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 the DPDK. 8The application performs L3 forwarding. 9 10Overview 11-------- 12 13The application demonstrates the use of the hash and LPM libraries in the DPDK to implement packet forwarding. 14The initialization and run-time paths are very similar to those of the :doc:`l2_forward_real_virtual`. 15The main difference from the L2 Forwarding sample application is that the forwarding decision 16is made based on information read from the input packet. 17 18The lookup method is either hash-based or LPM-based and is selected at run time. When the selected lookup method is hash-based, 19a hash object is used to emulate the flow classification stage. 20The hash object is used in correlation with a flow table to map each input packet to its flow at runtime. 21 22The hash lookup key is represented by a DiffServ 5-tuple composed of the following fields read from the input packet: 23Source IP Address, Destination IP Address, Protocol, Source Port and Destination Port. 24The ID of the output interface for the input packet is read from the identified flow table entry. 25The set of flows used by the application is statically configured and loaded into the hash at initialization time. 26When the selected lookup method is LPM based, an LPM object is used to emulate the forwarding stage for IPv4 packets. 27The LPM object is used as the routing table to identify the next hop for each input packet at runtime. 28 29The LPM lookup key is represented by the Destination IP Address field read from the input packet. 30The ID of the output interface for the input packet is the next hop returned by the LPM lookup. 31The set of LPM rules used by the application is statically configured and loaded into the LPM object at initialization time. 32 33In the sample application, hash-based forwarding supports IPv4 and IPv6. LPM-based forwarding supports IPv4 only. 34 35Compiling the Application 36------------------------- 37 38To compile the sample application see :doc:`compiling`. 39 40The application is located in the ``l3fwd`` sub-directory. 41 42Running the Application 43----------------------- 44 45The application has a number of command line options:: 46 47 ./l3fwd [EAL options] -- -p PORTMASK 48 [-P] 49 [-E] 50 [-L] 51 --config(port,queue,lcore)[,(port,queue,lcore)] 52 [--eth-dest=X,MM:MM:MM:MM:MM:MM] 53 [--enable-jumbo [--max-pkt-len PKTLEN]] 54 [--no-numa] 55 [--hash-entry-num] 56 [--ipv6] 57 [--parse-ptype] 58 [--per-port-pool] 59 60Where, 61 62* ``-p PORTMASK:`` Hexadecimal bitmask of ports to configure 63 64* ``-P:`` Optional, sets all ports to promiscuous mode so that packets are accepted regardless of the packet's Ethernet MAC destination address. 65 Without this option, only packets with the Ethernet MAC destination address set to the Ethernet address of the port are accepted. 66 67* ``-E:`` Optional, enable exact match. 68 69* ``-L:`` Optional, enable longest prefix match. 70 71* ``--config (port,queue,lcore)[,(port,queue,lcore)]:`` Determines which queues from which ports are mapped to which cores. 72 73* ``--eth-dest=X,MM:MM:MM:MM:MM:MM:`` Optional, ethernet destination for port X. 74 75* ``--enable-jumbo:`` Optional, enables jumbo frames. 76 77* ``--max-pkt-len:`` Optional, under the premise of enabling jumbo, maximum packet length in decimal (64-9600). 78 79* ``--no-numa:`` Optional, disables numa awareness. 80 81* ``--hash-entry-num:`` Optional, specifies the hash entry number in hexadecimal to be setup. 82 83* ``--ipv6:`` Optional, set if running ipv6 packets. 84 85* ``--parse-ptype:`` Optional, set to use software to analyze packet type. Without this option, hardware will check the packet type. 86 87* ``--per-port-pool:`` Optional, set to use independent buffer pools per port. Without this option, single buffer pool is used for all ports. 88 89For example, consider a dual processor socket platform with 8 physical cores, where cores 0-7 and 16-23 appear on socket 0, 90while cores 8-15 and 24-31 appear on socket 1. 