOAD_VXLAN_TNL_TSO 0x00000200 /**< Used for tunneling packet. */ 1409 #define DEV_TX_OFFLOAD_GRE_TNL_TSO 0x00000400 /**< Used for tunneling packet. */ 1410 #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */ 1411 #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */ 1412 #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000 1413 /** 1414 * Multiple threads can invoke rte_eth_tx_burst() concurrently on the same 1415 * Tx queue without SW lock. 1416 */ 1417 #define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000 1418 /** Device supports multi segment send. */ 1419 #define DEV_TX_OFFLOAD_MULTI_SEGS 0x00008000 1420 /** 1421 * Device supports optimization for fast release of mbufs. 1422 * When set application must guarantee that per-queue all mbufs comes from 1423 * the same mempool and has refcnt = 1. 1424 */ 1425 #define DEV_TX_OFFLOAD_MBUF_FAST_FREE 0x00010000 1426 #define DEV_TX_OFFLOAD_SECURITY 0x00020000 1427 /** 1428 * Device supports generic UDP tunneled packet TSO. 1429 * Application must set PKT_TX_TUNNEL_UDP and other mbuf fields required 1430 * for tunnel TSO. 1431 */ 1432 #define DEV_TX_OFFLOAD_UDP_TNL_TSO 0x00040000 1433 /** 1434 * Device supports generic IP tunneled packet TSO. 1435 * Application must set PKT_TX_TUNNEL_IP and other mbuf fields required 1436 * for tunnel TSO. 1437 */ 1438 #define DEV_TX_OFFLOAD_IP_TNL_TSO 0x00080000 1439 /** Device supports outer UDP checksum */ 1440 #define DEV_TX_OFFLOAD_OUTER_UDP_CKSUM 0x00100000 1441 /** 1442 * Device sends on time read from RTE_MBUF_DYNFIELD_TIMESTAMP_NAME 1443 * if RTE_MBUF_DYNFLAG_TX_TIMESTAMP_NAME is set in ol_flags. 1444 * The mbuf field and flag are registered when the offload is configured. 1445 */ 1446 #define DEV_TX_OFFLOAD_SEND_ON_TIMESTAMP 0x00200000 1447 /* 1448 * If new Tx offload capabilities are defined, they also must be 1449 * mentioned in rte_tx_offload_names in rte_ethdev.c file. 1450 */ 1451 1452 /**@{@name Device capabilities 1453 * Non-offload capabilities reported in rte_eth_dev_info.dev_capa. 1454 */ 1455 /** Device supports Rx queue setup after device started. */ 1456 #define RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP 0x00000001 1457 /** Device supports Tx queue setup after device started. */ 1458 #define RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP 0x00000002 1459 /** 1460 * Device supports shared Rx queue among ports within Rx domain and 1461 * switch domain. Mbufs are consumed by shared Rx queue instead of 1462 * each queue. Multiple groups are supported by share_group of Rx 1463 * queue configuration. Shared Rx queue is identified by PMD using 1464 * share_qid of Rx queue configuration. Polling any port in the group 1465 * receive packets of all member ports, source port identified by 1466 * mbuf->port field. 1467 */ 1468 #define RTE_ETH_DEV_CAPA_RXQ_SHARE RTE_BIT64(2) 1469 /**@}*/ 1470 1471 /* 1472 * Fallback default preferred Rx/Tx port parameters. 1473 * These are used if an application requests default parameters 1474 * but the PMD does not provide preferred values. 1475 */ 1476 #define RTE_ETH_DEV_FALLBACK_RX_RINGSIZE 512 1477 #define RTE_ETH_DEV_FALLBACK_TX_RINGSIZE 512 1478 #define RTE_ETH_DEV_FALLBACK_RX_NBQUEUES 1 1479 #define RTE_ETH_DEV_FALLBACK_TX_NBQUEUES 1 1480 1481 /** 1482 * Preferred Rx/Tx port parameters. 1483 * There are separate instances of this structure for transmission 1484 * and reception respectively. 1485 */ 1486 struct rte_eth_dev_portconf { 1487 uint16_t burst_size; /**< Device-preferred burst size */ 1488 uint16_t ring_size; /**< Device-preferred size of queue rings */ 1489 uint16_t nb_queues; /**< Device-preferred number of queues */ 1490 }; 1491 1492 /** 1493 * Default values for switch domain ID when ethdev does not support switch 1494 * domain definitions. 1495 */ 1496 #define RTE_ETH_DEV_SWITCH_DOMAIN_ID_INVALID (UINT16_MAX) 1497 1498 /** 1499 * Ethernet device associated switch information 1500 */ 1501 struct rte_eth_switch_info { 1502 const char *name; /**< switch name */ 1503 uint16_t domain_id; /**< switch domain ID */ 1504 /** 1505 * Mapping to the devices physical switch port as enumerated from the 1506 * perspective of the embedded interconnect/switch. For SR-IOV enabled 1507 * device this may correspond to the VF_ID of each virtual function, 1508 * but each driver should explicitly define the mapping of switch 1509 * port identifier to that physical interconnect/switch 1510 */ 1511 uint16_t port_id; 1512 /** 1513 * Shared Rx queue sub-domain boundary. Only ports in same Rx domain 1514 * and switch domain can share Rx queue. Valid only if device advertised 1515 * RTE_ETH_DEV_CAPA_RXQ_SHARE capability. 1516 */ 1517 uint16_t rx_domain; 1518 }; 1519 1520 /** 1521 * @warning 1522 * @b EXPERIMENTAL: this structure may change without prior notice. 1523 * 1524 * Ethernet device Rx buffer segmentation capabilities. 1525 */ 1526 struct rte_eth_rxseg_capa { 1527 __extension__ 1528 uint32_t multi_pools:1; /**< Supports receiving to multiple pools.*/ 1529 uint32_t offset_allowed:1; /**< Supports buffer offsets. */ 1530 uint32_t offset_align_log2:4; /**< Required offset alignment. */ 1531 uint16_t max_nseg; /**< Maximum amount of segments to split. */ 1532 uint16_t reserved; /**< Reserved field. */ 1533 }; 1534 1535 /** 1536 * Ethernet device information 1537 */ 1538 1539 /** 1540 * Ethernet device representor port type. 1541 */ 1542 enum rte_eth_representor_type { 1543 RTE_ETH_REPRESENTOR_NONE, /**< not a representor. */ 1544 RTE_ETH_REPRESENTOR_VF, /**< representor of Virtual Function. */ 1545 RTE_ETH_REPRESENTOR_SF, /**< representor of Sub Function. */ 1546 RTE_ETH_REPRESENTOR_PF, /**< representor of Physical Function. */ 1547 }; 1548 1549 /** 1550 * A structure used to retrieve the contextual information of 1551 * an Ethernet device, such as the controlling driver of the 1552 * device, etc... 1553 */ 1554 struct rte_eth_dev_info { 1555 struct rte_device *device; /** Generic device information */ 1556 const char *driver_name; /**< Device Driver name. */ 1557 unsigned int if_index; /**< Index to bound host interface, or 0 if none. 1558 Use if_indextoname() to translate into an interface name. */ 1559 uint16_t min_mtu; /**< Minimum MTU allowed */ 1560 uint16_t max_mtu; /**< Maximum MTU allowed */ 1561 const uint32_t *dev_flags; /**< Device flags */ 1562 uint32_t min_rx_bufsize; /**< Minimum size of Rx buffer. */ 1563 uint32_t max_rx_pktlen; /**< Maximum configurable length of Rx pkt. */ 1564 /** Maximum configurable size of LRO aggregated packet. */ 1565 uint32_t max_lro_pkt_size; 1566 uint16_t max_rx_queues; /**< Maximum number of Rx queues. */ 1567 uint16_t max_tx_queues; /**< Maximum number of Tx queues. */ 1568 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */ 1569 uint32_t max_hash_mac_addrs; 1570 /** Maximum number of hash MAC addresses for MTA and UTA. */ 1571 uint16_t max_vfs; /**< Maximum number of VFs. */ 1572 uint16_t max_vmdq_pools; /**< Maximum number of VMDq pools. */ 1573 struct rte_eth_rxseg_capa rx_seg_capa; /**< Segmentation capability.*/ 1574 /** All Rx offload capabilities including all per-queue ones */ 1575 uint64_t rx_offload_capa; 1576 /** All Tx offload capabilities including all per-queue ones */ 1577 uint64_t tx_offload_capa; 1578 /** Device per-queue Rx offload capabilities. */ 1579 uint64_t rx_queue_offload_capa; 1580 /** Device per-queue Tx offload capabilities. */ 1581 uint64_t tx_queue_offload_capa; 1582 /** Device redirection table size, the total number of entries. */ 1583 uint16_t reta_size; 1584 uint8_t hash_key_size; /**< Hash key size in bytes */ 1585 /** Bit mask of RSS offloads, the bit offset also means flow type */ 1586 uint64_t flow_type_rss_offloads; 1587 struct rte_eth_rxconf default_rxconf; /**< Default Rx configuration */ 1588 struct rte_eth_txconf default_txconf; /**< Default Tx configuration */ 1589 uint16_t vmdq_queue_base; /**< First queue ID for VMDq pools. */ 1590 uint16_t vmdq_queue_num; /**< Queue number for VMDq pools. */ 1591 uint16_t vmdq_pool_base; /**< First ID of VMDq pools. */ 1592 struct rte_eth_desc_lim rx_desc_lim; /**< Rx descriptors limits */ 1593 struct rte_eth_desc_lim tx_desc_lim; /**< Tx descriptors limits */ 1594 uint32_t speed_capa; /**< Supported speeds bitmap (ETH_LINK_SPEED_). */ 1595 /** Configured number of Rx/Tx queues */ 1596 uint16_t nb_rx_queues; /**< Number of Rx queues. */ 1597 uint16_t nb_tx_queues; /**< Number of Tx queues. */ 1598 /** Rx parameter recommendations */ 1599 struct rte_eth_dev_portconf default_rxportconf; 1600 /** Tx parameter recommendations */ 1601 struct rte_eth_dev_portconf default_txportconf; 1602 /** Generic device capabilities (RTE_ETH_DEV_CAPA_). */ 1603 uint64_t dev_capa; 1604 /** 1605 * Switching information for ports on a device with a 1606 * embedded managed interconnect/switch. 1607 */ 1608 struct rte_eth_switch_info switch_info; 1609 1610 uint64_t reserved_64s[2]; /**< Reserved for future fields */ 1611 void *reserved_ptrs[2]; /**< Reserved for future fields */ 1612 }; 1613 1614 /**@{@name Rx/Tx queue states */ 1615 #define RTE_ETH_QUEUE_STATE_STOPPED 0 /**< Queue stopped. */ 1616 #define RTE_ETH_QUEUE_STATE_STARTED 1 /**< Queue started. */ 1617 #define RTE_ETH_QUEUE_STATE_HAIRPIN 2 /**< Queue used for hairpin. */ 1618 /**@}*/ 1619 1620 /** 1621 * Ethernet device Rx queue information structure. 1622 * Used to retrieve information about configured queue. 1623 */ 1624 struct rte_eth_rxq_info { 1625 struct rte_mempool *mp; /**< mempool used by that queue. */ 1626 struct rte_eth_rxconf conf; /**< queue config parameters. */ 1627 uint8_t scattered_rx; /**< scattered packets Rx supported. */ 1628 uint8_t queue_state; /**< one of RTE_ETH_QUEUE_STATE_*. */ 1629 uint16_t nb_desc; /**< configured number of RXDs. */ 1630 uint16_t rx_buf_size; /**< hardware receive buffer size. */ 1631 } __rte_cache_min_aligned; 1632 1633 /** 1634 * Ethernet device Tx queue information structure. 1635 * Used to retrieve information about configured queue. 1636 */ 1637 struct rte_eth_txq_info { 1638 struct rte_eth_txconf conf; /**< queue config parameters. */ 1639 uint16_t nb_desc; /**< configured number of TXDs. */ 1640 uint8_t queue_state; /**< one of RTE_ETH_QUEUE_STATE_*. */ 1641 } __rte_cache_min_aligned; 1642 1643 /* Generic Burst mode flag definition, values can be ORed. */ 1644 1645 /** 1646 * If the queues have different burst mode description, this bit will be set 1647 * by PMD, then the application can iterate to retrieve burst description for 1648 * all other queues. 1649 */ 1650 #define RTE_ETH_BURST_FLAG_PER_QUEUE RTE_BIT64(0) 1651 1652 /** 1653 * Ethernet device Rx/Tx queue packet burst mode information structure. 1654 * Used to retrieve information about packet burst mode setting. 1655 */ 1656 struct rte_eth_burst_mode { 1657 uint64_t flags; /**< The ORed values of RTE_ETH_BURST_FLAG_xxx */ 1658 1659 #define RTE_ETH_BURST_MODE_INFO_SIZE 1024 /**< Maximum size for information */ 1660 char info[RTE_ETH_BURST_MODE_INFO_SIZE]; /**< burst mode information */ 1661 }; 1662 1663 /** Maximum name length for extended statistics counters */ 1664 #define RTE_ETH_XSTATS_NAME_SIZE 64 1665 1666 /** 1667 * An Ethernet device extended statistic structure 1668 * 1669 * This structure is used by rte_eth_xstats_get() to provide 1670 * statistics that are not provided in the generic *rte_eth_stats* 1671 * structure. 1672 * It maps a name ID, corresponding to an index in the array returned 1673 * by rte_eth_xstats_get_names(), to a statistic value. 1674 */ 1675 struct rte_eth_xstat { 1676 uint64_t id; /**< The index in xstats name array. */ 1677 uint64_t value; /**< The statistic counter value. */ 1678 }; 1679 1680 /** 1681 * A name element for extended statistics. 1682 * 1683 * An array of this structure is returned by rte_eth_xstats_get_names(). 1684 * It lists the names of extended statistics for a PMD. The *rte_eth_xstat* 1685 * structure references these names by their array index. 1686 * 1687 * The xstats should follow a common naming scheme. 1688 * Some names are standardized in rte_stats_strings. 1689 * Examples: 1690 * - rx_missed_errors 1691 * - tx_q3_bytes 1692 * - tx_size_128_to_255_packets 1693 */ 1694 struct rte_eth_xstat_name { 1695 char name[RTE_ETH_XSTATS_NAME_SIZE]; /**< The statistic name. */ 1696 }; 1697 1698 #define ETH_DCB_NUM_TCS 8 1699 #define ETH_MAX_VMDQ_POOL 64 1700 1701 /** 1702 * A structure used to get the information of queue and 1703 * TC mapping on both Tx and Rx paths. 1704 */ 1705 struct rte_eth_dcb_tc_queue_mapping { 1706 /** Rx queues assigned to tc per Pool */ 1707 struct { 1708 uint16_t base; 1709 uint16_t nb_queue; 1710 } tc_rxq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS]; 1711 /** Rx queues assigned to tc per Pool */ 1712 struct { 1713 uint16_t base; 1714 uint16_t nb_queue; 1715 } tc_txq[ETH_MAX_VMDQ_POOL][ETH_DCB_NUM_TCS]; 1716 }; 1717 1718 /** 1719 * A structure used to get the information of DCB. 1720 * It includes TC UP mapping and queue TC mapping. 1721 */ 1722 struct rte_eth_dcb_info { 1723 uint8_t nb_tcs; /**< number of TCs */ 1724 uint8_t prio_tc[ETH_DCB_NUM_USER_PRIORITIES]; /**< Priority to tc */ 1725 uint8_t tc_bws[ETH_DCB_NUM_TCS]; /**< Tx BW percentage for each TC */ 1726 /** Rx queues assigned to tc */ 1727 struct rte_eth_dcb_tc_queue_mapping tc_queue; 1728 }; 1729 1730 /** 1731 * This enum indicates the possible Forward Error Correction (FEC) modes 1732 * of an ethdev port. 1733 */ 1734 enum rte_eth_fec_mode { 1735 RTE_ETH_FEC_NOFEC = 0, /**< FEC is off */ 1736 RTE_ETH_FEC_AUTO, /**< FEC autonegotiation modes */ 1737 RTE_ETH_FEC_BASER, /**< FEC using common algorithm */ 1738 RTE_ETH_FEC_RS, /**< FEC using RS algorithm */ 1739 }; 1740 1741 /* Translate from FEC mode to FEC capa */ 1742 #define RTE_ETH_FEC_MODE_TO_CAPA(x) RTE_BIT32(x) 1743 1744 /* This macro indicates FEC capa mask */ 1745 #define RTE_ETH_FEC_MODE_CAPA_MASK(x) RTE_BIT32(RTE_ETH_FEC_ ## x) 1746 1747 /* A structure used to get capabilities per link speed */ 1748 struct rte_eth_fec_capa { 1749 uint32_t speed; /**< Link speed (see ETH_SPEED_NUM_*) */ 1750 uint32_t capa; /**< FEC capabilities bitmask */ 1751 }; 1752 1753 #define RTE_ETH_ALL RTE_MAX_ETHPORTS 1754 1755 /* Macros to check for valid port */ 1756 #define RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, retval) do { \ 1757 if (!rte_eth_dev_is_valid_port(port_id)) { \ 1758 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \ 1759 return retval; \ 1760 } \ 1761 } while (0) 1762 1763 #define RTE_ETH_VALID_PORTID_OR_RET(port_id) do { \ 1764 if (!rte_eth_dev_is_valid_port(port_id)) { \ 1765 RTE_ETHDEV_LOG(ERR, "Invalid port_id=%u\n", port_id); \ 1766 return; \ 1767 } \ 1768 } while (0) 1769 1770 /**@{@name L2 tunnel configuration */ 1771 /** L2 tunnel enable mask */ 1772 #define ETH_L2_TUNNEL_ENABLE_MASK 0x00000001 1773 /** L2 tunnel insertion mask */ 1774 #define ETH_L2_TUNNEL_INSERTION_MASK 0x00000002 1775 /** L2 tunnel stripping mask */ 1776 #define ETH_L2_TUNNEL_STRIPPING_MASK 0x00000004 1777 /** L2 tunnel forwarding mask */ 1778 #define ETH_L2_TUNNEL_FORWARDING_MASK 0x00000008 1779 /**@}*/ 1780 1781 /** 1782 * Function type used for Rx packet processing packet callbacks. 1783 * 1784 * The callback function is called on Rx with a burst of packets that have 1785 * been received on the given port and queue. 1786 * 1787 * @param port_id 1788 * The Ethernet port on which Rx is being performed. 1789 * @param queue 1790 * The queue on the Ethernet port which is being used to receive the packets. 1791 * @param pkts 1792 * The burst of packets that have just been received. 1793 * @param nb_pkts 1794 * The number of packets in the burst pointed to by "pkts". 1795 * @param max_pkts 1796 * The max number of packets that can be stored in the "pkts" array. 1797 * @param user_param 1798 * The arbitrary user parameter passed in by the application when the callback 1799 * was originally configured. 1800 * @return 1801 * The number of packets returned to the user. 1802 */ 1803 typedef uint16_t (*rte_rx_callback_fn)(uint16_t port_id, uint16_t queue, 1804 struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t max_pkts, 1805 void *user_param); 1806 1807 /** 1808 * Function type used for Tx packet processing packet callbacks. 1809 * 1810 * The callback function is called on Tx with a burst of packets immediately 1811 * before the packets are put onto the hardware queue for transmission. 1812 * 1813 * @param port_id 1814 * The Ethernet port on which Tx is being performed. 1815 * @param queue 1816 * The queue on the Ethernet port which is being used to transmit the packets. 1817 * @param pkts 1818 * The burst of packets that are about to be transmitted. 1819 * @param nb_pkts 1820 * The number of packets in the burst pointed to by "pkts". 1821 * @param user_param 1822 * The arbitrary user parameter passed in by the application when the callback 1823 * was originally configured. 1824 * @return 1825 * The number of packets to be written to the NIC. 1826 */ 1827 typedef uint16_t (*rte_tx_callback_fn)(uint16_t port_id, uint16_t queue, 1828 struct rte_mbuf *pkts[], uint16_t nb_pkts, void *user_param); 1829 1830 /** 1831 * Possible states of an ethdev port. 1832 */ 1833 enum rte_eth_dev_state { 1834 /** Device is unused before being probed. */ 1835 RTE_ETH_DEV_UNUSED = 0, 1836 /** Device is attached when allocated in probing. */ 1837 RTE_ETH_DEV_ATTACHED, 1838 /** Device is in removed state when plug-out is detected. */ 1839 RTE_ETH_DEV_REMOVED, 1840 }; 1841 1842 struct rte_eth_dev_sriov { 1843 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */ 1844 uint8_t nb_q_per_pool; /**< Rx queue number per pool */ 1845 uint16_t def_vmdq_idx; /**< Default pool num used for PF */ 1846 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */ 1847 }; 1848 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov) 1849 1850 #define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN 1851 1852 #define RTE_ETH_DEV_NO_OWNER 0 1853 1854 #define RTE_ETH_MAX_OWNER_NAME_LEN 64 1855 1856 struct rte_eth_dev_owner { 1857 uint64_t id; /**< The owner unique identifier. */ 1858 char name[RTE_ETH_MAX_OWNER_NAME_LEN]; /**< The owner name. */ 1859 }; 1860 1861 /**@{@name Device flags 1862 * Flags internally saved in rte_eth_dev_data.dev_flags 1863 * and reported in rte_eth_dev_info.dev_flags. 1864 */ 1865 /** PMD supports thread-safe flow operations */ 1866 #define RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE 0x0001 1867 /** Device supports link state interrupt */ 1868 #define RTE_ETH_DEV_INTR_LSC 0x0002 1869 /** Device is a bonded slave */ 1870 #define RTE_ETH_DEV_BONDED_SLAVE 0x0004 1871 /** Device supports device removal interrupt */ 1872 #define RTE_ETH_DEV_INTR_RMV 0x0008 1873 /** Device is port representor */ 1874 #define RTE_ETH_DEV_REPRESENTOR 0x0010 1875 /** Device does not support MAC change after started */ 1876 #define RTE_ETH_DEV_NOLIVE_MAC_ADDR 0x0020 1877 /** 1878 * Queue xstats filled automatically by ethdev layer. 1879 * PMDs filling the queue xstats themselves should not set this flag 1880 */ 1881 #define RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS 0x0040 1882 /**@}*/ 1883 1884 /** 1885 * Iterates over valid ethdev ports owned by a specific owner. 1886 * 1887 * @param port_id 1888 * The ID of the next possible valid owned port. 1889 * @param owner_id 1890 * The owner identifier. 1891 * RTE_ETH_DEV_NO_OWNER means iterate over all valid ownerless ports. 1892 * @return 1893 * Next valid port ID owned by owner_id, RTE_MAX_ETHPORTS if there is none. 1894 */ 1895 uint64_t rte_eth_find_next_owned_by(uint16_t port_id, 1896 const uint64_t owner_id); 1897 1898 /** 1899 * Macro to iterate over all enabled ethdev ports owned by a specific owner. 1900 */ 1901 #define RTE_ETH_FOREACH_DEV_OWNED_BY(p, o) \ 1902 for (p = rte_eth_find_next_owned_by(0, o); \ 1903 (unsigned int)p < (unsigned int)RTE_MAX_ETHPORTS; \ 1904 p = rte_eth_find_next_owned_by(p + 1, o)) 1905 1906 /** 1907 * Iterates over valid ethdev ports. 