1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2014 Intel Corporation. 3 4L3 Forwarding Sample Application 5================================ 6 7The L3 Forwarding application is a simple example of packet processing using 8DPDK to demonstrate usage of poll and event mode packet I/O mechanism. 9The application performs L3 forwarding. 10 11Overview 12-------- 13 14The application demonstrates the use of the hash, LPM and FIB libraries in DPDK 15to implement packet forwarding using poll or event mode PMDs for packet I/O. 16The initialization and run-time paths are very similar to those of the 17:doc:`l2_forward_real_virtual` and :doc:`l2_forward_event`. 18The main difference from the L2 Forwarding sample application is that optionally 19packet can be Rx/Tx from/to eventdev instead of port directly and forwarding 20decision is made based on information read from the input packet. 21 22Eventdev can optionally use S/W or H/W (if supported by platform) scheduler 23implementation for packet I/O based on run time parameters. 24 25The lookup method is hash-based, LPM-based or FIB-based 26and is selected at run time. 27When the selected lookup method is hash-based, 28a hash object is used to emulate the flow classification stage. 29The hash object is used in correlation with a flow table to map each input packet to its flow at runtime. 30 31The hash lookup key is represented by a DiffServ 5-tuple composed of the following fields read from the input packet: 32Source IP Address, Destination IP Address, Protocol, Source Port and Destination Port. 33The ID of the output interface for the input packet is read from the identified flow table entry. 34The set of flows used by the application is statically configured and loaded into the hash at initialization time. 35When the selected lookup method is LPM or FIB based, 36an LPM or FIB object is used to emulate the forwarding stage for IPv4 packets. 37The LPM or FIB object is used as the routing table 38to identify the next hop for each input packet at runtime. 39 40The LPM and FIB lookup keys are represented by the destination IP address field 41read from the input packet. 42The ID of the output interface for the input packet is the next hop 43returned by the LPM or FIB lookup. 44The set of LPM and FIB rules used by the application is statically configured 45and loaded into the LPM or FIB object at initialization time. 46 47In the sample application, hash-based and FIB-based forwarding supports 48both IPv4 and IPv6. 49LPM-based forwarding supports IPv4 only. 50 51Compiling the Application 52------------------------- 53 54To compile the sample application see :doc:`compiling`. 55 56The application is located in the ``l3fwd`` sub-directory. 57 58Running the Application 59----------------------- 60 61The application has a number of command line options:: 62 63 ./dpdk-l3fwd [EAL options] -- -p PORTMASK 64 [-P] 65 [--lookup LOOKUP_METHOD] 66 --config(port,queue,lcore)[,(port,queue,lcore)] 67 [--eth-dest=X,MM:MM:MM:MM:MM:MM] 68 [--max-pkt-len PKTLEN] 69 [--no-numa] 70 [--hash-entry-num] 71 [--ipv6] 72 [--parse-ptype] 73 [--per-port-pool] 74 [--mode] 75 [--eventq-sched] 76 [--event-eth-rxqs] 77 [--event-vector [--event-vector-size SIZE] [--event-vector-tmo NS]] 78 [-E] 79 [-L] 80 81Where, 82 83* ``-p PORTMASK:`` Hexadecimal bitmask of ports to configure 84 85* ``-P:`` Optional, sets all ports to promiscuous mode so that packets are accepted regardless of the packet's Ethernet MAC destination address. 86 Without this option, only packets with the Ethernet MAC destination address set to the Ethernet address of the port are accepted. 87 88* ``--lookup:`` Optional, select the lookup method. 89 Accepted options: 90 ``em`` (Exact Match), 91 ``lpm`` (Longest Prefix Match), 92 ``fib`` (Forwarding Information Base). 93 Default is ``lpm``. 94 95* ``--config (port,queue,lcore)[,(port,queue,lcore)]:`` Determines which queues from which ports are mapped to which cores. 