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 VMDQ and DCB Forwarding Sample Application
32 ==========================================
34 The VMDQ and DCB Forwarding sample application is a simple example of packet processing using the DPDK.
35 The application performs L2 forwarding using VMDQ and DCB to divide the incoming traffic into queues.
36 The traffic splitting is performed in hardware by the VMDQ and DCB features of the Intel® 82599 and X710/XL710 Ethernet Controllers.
41 This sample application can be used as a starting point for developing a new application that is based on the DPDK and
42 uses VMDQ and DCB for traffic partitioning.
44 The VMDQ and DCB filters work on MAC and VLAN traffic to divide the traffic into input queues on the basis of the Destination MAC
45 address, VLAN ID and VLAN user priority fields.
46 VMDQ filters split the traffic into 16 or 32 groups based on the Destination MAC and VLAN ID.
47 Then, DCB places each packet into one of queues within that group, based upon the VLAN user priority field.
49 All traffic is read from a single incoming port (port 0) and output on port 1, without any processing being performed.
50 With Intel® 82599 NIC, for example, the traffic is split into 128 queues on input, where each thread of the application reads from
51 multiple queues. When run with 8 threads, that is, with the -c FF option, each thread receives and forwards packets from 16 queues.
53 As supplied, the sample application configures the VMDQ feature to have 32 pools with 4 queues each as indicated in :numref:`figure_vmdq_dcb_example`.
54 The Intel® 82599 10 Gigabit Ethernet Controller NIC also supports the splitting of traffic into 16 pools of 8 queues. While the
55 Intel® X710 or XL710 Ethernet Controller NICs support many configurations of VMDQ pools of 4 or 8 queues each. For simplicity, only 16
56 or 32 pools is supported in this sample. And queues numbers for each VMDQ pool can be changed by setting CONFIG_RTE_LIBRTE_I40E_QUEUE_NUM_PER_VM
57 in config/common_* file.
58 The nb-pools, nb-tcs and enable-rss parameters can be passed on the command line, after the EAL parameters:
60 .. code-block:: console
62 ./build/vmdq_dcb [EAL options] -- -p PORTMASK --nb-pools NP --nb-tcs TC --enable-rss
64 where, NP can be 16 or 32, TC can be 4 or 8, rss is disabled by default.
66 .. _figure_vmdq_dcb_example:
68 .. figure:: img/vmdq_dcb_example.*
70 Packet Flow Through the VMDQ and DCB Sample Application
73 In Linux* user space, the application can display statistics with the number of packets received on each queue.
74 To have the application display the statistics, send a SIGHUP signal to the running application process.
76 The VMDQ and DCB Forwarding sample application is in many ways simpler than the L2 Forwarding application
77 (see :doc:`l2_forward_real_virtual`)
78 as it performs unidirectional L2 forwarding of packets from one port to a second port.
79 No command-line options are taken by this application apart from the standard EAL command-line options.
83 Since VMD queues are being used for VMM, this application works correctly
84 when VTd is disabled in the BIOS or Linux* kernel (intel_iommu=off).
86 Compiling the Application
87 -------------------------
89 #. Go to the examples directory:
91 .. code-block:: console
93 export RTE_SDK=/path/to/rte_sdk
94 cd ${RTE_SDK}/examples/vmdq_dcb
96 #. Set the target (a default target is used if not specified). For example:
98 .. code-block:: console
100 export RTE_TARGET=x86_64-native-linuxapp-gcc
102 See the *DPDK Getting Started Guide* for possible RTE_TARGET values.
104 #. Build the application:
106 .. code-block:: console
110 Running the Application
111 -----------------------
113 To run the example in a linuxapp environment:
115 .. code-block:: console
117 user@target:~$ ./build/vmdq_dcb -l 0-3 -n 4 -- -p 0x3 --nb-pools 32 --nb-tcs 4
119 Refer to the *DPDK Getting Started Guide* for general information on running applications and
120 the Environment Abstraction Layer (EAL) options.