1908 * 1909 * @param port_id 1910 * The ID of the next possible valid port. 1911 * @return 1912 * Next valid port ID, RTE_MAX_ETHPORTS if there is none. 1913 */ 1914 uint16_t rte_eth_find_next(uint16_t port_id); 1915 1916 /** 1917 * Macro to iterate over all enabled and ownerless ethdev ports. 1918 */ 1919 #define RTE_ETH_FOREACH_DEV(p) \ 1920 RTE_ETH_FOREACH_DEV_OWNED_BY(p, RTE_ETH_DEV_NO_OWNER) 1921 1922 /** 1923 * Iterates over ethdev ports of a specified device. 1924 * 1925 * @param port_id_start 1926 * The ID of the next possible valid port. 1927 * @param parent 1928 * The generic device behind the ports to iterate. 1929 * @return 1930 * Next port ID of the device, possibly port_id_start, 1931 * RTE_MAX_ETHPORTS if there is none. 1932 */ 1933 uint16_t 1934 rte_eth_find_next_of(uint16_t port_id_start, 1935 const struct rte_device *parent); 1936 1937 /** 1938 * Macro to iterate over all ethdev ports of a specified device. 1939 * 1940 * @param port_id 1941 * The ID of the matching port being iterated. 1942 * @param parent 1943 * The rte_device pointer matching the iterated ports. 1944 */ 1945 #define RTE_ETH_FOREACH_DEV_OF(port_id, parent) \ 1946 for (port_id = rte_eth_find_next_of(0, parent); \ 1947 port_id < RTE_MAX_ETHPORTS; \ 1948 port_id = rte_eth_find_next_of(port_id + 1, parent)) 1949 1950 /** 1951 * Iterates over sibling ethdev ports (i.e. sharing the same rte_device). 1952 * 1953 * @param port_id_start 1954 * The ID of the next possible valid sibling port. 1955 * @param ref_port_id 1956 * The ID of a reference port to compare rte_device with. 1957 * @return 1958 * Next sibling port ID, possibly port_id_start or ref_port_id itself, 1959 * RTE_MAX_ETHPORTS if there is none. 1960 */ 1961 uint16_t 1962 rte_eth_find_next_sibling(uint16_t port_id_start, uint16_t ref_port_id); 1963 1964 /** 1965 * Macro to iterate over all ethdev ports sharing the same rte_device 1966 * as the specified port. 1967 * Note: the specified reference port is part of the loop iterations. 1968 * 1969 * @param port_id 1970 * The ID of the matching port being iterated. 1971 * @param ref_port_id 1972 * The ID of the port being compared. 1973 */ 1974 #define RTE_ETH_FOREACH_DEV_SIBLING(port_id, ref_port_id) \ 1975 for (port_id = rte_eth_find_next_sibling(0, ref_port_id); \ 1976 port_id < RTE_MAX_ETHPORTS; \ 1977 port_id = rte_eth_find_next_sibling(port_id + 1, ref_port_id)) 1978 1979 /** 1980 * @warning 1981 * @b EXPERIMENTAL: this API may change without prior notice. 1982 * 1983 * Get a new unique owner identifier. 1984 * An owner identifier is used to owns Ethernet devices by only one DPDK entity 1985 * to avoid multiple management of device by different entities. 1986 * 1987 * @param owner_id 1988 * Owner identifier pointer. 1989 * @return 1990 * Negative errno value on error, 0 on success. 1991 */ 1992 __rte_experimental 1993 int rte_eth_dev_owner_new(uint64_t *owner_id); 1994 1995 /** 1996 * @warning 1997 * @b EXPERIMENTAL: this API may change without prior notice. 1998 * 1999 * Set an Ethernet device owner. 2000 * 2001 * @param port_id 2002 * The identifier of the port to own. 2003 * @param owner 2004 * The owner pointer. 2005 * @return 2006 * Negative errno value on error, 0 on success. 2007 */ 2008 __rte_experimental 2009 int rte_eth_dev_owner_set(const uint16_t port_id, 2010 const struct rte_eth_dev_owner *owner); 2011 2012 /** 2013 * @warning 2014 * @b EXPERIMENTAL: this API may change without prior notice. 2015 * 2016 * Unset Ethernet device owner to make the device ownerless. 2017 * 2018 * @param port_id 2019 * The identifier of port to make ownerless. 2020 * @param owner_id 2021 * The owner identifier. 2022 * @return 2023 * 0 on success, negative errno value on error. 2024 */ 2025 __rte_experimental 2026 int rte_eth_dev_owner_unset(const uint16_t port_id, 2027 const uint64_t owner_id); 2028 2029 /** 2030 * @warning 2031 * @b EXPERIMENTAL: this API may change without prior notice. 2032 * 2033 * Remove owner from all Ethernet devices owned by a specific owner. 2034 * 2035 * @param owner_id 2036 * The owner identifier. 2037 * @return 2038 * 0 on success, negative errno value on error. 2039 */ 2040 __rte_experimental 2041 int rte_eth_dev_owner_delete(const uint64_t owner_id); 2042 2043 /** 2044 * @warning 2045 * @b EXPERIMENTAL: this API may change without prior notice. 2046 * 2047 * Get the owner of an Ethernet device. 2048 * 2049 * @param port_id 2050 * The port identifier. 2051 * @param owner 2052 * The owner structure pointer to fill. 2053 * @return 2054 * 0 on success, negative errno value on error.. 2055 */ 2056 __rte_experimental 2057 int rte_eth_dev_owner_get(const uint16_t port_id, 2058 struct rte_eth_dev_owner *owner); 2059 2060 /** 2061 * Get the number of ports which are usable for the application. 2062 * 2063 * These devices must be iterated by using the macro 2064 * ``RTE_ETH_FOREACH_DEV`` or ``RTE_ETH_FOREACH_DEV_OWNED_BY`` 2065 * to deal with non-contiguous ranges of devices. 2066 * 2067 * @return 2068 * The count of available Ethernet devices. 2069 */ 2070 uint16_t rte_eth_dev_count_avail(void); 2071 2072 /** 2073 * Get the total number of ports which are allocated. 2074 * 2075 * Some devices may not be available for the application. 2076 * 2077 * @return 2078 * The total count of Ethernet devices. 2079 */ 2080 uint16_t rte_eth_dev_count_total(void); 2081 2082 /** 2083 * Convert a numerical speed in Mbps to a bitmap flag that can be used in 2084 * the bitmap link_speeds of the struct rte_eth_conf 2085 * 2086 * @param speed 2087 * Numerical speed value in Mbps 2088 * @param duplex 2089 * ETH_LINK_[HALF/FULL]_DUPLEX (only for 10/100M speeds) 2090 * @return 2091 * 0 if the speed cannot be mapped 2092 */ 2093 uint32_t rte_eth_speed_bitflag(uint32_t speed, int duplex); 2094 2095 /** 2096 * Get DEV_RX_OFFLOAD_* flag name. 2097 * 2098 * @param offload 2099 * Offload flag. 2100 * @return 2101 * Offload name or 'UNKNOWN' if the flag cannot be recognised. 2102 */ 2103 const char *rte_eth_dev_rx_offload_name(uint64_t offload); 2104 2105 /** 2106 * Get DEV_TX_OFFLOAD_* flag name. 2107 * 2108 * @param offload 2109 * Offload flag. 2110 * @return 2111 * Offload name or 'UNKNOWN' if the flag cannot be recognised. 2112 */ 2113 const char *rte_eth_dev_tx_offload_name(uint64_t offload); 2114 2115 /** 2116 * @warning 2117 * @b EXPERIMENTAL: this API may change without prior notice. 2118 * 2119 * Get RTE_ETH_DEV_CAPA_* flag name. 2120 * 2121 * @param capability 2122 * Capability flag. 2123 * @return 2124 * Capability name or 'UNKNOWN' if the flag cannot be recognized. 2125 */ 2126 __rte_experimental 2127 const char *rte_eth_dev_capability_name(uint64_t capability); 2128 2129 /** 2130 * Configure an Ethernet device. 2131 * This function must be invoked first before any other function in the 2132 * Ethernet API. This function can also be re-invoked when a device is in the 2133 * stopped state. 2134 * 2135 * @param port_id 2136 * The port identifier of the Ethernet device to configure. 2137 * @param nb_rx_queue 2138 * The number of receive queues to set up for the Ethernet device. 2139 * @param nb_tx_queue 2140 * The number of transmit queues to set up for the Ethernet device. 2141 * @param eth_conf 2142 * The pointer to the configuration data to be used for the Ethernet device. 2143 * The *rte_eth_conf* structure includes: 2144 * - the hardware offload features to activate, with dedicated fields for 2145 * each statically configurable offload hardware feature provided by 2146 * Ethernet devices, such as IP checksum or VLAN tag stripping for 2147 * example. 2148 * The Rx offload bitfield API is obsolete and will be deprecated. 2149 * Applications should set the ignore_bitfield_offloads bit on *rxmode* 2150 * structure and use offloads field to set per-port offloads instead. 2151 * - Any offloading set in eth_conf->[rt]xmode.offloads must be within 2152 * the [rt]x_offload_capa returned from rte_eth_dev_info_get(). 2153 * Any type of device supported offloading set in the input argument 2154 * eth_conf->[rt]xmode.offloads to rte_eth_dev_configure() is enabled 2155 * on all queues and it can't be disabled in rte_eth_[rt]x_queue_setup() 2156 * - the Receive Side Scaling (RSS) configuration when using multiple Rx 2157 * queues per port. Any RSS hash function set in eth_conf->rss_conf.rss_hf 2158 * must be within the flow_type_rss_offloads provided by drivers via 2159 * rte_eth_dev_info_get() API. 2160 * 2161 * Embedding all configuration information in a single data structure 2162 * is the more flexible method that allows the addition of new features 2163 * without changing the syntax of the API. 2164 * @return 2165 * - 0: Success, device configured. 2166 * - <0: Error code returned by the driver configuration function. 2167 */ 2168 int rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_queue, 2169 uint16_t nb_tx_queue, const struct rte_eth_conf *eth_conf); 2170 2171 /** 2172 * @warning 2173 * @b EXPERIMENTAL: this API may change without prior notice. 2174 * 2175 * Check if an Ethernet device was physically removed. 2176 * 2177 * @param port_id 2178 * The port identifier of the Ethernet device. 2179 * @return 2180 * 1 when the Ethernet device is removed, otherwise 0. 2181 */ 2182 __rte_experimental 2183 int 2184 rte_eth_dev_is_removed(uint16_t port_id); 2185 2186 /** 2187 * Allocate and set up a receive queue for an Ethernet device. 2188 * 2189 * The function allocates a contiguous block of memory for *nb_rx_desc* 2190 * receive descriptors from a memory zone associated with *socket_id* 2191 * and initializes each receive descriptor with a network buffer allocated 2192 * from the memory pool *mb_pool*. 2193 * 2194 * @param port_id 2195 * The port identifier of the Ethernet device. 2196 * @param rx_queue_id 2197 * The index of the receive queue to set up. 2198 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2199 * to rte_eth_dev_configure(). 2200 * @param nb_rx_desc 2201 * The number of receive descriptors to allocate for the receive ring. 2202 * @param socket_id 2203 * The *socket_id* argument is the socket identifier in case of NUMA. 2204 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for 2205 * the DMA memory allocated for the receive descriptors of the ring. 2206 * @param rx_conf 2207 * The pointer to the configuration data to be used for the receive queue. 2208 * NULL value is allowed, in which case default Rx configuration 2209 * will be used. 2210 * The *rx_conf* structure contains an *rx_thresh* structure with the values 2211 * of the Prefetch, Host, and Write-Back threshold registers of the receive 2212 * ring. 2213 * In addition it contains the hardware offloads features to activate using 2214 * the DEV_RX_OFFLOAD_* flags. 2215 * If an offloading set in rx_conf->offloads 2216 * hasn't been set in the input argument eth_conf->rxmode.offloads 2217 * to rte_eth_dev_configure(), it is a new added offloading, it must be 2218 * per-queue type and it is enabled for the queue. 2219 * No need to repeat any bit in rx_conf->offloads which has already been 2220 * enabled in rte_eth_dev_configure() at port level. An offloading enabled 2221 * at port level can't be disabled at queue level. 2222 * The configuration structure also contains the pointer to the array 2223 * of the receiving buffer segment descriptions, see rx_seg and rx_nseg 2224 * fields, this extended configuration might be used by split offloads like 2225 * RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT. If mb_pool is not NULL, 2226 * the extended configuration fields must be set to NULL and zero. 2227 * @param mb_pool 2228 * The pointer to the memory pool from which to allocate *rte_mbuf* network 2229 * memory buffers to populate each descriptor of the receive ring. There are 2230 * two options to provide Rx buffer configuration: 2231 * - single pool: 2232 * mb_pool is not NULL, rx_conf.rx_nseg is 0. 2233 * - multiple segments description: 2234 * mb_pool is NULL, rx_conf.rx_seg is not NULL, rx_conf.rx_nseg is not 0. 2235 * Taken only if flag RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is set in offloads. 2236 * 2237 * @return 2238 * - 0: Success, receive queue correctly set up. 2239 * - -EIO: if device is removed. 2240 * - -ENODEV: if *port_id* is invalid. 2241 * - -EINVAL: The memory pool pointer is null or the size of network buffers 2242 * which can be allocated from this memory pool does not fit the various 2243 * buffer sizes allowed by the device controller. 2244 * - -ENOMEM: Unable to allocate the receive ring descriptors or to 2245 * allocate network memory buffers from the memory pool when 2246 * initializing receive descriptors. 2247 */ 2248 int rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id, 2249 uint16_t nb_rx_desc, unsigned int socket_id, 2250 const struct rte_eth_rxconf *rx_conf, 2251 struct rte_mempool *mb_pool); 2252 2253 /** 2254 * @warning 2255 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2256 * 2257 * Allocate and set up a hairpin receive queue for an Ethernet device. 2258 * 2259 * The function set up the selected queue to be used in hairpin. 2260 * 2261 * @param port_id 2262 * The port identifier of the Ethernet device. 2263 * @param rx_queue_id 2264 * The index of the receive queue to set up. 2265 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2266 * to rte_eth_dev_configure(). 2267 * @param nb_rx_desc 2268 * The number of receive descriptors to allocate for the receive ring. 2269 * 0 means the PMD will use default value. 2270 * @param conf 2271 * The pointer to the hairpin configuration. 2272 * 2273 * @return 2274 * - (0) if successful. 2275 * - (-ENODEV) if *port_id* is invalid. 2276 * - (-ENOTSUP) if hardware doesn't support. 2277 * - (-EINVAL) if bad parameter. 2278 * - (-ENOMEM) if unable to allocate the resources. 2279 */ 2280 __rte_experimental 2281 int rte_eth_rx_hairpin_queue_setup 2282 (uint16_t port_id, uint16_t rx_queue_id, uint16_t nb_rx_desc, 2283 const struct rte_eth_hairpin_conf *conf); 2284 2285 /** 2286 * Allocate and set up a transmit queue for an Ethernet device. 2287 * 2288 * @param port_id 2289 * The port identifier of the Ethernet device. 2290 * @param tx_queue_id 2291 * The index of the transmit queue to set up. 2292 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2293 * to rte_eth_dev_configure(). 2294 * @param nb_tx_desc 2295 * The number of transmit descriptors to allocate for the transmit ring. 2296 * @param socket_id 2297 * The *socket_id* argument is the socket identifier in case of NUMA. 2298 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for 2299 * the DMA memory allocated for the transmit descriptors of the ring. 2300 * @param tx_conf 2301 * The pointer to the configuration data to be used for the transmit queue. 2302 * NULL value is allowed, in which case default Tx configuration 2303 * will be used. 2304 * The *tx_conf* structure contains the following data: 2305 * - The *tx_thresh* structure with the values of the Prefetch, Host, and 2306 * Write-Back threshold registers of the transmit ring. 2307 * When setting Write-Back threshold to the value greater then zero, 2308 * *tx_rs_thresh* value should be explicitly set to one. 2309 * - The *tx_free_thresh* value indicates the [minimum] number of network 2310 * buffers that must be pending in the transmit ring to trigger their 2311 * [implicit] freeing by the driver transmit function. 2312 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit 2313 * descriptors that must be pending in the transmit ring before setting the 2314 * RS bit on a descriptor by the driver transmit function. 2315 * The *tx_rs_thresh* value should be less or equal then 2316 * *tx_free_thresh* value, and both of them should be less then 2317 * *nb_tx_desc* - 3. 2318 * - The *offloads* member contains Tx offloads to be enabled. 2319 * If an offloading set in tx_conf->offloads 2320 * hasn't been set in the input argument eth_conf->txmode.offloads 2321 * to rte_eth_dev_configure(), it is a new added offloading, it must be 2322 * per-queue type and it is enabled for the queue. 2323 * No need to repeat any bit in tx_conf->offloads which has already been 2324 * enabled in rte_eth_dev_configure() at port level. An offloading enabled 2325 * at port level can't be disabled at queue level. 2326 * 2327 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces 2328 * the transmit function to use default values. 2329 * @return 2330 * - 0: Success, the transmit queue is correctly set up. 2331 * - -ENOMEM: Unable to allocate the transmit ring descriptors. 2332 */ 2333 int rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id, 2334 uint16_t nb_tx_desc, unsigned int socket_id, 2335 const struct rte_eth_txconf *tx_conf); 2336 2337 /** 2338 * @warning 2339 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2340 * 2341 * Allocate and set up a transmit hairpin queue for an Ethernet device. 2342 * 2343 * @param port_id 2344 * The port identifier of the Ethernet device. 2345 * @param tx_queue_id 2346 * The index of the transmit queue to set up. 2347 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2348 * to rte_eth_dev_configure(). 2349 * @param nb_tx_desc 2350 * The number of transmit descriptors to allocate for the transmit ring. 2351 * 0 to set default PMD value. 2352 * @param conf 2353 * The hairpin configuration. 2354 * 2355 * @return 2356 * - (0) if successful. 2357 * - (-ENODEV) if *port_id* is invalid. 2358 * - (-ENOTSUP) if hardware doesn't support. 2359 * - (-EINVAL) if bad parameter. 2360 * - (-ENOMEM) if unable to allocate the resources. 2361 */ 2362 __rte_experimental 2363 int rte_eth_tx_hairpin_queue_setup 2364 (uint16_t port_id, uint16_t tx_queue_id, uint16_t nb_tx_desc, 2365 const struct rte_eth_hairpin_conf *conf); 2366 2367 /** 2368 * @warning 2369 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2370 * 2371 * Get all the hairpin peer Rx / Tx ports of the current port. 2372 * The caller should ensure that the array is large enough to save the ports 2373 * list. 2374 * 2375 * @param port_id 2376 * The port identifier of the Ethernet device. 2377 * @param peer_ports 2378 * Pointer to the array to store the peer ports list. 2379 * @param len 2380 * Length of the array to store the port identifiers. 2381 * @param direction 2382 * Current port to peer port direction 2383 * positive - current used as Tx to get all peer Rx ports. 2384 * zero - current used as Rx to get all peer Tx ports. 2385 * 2386 * @return 2387 * - (0 or positive) actual peer ports number. 2388 * - (-EINVAL) if bad parameter. 2389 * - (-ENODEV) if *port_id* invalid 2390 * - (-ENOTSUP) if hardware doesn't support. 2391 * - Others detailed errors from PMD drivers. 2392 */ 2393 __rte_experimental 2394 int rte_eth_hairpin_get_peer_ports(uint16_t port_id, uint16_t *peer_ports, 2395 size_t len, uint32_t direction); 2396 2397 /** 2398 * @warning 2399 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2400 * 2401 * Bind all hairpin Tx queues of one port to the Rx queues of the peer port. 2402 * It is only allowed to call this function after all hairpin queues are 2403 * configured properly and the devices are in started state. 2404 * 2405 * @param tx_port 2406 * The identifier of the Tx port. 2407 * @param rx_port 2408 * The identifier of peer Rx port. 2409 * RTE_MAX_ETHPORTS is allowed for the traversal of all devices. 2410 * Rx port ID could have the same value as Tx port ID. 2411 * 2412 * @return 2413 * - (0) if successful. 2414 * - (-ENODEV) if Tx port ID is invalid. 2415 * - (-EBUSY) if device is not in started state. 2416 * - (-ENOTSUP) if hardware doesn't support. 2417 * - Others detailed errors from PMD drivers. 2418 */ 2419 __rte_experimental 2420 int rte_eth_hairpin_bind(uint16_t tx_port, uint16_t rx_port); 2421 2422 /** 2423 * @warning 2424 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 2425 * 2426 * Unbind all hairpin Tx queues of one port from the Rx queues of the peer port. 2427 * This should be called before closing the Tx or Rx devices, if the bind 2428 * function is called before. 