examples/ip_fragmentation/main.c

   1 /* SPDX-License-Identifier: BSD-3-Clause
   2  * Copyright(c) 2010-2014 Intel Corporation
   3  */
   4
   5 #include <stdio.h>
   6 #include <stdlib.h>
   7 #include <stdint.h>
   8 #include <inttypes.h>
   9 #include <sys/types.h>
  10 #include <sys/param.h>
  11 #include <string.h>
  12 #include <sys/queue.h>
  13 #include <stdarg.h>
  14 #include <errno.h>
  15 #include <getopt.h>
  16
  17 #include <rte_common.h>
  18 #include <rte_byteorder.h>
  19 #include <rte_log.h>
  20 #include <rte_memory.h>
  21 #include <rte_memcpy.h>
  22 #include <rte_eal.h>
  23 #include <rte_launch.h>
  24 #include <rte_atomic.h>
  25 #include <rte_cycles.h>
  26 #include <rte_prefetch.h>
  27 #include <rte_lcore.h>
  28 #include <rte_per_lcore.h>
  29 #include <rte_branch_prediction.h>
  30 #include <rte_interrupts.h>
  31 #include <rte_random.h>
  32 #include <rte_debug.h>
  33 #include <rte_ether.h>
  34 #include <rte_ethdev.h>
  35 #include <rte_mempool.h>
  36 #include <rte_mbuf.h>
  37 #include <rte_lpm.h>
  38 #include <rte_lpm6.h>
  39 #include <rte_ip.h>
  40 #include <rte_string_fns.h>
  41
  42 #include <rte_ip_frag.h>
  43
  44 #define RTE_LOGTYPE_IP_FRAG RTE_LOGTYPE_USER1
  45
  46 /* allow max jumbo frame 9.5 KB */
  47 #define JUMBO_FRAME_MAX_SIZE    0x2600
  48
  49 #define ROUNDUP_DIV(a, b)       (((a) + (b) - 1) / (b))
  50
  51 /*
  52  * Default byte size for the IPv6 Maximum Transfer Unit (MTU).
  53  * This value includes the size of IPv6 header.
  54  */
  55 #define IPV4_MTU_DEFAULT        ETHER_MTU
  56 #define IPV6_MTU_DEFAULT        ETHER_MTU
  57
  58 /*
  59  * The overhead from max frame size to MTU.
  60  * We have to consider the max possible overhead.
  61  */
  62 #define MTU_OVERHEAD    \
  63         (ETHER_HDR_LEN + ETHER_CRC_LEN + 2 * sizeof(struct vlan_hdr))
  64
  65 /*
  66  * Default payload in bytes for the IPv6 packet.
  67  */
  68 #define IPV4_DEFAULT_PAYLOAD    (IPV4_MTU_DEFAULT - sizeof(struct ipv4_hdr))
  69 #define IPV6_DEFAULT_PAYLOAD    (IPV6_MTU_DEFAULT - sizeof(struct ipv6_hdr))
  70
  71 /*
  72  * Max number of fragments per packet expected - defined by config file.
  73  */
  74 #define MAX_PACKET_FRAG RTE_LIBRTE_IP_FRAG_MAX_FRAG
  75
  76 #define NB_MBUF   8192
  77
  78 #define MAX_PKT_BURST   32
  79 #define BURST_TX_DRAIN_US 100 /* TX drain every ~100us */
  80
  81 /* Configure how many packets ahead to prefetch, when reading packets */
  82 #define PREFETCH_OFFSET 3
  83
  84 /*
  85  * Configurable number of RX/TX ring descriptors
  86  */
  87 #define RTE_TEST_RX_DESC_DEFAULT 1024
  88 #define RTE_TEST_TX_DESC_DEFAULT 1024
  89 static uint16_t nb_rxd = RTE_TEST_RX_DESC_DEFAULT;
  90 static uint16_t nb_txd = RTE_TEST_TX_DESC_DEFAULT;
  91
  92 /* ethernet addresses of ports */
  93 static struct ether_addr ports_eth_addr[RTE_MAX_ETHPORTS];
  94
  95 #ifndef IPv4_BYTES
  96 #define IPv4_BYTES_FMT "%" PRIu8 ".%" PRIu8 ".%" PRIu8 ".%" PRIu8
  97 #define IPv4_BYTES(addr) \
  98                 (uint8_t) (((addr) >> 24) & 0xFF),\
  99                 (uint8_t) (((addr) >> 16) & 0xFF),\
 100                 (uint8_t) (((addr) >> 8) & 0xFF),\
 101                 (uint8_t) ((addr) & 0xFF)
 102 #endif
 103
 104 #ifndef IPv6_BYTES
 105 #define IPv6_BYTES_FMT "%02x%02x:%02x%02x:%02x%02x:%02x%02x:"\
 106                        "%02x%02x:%02x%02x:%02x%02x:%02x%02x"
 107 #define IPv6_BYTES(addr) \
 108         addr[0],  addr[1], addr[2],  addr[3], \
 109         addr[4],  addr[5], addr[6],  addr[7], \
 110         addr[8],  addr[9], addr[10], addr[11],\
 111         addr[12], addr[13],addr[14], addr[15]
 112 #endif
 113
 114 #define IPV6_ADDR_LEN 16
 115
 116 /* mask of enabled ports */
 117 static int enabled_port_mask = 0;
 118
 119 static int rx_queue_per_lcore = 1;
 120
 121 #define MBUF_TABLE_SIZE  (2 * MAX(MAX_PKT_BURST, MAX_PACKET_FRAG))
 122
 123 struct mbuf_table {
 124         uint16_t len;
 125         struct rte_mbuf *m_table[MBUF_TABLE_SIZE];
 126 };
 127
 128 struct rx_queue {
 129         struct rte_mempool *direct_pool;
 130         struct rte_mempool *indirect_pool;
 131         struct rte_lpm *lpm;
 132         struct rte_lpm6 *lpm6;
 133         uint16_t portid;
 134 };
 135
 136 #define MAX_RX_QUEUE_PER_LCORE 16
 137 #define MAX_TX_QUEUE_PER_PORT 16
 138 struct lcore_queue_conf {
 139         uint16_t n_rx_queue;
 140         uint16_t tx_queue_id[RTE_MAX_ETHPORTS];
 141         struct rx_queue rx_queue_list[MAX_RX_QUEUE_PER_LCORE];
 142         struct mbuf_table tx_mbufs[RTE_MAX_ETHPORTS];
 143 } __rte_cache_aligned;
 144 struct lcore_queue_conf lcore_queue_conf[RTE_MAX_LCORE];
 145
 146 static struct rte_eth_conf port_conf = {
 147         .rxmode = {
 148                 .max_rx_pkt_len = JUMBO_FRAME_MAX_SIZE,
 149                 .split_hdr_size = 0,
 150                 .offloads = (DEV_RX_OFFLOAD_CHECKSUM |
 151                              DEV_RX_OFFLOAD_SCATTER |
 152                              DEV_RX_OFFLOAD_JUMBO_FRAME),
 153         },
 154         .txmode = {
 155                 .mq_mode = ETH_MQ_TX_NONE,
 156                 .offloads = (DEV_TX_OFFLOAD_IPV4_CKSUM |
 157                              DEV_TX_OFFLOAD_MULTI_SEGS),
 158         },
 159 };
 160
 161 /*
 162  * IPv4 forwarding table