91 92To enable L3 forwarding between two ports, assuming that both ports are in the same socket, using two cores, cores 1 and 2, 93(which are in the same socket too), use the following command: 94 95.. code-block:: console 96 97 ./build/l3fwd -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" 98 99In this command: 100 101* The -l option enables cores 1, 2 102 103* The -p option enables ports 0 and 1 104 105* The --config option enables one queue on each port and maps each (port,queue) pair to a specific core. 106 The following table shows the mapping in this example: 107 108+----------+-----------+-----------+-------------------------------------+ 109| **Port** | **Queue** | **lcore** | **Description** | 110| | | | | 111+----------+-----------+-----------+-------------------------------------+ 112| 0 | 0 | 1 | Map queue 0 from port 0 to lcore 1. | 113| | | | | 114+----------+-----------+-----------+-------------------------------------+ 115| 1 | 0 | 2 | Map queue 0 from port 1 to lcore 2. | 116| | | | | 117+----------+-----------+-----------+-------------------------------------+ 118 119Refer to the *DPDK Getting Started Guide* for general information on running applications and 120the Environment Abstraction Layer (EAL) options. 121 122.. _l3_fwd_explanation: 123 124Explanation 125----------- 126 127The following sections provide some explanation of the sample application code. As mentioned in the overview section, 128the initialization and run-time paths are very similar to those of the :doc:`l2_forward_real_virtual`. 129The following sections describe aspects that are specific to the L3 Forwarding sample application. 130 131Hash Initialization 132~~~~~~~~~~~~~~~~~~~ 133 134The hash object is created and loaded with the pre-configured entries read from a global array, 135and then generate the expected 5-tuple as key to keep consistence with those of real flow 136for the convenience to execute hash performance test on 4M/8M/16M flows. 137 138.. note:: 139 140 The Hash initialization will setup both ipv4 and ipv6 hash table, 141 and populate the either table depending on the value of variable ipv6. 142 To support the hash performance test with up to 8M single direction flows/16M bi-direction flows, 143 populate_ipv4_many_flow_into_table() function will populate the hash table with specified hash table entry number(default 4M). 144 145.. note:: 146 147 Value of global variable ipv6 can be specified with --ipv6 in the command line. 148 Value of global variable hash_entry_number, 149 which is used to specify the total hash entry number for all used ports in hash performance test, 150 can be specified with --hash-entry-num VALUE in command line, being its default value 4. 151 152.. code-block:: c 153 154 #if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) 155 156 static void 157 setup_hash(int socketid) 158 { 159 // ... 160 161 if (hash_entry_number != HASH_ENTRY_NUMBER_DEFAULT) { 162 if (ipv6 == 0) { 163 /* populate the ipv4 hash */ 164 populate_ipv4_many_flow_into_table(ipv4_l3fwd_lookup_struct[socketid], hash_entry_number); 165 } else { 166 /* populate the ipv6 hash */ 167 populate_ipv6_many_flow_into_table( ipv6_l3fwd_lookup_struct[socketid], hash_entry_number); 168 } 169 } else 170 if (ipv6 == 0) { 171 /* populate the ipv4 hash */ 172 populate_ipv4_few_flow_into_table(ipv4_l3fwd_lookup_struct[socketid]); 173 } else { 174 /* populate the ipv6 hash */ 175 populate_ipv6_few_flow_into_table(ipv6_l3fwd_lookup_struct[socketid]); 176 } 177 } 178 } 179 #endif 180 181LPM Initialization 182~~~~~~~~~~~~~~~~~~ 183 184The LPM object is created and loaded with the pre-configured entries read from a global array. 185 186.. code-block:: c 187 188 #if (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) 189 190 static void 191 setup_lpm(int socketid) 192 { 193 unsigned i; 194 int ret; 195 char s[64]; 196 197 /* create the LPM table */ 198 199 snprintf(s, sizeof(s), "IPV4_L3FWD_LPM_%d", socketid); 200 201 ipv4_l3fwd_lookup_struct[socketid] = rte_lpm_create(s, socketid, IPV4_L3FWD_LPM_MAX_RULES, 0); 202 203 if (ipv4_l3fwd_lookup_struct[socketid] == NULL) 204 rte_exit(EXIT_FAILURE, "Unable to create the l3fwd LPM table" 205 " on socket %d\n", socketid); 206 207 /* populate the LPM table */ 208 209 for (i = 0; i < IPV4_L3FWD_NUM_ROUTES; i++) { 210 /* skip unused ports */ 211 212 if ((1 << ipv4_l3fwd_route_array[i].if_out & enabled_port_mask) == 0) 213 continue; 214 215 ret = rte_lpm_add(ipv4_l3fwd_lookup_struct[socketid], ipv4_l3fwd_route_array[i].ip, 