4 * Copyright(c) 2010-2013 Intel Corporation. All rights reserved.
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
11 * * Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * * Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in
15 * the documentation and/or other materials provided with the
17 * * Neither the name of Intel Corporation nor the names of its
18 * contributors may be used to endorse or promote products derived
19 * from this software without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
22 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
23 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
24 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
25 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
26 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
27 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
28 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
29 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
30 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
31 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
35 #ifndef _RTE_ETHDEV_H_
36 #define _RTE_ETHDEV_H_
41 * RTE Ethernet Device API
43 * The Ethernet Device API is composed of two parts:
45 * - The application-oriented Ethernet API that includes functions to setup
46 * an Ethernet device (configure it, setup its RX and TX queues and start it),
47 * to get its MAC address, the speed and the status of its physical link,
48 * to receive and to transmit packets, and so on.
50 * - The driver-oriented Ethernet API that exports a function allowing
51 * an Ethernet Poll Mode Driver (PMD) to simultaneously register itself as
52 * an Ethernet device driver and as a PCI driver for a set of matching PCI
53 * [Ethernet] devices classes.
55 * By default, all the functions of the Ethernet Device API exported by a PMD
56 * are lock-free functions which assume to not be invoked in parallel on
57 * different logical cores to work on the same target object. For instance,
58 * the receive function of a PMD cannot be invoked in parallel on two logical
59 * cores to poll the same RX queue [of the same port]. Of course, this function
60 * can be invoked in parallel by different logical cores on different RX queues.
61 * It is the responsibility of the upper level application to enforce this rule.
63 * If needed, parallel accesses by multiple logical cores to shared queues
64 * shall be explicitly protected by dedicated inline lock-aware functions
65 * built on top of their corresponding lock-free functions of the PMD API.
67 * In all functions of the Ethernet API, the Ethernet device is
68 * designated by an integer >= 0 named the device port identifier.
70 * At the Ethernet driver level, Ethernet devices are represented by a generic
71 * data structure of type *rte_eth_dev*.
73 * Ethernet devices are dynamically registered during the PCI probing phase
74 * performed at EAL initialization time.
75 * When an Ethernet device is being probed, an *rte_eth_dev* structure and
76 * a new port identifier are allocated for that device. Then, the eth_dev_init()
77 * function supplied by the Ethernet driver matching the probed PCI
78 * device is invoked to properly initialize the device.
80 * The role of the device init function consists of resetting the hardware,
81 * checking access to Non-volatile Memory (NVM), reading the MAC address
84 * If the device init operation is successful, the correspondence between
85 * the port identifier assigned to the new device and its associated
86 * *rte_eth_dev* structure is effectively registered.
87 * Otherwise, both the *rte_eth_dev* structure and the port identifier are
90 * The functions exported by the application Ethernet API to setup a device
91 * designated by its port identifier must be invoked in the following order:
92 * - rte_eth_dev_configure()
93 * - rte_eth_tx_queue_setup()
94 * - rte_eth_rx_queue_setup()
95 * - rte_eth_dev_start()
97 * Then, the network application can invoke, in any order, the functions
98 * exported by the Ethernet API to get the MAC address of a given device, to
99 * get the speed and the status of a device physical link, to receive/transmit
100 * [burst of] packets, and so on.
102 * If the application wants to change the configuration (i.e. call
103 * rte_eth_dev_configure(), rte_eth_tx_queue_setup(), or
104 * rte_eth_rx_queue_setup()), it must call rte_eth_dev_stop() first to stop the
105 * device and then do the reconfiguration before calling rte_eth_dev_start()
106 * again. The tramsit and receive functions should not be invoked when the
109 * Please note that some configuration is not stored between calls to
110 * rte_eth_dev_stop()/rte_eth_dev_start(). The following configuration will
113 * - flow control settings
114 * - receive mode configuration (promiscuous mode, hardware checksum mode,
115 * RSS/VMDQ settings etc.)
116 * - VLAN filtering configuration
117 * - MAC addresses supplied to MAC address array
118 * - flow director filtering mode (but not filtering rules)
119 * - NIC queue statistics mappings
121 * Any other configuration will not be stored and will need to be re-entered
122 * after a call to rte_eth_dev_start().
124 * Finally, a network application can close an Ethernet device by invoking the
125 * rte_eth_dev_close() function.
127 * Each function of the application Ethernet API invokes a specific function
128 * of the PMD that controls the target device designated by its port
130 * For this purpose, all device-specific functions of an Ethernet driver are
131 * supplied through a set of pointers contained in a generic structure of type
133 * The address of the *eth_dev_ops* structure is stored in the *rte_eth_dev*
134 * structure by the device init function of the Ethernet driver, which is
135 * invoked during the PCI probing phase, as explained earlier.
137 * In other words, each function of the Ethernet API simply retrieves the
138 * *rte_eth_dev* structure associated with the device port identifier and
139 * performs an indirect invocation of the corresponding driver function
140 * supplied in the *eth_dev_ops* structure of the *rte_eth_dev* structure.
142 * For performance reasons, the address of the burst-oriented RX and TX
143 * functions of the Ethernet driver are not contained in the *eth_dev_ops*
144 * structure. Instead, they are directly stored at the beginning of the
145 * *rte_eth_dev* structure to avoid an extra indirect memory access during
148 * RTE ethernet device drivers do not use interrupts for transmitting or
149 * receiving. Instead, Ethernet drivers export Poll-Mode receive and transmit
150 * functions to applications.
151 * Both receive and transmit functions are packet-burst oriented to minimize
152 * their cost per packet through the following optimizations:
154 * - Sharing among multiple packets the incompressible cost of the
155 * invocation of receive/transmit functions.
157 * - Enabling receive/transmit functions to take advantage of burst-oriented
158 * hardware features (L1 cache, prefetch instructions, NIC head/tail
159 * registers) to minimize the number of CPU cycles per packet, for instance,
160 * by avoiding useless read memory accesses to ring descriptors, or by
161 * systematically using arrays of pointers that exactly fit L1 cache line
162 * boundaries and sizes.
164 * The burst-oriented receive function does not provide any error notification,
165 * to avoid the corresponding overhead. As a hint, the upper-level application
166 * might check the status of the device link once being systematically returned
167 * a 0 value by the receive function of the driver for a given number of tries.
177 #include <rte_interrupts.h>
179 #include <rte_mbuf.h>
180 #include "rte_ether.h"
183 * A structure used to retrieve statistics for an Ethernet port.
185 struct rte_eth_stats {
186 uint64_t ipackets; /**< Total number of successfully received packets. */
187 uint64_t opackets; /**< Total number of successfully transmitted packets.*/
188 uint64_t ibytes; /**< Total number of successfully received bytes. */
189 uint64_t obytes; /**< Total number of successfully transmitted bytes. */
190 uint64_t ierrors; /**< Total number of erroneous received packets. */
191 uint64_t oerrors; /**< Total number of failed transmitted packets. */
192 uint64_t imcasts; /**< Total number of multicast received packets. */
193 uint64_t rx_nombuf; /**< Total number of RX mbuf allocation failures. */
194 uint64_t fdirmatch; /**< Total number of RX packets matching a filter. */
195 uint64_t fdirmiss; /**< Total number of RX packets not matching any filter. */
196 uint64_t q_ipackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
197 /**< Total number of queue RX packets. */
198 uint64_t q_opackets[RTE_ETHDEV_QUEUE_STAT_CNTRS];
199 /**< Total number of queue TX packets. */
200 uint64_t q_ibytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
201 /**< Total number of successfully received queue bytes. */
202 uint64_t q_obytes[RTE_ETHDEV_QUEUE_STAT_CNTRS];
203 /**< Total number of successfully transmitted queue bytes. */
204 uint64_t q_errors[RTE_ETHDEV_QUEUE_STAT_CNTRS];
205 /**< Total number of queue packets received that are dropped. */
207 /**< Total number of good packets received from loopback,VF Only */
209 /**< Total number of good packets transmitted to loopback,VF Only */
211 /**< Total number of good bytes received from loopback,VF Only */
213 /**< Total number of good bytes transmitted to loopback,VF Only */
217 * A structure used to retrieve link-level information of an Ethernet port.
219 struct rte_eth_link {
220 uint16_t link_speed; /**< ETH_LINK_SPEED_[10, 100, 1000, 10000] */
221 uint16_t link_duplex; /**< ETH_LINK_[HALF_DUPLEX, FULL_DUPLEX] */
222 uint8_t link_status : 1; /**< 1 -> link up, 0 -> link down */
223 }__attribute__((aligned(8))); /**< aligned for atomic64 read/write */
225 #define ETH_LINK_SPEED_AUTONEG 0 /**< Auto-negotiate link speed. */
226 #define ETH_LINK_SPEED_10 10 /**< 10 megabits/second. */
227 #define ETH_LINK_SPEED_100 100 /**< 100 megabits/second. */
228 #define ETH_LINK_SPEED_1000 1000 /**< 1 gigabits/second. */
229 #define ETH_LINK_SPEED_10000 10000 /**< 10 gigabits/second. */
231 #define ETH_LINK_AUTONEG_DUPLEX 0 /**< Auto-negotiate duplex. */
232 #define ETH_LINK_HALF_DUPLEX 1 /**< Half-duplex connection. */
233 #define ETH_LINK_FULL_DUPLEX 2 /**< Full-duplex connection. */
236 * A structure used to configure the ring threshold registers of an RX/TX
237 * queue for an Ethernet port.
239 struct rte_eth_thresh {
240 uint8_t pthresh; /**< Ring prefetch threshold. */
241 uint8_t hthresh; /**< Ring host threshold. */
242 uint8_t wthresh; /**< Ring writeback threshold. */
246 * A set of values to identify what method is to be used to route
247 * packets to multiple queues.
249 enum rte_eth_rx_mq_mode {
250 ETH_MQ_RX_NONE = 0, /**< None of DCB,RSS or VMDQ mode */
252 ETH_MQ_RX_RSS, /**< For RX side, only RSS is on */
253 ETH_MQ_RX_DCB, /**< For RX side,only DCB is on. */
254 ETH_MQ_RX_DCB_RSS, /**< Both DCB and RSS enable */
256 ETH_MQ_RX_VMDQ_ONLY, /**< Only VMDQ, no RSS nor DCB */
257 ETH_MQ_RX_VMDQ_RSS, /**< RSS mode with VMDQ */
258 ETH_MQ_RX_VMDQ_DCB, /**< Use VMDQ+DCB to route traffic to queues */
259 ETH_MQ_RX_VMDQ_DCB_RSS, /**< Enable both VMDQ and DCB in VMDq */
263 * for rx mq mode backward compatible
265 #define ETH_RSS ETH_MQ_RX_RSS
266 #define VMDQ_DCB ETH_MQ_RX_VMDQ_DCB
267 #define ETH_DCB_RX ETH_MQ_RX_DCB
270 * A set of values to identify what method is to be used to transmit
271 * packets using multi-TCs.
