1 /* 2 * This program is free software; you can redistribute it and/or 3 * modify it under the terms of the GNU General Public License 4 * as published by the Free Software Foundation; either version 5 * 2 of the License, or (at your option) any later version. 6 * 7 * Robert Olsson <robert.olsson@its.uu.se> Uppsala Universitet 8 * & Swedish University of Agricultural Sciences. 9 * 10 * Jens Laas <jens.laas@data.slu.se> Swedish University of 11 * Agricultural Sciences. 12 * 13 * Hans Liss <hans.liss@its.uu.se> Uppsala Universitet 14 * 15 * This work is based on the LPC-trie which is originally descibed in: 16 * 17 * An experimental study of compression methods for dynamic tries 18 * Stefan Nilsson and Matti Tikkanen. Algorithmica, 33(1):19-33, 2002. 19 * http://www.nada.kth.se/~snilsson/public/papers/dyntrie2/ 20 * 21 * 22 * IP-address lookup using LC-tries. Stefan Nilsson and Gunnar Karlsson 23 * IEEE Journal on Selected Areas in Communications, 17(6):1083-1092, June 1999 24 * 25 * Version: $Id: fib_trie.c,v 1.3 2005/06/08 14:20:01 robert Exp $ 26 * 27 * 28 * Code from fib_hash has been reused which includes the following header: 29 * 30 * 31 * INET An implementation of the TCP/IP protocol suite for the LINUX 32 * operating system. INET is implemented using the BSD Socket 33 * interface as the means of communication with the user level. 34 * 35 * IPv4 FIB: lookup engine and maintenance routines. 36 * 37 * 38 * Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru> 39 * 40 * This program is free software; you can redistribute it and/or 41 * modify it under the terms of the GNU General Public License 42 * as published by the Free Software Foundation; either version 43 * 2 of the License, or (at your option) any later version. 44 * 45 * Substantial contributions to this work comes from: 46 * 47 * David S. Miller, <davem@davemloft.net> 48 * Stephen Hemminger <shemminger@osdl.org> 49 * Paul E. McKenney <paulmck@us.ibm.com> 50 * Patrick McHardy <kaber@trash.net> 51 */ 52 53 #define VERSION "0.408" 54 55 #include <asm/uaccess.h> 56 #include <asm/system.h> 57 #include <asm/bitops.h> 58 #include <linux/types.h> 59 #include <linux/kernel.h> 60 #include <linux/mm.h> 61 #include <linux/string.h> 62 #include <linux/socket.h> 63 #include <linux/sockios.h> 64 #include <linux/errno.h> 65 #include <linux/in.h> 66 #include <linux/inet.h> 67 #include <linux/inetdevice.h> 68 #include <linux/netdevice.h> 69 #include <linux/if_arp.h> 70 #include <linux/proc_fs.h> 71 #include <linux/rcupdate.h> 72 #include <linux/skbuff.h> 73 #include <linux/netlink.h> 74 #include <linux/init.h> 75 #include <linux/list.h> 76 #include <net/ip.h> 77 #include <net/protocol.h> 78 #include <net/route.h> 79 #include <net/tcp.h> 80 #include <net/sock.h> 81 #include <net/ip_fib.h> 82 #include "fib_lookup.h" 83 84 #undef CONFIG_IP_FIB_TRIE_STATS 85 #define MAX_STAT_DEPTH 32 86 87 #define KEYLENGTH (8*sizeof(t_key)) 88 #define MASK_PFX(k, l) (((l)==0)?0:(k >> (KEYLENGTH-l)) << (KEYLENGTH-l)) 89 #define TKEY_GET_MASK(offset, bits) (((bits)==0)?0:((t_key)(-1) << (KEYLENGTH - bits) >> offset)) 90 91 typedef unsigned int t_key; 92 93 #define T_TNODE 0 94 #define T_LEAF 1 95 #define NODE_TYPE_MASK 0x1UL 96 #define NODE_PARENT(node) \ 97 ((struct tnode *)rcu_dereference(((node)->parent & ~NODE_TYPE_MASK))) 98 99 #define NODE_TYPE(node) ((node)->parent & NODE_TYPE_MASK) 100 101 #define NODE_SET_PARENT(node, ptr) \ 102 rcu_assign_pointer((node)->parent, \ 103 ((unsigned long)(ptr)) | NODE_TYPE(node)) 104 105 #define IS_TNODE(n) (!(n->parent & T_LEAF)) 106 #define IS_LEAF(n) (n->parent & T_LEAF) 107 108 struct node { 109 t_key key; 110 unsigned long parent; 111 }; 112 113 struct leaf { 114 t_key key; 115 unsigned long parent; 116 struct hlist_head list; 117 struct rcu_head rcu; 118 }; 119 120 struct leaf_info { 121 struct hlist_node hlist; 122 struct rcu_head rcu; 123 int plen; 124 struct list_head falh; 125 }; 126 127 struct tnode { 128 t_key key; 129 unsigned long parent; 130 unsigned short pos:5; /* 2log(KEYLENGTH) bits needed */ 131 unsigned short bits:5; /* 2log(KEYLENGTH) bits needed */ 132 unsigned short full_children; /* KEYLENGTH bits needed */ 133 unsigned short empty_children; /* KEYLENGTH bits needed */ 134 struct rcu_head rcu; 135 struct node *child[0]; 136 }; 137 138 #ifdef CONFIG_IP_FIB_TRIE_STATS 139 struct trie_use_stats { 140 unsigned int gets; 141 unsigned int backtrack; 142 unsigned int semantic_match_passed; 143 unsigned int semantic_match_miss; 144 unsigned int null_node_hit; 145 unsigned int resize_node_skipped; 146 }; 147 #endif 148 149 struct trie_stat { 150 unsigned int totdepth; 151 unsigned int maxdepth; 152 unsigned int tnodes; 153 unsigned int leaves; 154 unsigned int nullpointers; 155 unsigned int nodesizes[MAX_STAT_DEPTH]; 156 }; 157 158 struct trie { 159 struct node *trie; 160 #ifdef CONFIG_IP_FIB_TRIE_STATS 161 struct trie_use_stats stats; 162 #endif 163 int size; 164 unsigned int revision; 165 }; 166 167 static void put_child(struct trie *t, struct tnode *tn, int i, struct node *n); 168 static void tnode_put_child_reorg(struct tnode *tn, int i, struct node *n, int wasfull); 169 static struct node *resize(struct trie *t, struct tnode *tn); 170 static struct tnode *inflate(struct trie *t, struct tnode *tn); 171 static struct tnode *halve(struct trie *t, struct tnode *tn); 172 static void tnode_free(struct tnode *tn); 173 174 static struct kmem_cache *fn_alias_kmem __read_mostly; 175 static struct trie *trie_local = NULL, *trie_main = NULL; 176 177 178 /* rcu_read_lock needs to be hold by caller from readside */ 179 180 static inline struct node *tnode_get_child(struct tnode *tn, int i) 181 { 182 BUG_ON(i >= 1 << tn->bits); 183 184 return rcu_dereference(tn->child[i]); 185 } 186 187 static inline int tnode_child_length(const struct tnode *tn) 188 { 189 return 1 << tn->bits; 190 } 191 192 static inline t_key tkey_extract_bits(t_key a, int offset, int bits) 193 { 194 if (offset < KEYLENGTH) 195 return ((t_key)(a << offset)) >> (KEYLENGTH - bits); 196 else 197 return 0; 198 } 199 200 static inline int tkey_equals(t_key a, t_key b) 201 { 202 return a == b; 203 } 204 205 static inline int tkey_sub_equals(t_key a, int offset, int bits, t_key b) 206 { 207 if (bits == 0 || offset >= KEYLENGTH) 208 return 1; 209 bits = bits > KEYLENGTH ? KEYLENGTH : bits; 210 return ((a ^ b) << offset) >> (KEYLENGTH - bits) == 0; 211 } 212 213 static inline int tkey_mismatch(t_key a, int offset, t_key b) 214 { 215 t_key diff = a ^ b; 216 int i = offset; 217 218 if (!diff) 219 return 0; 220 while ((diff << i) >> (KEYLENGTH-1) == 0) 221 i++; 222 return i; 223 } 224 225 /* 226 To understand this stuff, an understanding of keys and all their bits is 227 necessary. Every node in the trie has a key associated with it, but not 228 all of the bits in that key are significant. 229 230 Consider a node 'n' and its parent 'tp'. 231 232 If n is a leaf, every bit in its key is significant. Its presence is 233 necessitated by path compression, since during a tree traversal (when 234 searching for a leaf - unless we are doing an insertion) we will completely 235 ignore all skipped bits we encounter. Thus we need to verify, at the end of 236 a potentially successful search, that we have indeed been walking the 237 correct key path. 238 239 Note that we can never "miss" the correct key in the tree if present by 240 following the wrong path. Path compression ensures that segments of the key 241 that are the same for all keys with a given prefix are skipped, but the 242 skipped part *is* identical for each node in the subtrie below the skipped 243 bit! trie_insert() in this implementation takes care of that - note the 244 call to tkey_sub_equals() in trie_insert(). 245 246 if n is an internal node - a 'tnode' here, the various parts of its key 247 have many different meanings. 248 249 Example: 250 _________________________________________________________________ 251 | i | i | i | i | i | i | i | N | N | N | S | S | S | S | S | C | 252 ----------------------------------------------------------------- 253 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 254 255 _________________________________________________________________ 256 | C | C | C | u | u | u | u | u | u | u | u | u | u | u | u | u | 257 ----------------------------------------------------------------- 258 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 259 260 tp->pos = 7 261 tp->bits = 3 262 n->pos = 15 263 n->bits = 4 264 265 First, let's just ignore the bits that come before the parent tp, that is 266 the bits from 0 to (tp->pos-1). They are *known* but at this point we do 267 not use them for anything. 