2429 * After unbinding the hairpin ports pair, it is allowed to bind them again. 2430 * Changing queues configuration should be after stopping the device(s). 2431 * 2432 * @param tx_port 2433 * The identifier of the Tx port. 2434 * @param rx_port 2435 * The identifier of peer Rx port. 2436 * RTE_MAX_ETHPORTS is allowed for traversal of all devices. 2437 * Rx port ID could have the same value as Tx port ID. 2438 * 2439 * @return 2440 * - (0) if successful. 2441 * - (-ENODEV) if Tx port ID is invalid. 2442 * - (-EBUSY) if device is in stopped state. 2443 * - (-ENOTSUP) if hardware doesn't support. 2444 * - Others detailed errors from PMD drivers. 2445 */ 2446 __rte_experimental 2447 int rte_eth_hairpin_unbind(uint16_t tx_port, uint16_t rx_port); 2448 2449 /** 2450 * Return the NUMA socket to which an Ethernet device is connected 2451 * 2452 * @param port_id 2453 * The port identifier of the Ethernet device 2454 * @return 2455 * The NUMA socket ID to which the Ethernet device is connected or 2456 * a default of zero if the socket could not be determined. 2457 * -1 is returned is the port_id value is out of range. 2458 */ 2459 int rte_eth_dev_socket_id(uint16_t port_id); 2460 2461 /** 2462 * Check if port_id of device is attached 2463 * 2464 * @param port_id 2465 * The port identifier of the Ethernet device 2466 * @return 2467 * - 0 if port is out of range or not attached 2468 * - 1 if device is attached 2469 */ 2470 int rte_eth_dev_is_valid_port(uint16_t port_id); 2471 2472 /** 2473 * Start specified Rx queue of a port. It is used when rx_deferred_start 2474 * flag of the specified queue is true. 2475 * 2476 * @param port_id 2477 * The port identifier of the Ethernet device 2478 * @param rx_queue_id 2479 * The index of the Rx queue to update the ring. 2480 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2481 * to rte_eth_dev_configure(). 2482 * @return 2483 * - 0: Success, the receive queue is started. 2484 * - -ENODEV: if *port_id* is invalid. 2485 * - -EINVAL: The queue_id out of range or belong to hairpin. 2486 * - -EIO: if device is removed. 2487 * - -ENOTSUP: The function not supported in PMD driver. 2488 */ 2489 int rte_eth_dev_rx_queue_start(uint16_t port_id, uint16_t rx_queue_id); 2490 2491 /** 2492 * Stop specified Rx queue of a port 2493 * 2494 * @param port_id 2495 * The port identifier of the Ethernet device 2496 * @param rx_queue_id 2497 * The index of the Rx queue to update the ring. 2498 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 2499 * to rte_eth_dev_configure(). 2500 * @return 2501 * - 0: Success, the receive queue is stopped. 2502 * - -ENODEV: if *port_id* is invalid. 2503 * - -EINVAL: The queue_id out of range or belong to hairpin. 2504 * - -EIO: if device is removed. 2505 * - -ENOTSUP: The function not supported in PMD driver. 2506 */ 2507 int rte_eth_dev_rx_queue_stop(uint16_t port_id, uint16_t rx_queue_id); 2508 2509 /** 2510 * Start Tx for specified queue of a port. It is used when tx_deferred_start 2511 * flag of the specified queue is true. 2512 * 2513 * @param port_id 2514 * The port identifier of the Ethernet device 2515 * @param tx_queue_id 2516 * The index of the Tx queue to update the ring. 2517 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2518 * to rte_eth_dev_configure(). 2519 * @return 2520 * - 0: Success, the transmit queue is started. 2521 * - -ENODEV: if *port_id* is invalid. 2522 * - -EINVAL: The queue_id out of range or belong to hairpin. 2523 * - -EIO: if device is removed. 2524 * - -ENOTSUP: The function not supported in PMD driver. 2525 */ 2526 int rte_eth_dev_tx_queue_start(uint16_t port_id, uint16_t tx_queue_id); 2527 2528 /** 2529 * Stop specified Tx queue of a port 2530 * 2531 * @param port_id 2532 * The port identifier of the Ethernet device 2533 * @param tx_queue_id 2534 * The index of the Tx queue to update the ring. 2535 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 2536 * to rte_eth_dev_configure(). 2537 * @return 2538 * - 0: Success, the transmit queue is stopped. 2539 * - -ENODEV: if *port_id* is invalid. 2540 * - -EINVAL: The queue_id out of range or belong to hairpin. 2541 * - -EIO: if device is removed. 2542 * - -ENOTSUP: The function not supported in PMD driver. 2543 */ 2544 int rte_eth_dev_tx_queue_stop(uint16_t port_id, uint16_t tx_queue_id); 2545 2546 /** 2547 * Start an Ethernet device. 2548 * 2549 * The device start step is the last one and consists of setting the configured 2550 * offload features and in starting the transmit and the receive units of the 2551 * device. 2552 * 2553 * Device RTE_ETH_DEV_NOLIVE_MAC_ADDR flag causes MAC address to be set before 2554 * PMD port start callback function is invoked. 2555 * 2556 * On success, all basic functions exported by the Ethernet API (link status, 2557 * receive/transmit, and so on) can be invoked. 2558 * 2559 * @param port_id 2560 * The port identifier of the Ethernet device. 2561 * @return 2562 * - 0: Success, Ethernet device started. 2563 * - <0: Error code of the driver device start function. 2564 */ 2565 int rte_eth_dev_start(uint16_t port_id); 2566 2567 /** 2568 * Stop an Ethernet device. The device can be restarted with a call to 2569 * rte_eth_dev_start() 2570 * 2571 * @param port_id 2572 * The port identifier of the Ethernet device. 2573 * @return 2574 * - 0: Success, Ethernet device stopped. 2575 * - <0: Error code of the driver device stop function. 2576 */ 2577 int rte_eth_dev_stop(uint16_t port_id); 2578 2579 /** 2580 * Link up an Ethernet device. 2581 * 2582 * Set device link up will re-enable the device Rx/Tx 2583 * functionality after it is previously set device linked down. 2584 * 2585 * @param port_id 2586 * The port identifier of the Ethernet device. 2587 * @return 2588 * - 0: Success, Ethernet device linked up. 2589 * - <0: Error code of the driver device link up function. 2590 */ 2591 int rte_eth_dev_set_link_up(uint16_t port_id); 2592 2593 /** 2594 * Link down an Ethernet device. 2595 * The device Rx/Tx functionality will be disabled if success, 2596 * and it can be re-enabled with a call to 2597 * rte_eth_dev_set_link_up() 2598 * 2599 * @param port_id 2600 * The port identifier of the Ethernet device. 2601 */ 2602 int rte_eth_dev_set_link_down(uint16_t port_id); 2603 2604 /** 2605 * Close a stopped Ethernet device. The device cannot be restarted! 2606 * The function frees all port resources. 2607 * 2608 * @param port_id 2609 * The port identifier of the Ethernet device. 2610 * @return 2611 * - Zero if the port is closed successfully. 2612 * - Negative if something went wrong. 2613 */ 2614 int rte_eth_dev_close(uint16_t port_id); 2615 2616 /** 2617 * Reset a Ethernet device and keep its port ID. 2618 * 2619 * When a port has to be reset passively, the DPDK application can invoke 2620 * this function. For example when a PF is reset, all its VFs should also 2621 * be reset. Normally a DPDK application can invoke this function when 2622 * RTE_ETH_EVENT_INTR_RESET event is detected, but can also use it to start 2623 * a port reset in other circumstances. 2624 * 2625 * When this function is called, it first stops the port and then calls the 2626 * PMD specific dev_uninit( ) and dev_init( ) to return the port to initial 2627 * state, in which no Tx and Rx queues are setup, as if the port has been 2628 * reset and not started. The port keeps the port ID it had before the 2629 * function call. 2630 * 2631 * After calling rte_eth_dev_reset( ), the application should use 2632 * rte_eth_dev_configure( ), rte_eth_rx_queue_setup( ), 2633 * rte_eth_tx_queue_setup( ), and rte_eth_dev_start( ) 2634 * to reconfigure the device as appropriate. 2635 * 2636 * Note: To avoid unexpected behavior, the application should stop calling 2637 * Tx and Rx functions before calling rte_eth_dev_reset( ). For thread 2638 * safety, all these controlling functions should be called from the same 2639 * thread. 2640 * 2641 * @param port_id 2642 * The port identifier of the Ethernet device. 2643 * 2644 * @return 2645 * - (0) if successful. 2646 * - (-ENODEV) if *port_id* is invalid. 2647 * - (-ENOTSUP) if hardware doesn't support this function. 2648 * - (-EPERM) if not ran from the primary process. 2649 * - (-EIO) if re-initialisation failed or device is removed. 2650 * - (-ENOMEM) if the reset failed due to OOM. 2651 * - (-EAGAIN) if the reset temporarily failed and should be retried later. 2652 */ 2653 int rte_eth_dev_reset(uint16_t port_id); 2654 2655 /** 2656 * Enable receipt in promiscuous mode for an Ethernet device. 2657 * 2658 * @param port_id 2659 * The port identifier of the Ethernet device. 2660 * @return 2661 * - (0) if successful. 2662 * - (-ENOTSUP) if support for promiscuous_enable() does not exist 2663 * for the device. 2664 * - (-ENODEV) if *port_id* invalid. 2665 */ 2666 int rte_eth_promiscuous_enable(uint16_t port_id); 2667 2668 /** 2669 * Disable receipt in promiscuous mode for an Ethernet device. 2670 * 2671 * @param port_id 2672 * The port identifier of the Ethernet device. 2673 * @return 2674 * - (0) if successful. 2675 * - (-ENOTSUP) if support for promiscuous_disable() does not exist 2676 * for the device. 2677 * - (-ENODEV) if *port_id* invalid. 2678 */ 2679 int rte_eth_promiscuous_disable(uint16_t port_id); 2680 2681 /** 2682 * Return the value of promiscuous mode for an Ethernet device. 2683 * 2684 * @param port_id 2685 * The port identifier of the Ethernet device. 2686 * @return 2687 * - (1) if promiscuous is enabled 2688 * - (0) if promiscuous is disabled. 2689 * - (-1) on error 2690 */ 2691 int rte_eth_promiscuous_get(uint16_t port_id); 2692 2693 /** 2694 * Enable the receipt of any multicast frame by an Ethernet device. 2695 * 2696 * @param port_id 2697 * The port identifier of the Ethernet device. 2698 * @return 2699 * - (0) if successful. 2700 * - (-ENOTSUP) if support for allmulticast_enable() does not exist 2701 * for the device. 2702 * - (-ENODEV) if *port_id* invalid. 2703 */ 2704 int rte_eth_allmulticast_enable(uint16_t port_id); 2705 2706 /** 2707 * Disable the receipt of all multicast frames by an Ethernet device. 2708 * 2709 * @param port_id 2710 * The port identifier of the Ethernet device. 2711 * @return 2712 * - (0) if successful. 2713 * - (-ENOTSUP) if support for allmulticast_disable() does not exist 2714 * for the device. 2715 * - (-ENODEV) if *port_id* invalid. 2716 */ 2717 int rte_eth_allmulticast_disable(uint16_t port_id); 2718 2719 /** 2720 * Return the value of allmulticast mode for an Ethernet device. 2721 * 2722 * @param port_id 2723 * The port identifier of the Ethernet device. 2724 * @return 2725 * - (1) if allmulticast is enabled 2726 * - (0) if allmulticast is disabled. 2727 * - (-1) on error 2728 */ 2729 int rte_eth_allmulticast_get(uint16_t port_id); 2730 2731 /** 2732 * Retrieve the link status (up/down), the duplex mode (half/full), 2733 * the negotiation (auto/fixed), and if available, the speed (Mbps). 2734 * 2735 * It might need to wait up to 9 seconds. 2736 * @see rte_eth_link_get_nowait. 2737 * 2738 * @param port_id 2739 * The port identifier of the Ethernet device. 2740 * @param link 2741 * Link information written back. 2742 * @return 2743 * - (0) if successful. 2744 * - (-ENOTSUP) if the function is not supported in PMD driver. 2745 * - (-ENODEV) if *port_id* invalid. 2746 * - (-EINVAL) if bad parameter. 2747 */ 2748 int rte_eth_link_get(uint16_t port_id, struct rte_eth_link *link); 2749 2750 /** 2751 * Retrieve the link status (up/down), the duplex mode (half/full), 2752 * the negotiation (auto/fixed), and if available, the speed (Mbps). 2753 * 2754 * @param port_id 2755 * The port identifier of the Ethernet device. 2756 * @param link 2757 * Link information written back. 2758 * @return 2759 * - (0) if successful. 2760 * - (-ENOTSUP) if the function is not supported in PMD driver. 2761 * - (-ENODEV) if *port_id* invalid. 2762 * - (-EINVAL) if bad parameter. 2763 */ 2764 int rte_eth_link_get_nowait(uint16_t port_id, struct rte_eth_link *link); 2765 2766 /** 2767 * @warning 2768 * @b EXPERIMENTAL: this API may change without prior notice. 2769 * 2770 * The function converts a link_speed to a string. It handles all special 2771 * values like unknown or none speed. 2772 * 2773 * @param link_speed 2774 * link_speed of rte_eth_link struct 2775 * @return 2776 * Link speed in textual format. It's pointer to immutable memory. 2777 * No free is required. 2778 */ 2779 __rte_experimental 2780 const char *rte_eth_link_speed_to_str(uint32_t link_speed); 2781 2782 /** 2783 * @warning 2784 * @b EXPERIMENTAL: this API may change without prior notice. 2785 * 2786 * The function converts a rte_eth_link struct representing a link status to 2787 * a string. 2788 * 2789 * @param str 2790 * A pointer to a string to be filled with textual representation of 2791 * device status. At least ETH_LINK_MAX_STR_LEN bytes should be allocated to 2792 * store default link status text. 2793 * @param len 2794 * Length of available memory at 'str' string. 2795 * @param eth_link 2796 * Link status returned by rte_eth_link_get function 2797 * @return 2798 * Number of bytes written to str array or -EINVAL if bad parameter. 2799 */ 2800 __rte_experimental 2801 int rte_eth_link_to_str(char *str, size_t len, 2802 const struct rte_eth_link *eth_link); 2803 2804 /** 2805 * Retrieve the general I/O statistics of an Ethernet device. 2806 * 2807 * @param port_id 2808 * The port identifier of the Ethernet device. 2809 * @param stats 2810 * A pointer to a structure of type *rte_eth_stats* to be filled with 2811 * the values of device counters for the following set of statistics: 2812 * - *ipackets* with the total of successfully received packets. 2813 * - *opackets* with the total of successfully transmitted packets. 2814 * - *ibytes* with the total of successfully received bytes. 2815 * - *obytes* with the total of successfully transmitted bytes. 2816 * - *ierrors* with the total of erroneous received packets. 2817 * - *oerrors* with the total of failed transmitted packets. 2818 * @return 2819 * Zero if successful. Non-zero otherwise. 2820 */ 2821 int rte_eth_stats_get(uint16_t port_id, struct rte_eth_stats *stats); 2822 2823 /** 2824 * Reset the general I/O statistics of an Ethernet device. 2825 * 2826 * @param port_id 2827 * The port identifier of the Ethernet device. 2828 * @return 2829 * - (0) if device notified to reset stats. 2830 * - (-ENOTSUP) if hardware doesn't support. 2831 * - (-ENODEV) if *port_id* invalid. 2832 * - (<0): Error code of the driver stats reset function. 2833 */ 2834 int rte_eth_stats_reset(uint16_t port_id); 2835 2836 /** 2837 * Retrieve names of extended statistics of an Ethernet device. 2838 * 2839 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched 2840 * by array index: 2841 * xstats_names[i].name => xstats[i].value 2842 * 2843 * And the array index is same with id field of 'struct rte_eth_xstat': 2844 * xstats[i].id == i 2845 * 2846 * This assumption makes key-value pair matching less flexible but simpler. 2847 * 2848 * @param port_id 2849 * The port identifier of the Ethernet device. 2850 * @param xstats_names 2851 * An rte_eth_xstat_name array of at least *size* elements to 2852 * be filled. If set to NULL, the function returns the required number 2853 * of elements. 2854 * @param size 2855 * The size of the xstats_names array (number of elements). 2856 * @return 2857 * - A positive value lower or equal to size: success. The return value 2858 * is the number of entries filled in the stats table. 2859 * - A positive value higher than size: error, the given statistics table 2860 * is too small. The return value corresponds to the size that should 2861 * be given to succeed. The entries in the table are not valid and 2862 * shall not be used by the caller. 2863 * - A negative value on error (invalid port ID). 2864 */ 2865 int rte_eth_xstats_get_names(uint16_t port_id, 2866 struct rte_eth_xstat_name *xstats_names, 2867 unsigned int size); 2868 2869 /** 2870 * Retrieve extended statistics of an Ethernet device. 2871 * 2872 * There is an assumption that 'xstat_names' and 'xstats' arrays are matched 2873 * by array index: 2874 * xstats_names[i].name => xstats[i].value 2875 * 2876 * And the array index is same with id field of 'struct rte_eth_xstat': 2877 * xstats[i].id == i 2878 * 2879 * This assumption makes key-value pair matching less flexible but simpler. 2880 * 2881 * @param port_id 2882 * The port identifier of the Ethernet device. 2883 * @param xstats 2884 * A pointer to a table of structure of type *rte_eth_xstat* 2885 * to be filled with device statistics ids and values. 2886 * This parameter can be set to NULL if n is 0. 2887 * @param n 2888 * The size of the xstats array (number of elements). 2889 * @return 2890 * - A positive value lower or equal to n: success. The return value 2891 * is the number of entries filled in the stats table. 2892 * - A positive value higher than n: error, the given statistics table 2893 * is too small. The return value corresponds to the size that should 2894 * be given to succeed. The entries in the table are not valid and 2895 * shall not be used by the caller. 2896 * - A negative value on error (invalid port ID). 2897 */ 2898 int rte_eth_xstats_get(uint16_t port_id, struct rte_eth_xstat *xstats, 2899 unsigned int n); 2900 2901 /** 2902 * Retrieve names of extended statistics of an Ethernet device. 2903 * 2904 * @param port_id 2905 * The port identifier of the Ethernet device. 2906 * @param xstats_names 2907 * Array to be filled in with names of requested device statistics. 2908 * Must not be NULL if @p ids are specified (not NULL). 2909 * @param size 2910 * Number of elements in @p xstats_names array (if not NULL) and in 2911 * @p ids array (if not NULL). Must be 0 if both array pointers are NULL. 2912 * @param ids 2913 * IDs array given by app to retrieve specific statistics. May be NULL to 2914 * retrieve names of all available statistics or, if @p xstats_names is 2915 * NULL as well, just the number of available statistics. 2916 * @return 2917 * - A positive value lower or equal to size: success. The return value 2918 * is the number of entries filled in the stats table. 2919 * - A positive value higher than size: success. The given statistics table 2920 * is too small. The return value corresponds to the size that should 2921 * be given to succeed. The entries in the table are not valid and 2922 * shall not be used by the caller. 2923 * - A negative value on error. 2924 */ 2925 int 2926 rte_eth_xstats_get_names_by_id(uint16_t port_id, 2927 struct rte_eth_xstat_name *xstats_names, unsigned int size, 2928 uint64_t *ids); 2929 2930 /** 2931 * Retrieve extended statistics of an Ethernet device. 2932 * 2933 * @param port_id 2934 * The port identifier of the Ethernet device. 2935 * @param ids 2936 * IDs array given by app to retrieve specific statistics. May be NULL to 2937 * retrieve all available statistics or, if @p values is NULL as well, 2938 * just the number of available statistics. 2939 * @param values 2940 * Array to be filled in with requested device statistics. 2941 * Must not be NULL if ids are specified (not NULL). 2942 * @param size 2943 * Number of elements in @p values array (if not NULL) and in @p ids 2944 * array (if not NULL). Must be 0 if both array pointers are NULL. 2945 * @return 2946 * - A positive value lower or equal to size: success. The return value 2947 * is the number of entries filled in the stats table. 2948 * - A positive value higher than size: success: The given statistics table 2949 * is too small. The return value corresponds to the size that should 2950 * be given to succeed. The entries in the table are not valid and 2951 * shall not be used by the caller. 2952 * - A negative value on error. 