273 enum rte_eth_tx_mq_mode {
274 ETH_MQ_TX_NONE = 0, /**< It is in neither DCB nor VT mode. */
275 ETH_MQ_TX_DCB, /**< For TX side,only DCB is on. */
276 ETH_MQ_TX_VMDQ_DCB, /**< For TX side,both DCB and VT is on. */
277 ETH_MQ_TX_VMDQ_ONLY, /**< Only VT on, no DCB */
281 * for tx mq mode backward compatible
283 #define ETH_DCB_NONE ETH_MQ_TX_NONE
284 #define ETH_VMDQ_DCB_TX ETH_MQ_TX_VMDQ_DCB
285 #define ETH_DCB_TX ETH_MQ_TX_DCB
288 * A structure used to configure the RX features of an Ethernet port.
290 struct rte_eth_rxmode {
291 /** The multi-queue packet distribution mode to be used, e.g. RSS. */
292 enum rte_eth_rx_mq_mode mq_mode;
293 uint32_t max_rx_pkt_len; /**< Only used if jumbo_frame enabled. */
294 uint16_t split_hdr_size; /**< hdr buf size (header_split enabled).*/
295 uint8_t header_split : 1, /**< Header Split enable. */
296 hw_ip_checksum : 1, /**< IP/UDP/TCP checksum offload enable. */
297 hw_vlan_filter : 1, /**< VLAN filter enable. */
298 hw_vlan_strip : 1, /**< VLAN strip enable. */
299 hw_vlan_extend : 1, /**< Extended VLAN enable. */
300 jumbo_frame : 1, /**< Jumbo Frame Receipt enable. */
301 hw_strip_crc : 1; /**< Enable CRC stripping by hardware. */
305 * A structure used to configure the Receive Side Scaling (RSS) feature
306 * of an Ethernet port.
307 * If not NULL, the *rss_key* pointer of the *rss_conf* structure points
308 * to an array of 40 bytes holding the RSS key to use for hashing specific
309 * header fields of received packets.
310 * Otherwise, a default random hash key is used by the device driver.
312 * The *rss_hf* field of the *rss_conf* structure indicates the different
313 * types of IPv4/IPv6 packets to which the RSS hashing must be applied.
314 * Supplying an *rss_hf* equal to zero disables the RSS feature.
316 struct rte_eth_rss_conf {
317 uint8_t *rss_key; /**< If not NULL, 40-byte hash key. */
318 uint16_t rss_hf; /**< Hash functions to apply - see below. */
321 #define ETH_RSS_IPV4 0x0001 /**< IPv4 packet. */
322 #define ETH_RSS_IPV4_TCP 0x0002 /**< IPv4/TCP packet. */
323 #define ETH_RSS_IPV6 0x0004 /**< IPv6 packet. */
324 #define ETH_RSS_IPV6_EX 0x0008 /**< IPv6 packet with extension headers.*/
325 #define ETH_RSS_IPV6_TCP 0x0010 /**< IPv6/TCP packet. */
326 #define ETH_RSS_IPV6_TCP_EX 0x0020 /**< IPv6/TCP with extension headers. */
327 /* Intel RSS extensions to UDP packets */
328 #define ETH_RSS_IPV4_UDP 0x0040 /**< IPv4/UDP packet. */
329 #define ETH_RSS_IPV6_UDP 0x0080 /**< IPv6/UDP packet. */
330 #define ETH_RSS_IPV6_UDP_EX 0x0100 /**< IPv6/UDP with extension headers. */
331 /* Definitions used for redirection table entry size */
332 #define ETH_RSS_RETA_NUM_ENTRIES 128
333 #define ETH_RSS_RETA_MAX_QUEUE 16
335 /* Definitions used for VMDQ and DCB functionality */
336 #define ETH_VMDQ_MAX_VLAN_FILTERS 64 /**< Maximum nb. of VMDQ vlan filters. */
337 #define ETH_DCB_NUM_USER_PRIORITIES 8 /**< Maximum nb. of DCB priorities. */
338 #define ETH_VMDQ_DCB_NUM_QUEUES 128 /**< Maximum nb. of VMDQ DCB queues. */
339 #define ETH_DCB_NUM_QUEUES 128 /**< Maximum nb. of DCB queues. */
341 /* DCB capability defines */
342 #define ETH_DCB_PG_SUPPORT 0x00000001 /**< Priority Group(ETS) support. */
343 #define ETH_DCB_PFC_SUPPORT 0x00000002 /**< Priority Flow Control support. */
345 /* Definitions used for VLAN Offload functionality */
346 #define ETH_VLAN_STRIP_OFFLOAD 0x0001 /**< VLAN Strip On/Off */
347 #define ETH_VLAN_FILTER_OFFLOAD 0x0002 /**< VLAN Filter On/Off */
348 #define ETH_VLAN_EXTEND_OFFLOAD 0x0004 /**< VLAN Extend On/Off */
350 /* Definitions used for mask VLAN setting */
351 #define ETH_VLAN_STRIP_MASK 0x0001 /**< VLAN Strip setting mask */
352 #define ETH_VLAN_FILTER_MASK 0x0002 /**< VLAN Filter setting mask*/
353 #define ETH_VLAN_EXTEND_MASK 0x0004 /**< VLAN Extend setting mask*/
356 * A structure used to configure Redirection Table of the Receive Side
357 * Scaling (RSS) feature of an Ethernet port.
359 struct rte_eth_rss_reta {
360 /** First 64 mask bits indicate which entry(s) need to updated/queried. */
362 /** Second 64 mask bits indicate which entry(s) need to updated/queried. */
364 uint8_t reta[ETH_RSS_RETA_NUM_ENTRIES]; /**< 128 RETA entries*/
368 * This enum indicates the possible number of traffic classes
369 * in DCB configratioins
371 enum rte_eth_nb_tcs {
372 ETH_4_TCS = 4, /**< 4 TCs with DCB. */
373 ETH_8_TCS = 8 /**< 8 TCs with DCB. */
377 * This enum indicates the possible number of queue pools
378 * in VMDQ+DCB configurations.
380 enum rte_eth_nb_pools {
381 ETH_8_POOLS = 8, /**< 8 VMDq pools. */
382 ETH_16_POOLS = 16, /**< 16 VMDq pools. */
383 ETH_32_POOLS = 32, /**< 32 VMDq pools. */
384 ETH_64_POOLS = 64 /**< 64 VMDq pools. */
387 /* This structure may be extended in future. */
388 struct rte_eth_dcb_rx_conf {
389 enum rte_eth_nb_tcs nb_tcs; /**< Possible DCB TCs, 4 or 8 TCs */
390 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
391 /**< Possible DCB queue,4 or 8. */
394 struct rte_eth_vmdq_dcb_tx_conf {
395 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools. */
396 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
397 /**< Possible DCB queue,4 or 8. */
400 struct rte_eth_dcb_tx_conf {
401 enum rte_eth_nb_tcs nb_tcs; /**< possible DCB TCs, 4 or 8 TCs. */
402 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
403 /**< Possible DCB queue,4 or 8. */
407 * A structure used to configure the VMDQ+DCB feature
408 * of an Ethernet port.
410 * Using this feature, packets are routed to a pool of queues, based
411 * on the vlan id in the vlan tag, and then to a specific queue within
412 * that pool, using the user priority vlan tag field.
414 * A default pool may be used, if desired, to route all traffic which
415 * does not match the vlan filter rules.
417 struct rte_eth_vmdq_dcb_conf {
418 enum rte_eth_nb_pools nb_queue_pools; /**< With DCB, 16 or 32 pools */
419 uint8_t enable_default_pool; /**< If non-zero, use a default pool */
420 uint8_t default_pool; /**< The default pool, if applicable */
421 uint8_t nb_pool_maps; /**< We can have up to 64 filters/mappings */
423 uint16_t vlan_id; /**< The vlan id of the received frame */
424 uint64_t pools; /**< Bitmask of pools for packet rx */
425 } pool_map[ETH_VMDQ_MAX_VLAN_FILTERS]; /**< VMDq vlan pool maps. */
426 uint8_t dcb_queue[ETH_DCB_NUM_USER_PRIORITIES];
427 /**< Selects a queue in a pool */
431 * A structure used to configure the TX features of an Ethernet port.
433 struct rte_eth_txmode {
434 enum rte_eth_tx_mq_mode mq_mode; /**< TX multi-queues mode. */
438 * A structure used to configure an RX ring of an Ethernet port.
440 struct rte_eth_rxconf {
441 struct rte_eth_thresh rx_thresh; /**< RX ring threshold registers. */
442 uint16_t rx_free_thresh; /**< Drives the freeing of RX descriptors. */
443 uint8_t rx_drop_en; /**< Drop packets if no descriptors are available. */
446 #define ETH_TXQ_FLAGS_NOMULTSEGS 0x0001 /**< nb_segs=1 for all mbufs */
447 #define ETH_TXQ_FLAGS_NOREFCOUNT 0x0002 /**< refcnt can be ignored */
448 #define ETH_TXQ_FLAGS_NOMULTMEMP 0x0004 /**< all bufs come from same mempool */
449 #define ETH_TXQ_FLAGS_NOVLANOFFL 0x0100 /**< disable VLAN offload */
450 #define ETH_TXQ_FLAGS_NOXSUMSCTP 0x0200 /**< disable SCTP checksum offload */
451 #define ETH_TXQ_FLAGS_NOXSUMUDP 0x0400 /**< disable UDP checksum offload */
452 #define ETH_TXQ_FLAGS_NOXSUMTCP 0x0800 /**< disable TCP checksum offload */
453 #define ETH_TXQ_FLAGS_NOOFFLOADS \
454 (ETH_TXQ_FLAGS_NOVLANOFFL | ETH_TXQ_FLAGS_NOXSUMSCTP | \
455 ETH_TXQ_FLAGS_NOXSUMUDP | ETH_TXQ_FLAGS_NOXSUMTCP)
457 * A structure used to configure a TX ring of an Ethernet port.
459 struct rte_eth_txconf {
460 struct rte_eth_thresh tx_thresh; /**< TX ring threshold registers. */
461 uint16_t tx_rs_thresh; /**< Drives the setting of RS bit on TXDs. */
462 uint16_t tx_free_thresh; /**< Drives the freeing of TX buffers. */
463 uint32_t txq_flags; /**< Set flags for the Tx queue */
467 * This enum indicates the flow control mode
469 enum rte_eth_fc_mode {
470 RTE_FC_NONE = 0, /**< Disable flow control. */
471 RTE_FC_RX_PAUSE, /**< RX pause frame, enable flowctrl on TX side. */
472 RTE_FC_TX_PAUSE, /**< TX pause frame, enable flowctrl on RX side. */
473 RTE_FC_FULL /**< Enable flow control on both side. */
477 * A structure used to configure Ethernet flow control parameter.
478 * These parameters will be configured into the register of the NIC.
479 * Please refer to the corresponding data sheet for proper value.
481 struct rte_eth_fc_conf {
482 uint32_t high_water; /**< High threshold value to trigger XOFF */
483 uint32_t low_water; /**< Low threshold value to trigger XON */
484 uint16_t pause_time; /**< Pause quota in the Pause frame */
485 uint16_t send_xon; /**< Is XON frame need be sent */
486 enum rte_eth_fc_mode mode; /**< Link flow control mode */
490 * A structure used to configure Ethernet priority flow control parameter.
