1 /* $OpenBSD: rtable.c,v 1.82 2023/04/19 17:42:47 bluhm Exp $ */ 2 3 /* 4 * Copyright (c) 2014-2016 Martin Pieuchot 5 * 6 * Permission to use, copy, modify, and distribute this software for any 7 * purpose with or without fee is hereby granted, provided that the above 8 * copyright notice and this permission notice appear in all copies. 9 * 10 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES 11 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF 12 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR 13 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES 14 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN 15 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF 16 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. 17 */ 18 19 #ifndef _KERNEL 20 #include "kern_compat.h" 21 #else 22 #include <sys/param.h> 23 #include <sys/systm.h> 24 #include <sys/socket.h> 25 #include <sys/malloc.h> 26 #include <sys/queue.h> 27 #include <sys/domain.h> 28 #include <sys/srp.h> 29 #endif 30 31 #include <net/rtable.h> 32 #include <net/route.h> 33 34 /* 35 * Structures used by rtable_get() to retrieve the corresponding 36 * routing table for a given pair of ``af'' and ``rtableid''. 37 * 38 * Note that once allocated routing table heads are never freed. 39 * This way we do not need to reference count them. 40 * 41 * afmap rtmap/dommp 42 * ----------- --------- ----- 43 * | 0 |--------> | 0 | 0 | ... | 0 | Array mapping rtableid (=index) 44 * ----------- --------- ----- to rdomain/loopback (=value). 45 * | AF_INET |. 46 * ----------- `. .---------. .---------. 47 * ... `----> | rtable0 | ... | rtableN | Array of pointers for 48 * ----------- '---------' '---------' IPv4 routing tables 49 * | AF_MPLS | indexed by ``rtableid''. 50 * ----------- 51 */ 52 struct srp *afmap; 53 uint8_t af2idx[AF_MAX+1]; /* To only allocate supported AF */ 54 uint8_t af2idx_max; 55 56 /* Array of routing table pointers. */ 57 struct rtmap { 58 unsigned int limit; 59 void **tbl; 60 }; 61 62 /* 63 * Array of rtableid -> rdomain mapping. 64 * 65 * Only used for the first index as described above. 66 */ 67 struct dommp { 68 unsigned int limit; 69 /* 70 * Array to get the routing domain and loopback interface related to 71 * a routing table. Format: 72 * 73 * 8 unused bits | 16 bits for loopback index | 8 bits for rdomain 74 */ 75 unsigned int *value; 76 }; 77 78 unsigned int rtmap_limit = 0; 79 80 void rtmap_init(void); 81 void rtmap_grow(unsigned int, sa_family_t); 82 void rtmap_dtor(void *, void *); 83 84 struct srp_gc rtmap_gc = SRP_GC_INITIALIZER(rtmap_dtor, NULL); 85 86 void rtable_init_backend(void); 87 void *rtable_alloc(unsigned int, unsigned int, unsigned int); 88 void *rtable_get(unsigned int, sa_family_t); 89 90 void 91 rtmap_init(void) 92 { 93 const struct domain *dp; 94 int i; 95 96 /* Start with a single table for every domain that requires it. */ 97 for (i = 0; (dp = domains[i]) != NULL; i++) { 98 if (dp->dom_rtoffset == 0) 99 continue; 100 101 rtmap_grow(1, dp->dom_family); 102 } 103 104 /* Initialize the rtableid->rdomain mapping table. */ 105 rtmap_grow(1, 0); 106 107 rtmap_limit = 1; 108 } 109 110 /* 111 * Grow the size of the array of routing table for AF ``af'' to ``nlimit''. 