268 269 The bits from (tp->pos) to (tp->pos + tp->bits - 1) - "N", above - are the 270 index into the parent's child array. That is, they will be used to find 271 'n' among tp's children. 272 273 The bits from (tp->pos + tp->bits) to (n->pos - 1) - "S" - are skipped bits 274 for the node n. 275 276 All the bits we have seen so far are significant to the node n. The rest 277 of the bits are really not needed or indeed known in n->key. 278 279 The bits from (n->pos) to (n->pos + n->bits - 1) - "C" - are the index into 280 n's child array, and will of course be different for each child. 281 282 283 The rest of the bits, from (n->pos + n->bits) onward, are completely unknown 284 at this point. 285 286 */ 287 288 static inline void check_tnode(const struct tnode *tn) 289 { 290 WARN_ON(tn && tn->pos+tn->bits > 32); 291 } 292 293 static int halve_threshold = 25; 294 static int inflate_threshold = 50; 295 static int halve_threshold_root = 8; 296 static int inflate_threshold_root = 15; 297 298 299 static void __alias_free_mem(struct rcu_head *head) 300 { 301 struct fib_alias *fa = container_of(head, struct fib_alias, rcu); 302 kmem_cache_free(fn_alias_kmem, fa); 303 } 304 305 static inline void alias_free_mem_rcu(struct fib_alias *fa) 306 { 307 call_rcu(&fa->rcu, __alias_free_mem); 308 } 309 310 static void __leaf_free_rcu(struct rcu_head *head) 311 { 312 kfree(container_of(head, struct leaf, rcu)); 313 } 314 315 static void __leaf_info_free_rcu(struct rcu_head *head) 316 { 317 kfree(container_of(head, struct leaf_info, rcu)); 318 } 319 320 static inline void free_leaf_info(struct leaf_info *leaf) 321 { 322 call_rcu(&leaf->rcu, __leaf_info_free_rcu); 323 } 324 325 static struct tnode *tnode_alloc(unsigned int size) 326 { 327 struct page *pages; 328 329 if (size <= PAGE_SIZE) 330 return kcalloc(size, 1, GFP_KERNEL); 331 332 pages = alloc_pages(GFP_KERNEL|__GFP_ZERO, get_order(size)); 333 if (!pages) 334 return NULL; 335 336 return page_address(pages); 337 } 338 339 static void __tnode_free_rcu(struct rcu_head *head) 340 { 341 struct tnode *tn = container_of(head, struct tnode, rcu); 342 unsigned int size = sizeof(struct tnode) + 343 (1 << tn->bits) * sizeof(struct node *); 344 345 if (size <= PAGE_SIZE) 346 kfree(tn); 347 else 348 free_pages((unsigned long)tn, get_order(size)); 349 } 350 351 static inline void tnode_free(struct tnode *tn) 352 { 353 if (IS_LEAF(tn)) { 354 struct leaf *l = (struct leaf *) tn; 355 call_rcu_bh(&l->rcu, __leaf_free_rcu); 356 } else 357 call_rcu(&tn->rcu, __tnode_free_rcu); 358 } 359 360 static struct leaf *leaf_new(void) 361 { 362 struct leaf *l = kmalloc(sizeof(struct leaf), GFP_KERNEL); 363 if (l) { 364 l->parent = T_LEAF; 365 INIT_HLIST_HEAD(&l->list); 366 } 367 return l; 368 } 369 370 static struct leaf_info *leaf_info_new(int plen) 371 { 372 struct leaf_info *li = kmalloc(sizeof(struct leaf_info), GFP_KERNEL); 373 if (li) { 374 li->plen = plen; 375 INIT_LIST_HEAD(&li->falh); 376 } 377 return li; 378 } 379 380 static struct tnode* tnode_new(t_key key, int pos, int bits) 381 { 382 int nchildren = 1<<bits; 383 int sz = sizeof(struct tnode) + nchildren * sizeof(struct node *); 384 struct tnode *tn = tnode_alloc(sz); 385 386 if (tn) { 387 memset(tn, 0, sz); 388 tn->parent = T_TNODE; 389 tn->pos = pos; 390 tn->bits = bits; 391 tn->key = key; 392 tn->full_children = 0; 393 tn->empty_children = 1<<bits; 394 } 395 396 pr_debug("AT %p s=%u %u\n", tn, (unsigned int) sizeof(struct tnode), 397 (unsigned int) (sizeof(struct node) * 1<<bits)); 398 return tn; 399 } 400 401 /* 402 * Check whether a tnode 'n' is "full", i.e. it is an internal node 403 * and no bits are skipped. See discussion in dyntree paper p. 6 404 */ 405 406 static inline int tnode_full(const struct tnode *tn, const struct node *n) 407 { 408 if (n == NULL || IS_LEAF(n)) 409 return 0; 410 411 return ((struct tnode *) n)->pos == tn->pos + tn->bits; 412 } 413 414 static inline void put_child(struct trie *t, struct tnode *tn, int i, struct node *n) 415 { 416 tnode_put_child_reorg(tn, i, n, -1); 417 } 418 419 /* 420 * Add a child at position i overwriting the old value. 421 * Update the value of full_children and empty_children. 422 */ 423 424 static void tnode_put_child_reorg(struct tnode *tn, int i, struct node *n, int wasfull) 425 { 426 struct node *chi = tn->child[i]; 427 int isfull; 428 429 BUG_ON(i >= 1<<tn->bits); 430 431 432 /* update emptyChildren */ 433 if (n == NULL && chi != NULL) 434 tn->empty_children++; 435 else if (n != NULL && chi == NULL) 436 tn->empty_children--; 437 438 /* update fullChildren */ 439 if (wasfull == -1) 440 wasfull = tnode_full(tn, chi); 441 442 isfull = tnode_full(tn, n); 443 if (wasfull && !isfull) 444 tn->full_children--; 445 else if (!wasfull && isfull) 446 tn->full_children++; 447 448 if (n) 449 NODE_SET_PARENT(n, tn); 450 451 rcu_assign_pointer(tn->child[i], n); 452 } 453 454 static struct node *resize(struct trie *t, struct tnode *tn) 455 { 456 int i; 457 int err = 0; 458 struct tnode *old_tn; 459 int inflate_threshold_use; 460 int halve_threshold_use; 461 int max_resize; 462 463 if (!tn) 464 return NULL; 465 466 pr_debug("In tnode_resize %p inflate_threshold=%d threshold=%d\n", 467 tn, inflate_threshold, halve_threshold); 468 469 /* No children */ 470 if (tn->empty_children == tnode_child_length(tn)) { 471 tnode_free(tn); 472 return NULL; 473 } 474 /* One child */ 475 if (tn->empty_children == tnode_child_length(tn) - 1) 476 for (i = 0; i < tnode_child_length(tn); i++) { 477 struct node *n; 478 479 n = tn->child[i]; 480 if (!n) 481 continue; 482 483 /* compress one level */ 484 NODE_SET_PARENT(n, NULL); 485 tnode_free(tn); 486 return n; 487 } 488 /* 489 * Double as long as the resulting node has a number of 490 * nonempty nodes that are above the threshold. 491 */ 492 493 /* 494 * From "Implementing a dynamic compressed trie" by Stefan Nilsson of 495 * the Helsinki University of Technology and Matti Tikkanen of Nokia 496 * Telecommunications, page 6: 497 * "A node is doubled if the ratio of non-empty children to all 498 * children in the *doubled* node is at least 'high'." 499 * 500 * 'high' in this instance is the variable 'inflate_threshold'. It 501 * is expressed as a percentage, so we multiply it with 502 * tnode_child_length() and instead of multiplying by 2 (since the 503 * child array will be doubled by inflate()) and multiplying 504 * the left-hand side by 100 (to handle the percentage thing) we 505 * multiply the left-hand side by 50. 506 * 507 * The left-hand side may look a bit weird: tnode_child_length(tn) 508 * - tn->empty_children is of course the number of non-null children 509 * in the current node. tn->full_children is the number of "full" 510 * children, that is non-null tnodes with a skip value of 0. 