491 * These parameters will be configured into the register of the NIC.
492 * Please refer to the corresponding data sheet for proper value.
494 struct rte_eth_pfc_conf {
495 struct rte_eth_fc_conf fc; /**< General flow control parameter. */
496 uint8_t priority; /**< VLAN User Priority. */
500 * Flow Director setting modes: none (default), signature or perfect.
503 RTE_FDIR_MODE_NONE = 0, /**< Disable FDIR support. */
504 RTE_FDIR_MODE_SIGNATURE, /**< Enable FDIR signature filter mode. */
505 RTE_FDIR_MODE_PERFECT, /**< Enable FDIR perfect filter mode. */
509 * Memory space that can be configured to store Flow Director filters
510 * in the board memory.
512 enum rte_fdir_pballoc_type {
513 RTE_FDIR_PBALLOC_64K = 0, /**< 64k. */
514 RTE_FDIR_PBALLOC_128K, /**< 128k. */
515 RTE_FDIR_PBALLOC_256K, /**< 256k. */
519 * Select report mode of FDIR hash information in RX descriptors.
521 enum rte_fdir_status_mode {
522 RTE_FDIR_NO_REPORT_STATUS = 0, /**< Never report FDIR hash. */
523 RTE_FDIR_REPORT_STATUS, /**< Only report FDIR hash for matching pkts. */
524 RTE_FDIR_REPORT_STATUS_ALWAYS, /**< Always report FDIR hash. */
528 * A structure used to configure the Flow Director (FDIR) feature
529 * of an Ethernet port.
531 * If mode is RTE_FDIR_DISABLE, the pballoc value is ignored.
533 struct rte_fdir_conf {
534 enum rte_fdir_mode mode; /**< Flow Director mode. */
535 enum rte_fdir_pballoc_type pballoc; /**< Space for FDIR filters. */
536 enum rte_fdir_status_mode status; /**< How to report FDIR hash. */
537 /** Offset of flexbytes field in RX packets (in 16-bit word units). */
538 uint8_t flexbytes_offset;
539 /** RX queue of packets matching a "drop" filter in perfect mode. */
544 * Possible l4type of FDIR filters.
547 RTE_FDIR_L4TYPE_NONE = 0, /**< None. */
548 RTE_FDIR_L4TYPE_UDP, /**< UDP. */
549 RTE_FDIR_L4TYPE_TCP, /**< TCP. */
550 RTE_FDIR_L4TYPE_SCTP, /**< SCTP. */
554 * Select IPv4 or IPv6 FDIR filters.
557 RTE_FDIR_IPTYPE_IPV4 = 0, /**< IPv4. */
558 RTE_FDIR_IPTYPE_IPV6 , /**< IPv6. */
562 * A structure used to define a FDIR packet filter.
564 struct rte_fdir_filter {
565 uint16_t flex_bytes; /**< Flex bytes value to match. */
566 uint16_t vlan_id; /**< VLAN ID value to match, 0 otherwise. */
567 uint16_t port_src; /**< Source port to match, 0 otherwise. */
568 uint16_t port_dst; /**< Destination port to match, 0 otherwise. */
570 uint32_t ipv4_addr; /**< IPv4 source address to match. */
571 uint32_t ipv6_addr[4]; /**< IPv6 source address to match. */
572 } ip_src; /**< IPv4/IPv6 source address to match (union of above). */
574 uint32_t ipv4_addr; /**< IPv4 destination address to match. */
575 uint32_t ipv6_addr[4]; /**< IPv6 destination address to match */
576 } ip_dst; /**< IPv4/IPv6 destination address to match (union of above). */
577 enum rte_l4type l4type; /**< l4type to match: NONE/UDP/TCP/SCTP. */
578 enum rte_iptype iptype; /**< IP packet type to match: IPv4 or IPv6. */
582 * A structure used to configure FDIR masks that are used by the device
583 * to match the various fields of RX packet headers.
584 * @note The only_ip_flow field has the opposite meaning compared to other
587 struct rte_fdir_masks {
588 /** When set to 1, packet l4type is \b NOT relevant in filters, and
589 source and destination port masks must be set to zero. */
590 uint8_t only_ip_flow;
591 /** If set to 1, vlan_id is relevant in filters. */
593 /** If set to 1, vlan_prio is relevant in filters. */
595 /** If set to 1, flexbytes is relevant in filters. */
597 /** If set to 1, set the IPv6 masks. Otherwise set the IPv4 masks. */
598 uint8_t set_ipv6_mask;
599 /** When set to 1, comparison of destination IPv6 address with IP6AT
600 registers is meaningful. */
601 uint8_t comp_ipv6_dst;
602 /** Mask of Destination IPv4 Address. All bits set to 1 define the
603 relevant bits to use in the destination address of an IPv4 packet
604 when matching it against FDIR filters. */
605 uint32_t dst_ipv4_mask;
606 /** Mask of Source IPv4 Address. All bits set to 1 define
607 the relevant bits to use in the source address of an IPv4 packet
608 when matching it against FDIR filters. */
609 uint32_t src_ipv4_mask;
610 /** Mask of Source IPv6 Address. All bits set to 1 define the
611 relevant BYTES to use in the source address of an IPv6 packet
612 when matching it against FDIR filters. */
613 uint16_t dst_ipv6_mask;
614 /** Mask of Destination IPv6 Address. All bits set to 1 define the
615 relevant BYTES to use in the destination address of an IPv6 packet
616 when matching it against FDIR filters. */
617 uint16_t src_ipv6_mask;
618 /** Mask of Source Port. All bits set to 1 define the relevant
619 bits to use in the source port of an IP packets when matching it
620 against FDIR filters. */
621 uint16_t src_port_mask;
622 /** Mask of Destination Port. All bits set to 1 define the relevant
623 bits to use in the destination port of an IP packet when matching it
624 against FDIR filters. */
625 uint16_t dst_port_mask;
629 * A structure used to report the status of the flow director filters in use.
631 struct rte_eth_fdir {
632 /** Number of filters with collision indication. */
634 /** Number of free (non programmed) filters. */
636 /** The Lookup hash value of the added filter that updated the value
637 of the MAXLEN field */
639 /** Longest linked list of filters in the table. */
641 /** Number of added filters. */
643 /** Number of removed filters. */
645 /** Number of failed added filters (no more space in device). */
647 /** Number of failed removed filters. */
652 * A structure used to enable/disable specific device interrupts.
654 struct rte_intr_conf {
655 /** enable/disable lsc interrupt. 0 (default) - disable, 1 enable */
660 * A structure used to configure an Ethernet port.
661 * Depending upon the RX multi-queue mode, extra advanced
662 * configuration settings may be needed.
664 struct rte_eth_conf {
666 /**< ETH_LINK_SPEED_10[0|00|000], or 0 for autonegotation */
667 uint16_t link_duplex;
668 /**< ETH_LINK_[HALF_DUPLEX|FULL_DUPLEX], or 0 for autonegotation */
669 struct rte_eth_rxmode rxmode; /**< Port RX configuration. */
670 struct rte_eth_txmode txmode; /**< Port TX configuration. */
672 struct rte_eth_rss_conf rss_conf; /**< Port RSS configuration */
673 struct rte_eth_vmdq_dcb_conf vmdq_dcb_conf;
674 /**< Port vmdq+dcb configuration. */
675 struct rte_eth_dcb_rx_conf dcb_rx_conf;
676 /**< Port dcb RX configuration. */
677 } rx_adv_conf; /**< Port RX filtering configuration (union). */
679 struct rte_eth_vmdq_dcb_tx_conf vmdq_dcb_tx_conf;
680 /**< Port vmdq+dcb TX configuration. */
681 struct rte_eth_dcb_tx_conf dcb_tx_conf;
682 /**< Port dcb TX configuration. */
683 } tx_adv_conf; /**< Port TX DCB configuration (union). */
684 /** Currently,Priority Flow Control(PFC) are supported,if DCB with PFC
685 is needed,and the variable must be set ETH_DCB_PFC_SUPPORT. */
686 uint32_t dcb_capability_en;
687 struct rte_fdir_conf fdir_conf; /**< FDIR configuration. */
688 struct rte_intr_conf intr_conf; /**< Interrupt mode configuration. */
692 * A structure used to retrieve the contextual information of
693 * an Ethernet device, such as the controlling driver of the device,
694 * its PCI context, etc...
696 struct rte_eth_dev_info {
697 struct rte_pci_device *pci_dev; /**< Device PCI information. */
698 const char *driver_name; /**< Device Driver name. */
699 uint32_t min_rx_bufsize; /**< Minimum size of RX buffer. */
700 uint32_t max_rx_pktlen; /**< Maximum configurable length of RX pkt. */
701 uint16_t max_rx_queues; /**< Maximum number of RX queues. */
702 uint16_t max_tx_queues; /**< Maximum number of TX queues. */
703 uint32_t max_mac_addrs; /**< Maximum number of MAC addresses. */
708 struct rte_eth_dev_callback;
709 /** @internal Structure to keep track of registered callbacks */
710 TAILQ_HEAD(rte_eth_dev_cb_list, rte_eth_dev_callback);
713 * Definitions of all functions exported by an Ethernet driver through the
714 * the generic structure of type *eth_dev_ops* supplied in the *rte_eth_dev*
715 * structure associated with an Ethernet device.