 163  */
 164 struct l3fwd_ipv4_route {
 165         uint32_t ip;
 166         uint8_t  depth;
 167         uint8_t  if_out;
 168 };
 169
 170 struct l3fwd_ipv4_route l3fwd_ipv4_route_array[] = {
 171                 {IPv4(100,10,0,0), 16, 0},
 172                 {IPv4(100,20,0,0), 16, 1},
 173                 {IPv4(100,30,0,0), 16, 2},
 174                 {IPv4(100,40,0,0), 16, 3},
 175                 {IPv4(100,50,0,0), 16, 4},
 176                 {IPv4(100,60,0,0), 16, 5},
 177                 {IPv4(100,70,0,0), 16, 6},
 178                 {IPv4(100,80,0,0), 16, 7},
 179 };
 180
 181 /*
 182  * IPv6 forwarding table
 183  */
 184
 185 struct l3fwd_ipv6_route {
 186         uint8_t ip[IPV6_ADDR_LEN];
 187         uint8_t depth;
 188         uint8_t if_out;
 189 };
 190
 191 static struct l3fwd_ipv6_route l3fwd_ipv6_route_array[] = {
 192         {{1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 0},
 193         {{2,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 1},
 194         {{3,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 2},
 195         {{4,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 3},
 196         {{5,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 4},
 197         {{6,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 5},
 198         {{7,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 6},
 199         {{8,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}, 48, 7},
 200 };
 201
 202 #define LPM_MAX_RULES         1024
 203 #define LPM6_MAX_RULES         1024
 204 #define LPM6_NUMBER_TBL8S (1 << 16)
 205
 206 struct rte_lpm6_config lpm6_config = {
 207                 .max_rules = LPM6_MAX_RULES,
 208                 .number_tbl8s = LPM6_NUMBER_TBL8S,
 209                 .flags = 0
 210 };
 211
 212 static struct rte_mempool *socket_direct_pool[RTE_MAX_NUMA_NODES];
 213 static struct rte_mempool *socket_indirect_pool[RTE_MAX_NUMA_NODES];
 214 static struct rte_lpm *socket_lpm[RTE_MAX_NUMA_NODES];
 215 static struct rte_lpm6 *socket_lpm6[RTE_MAX_NUMA_NODES];
 216
 217 /* Send burst of packets on an output interface */
 218 static inline int
 219 send_burst(struct lcore_queue_conf *qconf, uint16_t n, uint16_t port)
 220 {
 221         struct rte_mbuf **m_table;
 222         int ret;
 223         uint16_t queueid;
 224
 225         queueid = qconf->tx_queue_id[port];
 226         m_table = (struct rte_mbuf **)qconf->tx_mbufs[port].m_table;
 227
 228         ret = rte_eth_tx_burst(port, queueid, m_table, n);
 229         if (unlikely(ret < n)) {
 230                 do {
 231                         rte_pktmbuf_free(m_table[ret]);
 232                 } while (++ret < n);
 233         }
 234
 235         return 0;
 236 }
 237
 238 static inline void
 239 l3fwd_simple_forward(struct rte_mbuf *m, struct lcore_queue_conf *qconf,
 240                 uint8_t queueid, uint16_t port_in)
 241 {
 242         struct rx_queue *rxq;
 243         uint32_t i, len, next_hop;
 244         uint8_t ipv6;
 245         uint16_t port_out;
 246         int32_t len2;
 247
 248         ipv6 = 0;
 249         rxq = &qconf->rx_queue_list[queueid];
 250
 251         /* by default, send everything back to the source port */
 252         port_out = port_in;
 253
 254         /* Remove the Ethernet header and trailer from the input packet */
 255         rte_pktmbuf_adj(m, (uint16_t)sizeof(struct ether_hdr));
 256
 257         /* Build transmission burst */
 258         len = qconf->tx_mbufs[port_out].len;
 259
 260         /* if this is an IPv4 packet */
 261         if (RTE_ETH_IS_IPV4_HDR(m->packet_type)) {
 262                 struct ipv4_hdr *ip_hdr;
 263                 uint32_t ip_dst;
 264                 /* Read the lookup key (i.e. ip_dst) from the input packet */
 265                 ip_hdr = rte_pktmbuf_mtod(m, struct ipv4_hdr *);
 266                 ip_dst = rte_be_to_cpu_32(ip_hdr->dst_addr);
 267
 268                 /* Find destination port */
 269                 if (rte_lpm_lookup(rxq->lpm, ip_dst, &next_hop) == 0 &&
 270                                 (enabled_port_mask & 1 << next_hop) != 0) {
 271                         port_out = next_hop;
 272
 273                         /* Build transmission burst for new port */
 274                         len = qconf->tx_mbufs[port_out].len;
 275                 }
 276
 277                 /* if we don't need to do any fragmentation */
 278                 if (likely (IPV4_MTU_DEFAULT >= m->pkt_len)) {
 279                         qconf->tx_mbufs[port_out].m_table[len] = m;
 280                         len2 = 1;
 281                 } else {
 282                         len2 = rte_ipv4_fragment_packet(m,
 283                                 &qconf->tx_mbufs[port_out].m_table[len],
 284                                 (uint16_t)(MBUF_TABLE_SIZE - len),
 285                                 IPV4_MTU_DEFAULT,
 286                                 rxq->direct_pool, rxq->indirect_pool);
 287
 288                         /* Free input packet */
 289                         rte_pktmbuf_free(m);
 290
 291                         /* If we fail to fragment the packet */
 292                         if (unlikely (len2 < 0))
 293                                 return;
 294                 }