511 * All of those will be doubled in the resulting inflated tnode, so 512 * we just count them one extra time here. 513 * 514 * A clearer way to write this would be: 515 * 516 * to_be_doubled = tn->full_children; 517 * not_to_be_doubled = tnode_child_length(tn) - tn->empty_children - 518 * tn->full_children; 519 * 520 * new_child_length = tnode_child_length(tn) * 2; 521 * 522 * new_fill_factor = 100 * (not_to_be_doubled + 2*to_be_doubled) / 523 * new_child_length; 524 * if (new_fill_factor >= inflate_threshold) 525 * 526 * ...and so on, tho it would mess up the while () loop. 527 * 528 * anyway, 529 * 100 * (not_to_be_doubled + 2*to_be_doubled) / new_child_length >= 530 * inflate_threshold 531 * 532 * avoid a division: 533 * 100 * (not_to_be_doubled + 2*to_be_doubled) >= 534 * inflate_threshold * new_child_length 535 * 536 * expand not_to_be_doubled and to_be_doubled, and shorten: 537 * 100 * (tnode_child_length(tn) - tn->empty_children + 538 * tn->full_children) >= inflate_threshold * new_child_length 539 * 540 * expand new_child_length: 541 * 100 * (tnode_child_length(tn) - tn->empty_children + 542 * tn->full_children) >= 543 * inflate_threshold * tnode_child_length(tn) * 2 544 * 545 * shorten again: 546 * 50 * (tn->full_children + tnode_child_length(tn) - 547 * tn->empty_children) >= inflate_threshold * 548 * tnode_child_length(tn) 549 * 550 */ 551 552 check_tnode(tn); 553 554 /* Keep root node larger */ 555 556 if (!tn->parent) 557 inflate_threshold_use = inflate_threshold_root; 558 else 559 inflate_threshold_use = inflate_threshold; 560 561 err = 0; 562 max_resize = 10; 563 while ((tn->full_children > 0 && max_resize-- && 564 50 * (tn->full_children + tnode_child_length(tn) - tn->empty_children) >= 565 inflate_threshold_use * tnode_child_length(tn))) { 566 567 old_tn = tn; 568 tn = inflate(t, tn); 569 if (IS_ERR(tn)) { 570 tn = old_tn; 571 #ifdef CONFIG_IP_FIB_TRIE_STATS 572 t->stats.resize_node_skipped++; 573 #endif 574 break; 575 } 576 } 577 578 if (max_resize < 0) { 579 if (!tn->parent) 580 printk(KERN_WARNING "Fix inflate_threshold_root. Now=%d size=%d bits\n", 581 inflate_threshold_root, tn->bits); 582 else 583 printk(KERN_WARNING "Fix inflate_threshold. Now=%d size=%d bits\n", 584 inflate_threshold, tn->bits); 585 } 586 587 check_tnode(tn); 588 589 /* 590 * Halve as long as the number of empty children in this 591 * node is above threshold. 592 */ 593 594 595 /* Keep root node larger */ 596 597 if (!tn->parent) 598 halve_threshold_use = halve_threshold_root; 599 else 600 halve_threshold_use = halve_threshold; 601 602 err = 0; 603 max_resize = 10; 604 while (tn->bits > 1 && max_resize-- && 605 100 * (tnode_child_length(tn) - tn->empty_children) < 606 halve_threshold_use * tnode_child_length(tn)) { 607 608 old_tn = tn; 609 tn = halve(t, tn); 610 if (IS_ERR(tn)) { 611 tn = old_tn; 612 #ifdef CONFIG_IP_FIB_TRIE_STATS 613 t->stats.resize_node_skipped++; 614 #endif 615 break; 616 } 617 } 618 619 if (max_resize < 0) { 620 if (!tn->parent) 621 printk(KERN_WARNING "Fix halve_threshold_root. Now=%d size=%d bits\n", 622 halve_threshold_root, tn->bits); 623 else 624 printk(KERN_WARNING "Fix halve_threshold. Now=%d size=%d bits\n", 625 halve_threshold, tn->bits); 626 } 627 628 /* Only one child remains */ 629 if (tn->empty_children == tnode_child_length(tn) - 1) 630 for (i = 0; i < tnode_child_length(tn); i++) { 631 struct node *n; 632 633 n = tn->child[i]; 634 if (!n) 635 continue; 636 637 /* compress one level */ 638 639 NODE_SET_PARENT(n, NULL); 640 tnode_free(tn); 641 return n; 642 } 643 644 return (struct node *) tn; 645 } 646 647 static struct tnode *inflate(struct trie *t, struct tnode *tn) 648 { 649 struct tnode *inode; 650 struct tnode *oldtnode = tn; 651 int olen = tnode_child_length(tn); 652 int i; 653 654 pr_debug("In inflate\n"); 655 656 tn = tnode_new(oldtnode->key, oldtnode->pos, oldtnode->bits + 1); 657 658 if (!tn) 659 return ERR_PTR(-ENOMEM); 660 661 /* 662 * Preallocate and store tnodes before the actual work so we 663 * don't get into an inconsistent state if memory allocation 664 * fails. In case of failure we return the oldnode and inflate 665 * of tnode is ignored. 666 */ 667 668 for (i = 0; i < olen; i++) { 669 struct tnode *inode = (struct tnode *) tnode_get_child(oldtnode, i); 670 671 if (inode && 672 IS_TNODE(inode) && 673 inode->pos == oldtnode->pos + oldtnode->bits && 674 inode->bits > 1) { 675 struct tnode *left, *right; 676 t_key m = TKEY_GET_MASK(inode->pos, 1); 677 678 left = tnode_new(inode->key&(~m), inode->pos + 1, 679 inode->bits - 1); 680 if (!left) 681 goto nomem; 682 683 right = tnode_new(inode->key|m, inode->pos + 1, 684 inode->bits - 1); 685 686 if (!right) { 687 tnode_free(left); 688 goto nomem; 689 } 690 691 put_child(t, tn, 2*i, (struct node *) left); 692 put_child(t, tn, 2*i+1, (struct node *) right); 693 } 694 } 695 696 for (i = 0; i < olen; i++) { 697 struct node *node = tnode_get_child(oldtnode, i); 698 struct tnode *left, *right; 699 int size, j; 700 701 /* An empty child */ 702 if (node == NULL) 703 continue; 704 705 /* A leaf or an internal node with skipped bits */ 706 707 if (IS_LEAF(node) || ((struct tnode *) node)->pos > 708 tn->pos + tn->bits - 1) { 709 if (tkey_extract_bits(node->key, oldtnode->pos + oldtnode->bits, 710 1) == 0) 711 put_child(t, tn, 2*i, node); 712 else 713 put_child(t, tn, 2*i+1, node); 714 continue; 715 } 716 717 /* An internal node with two children */ 718 inode = (struct tnode *) node; 719 720 if (inode->bits == 1) { 721 put_child(t, tn, 2*i, inode->child[0]); 722 put_child(t, tn, 2*i+1, inode->child[1]); 723 724 tnode_free(inode); 725 continue; 726 } 727 728 /* An internal node with more than two children */ 729 730 /* We will replace this node 'inode' with two new 731 * ones, 'left' and 'right', each with half of the 732 * original children. The two new nodes will have 733 * a position one bit further down the key and this 734 * means that the "significant" part of their keys 735 * (see the discussion near the top of this file) 736 * will differ by one bit, which will be "0" in 737 * left's key and "1" in right's key. Since we are 738 * moving the key position by one step, the bit that 739 * we are moving away from - the bit at position 740 * (inode->pos) - is the one that will differ between 741 * left and right. So... we synthesize that bit in the 742 * two new keys. 743 * The mask 'm' below will be a single "one" bit at 744 * the position (inode->pos) 745 */ 746 747 /* Use the old key, but set the new significant 748 * bit to zero. 749 */ 750 751 left = (struct tnode *) tnode_get_child(tn, 2*i); 752 put_child(t, tn, 2*i, NULL); 753 754 BUG_ON(!left); 755 756 right = (struct tnode *) tnode_get_child(tn, 2*i+1); 757 put_child(t, tn, 2*i+1, NULL); 758 759 BUG_ON(!right); 760 761 size = tnode_child_length(left); 762 for (j = 0; j < size; j++) { 763 put_child(t, left, j, inode->child[j]); 764 put_child(t, right, j, inode->child[j + size]); 765 } 766 put_child(t, tn, 2*i, resize(t, left)); 767 put_child(t, tn, 2*i+1, resize(t, right)); 768 769 tnode_free(inode); 770 } 771 tnode_free(oldtnode); 772 return tn; 773 nomem: 774 { 775 int size = tnode_child_length(tn); 776 int j; 777 778 for (j = 0; j < size; j++) 779 if (tn->child[j]) 780 tnode_free((struct tnode *)tn->child[j]); 781 782 tnode_free(tn); 783 784 return ERR_PTR(-ENOMEM); 785 } 786 } 787 788 static struct tnode *halve(struct trie *t, struct tnode *tn) 789 { 790 struct tnode *oldtnode = tn; 791 struct node *left, *right; 792 int i; 793 int olen = tnode_child_length(tn); 794 795 pr_debug("In halve\n"); 796 797 tn = tnode_new(oldtnode->key, oldtnode->pos, oldtnode->bits - 1); 798 799 if (!tn) 800 return ERR_PTR(-ENOMEM); 801 802 /* 803 * Preallocate and store tnodes before the actual work so we 804 * don't get into an inconsistent state if memory allocation 805 * fails. In case of failure we return the oldnode and halve 806 * of tnode is ignored. 807 */ 808 809 for (i = 0; i < olen; i += 2) { 810 left = tnode_get_child(oldtnode, i); 811 right = tnode_get_child(oldtnode, i+1); 812 813 /* Two nonempty children */ 814 if (left && right) { 815 struct tnode *newn; 816 817 newn = tnode_new(left->key, tn->pos + tn->bits, 1); 818 819 if (!newn) 820 goto nomem; 821 822 put_child(t, tn, i/2, (struct node *)newn); 823 } 824 825 } 826 827 for (i = 0; i < olen; i += 2) { 828 struct tnode *newBinNode; 829 830 left = tnode_get_child(oldtnode, i); 831 right = tnode_get_child(oldtnode, i+1); 832 833 /* At least one of the children is empty */ 834 if (left == NULL) { 835 if (right == NULL) /* Both are empty */ 836 continue; 837 put_child(t, tn, i/2, right); 838 continue; 839 } 840 841 if (right == NULL) { 842 put_child(t, tn, i/2, left); 843 continue; 844 } 845 846 /* Two nonempty children */ 847 newBinNode = (struct tnode *) tnode_get_child(tn, i/2); 848 put_child(t, tn, i/2, NULL); 849 put_child(t, newBinNode, 0, left); 850 put_child(t, newBinNode, 1, right); 851 put_child(t, tn, i/2, resize(t, newBinNode)); 852 } 853 tnode_free(oldtnode); 854 return tn; 855 nomem: 856 { 857 int size = tnode_child_length(tn); 858 int j; 859 860 for (j = 0; j < size; j++) 861 if (tn->child[j]) 862 tnode_free((struct tnode *)tn->child[j]); 863 864 tnode_free(tn); 865 866 return ERR_PTR(-ENOMEM); 867 } 868 } 869 870 static void trie_init(struct trie *t) 871 { 872 if (!t) 873 return; 874 875 t->size = 0; 876 rcu_assign_pointer(t->trie, NULL); 