718 typedef int (*eth_dev_configure_t)(struct rte_eth_dev *dev);
719 /**< @internal Ethernet device configuration. */
721 typedef int (*eth_dev_start_t)(struct rte_eth_dev *dev);
722 /**< @internal Function used to start a configured Ethernet device. */
724 typedef void (*eth_dev_stop_t)(struct rte_eth_dev *dev);
725 /**< @internal Function used to stop a configured Ethernet device. */
727 typedef void (*eth_dev_close_t)(struct rte_eth_dev *dev);
728 /**< @internal Function used to close a configured Ethernet device. */
730 typedef void (*eth_promiscuous_enable_t)(struct rte_eth_dev *dev);
731 /**< @internal Function used to enable the RX promiscuous mode of an Ethernet device. */
733 typedef void (*eth_promiscuous_disable_t)(struct rte_eth_dev *dev);
734 /**< @internal Function used to disable the RX promiscuous mode of an Ethernet device. */
736 typedef void (*eth_allmulticast_enable_t)(struct rte_eth_dev *dev);
737 /**< @internal Enable the receipt of all multicast packets by an Ethernet device. */
739 typedef void (*eth_allmulticast_disable_t)(struct rte_eth_dev *dev);
740 /**< @internal Disable the receipt of all multicast packets by an Ethernet device. */
742 typedef int (*eth_link_update_t)(struct rte_eth_dev *dev,
743 int wait_to_complete);
744 /**< @internal Get link speed, duplex mode and state (up/down) of an Ethernet device. */
746 typedef void (*eth_stats_get_t)(struct rte_eth_dev *dev,
747 struct rte_eth_stats *igb_stats);
748 /**< @internal Get global I/O statistics of an Ethernet device. */
750 typedef void (*eth_stats_reset_t)(struct rte_eth_dev *dev);
751 /**< @internal Reset global I/O statistics of an Ethernet device to 0. */
753 typedef int (*eth_queue_stats_mapping_set_t)(struct rte_eth_dev *dev,
757 /**< @internal Set a queue statistics mapping for a tx/rx queue of an Ethernet device. */
759 typedef void (*eth_dev_infos_get_t)(struct rte_eth_dev *dev,
760 struct rte_eth_dev_info *dev_info);
761 /**< @internal Get specific informations of an Ethernet device. */
763 typedef int (*eth_rx_queue_setup_t)(struct rte_eth_dev *dev,
764 uint16_t rx_queue_id,
766 unsigned int socket_id,
767 const struct rte_eth_rxconf *rx_conf,
768 struct rte_mempool *mb_pool);
769 /**< @internal Set up a receive queue of an Ethernet device. */
771 typedef int (*eth_tx_queue_setup_t)(struct rte_eth_dev *dev,
772 uint16_t tx_queue_id,
774 unsigned int socket_id,
775 const struct rte_eth_txconf *tx_conf);
776 /**< @internal Setup a transmit queue of an Ethernet device. */
778 typedef void (*eth_queue_release_t)(void *queue);
779 /**< @internal Release memory resources allocated by given RX/TX queue. */
781 typedef uint32_t (*eth_rx_queue_count_t)(struct rte_eth_dev *dev,
782 uint16_t rx_queue_id);
783 /**< @Get number of available descriptors on a receive queue of an Ethernet device. */
785 typedef int (*eth_rx_descriptor_done_t)(void *rxq, uint16_t offset);
786 /**< @Check DD bit of specific RX descriptor */
788 typedef int (*vlan_filter_set_t)(struct rte_eth_dev *dev,
791 /**< @internal filtering of a VLAN Tag Identifier by an Ethernet device. */
793 typedef void (*vlan_tpid_set_t)(struct rte_eth_dev *dev,
795 /**< @internal set the outer VLAN-TPID by an Ethernet device. */
797 typedef void (*vlan_offload_set_t)(struct rte_eth_dev *dev, int mask);
798 /**< @internal set VLAN offload function by an Ethernet device. */
800 typedef void (*vlan_strip_queue_set_t)(struct rte_eth_dev *dev,
801 uint16_t rx_queue_id,
803 /**< @internal VLAN stripping enable/disable by an queue of Ethernet device. */
805 typedef uint16_t (*eth_rx_burst_t)(void *rxq,
806 struct rte_mbuf **rx_pkts,
808 /**< @internal Retrieve input packets from a receive queue of an Ethernet device. */
810 typedef uint16_t (*eth_tx_burst_t)(void *txq,
811 struct rte_mbuf **tx_pkts,
813 /**< @internal Send output packets on a transmit queue of an Ethernet device. */
815 typedef int (*fdir_add_signature_filter_t)(struct rte_eth_dev *dev,
816 struct rte_fdir_filter *fdir_ftr,
818 /**< @internal Setup a new signature filter rule on an Ethernet device */
820 typedef int (*fdir_update_signature_filter_t)(struct rte_eth_dev *dev,
821 struct rte_fdir_filter *fdir_ftr,
823 /**< @internal Update a signature filter rule on an Ethernet device */
825 typedef int (*fdir_remove_signature_filter_t)(struct rte_eth_dev *dev,
826 struct rte_fdir_filter *fdir_ftr);
827 /**< @internal Remove a signature filter rule on an Ethernet device */
829 typedef void (*fdir_infos_get_t)(struct rte_eth_dev *dev,
830 struct rte_eth_fdir *fdir);
831 /**< @internal Get information about fdir status */
833 typedef int (*fdir_add_perfect_filter_t)(struct rte_eth_dev *dev,
834 struct rte_fdir_filter *fdir_ftr,
835 uint16_t soft_id, uint8_t rx_queue,
837 /**< @internal Setup a new perfect filter rule on an Ethernet device */
839 typedef int (*fdir_update_perfect_filter_t)(struct rte_eth_dev *dev,
840 struct rte_fdir_filter *fdir_ftr,
841 uint16_t soft_id, uint8_t rx_queue,
843 /**< @internal Update a perfect filter rule on an Ethernet device */
845 typedef int (*fdir_remove_perfect_filter_t)(struct rte_eth_dev *dev,
846 struct rte_fdir_filter *fdir_ftr,
848 /**< @internal Remove a perfect filter rule on an Ethernet device */
850 typedef int (*fdir_set_masks_t)(struct rte_eth_dev *dev,
851 struct rte_fdir_masks *fdir_masks);
852 /**< @internal Setup flow director masks on an Ethernet device */
854 typedef int (*flow_ctrl_set_t)(struct rte_eth_dev *dev,
855 struct rte_eth_fc_conf *fc_conf);
856 /**< @internal Setup flow control parameter on an Ethernet device */
858 typedef int (*priority_flow_ctrl_set_t)(struct rte_eth_dev *dev,
859 struct rte_eth_pfc_conf *pfc_conf);
860 /**< @internal Setup priority flow control parameter on an Ethernet device */
862 typedef int (*reta_update_t)(struct rte_eth_dev *dev,
863 struct rte_eth_rss_reta *reta_conf);
864 /**< @internal Update RSS redirection table on an Ethernet device */
866 typedef int (*reta_query_t)(struct rte_eth_dev *dev,
867 struct rte_eth_rss_reta *reta_conf);
868 /**< @internal Query RSS redirection table on an Ethernet device */
870 typedef int (*eth_dev_led_on_t)(struct rte_eth_dev *dev);
871 /**< @internal Turn on SW controllable LED on an Ethernet device */
873 typedef int (*eth_dev_led_off_t)(struct rte_eth_dev *dev);
874 /**< @internal Turn off SW controllable LED on an Ethernet device */
876 typedef void (*eth_mac_addr_remove_t)(struct rte_eth_dev *dev, uint32_t index);
877 /**< @internal Remove MAC address from receive address register */
879 typedef void (*eth_mac_addr_add_t)(struct rte_eth_dev *dev,
880 struct ether_addr *mac_addr,
883 /**< @internal Set a MAC address into Receive Address Address Register */
886 * @internal A structure containing the functions exported by an Ethernet driver.
889 eth_dev_configure_t dev_configure; /**< Configure device. */
890 eth_dev_start_t dev_start; /**< Start device. */
891 eth_dev_stop_t dev_stop; /**< Stop device. */
892 eth_dev_close_t dev_close; /**< Close device. */
893 eth_promiscuous_enable_t promiscuous_enable; /**< Promiscuous ON. */
894 eth_promiscuous_disable_t promiscuous_disable;/**< Promiscuous OFF. */
895 eth_allmulticast_enable_t allmulticast_enable;/**< RX multicast ON. */
896 eth_allmulticast_disable_t allmulticast_disable;/**< RX multicast OF. */
897 eth_link_update_t link_update; /**< Get device link state. */
898 eth_stats_get_t stats_get; /**< Get device statistics. */
899 eth_stats_reset_t stats_reset; /**< Reset device statistics. */
900 eth_queue_stats_mapping_set_t queue_stats_mapping_set;
901 /**< Configure per queue stat counter mapping. */
902 eth_dev_infos_get_t dev_infos_get; /**< Get device info. */
903 vlan_filter_set_t vlan_filter_set; /**< Filter VLAN Setup. */
904 vlan_tpid_set_t vlan_tpid_set; /**< Outer VLAN TPID Setup. */
905 vlan_strip_queue_set_t vlan_strip_queue_set; /**< VLAN Stripping on queue. */
906 vlan_offload_set_t vlan_offload_set; /**< Set VLAN Offload. */
907 eth_rx_queue_setup_t rx_queue_setup;/**< Set up device RX queue.*/
908 eth_queue_release_t rx_queue_release;/**< Release RX queue.*/
909 eth_rx_queue_count_t rx_queue_count; /**< Get Rx queue count. */
910 eth_rx_descriptor_done_t rx_descriptor_done; /**< Check rxd DD bit */
911 eth_tx_queue_setup_t tx_queue_setup;/**< Set up device TX queue.*/
912 eth_queue_release_t tx_queue_release;/**< Release TX queue.*/
913 eth_dev_led_on_t dev_led_on; /**< Turn on LED. */
914 eth_dev_led_off_t dev_led_off; /**< Turn off LED. */
915 flow_ctrl_set_t flow_ctrl_set; /**< Setup flow control. */
916 priority_flow_ctrl_set_t priority_flow_ctrl_set; /**< Setup priority flow control.*/
917 eth_mac_addr_remove_t mac_addr_remove; /**< Remove MAC address */
918 eth_mac_addr_add_t mac_addr_add; /**< Add a MAC address */
920 /** Add a signature filter. */
921 fdir_add_signature_filter_t fdir_add_signature_filter;
922 /** Update a signature filter. */
923 fdir_update_signature_filter_t fdir_update_signature_filter;
924 /** Remove a signature filter. */
925 fdir_remove_signature_filter_t fdir_remove_signature_filter;
926 /** Get information about FDIR status. */
927 fdir_infos_get_t fdir_infos_get;
928 /** Add a perfect filter. */
929 fdir_add_perfect_filter_t fdir_add_perfect_filter;
930 /** Update a perfect filter. */
931 fdir_update_perfect_filter_t fdir_update_perfect_filter;
932 /** Remove a perfect filter. */
933 fdir_remove_perfect_filter_t fdir_remove_perfect_filter;
934 /** Setup masks for FDIR filtering. */
935 fdir_set_masks_t fdir_set_masks;
936 /** Update redirection table. */
937 reta_update_t reta_update;
938 /** Query redirection table. */
939 reta_query_t reta_query;
944 * The generic data structure associated with each ethernet device.
946 * Pointers to burst-oriented packet receive and transmit functions are
947 * located at the beginning of the structure, along with the pointer to
948 * where all the data elements for the particular device are stored in shared
949 * memory. This split allows the function pointer and driver data to be per-
950 * process, while the actual configuration data for the device is shared.
953 eth_rx_burst_t rx_pkt_burst; /**< Pointer to PMD receive function. */
954 eth_tx_burst_t tx_pkt_burst; /**< Pointer to PMD transmit function. */
955 struct rte_eth_dev_data *data; /**< Pointer to device data */
956 const struct eth_driver *driver;/**< Driver for this device */
957 struct eth_dev_ops *dev_ops; /**< Functions exported by PMD */
958 struct rte_pci_device *pci_dev; /**< PCI info. supplied by probing */
959 struct rte_eth_dev_cb_list callbacks; /**< User application callbacks */
962 struct rte_eth_dev_sriov {
963 uint8_t active; /**< SRIOV is active with 16, 32 or 64 pools */
964 uint8_t nb_q_per_pool; /**< rx queue number per pool */
965 uint16_t def_vmdq_idx; /**< Default pool num used for PF */
966 uint16_t def_pool_q_idx; /**< Default pool queue start reg index */
968 #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov)
972 * The data part, with no function pointers, associated with each ethernet device.
974 * This structure is safe to place in shared memory to be common among different
975 * processes in a multi-process configuration.