2953 */ 2954 int rte_eth_xstats_get_by_id(uint16_t port_id, const uint64_t *ids, 2955 uint64_t *values, unsigned int size); 2956 2957 /** 2958 * Gets the ID of a statistic from its name. 2959 * 2960 * This function searches for the statistics using string compares, and 2961 * as such should not be used on the fast-path. For fast-path retrieval of 2962 * specific statistics, store the ID as provided in *id* from this function, 2963 * and pass the ID to rte_eth_xstats_get() 2964 * 2965 * @param port_id The port to look up statistics from 2966 * @param xstat_name The name of the statistic to return 2967 * @param[out] id A pointer to an app-supplied uint64_t which should be 2968 * set to the ID of the stat if the stat exists. 2969 * @return 2970 * 0 on success 2971 * -ENODEV for invalid port_id, 2972 * -EIO if device is removed, 2973 * -EINVAL if the xstat_name doesn't exist in port_id 2974 * -ENOMEM if bad parameter. 2975 */ 2976 int rte_eth_xstats_get_id_by_name(uint16_t port_id, const char *xstat_name, 2977 uint64_t *id); 2978 2979 /** 2980 * Reset extended statistics of an Ethernet device. 2981 * 2982 * @param port_id 2983 * The port identifier of the Ethernet device. 2984 * @return 2985 * - (0) if device notified to reset extended stats. 2986 * - (-ENOTSUP) if pmd doesn't support both 2987 * extended stats and basic stats reset. 2988 * - (-ENODEV) if *port_id* invalid. 2989 * - (<0): Error code of the driver xstats reset function. 2990 */ 2991 int rte_eth_xstats_reset(uint16_t port_id); 2992 2993 /** 2994 * Set a mapping for the specified transmit queue to the specified per-queue 2995 * statistics counter. 2996 * 2997 * @param port_id 2998 * The port identifier of the Ethernet device. 2999 * @param tx_queue_id 3000 * The index of the transmit queue for which a queue stats mapping is required. 3001 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 3002 * to rte_eth_dev_configure(). 3003 * @param stat_idx 3004 * The per-queue packet statistics functionality number that the transmit 3005 * queue is to be assigned. 3006 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1]. 3007 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256. 3008 * @return 3009 * Zero if successful. Non-zero otherwise. 3010 */ 3011 int rte_eth_dev_set_tx_queue_stats_mapping(uint16_t port_id, 3012 uint16_t tx_queue_id, uint8_t stat_idx); 3013 3014 /** 3015 * Set a mapping for the specified receive queue to the specified per-queue 3016 * statistics counter. 3017 * 3018 * @param port_id 3019 * The port identifier of the Ethernet device. 3020 * @param rx_queue_id 3021 * The index of the receive queue for which a queue stats mapping is required. 3022 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3023 * to rte_eth_dev_configure(). 3024 * @param stat_idx 3025 * The per-queue packet statistics functionality number that the receive 3026 * queue is to be assigned. 3027 * The value must be in the range [0, RTE_ETHDEV_QUEUE_STAT_CNTRS - 1]. 3028 * Max RTE_ETHDEV_QUEUE_STAT_CNTRS being 256. 3029 * @return 3030 * Zero if successful. Non-zero otherwise. 3031 */ 3032 int rte_eth_dev_set_rx_queue_stats_mapping(uint16_t port_id, 3033 uint16_t rx_queue_id, 3034 uint8_t stat_idx); 3035 3036 /** 3037 * Retrieve the Ethernet address of an Ethernet device. 3038 * 3039 * @param port_id 3040 * The port identifier of the Ethernet device. 3041 * @param mac_addr 3042 * A pointer to a structure of type *ether_addr* to be filled with 3043 * the Ethernet address of the Ethernet device. 3044 * @return 3045 * - (0) if successful 3046 * - (-ENODEV) if *port_id* invalid. 3047 * - (-EINVAL) if bad parameter. 3048 */ 3049 int rte_eth_macaddr_get(uint16_t port_id, struct rte_ether_addr *mac_addr); 3050 3051 /** 3052 * @warning 3053 * @b EXPERIMENTAL: this API may change without prior notice 3054 * 3055 * Retrieve the Ethernet addresses of an Ethernet device. 3056 * 3057 * @param port_id 3058 * The port identifier of the Ethernet device. 3059 * @param ma 3060 * A pointer to an array of structures of type *ether_addr* to be filled with 3061 * the Ethernet addresses of the Ethernet device. 3062 * @param num 3063 * Number of elements in the @p ma array. 3064 * Note that rte_eth_dev_info::max_mac_addrs can be used to retrieve 3065 * max number of Ethernet addresses for given port. 3066 * @return 3067 * - number of retrieved addresses if successful 3068 * - (-ENODEV) if *port_id* invalid. 3069 * - (-EINVAL) if bad parameter. 3070 */ 3071 __rte_experimental 3072 int rte_eth_macaddrs_get(uint16_t port_id, struct rte_ether_addr *ma, 3073 unsigned int num); 3074 3075 /** 3076 * Retrieve the contextual information of an Ethernet device. 3077 * 3078 * As part of this function, a number of of fields in dev_info will be 3079 * initialized as follows: 3080 * 3081 * rx_desc_lim = lim 3082 * tx_desc_lim = lim 3083 * 3084 * Where lim is defined within the rte_eth_dev_info_get as 3085 * 3086 * const struct rte_eth_desc_lim lim = { 3087 * .nb_max = UINT16_MAX, 3088 * .nb_min = 0, 3089 * .nb_align = 1, 3090 * .nb_seg_max = UINT16_MAX, 3091 * .nb_mtu_seg_max = UINT16_MAX, 3092 * }; 3093 * 3094 * device = dev->device 3095 * min_mtu = RTE_ETHER_MIN_LEN - RTE_ETHER_HDR_LEN - RTE_ETHER_CRC_LEN 3096 * max_mtu = UINT16_MAX 3097 * 3098 * The following fields will be populated if support for dev_infos_get() 3099 * exists for the device and the rte_eth_dev 'dev' has been populated 3100 * successfully with a call to it: 3101 * 3102 * driver_name = dev->device->driver->name 3103 * nb_rx_queues = dev->data->nb_rx_queues 3104 * nb_tx_queues = dev->data->nb_tx_queues 3105 * dev_flags = &dev->data->dev_flags 3106 * 3107 * @param port_id 3108 * The port identifier of the Ethernet device. 3109 * @param dev_info 3110 * A pointer to a structure of type *rte_eth_dev_info* to be filled with 3111 * the contextual information of the Ethernet device. 3112 * @return 3113 * - (0) if successful. 3114 * - (-ENOTSUP) if support for dev_infos_get() does not exist for the device. 3115 * - (-ENODEV) if *port_id* invalid. 3116 * - (-EINVAL) if bad parameter. 3117 */ 3118 int rte_eth_dev_info_get(uint16_t port_id, struct rte_eth_dev_info *dev_info); 3119 3120 /** 3121 * @warning 3122 * @b EXPERIMENTAL: this API may change without prior notice. 3123 * 3124 * Retrieve the configuration of an Ethernet device. 3125 * 3126 * @param port_id 3127 * The port identifier of the Ethernet device. 3128 * @param dev_conf 3129 * Location for Ethernet device configuration to be filled in. 3130 * @return 3131 * - (0) if successful. 3132 * - (-ENODEV) if *port_id* invalid. 3133 * - (-EINVAL) if bad parameter. 3134 */ 3135 __rte_experimental 3136 int rte_eth_dev_conf_get(uint16_t port_id, struct rte_eth_conf *dev_conf); 3137 3138 /** 3139 * Retrieve the firmware version of a device. 3140 * 3141 * @param port_id 3142 * The port identifier of the device. 3143 * @param fw_version 3144 * A pointer to a string array storing the firmware version of a device, 3145 * the string includes terminating null. This pointer is allocated by caller. 3146 * @param fw_size 3147 * The size of the string array pointed by fw_version, which should be 3148 * large enough to store firmware version of the device. 3149 * @return 3150 * - (0) if successful. 3151 * - (-ENOTSUP) if operation is not supported. 3152 * - (-ENODEV) if *port_id* invalid. 3153 * - (-EIO) if device is removed. 3154 * - (-EINVAL) if bad parameter. 3155 * - (>0) if *fw_size* is not enough to store firmware version, return 3156 * the size of the non truncated string. 3157 */ 3158 int rte_eth_dev_fw_version_get(uint16_t port_id, 3159 char *fw_version, size_t fw_size); 3160 3161 /** 3162 * Retrieve the supported packet types of an Ethernet device. 3163 * 3164 * When a packet type is announced as supported, it *must* be recognized by 3165 * the PMD. For instance, if RTE_PTYPE_L2_ETHER, RTE_PTYPE_L2_ETHER_VLAN 3166 * and RTE_PTYPE_L3_IPV4 are announced, the PMD must return the following 3167 * packet types for these packets: 3168 * - Ether/IPv4 -> RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4 3169 * - Ether/VLAN/IPv4 -> RTE_PTYPE_L2_ETHER_VLAN | RTE_PTYPE_L3_IPV4 3170 * - Ether/[anything else] -> RTE_PTYPE_L2_ETHER 3171 * - Ether/VLAN/[anything else] -> RTE_PTYPE_L2_ETHER_VLAN 3172 * 3173 * When a packet is received by a PMD, the most precise type must be 3174 * returned among the ones supported. However a PMD is allowed to set 3175 * packet type that is not in the supported list, at the condition that it 3176 * is more precise. Therefore, a PMD announcing no supported packet types 3177 * can still set a matching packet type in a received packet. 3178 * 3179 * @note 3180 * Better to invoke this API after the device is already started or Rx burst 3181 * function is decided, to obtain correct supported ptypes. 3182 * @note 3183 * if a given PMD does not report what ptypes it supports, then the supported 3184 * ptype count is reported as 0. 3185 * @param port_id 3186 * The port identifier of the Ethernet device. 3187 * @param ptype_mask 3188 * A hint of what kind of packet type which the caller is interested in. 3189 * @param ptypes 3190 * An array pointer to store adequate packet types, allocated by caller. 3191 * @param num 3192 * Size of the array pointed by param ptypes. 3193 * @return 3194 * - (>=0) Number of supported ptypes. If the number of types exceeds num, 3195 * only num entries will be filled into the ptypes array, but the full 3196 * count of supported ptypes will be returned. 3197 * - (-ENODEV) if *port_id* invalid. 3198 * - (-EINVAL) if bad parameter. 3199 */ 3200 int rte_eth_dev_get_supported_ptypes(uint16_t port_id, uint32_t ptype_mask, 3201 uint32_t *ptypes, int num); 3202 /** 3203 * Inform Ethernet device about reduced range of packet types to handle. 3204 * 3205 * Application can use this function to set only specific ptypes that it's 3206 * interested. This information can be used by the PMD to optimize Rx path. 3207 * 3208 * The function accepts an array `set_ptypes` allocated by the caller to 3209 * store the packet types set by the driver, the last element of the array 3210 * is set to RTE_PTYPE_UNKNOWN. The size of the `set_ptype` array should be 3211 * `rte_eth_dev_get_supported_ptypes() + 1` else it might only be filled 3212 * partially. 3213 * 3214 * @param port_id 3215 * The port identifier of the Ethernet device. 3216 * @param ptype_mask 3217 * The ptype family that application is interested in should be bitwise OR of 3218 * RTE_PTYPE_*_MASK or 0. 3219 * @param set_ptypes 3220 * An array pointer to store set packet types, allocated by caller. The 3221 * function marks the end of array with RTE_PTYPE_UNKNOWN. 3222 * @param num 3223 * Size of the array pointed by param ptypes. 3224 * Should be rte_eth_dev_get_supported_ptypes() + 1 to accommodate the 3225 * set ptypes. 3226 * @return 3227 * - (0) if Success. 3228 * - (-ENODEV) if *port_id* invalid. 3229 * - (-EINVAL) if *ptype_mask* is invalid (or) set_ptypes is NULL and 3230 * num > 0. 3231 */ 3232 int rte_eth_dev_set_ptypes(uint16_t port_id, uint32_t ptype_mask, 3233 uint32_t *set_ptypes, unsigned int num); 3234 3235 /** 3236 * Retrieve the MTU of an Ethernet device. 3237 * 3238 * @param port_id 3239 * The port identifier of the Ethernet device. 3240 * @param mtu 3241 * A pointer to a uint16_t where the retrieved MTU is to be stored. 3242 * @return 3243 * - (0) if successful. 3244 * - (-ENODEV) if *port_id* invalid. 3245 * - (-EINVAL) if bad parameter. 3246 */ 3247 int rte_eth_dev_get_mtu(uint16_t port_id, uint16_t *mtu); 3248 3249 /** 3250 * Change the MTU of an Ethernet device. 3251 * 3252 * @param port_id 3253 * The port identifier of the Ethernet device. 3254 * @param mtu 3255 * A uint16_t for the MTU to be applied. 3256 * @return 3257 * - (0) if successful. 3258 * - (-ENOTSUP) if operation is not supported. 3259 * - (-ENODEV) if *port_id* invalid. 3260 * - (-EIO) if device is removed. 3261 * - (-EINVAL) if *mtu* invalid, validation of mtu can occur within 3262 * rte_eth_dev_set_mtu if dev_infos_get is supported by the device or 3263 * when the mtu is set using dev->dev_ops->mtu_set. 3264 * - (-EBUSY) if operation is not allowed when the port is running 3265 */ 3266 int rte_eth_dev_set_mtu(uint16_t port_id, uint16_t mtu); 3267 3268 /** 3269 * Enable/Disable hardware filtering by an Ethernet device of received 3270 * VLAN packets tagged with a given VLAN Tag Identifier. 3271 * 3272 * @param port_id 3273 * The port identifier of the Ethernet device. 3274 * @param vlan_id 3275 * The VLAN Tag Identifier whose filtering must be enabled or disabled. 3276 * @param on 3277 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*. 3278 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*. 3279 * @return 3280 * - (0) if successful. 3281 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured. 3282 * - (-ENODEV) if *port_id* invalid. 3283 * - (-EIO) if device is removed. 3284 * - (-ENOSYS) if VLAN filtering on *port_id* disabled. 3285 * - (-EINVAL) if *vlan_id* > 4095. 3286 */ 3287 int rte_eth_dev_vlan_filter(uint16_t port_id, uint16_t vlan_id, int on); 3288 3289 /** 3290 * Enable/Disable hardware VLAN Strip by a Rx queue of an Ethernet device. 3291 * 3292 * @param port_id 3293 * The port identifier of the Ethernet device. 3294 * @param rx_queue_id 3295 * The index of the receive queue for which a queue stats mapping is required. 3296 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3297 * to rte_eth_dev_configure(). 3298 * @param on 3299 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port. 3300 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port. 3301 * @return 3302 * - (0) if successful. 3303 * - (-ENOTSUP) if hardware-assisted VLAN stripping not configured. 3304 * - (-ENODEV) if *port_id* invalid. 3305 * - (-EINVAL) if *rx_queue_id* invalid. 3306 */ 3307 int rte_eth_dev_set_vlan_strip_on_queue(uint16_t port_id, uint16_t rx_queue_id, 3308 int on); 3309 3310 /** 3311 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to 3312 * the VLAN header. 3313 * 3314 * @param port_id 3315 * The port identifier of the Ethernet device. 3316 * @param vlan_type 3317 * The VLAN type. 3318 * @param tag_type 3319 * The Tag Protocol ID 3320 * @return 3321 * - (0) if successful. 3322 * - (-ENOTSUP) if hardware-assisted VLAN TPID setup is not supported. 3323 * - (-ENODEV) if *port_id* invalid. 3324 * - (-EIO) if device is removed. 3325 */ 3326 int rte_eth_dev_set_vlan_ether_type(uint16_t port_id, 3327 enum rte_vlan_type vlan_type, 3328 uint16_t tag_type); 3329 3330 /** 3331 * Set VLAN offload configuration on an Ethernet device. 3332 * 3333 * @param port_id 3334 * The port identifier of the Ethernet device. 3335 * @param offload_mask 3336 * The VLAN Offload bit mask can be mixed use with "OR" 3337 * ETH_VLAN_STRIP_OFFLOAD 3338 * ETH_VLAN_FILTER_OFFLOAD 3339 * ETH_VLAN_EXTEND_OFFLOAD 3340 * ETH_QINQ_STRIP_OFFLOAD 3341 * @return 3342 * - (0) if successful. 3343 * - (-ENOTSUP) if hardware-assisted VLAN filtering not configured. 3344 * - (-ENODEV) if *port_id* invalid. 3345 * - (-EIO) if device is removed. 3346 */ 3347 int rte_eth_dev_set_vlan_offload(uint16_t port_id, int offload_mask); 3348 3349 /** 3350 * Read VLAN Offload configuration from an Ethernet device 3351 * 3352 * @param port_id 3353 * The port identifier of the Ethernet device. 3354 * @return 3355 * - (>0) if successful. Bit mask to indicate 3356 * ETH_VLAN_STRIP_OFFLOAD 3357 * ETH_VLAN_FILTER_OFFLOAD 3358 * ETH_VLAN_EXTEND_OFFLOAD 3359 * ETH_QINQ_STRIP_OFFLOAD 3360 * - (-ENODEV) if *port_id* invalid. 3361 */ 3362 int rte_eth_dev_get_vlan_offload(uint16_t port_id); 3363 3364 /** 3365 * Set port based Tx VLAN insertion on or off. 3366 * 3367 * @param port_id 3368 * The port identifier of the Ethernet device. 3369 * @param pvid 3370 * Port based Tx VLAN identifier together with user priority. 3371 * @param on 3372 * Turn on or off the port based Tx VLAN insertion. 3373 * 3374 * @return 3375 * - (0) if successful. 3376 * - negative if failed. 3377 */ 3378 int rte_eth_dev_set_vlan_pvid(uint16_t port_id, uint16_t pvid, int on); 3379 3380 typedef void (*buffer_tx_error_fn)(struct rte_mbuf **unsent, uint16_t count, 3381 void *userdata); 3382 3383 /** 3384 * Structure used to buffer packets for future Tx 3385 * Used by APIs rte_eth_tx_buffer and rte_eth_tx_buffer_flush 3386 */ 3387 struct rte_eth_dev_tx_buffer { 3388 buffer_tx_error_fn error_callback; 3389 void *error_userdata; 3390 uint16_t size; /**< Size of buffer for buffered Tx */ 3391 uint16_t length; /**< Number of packets in the array */ 3392 /** Pending packets to be sent on explicit flush or when full */ 3393 struct rte_mbuf *pkts[]; 3394 }; 3395 3396 /** 3397 * Calculate the size of the Tx buffer. 3398 * 3399 * @param sz 3400 * Number of stored packets. 3401 */ 3402 #define RTE_ETH_TX_BUFFER_SIZE(sz) \ 3403 (sizeof(struct rte_eth_dev_tx_buffer) + (sz) * sizeof(struct rte_mbuf *)) 3404 3405 /** 3406 * Initialize default values for buffered transmitting 3407 * 3408 * @param buffer 3409 * Tx buffer to be initialized. 3410 * @param size 3411 * Buffer size 3412 * @return 3413 * 0 if no error 3414 */ 3415 int 3416 rte_eth_tx_buffer_init(struct rte_eth_dev_tx_buffer *buffer, uint16_t size); 3417 3418 /** 3419 * Configure a callback for buffered packets which cannot be sent 3420 * 3421 * Register a specific callback to be called when an attempt is made to send 3422 * all packets buffered on an Ethernet port, but not all packets can 3423 * successfully be sent. The callback registered here will be called only 3424 * from calls to rte_eth_tx_buffer() and rte_eth_tx_buffer_flush() APIs. 3425 * The default callback configured for each queue by default just frees the 3426 * packets back to the calling mempool. If additional behaviour is required, 3427 * for example, to count dropped packets, or to retry transmission of packets 3428 * which cannot be sent, this function should be used to register a suitable 3429 * callback function to implement the desired behaviour. 3430 * The example callback "rte_eth_count_unsent_packet_callback()" is also 3431 * provided as reference. 3432 * 3433 * @param buffer 3434 * The port identifier of the Ethernet device. 3435 * @param callback 3436 * The function to be used as the callback. 3437 * @param userdata 3438 * Arbitrary parameter to be passed to the callback function 3439 * @return 3440 * 0 on success, or -EINVAL if bad parameter 3441 */ 3442 int 3443 rte_eth_tx_buffer_set_err_callback(struct rte_eth_dev_tx_buffer *buffer, 3444 buffer_tx_error_fn callback, void *userdata); 3445 3446 /** 3447 * Callback function for silently dropping unsent buffered packets. 3448 * 3449 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to 3450 * adjust the default behavior when buffered packets cannot be sent. This 3451 * function drops any unsent packets silently and is used by Tx buffered 3452 * operations as default behavior. 3453 * 3454 * NOTE: this function should not be called directly, instead it should be used 3455 * as a callback for packet buffering. 3456 * 3457 * NOTE: when configuring this function as a callback with 3458 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter 3459 * should point to an uint64_t value. 3460 * 3461 * @param pkts 3462 * The previously buffered packets which could not be sent 3463 * @param unsent 3464 * The number of unsent packets in the pkts array 3465 * @param userdata 3466 * Not used 3467 */ 3468 void 3469 rte_eth_tx_buffer_drop_callback(struct rte_mbuf **pkts, uint16_t unsent, 3470 void *userdata); 3471 3472 /** 3473 * Callback function for tracking unsent buffered packets. 3474 * 3475 * This function can be passed to rte_eth_tx_buffer_set_err_callback() to 3476 * adjust the default behavior when buffered packets cannot be sent. This 3477 * function drops any unsent packets, but also updates a user-supplied counter 3478 * to track the overall number of packets dropped. The counter should be an 3479 * uint64_t variable. 3480 * 3481 * NOTE: this function should not be called directly, instead it should be used 3482 * as a callback for packet buffering. 