 295         } else if (RTE_ETH_IS_IPV6_HDR(m->packet_type)) {
 296                 /* if this is an IPv6 packet */
 297                 struct ipv6_hdr *ip_hdr;
 298
 299                 ipv6 = 1;
 300
 301                 /* Read the lookup key (i.e. ip_dst) from the input packet */
 302                 ip_hdr = rte_pktmbuf_mtod(m, struct ipv6_hdr *);
 303
 304                 /* Find destination port */
 305                 if (rte_lpm6_lookup(rxq->lpm6, ip_hdr->dst_addr,
 306                                                 &next_hop) == 0 &&
 307                                 (enabled_port_mask & 1 << next_hop) != 0) {
 308                         port_out = next_hop;
 309
 310                         /* Build transmission burst for new port */
 311                         len = qconf->tx_mbufs[port_out].len;
 312                 }
 313
 314                 /* if we don't need to do any fragmentation */
 315                 if (likely (IPV6_MTU_DEFAULT >= m->pkt_len)) {
 316                         qconf->tx_mbufs[port_out].m_table[len] = m;
 317                         len2 = 1;
 318                 } else {
 319                         len2 = rte_ipv6_fragment_packet(m,
 320                                 &qconf->tx_mbufs[port_out].m_table[len],
 321                                 (uint16_t)(MBUF_TABLE_SIZE - len),
 322                                 IPV6_MTU_DEFAULT,
 323                                 rxq->direct_pool, rxq->indirect_pool);
 324
 325                         /* Free input packet */
 326                         rte_pktmbuf_free(m);
 327
 328                         /* If we fail to fragment the packet */
 329                         if (unlikely (len2 < 0))
 330                                 return;
 331                 }
 332         }
 333         /* else, just forward the packet */
 334         else {
 335                 qconf->tx_mbufs[port_out].m_table[len] = m;
 336                 len2 = 1;
 337         }
 338
 339         for (i = len; i < len + len2; i ++) {
 340                 void *d_addr_bytes;
 341
 342                 m = qconf->tx_mbufs[port_out].m_table[i];
 343                 struct ether_hdr *eth_hdr = (struct ether_hdr *)
 344                         rte_pktmbuf_prepend(m, (uint16_t)sizeof(struct ether_hdr));
 345                 if (eth_hdr == NULL) {
 346                         rte_panic("No headroom in mbuf.\n");
 347                 }
 348
 349                 m->l2_len = sizeof(struct ether_hdr);
 350
 351                 /* 02:00:00:00:00:xx */
 352                 d_addr_bytes = &eth_hdr->d_addr.addr_bytes[0];
 353                 *((uint64_t *)d_addr_bytes) = 0x000000000002 + ((uint64_t)port_out << 40);
 354
 355                 /* src addr */
 356                 ether_addr_copy(&ports_eth_addr[port_out], &eth_hdr->s_addr);
 357                 if (ipv6)
 358                         eth_hdr->ether_type = rte_be_to_cpu_16(ETHER_TYPE_IPv6);
 359                 else
 360                         eth_hdr->ether_type = rte_be_to_cpu_16(ETHER_TYPE_IPv4);
 361         }
 362
 363         len += len2;
 364
 365         if (likely(len < MAX_PKT_BURST)) {
 366                 qconf->tx_mbufs[port_out].len = (uint16_t)len;
 367                 return;
 368         }
 369
 370         /* Transmit packets */
 371         send_burst(qconf, (uint16_t)len, port_out);
 372         qconf->tx_mbufs[port_out].len = 0;
 373 }
 374
 375 /* main processing loop */
 376 static int
 377 main_loop(__attribute__((unused)) void *dummy)
 378 {
 379         struct rte_mbuf *pkts_burst[MAX_PKT_BURST];
 380         unsigned lcore_id;
 381         uint64_t prev_tsc, diff_tsc, cur_tsc;
 382         int i, j, nb_rx;
 383         uint16_t portid;
 384         struct lcore_queue_conf *qconf;
 385         const uint64_t drain_tsc = (rte_get_tsc_hz() + US_PER_S - 1) / US_PER_S * BURST_TX_DRAIN_US;
 386
 387         prev_tsc = 0;
 388
 389         lcore_id = rte_lcore_id();
 390         qconf = &lcore_queue_conf[lcore_id];
 391
 392         if (qconf->n_rx_queue == 0) {
 393                 RTE_LOG(INFO, IP_FRAG, "lcore %u has nothing to do\n", lcore_id);
 394                 return 0;
 395         }
 396
 397         RTE_LOG(INFO, IP_FRAG, "entering main loop on lcore %u\n", lcore_id);
 398
 399         for (i = 0; i < qconf->n_rx_queue; i++) {
 400
 401                 portid = qconf->rx_queue_list[i].portid;
 402                 RTE_LOG(INFO, IP_FRAG, " -- lcoreid=%u portid=%d\n", lcore_id,
 403                                 portid);
 404         }
 405
 406         while (1) {
 407
 408                 cur_tsc = rte_rdtsc();
 409
 410                 /*
 411                  * TX burst queue drain
 412                  */
 413                 diff_tsc = cur_tsc - prev_tsc;
 414                 if (unlikely(diff_tsc > drain_tsc)) {
 415
 416                         /*
 417                          * This could be optimized (use queueid instead of
 418                          * portid), but it is not called so often
 419                          */
 420                         for (portid = 0; portid < RTE_MAX_ETHPORTS; portid++) {
 421                                 if (qconf->tx_mbufs[portid].len == 0)
 422                                         continue;