977 struct rte_eth_dev_data {
978 void **rx_queues; /**< Array of pointers to RX queues. */
979 void **tx_queues; /**< Array of pointers to TX queues. */
980 uint16_t nb_rx_queues; /**< Number of RX queues. */
981 uint16_t nb_tx_queues; /**< Number of TX queues. */
983 struct rte_eth_dev_sriov sriov; /**< SRIOV data */
985 void *dev_private; /**< PMD-specific private data */
987 struct rte_eth_link dev_link;
988 /**< Link-level information & status */
990 struct rte_eth_conf dev_conf; /**< Configuration applied to device. */
991 uint16_t max_frame_size; /**< Default is ETHER_MAX_LEN (1518). */
993 uint64_t rx_mbuf_alloc_failed; /**< RX ring mbuf allocation failures. */
994 struct ether_addr* mac_addrs;/**< Device Ethernet Link address. */
995 uint8_t port_id; /**< Device [external] port identifier. */
996 uint8_t promiscuous : 1, /**< RX promiscuous mode ON(1) / OFF(0). */
997 scattered_rx : 1, /**< RX of scattered packets is ON(1) / OFF(0) */
998 all_multicast : 1, /**< RX all multicast mode ON(1) / OFF(0). */
999 dev_started : 1; /**< Device state: STARTED(1) / STOPPED(0). */
1004 * The pool of *rte_eth_dev* structures. The size of the pool
1005 * is configured at compile-time in the <rte_ethdev.c> file.
1007 extern struct rte_eth_dev rte_eth_devices[];
1010 * Get the total number of Ethernet devices that have been successfully
1011 * initialized by the [matching] Ethernet driver during the PCI probing phase.
1012 * All devices whose port identifier is in the range
1013 * [0, rte_eth_dev_count() - 1] can be operated on by network applications.
1016 * - The total number of usable Ethernet devices.
1018 extern uint8_t rte_eth_dev_count(void);
1023 * Initialization function of an Ethernet driver invoked for each matching
1024 * Ethernet PCI device detected during the PCI probing phase.
1027 * The pointer to the [matching] Ethernet driver structure supplied by
1028 * the PMD when it registered itself.
1030 * The *eth_dev* pointer is the address of the *rte_eth_dev* structure
1031 * associated with the matching device and which have been [automatically]
1032 * allocated in the *rte_eth_devices* array.
1033 * The *eth_dev* structure is supplied to the driver initialization function
1034 * with the following fields already initialized:
1036 * - *pci_dev*: Holds the pointers to the *rte_pci_device* structure which
1037 * contains the generic PCI information of the matching device.
1039 * - *dev_private*: Holds a pointer to the device private data structure.
1041 * - *max_frame_size*: Contains the default Ethernet maximum frame length
1044 * - *port_id*: Contains the port index of the device (actually the index
1045 * of the *eth_dev* structure in the *rte_eth_devices* array).
1048 * - 0: Success, the device is properly initialized by the driver.
1049 * In particular, the driver MUST have set up the *dev_ops* pointer
1050 * of the *eth_dev* structure.
1051 * - <0: Error code of the device initialization failure.
1053 typedef int (*eth_dev_init_t)(struct eth_driver *eth_drv,
1054 struct rte_eth_dev *eth_dev);
1058 * The structure associated with a PMD Ethernet driver.
1060 * Each Ethernet driver acts as a PCI driver and is represented by a generic
1061 * *eth_driver* structure that holds:
1063 * - An *rte_pci_driver* structure (which must be the first field).
1065 * - The *eth_dev_init* function invoked for each matching PCI device.
1067 * - The size of the private data to allocate for each matching device.
1070 struct rte_pci_driver pci_drv; /**< The PMD is also a PCI driver. */
1071 eth_dev_init_t eth_dev_init; /**< Device init function. */
1072 unsigned int dev_private_size; /**< Size of device private data. */
1077 * A function invoked by the initialization function of an Ethernet driver
1078 * to simultaneously register itself as a PCI driver and as an Ethernet
1079 * Poll Mode Driver (PMD).
1082 * The pointer to the *eth_driver* structure associated with
1083 * the Ethernet driver.
1085 extern void rte_eth_driver_register(struct eth_driver *eth_drv);
1088 * The initialization function of the driver for
1089 * Intel(r) IGB Gigabit Ethernet Controller devices.
1090 * This function is invoked once at EAL start time.
1094 extern int rte_igb_pmd_init(void);
1097 * The initialization function of the driver for
1098 * Intel(r) EM Gigabit Ethernet Controller devices.
1099 * This function is invoked once at EAL start time.
1103 extern int rte_em_pmd_init(void);
1106 * The initialization function of the driver for 1Gbps Intel IGB_VF
1108 * Invoked once at EAL start time.
1112 extern int rte_igbvf_pmd_init(void);
1115 * The initialization function of the driver for 10Gbps Intel IXGBE
1117 * Invoked once at EAL start time.
1121 extern int rte_ixgbe_pmd_init(void);
1124 * The initialization function of the driver for 10Gbps Intel IXGBE_VF
1126 * Invoked once at EAL start time.
1130 extern int rte_ixgbevf_pmd_init(void);
1133 * The initialization function of *all* supported and enabled drivers.
1134 * Right now, the following PMDs are supported:
1140 * This function is invoked once at EAL start time.
1143 * Error code of the device initialization failure,
1144 * -ENODEV if there are no drivers available
1145 * (e.g. if all driver config options are = n).
1148 int rte_pmd_init_all(void)
1152 #ifdef RTE_LIBRTE_IGB_PMD
1153 if ((ret = rte_igb_pmd_init()) != 0) {
1154 RTE_LOG(ERR, PMD, "Cannot init igb PMD\n");
1157 if ((ret = rte_igbvf_pmd_init()) != 0) {
1158 RTE_LOG(ERR, PMD, "Cannot init igbvf PMD\n");
1161 #endif /* RTE_LIBRTE_IGB_PMD */
1163 #ifdef RTE_LIBRTE_EM_PMD
1164 if ((ret = rte_em_pmd_init()) != 0) {
1165 RTE_LOG(ERR, PMD, "Cannot init em PMD\n");
1168 #endif /* RTE_LIBRTE_EM_PMD */
1170 #ifdef RTE_LIBRTE_IXGBE_PMD
1171 if ((ret = rte_ixgbe_pmd_init()) != 0) {
1172 RTE_LOG(ERR, PMD, "Cannot init ixgbe PMD\n");
1175 if ((ret = rte_ixgbevf_pmd_init()) != 0) {
1176 RTE_LOG(ERR, PMD, "Cannot init ixgbevf PMD\n");
1179 #endif /* RTE_LIBRTE_IXGBE_PMD */
1182 RTE_LOG(ERR, PMD, "No PMD(s) are configured\n");
1187 * Configure an Ethernet device.
1188 * This function must be invoked first before any other function in the
1189 * Ethernet API. This function can also be re-invoked when a device is in the
1193 * The port identifier of the Ethernet device to configure.
1194 * @param nb_rx_queue
1195 * The number of receive queues to set up for the Ethernet device.
1196 * @param nb_tx_queue
1197 * The number of transmit queues to set up for the Ethernet device.
1199 * The pointer to the configuration data to be used for the Ethernet device.
1200 * The *rte_eth_conf* structure includes:
1201 * - the hardware offload features to activate, with dedicated fields for
1202 * each statically configurable offload hardware feature provided by
1203 * Ethernet devices, such as IP checksum or VLAN tag stripping for
1205 * - the Receive Side Scaling (RSS) configuration when using multiple RX
1208 * Embedding all configuration information in a single data structure
1209 * is the more flexible method that allows the addition of new features
1210 * without changing the syntax of the API.
1212 * - 0: Success, device configured.
1213 * - <0: Error code returned by the driver configuration function.
1215 extern int rte_eth_dev_configure(uint8_t port_id,
1216 uint16_t nb_rx_queue,
1217 uint16_t nb_tx_queue,
1218 const struct rte_eth_conf *eth_conf);
1221 * Allocate and set up a receive queue for an Ethernet device.
1223 * The function allocates a contiguous block of memory for *nb_rx_desc*
1224 * receive descriptors from a memory zone associated with *socket_id*
1225 * and initializes each receive descriptor with a network buffer allocated
1226 * from the memory pool *mb_pool*.
1229 * The port identifier of the Ethernet device.
1230 * @param rx_queue_id
1231 * The index of the receive queue to set up.
1232 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1233 * to rte_eth_dev_configure().
1235 * The number of receive descriptors to allocate for the receive ring.
1237 * The *socket_id* argument is the socket identifier in case of NUMA.
1238 * The value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1239 * the DMA memory allocated for the receive descriptors of the ring.
1241 * The pointer to the configuration data to be used for the receive queue.
1242 * The *rx_conf* structure contains an *rx_thresh* structure with the values
1243 * of the Prefetch, Host, and Write-Back threshold registers of the receive
1246 * The pointer to the memory pool from which to allocate *rte_mbuf* network
1247 * memory buffers to populate each descriptor of the receive ring.
1249 * - 0: Success, receive queue correctly set up.
1250 * - -EINVAL: The size of network buffers which can be allocated from the
1251 * memory pool does not fit the various buffer sizes allowed by the
1252 * device controller.
1253 * - -ENOMEM: Unable to allocate the receive ring descriptors or to
1254 * allocate network memory buffers from the memory pool when
1255 * initializing receive descriptors.
1257 extern int rte_eth_rx_queue_setup(uint8_t port_id, uint16_t rx_queue_id,
1258 uint16_t nb_rx_desc, unsigned int socket_id,
1259 const struct rte_eth_rxconf *rx_conf,
1260 struct rte_mempool *mb_pool);
1263 * Allocate and set up a transmit queue for an Ethernet device.
1266 * The port identifier of the Ethernet device.
1267 * @param tx_queue_id
1268 * The index of the transmit queue to set up.
1269 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1270 * to rte_eth_dev_configure().
1272 * The number of transmit descriptors to allocate for the transmit ring.
1274 * The *socket_id* argument is the socket identifier in case of NUMA.
1275 * Its value can be *SOCKET_ID_ANY* if there is no NUMA constraint for
1276 * the DMA memory allocated for the transmit descriptors of the ring.
1278 * The pointer to the configuration data to be used for the transmit queue.
1279 * The *tx_conf* structure contains the following data:
1280 * - The *tx_thresh* structure with the values of the Prefetch, Host, and
1281 * Write-Back threshold registers of the transmit ring.
1282 * When setting Write-Back threshold to the value greater then zero,
1283 * *tx_rs_thresh* value should be explicitly set to one.
1284 * - The *tx_free_thresh* value indicates the [minimum] number of network
1285 * buffers that must be pending in the transmit ring to trigger their
1286 * [implicit] freeing by the driver transmit function.
1287 * - The *tx_rs_thresh* value indicates the [minimum] number of transmit
1288 * descriptors that must be pending in the transmit ring before setting the
1289 * RS bit on a descriptor by the driver transmit function.
1290 * The *tx_rs_thresh* value should be less or equal then
1291 * *tx_free_thresh* value, and both of them should be less then
1293 * - The *txq_flags* member contains flags to pass to the TX queue setup
1294 * function to configure the behavior of the TX queue. This should be set
1295 * to 0 if no special configuration is required.