216 ipv4_l3fwd_route_array[i].depth, ipv4_l3fwd_route_array[i].if_out); 217 218 if (ret < 0) { 219 rte_exit(EXIT_FAILURE, "Unable to add entry %u to the " 220 "l3fwd LPM table on socket %d\n", i, socketid); 221 } 222 223 printf("LPM: Adding route 0x%08x / %d (%d)\n", 224 (unsigned)ipv4_l3fwd_route_array[i].ip, ipv4_l3fwd_route_array[i].depth, ipv4_l3fwd_route_array[i].if_out); 225 } 226 } 227 #endif 228 229Packet Forwarding for Hash-based Lookups 230~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 231 232For each input packet, the packet forwarding operation is done by the l3fwd_simple_forward() 233or simple_ipv4_fwd_4pkts() function for IPv4 packets or the simple_ipv6_fwd_4pkts() function for IPv6 packets. 234The l3fwd_simple_forward() function provides the basic functionality for both IPv4 and IPv6 packet forwarding 235for any number of burst packets received, 236and the packet forwarding decision (that is, the identification of the output interface for the packet) 237for hash-based lookups is done by the get_ipv4_dst_port() or get_ipv6_dst_port() function. 238The get_ipv4_dst_port() function is shown below: 239 240.. code-block:: c 241 242 static inline uint8_t 243 get_ipv4_dst_port(void *ipv4_hdr, uint16_t portid, lookup_struct_t *ipv4_l3fwd_lookup_struct) 244 { 245 int ret = 0; 246 union ipv4_5tuple_host key; 247 248 ipv4_hdr = (uint8_t *)ipv4_hdr + offsetof(struct rte_ipv4_hdr, time_to_live); 249 250 m128i data = _mm_loadu_si128(( m128i*)(ipv4_hdr)); 251 252 /* Get 5 tuple: dst port, src port, dst IP address, src IP address and protocol */ 253 254 key.xmm = _mm_and_si128(data, mask0); 255 256 /* Find destination port */ 257 258 ret = rte_hash_lookup(ipv4_l3fwd_lookup_struct, (const void *)&key); 259 260 return (uint8_t)((ret < 0)? portid : ipv4_l3fwd_out_if[ret]); 261 } 262 263The get_ipv6_dst_port() function is similar to the get_ipv4_dst_port() function. 264 265The simple_ipv4_fwd_4pkts() and simple_ipv6_fwd_4pkts() function are optimized for continuous 4 valid ipv4 and ipv6 packets, 266they leverage the multiple buffer optimization to boost the performance of forwarding packets with the exact match on hash table. 267The key code snippet of simple_ipv4_fwd_4pkts() is shown below: 268 269.. code-block:: c 270 271 static inline void 272 simple_ipv4_fwd_4pkts(struct rte_mbuf* m[4], uint16_t portid, struct lcore_conf *qconf) 273 { 274 // ... 275 276 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))); 277 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))); 278 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))); 279 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))); 280 281 key[0].xmm = _mm_and_si128(data[0], mask0); 282 key[1].xmm = _mm_and_si128(data[1], mask0); 283 key[2].xmm = _mm_and_si128(data[2], mask0); 284 key[3].xmm = _mm_and_si128(data[3], mask0); 285 286 const void *key_array[4] = {&key[0], &key[1], &key[2],&key[3]}; 287 288 rte_hash_lookup_bulk(qconf->ipv4_lookup_struct, &key_array[0], 4, ret); 289 290 dst_port[0] = (ret[0] < 0)? portid:ipv4_l3fwd_out_if[ret[0]]; 291 dst_port[1] = (ret[1] < 0)? portid:ipv4_l3fwd_out_if[ret[1]]; 292 dst_port[2] = (ret[2] < 0)? portid:ipv4_l3fwd_out_if[ret[2]]; 293 dst_port[3] = (ret[3] < 0)? portid:ipv4_l3fwd_out_if[ret[3]]; 294 295 // ... 296 } 297 298The simple_ipv6_fwd_4pkts() function is similar to the simple_ipv4_fwd_4pkts() function. 299 300Known issue: IP packets with extensions or IP packets which are not TCP/UDP cannot work well at this mode. 301 302Packet Forwarding for LPM-based Lookups 303~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 304 305For each input packet, the packet forwarding operation is done by the l3fwd_simple_forward() function, 306but the packet forwarding decision (that is, the identification of the output interface for the packet) 307for LPM-based lookups is done by the get_ipv4_dst_port() function below: 308 309.. code-block:: c 310 311 static inline uint16_t 312 get_ipv4_dst_port(struct rte_ipv4_hdr *ipv4_hdr, uint16_t portid, lookup_struct_t *ipv4_l3fwd_lookup_struct) 313 { 314 uint8_t next_hop; 315 316 return ((rte_lpm_lookup(ipv4_l3fwd_lookup_struct, rte_be_to_cpu_32(ipv4_hdr->dst_addr), &next_hop) == 0)? next_hop : portid); 317 } 318