 423                                 send_burst(&lcore_queue_conf[lcore_id],
 424                                            qconf->tx_mbufs[portid].len,
 425                                            portid);
 426                                 qconf->tx_mbufs[portid].len = 0;
 427                         }
 428
 429                         prev_tsc = cur_tsc;
 430                 }
 431
 432                 /*
 433                  * Read packet from RX queues
 434                  */
 435                 for (i = 0; i < qconf->n_rx_queue; i++) {
 436
 437                         portid = qconf->rx_queue_list[i].portid;
 438                         nb_rx = rte_eth_rx_burst(portid, 0, pkts_burst,
 439                                                  MAX_PKT_BURST);
 440
 441                         /* Prefetch first packets */
 442                         for (j = 0; j < PREFETCH_OFFSET && j < nb_rx; j++) {
 443                                 rte_prefetch0(rte_pktmbuf_mtod(
 444                                                 pkts_burst[j], void *));
 445                         }
 446
 447                         /* Prefetch and forward already prefetched packets */
 448                         for (j = 0; j < (nb_rx - PREFETCH_OFFSET); j++) {
 449                                 rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[
 450                                                 j + PREFETCH_OFFSET], void *));
 451                                 l3fwd_simple_forward(pkts_burst[j], qconf, i, portid);
 452                         }
 453
 454                         /* Forward remaining prefetched packets */
 455                         for (; j < nb_rx; j++) {
 456                                 l3fwd_simple_forward(pkts_burst[j], qconf, i, portid);
 457                         }
 458                 }
 459         }
 460 }
 461
 462 /* display usage */
 463 static void
 464 print_usage(const char *prgname)
 465 {
 466         printf("%s [EAL options] -- -p PORTMASK [-q NQ]\n"
 467                "  -p PORTMASK: hexadecimal bitmask of ports to configure\n"
 468                "  -q NQ: number of queue (=ports) per lcore (default is 1)\n",
 469                prgname);
 470 }
 471
 472 static int
 473 parse_portmask(const char *portmask)
 474 {
 475         char *end = NULL;
 476         unsigned long pm;
 477
 478         /* parse hexadecimal string */
 479         pm = strtoul(portmask, &end, 16);
 480         if ((portmask[0] == '\0') || (end == NULL) || (*end != '\0'))
 481                 return -1;
 482
 483         if (pm == 0)
 484                 return -1;
 485
 486         return pm;
 487 }
 488
 489 static int
 490 parse_nqueue(const char *q_arg)
 491 {
 492         char *end = NULL;
 493         unsigned long n;
 494
 495         /* parse hexadecimal string */
 496         n = strtoul(q_arg, &end, 10);
 497         if ((q_arg[0] == '\0') || (end == NULL) || (*end != '\0'))
 498                 return -1;
 499         if (n == 0)
 500                 return -1;
 501         if (n >= MAX_RX_QUEUE_PER_LCORE)
 502                 return -1;
 503
 504         return n;
 505 }
 506
 507 /* Parse the argument given in the command line of the application */
 508 static int
 509 parse_args(int argc, char **argv)
 510 {
 511         int opt, ret;
 512         char **argvopt;
 513         int option_index;
 514         char *prgname = argv[0];
 515         static struct option lgopts[] = {
 516                 {NULL, 0, 0, 0}
 517         };
 518
 519         argvopt = argv;
 520
 521         while ((opt = getopt_long(argc, argvopt, "p:q:",
 522                                   lgopts, &option_index)) != EOF) {
 523
 524                 switch (opt) {
 525                 /* portmask */
 526                 case 'p':
 527                         enabled_port_mask = parse_portmask(optarg);
 528                         if (enabled_port_mask < 0) {
 529                                 printf("invalid portmask\n");
 530                                 print_usage(prgname);
 531                                 return -1;
 532                         }
 533                         break;
 534
 535                 /* nqueue */
 536                 case 'q':
 537                         rx_queue_per_lcore = parse_nqueue(optarg);
 538                         if (rx_queue_per_lcore < 0) {
 539                                 printf("invalid queue number\n");
 540                                 print_usage(prgname);
 541                                 return -1;
 542                         }
 543                         break;
 544
 545                 /* long options */
 546                 case 0:
 547                         print_usage(prgname);
 548                         return -1;
 549
 550                 default:
 551                         print_usage(prgname);
 552                         return -1;
 553                 }
 554         }
 555
 556         if (enabled_port_mask == 0) {
 557                 printf("portmask not specified\n");
 558                 print_usage(prgname);
 559                 return -1;
 560         }
 561
 562         if (optind >= 0)
 563                 argv[optind-1] = prgname;
 564
 565         ret = optind-1;
 566         optind = 1; /* reset getopt lib */