877 t->revision = 0; 878 #ifdef CONFIG_IP_FIB_TRIE_STATS 879 memset(&t->stats, 0, sizeof(struct trie_use_stats)); 880 #endif 881 } 882 883 /* readside must use rcu_read_lock currently dump routines 884 via get_fa_head and dump */ 885 886 static struct leaf_info *find_leaf_info(struct leaf *l, int plen) 887 { 888 struct hlist_head *head = &l->list; 889 struct hlist_node *node; 890 struct leaf_info *li; 891 892 hlist_for_each_entry_rcu(li, node, head, hlist) 893 if (li->plen == plen) 894 return li; 895 896 return NULL; 897 } 898 899 static inline struct list_head * get_fa_head(struct leaf *l, int plen) 900 { 901 struct leaf_info *li = find_leaf_info(l, plen); 902 903 if (!li) 904 return NULL; 905 906 return &li->falh; 907 } 908 909 static void insert_leaf_info(struct hlist_head *head, struct leaf_info *new) 910 { 911 struct leaf_info *li = NULL, *last = NULL; 912 struct hlist_node *node; 913 914 if (hlist_empty(head)) { 915 hlist_add_head_rcu(&new->hlist, head); 916 } else { 917 hlist_for_each_entry(li, node, head, hlist) { 918 if (new->plen > li->plen) 919 break; 920 921 last = li; 922 } 923 if (last) 924 hlist_add_after_rcu(&last->hlist, &new->hlist); 925 else 926 hlist_add_before_rcu(&new->hlist, &li->hlist); 927 } 928 } 929 930 /* rcu_read_lock needs to be hold by caller from readside */ 931 932 static struct leaf * 933 fib_find_node(struct trie *t, u32 key) 934 { 935 int pos; 936 struct tnode *tn; 937 struct node *n; 938 939 pos = 0; 940 n = rcu_dereference(t->trie); 941 942 while (n != NULL && NODE_TYPE(n) == T_TNODE) { 943 tn = (struct tnode *) n; 944 945 check_tnode(tn); 946 947 if (tkey_sub_equals(tn->key, pos, tn->pos-pos, key)) { 948 pos = tn->pos + tn->bits; 949 n = tnode_get_child(tn, tkey_extract_bits(key, tn->pos, tn->bits)); 950 } else 951 break; 952 } 953 /* Case we have found a leaf. Compare prefixes */ 954 955 if (n != NULL && IS_LEAF(n) && tkey_equals(key, n->key)) 956 return (struct leaf *)n; 957 958 return NULL; 959 } 960 961 static struct node *trie_rebalance(struct trie *t, struct tnode *tn) 962 { 963 int wasfull; 964 t_key cindex, key; 965 struct tnode *tp = NULL; 966 967 key = tn->key; 968 969 while (tn != NULL && NODE_PARENT(tn) != NULL) { 970 971 tp = NODE_PARENT(tn); 972 cindex = tkey_extract_bits(key, tp->pos, tp->bits); 973 wasfull = tnode_full(tp, tnode_get_child(tp, cindex)); 974 tn = (struct tnode *) resize (t, (struct tnode *)tn); 975 tnode_put_child_reorg((struct tnode *)tp, cindex,(struct node*)tn, wasfull); 976 977 if (!NODE_PARENT(tn)) 978 break; 979 980 tn = NODE_PARENT(tn); 981 } 982 /* Handle last (top) tnode */ 983 if (IS_TNODE(tn)) 984 tn = (struct tnode*) resize(t, (struct tnode *)tn); 985 986 return (struct node*) tn; 987 } 988 989 /* only used from updater-side */ 990 991 static struct list_head * 992 fib_insert_node(struct trie *t, int *err, u32 key, int plen) 993 { 994 int pos, newpos; 995 struct tnode *tp = NULL, *tn = NULL; 996 struct node *n; 997 struct leaf *l; 998 int missbit; 999 struct list_head *fa_head = NULL; 1000 struct leaf_info *li; 1001 t_key cindex; 1002 1003 pos = 0; 1004 n = t->trie; 1005 1006 /* If we point to NULL, stop. Either the tree is empty and we should 1007 * just put a new leaf in if, or we have reached an empty child slot, 1008 * and we should just put our new leaf in that. 1009 * If we point to a T_TNODE, check if it matches our key. Note that 1010 * a T_TNODE might be skipping any number of bits - its 'pos' need 1011 * not be the parent's 'pos'+'bits'! 1012 * 1013 * If it does match the current key, get pos/bits from it, extract 1014 * the index from our key, push the T_TNODE and walk the tree. 1015 * 1016 * If it doesn't, we have to replace it with a new T_TNODE. 1017 * 1018 * If we point to a T_LEAF, it might or might not have the same key 1019 * as we do. If it does, just change the value, update the T_LEAF's 1020 * value, and return it. 1021 * If it doesn't, we need to replace it with a T_TNODE. 1022 */ 1023 1024 while (n != NULL && NODE_TYPE(n) == T_TNODE) { 1025 tn = (struct tnode *) n; 1026 1027 check_tnode(tn); 1028 1029 if (tkey_sub_equals(tn->key, pos, tn->pos-pos, key)) { 1030 tp = tn; 1031 pos = tn->pos + tn->bits; 1032 n = tnode_get_child(tn, tkey_extract_bits(key, tn->pos, tn->bits)); 1033 1034 BUG_ON(n && NODE_PARENT(n) != tn); 1035 } else 1036 break; 1037 } 1038 1039 /* 1040 * n ----> NULL, LEAF or TNODE 1041 * 1042 * tp is n's (parent) ----> NULL or TNODE 1043 */ 1044 1045 BUG_ON(tp && IS_LEAF(tp)); 1046 1047 /* Case 1: n is a leaf. Compare prefixes */ 1048 1049 if (n != NULL && IS_LEAF(n) && tkey_equals(key, n->key)) { 1050 struct leaf *l = (struct leaf *) n; 1051 1052 li = leaf_info_new(plen); 1053 1054 if (!li) { 1055 *err = -ENOMEM; 1056 goto err; 1057 } 1058 1059 fa_head = &li->falh; 1060 insert_leaf_info(&l->list, li); 1061 goto done; 1062 } 1063 t->size++; 1064 l = leaf_new(); 1065 1066 if (!l) { 1067 *err = -ENOMEM; 1068 goto err; 1069 } 1070 1071 l->key = key; 1072 li = leaf_info_new(plen); 1073 1074 if (!li) { 1075 tnode_free((struct tnode *) l); 1076 *err = -ENOMEM; 1077 goto err; 1078 } 1079 1080 fa_head = &li->falh; 1081 insert_leaf_info(&l->list, li); 1082 1083 if (t->trie && n == NULL) { 1084 /* Case 2: n is NULL, and will just insert a new leaf */ 1085 1086 NODE_SET_PARENT(l, tp); 1087 1088 cindex = tkey_extract_bits(key, tp->pos, tp->bits); 1089 put_child(t, (struct tnode *)tp, cindex, (struct node *)l); 1090 } else { 1091 /* Case 3: n is a LEAF or a TNODE and the key doesn't match. */ 1092 /* 1093 * Add a new tnode here 1094 * first tnode need some special handling 1095 */ 1096 1097 if (tp) 1098 pos = tp->pos+tp->bits; 1099 else 1100 pos = 0; 1101 1102 if (n) { 1103 newpos = tkey_mismatch(key, pos, n->key); 1104 tn = tnode_new(n->key, newpos, 1); 1105 } else { 1106 newpos = 0; 1107 tn = tnode_new(key, newpos, 1); /* First tnode */ 1108 } 1109 1110 if (!tn) { 1111 free_leaf_info(li); 1112 tnode_free((struct tnode *) l); 1113 *err = -ENOMEM; 1114 goto err; 1115 } 1116 1117 NODE_SET_PARENT(tn, tp); 1118 1119 missbit = tkey_extract_bits(key, newpos, 1); 1120 put_child(t, tn, missbit, (struct node *)l); 1121 put_child(t, tn, 1-missbit, n); 1122 1123 if (tp) { 1124 cindex = tkey_extract_bits(key, tp->pos, tp->bits); 1125 put_child(t, (struct tnode *)tp, cindex, (struct node *)tn); 1126 } else { 1127 rcu_assign_pointer(t->trie, (struct node *)tn); /* First tnode */ 1128 tp = tn; 1129 } 1130 } 1131 1132 if (tp && tp->pos + tp->bits > 32) 1133 printk(KERN_WARNING "fib_trie tp=%p pos=%d, bits=%d, key=%0x plen=%d\n", 1134 tp, tp->pos, tp->bits, key, plen); 1135 1136 /* Rebalance the trie */ 1137 1138 rcu_assign_pointer(t->trie, trie_rebalance(t, tp)); 1139 done: 1140 t->revision++; 1141 err: 1142 return fa_head; 1143 } 1144 1145 /* 1146 * Caller must hold RTNL. 1147 */ 1148 static int fn_trie_insert(struct fib_table *tb, struct fib_config *cfg) 1149 { 1150 struct trie *t = (struct trie *) tb->tb_data; 1151 struct fib_alias *fa, *new_fa; 1152 struct list_head *fa_head = NULL; 1153 struct fib_info *fi; 1154 int plen = cfg->fc_dst_len; 1155 u8 tos = cfg->fc_tos; 1156 u32 key, mask; 1157 int err; 1158 struct leaf *l; 1159 1160 if (plen > 32) 1161 return -EINVAL; 1162 1163 key = ntohl(cfg->fc_dst); 1164 1165 pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen); 1166 1167 mask = ntohl(inet_make_mask(plen)); 1168 1169 if (key & ~mask) 1170 return -EINVAL; 1171 1172 key = key & mask; 1173 1174 fi = fib_create_info(cfg); 1175 if (IS_ERR(fi)) { 1176 err = PTR_ERR(fi); 1177 goto err; 1178 } 1179 1180 l = fib_find_node(t, key); 1181 fa = NULL; 1182 1183 if (l) { 1184 fa_head = get_fa_head(l, plen); 1185 fa = fib_find_alias(fa_head, tos, fi->fib_priority); 1186 } 1187 1188 /* Now fa, if non-NULL, points to the first fib alias 1189 * with the same keys [prefix,tos,priority], if such key already 1190 * exists or to the node before which we will insert new one. 1191 * 1192 * If fa is NULL, we will need to allocate a new one and 1193 * insert to the head of f. 1194 * 1195 * If f is NULL, no fib node matched the destination key 1196 * and we need to allocate a new one of those as well. 1197 */ 1198 1199 if (fa && fa->fa_info->fib_priority == fi->fib_priority) { 1200 struct fib_alias *fa_orig; 1201 1202 err = -EEXIST; 1203 if (cfg->fc_nlflags & NLM_F_EXCL) 1204 goto out; 1205 1206 if (cfg->fc_nlflags & NLM_F_REPLACE) { 1207 struct fib_info *fi_drop; 1208 u8 state; 1209 1210 err = -ENOBUFS; 1211 new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL); 1212 if (new_fa == NULL) 1213 goto out; 1214 1215 fi_drop = fa->fa_info; 1216 new_fa->fa_tos = fa->fa_tos; 1217 new_fa->fa_info = fi; 1218 new_fa->fa_type = cfg->fc_type; 1219 new_fa->fa_scope = cfg->fc_scope; 1220 state = fa->fa_state; 1221 new_fa->fa_state &= ~FA_S_ACCESSED; 1222 1223 list_replace_rcu(&fa->fa_list, &new_fa->fa_list); 1224 alias_free_mem_rcu(fa); 1225 1226 fib_release_info(fi_drop); 1227 if (state & FA_S_ACCESSED) 1228 rt_cache_flush(-1); 1229 rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, 1230 tb->tb_id, &cfg->fc_nlinfo, NLM_F_REPLACE); 1231 1232 goto succeeded; 1233 } 1234 /* Error if we find a perfect match which 1235 * uses the same scope, type, and nexthop 1236 * information. 