96 97* ``--eth-dest=X,MM:MM:MM:MM:MM:MM:`` Optional, ethernet destination for port X. 98 99* ``--max-pkt-len:`` Optional, maximum packet length in decimal (64-9600). 100 101* ``--no-numa:`` Optional, disables numa awareness. 102 103* ``--hash-entry-num:`` Optional, specifies the hash entry number in hexadecimal to be setup. 104 105* ``--ipv6:`` Optional, set if running ipv6 packets. 106 107* ``--parse-ptype:`` Optional, set to use software to analyze packet type. Without this option, hardware will check the packet type. 108 109* ``--per-port-pool:`` Optional, set to use independent buffer pools per port. Without this option, single buffer pool is used for all ports. 110 111* ``--mode:`` Optional, Packet transfer mode for I/O, poll or eventdev. 112 113* ``--eventq-sched:`` Optional, Event queue synchronization method, Ordered, Atomic or Parallel. Only valid if --mode=eventdev. 114 115* ``--event-eth-rxqs:`` Optional, Number of ethernet RX queues per device. Only valid if --mode=eventdev. 116 117* ``--event-vector:`` Optional, Enable event vectorization. Only valid if --mode=eventdev. 118 119* ``--event-vector-size:`` Optional, Max vector size if event vectorization is enabled. 120 121* ``--event-vector-tmo:`` Optional, Max timeout to form vector in nanoseconds if event vectorization is enabled. 122 123* ``-E:`` Optional, enable exact match, 124 legacy flag, please use ``--lookup=em`` instead. 125 126* ``-L:`` Optional, enable longest prefix match, 127 legacy flag, please use ``--lookup=lpm`` instead. 128 129 130For example, consider a dual processor socket platform with 8 physical cores, where cores 0-7 and 16-23 appear on socket 0, 131while cores 8-15 and 24-31 appear on socket 1. 132 133To enable L3 forwarding between two ports, assuming that both ports are in the same socket, using two cores, cores 1 and 2, 134(which are in the same socket too), use the following command: 135 136.. code-block:: console 137 138 ./<build_dir>/examples/dpdk-l3fwd -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" 139 140In this command: 141 142* The -l option enables cores 1, 2 143 144* The -p option enables ports 0 and 1 145 146* The --config option enables one queue on each port and maps each (port,queue) pair to a specific core. 147 The following table shows the mapping in this example: 148 149+----------+-----------+-----------+-------------------------------------+ 150| **Port** | **Queue** | **lcore** | **Description** | 151| | | | | 152+----------+-----------+-----------+-------------------------------------+ 153| 0 | 0 | 1 | Map queue 0 from port 0 to lcore 1. | 154| | | | | 155+----------+-----------+-----------+-------------------------------------+ 156| 1 | 0 | 2 | Map queue 0 from port 1 to lcore 2. | 157| | | | | 158+----------+-----------+-----------+-------------------------------------+ 159 160To use eventdev mode with sync method **ordered** on above mentioned environment, 161Following is the sample command: 162 163.. code-block:: console 164 165 ./<build_dir>/examples/dpdk-l3fwd -l 0-3 -n 4 -a <event device> -- -p 0x3 --eventq-sched=ordered 166 167or 168 169.. code-block:: console 170 171 ./<build_dir>/examples/dpdk-l3fwd -l 0-3 -n 4 -a <event device> \ 172 -- -p 0x03 --mode=eventdev --eventq-sched=ordered 173 174In this command: 175 176* -a option allows the event device supported by platform. 177 The syntax used to indicate this device may vary based on platform. 178 179* The --mode option defines PMD to be used for packet I/O. 180 181* The --eventq-sched option enables synchronization menthod of event queue so that packets will be scheduled accordingly. 