Loading...
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 | /* * Copyright (c) 2018 Intel Corporation * * SPDX-License-Identifier: Apache-2.0 */ #include <zephyr/types.h> #include <stddef.h> #include <string.h> #include <errno.h> #include <net/net_pkt.h> #include <net/net_if.h> #include <net/net_ip.h> #include <net/ethernet.h> #include <ztest.h> static u8_t mac_addr[sizeof(struct net_eth_addr)]; static struct net_if *eth_if; static u8_t small_buffer[512]; /************************\ * FAKE ETHERNET DEVICE * \************************/ static void fake_dev_iface_init(struct net_if *iface) { if (mac_addr[2] == 0U) { /* 00-00-5E-00-53-xx Documentation RFC 7042 */ mac_addr[0] = 0x00; mac_addr[1] = 0x00; mac_addr[2] = 0x5E; mac_addr[3] = 0x00; mac_addr[4] = 0x53; mac_addr[5] = sys_rand32_get(); } net_if_set_link_addr(iface, mac_addr, 6, NET_LINK_ETHERNET); eth_if = iface; } static int fake_dev_send(struct device *dev, struct net_pkt *pkt) { return 0; } int fake_dev_init(struct device *dev) { ARG_UNUSED(dev); return 0; } #if defined(CONFIG_NET_L2_ETHERNET) static const struct ethernet_api fake_dev_api = { .iface_api.init = fake_dev_iface_init, .send = fake_dev_send, }; #define _ETH_L2_LAYER ETHERNET_L2 #define _ETH_L2_CTX_TYPE NET_L2_GET_CTX_TYPE(ETHERNET_L2) #define L2_HDR_SIZE sizeof(struct net_eth_hdr) #else static const struct dummy_api fake_dev_api = { .iface_api.init = fake_dev_iface_init, .send = fake_dev_send, }; #define _ETH_L2_LAYER DUMMY_L2 #define _ETH_L2_CTX_TYPE NET_L2_GET_CTX_TYPE(DUMMY_L2) #define L2_HDR_SIZE 0 #endif NET_DEVICE_INIT(fake_dev, "fake_dev", fake_dev_init, NULL, NULL, CONFIG_KERNEL_INIT_PRIORITY_DEFAULT, &fake_dev_api, _ETH_L2_LAYER, _ETH_L2_CTX_TYPE, NET_ETH_MTU); /*********************\ * UTILITY FUNCTIONS * \*********************/ static bool pkt_is_of_size(struct net_pkt *pkt, size_t size) { return (net_pkt_available_buffer(pkt) == size); } static void pkt_print_cursor(struct net_pkt *pkt) { if (!pkt || !pkt->cursor.buf || !pkt->cursor.pos) { printk("Unknown position\n"); } else { printk("Position %zu (%p) in net_buf %p (data %p)\n", pkt->cursor.pos - pkt->cursor.buf->data, pkt->cursor.pos, pkt->cursor.buf, pkt->cursor.buf->data); } } /*****************************\ * HOW TO ALLOCATE - 2 TESTS * \*****************************/ static void test_net_pkt_allocate_wo_buffer(void) { struct net_pkt *pkt; /* How to allocate a packet, with no buffer */ pkt = net_pkt_alloc(K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); /* Note that, if you already know the iface to which the packet * belongs to, you will be able to use net_pkt_alloc_on_iface(). */ pkt = net_pkt_alloc_on_iface(eth_if, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); } static void test_net_pkt_allocate_with_buffer(void) { struct net_pkt *pkt; /* How to allocate a packet, with buffer * a) - with a size that will fit MTU, let's say 512 bytes * Note: we don't care of the family/protocol for now */ pkt = net_pkt_alloc_with_buffer(eth_if, 512, AF_UNSPEC, 0, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); /* Did we get the requested size? */ zassert_true(pkt_is_of_size(pkt, 512), "Pkt size is not right"); /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); /* * b) - with a size that will not fit MTU, let's say 1800 bytes * Note: again we don't care of family/protocol for now. */ pkt = net_pkt_alloc_with_buffer(eth_if, 1800, AF_UNSPEC, 0, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); zassert_false(pkt_is_of_size(pkt, 1800), "Pkt size is not right"); zassert_true(pkt_is_of_size(pkt, net_if_get_mtu(eth_if) + L2_HDR_SIZE), "Pkt size is not right"); /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); /* * c) - Now with 512 bytes but on IPv4/UDP */ pkt = net_pkt_alloc_with_buffer(eth_if, 512, AF_INET, IPPROTO_UDP, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); /* Because 512 + NET_IPV4UDPH_LEN fits MTU, total must be that one */ zassert_true(pkt_is_of_size(pkt, 512 + NET_IPV4UDPH_LEN), "Pkt overall size does not match"); /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); /* * c) - Now