3483 * 3484 * NOTE: when configuring this function as a callback with 3485 * rte_eth_tx_buffer_set_err_callback(), the final, userdata parameter 3486 * should point to an uint64_t value. 3487 * 3488 * @param pkts 3489 * The previously buffered packets which could not be sent 3490 * @param unsent 3491 * The number of unsent packets in the pkts array 3492 * @param userdata 3493 * Pointer to an uint64_t value, which will be incremented by unsent 3494 */ 3495 void 3496 rte_eth_tx_buffer_count_callback(struct rte_mbuf **pkts, uint16_t unsent, 3497 void *userdata); 3498 3499 /** 3500 * Request the driver to free mbufs currently cached by the driver. The 3501 * driver will only free the mbuf if it is no longer in use. It is the 3502 * application's responsibility to ensure rte_eth_tx_buffer_flush(..) is 3503 * called if needed. 3504 * 3505 * @param port_id 3506 * The port identifier of the Ethernet device. 3507 * @param queue_id 3508 * The index of the transmit queue through which output packets must be 3509 * sent. 3510 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 3511 * to rte_eth_dev_configure(). 3512 * @param free_cnt 3513 * Maximum number of packets to free. Use 0 to indicate all possible packets 3514 * should be freed. Note that a packet may be using multiple mbufs. 3515 * @return 3516 * Failure: < 0 3517 * -ENODEV: Invalid interface 3518 * -EIO: device is removed 3519 * -ENOTSUP: Driver does not support function 3520 * Success: >= 0 3521 * 0-n: Number of packets freed. More packets may still remain in ring that 3522 * are in use. 3523 */ 3524 int 3525 rte_eth_tx_done_cleanup(uint16_t port_id, uint16_t queue_id, uint32_t free_cnt); 3526 3527 /** 3528 * Subtypes for IPsec offload event(@ref RTE_ETH_EVENT_IPSEC) raised by 3529 * eth device. 3530 */ 3531 enum rte_eth_event_ipsec_subtype { 3532 /** Unknown event type */ 3533 RTE_ETH_EVENT_IPSEC_UNKNOWN = 0, 3534 /** Sequence number overflow */ 3535 RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW, 3536 /** Soft time expiry of SA */ 3537 RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY, 3538 /** Soft byte expiry of SA */ 3539 RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY, 3540 /** Max value of this enum */ 3541 RTE_ETH_EVENT_IPSEC_MAX 3542 }; 3543 3544 /** 3545 * Descriptor for @ref RTE_ETH_EVENT_IPSEC event. Used by eth dev to send extra 3546 * information of the IPsec offload event. 3547 */ 3548 struct rte_eth_event_ipsec_desc { 3549 /** Type of RTE_ETH_EVENT_IPSEC_* event */ 3550 enum rte_eth_event_ipsec_subtype subtype; 3551 /** 3552 * Event specific metadata. 3553 * 3554 * For the following events, *userdata* registered 3555 * with the *rte_security_session* would be returned 3556 * as metadata, 3557 * 3558 * - @ref RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW 3559 * - @ref RTE_ETH_EVENT_IPSEC_SA_TIME_EXPIRY 3560 * - @ref RTE_ETH_EVENT_IPSEC_SA_BYTE_EXPIRY 3561 * 3562 * @see struct rte_security_session_conf 3563 * 3564 */ 3565 uint64_t metadata; 3566 }; 3567 3568 /** 3569 * The eth device event type for interrupt, and maybe others in the future. 3570 */ 3571 enum rte_eth_event_type { 3572 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */ 3573 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */ 3574 /** queue state event (enabled/disabled) */ 3575 RTE_ETH_EVENT_QUEUE_STATE, 3576 /** reset interrupt event, sent to VF on PF reset */ 3577 RTE_ETH_EVENT_INTR_RESET, 3578 RTE_ETH_EVENT_VF_MBOX, /**< message from the VF received by PF */ 3579 RTE_ETH_EVENT_MACSEC, /**< MACsec offload related event */ 3580 RTE_ETH_EVENT_INTR_RMV, /**< device removal event */ 3581 RTE_ETH_EVENT_NEW, /**< port is probed */ 3582 RTE_ETH_EVENT_DESTROY, /**< port is released */ 3583 RTE_ETH_EVENT_IPSEC, /**< IPsec offload related event */ 3584 RTE_ETH_EVENT_FLOW_AGED,/**< New aged-out flows is detected */ 3585 RTE_ETH_EVENT_MAX /**< max value of this enum */ 3586 }; 3587 3588 /** User application callback to be registered for interrupts. */ 3589 typedef int (*rte_eth_dev_cb_fn)(uint16_t port_id, 3590 enum rte_eth_event_type event, void *cb_arg, void *ret_param); 3591 3592 /** 3593 * Register a callback function for port event. 3594 * 3595 * @param port_id 3596 * Port ID. 3597 * RTE_ETH_ALL means register the event for all port ids. 3598 * @param event 3599 * Event interested. 3600 * @param cb_fn 3601 * User supplied callback function to be called. 3602 * @param cb_arg 3603 * Pointer to the parameters for the registered callback. 3604 * 3605 * @return 3606 * - On success, zero. 3607 * - On failure, a negative value. 3608 */ 3609 int rte_eth_dev_callback_register(uint16_t port_id, 3610 enum rte_eth_event_type event, 3611 rte_eth_dev_cb_fn cb_fn, void *cb_arg); 3612 3613 /** 3614 * Unregister a callback function for port event. 3615 * 3616 * @param port_id 3617 * Port ID. 3618 * RTE_ETH_ALL means unregister the event for all port ids. 3619 * @param event 3620 * Event interested. 3621 * @param cb_fn 3622 * User supplied callback function to be called. 3623 * @param cb_arg 3624 * Pointer to the parameters for the registered callback. -1 means to 3625 * remove all for the same callback address and same event. 3626 * 3627 * @return 3628 * - On success, zero. 3629 * - On failure, a negative value. 3630 */ 3631 int rte_eth_dev_callback_unregister(uint16_t port_id, 3632 enum rte_eth_event_type event, 3633 rte_eth_dev_cb_fn cb_fn, void *cb_arg); 3634 3635 /** 3636 * When there is no Rx packet coming in Rx Queue for a long time, we can 3637 * sleep lcore related to Rx Queue for power saving, and enable Rx interrupt 3638 * to be triggered when Rx packet arrives. 3639 * 3640 * The rte_eth_dev_rx_intr_enable() function enables Rx queue 3641 * interrupt on specific Rx queue of a port. 3642 * 3643 * @param port_id 3644 * The port identifier of the Ethernet device. 3645 * @param queue_id 3646 * The index of the receive queue from which to retrieve input packets. 3647 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3648 * to rte_eth_dev_configure(). 3649 * @return 3650 * - (0) if successful. 3651 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3652 * that operation. 3653 * - (-ENODEV) if *port_id* invalid. 3654 * - (-EIO) if device is removed. 3655 */ 3656 int rte_eth_dev_rx_intr_enable(uint16_t port_id, uint16_t queue_id); 3657 3658 /** 3659 * When lcore wakes up from Rx interrupt indicating packet coming, disable Rx 3660 * interrupt and returns to polling mode. 3661 * 3662 * The rte_eth_dev_rx_intr_disable() function disables Rx queue 3663 * interrupt on specific Rx queue of a port. 3664 * 3665 * @param port_id 3666 * The port identifier of the Ethernet device. 3667 * @param queue_id 3668 * The index of the receive queue from which to retrieve input packets. 3669 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3670 * to rte_eth_dev_configure(). 3671 * @return 3672 * - (0) if successful. 3673 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3674 * that operation. 3675 * - (-ENODEV) if *port_id* invalid. 3676 * - (-EIO) if device is removed. 3677 */ 3678 int rte_eth_dev_rx_intr_disable(uint16_t port_id, uint16_t queue_id); 3679 3680 /** 3681 * Rx Interrupt control per port. 3682 * 3683 * @param port_id 3684 * The port identifier of the Ethernet device. 3685 * @param epfd 3686 * Epoll instance fd which the intr vector associated to. 3687 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance. 3688 * @param op 3689 * The operation be performed for the vector. 3690 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}. 3691 * @param data 3692 * User raw data. 3693 * @return 3694 * - On success, zero. 3695 * - On failure, a negative value. 3696 */ 3697 int rte_eth_dev_rx_intr_ctl(uint16_t port_id, int epfd, int op, void *data); 3698 3699 /** 3700 * Rx Interrupt control per queue. 3701 * 3702 * @param port_id 3703 * The port identifier of the Ethernet device. 3704 * @param queue_id 3705 * The index of the receive queue from which to retrieve input packets. 3706 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3707 * to rte_eth_dev_configure(). 3708 * @param epfd 3709 * Epoll instance fd which the intr vector associated to. 3710 * Using RTE_EPOLL_PER_THREAD allows to use per thread epoll instance. 3711 * @param op 3712 * The operation be performed for the vector. 3713 * Operation type of {RTE_INTR_EVENT_ADD, RTE_INTR_EVENT_DEL}. 3714 * @param data 3715 * User raw data. 3716 * @return 3717 * - On success, zero. 3718 * - On failure, a negative value. 3719 */ 3720 int rte_eth_dev_rx_intr_ctl_q(uint16_t port_id, uint16_t queue_id, 3721 int epfd, int op, void *data); 3722 3723 /** 3724 * Get interrupt fd per Rx queue. 3725 * 3726 * @param port_id 3727 * The port identifier of the Ethernet device. 3728 * @param queue_id 3729 * The index of the receive queue from which to retrieve input packets. 3730 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 3731 * to rte_eth_dev_configure(). 3732 * @return 3733 * - (>=0) the interrupt fd associated to the requested Rx queue if 3734 * successful. 3735 * - (-1) on error. 3736 */ 3737 int 3738 rte_eth_dev_rx_intr_ctl_q_get_fd(uint16_t port_id, uint16_t queue_id); 3739 3740 /** 3741 * Turn on the LED on the Ethernet device. 3742 * This function turns on the LED on the Ethernet device. 3743 * 3744 * @param port_id 3745 * The port identifier of the Ethernet device. 3746 * @return 3747 * - (0) if successful. 3748 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3749 * that operation. 3750 * - (-ENODEV) if *port_id* invalid. 3751 * - (-EIO) if device is removed. 3752 */ 3753 int rte_eth_led_on(uint16_t port_id); 3754 3755 /** 3756 * Turn off the LED on the Ethernet device. 3757 * This function turns off the LED on the Ethernet device. 3758 * 3759 * @param port_id 3760 * The port identifier of the Ethernet device. 3761 * @return 3762 * - (0) if successful. 3763 * - (-ENOTSUP) if underlying hardware OR driver doesn't support 3764 * that operation. 3765 * - (-ENODEV) if *port_id* invalid. 3766 * - (-EIO) if device is removed. 3767 */ 3768 int rte_eth_led_off(uint16_t port_id); 3769 3770 /** 3771 * @warning 3772 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 3773 * 3774 * Get Forward Error Correction(FEC) capability. 3775 * 3776 * @param port_id 3777 * The port identifier of the Ethernet device. 3778 * @param speed_fec_capa 3779 * speed_fec_capa is out only with per-speed capabilities. 3780 * If set to NULL, the function returns the required number 3781 * of required array entries. 3782 * @param num 3783 * a number of elements in an speed_fec_capa array. 3784 * 3785 * @return 3786 * - A non-negative value lower or equal to num: success. The return value 3787 * is the number of entries filled in the fec capa array. 3788 * - A non-negative value higher than num: error, the given fec capa array 3789 * is too small. The return value corresponds to the num that should 3790 * be given to succeed. The entries in fec capa array are not valid and 3791 * shall not be used by the caller. 3792 * - (-ENOTSUP) if underlying hardware OR driver doesn't support. 3793 * that operation. 3794 * - (-EIO) if device is removed. 3795 * - (-ENODEV) if *port_id* invalid. 3796 * - (-EINVAL) if *num* or *speed_fec_capa* invalid 3797 */ 3798 __rte_experimental 3799 int rte_eth_fec_get_capability(uint16_t port_id, 3800 struct rte_eth_fec_capa *speed_fec_capa, 3801 unsigned int num); 3802 3803 /** 3804 * @warning 3805 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 3806 * 3807 * Get current Forward Error Correction(FEC) mode. 3808 * If link is down and AUTO is enabled, AUTO is returned, otherwise, 3809 * configured FEC mode is returned. 3810 * If link is up, current FEC mode is returned. 3811 * 3812 * @param port_id 3813 * The port identifier of the Ethernet device. 3814 * @param fec_capa 3815 * A bitmask of enabled FEC modes. If AUTO bit is set, other 3816 * bits specify FEC modes which may be negotiated. If AUTO 3817 * bit is clear, specify FEC modes to be used (only one valid 3818 * mode per speed may be set). 3819 * @return 3820 * - (0) if successful. 3821 * - (-ENOTSUP) if underlying hardware OR driver doesn't support. 3822 * that operation. 3823 * - (-EIO) if device is removed. 3824 * - (-ENODEV) if *port_id* invalid. 3825 */ 3826 __rte_experimental 3827 int rte_eth_fec_get(uint16_t port_id, uint32_t *fec_capa); 3828 3829 /** 3830 * @warning 3831 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 3832 * 3833 * Set Forward Error Correction(FEC) mode. 3834 * 3835 * @param port_id 3836 * The port identifier of the Ethernet device. 3837 * @param fec_capa 3838 * A bitmask of allowed FEC modes. If AUTO bit is set, other 3839 * bits specify FEC modes which may be negotiated. If AUTO 3840 * bit is clear, specify FEC modes to be used (only one valid 3841 * mode per speed may be set). 3842 * @return 3843 * - (0) if successful. 3844 * - (-EINVAL) if the FEC mode is not valid. 3845 * - (-ENOTSUP) if underlying hardware OR driver doesn't support. 3846 * - (-EIO) if device is removed. 3847 * - (-ENODEV) if *port_id* invalid. 3848 */ 3849 __rte_experimental 3850 int rte_eth_fec_set(uint16_t port_id, uint32_t fec_capa); 3851 3852 /** 3853 * Get current status of the Ethernet link flow control for Ethernet device 3854 * 3855 * @param port_id 3856 * The port identifier of the Ethernet device. 3857 * @param fc_conf 3858 * The pointer to the structure where to store the flow control parameters. 3859 * @return 3860 * - (0) if successful. 3861 * - (-ENOTSUP) if hardware doesn't support flow control. 3862 * - (-ENODEV) if *port_id* invalid. 3863 * - (-EIO) if device is removed. 3864 * - (-EINVAL) if bad parameter. 3865 */ 3866 int rte_eth_dev_flow_ctrl_get(uint16_t port_id, 3867 struct rte_eth_fc_conf *fc_conf); 3868 3869 /** 3870 * Configure the Ethernet link flow control for Ethernet device 3871 * 3872 * @param port_id 3873 * The port identifier of the Ethernet device. 3874 * @param fc_conf 3875 * The pointer to the structure of the flow control parameters. 3876 * @return 3877 * - (0) if successful. 3878 * - (-ENOTSUP) if hardware doesn't support flow control mode. 3879 * - (-ENODEV) if *port_id* invalid. 3880 * - (-EINVAL) if bad parameter 3881 * - (-EIO) if flow control setup failure or device is removed. 3882 */ 3883 int rte_eth_dev_flow_ctrl_set(uint16_t port_id, 3884 struct rte_eth_fc_conf *fc_conf); 3885 3886 /** 3887 * Configure the Ethernet priority flow control under DCB environment 3888 * for Ethernet device. 3889 * 3890 * @param port_id 3891 * The port identifier of the Ethernet device. 3892 * @param pfc_conf 3893 * The pointer to the structure of the priority flow control parameters. 3894 * @return 3895 * - (0) if successful. 3896 * - (-ENOTSUP) if hardware doesn't support priority flow control mode. 3897 * - (-ENODEV) if *port_id* invalid. 3898 * - (-EINVAL) if bad parameter 3899 * - (-EIO) if flow control setup failure or device is removed. 3900 */ 3901 int rte_eth_dev_priority_flow_ctrl_set(uint16_t port_id, 3902 struct rte_eth_pfc_conf *pfc_conf); 3903 3904 /** 3905 * Add a MAC address to the set used for filtering incoming packets. 3906 * 3907 * @param port_id 3908 * The port identifier of the Ethernet device. 3909 * @param mac_addr 3910 * The MAC address to add. 3911 * @param pool 3912 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is 3913 * not enabled, this should be set to 0. 3914 * @return 3915 * - (0) if successfully added or *mac_addr* was already added. 3916 * - (-ENOTSUP) if hardware doesn't support this feature. 3917 * - (-ENODEV) if *port* is invalid. 3918 * - (-EIO) if device is removed. 3919 * - (-ENOSPC) if no more MAC addresses can be added. 3920 * - (-EINVAL) if MAC address is invalid. 3921 */ 3922 int rte_eth_dev_mac_addr_add(uint16_t port_id, struct rte_ether_addr *mac_addr, 3923 uint32_t pool); 3924 3925 /** 3926 * Remove a MAC address from the internal array of addresses. 3927 * 3928 * @param port_id 3929 * The port identifier of the Ethernet device. 3930 * @param mac_addr 3931 * MAC address to remove. 3932 * @return 3933 * - (0) if successful, or *mac_addr* didn't exist. 3934 * - (-ENOTSUP) if hardware doesn't support. 3935 * - (-ENODEV) if *port* invalid. 3936 * - (-EADDRINUSE) if attempting to remove the default MAC address. 3937 * - (-EINVAL) if MAC address is invalid. 3938 */ 3939 int rte_eth_dev_mac_addr_remove(uint16_t port_id, 3940 struct rte_ether_addr *mac_addr); 3941 3942 /** 3943 * Set the default MAC address. 3944 * 3945 * @param port_id 3946 * The port identifier of the Ethernet device. 3947 * @param mac_addr 3948 * New default MAC address. 3949 * @return 3950 * - (0) if successful, or *mac_addr* didn't exist. 3951 * - (-ENOTSUP) if hardware doesn't support. 3952 * - (-ENODEV) if *port* invalid. 3953 * - (-EINVAL) if MAC address is invalid. 3954 */ 3955 int rte_eth_dev_default_mac_addr_set(uint16_t port_id, 3956 struct rte_ether_addr *mac_addr); 3957 3958 /** 3959 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device. 3960 * 3961 * @param port_id 3962 * The port identifier of the Ethernet device. 3963 * @param reta_conf 3964 * RETA to update. 3965 * @param reta_size 3966 * Redirection table size. The table size can be queried by 3967 * rte_eth_dev_info_get(). 3968 * @return 3969 * - (0) if successful. 3970 * - (-ENODEV) if *port_id* is invalid. 3971 * - (-ENOTSUP) if hardware doesn't support. 3972 * - (-EINVAL) if bad parameter. 3973 * - (-EIO) if device is removed. 3974 */ 3975 int rte_eth_dev_rss_reta_update(uint16_t port_id, 3976 struct rte_eth_rss_reta_entry64 *reta_conf, 3977 uint16_t reta_size); 3978 3979 /** 3980 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device. 3981 * 3982 * @param port_id 3983 * The port identifier of the Ethernet device. 3984 * @param reta_conf 3985 * RETA to query. For each requested reta entry, corresponding bit 3986 * in mask must be set. 3987 * @param reta_size 3988 * Redirection table size. The table size can be queried by 3989 * rte_eth_dev_info_get(). 3990 * @return 3991 * - (0) if successful. 3992 * - (-ENODEV) if *port_id* is invalid. 3993 * - (-ENOTSUP) if hardware doesn't support. 3994 * - (-EINVAL) if bad parameter. 3995 * - (-EIO) if device is removed. 3996 */ 3997 int rte_eth_dev_rss_reta_query(uint16_t port_id, 3998 struct rte_eth_rss_reta_entry64 *reta_conf, 3999 uint16_t reta_size); 4000 4001 /** 4002 * Updates unicast hash table for receiving packet with the given destination 4003 * MAC address, and the packet is routed to all VFs for which the Rx mode is 4004 * accept packets that match the unicast hash table. 4005 * 4006 * @param port_id 4007 * The port identifier of the Ethernet device. 4008 * @param addr 4009 * Unicast MAC address. 4010 * @param on 4011 * 1 - Set an unicast hash bit for receiving packets with the MAC address. 4012 * 0 - Clear an unicast hash bit. 4013 * @return 4014 * - (0) if successful. 4015 * - (-ENOTSUP) if hardware doesn't support. 4016 * - (-ENODEV) if *port_id* invalid. 4017 * - (-EIO) if device is removed. 4018 * - (-EINVAL) if bad parameter. 4019 */ 4020 int rte_eth_dev_uc_hash_table_set(uint16_t port_id, struct rte_ether_addr *addr, 4021 uint8_t on); 4022 4023 /** 4024 * Updates all unicast hash bitmaps for receiving packet with any Unicast 4025 * Ethernet MAC addresses,the packet is routed to all VFs for which the Rx 4026 * mode is accept packets that match the unicast hash table. 4027 * 4028 * @param port_id 4029 * The port identifier of the Ethernet device. 4030 * @param on 4031 * 1 - Set all unicast hash bitmaps for receiving all the Ethernet 4032 * MAC addresses 4033 * 0 - Clear all unicast hash bitmaps 4034 * @return 4035 * - (0) if successful. 4036 * - (-ENOTSUP) if hardware doesn't support. 4037 * - (-ENODEV) if *port_id* invalid. 4038 * - (-EIO) if device is removed. 4039 * - (-EINVAL) if bad parameter. 4040 */ 4041 int rte_eth_dev_uc_all_hash_table_set(uint16_t port_id, uint8_t on); 4042 4043 /** 4044 * Set the rate limitation for a queue on an Ethernet device. 