1237 */ 1238 fa_orig = fa; 1239 list_for_each_entry(fa, fa_orig->fa_list.prev, fa_list) { 1240 if (fa->fa_tos != tos) 1241 break; 1242 if (fa->fa_info->fib_priority != fi->fib_priority) 1243 break; 1244 if (fa->fa_type == cfg->fc_type && 1245 fa->fa_scope == cfg->fc_scope && 1246 fa->fa_info == fi) { 1247 goto out; 1248 } 1249 } 1250 if (!(cfg->fc_nlflags & NLM_F_APPEND)) 1251 fa = fa_orig; 1252 } 1253 err = -ENOENT; 1254 if (!(cfg->fc_nlflags & NLM_F_CREATE)) 1255 goto out; 1256 1257 err = -ENOBUFS; 1258 new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL); 1259 if (new_fa == NULL) 1260 goto out; 1261 1262 new_fa->fa_info = fi; 1263 new_fa->fa_tos = tos; 1264 new_fa->fa_type = cfg->fc_type; 1265 new_fa->fa_scope = cfg->fc_scope; 1266 new_fa->fa_state = 0; 1267 /* 1268 * Insert new entry to the list. 1269 */ 1270 1271 if (!fa_head) { 1272 err = 0; 1273 fa_head = fib_insert_node(t, &err, key, plen); 1274 if (err) 1275 goto out_free_new_fa; 1276 } 1277 1278 list_add_tail_rcu(&new_fa->fa_list, 1279 (fa ? &fa->fa_list : fa_head)); 1280 1281 rt_cache_flush(-1); 1282 rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, 1283 &cfg->fc_nlinfo, 0); 1284 succeeded: 1285 return 0; 1286 1287 out_free_new_fa: 1288 kmem_cache_free(fn_alias_kmem, new_fa); 1289 out: 1290 fib_release_info(fi); 1291 err: 1292 return err; 1293 } 1294 1295 1296 /* should be called with rcu_read_lock */ 1297 static inline int check_leaf(struct trie *t, struct leaf *l, 1298 t_key key, int *plen, const struct flowi *flp, 1299 struct fib_result *res) 1300 { 1301 int err, i; 1302 __be32 mask; 1303 struct leaf_info *li; 1304 struct hlist_head *hhead = &l->list; 1305 struct hlist_node *node; 1306 1307 hlist_for_each_entry_rcu(li, node, hhead, hlist) { 1308 i = li->plen; 1309 mask = inet_make_mask(i); 1310 if (l->key != (key & ntohl(mask))) 1311 continue; 1312 1313 if ((err = fib_semantic_match(&li->falh, flp, res, htonl(l->key), mask, i)) <= 0) { 1314 *plen = i; 1315 #ifdef CONFIG_IP_FIB_TRIE_STATS 1316 t->stats.semantic_match_passed++; 1317 #endif 1318 return err; 1319 } 1320 #ifdef CONFIG_IP_FIB_TRIE_STATS 1321 t->stats.semantic_match_miss++; 1322 #endif 1323 } 1324 return 1; 1325 } 1326 1327 static int 1328 fn_trie_lookup(struct fib_table *tb, const struct flowi *flp, struct fib_result *res) 1329 { 1330 struct trie *t = (struct trie *) tb->tb_data; 1331 int plen, ret = 0; 1332 struct node *n; 1333 struct tnode *pn; 1334 int pos, bits; 1335 t_key key = ntohl(flp->fl4_dst); 1336 int chopped_off; 1337 t_key cindex = 0; 1338 int current_prefix_length = KEYLENGTH; 1339 struct tnode *cn; 1340 t_key node_prefix, key_prefix, pref_mismatch; 1341 int mp; 1342 1343 rcu_read_lock(); 1344 1345 n = rcu_dereference(t->trie); 1346 if (!n) 1347 goto failed; 1348 1349 #ifdef CONFIG_IP_FIB_TRIE_STATS 1350 t->stats.gets++; 1351 #endif 1352 1353 /* Just a leaf? */ 1354 if (IS_LEAF(n)) { 1355 if ((ret = check_leaf(t, (struct leaf *)n, key, &plen, flp, res)) <= 0) 1356 goto found; 1357 goto failed; 1358 } 1359 pn = (struct tnode *) n; 1360 chopped_off = 0; 1361 1362 while (pn) { 1363 pos = pn->pos; 1364 bits = pn->bits; 1365 1366 if (!chopped_off) 1367 cindex = tkey_extract_bits(MASK_PFX(key, current_prefix_length), pos, bits); 1368 1369 n = tnode_get_child(pn, cindex); 1370 1371 if (n == NULL) { 1372 #ifdef CONFIG_IP_FIB_TRIE_STATS 1373 t->stats.null_node_hit++; 1374 #endif 1375 goto backtrace; 1376 } 1377 1378 if (IS_LEAF(n)) { 1379 if ((ret = check_leaf(t, (struct leaf *)n, key, &plen, flp, res)) <= 0) 1380 goto found; 1381 else 1382 goto backtrace; 1383 } 1384 1385 #define HL_OPTIMIZE 1386 #ifdef HL_OPTIMIZE 1387 cn = (struct tnode *)n; 1388 1389 /* 1390 * It's a tnode, and we can do some extra checks here if we 1391 * like, to avoid descending into a dead-end branch. 1392 * This tnode is in the parent's child array at index 1393 * key[p_pos..p_pos+p_bits] but potentially with some bits 1394 * chopped off, so in reality the index may be just a 1395 * subprefix, padded with zero at the end. 1396 * We can also take a look at any skipped bits in this 1397 * tnode - everything up to p_pos is supposed to be ok, 1398 * and the non-chopped bits of the index (se previous 1399 * paragraph) are also guaranteed ok, but the rest is 1400 * considered unknown. 1401 * 1402 * The skipped bits are key[pos+bits..cn->pos]. 1403 */ 1404 1405 /* If current_prefix_length < pos+bits, we are already doing 1406 * actual prefix matching, which means everything from 1407 * pos+(bits-chopped_off) onward must be zero along some 1408 * branch of this subtree - otherwise there is *no* valid 1409 * prefix present. Here we can only check the skipped 1410 * bits. Remember, since we have already indexed into the 1411 * parent's child array, we know that the bits we chopped of 1412 * *are* zero. 1413 */ 1414 1415 /* NOTA BENE: CHECKING ONLY SKIPPED BITS FOR THE NEW NODE HERE */ 1416 1417 if (current_prefix_length < pos+bits) { 1418 if (tkey_extract_bits(cn->key, current_prefix_length, 1419 cn->pos - current_prefix_length) != 0 || 1420 !(cn->child[0])) 1421 goto backtrace; 1422 } 1423 1424 /* 1425 * If chopped_off=0, the index is fully validated and we 1426 * only need to look at the skipped bits for this, the new, 1427 * tnode. What we actually want to do is to find out if 1428 * these skipped bits match our key perfectly, or if we will 1429 * have to count on finding a matching prefix further down, 1430 * because if we do, we would like to have some way of 1431 * verifying the existence of such a prefix at this point. 1432 */ 1433 1434 /* The only thing we can do at this point is to verify that 1435 * any such matching prefix can indeed be a prefix to our 1436 * key, and if the bits in the node we are inspecting that 1437 * do not match our key are not ZERO, this cannot be true. 1438 * Thus, find out where there is a mismatch (before cn->pos) 1439 * and verify that all the mismatching bits are zero in the 1440 * new tnode's key. 1441 */ 1442 1443 /* Note: We aren't very concerned about the piece of the key 1444 * that precede pn->pos+pn->bits, since these have already been 1445 * checked. The bits after cn->pos aren't checked since these are 1446 * by definition "unknown" at this point. Thus, what we want to 1447 * see is if we are about to enter the "prefix matching" state, 1448 * and in that case verify that the skipped bits that will prevail 1449 * throughout this subtree are zero, as they have to be if we are 1450 * to find a matching prefix. 1451 */ 1452 1453 node_prefix = MASK_PFX(cn->key, cn->pos); 1454 key_prefix = MASK_PFX(key, cn->pos); 1455 pref_mismatch = key_prefix^node_prefix; 1456 mp = 0; 1457 1458 /* In short: If skipped bits in this node do not match the search 1459 * key, enter the "prefix matching" state.directly. 1460 */ 1461 if (pref_mismatch) { 1462 while (!(pref_mismatch & (1<<(KEYLENGTH-1)))) { 1463 mp++; 1464 pref_mismatch = pref_mismatch <<1; 1465 } 1466 key_prefix = tkey_extract_bits(cn->key, mp, cn->pos-mp); 1467 1468 if (key_prefix != 0) 1469 goto backtrace; 1470 1471 if (current_prefix_length >= cn->pos) 1472 current_prefix_length = mp; 1473 } 1474 #endif 1475 pn = (struct tnode *)n; /* Descend */ 1476 chopped_off = 0; 1477 continue; 1478 1479 backtrace: 1480 chopped_off++; 1481 1482 /* As zero don't change the child key (cindex) */ 1483 while ((chopped_off <= pn->bits) && !