with 1800 bytes but on IPv4/UDP */ pkt = net_pkt_alloc_with_buffer(eth_if, 1800, AF_INET, IPPROTO_UDP, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); /* Because 1800 + NET_IPV4UDPH_LEN won't fit MTU, payload size * should be MTU */ zassert_true(net_pkt_available_buffer(pkt) == net_if_get_mtu(eth_if), "Payload buf size does not match for ipv4/udp"); /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); } /********************************\ * HOW TO R/W A PACKET - TESTS * \********************************/ static void test_net_pkt_basics_of_rw(void) { struct net_pkt *pkt; int ret; pkt = net_pkt_alloc_with_buffer(eth_if, 512, AF_UNSPEC, 0, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); /* Once newly allocated with buffer, * a packet has no data accounted for in its buffer */ zassert_true(net_pkt_get_len(pkt) == 0, "Pkt initial length should be 0"); /* This is done through net_buf which can distinguish * the size of a buffer from the length of the data in it. */ /* Let's subsequently write 1 byte, then 2 bytes and 4 bytes * We write values made of 0s */ ret = net_pkt_write_u8(pkt, 0); zassert_true(ret == 0, "Pkt write failed"); /* Length should be 1 now */ zassert_true(net_pkt_get_len(pkt) == 1, "Pkt length mismatch"); ret = net_pkt_write_be16(pkt, 0); zassert_true(ret == 0, "Pkt write failed"); /* Length should be 3 now */ zassert_true(net_pkt_get_len(pkt) == 3, "Pkt length mismatch"); ret = net_pkt_write_be32(pkt, 0); zassert_true(ret == 0, "Pkt write failed"); /* Length should be 7 now */ zassert_true(net_pkt_get_len(pkt) == 7, "Pkt length mismatch"); /* All these writing functions use net_ptk_write(), which works * this way: */ ret = net_pkt_write(pkt, small_buffer, 9); zassert_true(ret == 0, "Pkt write failed"); /* Length should be 16 now */ zassert_true(net_pkt_get_len(pkt) == 16, "Pkt length mismatch"); /* Now let's say you want to memset some data */ ret = net_pkt_memset(pkt, 0, 4); zassert_true(ret == 0, "Pkt memset failed"); /* Length should be 20 now */ zassert_true(net_pkt_get_len(pkt) == 20, "Pkt length mismatch"); /* So memset affects the length exactly as write does */ /* Sometimes you might want to advance in the buffer without caring * what's written there since you'll eventually come back for that. * net_pkt_skip() is used for it. * Note: usally you will not have to use that function a lot yourself. */ ret = net_pkt_skip(pkt, 20); zassert_true(ret == 0, "Pkt skip failed"); /* Length should be 40 now */ zassert_true(net_pkt_get_len(pkt) == 40, "Pkt length mismatch"); /* Again, skip affected the length also, like a write * But wait a minute: how to get back then, in order to write at * the position we just skipped? * * So let's introduce the concept of buffer cursor. (which could * be named 'cursor' if such name has more relevancy. Basically, each * net_pkt embeds such 'cursor': it's like a head of a tape * recorder/reader, it holds the current position in the buffer where * you can r/w. All operations use and update it below. * There is, however, a catch: buffer is described through net_buf * and these are like a simple linked-list. * Which means that unlike a tape recorder/reader: you are not * able to go backward. Only back from starting point and forward. * Thus why there is a net_pkt_cursor_init(pkt) which will let you going * back from the start. We could hold more info in order to avoid that, * but that would mean growing each an every net_buf. */ net_pkt_cursor_init(pkt); /* But isn't it so that if I want to go at the previous position I * skipped, I'll use skip again but then won't it affect again the * length? * Answer is yes. Hopefully there is a mean to avoid that. Basically * for data that already "exists" in the buffer (aka: data accounted * for in the buffer, through the length) you'll need to set the packet * to overwrite: all subsequent operations will then work on existing * data and will not affect the length (it won't add more data) */ net_pkt_set_overwrite(pkt, true); zassert_true(net_pkt_is_being_overwritten(pkt), "Pkt is not set to overwrite"); /* Ok so previous