1297 * Note that setting *tx_free_thresh* or *tx_rs_thresh* value to 0 forces
1298 * the transmit function to use default values.
1300 * - 0: Success, the transmit queue is correctly set up.
1301 * - -ENOMEM: Unable to allocate the transmit ring descriptors.
1303 extern int rte_eth_tx_queue_setup(uint8_t port_id, uint16_t tx_queue_id,
1304 uint16_t nb_tx_desc, unsigned int socket_id,
1305 const struct rte_eth_txconf *tx_conf);
1308 * Return the NUMA socket to which an Ethernet device is connected
1311 * The port identifier of the Ethernet device
1313 * The NUMA socket id to which the Ethernet device is connected or
1314 * a default of zero if the socket could not be determined.
1315 * -1 is returned is the port_id value is out of range.
1317 extern int rte_eth_dev_socket_id(uint8_t port_id);
1321 * Start an Ethernet device.
1323 * The device start step is the last one and consists of setting the configured
1324 * offload features and in starting the transmit and the receive units of the
1326 * On success, all basic functions exported by the Ethernet API (link status,
1327 * receive/transmit, and so on) can be invoked.
1330 * The port identifier of the Ethernet device.
1332 * - 0: Success, Ethernet device started.
1333 * - <0: Error code of the driver device start function.
1335 extern int rte_eth_dev_start(uint8_t port_id);
1338 * Stop an Ethernet device. The device can be restarted with a call to
1339 * rte_eth_dev_start()
1342 * The port identifier of the Ethernet device.
1344 extern void rte_eth_dev_stop(uint8_t port_id);
1347 * Close an Ethernet device. The device cannot be restarted!
1350 * The port identifier of the Ethernet device.
1352 extern void rte_eth_dev_close(uint8_t port_id);
1355 * Enable receipt in promiscuous mode for an Ethernet device.
1358 * The port identifier of the Ethernet device.
1360 extern void rte_eth_promiscuous_enable(uint8_t port_id);
1363 * Disable receipt in promiscuous mode for an Ethernet device.
1366 * The port identifier of the Ethernet device.
1368 extern void rte_eth_promiscuous_disable(uint8_t port_id);
1371 * Return the value of promiscuous mode for an Ethernet device.
1374 * The port identifier of the Ethernet device.
1376 * - (1) if promiscuous is enabled
1377 * - (0) if promiscuous is disabled.
1380 extern int rte_eth_promiscuous_get(uint8_t port_id);
1383 * Enable the receipt of any multicast frame by an Ethernet device.
1386 * The port identifier of the Ethernet device.
1388 extern void rte_eth_allmulticast_enable(uint8_t port_id);
1391 * Disable the receipt of all multicast frames by an Ethernet device.
1394 * The port identifier of the Ethernet device.
1396 extern void rte_eth_allmulticast_disable(uint8_t port_id);
1399 * Return the value of allmulticast mode for an Ethernet device.
1402 * The port identifier of the Ethernet device.
1404 * - (1) if allmulticast is enabled
1405 * - (0) if allmulticast is disabled.
1408 extern int rte_eth_allmulticast_get(uint8_t port_id);
1411 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
1412 * or FULL-DUPLEX) of the physical link of an Ethernet device. It might need
1413 * to wait up to 9 seconds in it.
1416 * The port identifier of the Ethernet device.
1418 * A pointer to an *rte_eth_link* structure to be filled with
1419 * the status, the speed and the mode of the Ethernet device link.
1421 extern void rte_eth_link_get(uint8_t port_id, struct rte_eth_link *link);
1424 * Retrieve the status (ON/OFF), the speed (in Mbps) and the mode (HALF-DUPLEX
1425 * or FULL-DUPLEX) of the physical link of an Ethernet device. It is a no-wait
1426 * version of rte_eth_link_get().
1429 * The port identifier of the Ethernet device.
1431 * A pointer to an *rte_eth_link* structure to be filled with
1432 * the status, the speed and the mode of the Ethernet device link.
1434 extern void rte_eth_link_get_nowait(uint8_t port_id,
1435 struct rte_eth_link *link);
1438 * Retrieve the general I/O statistics of an Ethernet device.
1441 * The port identifier of the Ethernet device.
1443 * A pointer to a structure of type *rte_eth_stats* to be filled with
1444 * the values of device counters for the following set of statistics:
1445 * - *ipackets* with the total of successfully received packets.
1446 * - *opackets* with the total of successfully transmitted packets.
1447 * - *ibytes* with the total of successfully received bytes.
1448 * - *obytes* with the total of successfully transmitted bytes.
1449 * - *ierrors* with the total of erroneous received packets.
1450 * - *oerrors* with the total of failed transmitted packets.
1452 extern void rte_eth_stats_get(uint8_t port_id, struct rte_eth_stats *stats);
1455 * Reset the general I/O statistics of an Ethernet device.
1458 * The port identifier of the Ethernet device.
1460 extern void rte_eth_stats_reset(uint8_t port_id);
1463 * Set a mapping for the specified transmit queue to the specified per-queue
1464 * statistics counter.
1467 * The port identifier of the Ethernet device.
1468 * @param tx_queue_id
1469 * The index of the transmit queue for which a queue stats mapping is required.
1470 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1471 * to rte_eth_dev_configure().
1473 * The per-queue packet statistics functionality number that the transmit
1474 * queue is to be assigned.
1475 * The value must be in the range [0, RTE_MAX_ETHPORT_QUEUE_STATS_MAPS - 1].
1477 * Zero if successful. Non-zero otherwise.
1479 extern int rte_eth_dev_set_tx_queue_stats_mapping(uint8_t port_id,
1480 uint16_t tx_queue_id,
1484 * Set a mapping for the specified receive queue to the specified per-queue
1485 * statistics counter.
1488 * The port identifier of the Ethernet device.
1489 * @param rx_queue_id
1490 * The index of the receive queue for which a queue stats mapping is required.
1491 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1492 * to rte_eth_dev_configure().
1494 * The per-queue packet statistics functionality number that the receive
1495 * queue is to be assigned.
1496 * The value must be in the range [0, RTE_MAX_ETHPORT_QUEUE_STATS_MAPS - 1].
1498 * Zero if successful. Non-zero otherwise.
1500 extern int rte_eth_dev_set_rx_queue_stats_mapping(uint8_t port_id,
1501 uint16_t rx_queue_id,
1505 * Retrieve the Ethernet address of an Ethernet device.
1508 * The port identifier of the Ethernet device.
1510 * A pointer to a structure of type *ether_addr* to be filled with
1511 * the Ethernet address of the Ethernet device.
1513 extern void rte_eth_macaddr_get(uint8_t port_id, struct ether_addr *mac_addr);
1516 * Retrieve the contextual information of an Ethernet device.
1519 * The port identifier of the Ethernet device.
1521 * A pointer to a structure of type *rte_eth_dev_info* to be filled with
1522 * the contextual information of the Ethernet device.
1524 extern void rte_eth_dev_info_get(uint8_t port_id,
1525 struct rte_eth_dev_info *dev_info);
1528 * Enable/Disable hardware filtering by an Ethernet device of received
1529 * VLAN packets tagged with a given VLAN Tag Identifier.
1532 * The port identifier of the Ethernet device.
1534 * The VLAN Tag Identifier whose filtering must be enabled or disabled.
1536 * If > 0, enable VLAN filtering of VLAN packets tagged with *vlan_id*.
1537 * Otherwise, disable VLAN filtering of VLAN packets tagged with *vlan_id*.
1539 * - (0) if successful.
1540 * - (-ENOSUP) if hardware-assisted VLAN filtering not configured.
1541 * - (-ENODEV) if *port_id* invalid.
1542 * - (-ENOSYS) if VLAN filtering on *port_id* disabled.
1543 * - (-EINVAL) if *vlan_id* > 4095.
1545 extern int rte_eth_dev_vlan_filter(uint8_t port_id, uint16_t vlan_id , int on);
1548 * Enable/Disable hardware VLAN Strip by a rx queue of an Ethernet device.
1549 * 82599/X540 can support VLAN stripping at the rx queue level
1552 * The port identifier of the Ethernet device.
1553 * @param rx_queue_id
1554 * The index of the receive queue for which a queue stats mapping is required.
1555 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1556 * to rte_eth_dev_configure().
1558 * If 1, Enable VLAN Stripping of the receive queue of the Ethernet port.
1559 * If 0, Disable VLAN Stripping of the receive queue of the Ethernet port.
1561 * - (0) if successful.
1562 * - (-ENOSUP) if hardware-assisted VLAN stripping not configured.
1563 * - (-ENODEV) if *port_id* invalid.
1564 * - (-EINVAL) if *rx_queue_id* invalid.
1566 extern int rte_eth_dev_set_vlan_strip_on_queue(uint8_t port_id,
1567 uint16_t rx_queue_id, int on);
1570 * Set the Outer VLAN Ether Type by an Ethernet device, it can be inserted to
1571 * the VLAN Header. This is a register setup available on some Intel NIC, not
1572 * but all, please check the data sheet for availability.
1575 * The port identifier of the Ethernet device.
1577 * The Tag Protocol ID
1579 * - (0) if successful.
1580 * - (-ENOSUP) if hardware-assisted VLAN TPID setup is not supported.
1581 * - (-ENODEV) if *port_id* invalid.
1583 extern int rte_eth_dev_set_vlan_ether_type(uint8_t port_id, uint16_t tag_type);
1586 * Set VLAN offload configuration on an Ethernet device
1587 * Enable/Disable Extended VLAN by an Ethernet device, This is a register setup
1588 * available on some Intel NIC, not but all, please check the data sheet for
1590 * Enable/Disable VLAN Strip can be done on rx queue for certain NIC, but here
1591 * the configuration is applied on the port level.
1594 * The port identifier of the Ethernet device.
1595 * @param offload_mask
1596 * The VLAN Offload bit mask can be mixed use with "OR"
1597 * ETH_VLAN_STRIP_OFFLOAD
1598 * ETH_VLAN_FILTER_OFFLOAD
1599 * ETH_VLAN_EXTEND_OFFLOAD
1601 * - (0) if successful.
1602 * - (-ENOSUP) if hardware-assisted VLAN filtering not configured.
1603 * - (-ENODEV) if *port_id* invalid.
1605 extern int rte_eth_dev_set_vlan_offload(uint8_t port_id, int offload_mask);
1608 * Read VLAN Offload configuration from an Ethernet device
1611 * The port identifier of the Ethernet device.
1613 * - (>0) if successful. Bit mask to indicate
1614 * ETH_VLAN_STRIP_OFFLOAD
1615 * ETH_VLAN_FILTER_OFFLOAD
1616 * ETH_VLAN_EXTEND_OFFLOAD
1617 * - (-ENODEV) if *port_id* invalid.
1619 extern int rte_eth_dev_get_vlan_offload(uint8_t port_id);
1623 * Retrieve a burst of input packets from a receive queue of an Ethernet
1624 * device. The retrieved packets are stored in *rte_mbuf* structures whose
1625 * pointers are supplied in the *rx_pkts* array.