 567         return ret;
 568 }
 569
 570 static void
 571 print_ethaddr(const char *name, struct ether_addr *eth_addr)
 572 {
 573         char buf[ETHER_ADDR_FMT_SIZE];
 574         ether_format_addr(buf, ETHER_ADDR_FMT_SIZE, eth_addr);
 575         printf("%s%s", name, buf);
 576 }
 577
 578 /* Check the link status of all ports in up to 9s, and print them finally */
 579 static void
 580 check_all_ports_link_status(uint32_t port_mask)
 581 {
 582 #define CHECK_INTERVAL 100 /* 100ms */
 583 #define MAX_CHECK_TIME 90 /* 9s (90 * 100ms) in total */
 584         uint16_t portid;
 585         uint8_t count, all_ports_up, print_flag = 0;
 586         struct rte_eth_link link;
 587
 588         printf("\nChecking link status");
 589         fflush(stdout);
 590         for (count = 0; count <= MAX_CHECK_TIME; count++) {
 591                 all_ports_up = 1;
 592                 RTE_ETH_FOREACH_DEV(portid) {
 593                         if ((port_mask & (1 << portid)) == 0)
 594                                 continue;
 595                         memset(&link, 0, sizeof(link));
 596                         rte_eth_link_get_nowait(portid, &link);
 597                         /* print link status if flag set */
 598                         if (print_flag == 1) {
 599                                 if (link.link_status)
 600                                         printf(
 601                                         "Port%d Link Up .Speed %u Mbps - %s\n",
 602                                                 portid, link.link_speed,
 603                                 (link.link_duplex == ETH_LINK_FULL_DUPLEX) ?
 604                                         ("full-duplex") : ("half-duplex\n"));
 605                                 else
 606                                         printf("Port %d Link Down\n", portid);
 607                                 continue;
 608                         }
 609                         /* clear all_ports_up flag if any link down */
 610                         if (link.link_status == ETH_LINK_DOWN) {
 611                                 all_ports_up = 0;
 612                                 break;
 613                         }
 614                 }
 615                 /* after finally printing all link status, get out */
 616                 if (print_flag == 1)
 617                         break;
 618
 619                 if (all_ports_up == 0) {
 620                         printf(".");
 621                         fflush(stdout);
 622                         rte_delay_ms(CHECK_INTERVAL);
 623                 }
 624
 625                 /* set the print_flag if all ports up or timeout */
 626                 if (all_ports_up == 1 || count == (MAX_CHECK_TIME - 1)) {
 627                         print_flag = 1;
 628                         printf("\ndone\n");
 629                 }
 630         }
 631 }
 632
 633 /* Check L3 packet type detection capablity of the NIC port */
 634 static int
 635 check_ptype(int portid)
 636 {
 637         int i, ret;
 638         int ptype_l3_ipv4 = 0, ptype_l3_ipv6 = 0;
 639         uint32_t ptype_mask = RTE_PTYPE_L3_MASK;
 640
 641         ret = rte_eth_dev_get_supported_ptypes(portid, ptype_mask, NULL, 0);
 642         if (ret <= 0)
 643                 return 0;
 644
 645         uint32_t ptypes[ret];
 646
 647         ret = rte_eth_dev_get_supported_ptypes(portid, ptype_mask, ptypes, ret);
 648         for (i = 0; i < ret; ++i) {
 649                 if (ptypes[i] & RTE_PTYPE_L3_IPV4)
 650                         ptype_l3_ipv4 = 1;
 651                 if (ptypes[i] & RTE_PTYPE_L3_IPV6)
 652                         ptype_l3_ipv6 = 1;
 653         }
 654
 655         if (ptype_l3_ipv4 == 0)
 656                 printf("port %d cannot parse RTE_PTYPE_L3_IPV4\n", portid);
 657
 658         if (ptype_l3_ipv6 == 0)
 659                 printf("port %d cannot parse RTE_PTYPE_L3_IPV6\n", portid);
 660
 661         if (ptype_l3_ipv4 && ptype_l3_ipv6)
 662                 return 1;
 663
 664         return 0;
 665
 666 }
 667
 668 /* Parse packet type of a packet by SW */
 669 static inline void
 670 parse_ptype(struct rte_mbuf *m)
 671 {
 672         struct ether_hdr *eth_hdr;
 673         uint32_t packet_type = RTE_PTYPE_UNKNOWN;
 674         uint16_t ether_type;
 675
 676         eth_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *);
 677         ether_type = eth_hdr->ether_type;
 678         if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv4))
 679                 packet_type |= RTE_PTYPE_L3_IPV4_EXT_UNKNOWN;
 680         else if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv6))
 681                 packet_type |= RTE_PTYPE_L3_IPV6_EXT_UNKNOWN;
 682
 683         m->packet_type = packet_type;
 684 }
 685
 686 /* callback function to detect packet type for a queue of a port */
 687 static uint16_t
 688 cb_parse_ptype(uint16_t port __rte_unused, uint16_t queue __rte_unused,
 689                    struct rte_mbuf *pkts[], uint16_t nb_pkts,
 690                    uint16_t max_pkts __rte_unused,
 691                    void *user_param __rte_unused)
 692 {
 693         uint16_t i;
 694
 695         for (i = 0; i < nb_pkts; ++i)
 696                 parse_ptype(pkts[i]);
 697
 698         return nb_pkts;