112 */ 113 void 114 rtmap_grow(unsigned int nlimit, sa_family_t af) 115 { 116 struct rtmap *map, *nmap; 117 int i; 118 119 KERNEL_ASSERT_LOCKED(); 120 121 KASSERT(nlimit > rtmap_limit); 122 123 nmap = malloc(sizeof(*nmap), M_RTABLE, M_WAITOK); 124 nmap->limit = nlimit; 125 nmap->tbl = mallocarray(nlimit, sizeof(*nmap[0].tbl), M_RTABLE, 126 M_WAITOK|M_ZERO); 127 128 map = srp_get_locked(&afmap[af2idx[af]]); 129 if (map != NULL) { 130 KASSERT(map->limit == rtmap_limit); 131 132 for (i = 0; i < map->limit; i++) 133 nmap->tbl[i] = map->tbl[i]; 134 } 135 136 srp_update_locked(&rtmap_gc, &afmap[af2idx[af]], nmap); 137 } 138 139 void 140 rtmap_dtor(void *null, void *xmap) 141 { 142 struct rtmap *map = xmap; 143 144 /* 145 * doesn't need to be serialized since this is the last reference 146 * to this map. there's nothing to race against. 147 */ 148 free(map->tbl, M_RTABLE, map->limit * sizeof(*map[0].tbl)); 149 free(map, M_RTABLE, sizeof(*map)); 150 } 151 152 void 153 rtable_init(void) 154 { 155 const struct domain *dp; 156 int i; 157 158 KASSERT(sizeof(struct rtmap) == sizeof(struct dommp)); 159 160 /* We use index 0 for the rtable/rdomain map. */ 161 af2idx_max = 1; 162 memset(af2idx, 0, sizeof(af2idx)); 163 164 /* 165 * Compute the maximum supported key length in case the routing 166 * table backend needs it. 167 */ 168 for (i = 0; (dp = domains[i]) != NULL; i++) { 169 if (dp->dom_rtoffset == 0) 170 continue; 171 172 af2idx[dp->dom_family] = af2idx_max++; 173 } 174 rtable_init_backend(); 175 176 /* 177 * Allocate AF-to-id table now that we now how many AFs this 178 * kernel supports. 179 */ 180 afmap = mallocarray(af2idx_max + 1, sizeof(*afmap), M_RTABLE, 181 M_WAITOK|M_ZERO); 182 183 rtmap_init(); 184 185 if (rtable_add(0) != 0) 186 panic("unable to create default routing table"); 187 188 rt_timer_init(); 189 } 190 191 int 192 rtable_add(unsigned int id) 193 { 194 const struct domain *dp; 195 void *tbl; 196 struct rtmap *map; 197 struct dommp *dmm; 198 sa_family_t af; 199 unsigned int off, alen; 200 int i, error = 0; 201 202 if (id > RT_TABLEID_MAX) 203 return (EINVAL); 204 205 KERNEL_LOCK(); 206 207 if (rtable_exists(id)) 208 goto out; 209 210 for (i = 0; (dp = domains[i]) != NULL; i++) { 211 if (dp->dom_rtoffset == 0) 212 continue; 213 214 af = dp->dom_family; 215 off = dp->dom_rtoffset; 216 alen = dp->dom_maxplen; 217 218 if (id >= rtmap_limit) 219 rtmap_grow(id + 1, af); 220 221 tbl = rtable_alloc(id, alen, off); 222 if (tbl == NULL) { 223 error = ENOMEM; 224 goto out; 225 } 226 227 map = srp_get_locked(&afmap[af2idx[af]]); 228 map->tbl[id] = tbl; 229 } 230 231 /* Reflect possible growth. */ 232 if (id >= rtmap_limit) { 233 rtmap_grow(id + 1, 0); 234 rtmap_limit = id + 1; 235 } 236 237 /* Use main rtable/rdomain by default. */ 238 dmm = srp_get_locked(&afmap[0]); 239 dmm->value[id] = 0; 240 out: 241 KERNEL_UNLOCK(); 242 243 return (error); 244 } 245 246 void * 247 rtable_get(unsigned int rtableid, sa_family_t af) 248 { 249 struct rtmap *map; 250 void *tbl = NULL; 251 struct srp_ref sr; 252 253 if (af >= nitems(af2idx) || af2idx[af] == 0) 254 return (NULL); 255 256 map = srp_enter(&sr, &afmap[af2idx[af]]); 257 if (rtableid < map->limit) 258 tbl = map->tbl[rtableid]; 259 srp_leave(&sr); 260 261 return (tbl); 262 } 263 264 int 265 rtable_exists(unsigned int rtableid) 266 { 267 const struct domain *dp; 268 void *tbl; 269 int i; 270 271 for (i = 0; (dp = domains[i]) != NULL; i++) { 272 if (dp->dom_rtoffset == 0) 273 continue; 274 275 tbl = rtable_get(rtableid, dp->dom_family); 276 if (tbl != NULL) 277 return (1); 278 } 279 280 return (0); 281 } 282 283 int 284 rtable_empty(unsigned int rtableid) 285 { 286 const struct domain *dp; 287 int i; 288 struct art_root *tbl; 289 290 for (i = 0; (dp = domains[i]) != NULL; i++) { 291 if (dp->dom_rtoffset == 0) 292 continue; 293 294 tbl = rtable_get(rtableid, dp->dom_family); 295 if (tbl == NULL) 296 continue; 297 if (tbl->ar_root.ref != NULL) 298 return (0); 299 } 300 301 return (1); 302 } 303 304 unsigned int 305 rtable_l2(unsigned int rtableid) 306 { 307 struct dommp *dmm; 308 unsigned int rdomain = 0; 309 struct srp_ref sr; 310 311 dmm = srp_enter(&sr, &afmap[0]); 312 if (rtableid < dmm->limit) 313 rdomain = (dmm->value[rtableid] & RT_TABLEID_MASK); 314 srp_leave(&sr); 315 316 return (rdomain); 317 } 318 319 unsigned int 320 rtable_loindex(unsigned int rtableid) 321 { 322 struct dommp *dmm; 323 unsigned int loifidx = 0; 324 struct srp_ref sr; 325 326 dmm = srp_enter(&sr, &afmap[0]); 327 if (rtableid < dmm->limit) 328 loifidx = (dmm->value[rtableid] >> RT_TABLEID_BITS); 329 srp_leave(&sr); 330 331 return (loifidx); 332 } 333 334 void 335 rtable_l2set(unsigned int rtableid, unsigned int rdomain, unsigned int loifidx) 336 { 337 struct dommp *dmm; 338 unsigned int value; 339 340 KERNEL_ASSERT_LOCKED(); 341 342 if (!rtable_exists(rtableid) || !rtable_exists(rdomain)) 343 return; 344 345 value = (rdomain & RT_TABLEID_MASK) | (loifidx << RT_TABLEID_BITS); 346 347 dmm = srp_get_locked(&afmap[0]); 348 dmm->value[rtableid] = value; 349 } 350 351 352 static inline uint8_t *satoaddr(struct art_root *, struct sockaddr *); 353 354 int an_match(struct art_node *, struct sockaddr *, int); 355 void rtentry_ref(void *, void *); 356 void rtentry_unref(void *, void *); 357 358 void rtable_mpath_insert(struct art_node *, struct rtentry *); 359 360 struct srpl_rc rt_rc = SRPL_RC_INITIALIZER(rtentry_ref, rtentry_unref, NULL); 361 362 void 363 rtable_init_backend(void) 364 { 365 art_init(); 366 } 367 368 void * 369 rtable_alloc(unsigned int rtableid, unsigned int alen, unsigned