skipped position was at offset 20 */ ret = net_pkt_skip(pkt, 20); zassert_true(ret == 0, "Pkt skip failed"); /* Length should _still_ be 40 */ zassert_true(net_pkt_get_len(pkt) == 40, "Pkt length mismatch"); /* And you can write stuff */ ret = net_pkt_write_le32(pkt, 0); zassert_true(ret == 0, "Pkt write failed"); /* Again, length should _still_ be 40 */ zassert_true(net_pkt_get_len(pkt) == 40, "Pkt length mismatch"); /* Let's memset the rest */ ret = net_pkt_memset(pkt, 0, 16); zassert_true(ret == 0, "Pkt memset failed"); /* Again, length should _still_ be 40 */ zassert_true(net_pkt_get_len(pkt) == 40, "Pkt length mismatch"); /* We are now back at the end of the existing data in the buffer * Since overwrite is still on, we should not be able to r/w * anything. * This is completely nominal, as being set, overwrite allows r/w only * on existing data in the buffer: */ ret = net_pkt_write_be32(pkt, 0); zassert_true(ret != 0, "Pkt write succeeded where it shouldn't have"); /* Logically, in order to be able to add new data in the buffer, * overwrite should be disabled: */ net_pkt_set_overwrite(pkt, false); /* But it will fail: */ ret = net_pkt_write_le32(pkt, 0); zassert_true(ret != 0, "Pkt write succeeded?"); /* Why is that? * This is because in case of r/w error: the iterator is invalidated. * This a design choice, once you get a r/w error it means your code * messed up requesting smaller buffer than you actually needed, or * writing too much data than it should have been etc...). * So you must drop your packet entirely. */ /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); } void test_net_pkt_advanced_basics(void) { struct net_pkt_cursor backup; struct net_pkt *pkt; int ret; pkt = net_pkt_alloc_with_buffer(eth_if, 512, AF_INET, IPPROTO_UDP, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); pkt_print_cursor(pkt); /* As stated earlier, initializing the cursor, is the way to go * back from the start in the buffer (either header or payload then). * We also showed that using net_pkt_skip() could be used to move * forward in the buffer. * But what if you are far in the buffer, you need to go backward, * and back again to your previous position? * You could certainly do: */ ret = net_pkt_write(pkt, small_buffer, 20); zassert_true(ret == 0, "Pkt write failed"); pkt_print_cursor(pkt); net_pkt_cursor_init(pkt); pkt_print_cursor(pkt); /* ... do something here ... */ /* And finally go back with overwrite/skip: */ net_pkt_set_overwrite(pkt, true); ret = net_pkt_skip(pkt, 20); zassert_true(ret == 0, "Pkt skip failed"); net_pkt_set_overwrite(pkt, false); pkt_print_cursor(pkt); /* In this example, do not focus on the 20 bytes. It is just for * the sake of the example. * The other method is backup/restore the packet cursor. */ net_pkt_cursor_backup(pkt, &backup); net_pkt_cursor_init(pkt); /* ... do something here ... */ /* and restore: */ net_pkt_cursor_restore(pkt, &backup); pkt_print_cursor(pkt); /* Another feature, is how you access your data. Earlier was * presented basic r/w functions. But sometime you might want to * access your data directly through a structure/type etc... * Due to the "fragmented" possible nature of your buffer, you * need to know if the data you are trying to access is in * contiguous area. * For this, you'll use: */ ret = (int) net_pkt_is_contiguous(pkt, 4); zassert_true(ret == 1, "Pkt contiguity check failed"); /* If that's successful you should be able to get the actual * position in the buffer and cast it to the type you want. */ { u32_t *val = (u32_t *)net_pkt_cursor_get_pos(pkt); *val = 0U; /* etc... */ } /* However, to advance your cursor, since none of the usual r/w * functions got used: net_pkt_skip() should be called relevantly: */ net_pkt_skip(pkt, 4); /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); /* Obviously one will very rarely use these 2 last low level functions * - net_pkt_is_contiguous() * - net_pkt_cursor_update() * * Let's see why next. */ } void test_net_pkt_easier_rw_usage(void) { struct net_pkt *pkt; int ret; pkt = net_pkt_alloc_with_buffer(eth_if, 