 699 }
 700
 701 static int
 702 init_routing_table(void)
 703 {
 704         struct rte_lpm *lpm;
 705         struct rte_lpm6 *lpm6;
 706         int socket, ret;
 707         unsigned i;
 708
 709         for (socket = 0; socket < RTE_MAX_NUMA_NODES; socket++) {
 710                 if (socket_lpm[socket]) {
 711                         lpm = socket_lpm[socket];
 712                         /* populate the LPM table */
 713                         for (i = 0; i < RTE_DIM(l3fwd_ipv4_route_array); i++) {
 714                                 ret = rte_lpm_add(lpm,
 715                                         l3fwd_ipv4_route_array[i].ip,
 716                                         l3fwd_ipv4_route_array[i].depth,
 717                                         l3fwd_ipv4_route_array[i].if_out);
 718
 719                                 if (ret < 0) {
 720                                         RTE_LOG(ERR, IP_FRAG, "Unable to add entry %i to the l3fwd "
 721                                                 "LPM table\n", i);
 722                                         return -1;
 723                                 }
 724
 725                                 RTE_LOG(INFO, IP_FRAG, "Socket %i: adding route " IPv4_BYTES_FMT
 726                                                 "/%d (port %d)\n",
 727                                         socket,
 728                                         IPv4_BYTES(l3fwd_ipv4_route_array[i].ip),
 729                                         l3fwd_ipv4_route_array[i].depth,
 730                                         l3fwd_ipv4_route_array[i].if_out);
 731                         }
 732                 }
 733
 734                 if (socket_lpm6[socket]) {
 735                         lpm6 = socket_lpm6[socket];
 736                         /* populate the LPM6 table */
 737                         for (i = 0; i < RTE_DIM(l3fwd_ipv6_route_array); i++) {
 738                                 ret = rte_lpm6_add(lpm6,
 739                                         l3fwd_ipv6_route_array[i].ip,
 740                                         l3fwd_ipv6_route_array[i].depth,
 741                                         l3fwd_ipv6_route_array[i].if_out);
 742
 743                                 if (ret < 0) {
 744                                         RTE_LOG(ERR, IP_FRAG, "Unable to add entry %i to the l3fwd "
 745                                                 "LPM6 table\n", i);
 746                                         return -1;
 747                                 }
 748
 749                                 RTE_LOG(INFO, IP_FRAG, "Socket %i: adding route " IPv6_BYTES_FMT
 750                                                 "/%d (port %d)\n",
 751                                         socket,
 752                                         IPv6_BYTES(l3fwd_ipv6_route_array[i].ip),
 753                                         l3fwd_ipv6_route_array[i].depth,
 754                                         l3fwd_ipv6_route_array[i].if_out);
 755                         }
 756                 }
 757         }
 758         return 0;
 759 }
 760
 761 static int
 762 init_mem(void)
 763 {
 764         char buf[PATH_MAX];
 765         struct rte_mempool *mp;
 766         struct rte_lpm *lpm;
 767         struct rte_lpm6 *lpm6;
 768         struct rte_lpm_config lpm_config;
 769         int socket;
 770         unsigned lcore_id;
 771
 772         /* traverse through lcores and initialize structures on each socket */
 773
 774         for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
 775
 776                 if (rte_lcore_is_enabled(lcore_id) == 0)
 777                         continue;
 778
 779                 socket = rte_lcore_to_socket_id(lcore_id);
 780
 781                 if (socket == SOCKET_ID_ANY)
 782                         socket = 0;
 783
 784                 if (socket_direct_pool[socket] == NULL) {
 785                         RTE_LOG(INFO, IP_FRAG, "Creating direct mempool on socket %i\n",
 786                                         socket);
 787                         snprintf(buf, sizeof(buf), "pool_direct_%i", socket);
 788
 789                         mp = rte_pktmbuf_pool_create(buf, NB_MBUF, 32,
 790                                 0, RTE_MBUF_DEFAULT_BUF_SIZE, socket);
 791                         if (mp == NULL) {
 792                                 RTE_LOG(ERR, IP_FRAG, "Cannot create direct mempool\n");
 793                                 return -1;
 794                         }
 795                         socket_direct_pool[socket] = mp;
 796                 }
 797
 798                 if (socket_indirect_pool[socket] == NULL) {
 799                         RTE_LOG(INFO, IP_FRAG, "Creating indirect mempool on socket %i\n",
 800                                         socket);
 801                         snprintf(buf, sizeof(buf), "pool_indirect_%i", socket);
 802
 803                         mp = rte_pktmbuf_pool_create(buf, NB_MBUF, 32, 0, 0,
 804                                 socket);
 805                         if (mp == NULL) {
 806                                 RTE_LOG(ERR, IP_FRAG, "Cannot create indirect mempool\n");
 807                                 return -1;
 808                         }