4045 * 4046 * @param port_id 4047 * The port identifier of the Ethernet device. 4048 * @param queue_idx 4049 * The queue ID. 4050 * @param tx_rate 4051 * The Tx rate in Mbps. Allocated from the total port link speed. 4052 * @return 4053 * - (0) if successful. 4054 * - (-ENOTSUP) if hardware doesn't support this feature. 4055 * - (-ENODEV) if *port_id* invalid. 4056 * - (-EIO) if device is removed. 4057 * - (-EINVAL) if bad parameter. 4058 */ 4059 int rte_eth_set_queue_rate_limit(uint16_t port_id, uint16_t queue_idx, 4060 uint16_t tx_rate); 4061 4062 /** 4063 * Configuration of Receive Side Scaling hash computation of Ethernet device. 4064 * 4065 * @param port_id 4066 * The port identifier of the Ethernet device. 4067 * @param rss_conf 4068 * The new configuration to use for RSS hash computation on the port. 4069 * @return 4070 * - (0) if successful. 4071 * - (-ENODEV) if port identifier is invalid. 4072 * - (-EIO) if device is removed. 4073 * - (-ENOTSUP) if hardware doesn't support. 4074 * - (-EINVAL) if bad parameter. 4075 */ 4076 int rte_eth_dev_rss_hash_update(uint16_t port_id, 4077 struct rte_eth_rss_conf *rss_conf); 4078 4079 /** 4080 * Retrieve current configuration of Receive Side Scaling hash computation 4081 * of Ethernet device. 4082 * 4083 * @param port_id 4084 * The port identifier of the Ethernet device. 4085 * @param rss_conf 4086 * Where to store the current RSS hash configuration of the Ethernet device. 4087 * @return 4088 * - (0) if successful. 4089 * - (-ENODEV) if port identifier is invalid. 4090 * - (-EIO) if device is removed. 4091 * - (-ENOTSUP) if hardware doesn't support RSS. 4092 * - (-EINVAL) if bad parameter. 4093 */ 4094 int 4095 rte_eth_dev_rss_hash_conf_get(uint16_t port_id, 4096 struct rte_eth_rss_conf *rss_conf); 4097 4098 /** 4099 * Add UDP tunneling port for a type of tunnel. 4100 * 4101 * Some NICs may require such configuration to properly parse a tunnel 4102 * with any standard or custom UDP port. 4103 * The packets with this UDP port will be parsed for this type of tunnel. 4104 * The device parser will also check the rest of the tunnel headers 4105 * before classifying the packet. 4106 * 4107 * With some devices, this API will affect packet classification, i.e.: 4108 * - mbuf.packet_type reported on Rx 4109 * - rte_flow rules with tunnel items 4110 * 4111 * @param port_id 4112 * The port identifier of the Ethernet device. 4113 * @param tunnel_udp 4114 * UDP tunneling configuration. 4115 * 4116 * @return 4117 * - (0) if successful. 4118 * - (-ENODEV) if port identifier is invalid. 4119 * - (-EIO) if device is removed. 4120 * - (-ENOTSUP) if hardware doesn't support tunnel type. 4121 */ 4122 int 4123 rte_eth_dev_udp_tunnel_port_add(uint16_t port_id, 4124 struct rte_eth_udp_tunnel *tunnel_udp); 4125 4126 /** 4127 * Delete UDP tunneling port for a type of tunnel. 4128 * 4129 * The packets with this UDP port will not be classified as this type of tunnel 4130 * anymore if the device use such mapping for tunnel packet classification. 4131 * 4132 * @see rte_eth_dev_udp_tunnel_port_add 4133 * 4134 * @param port_id 4135 * The port identifier of the Ethernet device. 4136 * @param tunnel_udp 4137 * UDP tunneling configuration. 4138 * 4139 * @return 4140 * - (0) if successful. 4141 * - (-ENODEV) if port identifier is invalid. 4142 * - (-EIO) if device is removed. 4143 * - (-ENOTSUP) if hardware doesn't support tunnel type. 4144 */ 4145 int 4146 rte_eth_dev_udp_tunnel_port_delete(uint16_t port_id, 4147 struct rte_eth_udp_tunnel *tunnel_udp); 4148 4149 /** 4150 * Get DCB information on an Ethernet device. 4151 * 4152 * @param port_id 4153 * The port identifier of the Ethernet device. 4154 * @param dcb_info 4155 * DCB information. 4156 * @return 4157 * - (0) if successful. 4158 * - (-ENODEV) if port identifier is invalid. 4159 * - (-EIO) if device is removed. 4160 * - (-ENOTSUP) if hardware doesn't support. 4161 * - (-EINVAL) if bad parameter. 4162 */ 4163 int rte_eth_dev_get_dcb_info(uint16_t port_id, 4164 struct rte_eth_dcb_info *dcb_info); 4165 4166 struct rte_eth_rxtx_callback; 4167 4168 /** 4169 * Add a callback to be called on packet Rx on a given port and queue. 4170 * 4171 * This API configures a function to be called for each burst of 4172 * packets received on a given NIC port queue. The return value is a pointer 4173 * that can be used to later remove the callback using 4174 * rte_eth_remove_rx_callback(). 4175 * 4176 * Multiple functions are called in the order that they are added. 4177 * 4178 * @param port_id 4179 * The port identifier of the Ethernet device. 4180 * @param queue_id 4181 * The queue on the Ethernet device on which the callback is to be added. 4182 * @param fn 4183 * The callback function 4184 * @param user_param 4185 * A generic pointer parameter which will be passed to each invocation of the 4186 * callback function on this port and queue. Inter-thread synchronization 4187 * of any user data changes is the responsibility of the user. 4188 * 4189 * @return 4190 * NULL on error. 4191 * On success, a pointer value which can later be used to remove the callback. 4192 */ 4193 const struct rte_eth_rxtx_callback * 4194 rte_eth_add_rx_callback(uint16_t port_id, uint16_t queue_id, 4195 rte_rx_callback_fn fn, void *user_param); 4196 4197 /** 4198 * Add a callback that must be called first on packet Rx on a given port 4199 * and queue. 4200 * 4201 * This API configures a first function to be called for each burst of 4202 * packets received on a given NIC port queue. The return value is a pointer 4203 * that can be used to later remove the callback using 4204 * rte_eth_remove_rx_callback(). 4205 * 4206 * Multiple functions are called in the order that they are added. 4207 * 4208 * @param port_id 4209 * The port identifier of the Ethernet device. 4210 * @param queue_id 4211 * The queue on the Ethernet device on which the callback is to be added. 4212 * @param fn 4213 * The callback function 4214 * @param user_param 4215 * A generic pointer parameter which will be passed to each invocation of the 4216 * callback function on this port and queue. Inter-thread synchronization 4217 * of any user data changes is the responsibility of the user. 4218 * 4219 * @return 4220 * NULL on error. 4221 * On success, a pointer value which can later be used to remove the callback. 4222 */ 4223 const struct rte_eth_rxtx_callback * 4224 rte_eth_add_first_rx_callback(uint16_t port_id, uint16_t queue_id, 4225 rte_rx_callback_fn fn, void *user_param); 4226 4227 /** 4228 * Add a callback to be called on packet Tx on a given port and queue. 4229 * 4230 * This API configures a function to be called for each burst of 4231 * packets sent on a given NIC port queue. The return value is a pointer 4232 * that can be used to later remove the callback using 4233 * rte_eth_remove_tx_callback(). 4234 * 4235 * Multiple functions are called in the order that they are added. 4236 * 4237 * @param port_id 4238 * The port identifier of the Ethernet device. 4239 * @param queue_id 4240 * The queue on the Ethernet device on which the callback is to be added. 4241 * @param fn 4242 * The callback function 4243 * @param user_param 4244 * A generic pointer parameter which will be passed to each invocation of the 4245 * callback function on this port and queue. Inter-thread synchronization 4246 * of any user data changes is the responsibility of the user. 4247 * 4248 * @return 4249 * NULL on error. 4250 * On success, a pointer value which can later be used to remove the callback. 4251 */ 4252 const struct rte_eth_rxtx_callback * 4253 rte_eth_add_tx_callback(uint16_t port_id, uint16_t queue_id, 4254 rte_tx_callback_fn fn, void *user_param); 4255 4256 /** 4257 * Remove an Rx packet callback from a given port and queue. 4258 * 4259 * This function is used to removed callbacks that were added to a NIC port 4260 * queue using rte_eth_add_rx_callback(). 4261 * 4262 * Note: the callback is removed from the callback list but it isn't freed 4263 * since the it may still be in use. The memory for the callback can be 4264 * subsequently freed back by the application by calling rte_free(): 4265 * 4266 * - Immediately - if the port is stopped, or the user knows that no 4267 * callbacks are in flight e.g. if called from the thread doing Rx/Tx 4268 * on that queue. 4269 * 4270 * - After a short delay - where the delay is sufficient to allow any 4271 * in-flight callbacks to complete. Alternately, the RCU mechanism can be 4272 * used to detect when data plane threads have ceased referencing the 4273 * callback memory. 4274 * 4275 * @param port_id 4276 * The port identifier of the Ethernet device. 4277 * @param queue_id 4278 * The queue on the Ethernet device from which the callback is to be removed. 4279 * @param user_cb 4280 * User supplied callback created via rte_eth_add_rx_callback(). 4281 * 4282 * @return 4283 * - 0: Success. Callback was removed. 4284 * - -ENODEV: If *port_id* is invalid. 4285 * - -ENOTSUP: Callback support is not available. 4286 * - -EINVAL: The queue_id is out of range, or the callback 4287 * is NULL or not found for the port/queue. 4288 */ 4289 int rte_eth_remove_rx_callback(uint16_t port_id, uint16_t queue_id, 4290 const struct rte_eth_rxtx_callback *user_cb); 4291 4292 /** 4293 * Remove a Tx packet callback from a given port and queue. 4294 * 4295 * This function is used to removed callbacks that were added to a NIC port 4296 * queue using rte_eth_add_tx_callback(). 4297 * 4298 * Note: the callback is removed from the callback list but it isn't freed 4299 * since the it may still be in use. The memory for the callback can be 4300 * subsequently freed back by the application by calling rte_free(): 4301 * 4302 * - Immediately - if the port is stopped, or the user knows that no 4303 * callbacks are in flight e.g. if called from the thread doing Rx/Tx 4304 * on that queue. 4305 * 4306 * - After a short delay - where the delay is sufficient to allow any 4307 * in-flight callbacks to complete. Alternately, the RCU mechanism can be 4308 * used to detect when data plane threads have ceased referencing the 4309 * callback memory. 4310 * 4311 * @param port_id 4312 * The port identifier of the Ethernet device. 4313 * @param queue_id 4314 * The queue on the Ethernet device from which the callback is to be removed. 4315 * @param user_cb 4316 * User supplied callback created via rte_eth_add_tx_callback(). 4317 * 4318 * @return 4319 * - 0: Success. Callback was removed. 4320 * - -ENODEV: If *port_id* is invalid. 4321 * - -ENOTSUP: Callback support is not available. 4322 * - -EINVAL: The queue_id is out of range, or the callback 4323 * is NULL or not found for the port/queue. 4324 */ 4325 int rte_eth_remove_tx_callback(uint16_t port_id, uint16_t queue_id, 4326 const struct rte_eth_rxtx_callback *user_cb); 4327 4328 /** 4329 * Retrieve information about given port's Rx queue. 4330 * 4331 * @param port_id 4332 * The port identifier of the Ethernet device. 4333 * @param queue_id 4334 * The Rx queue on the Ethernet device for which information 4335 * will be retrieved. 4336 * @param qinfo 4337 * A pointer to a structure of type *rte_eth_rxq_info_info* to be filled with 4338 * the information of the Ethernet device. 4339 * 4340 * @return 4341 * - 0: Success 4342 * - -ENODEV: If *port_id* is invalid. 4343 * - -ENOTSUP: routine is not supported by the device PMD. 4344 * - -EINVAL: The queue_id is out of range, or the queue 4345 * is hairpin queue. 4346 */ 4347 int rte_eth_rx_queue_info_get(uint16_t port_id, uint16_t queue_id, 4348 struct rte_eth_rxq_info *qinfo); 4349 4350 /** 4351 * Retrieve information about given port's Tx queue. 4352 * 4353 * @param port_id 4354 * The port identifier of the Ethernet device. 4355 * @param queue_id 4356 * The Tx queue on the Ethernet device for which information 4357 * will be retrieved. 4358 * @param qinfo 4359 * A pointer to a structure of type *rte_eth_txq_info_info* to be filled with 4360 * the information of the Ethernet device. 4361 * 4362 * @return 4363 * - 0: Success 4364 * - -ENODEV: If *port_id* is invalid. 4365 * - -ENOTSUP: routine is not supported by the device PMD. 4366 * - -EINVAL: The queue_id is out of range, or the queue 4367 * is hairpin queue. 4368 */ 4369 int rte_eth_tx_queue_info_get(uint16_t port_id, uint16_t queue_id, 4370 struct rte_eth_txq_info *qinfo); 4371 4372 /** 4373 * Retrieve information about the Rx packet burst mode. 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 mode 4381 * A pointer to a structure of type *rte_eth_burst_mode* to be filled 4382 * with the information of the packet burst mode. 4383 * 4384 * @return 4385 * - 0: Success 4386 * - -ENODEV: If *port_id* is invalid. 4387 * - -ENOTSUP: routine is not supported by the device PMD. 4388 * - -EINVAL: The queue_id is out of range. 4389 */ 4390 int rte_eth_rx_burst_mode_get(uint16_t port_id, uint16_t queue_id, 4391 struct rte_eth_burst_mode *mode); 4392 4393 /** 4394 * Retrieve information about the Tx packet burst mode. 4395 * 4396 * @param port_id 4397 * The port identifier of the Ethernet device. 4398 * @param queue_id 4399 * The Tx queue on the Ethernet device for which information 4400 * will be retrieved. 4401 * @param mode 4402 * A pointer to a structure of type *rte_eth_burst_mode* to be filled 4403 * with the information of the packet burst mode. 4404 * 4405 * @return 4406 * - 0: Success 4407 * - -ENODEV: If *port_id* is invalid. 4408 * - -ENOTSUP: routine is not supported by the device PMD. 4409 * - -EINVAL: The queue_id is out of range. 4410 */ 4411 int rte_eth_tx_burst_mode_get(uint16_t port_id, uint16_t queue_id, 4412 struct rte_eth_burst_mode *mode); 4413 4414 /** 4415 * @warning 4416 * @b EXPERIMENTAL: this API may change without prior notice. 4417 * 4418 * Retrieve the monitor condition for a given receive queue. 4419 * 4420 * @param port_id 4421 * The port identifier of the Ethernet device. 4422 * @param queue_id 4423 * The Rx queue on the Ethernet device for which information 4424 * will be retrieved. 4425 * @param pmc 4426 * The pointer to power-optimized monitoring condition structure. 4427 * 4428 * @return 4429 * - 0: Success. 4430 * -ENOTSUP: Operation not supported. 4431 * -EINVAL: Invalid parameters. 4432 * -ENODEV: Invalid port ID. 4433 */ 4434 __rte_experimental 4435 int rte_eth_get_monitor_addr(uint16_t port_id, uint16_t queue_id, 4436 struct rte_power_monitor_cond *pmc); 4437 4438 /** 4439 * Retrieve device registers and register attributes (number of registers and 4440 * register size) 4441 * 4442 * @param port_id 4443 * The port identifier of the Ethernet device. 4444 * @param info 4445 * Pointer to rte_dev_reg_info structure to fill in. If info->data is 4446 * NULL the function fills in the width and length fields. If non-NULL 4447 * the registers are put into the buffer pointed at by the data field. 4448 * @return 4449 * - (0) if successful. 4450 * - (-ENOTSUP) if hardware doesn't support. 4451 * - (-EINVAL) if bad parameter. 4452 * - (-ENODEV) if *port_id* invalid. 4453 * - (-EIO) if device is removed. 4454 * - others depends on the specific operations implementation. 4455 */ 4456 int rte_eth_dev_get_reg_info(uint16_t port_id, struct rte_dev_reg_info *info); 4457 4458 /** 4459 * Retrieve size of device EEPROM 4460 * 4461 * @param port_id 4462 * The port identifier of the Ethernet device. 4463 * @return 4464 * - (>=0) EEPROM size if successful. 4465 * - (-ENOTSUP) if hardware doesn't support. 4466 * - (-ENODEV) if *port_id* invalid. 4467 * - (-EIO) if device is removed. 4468 * - others depends on the specific operations implementation. 4469 */ 4470 int rte_eth_dev_get_eeprom_length(uint16_t port_id); 4471 4472 /** 4473 * Retrieve EEPROM and EEPROM attribute 4474 * 4475 * @param port_id 4476 * The port identifier of the Ethernet device. 4477 * @param info 4478 * The template includes buffer for return EEPROM data and 4479 * EEPROM attributes to be filled. 4480 * @return 4481 * - (0) if successful. 4482 * - (-ENOTSUP) if hardware doesn't support. 4483 * - (-EINVAL) if bad parameter. 4484 * - (-ENODEV) if *port_id* invalid. 4485 * - (-EIO) if device is removed. 4486 * - others depends on the specific operations implementation. 4487 */ 4488 int rte_eth_dev_get_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info); 4489 4490 /** 4491 * Program EEPROM with provided data 4492 * 4493 * @param port_id 4494 * The port identifier of the Ethernet device. 4495 * @param info 4496 * The template includes EEPROM data for programming and 4497 * EEPROM attributes to be filled 4498 * @return 4499 * - (0) if successful. 4500 * - (-ENOTSUP) if hardware doesn't support. 4501 * - (-ENODEV) if *port_id* invalid. 4502 * - (-EINVAL) if bad parameter. 4503 * - (-EIO) if device is removed. 4504 * - others depends on the specific operations implementation. 4505 */ 4506 int rte_eth_dev_set_eeprom(uint16_t port_id, struct rte_dev_eeprom_info *info); 4507 4508 /** 4509 * @warning 4510 * @b EXPERIMENTAL: this API may change without prior notice. 4511 * 4512 * Retrieve the type and size of plugin module EEPROM 4513 * 4514 * @param port_id 4515 * The port identifier of the Ethernet device. 4516 * @param modinfo 4517 * The type and size of plugin module EEPROM. 4518 * @return 4519 * - (0) if successful. 4520 * - (-ENOTSUP) if hardware doesn't support. 4521 * - (-ENODEV) if *port_id* invalid. 4522 * - (-EINVAL) if bad parameter. 4523 * - (-EIO) if device is removed. 4524 * - others depends on the specific operations implementation. 4525 */ 4526 __rte_experimental 4527 int 4528 rte_eth_dev_get_module_info(uint16_t port_id, 4529 struct rte_eth_dev_module_info *modinfo); 4530 4531 /** 4532 * @warning 4533 * @b EXPERIMENTAL: this API may change without prior notice. 4534 * 4535 * Retrieve the data of plugin module EEPROM 4536 * 4537 * @param port_id 4538 * The port identifier of the Ethernet device. 4539 * @param info 4540 * The template includes the plugin module EEPROM attributes, and the 4541 * buffer for return plugin module EEPROM data. 4542 * @return 4543 * - (0) if successful. 4544 * - (-ENOTSUP) if hardware doesn't support. 4545 * - (-EINVAL) if bad parameter. 4546 * - (-ENODEV) if *port_id* invalid. 4547 * - (-EIO) if device is removed. 4548 * - others depends on the specific operations implementation. 4549 */ 4550 __rte_experimental 4551 int 4552 rte_eth_dev_get_module_eeprom(uint16_t port_id, 4553 struct rte_dev_eeprom_info *info); 4554 4555 /** 4556 * Set the list of multicast addresses to filter on an Ethernet device. 4557 * 4558 * @param port_id 4559 * The port identifier of the Ethernet device. 4560 * @param mc_addr_set 4561 * The array of multicast addresses to set. Equal to NULL when the function 4562 * is invoked to flush the set of filtered addresses. 4563 * @param nb_mc_addr 4564 * The number of multicast addresses in the *mc_addr_set* array. Equal to 0 4565 * when the function is invoked to flush the set of filtered addresses. 4566 * @return 4567 * - (0) if successful. 4568 * - (-ENODEV) if *port_id* invalid. 4569 * - (-EIO) if device is removed. 4570 * - (-ENOTSUP) if PMD of *port_id* doesn't support multicast filtering. 4571 * - (-ENOSPC) if *port_id* has not enough multicast filtering resources. 4572 * - (-EINVAL) if bad parameter. 4573 */ 4574 int rte_eth_dev_set_mc_addr_list(uint16_t port_id, 4575 struct rte_ether_addr *mc_addr_set, 4576 uint32_t nb_mc_addr); 4577 4578 /** 4579 * Enable IEEE1588/802.1AS timestamping for an Ethernet device. 4580 * 4581 * @param port_id 4582 * The port identifier of the Ethernet device. 4583 * 4584 * @return 4585 * - 0: Success. 4586 * - -ENODEV: The port ID is invalid. 4587 * - -EIO: if device is removed. 4588 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4589 */ 4590 int rte_eth_timesync_enable(uint16_t port_id); 4591 4592 /** 4593 * Disable IEEE1588/802.1AS timestamping for an Ethernet device. 4594 * 4595 * @param port_id 4596 * The port identifier of the Ethernet device. 