int off) 370 { 371 return (art_alloc(rtableid, alen, off)); 372 } 373 374 int 375 rtable_setsource(unsigned int rtableid, int af, struct sockaddr *src) 376 { 377 struct art_root *ar; 378 379 NET_ASSERT_LOCKED_EXCLUSIVE(); 380 381 if ((ar = rtable_get(rtableid, af)) == NULL) 382 return (EAFNOSUPPORT); 383 384 ar->ar_source = src; 385 386 return (0); 387 } 388 389 struct sockaddr * 390 rtable_getsource(unsigned int rtableid, int af) 391 { 392 struct art_root *ar; 393 394 NET_ASSERT_LOCKED(); 395 396 ar = rtable_get(rtableid, af); 397 if (ar == NULL) 398 return (NULL); 399 400 return (ar->ar_source); 401 } 402 403 void 404 rtable_clearsource(unsigned int rtableid, struct sockaddr *src) 405 { 406 struct sockaddr *addr; 407 408 addr = rtable_getsource(rtableid, src->sa_family); 409 if (addr && (addr->sa_len == src->sa_len)) { 410 if (memcmp(src, addr, addr->sa_len) == 0) { 411 rtable_setsource(rtableid, src->sa_family, NULL); 412 } 413 } 414 } 415 416 struct rtentry * 417 rtable_lookup(unsigned int rtableid, struct sockaddr *dst, 418 struct sockaddr *mask, struct sockaddr *gateway, uint8_t prio) 419 { 420 struct art_root *ar; 421 struct art_node *an; 422 struct rtentry *rt = NULL; 423 struct srp_ref sr, nsr; 424 uint8_t *addr; 425 int plen; 426 427 ar = rtable_get(rtableid, dst->sa_family); 428 if (ar == NULL) 429 return (NULL); 430 431 addr = satoaddr(ar, dst); 432 433 /* No need for a perfect match. */ 434 if (mask == NULL) { 435 an = art_match(ar, addr, &nsr); 436 if (an == NULL) 437 goto out; 438 } else { 439 plen = rtable_satoplen(dst->sa_family, mask); 440 if (plen == -1) 441 return (NULL); 442 443 an = art_lookup(ar, addr, plen, &nsr); 444 445 /* Make sure we've got a perfect match. */ 446 if (!an_match(an, dst, plen)) 447 goto out; 448 } 449 450 SRPL_FOREACH(rt, &sr, &an->an_rtlist, rt_next) { 451 if (prio != RTP_ANY && 452 (rt->rt_priority & RTP_MASK) != (prio & RTP_MASK)) 453 continue; 454 455 if (gateway == NULL) 456 break; 457 458 if (rt->rt_gateway->sa_len == gateway->sa_len && 459 memcmp(rt->rt_gateway, gateway, gateway->sa_len) == 0) 460 break; 461 } 462 if (rt != NULL) 463 rtref(rt); 464 465 SRPL_LEAVE(&sr); 466 out: 467 srp_leave(&nsr); 468 469 return (rt); 470 } 471 472 struct rtentry * 473 rtable_match(unsigned int rtableid, struct sockaddr *dst, uint32_t *src) 474 { 475 struct art_root *ar; 476 struct art_node *an; 477 struct rtentry *rt = NULL; 478 struct srp_ref sr, nsr; 479 uint8_t *addr; 480 int hash; 481 482 ar = rtable_get(rtableid, dst->sa_family); 483 if (ar == NULL) 484 return (NULL); 485 486 addr = satoaddr(ar, dst); 487 488 an = art_match(ar, addr, &nsr); 489 if (an == NULL) 490 goto out; 491 492 rt = SRPL_FIRST(&sr, &an->an_rtlist); 493 if (rt == NULL) { 494 SRPL_LEAVE(&sr); 495 goto out; 496 } 497 rtref(rt); 498 SRPL_LEAVE(&sr); 499 500 /* Gateway selection by Hash-Threshold (RFC 2992) */ 501 if ((hash = rt_hash(rt, dst, src)) != -1) { 502 struct rtentry *mrt; 503 int threshold, npaths = 0; 504 505 KASSERT(hash <= 0xffff); 506 507 SRPL_FOREACH(mrt, &sr, &an->an_rtlist, rt_next) { 508 /* Only count nexthops with the same priority. */ 509 if (mrt->rt_priority == rt->rt_priority) 510 npaths++; 511 } 512 SRPL_LEAVE(&sr); 513 514 threshold = (0xffff / npaths) + 1; 515 516 /* 517 * we have no protection against concurrent modification of the 518 * route list attached to the node, so we won't necessarily 519 * have the same number of routes. for most modifications, 520 * we'll pick a route that we wouldn't have if we only saw the 521 * list before or after the change. if we were going to use 522 * the last available route, but it got removed, we'll hit 523 * the end of the list and then pick the first route. 524 */ 525 526 mrt = SRPL_FIRST(&sr, &an->an_rtlist); 527 while (hash > threshold && mrt != NULL) { 528 if (mrt->rt_priority == rt->rt_priority) 529 hash -= threshold; 530 mrt = SRPL_FOLLOW(&sr, mrt, rt_next); 531 } 532 533 if (mrt != NULL) { 534 rtref(mrt); 535 rtfree(rt); 536 rt = mrt; 537 } 538 SRPL_LEAVE(&sr); 539 } 540 out: 541 srp_leave(&nsr); 542 return (rt); 543 } 544 545 int 546 rtable_insert(unsigned int rtableid, struct sockaddr *dst, 547 struct sockaddr *mask, struct sockaddr *gateway, uint8_t prio, 548 struct rtentry *rt) 549 { 550 struct rtentry *mrt; 551 struct srp_ref sr; 552 struct art_root *ar; 553 struct art_node *an, *prev; 554 uint8_t *addr; 555 int plen; 556 unsigned int rt_flags; 557 int error = 0; 558 559 ar = rtable_get(rtableid, dst->sa_family); 560 if (ar == NULL) 561 return (EAFNOSUPPORT); 562 563 addr = satoaddr(ar, dst); 564 plen = rtable_satoplen(dst->sa_family, mask); 565 if (plen == -1) 566 return (EINVAL); 567 568 rtref(rt); /* guarantee rtfree won't do anything during insert */ 569 rw_enter_write(&ar->ar_lock); 570 571 /* Do not permit exactly the same dst/mask/gw pair. */ 572 an = art_lookup(ar, addr, plen, &sr); 573 srp_leave(&sr); /* an can't go away while we have the lock */ 574 if (an_match(an, dst, plen)) { 575 struct rtentry *mrt; 576 int mpathok = ISSET(rt->rt_flags, RTF_MPATH); 577 578 SRPL_FOREACH_LOCKED(mrt, &an->an_rtlist, rt_next) { 579 if (prio != RTP_ANY && 580 (mrt->rt_priority & RTP_MASK) != (prio & RTP_MASK)) 581 continue; 582 583 if (!mpathok || 584 (mrt->rt_gateway->sa_len == gateway->sa_len && 585 memcmp(mrt->rt_gateway, gateway, 586 gateway->sa_len) == 0)) { 587 error = EEXIST; 588 goto leave; 589 } 590 } 591 } 592 593 an = art_get(dst, plen); 594 if (an == NULL) { 595 error = ENOBUFS; 596 goto leave; 597 } 598 599 /* prepare for immediate operation if insert succeeds */ 600 rt_flags = rt->rt_flags; 601 rt->rt_flags &= ~RTF_MPATH; 602 rt->rt_dest = dst; 603 rt->rt_plen = plen; 604 SRPL_INSERT_HEAD_LOCKED(&rt_rc, &an->an_rtlist, rt, rt_next); 605 606 prev = art_insert(ar, an, addr, plen); 607 if (prev != an) { 608 SRPL_REMOVE_LOCKED(&rt_rc, &an->an_rtlist, rt, rtentry, 609 rt_next); 610 rt->rt_flags = rt_flags; 611 art_put(an); 612 613 if (prev == NULL) { 614 error = ESRCH; 615 goto leave; 616 } 617 618 an = prev; 619 620 mrt = SRPL_FIRST_LOCKED(&an->an_rtlist); 621 KASSERT(mrt != NULL); 622 KASSERT((rt->rt_flags & RTF_MPATH) || mrt->rt_priority != prio); 623 624 /* 625 * An ART node with the same destination/netmask already 626 * exists, MPATH conflict must have been already checked. 