 809                         socket_indirect_pool[socket] = mp;
 810                 }
 811
 812                 if (socket_lpm[socket] == NULL) {
 813                         RTE_LOG(INFO, IP_FRAG, "Creating LPM table on socket %i\n", socket);
 814                         snprintf(buf, sizeof(buf), "IP_FRAG_LPM_%i", socket);
 815
 816                         lpm_config.max_rules = LPM_MAX_RULES;
 817                         lpm_config.number_tbl8s = 256;
 818                         lpm_config.flags = 0;
 819
 820                         lpm = rte_lpm_create(buf, socket, &lpm_config);
 821                         if (lpm == NULL) {
 822                                 RTE_LOG(ERR, IP_FRAG, "Cannot create LPM table\n");
 823                                 return -1;
 824                         }
 825                         socket_lpm[socket] = lpm;
 826                 }
 827
 828                 if (socket_lpm6[socket] == NULL) {
 829                         RTE_LOG(INFO, IP_FRAG, "Creating LPM6 table on socket %i\n", socket);
 830                         snprintf(buf, sizeof(buf), "IP_FRAG_LPM_%i", socket);
 831
 832                         lpm6 = rte_lpm6_create(buf, socket, &lpm6_config);
 833                         if (lpm6 == NULL) {
 834                                 RTE_LOG(ERR, IP_FRAG, "Cannot create LPM table\n");
 835                                 return -1;
 836                         }
 837                         socket_lpm6[socket] = lpm6;
 838                 }
 839         }
 840
 841         return 0;
 842 }
 843
 844 int
 845 main(int argc, char **argv)
 846 {
 847         struct lcore_queue_conf *qconf;
 848         struct rte_eth_dev_info dev_info;
 849         struct rte_eth_txconf *txconf;
 850         struct rx_queue *rxq;
 851         int socket, ret;
 852         uint16_t nb_ports;
 853         uint16_t queueid = 0;
 854         unsigned lcore_id = 0, rx_lcore_id = 0;
 855         uint32_t n_tx_queue, nb_lcores;
 856         uint16_t portid;
 857
 858         /* init EAL */
 859         ret = rte_eal_init(argc, argv);
 860         if (ret < 0)
 861                 rte_exit(EXIT_FAILURE, "rte_eal_init failed");
 862         argc -= ret;
 863         argv += ret;
 864
 865         /* parse application arguments (after the EAL ones) */
 866         ret = parse_args(argc, argv);
 867         if (ret < 0)
 868                 rte_exit(EXIT_FAILURE, "Invalid arguments");
 869
 870         nb_ports = rte_eth_dev_count_avail();
 871         if (nb_ports == 0)
 872                 rte_exit(EXIT_FAILURE, "No ports found!\n");
 873
 874         nb_lcores = rte_lcore_count();
 875
 876         /* initialize structures (mempools, lpm etc.) */
 877         if (init_mem() < 0)
 878                 rte_panic("Cannot initialize memory structures!\n");
 879
 880         /* check if portmask has non-existent ports */
 881         if (enabled_port_mask & ~(RTE_LEN2MASK(nb_ports, unsigned)))
 882                 rte_exit(EXIT_FAILURE, "Non-existent ports in portmask!\n");
 883
 884         /* initialize all ports */
 885         RTE_ETH_FOREACH_DEV(portid) {
 886                 struct rte_eth_conf local_port_conf = port_conf;
 887                 struct rte_eth_rxconf rxq_conf;
 888
 889                 /* skip ports that are not enabled */
 890                 if ((enabled_port_mask & (1 << portid)) == 0) {
 891                         printf("Skipping disabled port %d\n", portid);
 892                         continue;
 893                 }
 894
 895                 qconf = &lcore_queue_conf[rx_lcore_id];
 896
 897                 /* limit the frame size to the maximum supported by NIC */
 898                 rte_eth_dev_info_get(portid, &dev_info);
 899                 local_port_conf.rxmode.max_rx_pkt_len = RTE_MIN(
 900                     dev_info.max_rx_pktlen,
 901                     local_port_conf.rxmode.max_rx_pkt_len);
 902
 903                 /* get the lcore_id for this port */
 904                 while (rte_lcore_is_enabled(rx_lcore_id) == 0 ||
 905                        qconf->n_rx_queue == (unsigned)rx_queue_per_lcore) {
 906
 907                         rx_lcore_id ++;
 908                         if (rx_lcore_id >= RTE_MAX_LCORE)
 909                                 rte_exit(EXIT_FAILURE, "Not enough cores\n");
 910
 911                         qconf = &lcore_queue_conf[rx_lcore_id];
 912                 }
 913
 914                 socket = (int) rte_lcore_to_socket_id(rx_lcore_id);
 915                 if (socket == SOCKET_ID_ANY)
 916                         socket = 0;
 917
 918                 rxq = &qconf->rx_queue_list[qconf->n_rx_queue];
 919                 rxq->portid = portid;
 920                 rxq->direct_pool = socket_direct_pool[socket];
 921                 rxq->indirect_pool = socket_indirect_pool[socket];
 922                 rxq->lpm = socket_lpm[socket];
 923                 rxq->lpm6 = socket_lpm6[socket];
 924                 qconf->n_rx_queue++;
 925
 926                 /* init port */
 927                 printf("Initializing port %d on lcore %u...", portid,
 928                        rx_lcore_id);
 929                 fflush(stdout);
 930
 931                 n_tx_queue = nb_lcores;
 932                 if (n_tx_queue > MAX_TX_QUEUE_PER_PORT)