512, AF_INET, IPPROTO_UDP, K_NO_WAIT); zassert_true(pkt != NULL, "Pkt not allocated"); /* In net core, all goes down in fine to header manipulation. * Either it's an IP header, UDP, ICMP, TCP one etc... * One would then prefer to access those directly via there * descriptors (struct net_udp_hdr, struct net_icmp_hdr, ...) * rather than building it byte by bytes etc... * * As seen earlier, it is possible to cast on current position. * However, due to the "fragmented" possible nature of the buffer, * it should also be possible to handle the case the data being * accessed is scattered on 1+ net_buf. * * To avoid redoing the contiguity check, cast or copy on failure, * a complex type named struct net_pkt_header_access exists. * It solves both cases (accessing data contiguous or not), without * the need for runtime allocation (all is on stack) */ { NET_PKT_DATA_ACCESS_DEFINE(ip_access, struct net_ipv4_hdr); struct net_ipv4_hdr *ip_hdr; ip_hdr = (struct net_ipv4_hdr *) net_pkt_get_data(pkt, &ip_access); zassert_not_null(ip_hdr, "Accessor failed"); ip_hdr->tos = 0x00; ret = net_pkt_set_data(pkt, &ip_access); zassert_true(ret == 0, "Accessor failed"); zassert_true(net_pkt_get_len(pkt) == NET_IPV4H_LEN, "Pkt length mismatch"); } /* As you can notice: get/set take also care of handling the cursor * and updating the packet length relevantly thus why packet length * has properly grown. */ /* Freeing the packet */ net_pkt_unref(pkt); zassert_true(atomic_get(&pkt->atomic_ref) == 0, "Pkt not properly unreferenced"); } u8_t b5_data[10] = "qrstuvwxyz"; struct net_buf b5 = { .ref = 1, .data = b5_data, .len = 0, .size = 0, }; u8_t b4_data[4] = "mnop"; struct net_buf b4 = { .frags = &b5, .ref = 1, .data = b4_data, .len = sizeof(b4_data) - 2, .size = sizeof(b4_data), }; struct net_buf b3 = { .frags = &b4, .ref = 1, }; u8_t b2_data[8] = "efghijkl"; struct net_buf b2 = { .frags = &b3, .ref = 1, .data = b2_data, .len = 0, .size = sizeof(b2_data), }; u8_t b1_data[4] = "abcd"; struct net_buf b1 = { .frags = &b2, .ref = 1, .data = b1_data, .len = sizeof(b1_data) - 2, .size = sizeof(b1_data), }; void test_net_pkt_copy(void) { struct net_pkt *pkt_src; struct net_pkt *pkt_dst; pkt_src = net_pkt_alloc_on_iface(eth_if, K_NO_WAIT); zassert_true(pkt_src != NULL, "Pkt not allocated"); pkt_print_cursor(pkt_src); /* Let's append the buffers */ net_pkt_append_buffer(pkt_src, &b1); net_pkt_set_overwrite(pkt_src, true); /* There should be some space left */ zassert_true(net_pkt_available_buffer(pkt_src) != 0, "No space left?"); /* Length should be 4 */ zassert_true(net_pkt_get_len(pkt_src) == 4, "Wrong length"); /* Actual space left is 12 (in b1, b2 and b4) */ zassert_true(net_pkt_available_buffer(pkt_src) == 12, "Wrong space left?"); pkt_print_cursor(pkt_src); /* Now let's clone the pkt * This will test net_pkt_copy_new() as it uses it for the buffers */ pkt_dst = net_pkt_clone(pkt_src, K_NO_WAIT); zassert_true(pkt_dst != NULL, "Pkt not clone"); /* Cloning does not take into account left space, * but only occupied one */ zassert_true(net_pkt_available_buffer(pkt_dst) == 0, "Space left"); zassert_true(net_pkt_get_len(pkt_src) == net_pkt_get_len(pkt_dst), "Not same amount?"); /* It also did not care to copy the net_buf itself, only the content * so, knowing that the base buffer size is bigger than necessary, * pkt_dst has only one net_buf */ zassert_true(pkt_dst->buffer->frags == NULL, "Not only one buffer?"); /* Freeing the packet */ pkt_src->buffer = NULL; net_pkt_unref(pkt_src); zassert_true(atomic_get(&pkt_src->atomic_ref) == 0, "Pkt not properly unreferenced"); net_pkt_unref(pkt_dst); zassert_true(atomic_get(&pkt_dst->atomic_ref) == 0, "Pkt not properly unreferenced"); } void test_main(void) { eth_if = net_if_get_default(); ztest_test_suite(net_pkt_tests, ztest_unit_test(test_net_pkt_allocate_wo_buffer), ztest_unit_test(test_net_pkt_allocate_with_buffer), ztest_unit_test(test_net_pkt_basics_of_rw), ztest_unit_test(test_net_pkt_advanced_basics), ztest_unit_test(test_net_pkt_easier_rw_usage), ztest_unit_test(test_net_pkt_copy) ); ztest_run_test_suite(net_pkt_tests); } |