1627 * The rte_eth_rx_burst() function loops, parsing the RX ring of the
1628 * receive queue, up to *nb_pkts* packets, and for each completed RX
1629 * descriptor in the ring, it performs the following operations:
1631 * - Initialize the *rte_mbuf* data structure associated with the
1632 * RX descriptor according to the information provided by the NIC into
1633 * that RX descriptor.
1635 * - Store the *rte_mbuf* data structure into the next entry of the
1638 * - Replenish the RX descriptor with a new *rte_mbuf* buffer
1639 * allocated from the memory pool associated with the receive queue at
1640 * initialization time.
1642 * When retrieving an input packet that was scattered by the controller
1643 * into multiple receive descriptors, the rte_eth_rx_burst() function
1644 * appends the associated *rte_mbuf* buffers to the first buffer of the
1647 * The rte_eth_rx_burst() function returns the number of packets
1648 * actually retrieved, which is the number of *rte_mbuf* data structures
1649 * effectively supplied into the *rx_pkts* array.
1650 * A return value equal to *nb_pkts* indicates that the RX queue contained
1651 * at least *rx_pkts* packets, and this is likely to signify that other
1652 * received packets remain in the input queue. Applications implementing
1653 * a "retrieve as much received packets as possible" policy can check this
1654 * specific case and keep invoking the rte_eth_rx_burst() function until
1655 * a value less than *nb_pkts* is returned.
1657 * This receive method has the following advantages:
1659 * - It allows a run-to-completion network stack engine to retrieve and
1660 * to immediately process received packets in a fast burst-oriented
1661 * approach, avoiding the overhead of unnecessary intermediate packet
1662 * queue/dequeue operations.
1664 * - Conversely, it also allows an asynchronous-oriented processing
1665 * method to retrieve bursts of received packets and to immediately
1666 * queue them for further parallel processing by another logical core,
1667 * for instance. However, instead of having received packets being
1668 * individually queued by the driver, this approach allows the invoker
1669 * of the rte_eth_rx_burst() function to queue a burst of retrieved
1670 * packets at a time and therefore dramatically reduce the cost of
1671 * enqueue/dequeue operations per packet.
1673 * - It allows the rte_eth_rx_burst() function of the driver to take
1674 * advantage of burst-oriented hardware features (CPU cache,
1675 * prefetch instructions, and so on) to minimize the number of CPU
1676 * cycles per packet.
1678 * To summarize, the proposed receive API enables many
1679 * burst-oriented optimizations in both synchronous and asynchronous
1680 * packet processing environments with no overhead in both cases.
1682 * The rte_eth_rx_burst() function does not provide any error
1683 * notification to avoid the corresponding overhead. As a hint, the
1684 * upper-level application might check the status of the device link once
1685 * being systematically returned a 0 value for a given number of tries.
1688 * The port identifier of the Ethernet device.
1690 * The index of the receive queue from which to retrieve input packets.
1691 * The value must be in the range [0, nb_rx_queue - 1] previously supplied
1692 * to rte_eth_dev_configure().
1694 * The address of an array of pointers to *rte_mbuf* structures that
1695 * must be large enough to store *nb_pkts* pointers in it.
1697 * The maximum number of packets to retrieve.
1699 * The number of packets actually retrieved, which is the number
1700 * of pointers to *rte_mbuf* structures effectively supplied to the
1703 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
1704 extern uint16_t rte_eth_rx_burst(uint8_t port_id, uint16_t queue_id,
1705 struct rte_mbuf **rx_pkts, uint16_t nb_pkts);
1707 static inline uint16_t
1708 rte_eth_rx_burst(uint8_t port_id, uint16_t queue_id,
1709 struct rte_mbuf **rx_pkts, uint16_t nb_pkts)
1711 struct rte_eth_dev *dev;
1713 dev = &rte_eth_devices[port_id];
1714 return (*dev->rx_pkt_burst)(dev->data->rx_queues[queue_id], rx_pkts, nb_pkts);
1719 * Get the number of used descriptors in a specific queue
1722 * The port identifier of the Ethernet device.
1724 * The queue id on the specific port.
1726 * The number of used descriptors in the specific queue.
1728 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
1729 extern uint32_t rte_eth_rx_queue_count(uint8_t port_id, uint16_t queue_id);
1731 static inline uint32_t
1732 rte_eth_rx_queue_count(uint8_t port_id, uint16_t queue_id)
1734 struct rte_eth_dev *dev;
1736 dev = &rte_eth_devices[port_id];
1737 return (*dev->dev_ops->rx_queue_count)(dev, queue_id);
1742 * Check if the DD bit of the specific RX descriptor in the queue has been set
1745 * The port identifier of the Ethernet device.
1747 * The queue id on the specific port.
1749 * The offset of the descriptor ID from tail.
1751 * - (1) if the specific DD bit is set.
1752 * - (0) if the specific DD bit is not set.
1753 * - (-ENODEV) if *port_id* invalid.
1755 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
1756 extern int rte_eth_rx_descriptor_done(uint8_t port_id,
1761 rte_eth_rx_descriptor_done(uint8_t port_id, uint16_t queue_id, uint16_t offset)
1763 struct rte_eth_dev *dev;
1765 dev = &rte_eth_devices[port_id];
1766 return (*dev->dev_ops->rx_descriptor_done)( \
1767 dev->data->rx_queues[queue_id], offset);
1772 * Send a burst of output packets on a transmit queue of an Ethernet device.
1774 * The rte_eth_tx_burst() function is invoked to transmit output packets
1775 * on the output queue *queue_id* of the Ethernet device designated by its
1777 * The *nb_pkts* parameter is the number of packets to send which are
1778 * supplied in the *tx_pkts* array of *rte_mbuf* structures.
1779 * The rte_eth_tx_burst() function loops, sending *nb_pkts* packets,
1780 * up to the number of transmit descriptors available in the TX ring of the
1782 * For each packet to send, the rte_eth_tx_burst() function performs
1783 * the following operations:
1785 * - Pick up the next available descriptor in the transmit ring.
1787 * - Free the network buffer previously sent with that descriptor, if any.
1789 * - Initialize the transmit descriptor with the information provided
1790 * in the *rte_mbuf data structure.
1792 * In the case of a segmented packet composed of a list of *rte_mbuf* buffers,
1793 * the rte_eth_tx_burst() function uses several transmit descriptors
1796 * The rte_eth_tx_burst() function returns the number of packets it
1797 * actually sent. A return value equal to *nb_pkts* means that all packets
1798 * have been sent, and this is likely to signify that other output packets
1799 * could be immediately transmitted again. Applications that implement a
1800 * "send as many packets to transmit as possible" policy can check this
1801 * specific case and keep invoking the rte_eth_tx_burst() function until
1802 * a value less than *nb_pkts* is returned.
1804 * It is the responsibility of the rte_eth_tx_burst() function to
1805 * transparently free the memory buffers of packets previously sent.
1806 * This feature is driven by the *tx_free_thresh* value supplied to the
1807 * rte_eth_dev_configure() function at device configuration time.
1808 * When the number of previously sent packets reached the "minimum transmit
1809 * packets to free" threshold, the rte_eth_tx_burst() function must
1810 * [attempt to] free the *rte_mbuf* buffers of those packets whose
1811 * transmission was effectively completed.
1814 * The port identifier of the Ethernet device.
1816 * The index of the transmit queue through which output packets must be
1818 * The value must be in the range [0, nb_tx_queue - 1] previously supplied
1819 * to rte_eth_dev_configure().
1821 * The address of an array of *nb_pkts* pointers to *rte_mbuf* structures
1822 * which contain the output packets.
1824 * The maximum number of packets to transmit.
1826 * The number of output packets actually stored in transmit descriptors of
1827 * the transmit ring. The return value can be less than the value of the
1828 * *tx_pkts* parameter when the transmit ring is full or has been filled up.
1830 #ifdef RTE_LIBRTE_ETHDEV_DEBUG
1831 extern uint16_t rte_eth_tx_burst(uint8_t port_id, uint16_t queue_id,
1832 struct rte_mbuf **tx_pkts, uint16_t nb_pkts);
1834 static inline uint16_t
1835 rte_eth_tx_burst(uint8_t port_id, uint16_t queue_id,
1836 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
1838 struct rte_eth_dev *dev;
1840 dev = &rte_eth_devices[port_id];
1841 return (*dev->tx_pkt_burst)(dev->data->tx_queues[queue_id], tx_pkts, nb_pkts);
1846 * Setup a new signature filter rule on an Ethernet device
1849 * The port identifier of the Ethernet device.
1850 * @param fdir_filter
1851 * The pointer to the fdir filter structure describing the signature filter
1853 * The *rte_fdir_filter* structure includes the values of the different fields
1854 * to match: source and destination IP addresses, vlan id, flexbytes, source
1855 * and destination ports, and so on.
1857 * The index of the RX queue where to store RX packets matching the added
1858 * signature filter defined in fdir_filter.
1860 * - (0) if successful.
1861 * - (-ENOTSUP) if hardware doesn't support flow director mode.
1862 * - (-ENODEV) if *port_id* invalid.
1863 * - (-ENOSYS) if the FDIR mode is not configured in signature mode
1865 * - (-EINVAL) if the fdir_filter information is not correct.
1867 int rte_eth_dev_fdir_add_signature_filter(uint8_t port_id,
1868 struct rte_fdir_filter *fdir_filter,
1872 * Update a signature filter rule on an Ethernet device.
1873 * If the rule doesn't exits, it is created.
1876 * The port identifier of the Ethernet device.
1878 * The pointer to the structure describing the signature filter rule.
1879 * The *rte_fdir_filter* structure includes the values of the different fields
1880 * to match: source and destination IP addresses, vlan id, flexbytes, source
1881 * and destination ports, and so on.
1883 * The index of the RX queue where to store RX packets matching the added
1884 * signature filter defined in fdir_ftr.
1886 * - (0) if successful.
1887 * - (-ENOTSUP) if hardware doesn't support flow director mode.
1888 * - (-ENODEV) if *port_id* invalid.
1889 * - (-ENOSYS) if the flow director mode is not configured in signature mode
1891 * - (-EINVAL) if the fdir_filter information is not correct.
1893 int rte_eth_dev_fdir_update_signature_filter(uint8_t port_id,
1894 struct rte_fdir_filter *fdir_ftr,
1898 * Remove a signature filter rule on an Ethernet device.
1901 * The port identifier of the Ethernet device.
1903 * The pointer to the structure describing the signature filter rule.
1904 * The *rte_fdir_filter* structure includes the values of the different fields
1905 * to match: source and destination IP addresses, vlan id, flexbytes, source
1906 * and destination ports, and so on.
1908 * - (0) if successful.
1909 * - (-ENOTSUP) if hardware doesn't support flow director mode.
1910 * - (-ENODEV) if *port_id* invalid.
1911 * - (-ENOSYS) if the flow director mode is not configured in signature mode
1913 * - (-EINVAL) if the fdir_filter information is not correct.
1915 int rte_eth_dev_fdir_remove_signature_filter(uint8_t port_id,
1916 struct rte_fdir_filter *fdir_ftr);
1919 * Retrieve the flow director information of an Ethernet device.