125 The following sections provide some explanation of the code.
130 The EAL, driver and PCI configuration is performed largely as in the L2 Forwarding sample application,
131 as is the creation of the mbuf pool.
132 See :doc:`l2_forward_real_virtual`.
133 Where this example application differs is in the configuration of the NIC port for RX.
135 The VMDQ and DCB hardware feature is configured at port initialization time by setting the appropriate values in the
136 rte_eth_conf structure passed to the rte_eth_dev_configure() API.
137 Initially in the application,
138 a default structure is provided for VMDQ and DCB configuration to be filled in later by the application.
142 /* empty vmdq+dcb configuration structure. Filled in programmatically */
143 static const struct rte_eth_conf vmdq_dcb_conf_default = {
145 .mq_mode = ETH_MQ_RX_VMDQ_DCB,
147 .header_split = 0, /**< Header Split disabled */
148 .hw_ip_checksum = 0, /**< IP checksum offload disabled */
149 .hw_vlan_filter = 0, /**< VLAN filtering disabled */
150 .jumbo_frame = 0, /**< Jumbo Frame Support disabled */
153 .mq_mode = ETH_MQ_TX_VMDQ_DCB,
156 * should be overridden separately in code with
161 .nb_queue_pools = ETH_32_POOLS,
162 .enable_default_pool = 0,
165 .pool_map = {{0, 0},},
170 /** Traffic class each UP mapped to. */
174 .nb_queue_pools = ETH_32_POOLS,
175 .enable_default_pool = 0,
178 .pool_map = {{0, 0},},
182 .vmdq_dcb_tx_conf = {
183 .nb_queue_pools = ETH_32_POOLS,
189 The get_eth_conf() function fills in an rte_eth_conf structure with the appropriate values,
190 based on the global vlan_tags array,
191 and dividing up the possible user priority values equally among the individual queues
192 (also referred to as traffic classes) within each pool. With Intel® 82599 NIC,
193 if the number of pools is 32, then the user priority fields are allocated 2 to a queue.
194 If 16 pools are used, then each of the 8 user priority fields is allocated to its own queue within the pool.
195 With Intel® X710/XL710 NICs, if number of tcs is 4, and number of queues in pool is 8,
196 then the user priority fields are allocated 2 to one tc, and a tc has 2 queues mapping to it, then
197 RSS will determine the destination queue in 2.
198 For the VLAN IDs, each one can be allocated to possibly multiple pools of queues,
199 so the pools parameter in the rte_eth_vmdq_dcb_conf structure is specified as a bitmask value.
200 For destination MAC, each VMDQ pool will be assigned with a MAC address. In this sample, each VMDQ pool
201 is assigned to the MAC like 52:54:00:12:<port_id>:<pool_id>, that is,
202 the MAC of VMDQ pool 2 on port 1 is 52:54:00:12:01:02.
