2 Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31 Link Status Interrupt Sample Application
32 ========================================
34 The Link Status Interrupt sample application is a simple example of packet processing using
35 the Data Plane Development Kit (DPDK) that
36 demonstrates how network link status changes for a network port can be captured and
37 used by a DPDK application.
42 The Link Status Interrupt sample application registers a user space callback for the link status interrupt of each port
43 and performs L2 forwarding for each packet that is received on an RX_PORT.
44 The following operations are performed:
46 * RX_PORT and TX_PORT are paired with available ports one-by-one according to the core mask
48 * The source MAC address is replaced by the TX_PORT MAC address
50 * The destination MAC address is replaced by 02:00:00:00:00:TX_PORT_ID
52 This application can be used to demonstrate the usage of link status interrupt and its user space callbacks
53 and the behavior of L2 forwarding each time the link status changes.
55 Compiling the Application
56 -------------------------
58 #. Go to the example directory:
60 .. code-block:: console
62 export RTE_SDK=/path/to/rte_sdk
63 cd ${RTE_SDK}/examples/link_status_interrupt
65 #. Set the target (a default target is used if not specified). For example:
67 .. code-block:: console
69 export RTE_TARGET=x86_64-native-linuxapp-gcc
71 See the *DPDK Getting Started Guide* for possible RTE_TARGET values.
73 #. Build the application:
75 .. code-block:: console
81 The compiled application is written to the build subdirectory.
82 To have the application written to a different location,
83 the O=/path/to/build/directory option may be specified on the make command line.
85 Running the Application
86 -----------------------
88 The application requires a number of command line options:
90 .. code-block:: console
92 ./build/link_status_interrupt [EAL options] -- -p PORTMASK [-q NQ][-T PERIOD]
96 * -p PORTMASK: A hexadecimal bitmask of the ports to configure
98 * -q NQ: A number of queues (=ports) per lcore (default is 1)
100 * -T PERIOD: statistics will be refreshed each PERIOD seconds (0 to disable, 10 default)
102 To run the application in a linuxapp environment with 4 lcores, 4 memory channels, 16 ports and 8 RX queues per lcore,
105 .. code-block:: console
107 $ ./build/link_status_interrupt -l 0-3 -n 4-- -q 8 -p ffff
109 Refer to the *DPDK Getting Started Guide* for general information on running applications
110 and the Environment Abstraction Layer (EAL) options.
115 The following sections provide some explanation of the code.
117 Command Line Arguments
118 ~~~~~~~~~~~~~~~~~~~~~~
120 The Link Status Interrupt sample application takes specific parameters,
121 in addition to Environment Abstraction Layer (EAL) arguments (see Section `Running the Application`_).
123 Command line parsing is done in the same way as it is done in the L2 Forwarding Sample Application.
124 See :ref:`l2_fwd_app_cmd_arguments` for more information.
126 Mbuf Pool Initialization
127 ~~~~~~~~~~~~~~~~~~~~~~~~
129 Mbuf pool initialization is done in the same way as it is done in the L2 Forwarding Sample Application.
130 See :ref:`l2_fwd_app_mbuf_init` for more information.
132 Driver Initialization
133 ~~~~~~~~~~~~~~~~~~~~~
135 The main part of the code in the main() function relates to the initialization of the driver.
136 To fully understand this code, it is recommended to study the chapters that related to the Poll Mode Driver in the
137 *DPDK Programmer's Guide and the DPDK API Reference*.
141 if (rte_eal_pci_probe() < 0)
142 rte_exit(EXIT_FAILURE, "Cannot probe PCI\n");
144 nb_ports = rte_eth_dev_count();
146 rte_exit(EXIT_FAILURE, "No Ethernet ports - bye\n");
149 * Each logical core is assigned a dedicated TX queue on each port.
152 for (portid = 0; portid < nb_ports; portid++) {
153 /* skip ports that are not enabled */
155 if ((lsi_enabled_port_mask & (1 << portid)) == 0)
158 /* save the destination port id */
160 if (nb_ports_in_mask % 2) {
161 lsi_dst_ports[portid] = portid_last;
162 lsi_dst_ports[portid_last] = portid;
165 portid_last = portid;
169 rte_eth_dev_info_get((uint8_t) portid, &dev_info);
174 * rte_eal_pci_probe() parses the devices on the PCI bus and initializes recognized devices.