627 */ 628 if (rt->rt_flags & RTF_MPATH) { 629 /* 630 * Only keep the RTF_MPATH flag if two routes have 631 * the same gateway. 632 */ 633 rt->rt_flags &= ~RTF_MPATH; 634 SRPL_FOREACH_LOCKED(mrt, &an->an_rtlist, rt_next) { 635 if (mrt->rt_priority == prio) { 636 mrt->rt_flags |= RTF_MPATH; 637 rt->rt_flags |= RTF_MPATH; 638 } 639 } 640 } 641 642 /* Put newly inserted entry at the right place. */ 643 rtable_mpath_insert(an, rt); 644 } 645 leave: 646 rw_exit_write(&ar->ar_lock); 647 rtfree(rt); 648 return (error); 649 } 650 651 int 652 rtable_delete(unsigned int rtableid, struct sockaddr *dst, 653 struct sockaddr *mask, struct rtentry *rt) 654 { 655 struct art_root *ar; 656 struct art_node *an; 657 struct srp_ref sr; 658 uint8_t *addr; 659 int plen; 660 struct rtentry *mrt; 661 int npaths = 0; 662 int error = 0; 663 664 ar = rtable_get(rtableid, dst->sa_family); 665 if (ar == NULL) 666 return (EAFNOSUPPORT); 667 668 addr = satoaddr(ar, dst); 669 plen = rtable_satoplen(dst->sa_family, mask); 670 if (plen == -1) 671 return (EINVAL); 672 673 rtref(rt); /* guarantee rtfree won't do anything under ar_lock */ 674 rw_enter_write(&ar->ar_lock); 675 an = art_lookup(ar, addr, plen, &sr); 676 srp_leave(&sr); /* an can't go away while we have the lock */ 677 678 /* Make sure we've got a perfect match. */ 679 if (!an_match(an, dst, plen)) { 680 error = ESRCH; 681 goto leave; 682 } 683 684 /* 685 * If other multipath route entries are still attached to 686 * this ART node we only have to unlink it. 687 */ 688 SRPL_FOREACH_LOCKED(mrt, &an->an_rtlist, rt_next) 689 npaths++; 690 691 if (npaths > 1) { 692 KASSERT(refcnt_read(&rt->rt_refcnt) >= 1); 693 SRPL_REMOVE_LOCKED(&rt_rc, &an->an_rtlist, rt, rtentry, 694 rt_next); 695 696 mrt = SRPL_FIRST_LOCKED(&an->an_rtlist); 697 if (npaths == 2) 698 mrt->rt_flags &= ~RTF_MPATH; 699 700 goto leave; 701 } 702 703 if (art_delete(ar, an, addr, plen) == NULL) 704 panic("art_delete failed to find node %p", an); 705 706 KASSERT(refcnt_read(&rt->rt_refcnt) >= 1); 707 SRPL_REMOVE_LOCKED(&rt_rc, &an->an_rtlist, rt, rtentry, rt_next); 708 art_put(an); 709 710 leave: 711 rw_exit_write(&ar->ar_lock); 712 rtfree(rt); 713 714 return (error); 715 } 716 717 struct rtable_walk_cookie { 718 int (*rwc_func)(struct rtentry *, void *, unsigned int); 719 void *rwc_arg; 720 struct rtentry **rwc_prt; 721 unsigned int rwc_rid; 722 }; 723 724 /* 725 * Helper for rtable_walk to keep the ART code free from any "struct rtentry". 