206 const uint16_t vlan_tags[] = {
207 0, 1, 2, 3, 4, 5, 6, 7,
208 8, 9, 10, 11, 12, 13, 14, 15,
209 16, 17, 18, 19, 20, 21, 22, 23,
210 24, 25, 26, 27, 28, 29, 30, 31
213 /* pool mac addr template, pool mac addr is like: 52 54 00 12 port# pool# */
214 static struct ether_addr pool_addr_template = {
215 .addr_bytes = {0x52, 0x54, 0x00, 0x12, 0x00, 0x00}
218 /* Builds up the correct configuration for vmdq+dcb based on the vlan tags array
219 * given above, and the number of traffic classes available for use. */
221 get_eth_conf(struct rte_eth_conf *eth_conf)
223 struct rte_eth_vmdq_dcb_conf conf;
224 struct rte_eth_vmdq_rx_conf vmdq_conf;
225 struct rte_eth_dcb_rx_conf dcb_conf;
226 struct rte_eth_vmdq_dcb_tx_conf tx_conf;
229 conf.nb_queue_pools = (enum rte_eth_nb_pools)num_pools;
230 vmdq_conf.nb_queue_pools = (enum rte_eth_nb_pools)num_pools;
231 tx_conf.nb_queue_pools = (enum rte_eth_nb_pools)num_pools;
232 conf.nb_pool_maps = num_pools;
233 vmdq_conf.nb_pool_maps = num_pools;
234 conf.enable_default_pool = 0;
235 vmdq_conf.enable_default_pool = 0;
236 conf.default_pool = 0; /* set explicit value, even if not used */
237 vmdq_conf.default_pool = 0;
239 for (i = 0; i < conf.nb_pool_maps; i++) {
240 conf.pool_map[i].vlan_id = vlan_tags[i];
241 vmdq_conf.pool_map[i].vlan_id = vlan_tags[i];
242 conf.pool_map[i].pools = 1UL << i ;
243 vmdq_conf.pool_map[i].pools = 1UL << i;
245 for (i = 0; i < ETH_DCB_NUM_USER_PRIORITIES; i++){
246 conf.dcb_tc[i] = i % num_tcs;
247 dcb_conf.dcb_tc[i] = i % num_tcs;
248 tx_conf.dcb_tc[i] = i % num_tcs;
250 dcb_conf.nb_tcs = (enum rte_eth_nb_tcs)num_tcs;
251 (void)(rte_memcpy(eth_conf, &vmdq_dcb_conf_default, sizeof(*eth_conf)));
252 (void)(rte_memcpy(ð_conf->rx_adv_conf.vmdq_dcb_conf, &conf,
254 (void)(rte_memcpy(ð_conf->rx_adv_conf.dcb_rx_conf, &dcb_conf,
256 (void)(rte_memcpy(ð_conf->rx_adv_conf.vmdq_rx_conf, &vmdq_conf,
258 (void)(rte_memcpy(ð_conf->tx_adv_conf.vmdq_dcb_tx_conf, &tx_conf,
261 eth_conf->rxmode.mq_mode= ETH_MQ_RX_VMDQ_DCB_RSS;
262 eth_conf->rx_adv_conf.rss_conf.rss_hf = ETH_RSS_IP |
272 /* Set mac for each pool.*/
273 for (q = 0; q < num_pools; q++) {
274 struct ether_addr mac;
275 mac = pool_addr_template;
276 mac.addr_bytes[4] = port;
277 mac.addr_bytes[5] = q;
278 printf("Port %u vmdq pool %u set mac %02x:%02x:%02x:%02x:%02x:%02x\n",
280 mac.addr_bytes[0], mac.addr_bytes[1],
281 mac.addr_bytes[2], mac.addr_bytes[3],
282 mac.addr_bytes[4], mac.addr_bytes[5]);
283 retval = rte_eth_dev_mac_addr_add(port, &mac,
286 printf("mac addr add failed at pool %d\n", q);
291 Once the network port has been initialized using the correct VMDQ and DCB values,
292 the initialization of the port's RX and TX hardware rings is performed similarly to that
293 in the L2 Forwarding sample application.
294 See :doc:`l2_forward_real_virtual` for more information.
299 When run in a linuxapp environment,
300 the VMDQ and DCB Forwarding sample application can display statistics showing the number of packets read from each RX queue.
301 This is provided by way of a signal handler for the SIGHUP signal,
302 which simply prints to standard output the packet counts in grid form.
303 Each row of the output is a single pool with the columns being the queue number within that pool.
305 To generate the statistics output, use the following command:
307 .. code-block:: console
309 user@host$ sudo killall -HUP vmdq_dcb_app
311 Please note that the statistics output will appear on the terminal where the vmdq_dcb_app is running,
312 rather than the terminal from which the HUP signal was sent.