4597 * 4598 * @return 4599 * - 0: Success. 4600 * - -ENODEV: The port ID is invalid. 4601 * - -EIO: if device is removed. 4602 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4603 */ 4604 int rte_eth_timesync_disable(uint16_t port_id); 4605 4606 /** 4607 * Read an IEEE1588/802.1AS Rx timestamp from an Ethernet device. 4608 * 4609 * @param port_id 4610 * The port identifier of the Ethernet device. 4611 * @param timestamp 4612 * Pointer to the timestamp struct. 4613 * @param flags 4614 * Device specific flags. Used to pass the Rx timesync register index to 4615 * i40e. Unused in igb/ixgbe, pass 0 instead. 4616 * 4617 * @return 4618 * - 0: Success. 4619 * - -EINVAL: No timestamp is available. 4620 * - -ENODEV: The port ID is invalid. 4621 * - -EIO: if device is removed. 4622 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4623 */ 4624 int rte_eth_timesync_read_rx_timestamp(uint16_t port_id, 4625 struct timespec *timestamp, uint32_t flags); 4626 4627 /** 4628 * Read an IEEE1588/802.1AS Tx timestamp from an Ethernet device. 4629 * 4630 * @param port_id 4631 * The port identifier of the Ethernet device. 4632 * @param timestamp 4633 * Pointer to the timestamp struct. 4634 * 4635 * @return 4636 * - 0: Success. 4637 * - -EINVAL: No timestamp is available. 4638 * - -ENODEV: The port ID is invalid. 4639 * - -EIO: if device is removed. 4640 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4641 */ 4642 int rte_eth_timesync_read_tx_timestamp(uint16_t port_id, 4643 struct timespec *timestamp); 4644 4645 /** 4646 * Adjust the timesync clock on an Ethernet device. 4647 * 4648 * This is usually used in conjunction with other Ethdev timesync functions to 4649 * synchronize the device time using the IEEE1588/802.1AS protocol. 4650 * 4651 * @param port_id 4652 * The port identifier of the Ethernet device. 4653 * @param delta 4654 * The adjustment in nanoseconds. 4655 * 4656 * @return 4657 * - 0: Success. 4658 * - -ENODEV: The port ID is invalid. 4659 * - -EIO: if device is removed. 4660 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4661 */ 4662 int rte_eth_timesync_adjust_time(uint16_t port_id, int64_t delta); 4663 4664 /** 4665 * Read the time from the timesync clock on an Ethernet device. 4666 * 4667 * This is usually used in conjunction with other Ethdev timesync functions to 4668 * synchronize the device time using the IEEE1588/802.1AS protocol. 4669 * 4670 * @param port_id 4671 * The port identifier of the Ethernet device. 4672 * @param time 4673 * Pointer to the timespec struct that holds the time. 4674 * 4675 * @return 4676 * - 0: Success. 4677 * - -EINVAL: Bad parameter. 4678 */ 4679 int rte_eth_timesync_read_time(uint16_t port_id, struct timespec *time); 4680 4681 /** 4682 * Set the time of the timesync clock on an Ethernet device. 4683 * 4684 * This is usually used in conjunction with other Ethdev timesync functions to 4685 * synchronize the device time using the IEEE1588/802.1AS protocol. 4686 * 4687 * @param port_id 4688 * The port identifier of the Ethernet device. 4689 * @param time 4690 * Pointer to the timespec struct that holds the time. 4691 * 4692 * @return 4693 * - 0: Success. 4694 * - -EINVAL: No timestamp is available. 4695 * - -ENODEV: The port ID is invalid. 4696 * - -EIO: if device is removed. 4697 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4698 */ 4699 int rte_eth_timesync_write_time(uint16_t port_id, const struct timespec *time); 4700 4701 /** 4702 * @warning 4703 * @b EXPERIMENTAL: this API may change without prior notice. 4704 * 4705 * Read the current clock counter of an Ethernet device 4706 * 4707 * This returns the current raw clock value of an Ethernet device. It is 4708 * a raw amount of ticks, with no given time reference. 4709 * The value returned here is from the same clock than the one 4710 * filling timestamp field of Rx packets when using hardware timestamp 4711 * offload. Therefore it can be used to compute a precise conversion of 4712 * the device clock to the real time. 4713 * 4714 * E.g, a simple heuristic to derivate the frequency would be: 4715 * uint64_t start, end; 4716 * rte_eth_read_clock(port, start); 4717 * rte_delay_ms(100); 4718 * rte_eth_read_clock(port, end); 4719 * double freq = (end - start) * 10; 4720 * 4721 * Compute a common reference with: 4722 * uint64_t base_time_sec = current_time(); 4723 * uint64_t base_clock; 4724 * rte_eth_read_clock(port, base_clock); 4725 * 4726 * Then, convert the raw mbuf timestamp with: 4727 * base_time_sec + (double)(*timestamp_dynfield(mbuf) - base_clock) / freq; 4728 * 4729 * This simple example will not provide a very good accuracy. One must 4730 * at least measure multiple times the frequency and do a regression. 4731 * To avoid deviation from the system time, the common reference can 4732 * be repeated from time to time. The integer division can also be 4733 * converted by a multiplication and a shift for better performance. 4734 * 4735 * @param port_id 4736 * The port identifier of the Ethernet device. 4737 * @param clock 4738 * Pointer to the uint64_t that holds the raw clock value. 4739 * 4740 * @return 4741 * - 0: Success. 4742 * - -ENODEV: The port ID is invalid. 4743 * - -ENOTSUP: The function is not supported by the Ethernet driver. 4744 * - -EINVAL: if bad parameter. 4745 */ 4746 __rte_experimental 4747 int 4748 rte_eth_read_clock(uint16_t port_id, uint64_t *clock); 4749 4750 /** 4751 * Get the port ID from device name. The device name should be specified 4752 * as below: 4753 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:2:00.0 4754 * - SoC device name, for example- fsl-gmac0 4755 * - vdev dpdk name, for example- net_[pcap0|null0|tap0] 4756 * 4757 * @param name 4758 * pci address or name of the device 4759 * @param port_id 4760 * pointer to port identifier of the device 4761 * @return 4762 * - (0) if successful and port_id is filled. 4763 * - (-ENODEV or -EINVAL) on failure. 4764 */ 4765 int 4766 rte_eth_dev_get_port_by_name(const char *name, uint16_t *port_id); 4767 4768 /** 4769 * Get the device name from port ID. The device name is specified as below: 4770 * - PCIe address (Domain:Bus:Device.Function), for example- 0000:02:00.0 4771 * - SoC device name, for example- fsl-gmac0 4772 * - vdev dpdk name, for example- net_[pcap0|null0|tun0|tap0] 4773 * 4774 * @param port_id 4775 * Port identifier of the device. 4776 * @param name 4777 * Buffer of size RTE_ETH_NAME_MAX_LEN to store the name. 4778 * @return 4779 * - (0) if successful. 4780 * - (-ENODEV) if *port_id* is invalid. 4781 * - (-EINVAL) on failure. 4782 */ 4783 int 4784 rte_eth_dev_get_name_by_port(uint16_t port_id, char *name); 4785 4786 /** 4787 * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from 4788 * the Ethernet device information, otherwise adjust them to boundaries. 4789 * 4790 * @param port_id 4791 * The port identifier of the Ethernet device. 4792 * @param nb_rx_desc 4793 * A pointer to a uint16_t where the number of receive 4794 * descriptors stored. 4795 * @param nb_tx_desc 4796 * A pointer to a uint16_t where the number of transmit 4797 * descriptors stored. 4798 * @return 4799 * - (0) if successful. 4800 * - (-ENOTSUP, -ENODEV or -EINVAL) on failure. 4801 */ 4802 int rte_eth_dev_adjust_nb_rx_tx_desc(uint16_t port_id, 4803 uint16_t *nb_rx_desc, 4804 uint16_t *nb_tx_desc); 4805 4806 /** 4807 * Test if a port supports specific mempool ops. 4808 * 4809 * @param port_id 4810 * Port identifier of the Ethernet device. 4811 * @param [in] pool 4812 * The name of the pool operations to test. 4813 * @return 4814 * - 0: best mempool ops choice for this port. 4815 * - 1: mempool ops are supported for this port. 4816 * - -ENOTSUP: mempool ops not supported for this port. 4817 * - -ENODEV: Invalid port Identifier. 4818 * - -EINVAL: Pool param is null. 4819 */ 4820 int 4821 rte_eth_dev_pool_ops_supported(uint16_t port_id, const char *pool); 4822 4823 /** 4824 * Get the security context for the Ethernet device. 4825 * 4826 * @param port_id 4827 * Port identifier of the Ethernet device 4828 * @return 4829 * - NULL on error. 4830 * - pointer to security context on success. 4831 */ 4832 void * 4833 rte_eth_dev_get_sec_ctx(uint16_t port_id); 4834 4835 /** 4836 * @warning 4837 * @b EXPERIMENTAL: this API may change, or be removed, without prior notice 4838 * 4839 * Query the device hairpin capabilities. 4840 * 4841 * @param port_id 4842 * The port identifier of the Ethernet device. 4843 * @param cap 4844 * Pointer to a structure that will hold the hairpin capabilities. 4845 * @return 4846 * - (0) if successful. 4847 * - (-ENOTSUP) if hardware doesn't support. 4848 * - (-EINVAL) if bad parameter. 4849 */ 4850 __rte_experimental 4851 int rte_eth_dev_hairpin_capability_get(uint16_t port_id, 4852 struct rte_eth_hairpin_cap *cap); 4853 4854 /** 4855 * @warning 4856 * @b EXPERIMENTAL: this structure may change without prior notice. 4857 * 4858 * Ethernet device representor ID range entry 4859 */ 4860 struct rte_eth_representor_range { 4861 enum rte_eth_representor_type type; /**< Representor type */ 4862 int controller; /**< Controller index */ 4863 int pf; /**< Physical function index */ 4864 __extension__ 4865 union { 4866 int vf; /**< VF start index */ 4867 int sf; /**< SF start index */ 4868 }; 4869 uint32_t id_base; /**< Representor ID start index */ 4870 uint32_t id_end; /**< Representor ID end index */ 4871 char name[RTE_DEV_NAME_MAX_LEN]; /**< Representor name */ 4872 }; 4873 4874 /** 4875 * @warning 4876 * @b EXPERIMENTAL: this structure may change without prior notice. 4877 * 4878 * Ethernet device representor information 4879 */ 4880 struct rte_eth_representor_info { 4881 uint16_t controller; /**< Controller ID of caller device. */ 4882 uint16_t pf; /**< Physical function ID of caller device. */ 4883 uint32_t nb_ranges_alloc; /**< Size of the ranges array. */ 4884 uint32_t nb_ranges; /**< Number of initialized ranges. */ 4885 struct rte_eth_representor_range ranges[];/**< Representor ID range. */ 4886 }; 4887 4888 /** 4889 * Retrieve the representor info of the device. 4890 * 4891 * Get device representor info to be able to calculate a unique 4892 * representor ID. @see rte_eth_representor_id_get helper. 4893 * 4894 * @param port_id 4895 * The port identifier of the device. 4896 * @param info 4897 * A pointer to a representor info structure. 4898 * NULL to return number of range entries and allocate memory 4899 * for next call to store detail. 4900 * The number of ranges that were written into this structure 4901 * will be placed into its nb_ranges field. This number cannot be 4902 * larger than the nb_ranges_alloc that by the user before calling 4903 * this function. It can be smaller than the value returned by the 4904 * function, however. 4905 * @return 4906 * - (-ENOTSUP) if operation is not supported. 4907 * - (-ENODEV) if *port_id* invalid. 4908 * - (-EIO) if device is removed. 4909 * - (>=0) number of available representor range entries. 4910 */ 4911 __rte_experimental 4912 int rte_eth_representor_info_get(uint16_t port_id, 4913 struct rte_eth_representor_info *info); 4914 4915 /** The NIC is able to deliver flag (if set) with packets to the PMD. */ 4916 #define RTE_ETH_RX_METADATA_USER_FLAG RTE_BIT64(0) 4917 4918 /** The NIC is able to deliver mark ID with packets to the PMD. */ 4919 #define RTE_ETH_RX_METADATA_USER_MARK RTE_BIT64(1) 4920 4921 /** The NIC is able to deliver tunnel ID with packets to the PMD. */ 4922 #define RTE_ETH_RX_METADATA_TUNNEL_ID RTE_BIT64(2) 4923 4924 /** 4925 * @warning 4926 * @b EXPERIMENTAL: this API may change without prior notice 4927 * 4928 * Negotiate the NIC's ability to deliver specific kinds of metadata to the PMD. 4929 * 4930 * Invoke this API before the first rte_eth_dev_configure() invocation 4931 * to let the PMD make preparations that are inconvenient to do later. 4932 * 4933 * The negotiation process is as follows: 4934 * 4935 * - the application requests features intending to use at least some of them; 4936 * - the PMD responds with the guaranteed subset of the requested feature set; 4937 * - the application can retry negotiation with another set of features; 4938 * - the application can pass zero to clear the negotiation result; 4939 * - the last negotiated result takes effect upon 4940 * the ethdev configure and start. 4941 * 4942 * @note 4943 * The PMD is supposed to first consider enabling the requested feature set 4944 * in its entirety. Only if it fails to do so, does it have the right to 4945 * respond with a smaller set of the originally requested features. 4946 * 4947 * @note 4948 * Return code (-ENOTSUP) does not necessarily mean that the requested 4949 * features are unsupported. In this case, the application should just 4950 * assume that these features can be used without prior negotiations. 4951 * 4952 * @param port_id 4953 * Port (ethdev) identifier 4954 * 4955 * @param[inout] features 4956 * Feature selection buffer 4957 * 4958 * @return 4959 * - (-EBUSY) if the port can't handle this in its current state; 4960 * - (-ENOTSUP) if the method itself is not supported by the PMD; 4961 * - (-ENODEV) if *port_id* is invalid; 4962 * - (-EINVAL) if *features* is NULL; 4963 * - (-EIO) if the device is removed; 4964 * - (0) on success 4965 */ 4966 __rte_experimental 4967 int rte_eth_rx_metadata_negotiate(uint16_t port_id, uint64_t *features); 4968 4969 #include <rte_ethdev_core.h> 4970 4971 /** 4972 * @internal 4973 * Helper routine for rte_eth_rx_burst(). 4974 * Should be called at exit from PMD's rte_eth_rx_bulk implementation. 4975 * Does necessary post-processing - invokes Rx callbacks if any, etc. 4976 * 4977 * @param port_id 4978 * The port identifier of the Ethernet device. 4979 * @param queue_id 4980 * The index of the receive queue from which to retrieve input packets. 4981 * @param rx_pkts 4982 * The address of an array of pointers to *rte_mbuf* structures that 4983 * have been retrieved from the device. 4984 * @param nb_rx 4985 * The number of packets that were retrieved from the device. 4986 * @param nb_pkts 4987 * The number of elements in @p rx_pkts array. 4988 * @param opaque 4989 * Opaque pointer of Rx queue callback related data. 4990 * 4991 * @return 4992 * The number of packets effectively supplied to the @p rx_pkts array. 4993 */ 4994 uint16_t rte_eth_call_rx_callbacks(uint16_t port_id, uint16_t queue_id, 4995 struct rte_mbuf **rx_pkts, uint16_t nb_rx, uint16_t nb_pkts, 4996 void *opaque); 4997 4998 /** 4999 * 5000 * Retrieve a burst of input packets from a receive queue of an Ethernet 5001 * device. The retrieved packets are stored in *rte_mbuf* structures whose 5002 * pointers are supplied in the *rx_pkts* array. 5003 * 5004 * The rte_eth_rx_burst() function loops, parsing the Rx ring of the 5005 * receive queue, up to *nb_pkts* packets, and for each completed Rx 5006 * descriptor in the ring, it performs the following operations: 5007 * 5008 * - Initialize the *rte_mbuf* data structure associated with the 5009 * Rx descriptor according to the information provided by the NIC into 5010 * that Rx descriptor. 5011 * 5012 * - Store the *rte_mbuf* data structure into the next entry of the 5013 * *rx_pkts* array. 5014 * 5015 * - Replenish the Rx descriptor with a new *rte_mbuf* buffer 5016 * allocated from the memory pool associated with the receive queue at 5017 * initialization time. 5018 * 5019 * When retrieving an input packet that was scattered by the controller 5020 * into multiple receive descriptors, the rte_eth_rx_burst() function 5021 * appends the associated *rte_mbuf* buffers to the first buffer of the 5022 * packet. 5023 * 5024 * The rte_eth_rx_burst() function returns the number of packets 5025 * actually retrieved, which is the number of *rte_mbuf* data structures 5026 * effectively supplied into the *rx_pkts* array. 5027 * A return value equal to *nb_pkts* indicates that the Rx queue contained 5028 * at least *rx_pkts* packets, and this is likely to signify that other 5029 * received packets remain in the input queue. Applications implementing 5030 * a "retrieve as much received packets as possible" policy can check this 5031 * specific case and keep invoking the rte_eth_rx_burst() function until 5032 * a value less than *nb_pkts* is returned. 5033 * 5034 * This receive method has the following advantages: 5035 * 5036 * - It allows a run-to-completion network stack engine to retrieve and 5037 * to immediately process received packets in a fast burst-oriented 5038 * approach, avoiding the overhead of unnecessary intermediate packet 5039 * queue/dequeue operations. 5040 * 5041 * - Conversely, it also allows an asynchronous-oriented processing 5042 * method to retrieve bursts of received packets and to immediately 5043 * queue them for further parallel processing by another logical core, 5044 * for instance. However, instead of having received packets being 5045 * individually queued by the driver, this approach allows the caller 5046 * of the rte_eth_rx_burst() function to queue a burst of retrieved 5047 * packets at a time and therefore dramatically reduce the cost of 5048 * enqueue/dequeue operations per packet. 5049 * 5050 * - It allows the rte_eth_rx_burst() function of the driver to take 5051 * advantage of burst-oriented hardware features (CPU cache, 5052 * prefetch instructions, and so on) to minimize the number of CPU 5053 * cycles per packet. 5054 * 5055 * To summarize, the proposed receive API enables many 5056 * burst-oriented optimizations in both synchronous and asynchronous 5057 * packet processing environments with no overhead in both cases. 5058 * 5059 * @note 5060 * Some drivers using vector instructions require that *nb_pkts* is 5061 * divisible by 4 or 8, depending on the driver implementation. 5062 * 5063 * The rte_eth_rx_burst() function does not provide any error 5064 * notification to avoid the corresponding overhead. As a hint, the 5065 * upper-level application might check the status of the device link once 5066 * being systematically returned a 0 value for a given number of tries. 5067 * 5068 * @param port_id 5069 * The port identifier of the Ethernet device. 5070 * @param queue_id 5071 * The index of the receive queue from which to retrieve input packets. 5072 * The value must be in the range [0, nb_rx_queue - 1] previously supplied 5073 * to rte_eth_dev_configure(). 5074 * @param rx_pkts 5075 * The address of an array of pointers to *rte_mbuf* structures that 5076 * must be large enough to store *nb_pkts* pointers in it. 5077 * @param nb_pkts 5078 * The maximum number of packets to retrieve. 5079 * The value must be divisible by 8 in order to work with any driver. 5080 * @return 5081 * The number of packets actually retrieved, which is the number 5082 * of pointers to *rte_mbuf* structures effectively supplied to the 5083 * *rx_pkts* array. 5084 */ 5085 static inline uint16_t 5086 rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id, 5087 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts) 5088 { 5089 uint16_t nb_rx; 5090 struct rte_eth_fp_ops *p; 5091 void *qd; 5092 5093 #ifdef RTE_ETHDEV_DEBUG_RX 5094 if (port_id >= RTE_MAX_ETHPORTS || 5095 queue_id >= RTE_MAX_QUEUES_PER_PORT) { 5096 RTE_ETHDEV_LOG(ERR, 5097 "Invalid port_id=%u or queue_id=%u\n", 5098 port_id, queue_id); 5099 return 0; 5100 } 5101 #endif 5102 5103 /* fetch pointer to queue data */ 5104 p = &rte_eth_fp_ops[port_id]; 5105 qd = p->rxq.data[queue_id]; 5106 5107 #ifdef RTE_ETHDEV_DEBUG_RX 5108 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0); 5109 5110 if (qd == NULL) { 5111 RTE_ETHDEV_LOG(ERR, "Invalid Rx queue_id=%u for port_id=%u\n", 5112 queue_id, port_id); 5113 return 0; 5114 } 5115 #endif 5116 5117 nb_rx = p->rx_pkt_burst(qd, rx_pkts, nb_pkts); 5118 5119 #ifdef RTE_ETHDEV_RXTX_CALLBACKS 5120 { 5121 void *cb; 5122 5123 /* __ATOMIC_RELEASE memory order was used when the 5124 * call back was inserted into the list. 5125 * Since there is a clear dependency between loading 5126 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is 5127 * not required. 