(cindex & (1<<(chopped_off-1)))) 1484 chopped_off++; 1485 1486 /* Decrease current_... with bits chopped off */ 1487 if (current_prefix_length > pn->pos + pn->bits - chopped_off) 1488 current_prefix_length = pn->pos + pn->bits - chopped_off; 1489 1490 /* 1491 * Either we do the actual chop off according or if we have 1492 * chopped off all bits in this tnode walk up to our parent. 1493 */ 1494 1495 if (chopped_off <= pn->bits) { 1496 cindex &= ~(1 << (chopped_off-1)); 1497 } else { 1498 if (NODE_PARENT(pn) == NULL) 1499 goto failed; 1500 1501 /* Get Child's index */ 1502 cindex = tkey_extract_bits(pn->key, NODE_PARENT(pn)->pos, NODE_PARENT(pn)->bits); 1503 pn = NODE_PARENT(pn); 1504 chopped_off = 0; 1505 1506 #ifdef CONFIG_IP_FIB_TRIE_STATS 1507 t->stats.backtrack++; 1508 #endif 1509 goto backtrace; 1510 } 1511 } 1512 failed: 1513 ret = 1; 1514 found: 1515 rcu_read_unlock(); 1516 return ret; 1517 } 1518 1519 /* only called from updater side */ 1520 static int trie_leaf_remove(struct trie *t, t_key key) 1521 { 1522 t_key cindex; 1523 struct tnode *tp = NULL; 1524 struct node *n = t->trie; 1525 struct leaf *l; 1526 1527 pr_debug("entering trie_leaf_remove(%p)\n", n); 1528 1529 /* Note that in the case skipped bits, those bits are *not* checked! 1530 * When we finish this, we will have NULL or a T_LEAF, and the 1531 * T_LEAF may or may not match our key. 1532 */ 1533 1534 while (n != NULL && IS_TNODE(n)) { 1535 struct tnode *tn = (struct tnode *) n; 1536 check_tnode(tn); 1537 n = tnode_get_child(tn ,tkey_extract_bits(key, tn->pos, tn->bits)); 1538 1539 BUG_ON(n && NODE_PARENT(n) != tn); 1540 } 1541 l = (struct leaf *) n; 1542 1543 if (!n || !tkey_equals(l->key, key)) 1544 return 0; 1545 1546 /* 1547 * Key found. 1548 * Remove the leaf and rebalance the tree 1549 */ 1550 1551 t->revision++; 1552 t->size--; 1553 1554 tp = NODE_PARENT(n); 1555 tnode_free((struct tnode *) n); 1556 1557 if (tp) { 1558 cindex = tkey_extract_bits(key, tp->pos, tp->bits); 1559 put_child(t, (struct tnode *)tp, cindex, NULL); 1560 rcu_assign_pointer(t->trie, trie_rebalance(t, tp)); 1561 } else 1562 rcu_assign_pointer(t->trie, NULL); 1563 1564 return 1; 1565 } 1566 1567 /* 1568 * Caller must hold RTNL. 1569 */ 1570 static int fn_trie_delete(struct fib_table *tb, struct fib_config *cfg) 1571 { 1572 struct trie *t = (struct trie *) tb->tb_data; 1573 u32 key, mask; 1574 int plen = cfg->fc_dst_len; 1575 u8 tos = cfg->fc_tos; 1576 struct fib_alias *fa, *fa_to_delete; 1577 struct list_head *fa_head; 1578 struct leaf *l; 1579 struct leaf_info *li; 1580 1581 if (plen > 32) 1582 return -EINVAL; 1583 1584 key = ntohl(cfg->fc_dst); 1585 mask = ntohl(inet_make_mask(plen)); 1586 1587 if (key & ~mask) 1588 return -EINVAL; 1589 1590 key = key & mask; 1591 l = fib_find_node(t, key); 1592 1593 if (!l) 1594 return -ESRCH; 1595 1596 fa_head = get_fa_head(l, plen); 1597 fa = fib_find_alias(fa_head, tos, 0); 1598 1599 if (!fa) 1600 return -ESRCH; 1601 1602 pr_debug("Deleting %08x/%d tos=%d t=%p\n", key, plen, tos, t); 1603 1604 fa_to_delete = NULL; 1605 fa_head = fa->fa_list.prev; 1606 1607 list_for_each_entry(fa, fa_head, fa_list) { 1608 struct fib_info *fi = fa->fa_info; 1609 1610 if (fa->fa_tos != tos) 1611 break; 1612 1613 if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) && 1614 (cfg->fc_scope == RT_SCOPE_NOWHERE || 1615 fa->fa_scope == cfg->fc_scope) && 1616 (!cfg->fc_protocol || 1617 fi->fib_protocol == cfg->fc_protocol) && 1618 fib_nh_match(cfg, fi) == 0) { 1619 fa_to_delete = fa; 1620 break; 1621 } 1622 } 1623 1624 if (!fa_to_delete) 1625 return -ESRCH; 1626 1627 fa = fa_to_delete; 1628 rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, 1629 &cfg->fc_nlinfo, 0); 1630 1631 l = fib_find_node(t, key); 1632 li = find_leaf_info(l, plen); 1633 1634 list_del_rcu(&fa->fa_list); 1635 1636 if (list_empty(fa_head)) { 1637 hlist_del_rcu(&li->hlist); 1638 free_leaf_info(li); 1639 } 1640 1641 if (hlist_empty(&l->list)) 1642 trie_leaf_remove(t, key); 1643 1644 if (fa->fa_state & FA_S_ACCESSED) 1645 rt_cache_flush(-1); 1646 1647 fib_release_info(fa->fa_info); 1648 alias_free_mem_rcu(fa); 1649 return 0; 1650 } 1651 1652 static int trie_flush_list(struct trie *t, struct list_head *head) 1653 { 1654 struct fib_alias *fa, *fa_node; 1655 int found = 0; 1656 1657 list_for_each_entry_safe(fa, fa_node, head, fa_list) { 1658 struct fib_info *fi = fa->fa_info; 1659 1660 if (fi && (fi->fib_flags & RTNH_F_DEAD)) { 1661 list_del_rcu(&fa->fa_list); 1662 fib_release_info(fa->fa_info); 1663 alias_free_mem_rcu(fa); 1664 found++; 1665 } 1666 } 1667 return found; 1668 } 1669 1670 static int trie_flush_leaf(struct trie *t, struct leaf *l) 1671 { 1672 int found = 0; 1673 struct hlist_head *lih = &l->list; 1674 struct hlist_node *node, *tmp; 1675 struct leaf_info *li = NULL; 1676 1677 hlist_for_each_entry_safe(li, node, tmp, lih, hlist) { 1678 found += trie_flush_list(t, &li->falh); 1679 1680 if (list_empty(&li->falh)) { 1681 hlist_del_rcu(&li->hlist); 1682 free_leaf_info(li); 1683 } 1684 } 1685 return found; 1686 } 1687 1688 /* rcu_read_lock needs to be hold by caller from readside */ 1689 1690 static struct leaf *nextleaf(struct trie *t, struct leaf *thisleaf) 1691 { 1692 struct node *c = (struct node *) thisleaf; 1693 struct tnode *p; 1694 int idx; 1695 struct node *trie = rcu_dereference(t->trie); 1696 1697 if (c == NULL) { 1698 if (trie == NULL) 1699 return NULL; 1700 1701 if (IS_LEAF(trie)) /* trie w. just a leaf */ 1702 return (struct leaf *) trie; 1703 1704 p = (struct tnode*) trie; /* Start */ 1705 } else 1706 p = (struct tnode *) NODE_PARENT(c); 1707 1708 while (p) { 1709 int pos, last; 1710 1711 /* Find the next child of the parent */ 1712 if (c) 1713 pos = 1 + tkey_extract_bits(c->key, p->pos, p->bits); 1714 else 1715 pos = 0; 1716 1717 last = 1 << p->bits; 1718 for (idx = pos; idx < last ; idx++) { 1719 c = rcu_dereference(p->child[idx]); 1720 1721 if (!c) 1722 continue; 1723 1724 /* Decend if tnode */ 1725 while (IS_TNODE(c)) { 1726 p = (struct tnode *) c; 1727 idx = 0; 1728 1729 /* Rightmost non-NULL branch */ 1730 if (p && IS_TNODE(p)) 1731 while (!(c = rcu_dereference(p->child[idx])) 1732 && idx < (1<<p->bits)) idx++; 1733 1734 /* Done with this tnode? */ 1735 if (idx >= (1 << p->bits) || !c) 1736 goto up; 1737 } 1738 return (struct leaf *) c; 1739 } 1740 up: 1741 /* No more children go up one step */ 1742 c = (struct node *) p; 1743 p = (struct tnode *) NODE_PARENT(p); 1744 } 1745 return NULL; /* Ready. Root of trie */ 1746 } 1747 1748 /* 1749 * Caller must hold RTNL. 1750 */ 1751 static int fn_trie_flush(struct fib_table *tb) 1752 { 1753 struct trie *t = (struct trie *) tb->tb_data; 1754 struct leaf *ll = NULL, *l = NULL; 1755 int found = 0, h; 1756 1757 t->revision++; 1758 1759 for (h = 0; (l = nextleaf(t, l)) != NULL; h++) { 1760 found += trie_flush_leaf(t, l); 1761 1762 if (ll && hlist_empty(&ll->list)) 1763 trie_leaf_remove(t, ll->key); 1764 ll = l; 1765 } 1766 1767 if (ll && hlist_empty(&ll->list)) 1768 trie_leaf_remove(t, ll->key); 1769 1770 pr_debug("trie_flush found=%d\n", found); 1771 return found; 1772 } 1773 1774 static int trie_last_dflt = -1; 1775 1776 static void 1777 fn_trie_select_default(struct fib_table *tb, const struct flowi *flp, struct fib_result *res) 1778 { 1779 struct trie *t = (struct trie *) tb->tb_data; 1780 int order, last_idx; 1781 struct fib_info *fi = NULL; 1782 struct fib_info *last_resort; 1783 struct fib_alias *fa = NULL; 1784 struct list_head *fa_head; 1785 struct leaf *l; 1786 1787 last_idx = -1; 1788 last_resort = NULL; 1789 order = -1; 1790 1791 rcu_read_lock(); 1792 1793 l = fib_find_node(t, 0); 1794 if (!l) 1795 goto out; 1796 1797 fa_head = get_fa_head(l, 0); 1798 if (!fa_head) 1799 goto out; 1800 1801 if (list_empty(fa_head)) 1802 goto out; 1803 1804 list_for_each_entry_rcu(fa, fa_head, fa_list) { 1805 struct fib_info *next_fi = fa->fa_info; 1806 1807 if (fa->fa_scope != res->scope || 1808 fa->fa_type != RTN_UNICAST) 1809 continue; 1810 1811 if (next_fi->fib_priority > res->fi->fib_priority) 1812 break; 1813 if (!next_fi->fib_nh[0].nh_gw || 1814 next_fi->fib_nh[0].nh_scope != RT_SCOPE_LINK) 1815 continue; 1816 fa->fa_state |= FA_S_ACCESSED; 1817 1818 if (fi == NULL) { 1819 if (next_fi != res->fi) 1820 break; 1821 } else