726 */ 727 int 728 rtable_walk_helper(struct art_node *an, void *xrwc) 729 { 730 struct srp_ref sr; 731 struct rtable_walk_cookie *rwc = xrwc; 732 struct rtentry *rt; 733 int error = 0; 734 735 SRPL_FOREACH(rt, &sr, &an->an_rtlist, rt_next) { 736 error = (*rwc->rwc_func)(rt, rwc->rwc_arg, rwc->rwc_rid); 737 if (error != 0) 738 break; 739 } 740 if (rwc->rwc_prt != NULL && rt != NULL) { 741 rtref(rt); 742 *rwc->rwc_prt = rt; 743 } 744 SRPL_LEAVE(&sr); 745 746 return (error); 747 } 748 749 int 750 rtable_walk(unsigned int rtableid, sa_family_t af, struct rtentry **prt, 751 int (*func)(struct rtentry *, void *, unsigned int), void *arg) 752 { 753 struct art_root *ar; 754 struct rtable_walk_cookie rwc; 755 int error; 756 757 ar = rtable_get(rtableid, af); 758 if (ar == NULL) 759 return (EAFNOSUPPORT); 760 761 rwc.rwc_func = func; 762 rwc.rwc_arg = arg; 763 rwc.rwc_prt = prt; 764 rwc.rwc_rid = rtableid; 765 766 error = art_walk(ar, rtable_walk_helper, &rwc); 767 768 return (error); 769 } 770 771 struct rtentry * 772 rtable_iterate(struct rtentry *rt0) 773 { 774 struct rtentry *rt = NULL; 775 struct srp_ref sr; 776 777 rt = SRPL_NEXT(&sr, rt0, rt_next); 778 if (rt != NULL) 779 rtref(rt); 780 SRPL_LEAVE(&sr); 781 rtfree(rt0); 782 return (rt); 783 } 784 785 int 786 rtable_mpath_capable(unsigned int rtableid, sa_family_t af) 787 { 788 return (1); 789 } 790 791 int 792 rtable_mpath_reprio(unsigned int rtableid, struct sockaddr *dst, 793 int plen, uint8_t prio, struct rtentry *rt) 794 { 795 struct art_root *ar; 796 struct art_node *an; 797 struct srp_ref sr; 798 uint8_t *addr; 799 int error = 0; 800 801 ar = rtable_get(rtableid, dst->sa_family); 802 if (ar == NULL) 803 return (EAFNOSUPPORT); 804 805 addr = satoaddr(ar, dst); 806 807 rw_enter_write(&ar->ar_lock); 808 an = art_lookup(ar, addr, plen, &sr); 809 srp_leave(&sr); /* an can't go away while we have the lock */ 810 811 /* Make sure we've got a perfect match. */ 812 if (!an_match(an, dst, plen)) { 813 error = ESRCH; 814 } else if (SRPL_FIRST_LOCKED(&an->an_rtlist) == rt && 815 SRPL_NEXT_LOCKED(rt, rt_next) == NULL) { 816 /* 817 * If there's only one entry on the list do not go 818 * through an insert/remove cycle. This is done to 819 * guarantee that ``an->an_rtlist'' is never empty 820 * when a node is in the tree. 821 */ 822 rt->rt_priority = prio; 823 } else { 824 rtref(rt); /* keep rt alive in between remove and insert */ 825 SRPL_REMOVE_LOCKED(&rt_rc, &an->an_rtlist, 826 rt, rtentry, rt_next); 827 rt->rt_priority = prio; 828 rtable_mpath_insert(an, rt); 829 rtfree(rt); 830 error = EAGAIN; 831 } 832 rw_exit_write(&ar->ar_lock); 833 834 return (error); 835 } 836 837 void 838 rtable_mpath_insert(struct art_node *an, struct rtentry *rt) 839 { 840 struct rtentry *mrt, *prt = NULL; 841 uint8_t prio = rt->rt_priority; 842 843 if ((mrt = SRPL_FIRST_LOCKED(&an->an_rtlist)) == NULL) { 844 SRPL_INSERT_HEAD_LOCKED(&rt_rc, &an->an_rtlist, rt, rt_next); 845 return; 846 } 847 848 /* Iterate until we find the route to be