176 The next step is to configure the RX and TX queues.
177 For each port, there is only one RX queue (only one lcore is able to poll a given port).
178 The number of TX queues depends on the number of available lcores.
179 The rte_eth_dev_configure() function is used to configure the number of queues for a port:
183 ret = rte_eth_dev_configure((uint8_t) portid, 1, 1, &port_conf);
185 rte_exit(EXIT_FAILURE, "Cannot configure device: err=%d, port=%u\n", ret, portid);
187 The global configuration is stored in a static structure:
191 static const struct rte_eth_conf port_conf = {
194 .header_split = 0, /**< Header Split disabled */
195 .hw_ip_checksum = 0, /**< IP checksum offload disabled */
196 .hw_vlan_filter = 0, /**< VLAN filtering disabled */
197 .hw_strip_crc= 0, /**< CRC stripped by hardware */
201 .lsc = 1, /**< link status interrupt feature enabled */
205 Configuring lsc to 0 (the default) disables the generation of any link status change interrupts in kernel space
206 and no user space interrupt event is received.
207 The public interface rte_eth_link_get() accesses the NIC registers directly to update the link status.
208 Configuring lsc to non-zero enables the generation of link status change interrupts in kernel space
209 when a link status change is present and calls the user space callbacks registered by the application.
210 The public interface rte_eth_link_get() just reads the link status in a global structure
211 that would be updated in the interrupt host thread only.
213 Interrupt Callback Registration
214 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
216 The application can register one or more callbacks to a specific port and interrupt event.
217 An example callback function that has been written as indicated below.
222 lsi_event_callback(uint8_t port_id, enum rte_eth_event_type type, void *param)
224 struct rte_eth_link link;
228 printf("\n\nIn registered callback...\n");
230 printf("Event type: %s\n", type == RTE_ETH_EVENT_INTR_LSC ? "LSC interrupt" : "unknown event");
232 rte_eth_link_get_nowait(port_id, &link);
234 if (link.link_status) {
235 printf("Port %d Link Up - speed %u Mbps - %s\n\n", port_id, (unsigned)link.link_speed,
236 (link.link_duplex == ETH_LINK_FULL_DUPLEX) ? ("full-duplex") : ("half-duplex"));
238 printf("Port %d Link Down\n\n", port_id);
241 This function is called when a link status interrupt is present for the right port.
242 The port_id indicates which port the interrupt applies to.
243 The type parameter identifies the interrupt event type,
244 which currently can be RTE_ETH_EVENT_INTR_LSC only, but other types can be added in the future.
245 The param parameter is the address of the parameter for the callback.
246 This function should be implemented with care since it will be called in the interrupt host thread,
247 which is different from the main thread of its caller.
249 The application registers the lsi_event_callback and a NULL parameter to the link status interrupt event on each port:
253 rte_eth_dev_callback_register((uint8_t)portid, RTE_ETH_EVENT_INTR_LSC, lsi_event_callback, NULL);
255 This registration can be done only after calling the rte_eth_dev_configure() function and before calling any other function.
256 If lsc is initialized with 0, the callback is never called since no interrupt event would ever be present.
258 RX Queue Initialization
259 ~~~~~~~~~~~~~~~~~~~~~~~
261 The application uses one lcore to poll one or several ports, depending on the -q option,
262 which specifies the number of queues per lcore.
264 For example, if the user specifies -q 4, the application is able to poll four ports with one lcore.
265 If there are 16 ports on the target (and if the portmask argument is -p ffff),
266 the application will need four lcores to poll all the ports.
270 ret = rte_eth_rx_queue_setup((uint8_t) portid, 0, nb_rxd, SOCKET0, &rx_conf, lsi_pktmbuf_pool);
272 rte_exit(EXIT_FAILURE, "rte_eth_rx_queue_setup: err=%d, port=%u\n", ret, portid);
274 The list of queues that must be polled for a given lcore is stored in a private structure called struct lcore_queue_conf.
278 struct lcore_queue_conf {
280 unsigned rx_port_list[MAX_RX_QUEUE_PER_LCORE]; unsigned tx_queue_id;
281 struct mbuf_table tx_mbufs[LSI_MAX_PORTS];
284 struct lcore_queue_conf lcore_queue_conf[RTE_MAX_LCORE];
286 The n_rx_port and rx_port_list[] fields are used in the main packet processing loop
287 (see `Receive, Process and Transmit Packets`_).
289 The global configuration for the RX queues is stored in a static structure:
293 static const struct rte_eth_rxconf rx_conf = {
295 .pthresh = RX_PTHRESH,
296 .hthresh = RX_HTHRESH,
297 .wthresh = RX_WTHRESH,
301 TX Queue Initialization
302 ~~~~~~~~~~~~~~~~~~~~~~~
304 Each lcore should be able to transmit on any port.
305 For every port, a single TX queue is initialized.
309 /* init one TX queue logical core on each port */
313 ret = rte_eth_tx_queue_setup(portid, 0, nb_txd, rte_eth_dev_socket_id(portid), &tx_conf);
315 rte_exit(EXIT_FAILURE, "rte_eth_tx_queue_setup: err=%d,port=%u\n", ret, (unsigned) portid);