182 183If application uses S/W scheduler, it uses following DPDK services: 184 185* Software scheduler 186* Rx adapter service function 187* Tx adapter service function 188 189Application needs service cores to run above mentioned services. Service cores 190must be provided as EAL parameters along with the --vdev=event_sw0 to enable S/W 191scheduler. Following is the sample command: 192 193.. code-block:: console 194 195 ./<build_dir>/examples/dpdk-l3fwd -l 0-7 -s 0xf0000 -n 4 --vdev event_sw0 -- -p 0x3 --mode=eventdev --eventq-sched=ordered 196 197In case of eventdev mode, *--config* option is not used for ethernet port 198configuration. Instead each ethernet port will be configured with mentioned 199setup: 200 201* Single Rx/Tx queue 202 203* Each Rx queue will be connected to event queue via Rx adapter. 204 205* Each Tx queue will be connected via Tx adapter. 206 207Refer to the *DPDK Getting Started Guide* for general information on running applications and 208the Environment Abstraction Layer (EAL) options. 209 210.. _l3_fwd_explanation: 211 212Explanation 213----------- 214 215The following sections provide some explanation of the sample application code. As mentioned in the overview section, 216the initialization and run-time paths are very similar to those of the :doc:`l2_forward_real_virtual` and :doc:`l2_forward_event`. 217The following sections describe aspects that are specific to the L3 Forwarding sample application. 218 219Hash Initialization 220~~~~~~~~~~~~~~~~~~~ 221 222The hash object is created and loaded with the pre-configured entries read from a global array, 223and then generate the expected 5-tuple as key to keep consistence with those of real flow 224for the convenience to execute hash performance test on 4M/8M/16M flows. 225 226.. note:: 227 228 The Hash initialization will setup both ipv4 and ipv6 hash table, 229 and populate the either table depending on the value of variable ipv6. 230 To support the hash performance test with up to 8M single direction flows/16M bi-direction flows, 231 populate_ipv4_many_flow_into_table() function will populate the hash table with specified hash table entry number(default 4M). 232 233.. note:: 234 235 Value of global variable ipv6 can be specified with --ipv6 in the command line. 236 Value of global variable hash_entry_number, 237 which is used to specify the total hash entry number for all used ports in hash performance test, 238 can be specified with --hash-entry-num VALUE in command line, being its default value 4. 239 240.. code-block:: c 241 242 #if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) 243 244 static void 245 setup_hash(int socketid) 246 { 247 // ... 248 249 if (hash_entry_number != HASH_ENTRY_NUMBER_DEFAULT) { 250 if (ipv6 == 0) { 251 /* populate the ipv4 hash */ 252 populate_ipv4_many_flow_into_table(ipv4_l3fwd_lookup_struct[socketid], hash_entry_number); 253 } else { 254 /* populate the ipv6 hash */ 255 populate_ipv6_many_flow_into_table( ipv6_l3fwd_lookup_struct[socketid], hash_entry_number); 256 } 257 } else 258 if (ipv6 == 0) { 259 /* populate the ipv4 hash */ 260 populate_ipv4_few_flow_into_table(ipv4_l3fwd_lookup_struct[socketid]); 261 } else { 262 /* populate the ipv6 hash */ 263 populate_ipv6_few_flow_into_table(ipv6_l3fwd_lookup_struct[socketid]); 264 } 265 } 266 } 267 #endif 268 269LPM Initialization 270~~~~~~~~~~~~~~~~~~ 271 272The LPM object is created and loaded with the pre-configured entries read from a global array. 273 274.. literalinclude:: ../../../examples/l3fwd/l3fwd_em.c 275 :language: c 276 :start-after: Initialize exact match (hash) parameters. 8< 277 :end-before: >8 End of initialization of hash parameters. 278 279FIB Initialization 280~~~~~~~~~~~~~~~~~~ 281 282The FIB object is created and loaded with the pre-configured entries 283read from a global array. 284The abridged code snippet below shows the FIB initialization for IPv4, 285the full setup function including the IPv6 setup can be seen in the app code. 286 287.. literalinclude:: ../../../examples/l3fwd/l3fwd_fib.c 288 :language: c 289 :start-after: Function to setup fib. 8< 290 :end-before: >8 End of setup fib. 291 292Packet Forwarding for Hash-based Lookups 293~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 294 295For each input packet, the packet forwarding operation is done by the l3fwd_simple_forward() 296or simple_ipv4_fwd_4pkts() function for IPv4 packets or the simple_ipv6_fwd_4pkts() function for IPv6 packets. 