 933                         n_tx_queue = MAX_TX_QUEUE_PER_PORT;
 934                 if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_MBUF_FAST_FREE)
 935                         local_port_conf.txmode.offloads |=
 936                                 DEV_TX_OFFLOAD_MBUF_FAST_FREE;
 937                 ret = rte_eth_dev_configure(portid, 1, (uint16_t)n_tx_queue,
 938                                             &local_port_conf);
 939                 if (ret < 0) {
 940                         printf("\n");
 941                         rte_exit(EXIT_FAILURE, "Cannot configure device: "
 942                                 "err=%d, port=%d\n",
 943                                 ret, portid);
 944                 }
 945
 946                 /* set the mtu to the maximum received packet size */
 947                 ret = rte_eth_dev_set_mtu(portid,
 948                         local_port_conf.rxmode.max_rx_pkt_len - MTU_OVERHEAD);
 949                 if (ret < 0) {
 950                         printf("\n");
 951                         rte_exit(EXIT_FAILURE, "Set MTU failed: "
 952                                 "err=%d, port=%d\n",
 953                         ret, portid);
 954                 }
 955
 956                 ret = rte_eth_dev_adjust_nb_rx_tx_desc(portid, &nb_rxd,
 957                                             &nb_txd);
 958                 if (ret < 0) {
 959                         printf("\n");
 960                         rte_exit(EXIT_FAILURE, "Cannot adjust number of "
 961                                 "descriptors: err=%d, port=%d\n", ret, portid);
 962                 }
 963
 964                 /* init one RX queue */
 965                 rxq_conf = dev_info.default_rxconf;
 966                 rxq_conf.offloads = local_port_conf.rxmode.offloads;
 967                 ret = rte_eth_rx_queue_setup(portid, 0, nb_rxd,
 968                                              socket, &rxq_conf,
 969                                              socket_direct_pool[socket]);
 970                 if (ret < 0) {
 971                         printf("\n");
 972                         rte_exit(EXIT_FAILURE, "rte_eth_rx_queue_setup: "
 973                                 "err=%d, port=%d\n",
 974                                 ret, portid);
 975                 }
 976
 977                 rte_eth_macaddr_get(portid, &ports_eth_addr[portid]);
 978                 print_ethaddr(" Address:", &ports_eth_addr[portid]);
 979                 printf("\n");
 980
 981                 /* init one TX queue per couple (lcore,port) */
 982                 queueid = 0;
 983                 for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
 984                         if (rte_lcore_is_enabled(lcore_id) == 0)
 985                                 continue;
 986
 987                         socket = (int) rte_lcore_to_socket_id(lcore_id);
 988                         printf("txq=%u,%d ", lcore_id, queueid);
 989                         fflush(stdout);
 990
 991                         txconf = &dev_info.default_txconf;
 992                         txconf->offloads = local_port_conf.txmode.offloads;
 993                         ret = rte_eth_tx_queue_setup(portid, queueid, nb_txd,
 994                                                      socket, txconf);
 995                         if (ret < 0) {
 996                                 printf("\n");
 997                                 rte_exit(EXIT_FAILURE, "rte_eth_tx_queue_setup: "
 998                                         "err=%d, port=%d\n", ret, portid);
 999                         }
1000
1001                         qconf = &lcore_queue_conf[lcore_id];
1002                         qconf->tx_queue_id[portid] = queueid;
1003                         queueid++;
1004                 }
1005
1006                 printf("\n");
1007         }
1008
1009         printf("\n");
1010
1011         /* start ports */
1012         RTE_ETH_FOREACH_DEV(portid) {
1013                 if ((enabled_port_mask & (1 << portid)) == 0) {
1014                         continue;
1015                 }
1016                 /* Start device */
1017                 ret = rte_eth_dev_start(portid);
1018                 if (ret < 0)
1019                         rte_exit(EXIT_FAILURE, "rte_eth_dev_start: err=%d, port=%d\n",
1020                                 ret, portid);
1021
1022                 rte_eth_promiscuous_enable(portid);
1023
1024                 if (check_ptype(portid) == 0) {
1025                         rte_eth_add_rx_callback(portid, 0, cb_parse_ptype, NULL);
1026                         printf("Add Rx callback function to detect L3 packet type by SW :"
1027                                 " port = %d\n", portid);
1028                 }
1029         }
1030
1031         if (init_routing_table() < 0)
1032                 rte_exit(EXIT_FAILURE, "Cannot init routing table\n");
1033
1034         check_all_ports_link_status(enabled_port_mask);
1035
1036         /* launch per-lcore init on every lcore */
1037         rte_eal_mp_remote_launch(main_loop, NULL, CALL_MASTER);
1038         RTE_LCORE_FOREACH_SLAVE(lcore_id) {
1039                 if (rte_eal_wait_lcore(lcore_id) < 0)
1040                         return -1;
1041         }
1042
1043         return 0;
1044 }