5128 */ 5129 cb = __atomic_load_n((void **)&p->rxq.clbk[queue_id], 5130 __ATOMIC_RELAXED); 5131 if (unlikely(cb != NULL)) 5132 nb_rx = rte_eth_call_rx_callbacks(port_id, queue_id, 5133 rx_pkts, nb_rx, nb_pkts, cb); 5134 } 5135 #endif 5136 5137 rte_ethdev_trace_rx_burst(port_id, queue_id, (void **)rx_pkts, nb_rx); 5138 return nb_rx; 5139 } 5140 5141 /** 5142 * Get the number of used descriptors of a Rx queue 5143 * 5144 * @param port_id 5145 * The port identifier of the Ethernet device. 5146 * @param queue_id 5147 * The queue ID on the specific port. 5148 * @return 5149 * The number of used descriptors in the specific queue, or: 5150 * - (-ENODEV) if *port_id* is invalid. 5151 * (-EINVAL) if *queue_id* is invalid 5152 * (-ENOTSUP) if the device does not support this function 5153 */ 5154 static inline int 5155 rte_eth_rx_queue_count(uint16_t port_id, uint16_t queue_id) 5156 { 5157 struct rte_eth_fp_ops *p; 5158 void *qd; 5159 5160 if (port_id >= RTE_MAX_ETHPORTS || 5161 queue_id >= RTE_MAX_QUEUES_PER_PORT) { 5162 RTE_ETHDEV_LOG(ERR, 5163 "Invalid port_id=%u or queue_id=%u\n", 5164 port_id, queue_id); 5165 return -EINVAL; 5166 } 5167 5168 /* fetch pointer to queue data */ 5169 p = &rte_eth_fp_ops[port_id]; 5170 qd = p->rxq.data[queue_id]; 5171 5172 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); 5173 RTE_FUNC_PTR_OR_ERR_RET(*p->rx_queue_count, -ENOTSUP); 5174 if (qd == NULL) 5175 return -EINVAL; 5176 5177 return (int)(*p->rx_queue_count)(qd); 5178 } 5179 5180 /**@{@name Rx hardware descriptor states 5181 * @see rte_eth_rx_descriptor_status 5182 */ 5183 #define RTE_ETH_RX_DESC_AVAIL 0 /**< Desc available for hw. */ 5184 #define RTE_ETH_RX_DESC_DONE 1 /**< Desc done, filled by hw. */ 5185 #define RTE_ETH_RX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */ 5186 /**@}*/ 5187 5188 /** 5189 * Check the status of a Rx descriptor in the queue 5190 * 5191 * It should be called in a similar context than the Rx function: 5192 * - on a dataplane core 5193 * - not concurrently on the same queue 5194 * 5195 * Since it's a dataplane function, no check is performed on port_id and 5196 * queue_id. The caller must therefore ensure that the port is enabled 5197 * and the queue is configured and running. 5198 * 5199 * Note: accessing to a random descriptor in the ring may trigger cache 5200 * misses and have a performance impact. 5201 * 5202 * @param port_id 5203 * A valid port identifier of the Ethernet device which. 5204 * @param queue_id 5205 * A valid Rx queue identifier on this port. 5206 * @param offset 5207 * The offset of the descriptor starting from tail (0 is the next 5208 * packet to be received by the driver). 5209 * 5210 * @return 5211 * - (RTE_ETH_RX_DESC_AVAIL): Descriptor is available for the hardware to 5212 * receive a packet. 5213 * - (RTE_ETH_RX_DESC_DONE): Descriptor is done, it is filled by hw, but 5214 * not yet processed by the driver (i.e. in the receive queue). 5215 * - (RTE_ETH_RX_DESC_UNAVAIL): Descriptor is unavailable, either hold by 5216 * the driver and not yet returned to hw, or reserved by the hw. 5217 * - (-EINVAL) bad descriptor offset. 5218 * - (-ENOTSUP) if the device does not support this function. 5219 * - (-ENODEV) bad port or queue (only if compiled with debug). 5220 */ 5221 static inline int 5222 rte_eth_rx_descriptor_status(uint16_t port_id, uint16_t queue_id, 5223 uint16_t offset) 5224 { 5225 struct rte_eth_fp_ops *p; 5226 void *qd; 5227 5228 #ifdef RTE_ETHDEV_DEBUG_RX 5229 if (port_id >= RTE_MAX_ETHPORTS || 5230 queue_id >= RTE_MAX_QUEUES_PER_PORT) { 5231 RTE_ETHDEV_LOG(ERR, 5232 "Invalid port_id=%u or queue_id=%u\n", 5233 port_id, queue_id); 5234 return -EINVAL; 5235 } 5236 #endif 5237 5238 /* fetch pointer to queue data */ 5239 p = &rte_eth_fp_ops[port_id]; 5240 qd = p->rxq.data[queue_id]; 5241 5242 #ifdef RTE_ETHDEV_DEBUG_RX 5243 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); 5244 if (qd == NULL) 5245 return -ENODEV; 5246 #endif 5247 RTE_FUNC_PTR_OR_ERR_RET(*p->rx_descriptor_status, -ENOTSUP); 5248 return (*p->rx_descriptor_status)(qd, offset); 5249 } 5250 5251 /**@{@name Tx hardware descriptor states 5252 * @see rte_eth_tx_descriptor_status 5253 */ 5254 #define RTE_ETH_TX_DESC_FULL 0 /**< Desc filled for hw, waiting xmit. */ 5255 #define RTE_ETH_TX_DESC_DONE 1 /**< Desc done, packet is transmitted. */ 5256 #define RTE_ETH_TX_DESC_UNAVAIL 2 /**< Desc used by driver or hw. */ 5257 /**@}*/ 5258 5259 /** 5260 * Check the status of a Tx descriptor in the queue. 5261 * 5262 * It should be called in a similar context than the Tx function: 5263 * - on a dataplane core 5264 * - not concurrently on the same queue 5265 * 5266 * Since it's a dataplane function, no check is performed on port_id and 5267 * queue_id. The caller must therefore ensure that the port is enabled 5268 * and the queue is configured and running. 5269 * 5270 * Note: accessing to a random descriptor in the ring may trigger cache 5271 * misses and have a performance impact. 5272 * 5273 * @param port_id 5274 * A valid port identifier of the Ethernet device which. 5275 * @param queue_id 5276 * A valid Tx queue identifier on this port. 5277 * @param offset 5278 * The offset of the descriptor starting from tail (0 is the place where 5279 * the next packet will be send). 5280 * 5281 * @return 5282 * - (RTE_ETH_TX_DESC_FULL) Descriptor is being processed by the hw, i.e. 5283 * in the transmit queue. 5284 * - (RTE_ETH_TX_DESC_DONE) Hardware is done with this descriptor, it can 5285 * be reused by the driver. 5286 * - (RTE_ETH_TX_DESC_UNAVAIL): Descriptor is unavailable, reserved by the 5287 * driver or the hardware. 5288 * - (-EINVAL) bad descriptor offset. 5289 * - (-ENOTSUP) if the device does not support this function. 5290 * - (-ENODEV) bad port or queue (only if compiled with debug). 5291 */ 5292 static inline int rte_eth_tx_descriptor_status(uint16_t port_id, 5293 uint16_t queue_id, uint16_t offset) 5294 { 5295 struct rte_eth_fp_ops *p; 5296 void *qd; 5297 5298 #ifdef RTE_ETHDEV_DEBUG_TX 5299 if (port_id >= RTE_MAX_ETHPORTS || 5300 queue_id >= RTE_MAX_QUEUES_PER_PORT) { 5301 RTE_ETHDEV_LOG(ERR, 5302 "Invalid port_id=%u or queue_id=%u\n", 5303 port_id, queue_id); 5304 return -EINVAL; 5305 } 5306 #endif 5307 5308 /* fetch pointer to queue data */ 5309 p = &rte_eth_fp_ops[port_id]; 5310 qd = p->txq.data[queue_id]; 5311 5312 #ifdef RTE_ETHDEV_DEBUG_TX 5313 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); 5314 if (qd == NULL) 5315 return -ENODEV; 5316 #endif 5317 RTE_FUNC_PTR_OR_ERR_RET(*p->tx_descriptor_status, -ENOTSUP); 5318 return (*p->tx_descriptor_status)(qd, offset); 5319 } 5320 5321 /** 5322 * @internal 5323 * Helper routine for rte_eth_tx_burst(). 5324 * Should be called before entry PMD's rte_eth_tx_bulk implementation. 5325 * Does necessary pre-processing - invokes Tx callbacks if any, etc. 5326 * 5327 * @param port_id 5328 * The port identifier of the Ethernet device. 5329 * @param queue_id 5330 * The index of the transmit queue through which output packets must be 5331 * sent. 5332 * @param tx_pkts 5333 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures 5334 * which contain the output packets. 5335 * @param nb_pkts 5336 * The maximum number of packets to transmit. 5337 * @return 5338 * The number of output packets to transmit. 5339 */ 5340 uint16_t rte_eth_call_tx_callbacks(uint16_t port_id, uint16_t queue_id, 5341 struct rte_mbuf **tx_pkts, uint16_t nb_pkts, void *opaque); 5342 5343 /** 5344 * Send a burst of output packets on a transmit queue of an Ethernet device. 5345 * 5346 * The rte_eth_tx_burst() function is invoked to transmit output packets 5347 * on the output queue *queue_id* of the Ethernet device designated by its 5348 * *port_id*. 5349 * The *nb_pkts* parameter is the number of packets to send which are 5350 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them 5351 * allocated from a pool created with rte_pktmbuf_pool_create(). 5352 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets, 5353 * up to the number of transmit descriptors available in the Tx ring of the 5354 * transmit queue. 5355 * For each packet to send, the rte_eth_tx_burst() function performs 5356 * the following operations: 5357 * 5358 * - Pick up the next available descriptor in the transmit ring. 5359 * 5360 * - Free the network buffer previously sent with that descriptor, if any. 5361 * 5362 * - Initialize the transmit descriptor with the information provided 5363 * in the *rte_mbuf data structure. 5364 * 5365 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers, 5366 * the rte_eth_tx_burst() function uses several transmit descriptors 5367 * of the ring. 5368 * 5369 * The rte_eth_tx_burst() function returns the number of packets it 5370 * actually sent. A return value equal to *nb_pkts* means that all packets 5371 * have been sent, and this is likely to signify that other output packets 5372 * could be immediately transmitted again. Applications that implement a 5373 * "send as many packets to transmit as possible" policy can check this 5374 * specific case and keep invoking the rte_eth_tx_burst() function until 5375 * a value less than *nb_pkts* is returned. 5376 * 5377 * It is the responsibility of the rte_eth_tx_burst() function to 5378 * transparently free the memory buffers of packets previously sent. 5379 * This feature is driven by the *tx_free_thresh* value supplied to the 5380 * rte_eth_dev_configure() function at device configuration time. 5381 * When the number of free Tx descriptors drops below this threshold, the 5382 * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers 5383 * of those packets whose transmission was effectively completed. 5384 * 5385 * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can 5386 * invoke this function concurrently on the same Tx queue without SW lock. 5387 * @see rte_eth_dev_info_get, struct rte_eth_txconf::offloads 5388 * 5389 * @see rte_eth_tx_prepare to perform some prior checks or adjustments 5390 * for offloads. 5391 * 5392 * @param port_id 5393 * The port identifier of the Ethernet device. 5394 * @param queue_id 5395 * The index of the transmit queue through which output packets must be 5396 * sent. 5397 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5398 * to rte_eth_dev_configure(). 5399 * @param tx_pkts 5400 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures 5401 * which contain the output packets. 5402 * @param nb_pkts 5403 * The maximum number of packets to transmit. 5404 * @return 5405 * The number of output packets actually stored in transmit descriptors of 5406 * the transmit ring. The return value can be less than the value of the 5407 * *tx_pkts* parameter when the transmit ring is full or has been filled up. 5408 */ 5409 static inline uint16_t 5410 rte_eth_tx_burst(uint16_t port_id, uint16_t queue_id, 5411 struct rte_mbuf **tx_pkts, uint16_t nb_pkts) 5412 { 5413 struct rte_eth_fp_ops *p; 5414 void *qd; 5415 5416 #ifdef RTE_ETHDEV_DEBUG_TX 5417 if (port_id >= RTE_MAX_ETHPORTS || 5418 queue_id >= RTE_MAX_QUEUES_PER_PORT) { 5419 RTE_ETHDEV_LOG(ERR, 5420 "Invalid port_id=%u or queue_id=%u\n", 5421 port_id, queue_id); 5422 return 0; 5423 } 5424 #endif 5425 5426 /* fetch pointer to queue data */ 5427 p = &rte_eth_fp_ops[port_id]; 5428 qd = p->txq.data[queue_id]; 5429 5430 #ifdef RTE_ETHDEV_DEBUG_TX 5431 RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, 0); 5432 5433 if (qd == NULL) { 5434 RTE_ETHDEV_LOG(ERR, "Invalid Tx queue_id=%u for port_id=%u\n", 5435 queue_id, port_id); 5436 return 0; 5437 } 5438 #endif 5439 5440 #ifdef RTE_ETHDEV_RXTX_CALLBACKS 5441 { 5442 void *cb; 5443 5444 /* __ATOMIC_RELEASE memory order was used when the 5445 * call back was inserted into the list. 5446 * Since there is a clear dependency between loading 5447 * cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory order is 5448 * not required. 5449 */ 5450 cb = __atomic_load_n((void **)&p->txq.clbk[queue_id], 5451 __ATOMIC_RELAXED); 5452 if (unlikely(cb != NULL)) 5453 nb_pkts = rte_eth_call_tx_callbacks(port_id, queue_id, 5454 tx_pkts, nb_pkts, cb); 5455 } 5456 #endif 5457 5458 nb_pkts = p->tx_pkt_burst(qd, tx_pkts, nb_pkts); 5459 5460 rte_ethdev_trace_tx_burst(port_id, queue_id, (void **)tx_pkts, nb_pkts); 5461 return nb_pkts; 5462 } 5463 5464 /** 5465 * Process a burst of output packets on a transmit queue of an Ethernet device. 5466 * 5467 * The rte_eth_tx_prepare() function is invoked to prepare output packets to be 5468 * transmitted on the output queue *queue_id* of the Ethernet device designated 5469 * by its *port_id*. 5470 * The *nb_pkts* parameter is the number of packets to be prepared which are 5471 * supplied in the *tx_pkts* array of *rte_mbuf* structures, each of them 5472 * allocated from a pool created with rte_pktmbuf_pool_create(). 5473 * For each packet to send, the rte_eth_tx_prepare() function performs 5474 * the following operations: 5475 * 5476 * - Check if packet meets devices requirements for Tx offloads. 5477 * 5478 * - Check limitations about number of segments. 5479 * 5480 * - Check additional requirements when debug is enabled. 5481 * 5482 * - Update and/or reset required checksums when Tx offload is set for packet. 5483 * 5484 * Since this function can modify packet data, provided mbufs must be safely 5485 * writable (e.g. modified data cannot be in shared segment). 5486 * 5487 * The rte_eth_tx_prepare() function returns the number of packets ready to be 5488 * sent. A return value equal to *nb_pkts* means that all packets are valid and 5489 * ready to be sent, otherwise stops processing on the first invalid packet and 5490 * leaves the rest packets untouched. 5491 * 5492 * When this functionality is not implemented in the driver, all packets are 5493 * are returned untouched. 5494 * 5495 * @param port_id 5496 * The port identifier of the Ethernet device. 5497 * The value must be a valid port ID. 5498 * @param queue_id 5499 * The index of the transmit queue through which output packets must be 5500 * sent. 5501 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5502 * to rte_eth_dev_configure(). 5503 * @param tx_pkts 5504 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures 5505 * which contain the output packets. 5506 * @param nb_pkts 5507 * The maximum number of packets to process. 5508 * @return 5509 * The number of packets correct and ready to be sent. The return value can be 5510 * less than the value of the *tx_pkts* parameter when some packet doesn't 5511 * meet devices requirements with rte_errno set appropriately: 5512 * - EINVAL: offload flags are not correctly set 5513 * - ENOTSUP: the offload feature is not supported by the hardware 5514 * - ENODEV: if *port_id* is invalid (with debug enabled only) 5515 * 5516 */ 5517 5518 #ifndef RTE_ETHDEV_TX_PREPARE_NOOP 5519 5520 static inline uint16_t 5521 rte_eth_tx_prepare(uint16_t port_id, uint16_t queue_id, 5522 struct rte_mbuf **tx_pkts, uint16_t nb_pkts) 5523 { 5524 struct rte_eth_fp_ops *p; 5525 void *qd; 5526 5527 #ifdef RTE_ETHDEV_DEBUG_TX 5528 if (port_id >= RTE_MAX_ETHPORTS || 5529 queue_id >= RTE_MAX_QUEUES_PER_PORT) { 5530 RTE_ETHDEV_LOG(ERR, 5531 "Invalid port_id=%u or queue_id=%u\n", 5532 port_id, queue_id); 5533 rte_errno = ENODEV; 5534 return 0; 5535 } 5536 #endif 5537 5538 /* fetch pointer to queue data */ 5539 p = &rte_eth_fp_ops[port_id]; 5540 qd = p->txq.data[queue_id]; 5541 5542 #ifdef RTE_ETHDEV_DEBUG_TX 5543 if (!rte_eth_dev_is_valid_port(port_id)) { 5544 RTE_ETHDEV_LOG(ERR, "Invalid Tx port_id=%u\n", port_id); 5545 rte_errno = ENODEV; 5546 return 0; 5547 } 5548 if (qd == NULL) { 5549 RTE_ETHDEV_LOG(ERR, "Invalid Tx queue_id=%u for port_id=%u\n", 5550 queue_id, port_id); 5551 rte_errno = EINVAL; 5552 return 0; 5553 } 5554 #endif 5555 5556 if (!p->tx_pkt_prepare) 5557 return nb_pkts; 5558 5559 return p->tx_pkt_prepare(qd, tx_pkts, nb_pkts); 5560 } 5561 5562 #else 5563 5564 /* 5565 * Native NOOP operation for compilation targets which doesn't require any 5566 * preparations steps, and functional NOOP may introduce unnecessary performance 5567 * drop. 5568 * 5569 * Generally this is not a good idea to turn it on globally and didn't should 5570 * be used if behavior of tx_preparation can change. 5571 */ 5572 5573 static inline uint16_t 5574 rte_eth_tx_prepare(__rte_unused uint16_t port_id, 5575 __rte_unused uint16_t queue_id, 5576 __rte_unused struct rte_mbuf **tx_pkts, uint16_t nb_pkts) 5577 { 5578 return nb_pkts; 5579 } 5580 5581 #endif 5582 5583 /** 5584 * Send any packets queued up for transmission on a port and HW queue 5585 * 5586 * This causes an explicit flush of packets previously buffered via the 5587 * rte_eth_tx_buffer() function. It returns the number of packets successfully 5588 * sent to the NIC, and calls the error callback for any unsent packets. Unless 5589 * explicitly set up otherwise, the default callback simply frees the unsent 5590 * packets back to the owning mempool. 5591 * 5592 * @param port_id 5593 * The port identifier of the Ethernet device. 5594 * @param queue_id 5595 * The index of the transmit queue through which output packets must be 5596 * sent. 5597 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5598 * to rte_eth_dev_configure(). 5599 * @param buffer 5600 * Buffer of packets to be transmit. 5601 * @return 5602 * The number of packets successfully sent to the Ethernet device. The error 5603 * callback is called for any packets which could not be sent. 5604 */ 5605 static inline uint16_t 5606 rte_eth_tx_buffer_flush(uint16_t port_id, uint16_t queue_id, 5607 struct rte_eth_dev_tx_buffer *buffer) 5608 { 5609 uint16_t sent; 5610 uint16_t to_send = buffer->length; 5611 5612 if (to_send == 0) 5613 return 0; 5614 5615 sent = rte_eth_tx_burst(port_id, queue_id, buffer->pkts, to_send); 5616 5617 buffer->length = 0; 5618 5619 /* All packets sent, or to be dealt with by callback below */ 5620 if (unlikely(sent != to_send)) 5621 buffer->error_callback(&buffer->pkts[sent], 5622 (uint16_t)(to_send - sent), 5623 buffer->error_userdata); 5624 5625 return sent; 5626 } 5627 5628 /** 5629 * Buffer a single packet for future transmission on a port and queue 5630 * 5631 * This function takes a single mbuf/packet and buffers it for later 5632 * transmission on the particular port and queue specified. Once the buffer is 5633 * full of packets, an attempt will be made to transmit all the buffered 5634 * packets. In case of error, where not all packets can be transmitted, a 5635 * callback is called with the unsent packets as a parameter. If no callback 5636 * is explicitly set up, the unsent packets are just freed back to the owning 5637 * mempool. The function returns the number of packets actually sent i.e. 5638 * 0 if no buffer flush occurred, otherwise the number of packets successfully 5639 * flushed 5640 * 5641 * @param port_id 5642 * The port identifier of the Ethernet device. 5643 * @param queue_id 5644 * The index of the transmit queue through which output packets must be 5645 * sent. 5646 * The value must be in the range [0, nb_tx_queue - 1] previously supplied 5647 * to rte_eth_dev_configure(). 5648 * @param buffer 5649 * Buffer used to collect packets to be sent. 5650 * @param tx_pkt 5651 * Pointer to the packet mbuf to be sent. 5652 * @return 5653 * 0 = packet has been buffered for later transmission 5654 * N > 0 = packet has been buffered, and the buffer was subsequently flushed, 5655 * causing N packets to be sent, and the error callback to be called for 5656 * the rest. 5657 */ 5658 static __rte_always_inline uint16_t 5659 rte_eth_tx_buffer(uint16_t port_id, uint16_t queue_id, 5660 struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt) 5661 { 5662 buffer->pkts[buffer->length++] = tx_pkt; 5663 if (buffer->length < buffer->size) 5664 return 0; 5665 5666 return rte_eth_tx_buffer_flush(port_id, queue_id, buffer); 5667 } 5668 5669 #ifdef __cplusplus 5670 } 5671 #endif 5672 5673 #endif /* _RTE_ETHDEV_H_ */ 5674