if (!fib_detect_death(fi, order, &last_resort, 1822 &last_idx, &trie_last_dflt)) { 1823 if (res->fi) 1824 fib_info_put(res->fi); 1825 res->fi = fi; 1826 atomic_inc(&fi->fib_clntref); 1827 trie_last_dflt = order; 1828 goto out; 1829 } 1830 fi = next_fi; 1831 order++; 1832 } 1833 if (order <= 0 || fi == NULL) { 1834 trie_last_dflt = -1; 1835 goto out; 1836 } 1837 1838 if (!fib_detect_death(fi, order, &last_resort, &last_idx, &trie_last_dflt)) { 1839 if (res->fi) 1840 fib_info_put(res->fi); 1841 res->fi = fi; 1842 atomic_inc(&fi->fib_clntref); 1843 trie_last_dflt = order; 1844 goto out; 1845 } 1846 if (last_idx >= 0) { 1847 if (res->fi) 1848 fib_info_put(res->fi); 1849 res->fi = last_resort; 1850 if (last_resort) 1851 atomic_inc(&last_resort->fib_clntref); 1852 } 1853 trie_last_dflt = last_idx; 1854 out:; 1855 rcu_read_unlock(); 1856 } 1857 1858 static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fib_table *tb, 1859 struct sk_buff *skb, struct netlink_callback *cb) 1860 { 1861 int i, s_i; 1862 struct fib_alias *fa; 1863 1864 __be32 xkey = htonl(key); 1865 1866 s_i = cb->args[4]; 1867 i = 0; 1868 1869 /* rcu_read_lock is hold by caller */ 1870 1871 list_for_each_entry_rcu(fa, fah, fa_list) { 1872 if (i < s_i) { 1873 i++; 1874 continue; 1875 } 1876 BUG_ON(!fa->fa_info); 1877 1878 if (fib_dump_info(skb, NETLINK_CB(cb->skb).pid, 1879 cb->nlh->nlmsg_seq, 1880 RTM_NEWROUTE, 1881 tb->tb_id, 1882 fa->fa_type, 1883 fa->fa_scope, 1884 xkey, 1885 plen, 1886 fa->fa_tos, 1887 fa->fa_info, 0) < 0) { 1888 cb->args[4] = i; 1889 return -1; 1890 } 1891 i++; 1892 } 1893 cb->args[4] = i; 1894 return skb->len; 1895 } 1896 1897 static int fn_trie_dump_plen(struct trie *t, int plen, struct fib_table *tb, struct sk_buff *skb, 1898 struct netlink_callback *cb) 1899 { 1900 int h, s_h; 1901 struct list_head *fa_head; 1902 struct leaf *l = NULL; 1903 1904 s_h = cb->args[3]; 1905 1906 for (h = 0; (l = nextleaf(t, l)) != NULL; h++) { 1907 if (h < s_h) 1908 continue; 1909 if (h > s_h) 1910 memset(&cb->args[4], 0, 1911 sizeof(cb->args) - 4*sizeof(cb->args[0])); 1912 1913 fa_head = get_fa_head(l, plen); 1914 1915 if (!fa_head) 1916 continue; 1917 1918 if (list_empty(fa_head)) 1919 continue; 1920 1921 if (fn_trie_dump_fa(l->key, plen, fa_head, tb, skb, cb)<0) { 1922 cb->args[3] = h; 1923 return -1; 1924 } 1925 } 1926 cb->args[3] = h; 1927 return skb->len; 1928 } 1929 1930 static int fn_trie_dump(struct fib_table *tb, struct sk_buff *skb, struct netlink_callback *cb) 1931 { 1932 int m, s_m; 1933 struct trie *t = (struct trie *) tb->tb_data; 1934 1935 s_m = cb->args[2]; 1936 1937 rcu_read_lock(); 1938 for (m = 0; m <= 32; m++) { 1939 if (m < s_m) 1940 continue; 1941 if (m > s_m) 1942 memset(&cb->args[3], 0, 1943 sizeof(cb->args) - 3*sizeof(cb->args[0])); 1944 1945 if (fn_trie_dump_plen(t, 32-m, tb, skb, cb)<0) { 1946 cb->args[2] = m; 1947 goto out; 1948 } 1949 } 1950 rcu_read_unlock(); 1951 cb->args[2] = m; 1952 return skb->len; 1953 out: 1954 rcu_read_unlock(); 1955 return -1; 1956 } 1957 1958 /* Fix more generic FIB names for init later */ 1959 1960 #ifdef CONFIG_IP_MULTIPLE_TABLES 1961 struct fib_table * fib_hash_init(u32 id) 1962 #else 1963 struct fib_table * __init fib_hash_init(u32 id) 1964 #endif 1965 { 1966 struct fib_table *tb; 1967 struct trie *t; 1968 1969 if (fn_alias_kmem == NULL) 1970 fn_alias_kmem = kmem_cache_create("ip_fib_alias", 1971 sizeof(struct fib_alias), 1972 0, SLAB_HWCACHE_ALIGN, 1973 NULL, NULL); 1974 1975 tb = kmalloc(sizeof(struct fib_table) + sizeof(struct trie), 1976 GFP_KERNEL); 1977 if (tb == NULL) 1978 return NULL; 1979 1980 tb->tb_id = id; 1981 tb->tb_lookup = fn_trie_lookup; 1982 tb->tb_insert = fn_trie_insert; 1983 tb->tb_delete = fn_trie_delete; 1984 tb->tb_flush = fn_trie_flush; 1985 tb->tb_select_default = fn_trie_select_default; 1986 tb->tb_dump = fn_trie_dump; 1987 memset(tb->tb_data, 0, sizeof(struct trie)); 1988 1989 t = (struct trie *) tb->tb_data; 1990 1991 trie_init(t); 1992 1993 if (id == RT_TABLE_LOCAL) 1994 trie_local = t; 1995 else if (id == RT_TABLE_MAIN) 1996 trie_main = t; 1997 1998 if (id == RT_TABLE_LOCAL) 1999 printk(KERN_INFO "IPv4 FIB: Using LC-trie version %s\n", VERSION); 2000 2001 return tb; 2002 } 2003 2004 #ifdef CONFIG_PROC_FS 2005 /* Depth first Trie walk iterator */ 2006 struct fib_trie_iter { 2007 struct tnode *tnode; 2008 struct trie *trie; 2009 unsigned index; 2010 unsigned depth; 2011 }; 2012 2013 static struct node *fib_trie_get_next(struct fib_trie_iter *iter) 2014 { 2015 struct tnode *tn = iter->tnode; 2016 unsigned cindex = iter->index; 2017 struct tnode *p; 2018 2019 /* A single entry routing table */ 2020 if (!tn) 2021 return NULL; 2022 2023 pr_debug("get_next iter={node=%p index=%d depth=%d}\n", 2024 iter->tnode, iter->index, iter->depth); 2025 rescan: 2026 while (cindex < (1<<tn->bits)) { 2027 struct node *n = tnode_get_child(tn, cindex); 2028 2029 if (n) { 2030 if (IS_LEAF(n)) { 2031 iter->tnode = tn; 2032 iter->index = cindex + 1; 2033 } else { 2034 /* push down one level */ 2035 iter->tnode = (struct tnode *) n; 2036 iter->index = 0; 2037 ++iter->depth; 2038 } 2039 return n; 2040 } 2041 2042 ++cindex; 2043 } 2044 2045 /* Current node exhausted, pop back up */ 2046 p = NODE_PARENT(tn); 2047 if (p) { 2048 cindex = tkey_extract_bits(tn->key, p->pos, p->bits)+1; 2049 tn = p; 2050 --iter->depth; 2051 goto rescan; 2052 } 2053 2054 /* got root? */ 2055 return NULL; 2056 } 2057 2058 static struct node *fib_trie_get_first(struct fib_trie_iter *iter, 2059 struct trie *t) 2060 { 2061 struct node *n ; 2062 2063 if (!t) 2064 return NULL; 2065 2066 n = rcu_dereference(t->trie); 2067 2068 if (!iter) 2069 return NULL; 2070 2071 if (n) { 2072 if (IS_TNODE(n)) { 2073 iter->tnode = (struct tnode *) n; 2074 iter->trie = t; 2075 iter->index = 0; 2076 iter->depth = 1; 2077 } else { 2078 iter->tnode = NULL; 2079 iter->trie = t; 2080 iter->index = 0; 2081 iter->depth = 0; 2082 } 2083 return n; 2084 } 2085 return NULL; 2086 } 2087 2088 static void trie_collect_stats(struct trie *t, struct trie_stat *s) 2089 { 2090 struct node *n; 2091 struct fib_trie_iter iter; 2092 2093 memset(s, 0, sizeof(*s)); 2094 2095 rcu_read_lock(); 2096 for (n = fib_trie_get_first(&iter, t); n; 2097 n = fib_trie_get_next(&iter)) { 2098 if (IS_LEAF(n)) { 2099 s->leaves++; 2100 s->totdepth += iter.depth; 2101 if (iter.depth > s->maxdepth) 2102 s->maxdepth = iter.depth; 2103 } else { 2104 const struct tnode *tn = (const struct tnode *) n; 2105 int i; 2106 2107 s->tnodes++; 2108 if (tn->bits < MAX_STAT_DEPTH) 2109 s->nodesizes[tn->bits]++; 2110 2111 for (i = 0; i < (1<<tn->bits); i++) 2112 if (!tn->child[i]) 2113 s->nullpointers++; 2114 } 2115 } 2116 rcu_read_unlock(); 2117 } 2118 2119 /* 2120 * This outputs /proc/net/fib_triestats 2121 */ 2122 static void trie_show_stats(struct seq_file *seq, struct trie_stat *stat) 2123 { 2124 unsigned i, max, pointers, bytes, avdepth; 2125 2126 if (stat->leaves) 2127 avdepth = stat->totdepth*100 / stat->leaves; 2128 else 2129 avdepth = 0; 2130 2131 seq_printf(seq, "\tAver depth: %d.%02d\n", avdepth / 100, avdepth % 100 ); 2132 seq_printf(seq, "\tMax depth: %u\n", stat->maxdepth); 2133 2134 seq_printf(seq, "\tLeaves: %u\n", stat->leaves); 2135 2136 bytes = sizeof(struct leaf) * stat->leaves; 2137 seq_printf(seq, "\tInternal nodes: %d\n\t", stat->tnodes); 2138 bytes += sizeof(struct tnode) * stat->tnodes; 2139 2140 max = MAX_STAT_DEPTH; 2141 while (max > 0 && stat->nodesizes[max-1] == 0) 2142 max--; 2143 2144 pointers = 0; 2145 for (i = 1; i <= max; i++) 2146 if (stat->nodesizes[i] != 0) { 2147 seq_printf(seq, " %d: %d", i, stat->nodesizes[i]); 2148 pointers += (1<<i) * stat->nodesizes[i]; 2149 } 2150 seq_putc(seq, '\n'); 2151 seq_printf(seq, "\tPointers: %d\n", pointers); 2152 2153 bytes += sizeof(struct node *) * pointers; 2154 seq_printf(seq, "Null ptrs: %d\n", stat->nullpointers); 2155 seq_printf(seq, "Total size: %d kB\n", (bytes + 1023) / 1024); 2156 2157 #ifdef CONFIG_IP_FIB_TRIE_STATS 2158 seq_printf(seq, "Counters:\n---------\n"); 2159 seq_printf(seq,"gets = %d\n", t->stats.gets); 2160 seq_printf(seq,"backtracks = %d\n", t->stats.backtrack); 2161 seq_printf(seq,"semantic match passed = %d\n", t->stats.semantic_match_passed); 2162 seq_printf(seq,"semantic match miss = %d\n", t->stats.semantic_match_miss); 2163 seq_printf(seq,"null node hit= %d\n", t->stats.null_node_hit); 2164 seq_printf(seq,"skipped node resize = %d\n", t->stats.resize_node_skipped); 2165 #ifdef CLEAR_STATS 2166 memset(&(t->stats), 0, sizeof(t->stats)); 2167 #endif 2168 #endif /* CONFIG_IP_FIB_TRIE_STATS */ 2169 } 2170 2171 static int fib_triestat_seq_show(struct seq_file *seq, void *v) 2172 { 2173 struct trie_stat *stat; 2174 2175 stat = kmalloc(sizeof(*stat), GFP_KERNEL); 2176 if (!stat) 2177 return -ENOMEM; 2178 2179 seq_printf(seq, "Basic info: size of leaf: %Zd bytes, size of tnode: %Zd bytes.