297The l3fwd_simple_forward() function provides the basic functionality for both IPv4 and IPv6 packet forwarding 298for any number of burst packets received, 299and the packet forwarding decision (that is, the identification of the output interface for the packet) 300for hash-based lookups is done by the get_ipv4_dst_port() or get_ipv6_dst_port() function. 301The get_ipv4_dst_port() function is shown below: 302 303.. literalinclude:: ../../../examples/l3fwd/l3fwd_em.c 304 :language: c 305 :start-after: Performing hash-based lookups. 8< 306 :end-before: >8 End of performing hash-based lookups. 307 308The get_ipv6_dst_port() function is similar to the get_ipv4_dst_port() function. 309 310The simple_ipv4_fwd_4pkts() and simple_ipv6_fwd_4pkts() function are optimized for continuous 4 valid ipv4 and ipv6 packets, 311they leverage the multiple buffer optimization to boost the performance of forwarding packets with the exact match on hash table. 312The key code snippet of simple_ipv4_fwd_4pkts() is shown below: 313 314.. code-block:: c 315 316 static inline void 317 simple_ipv4_fwd_4pkts(struct rte_mbuf* m[4], uint16_t portid, struct lcore_conf *qconf) 318 { 319 // ... 320 321 data[0] = _mm_loadu_si128(( m128i*)(rte_pktmbuf_mtod(m[0], unsigned char *) + sizeof(struct rte_ether_hdr) + offsetof(struct rte_ipv4_hdr, time_to_live))); 322 data[1] = _mm_loadu_si128(( m128i*)(rte_pktmbuf_mtod(m[1], unsigned char *) + sizeof(struct rte_ether_hdr) + offsetof(struct rte_ipv4_hdr, time_to_live))); 323 data[2] = _mm_loadu_si128(( m128i*)(rte_pktmbuf_mtod(m[2], unsigned char *) + sizeof(struct rte_ether_hdr) + offsetof(struct rte_ipv4_hdr, time_to_live))); 324 data[3] = _mm_loadu_si128(( m128i*)(rte_pktmbuf_mtod(m[3], unsigned char *) + sizeof(struct rte_ether_hdr) + offsetof(struct rte_ipv4_hdr, time_to_live))); 325 326 key[0].xmm = _mm_and_si128(data[0], mask0); 327 key[1].xmm = _mm_and_si128(data[1], mask0); 328 key[2].xmm = _mm_and_si128(data[2], mask0); 329 key[3].xmm = _mm_and_si128(data[3], mask0); 330 331 const void *key_array[4] = {&key[0], &key[1], &key[2],&key[3]}; 332 333 rte_hash_lookup_bulk(qconf->ipv4_lookup_struct, &key_array[0], 4, ret); 334 335 dst_port[0] = (ret[0] < 0)? portid:ipv4_l3fwd_out_if[ret[0]]; 336 dst_port[1] = (ret[1] < 0)? portid:ipv4_l3fwd_out_if[ret[1]]; 337 dst_port[2] = (ret[2] < 0)? portid:ipv4_l3fwd_out_if[ret[2]]; 338 dst_port[3] = (ret[3] < 0)? portid:ipv4_l3fwd_out_if[ret[3]]; 339 340 // ... 341 } 342 343The simple_ipv6_fwd_4pkts() function is similar to the simple_ipv4_fwd_4pkts() function. 344 345Known issue: IP packets with extensions or IP packets which are not TCP/UDP cannot work well at this mode. 346 347Packet Forwarding for LPM-based Lookups 348~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 349 350For each input packet, the packet forwarding operation is done by the l3fwd_simple_forward() function, 351but the packet forwarding decision (that is, the identification of the output interface for the packet) 352for LPM-based lookups is done by the get_ipv4_dst_port() function below: 353 354.. literalinclude:: ../../../examples/l3fwd/l3fwd_lpm.c 355 :language: c 356 :start-after: Performing LPM-based lookups. 8< 357 :end-before: >8 End of performing LPM-based lookups. 358 359Packet Forwarding for FIB-based Lookups 360~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 361 362The FIB library was designed to process multiple packets at once, 363it does not have separate functions for single and bulk lookups. 364``rte_fib_lookup_bulk`` is used for IPv4 lookups 365and ``rte_fib6_lookup_bulk`` for IPv6. 366Various examples of these functions being used 367can be found in the sample app code. 368 369Eventdev Driver Initialization 370~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 371Eventdev driver initialization is same as L2 forwarding eventdev application. 372Refer :doc:`l2_forward_event` for more details. 373