317 The global configuration for TX queues is stored in a static structure:
321 static const struct rte_eth_txconf tx_conf = {
323 .pthresh = TX_PTHRESH,
324 .hthresh = TX_HTHRESH,
325 .wthresh = TX_WTHRESH,
327 .tx_free_thresh = RTE_TEST_TX_DESC_DEFAULT + 1, /* disable feature */
330 Receive, Process and Transmit Packets
331 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
333 In the lsi_main_loop() function, the main task is to read ingress packets from the RX queues.
334 This is done using the following code:
339 * Read packet from RX queues
342 for (i = 0; i < qconf->n_rx_port; i++) {
343 portid = qconf->rx_port_list[i];
344 nb_rx = rte_eth_rx_burst((uint8_t) portid, 0, pkts_burst, MAX_PKT_BURST);
345 port_statistics[portid].rx += nb_rx;
347 for (j = 0; j < nb_rx; j++) {
349 rte_prefetch0(rte_pktmbuf_mtod(m, void *));
350 lsi_simple_forward(m, portid);
354 Packets are read in a burst of size MAX_PKT_BURST.
355 The rte_eth_rx_burst() function writes the mbuf pointers in a local table and returns the number of available mbufs in the table.
357 Then, each mbuf in the table is processed by the lsi_simple_forward() function.
358 The processing is very simple: processes the TX port from the RX port and then replaces the source and destination MAC addresses.
362 In the following code, the two lines for calculating the output port require some explanation.
363 If portId is even, the first line does nothing (as portid & 1 will be 0), and the second line adds 1.
364 If portId is odd, the first line subtracts one and the second line does nothing.
365 Therefore, 0 goes to 1, and 1 to 0, 2 goes to 3 and 3 to 2, and so on.
370 lsi_simple_forward(struct rte_mbuf *m, unsigned portid)
372 struct ether_hdr *eth;
374 unsigned dst_port = lsi_dst_ports[portid];
376 eth = rte_pktmbuf_mtod(m, struct ether_hdr *);
378 /* 02:00:00:00:00:xx */
380 tmp = ð->d_addr.addr_bytes[0];
382 *((uint64_t *)tmp) = 0x000000000002 + (dst_port << 40);
385 ether_addr_copy(&lsi_ports_eth_addr[dst_port], ð->s_addr);
387 lsi_send_packet(m, dst_port);
390 Then, the packet is sent using the lsi_send_packet(m, dst_port) function.
391 For this test application, the processing is exactly the same for all packets arriving on the same RX port.
392 Therefore, it would have been possible to call the lsi_send_burst() function directly from the main loop
393 to send all the received packets on the same TX port using
394 the burst-oriented send function, which is more efficient.
396 However, in real-life applications (such as, L3 routing),
397 packet N is not necessarily forwarded on the same port as packet N-1.
398 The application is implemented to illustrate that so the same approach can be reused in a more complex application.
400 The lsi_send_packet() function stores the packet in a per-lcore and per-txport table.
401 If the table is full, the whole packets table is transmitted using the lsi_send_burst() function:
405 /* Send the packet on an output interface */
408 lsi_send_packet(struct rte_mbuf *m, uint8_t port)
410 unsigned lcore_id, len;
411 struct lcore_queue_conf *qconf;
413 lcore_id = rte_lcore_id();
414 qconf = &lcore_queue_conf[lcore_id];
415 len = qconf->tx_mbufs[port].len;
416 qconf->tx_mbufs[port].m_table[len] = m;
419 /* enough pkts to be sent */
421 if (unlikely(len == MAX_PKT_BURST)) {
422 lsi_send_burst(qconf, MAX_PKT_BURST, port);
425 qconf->tx_mbufs[port].len = len;
430 To ensure that no packets remain in the tables, each lcore does a draining of the TX queue in its main loop.
431 This technique introduces some latency when there are not many packets to send.
432 However, it improves performance:
436 cur_tsc = rte_rdtsc();
439 * TX burst queue drain
442 diff_tsc = cur_tsc - prev_tsc;
444 if (unlikely(diff_tsc > drain_tsc)) {
445 /* this could be optimized (use queueid instead of * portid), but it is not called so often */
447 for (portid = 0; portid < RTE_MAX_ETHPORTS; portid++) {
448 if (qconf->tx_mbufs[portid].len == 0)
451 lsi_send_burst(&lcore_queue_conf[lcore_id],
452 qconf->tx_mbufs[portid].len, (uint8_t) portid);
453 qconf->tx_mbufs[portid].len = 0;
456 /* if timer is enabled */
458 if (timer_period > 0) {
459 /* advance the timer */
461 timer_tsc += diff_tsc;
463 /* if timer has reached its timeout */
465 if (unlikely(timer_tsc >= (uint64_t) timer_period)) {
466 /* do this only on master core */
468 if (lcore_id == rte_get_master_lcore()) {
471 /* reset the timer */