placed after ``rt''. */ 849 while (mrt->rt_priority <= prio && SRPL_NEXT_LOCKED(mrt, rt_next)) { 850 prt = mrt; 851 mrt = SRPL_NEXT_LOCKED(mrt, rt_next); 852 } 853 854 if (mrt->rt_priority <= prio) { 855 SRPL_INSERT_AFTER_LOCKED(&rt_rc, mrt, rt, rt_next); 856 } else if (prt != NULL) { 857 SRPL_INSERT_AFTER_LOCKED(&rt_rc, prt, rt, rt_next); 858 } else { 859 SRPL_INSERT_HEAD_LOCKED(&rt_rc, &an->an_rtlist, rt, rt_next); 860 } 861 } 862 863 /* 864 * Returns 1 if ``an'' perfectly matches (``dst'', ``plen''), 0 otherwise. 865 */ 866 int 867 an_match(struct art_node *an, struct sockaddr *dst, int plen) 868 { 869 struct rtentry *rt; 870 struct srp_ref sr; 871 int match; 872 873 if (an == NULL || an->an_plen != plen) 874 return (0); 875 876 rt = SRPL_FIRST(&sr, &an->an_rtlist); 877 match = (memcmp(rt->rt_dest, dst, dst->sa_len) == 0); 878 SRPL_LEAVE(&sr); 879 880 return (match); 881 } 882 883 void 884 rtentry_ref(void *null, void *xrt) 885 { 886 struct rtentry *rt = xrt; 887 888 rtref(rt); 889 } 890 891 void 892 rtentry_unref(void *null, void *xrt) 893 { 894 struct rtentry *rt = xrt; 895 896 rtfree(rt); 897 } 898 899 /* 900 * Return a pointer to the address (key). This is an heritage from the 901 * BSD radix tree needed to skip the non-address fields from the flavor 902 * of "struct sockaddr" used by this routing table. 903 */ 904 static inline uint8_t * 905 satoaddr(struct art_root *at, struct sockaddr *sa) 906 { 907 return (((uint8_t *)sa) + at->ar_off); 908 } 909 910 /* 911 * Return the prefix length of a mask. 912 */ 913 int 914 rtable_satoplen(sa_family_t af, struct sockaddr *mask) 915 { 916 const struct domain *dp; 917 uint8_t *ap, *ep; 918 int mlen, plen = 0; 919 int i; 920 921 for (i = 0; (dp = domains[i]) != NULL; i++) { 922 if (dp->dom_rtoffset == 0) 923 continue; 924 925 if (af == dp->dom_family) 926 break; 927 } 928 if (dp == NULL) 929 return (-1); 930 931 /* Host route */ 932 if (mask == NULL) 933 return (dp->dom_maxplen); 934 935 mlen = mask->sa_len; 936 937 /* Default route */ 938 if (mlen == 0) 939 return (0); 940 941 ap = (uint8_t *)((uint8_t *)mask) + dp->dom_rtoffset; 942 ep = (uint8_t *)((uint8_t *)mask) + mlen; 943 if (ap > ep) 944 return (-1); 945 946 /* Trim trailing zeroes. */ 947 while (ap < ep && ep[-1] == 0) 948 ep--; 949 950 if (ap == ep) 951 return (0); 952 953 /* "Beauty" adapted from sbin/route/show.c ... */ 954 while (ap < ep) { 955 switch (*ap++) { 956 case 0xff: 957 plen += 8; 958 break; 959 case 0xfe: 960 plen += 7; 961 goto out; 962 case 0xfc: 963 plen += 6; 964 goto out; 965 case 0xf8: 966 plen += 5; 967 goto out; 968 case 0xf0: 969 plen += 4; 970 goto out; 971 case 0xe0: 972 plen += 3; 973 goto out; 974 case 0xc0: 975 plen += 2; 976 goto out; 977 case 0x80: 978 plen += 1; 979 goto out; 980 default: 981 /* Non contiguous mask. */ 982 return (-1); 983 } 984 } 985 986 out: 987 if (plen > dp->dom_maxplen || ap != ep) 988 return -1; 989 990 return (plen); 991 } 992