1922 * The port identifier of the Ethernet device.
1924 * A pointer to a structure of type *rte_eth_dev_fdir* to be filled with
1925 * the flow director information of the Ethernet device.
1927 * - (0) if successful.
1928 * - (-ENOTSUP) if hardware doesn't support flow director mode.
1929 * - (-ENODEV) if *port_id* invalid.
1930 * - (-ENOSYS) if the flow director mode is not configured on *port_id*.
1932 int rte_eth_dev_fdir_get_infos(uint8_t port_id, struct rte_eth_fdir *fdir);
1935 * Add a new perfect filter rule on an Ethernet device.
1938 * The port identifier of the Ethernet device.
1939 * @param fdir_filter
1940 * The pointer to the structure describing the perfect filter rule.
1941 * The *rte_fdir_filter* structure includes the values of the different fields
1942 * to match: source and destination IP addresses, vlan id, flexbytes, source
1943 * and destination ports, and so on.
1944 * IPv6 are not supported.
1946 * The 16-bit value supplied in the field hash.fdir.id of mbuf for RX
1947 * packets matching the perfect filter.
1949 * The index of the RX queue where to store RX packets matching the added
1950 * perfect filter defined in fdir_filter.
1952 * If drop is set to 1, matching RX packets are stored into the RX drop
1953 * queue defined in the rte_fdir_conf.
1955 * - (0) if successful.
1956 * - (-ENOTSUP) if hardware doesn't support flow director mode.
1957 * - (-ENODEV) if *port_id* invalid.
1958 * - (-ENOSYS) if the flow director mode is not configured in perfect mode
1960 * - (-EINVAL) if the fdir_filter information is not correct.
1962 int rte_eth_dev_fdir_add_perfect_filter(uint8_t port_id,
1963 struct rte_fdir_filter *fdir_filter,
1964 uint16_t soft_id, uint8_t rx_queue,
1968 * Update a perfect filter rule on an Ethernet device.
1969 * If the rule doesn't exits, it is created.
1972 * The port identifier of the Ethernet device.
1973 * @param fdir_filter
1974 * The pointer to the structure describing the perfect filter rule.
1975 * The *rte_fdir_filter* structure includes the values of the different fields
1976 * to match: source and destination IP addresses, vlan id, flexbytes, source
1977 * and destination ports, and so on.
1978 * IPv6 are not supported.
1980 * The 16-bit value supplied in the field hash.fdir.id of mbuf for RX
1981 * packets matching the perfect filter.
1983 * The index of the RX queue where to store RX packets matching the added
1984 * perfect filter defined in fdir_filter.
1986 * If drop is set to 1, matching RX packets are stored into the RX drop
1987 * queue defined in the rte_fdir_conf.
1989 * - (0) if successful.
1990 * - (-ENOTSUP) if hardware doesn't support flow director mode.
1991 * - (-ENODEV) if *port_id* invalid.
1992 * - (-ENOSYS) if the flow director mode is not configured in perfect mode
1994 * - (-EINVAL) if the fdir_filter information is not correct.
1996 int rte_eth_dev_fdir_update_perfect_filter(uint8_t port_id,
1997 struct rte_fdir_filter *fdir_filter,
1998 uint16_t soft_id, uint8_t rx_queue,
2002 * Remove a perfect filter rule on an Ethernet device.
2005 * The port identifier of the Ethernet device.
2006 * @param fdir_filter
2007 * The pointer to the structure describing the perfect filter rule.
2008 * The *rte_fdir_filter* structure includes the values of the different fields
2009 * to match: source and destination IP addresses, vlan id, flexbytes, source
2010 * and destination ports, and so on.
2011 * IPv6 are not supported.
2013 * The soft_id value provided when adding/updating the removed filter.
2015 * - (0) if successful.
2016 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2017 * - (-ENODEV) if *port_id* invalid.
2018 * - (-ENOSYS) if the flow director mode is not configured in perfect mode
2020 * - (-EINVAL) if the fdir_filter information is not correct.
2022 int rte_eth_dev_fdir_remove_perfect_filter(uint8_t port_id,
2023 struct rte_fdir_filter *fdir_filter,
2026 * Configure globally the masks for flow director mode for an Ethernet device.
2027 * For example, the device can match packets with only the first 24 bits of
2028 * the IPv4 source address.
2030 * The following fields can be masked: IPv4 addresses and L4 port numbers.
2031 * The following fields can be either enabled or disabled completely for the
2032 * matching functionality: VLAN ID tag; VLAN Priority + CFI bit; Flexible 2-byte
2034 * IPv6 masks are not supported.
2036 * All filters must comply with the masks previously configured.
2037 * For example, with a mask equal to 255.255.255.0 for the source IPv4 address,
2038 * all IPv4 filters must be created with a source IPv4 address that fits the
2041 * This function flushes all filters that have been previously added in
2045 * The port identifier of the Ethernet device.
2047 * The pointer to the fdir mask structure describing relevant headers fields
2048 * and relevant bits to use when matching packets addresses and ports.
2049 * IPv6 masks are not supported.
2051 * - (0) if successful.
2052 * - (-ENOTSUP) if hardware doesn't support flow director mode.
2053 * - (-ENODEV) if *port_id* invalid.
2054 * - (-ENOSYS) if the flow director mode is not configured in perfect
2055 * mode on *port_id*.
2056 * - (-EINVAL) if the fdir_filter information is not correct
2058 int rte_eth_dev_fdir_set_masks(uint8_t port_id,
2059 struct rte_fdir_masks *fdir_mask);
2062 * The eth device event type for interrupt, and maybe others in the future.
2064 enum rte_eth_event_type {
2065 RTE_ETH_EVENT_UNKNOWN, /**< unknown event type */
2066 RTE_ETH_EVENT_INTR_LSC, /**< lsc interrupt event */
2067 RTE_ETH_EVENT_MAX /**< max value of this enum */
2070 typedef void (*rte_eth_dev_cb_fn)(uint8_t port_id, \
2071 enum rte_eth_event_type event, void *cb_arg);
2072 /**< user application callback to be registered for interrupts */
2077 * Register a callback function for specific port id.
2084 * User supplied callback function to be called.
2086 * Pointer to the parameters for the registered callback.
2089 * - On success, zero.
2090 * - On failure, a negative value.
2092 int rte_eth_dev_callback_register(uint8_t port_id,
2093 enum rte_eth_event_type event,
2094 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
2097 * Unregister a callback function for specific port id.
2104 * User supplied callback function to be called.
2106 * Pointer to the parameters for the registered callback. -1 means to
2107 * remove all for the same callback address and same event.
2110 * - On success, zero.
2111 * - On failure, a negative value.
2113 int rte_eth_dev_callback_unregister(uint8_t port_id,
2114 enum rte_eth_event_type event,
2115 rte_eth_dev_cb_fn cb_fn, void *cb_arg);
2118 * @internal Executes all the user application registered callbacks for
2119 * the specific device. It is for DPDK internal user only. User
2120 * application should not call it directly.
2123 * Pointer to struct rte_eth_dev.
2125 * Eth device interrupt event type.
2130 void _rte_eth_dev_callback_process(struct rte_eth_dev *dev,
2131 enum rte_eth_event_type event);
2134 * Turn on the LED on the Ethernet device.
2135 * This function turns on the LED on the Ethernet device.
2138 * The port identifier of the Ethernet device.
2140 * - (0) if successful.
2141 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
2143 * - (-ENODEV) if *port_id* invalid.
2145 int rte_eth_led_on(uint8_t port_id);
2148 * Turn off the LED on the Ethernet device.
2149 * This function turns off the LED on the Ethernet device.
2152 * The port identifier of the Ethernet device.
2154 * - (0) if successful.
2155 * - (-ENOTSUP) if underlying hardware OR driver doesn't support
2157 * - (-ENODEV) if *port_id* invalid.
2159 int rte_eth_led_off(uint8_t port_id);
2162 * Configure the Ethernet link flow control for Ethernet device
2165 * The port identifier of the Ethernet device.
2167 * The pointer to the structure of the flow control parameters.
2169 * - (0) if successful.
2170 * - (-ENOTSUP) if hardware doesn't support flow control mode.
2171 * - (-ENODEV) if *port_id* invalid.
2172 * - (-EINVAL) if bad parameter
2173 * - (-EIO) if flow control setup failure
2175 int rte_eth_dev_flow_ctrl_set(uint8_t port_id,
2176 struct rte_eth_fc_conf *fc_conf);
2179 * Configure the Ethernet priority flow control under DCB environment
2180 * for Ethernet device.
2183 * The port identifier of the Ethernet device.
2185 * The pointer to the structure of the priority flow control parameters.
2187 * - (0) if successful.
2188 * - (-ENOTSUP) if hardware doesn't support priority flow control mode.
2189 * - (-ENODEV) if *port_id* invalid.
2190 * - (-EINVAL) if bad parameter
2191 * - (-EIO) if flow control setup failure
2193 int rte_eth_dev_priority_flow_ctrl_set(uint8_t port_id,
2194 struct rte_eth_pfc_conf *pfc_conf);
2197 * Add a MAC address to an internal array of addresses used to enable whitelist
2198 * filtering to accept packets only if the destination MAC address matches.
2201 * The port identifier of the Ethernet device.
2203 * The MAC address to add.
2205 * VMDq pool index to associate address with (if VMDq is enabled). If VMDq is
2206 * not enabled, this should be set to 0.
2208 * - (0) if successfully added or *mac_addr" was already added.
2209 * - (-ENOTSUP) if hardware doesn't support this feature.
2210 * - (-ENODEV) if *port* is invalid.
2211 * - (-ENOSPC) if no more MAC addresses can be added.
2212 * - (-EINVAL) if MAC address is invalid.
2214 int rte_eth_dev_mac_addr_add(uint8_t port, struct ether_addr *mac_addr,
2218 * Remove a MAC address from the internal array of addresses.
2221 * The port identifier of the Ethernet device.
2223 * MAC address to remove.
2225 * - (0) if successful, or *mac_addr* didn't exist.
2226 * - (-ENOTSUP) if hardware doesn't support.
2227 * - (-ENODEV) if *port* invalid.
2228 * - (-EADDRINUSE) if attempting to remove the default MAC address
2230 int rte_eth_dev_mac_addr_remove(uint8_t port, struct ether_addr *mac_addr);
2233 * Update Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
2236 * The port identifier of the Ethernet device.
2240 * - (0) if successful.
2241 * - (-ENOTSUP) if hardware doesn't support.
2242 * - (-EINVAL) if bad parameter.
2244 int rte_eth_dev_rss_reta_update(uint8_t port,
2245 struct rte_eth_rss_reta *reta_conf);
2248 * Query Redirection Table(RETA) of Receive Side Scaling of Ethernet device.
2251 * The port identifier of the Ethernet device.
2255 * - (0) if successful.
2256 * - (-ENOTSUP) if hardware doesn't support.
2257 * - (-EINVAL) if bad parameter.
2259 int rte_eth_dev_rss_reta_query(uint8_t port,
2260 struct rte_eth_rss_reta *reta_conf);
2266 #endif /* _RTE_ETHDEV_H_ */