\n", 2180 sizeof(struct leaf), sizeof(struct tnode)); 2181 2182 if (trie_local) { 2183 seq_printf(seq, "Local:\n"); 2184 trie_collect_stats(trie_local, stat); 2185 trie_show_stats(seq, stat); 2186 } 2187 2188 if (trie_main) { 2189 seq_printf(seq, "Main:\n"); 2190 trie_collect_stats(trie_main, stat); 2191 trie_show_stats(seq, stat); 2192 } 2193 kfree(stat); 2194 2195 return 0; 2196 } 2197 2198 static int fib_triestat_seq_open(struct inode *inode, struct file *file) 2199 { 2200 return single_open(file, fib_triestat_seq_show, NULL); 2201 } 2202 2203 static const struct file_operations fib_triestat_fops = { 2204 .owner = THIS_MODULE, 2205 .open = fib_triestat_seq_open, 2206 .read = seq_read, 2207 .llseek = seq_lseek, 2208 .release = single_release, 2209 }; 2210 2211 static struct node *fib_trie_get_idx(struct fib_trie_iter *iter, 2212 loff_t pos) 2213 { 2214 loff_t idx = 0; 2215 struct node *n; 2216 2217 for (n = fib_trie_get_first(iter, trie_local); 2218 n; ++idx, n = fib_trie_get_next(iter)) { 2219 if (pos == idx) 2220 return n; 2221 } 2222 2223 for (n = fib_trie_get_first(iter, trie_main); 2224 n; ++idx, n = fib_trie_get_next(iter)) { 2225 if (pos == idx) 2226 return n; 2227 } 2228 return NULL; 2229 } 2230 2231 static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos) 2232 { 2233 rcu_read_lock(); 2234 if (*pos == 0) 2235 return SEQ_START_TOKEN; 2236 return fib_trie_get_idx(seq->private, *pos - 1); 2237 } 2238 2239 static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos) 2240 { 2241 struct fib_trie_iter *iter = seq->private; 2242 void *l = v; 2243 2244 ++*pos; 2245 if (v == SEQ_START_TOKEN) 2246 return fib_trie_get_idx(iter, 0); 2247 2248 v = fib_trie_get_next(iter); 2249 BUG_ON(v == l); 2250 if (v) 2251 return v; 2252 2253 /* continue scan in next trie */ 2254 if (iter->trie == trie_local) 2255 return fib_trie_get_first(iter, trie_main); 2256 2257 return NULL; 2258 } 2259 2260 static void fib_trie_seq_stop(struct seq_file *seq, void *v) 2261 { 2262 rcu_read_unlock(); 2263 } 2264 2265 static void seq_indent(struct seq_file *seq, int n) 2266 { 2267 while (n-- > 0) seq_puts(seq, " "); 2268 } 2269 2270 static inline const char *rtn_scope(enum rt_scope_t s) 2271 { 2272 static char buf[32]; 2273 2274 switch (s) { 2275 case RT_SCOPE_UNIVERSE: return "universe"; 2276 case RT_SCOPE_SITE: return "site"; 2277 case RT_SCOPE_LINK: return "link"; 2278 case RT_SCOPE_HOST: return "host"; 2279 case RT_SCOPE_NOWHERE: return "nowhere"; 2280 default: 2281 snprintf(buf, sizeof(buf), "scope=%d", s); 2282 return buf; 2283 } 2284 } 2285 2286 static const char *rtn_type_names[__RTN_MAX] = { 2287 [RTN_UNSPEC] = "UNSPEC", 2288 [RTN_UNICAST] = "UNICAST", 2289 [RTN_LOCAL] = "LOCAL", 2290 [RTN_BROADCAST] = "BROADCAST", 2291 [RTN_ANYCAST] = "ANYCAST", 2292 [RTN_MULTICAST] = "MULTICAST", 2293 [RTN_BLACKHOLE] = "BLACKHOLE", 2294 [RTN_UNREACHABLE] = "UNREACHABLE", 2295 [RTN_PROHIBIT] = "PROHIBIT", 2296 [RTN_THROW] = "THROW", 2297 [RTN_NAT] = "NAT", 2298 [RTN_XRESOLVE] = "XRESOLVE", 2299 }; 2300 2301 static inline const char *rtn_type(unsigned t) 2302 { 2303 static char buf[32]; 2304 2305 if (t < __RTN_MAX && rtn_type_names[t]) 2306 return rtn_type_names[t]; 2307 snprintf(buf, sizeof(buf), "type %d", t); 2308 return buf; 2309 } 2310 2311 /* Pretty print the trie */ 2312 static int fib_trie_seq_show(struct seq_file *seq, void *v) 2313 { 2314 const struct fib_trie_iter *iter = seq->private; 2315 struct node *n = v; 2316 2317 if (v == SEQ_START_TOKEN) 2318 return 0; 2319 2320 if (!NODE_PARENT(n)) { 2321 if (iter->trie == trie_local) 2322 seq_puts(seq, "<local>:\n"); 2323 else 2324 seq_puts(seq, "<main>:\n"); 2325 } 2326 2327 if (IS_TNODE(n)) { 2328 struct tnode *tn = (struct tnode *) n; 2329 __be32 prf = htonl(MASK_PFX(tn->key, tn->pos)); 2330 2331 seq_indent(seq, iter->depth-1); 2332 seq_printf(seq, " +-- %d.%d.%d.%d/%d %d %d %d\n", 2333 NIPQUAD(prf), tn->pos, tn->bits, tn->full_children, 2334 tn->empty_children); 2335 2336 } else { 2337 struct leaf *l = (struct leaf *) n; 2338 int i; 2339 __be32 val = htonl(l->key); 2340 2341 seq_indent(seq, iter->depth); 2342 seq_printf(seq, " |-- %d.%d.%d.%d\n", NIPQUAD(val)); 2343 for (i = 32; i >= 0; i--) { 2344 struct leaf_info *li = find_leaf_info(l, i); 2345 if (li) { 2346 struct fib_alias *fa; 2347 list_for_each_entry_rcu(fa, &li->falh, fa_list) { 2348 seq_indent(seq, iter->depth+1); 2349 seq_printf(seq, " /%d %s %s", i, 2350 rtn_scope(fa->fa_scope), 2351 rtn_type(fa->fa_type)); 2352 if (fa->fa_tos) 2353 seq_printf(seq, "tos =%d\n", 2354 fa->fa_tos); 2355 seq_putc(seq, '\n'); 2356 } 2357 } 2358 } 2359 } 2360 2361 return 0; 2362 } 2363 2364 static const struct seq_operations fib_trie_seq_ops = { 2365 .start = fib_trie_seq_start, 2366 .next = fib_trie_seq_next, 2367 .stop = fib_trie_seq_stop, 2368 .show = fib_trie_seq_show, 2369 }; 2370 2371 static int fib_trie_seq_open(struct inode *inode, struct file *file) 2372 { 2373 struct seq_file *seq; 2374 int rc = -ENOMEM; 2375 struct fib_trie_iter *s = kmalloc(sizeof(*s), GFP_KERNEL); 2376 2377 if (!s) 2378 goto out; 2379 2380 rc = seq_open(file, &fib_trie_seq_ops); 2381 if (rc) 2382 goto out_kfree; 2383 2384 seq = file->private_data; 2385 seq->private = s; 2386 memset(s, 0, sizeof(*s)); 2387 out: 2388 return rc; 2389 out_kfree: 2390 kfree(s); 2391 goto out; 2392 } 2393 2394 static const struct file_operations fib_trie_fops = { 2395 .owner = THIS_MODULE, 2396 .open = fib_trie_seq_open, 2397 .read = seq_read, 2398 .llseek = seq_lseek, 2399 .release = seq_release_private, 2400 }; 2401 2402 static unsigned fib_flag_trans(int type, __be32 mask, const struct fib_info *fi) 2403 { 2404 static unsigned type2flags[RTN_MAX + 1] = { 2405 [7] = RTF_REJECT, [8] = RTF_REJECT, 2406 }; 2407 unsigned flags = type2flags[type]; 2408 2409 if (fi && fi->fib_nh->nh_gw) 2410 flags |= RTF_GATEWAY; 2411 if (mask == htonl(0xFFFFFFFF)) 2412 flags |= RTF_HOST; 2413 flags |= RTF_UP; 2414 return flags; 2415 } 2416 2417 /* 2418 * This outputs /proc/net/route. 2419 * The format of the file is not supposed to be changed 2420 * and needs to be same as fib_hash output to avoid breaking 2421 * legacy utilities 2422 */ 2423 static int fib_route_seq_show(struct seq_file *seq, void *v) 2424 { 2425 const struct fib_trie_iter *iter = seq->private; 2426 struct leaf *l = v; 2427 int i; 2428 char bf[128]; 2429 2430 if (v == SEQ_START_TOKEN) { 2431 seq_printf(seq, "%-127s\n", "Iface\tDestination\tGateway " 2432 "\tFlags\tRefCnt\tUse\tMetric\tMask\t\tMTU" 2433 "\tWindow\tIRTT"); 2434 return 0; 2435 } 2436 2437 if (iter->trie == trie_local) 2438 return 0; 2439 if (IS_TNODE(l)) 2440 return 0; 2441 2442 for (i=32; i>=0; i--) { 2443 struct leaf_info *li = find_leaf_info(l, i); 2444 struct fib_alias *fa; 2445 __be32 mask, prefix; 2446 2447 if (!li) 2448 continue; 2449 2450 mask = inet_make_mask(li->plen); 2451 prefix = htonl(l->key); 2452 2453 list_for_each_entry_rcu(fa, &li->falh, fa_list) { 2454 const struct fib_info *fi = fa->fa_info; 2455 unsigned flags = fib_flag_trans(fa->fa_type, mask, fi); 2456 2457 if (fa->fa_type == RTN_BROADCAST 2458 || fa->fa_type == RTN_MULTICAST) 2459 continue; 2460 2461 if (fi) 2462 snprintf(bf, sizeof(bf), 2463 "%s\t%08X\t%08X\t%04X\t%d\t%u\t%d\t%08X\t%d\t%u\t%u", 2464 fi->fib_dev ? fi->fib_dev->name : "*", 2465 prefix, 2466 fi->fib_nh->nh_gw, flags, 0, 0, 2467 fi->fib_priority, 2468 mask, 2469 (fi->fib_advmss ? fi->fib_advmss + 40 : 0), 2470 fi->fib_window, 2471 fi->fib_rtt >> 3); 2472 else 2473 snprintf(bf, sizeof(bf), 2474 "*\t%08X\t%08X\t%04X\t%d\t%u\t%d\t%08X\t%d\t%u\t%u", 2475 prefix, 0, flags, 0, 0, 0, 2476 mask, 0, 0, 0); 2477 2478 seq_printf(seq, "%-127s\n", bf); 2479 } 2480 } 2481 2482 return 0; 2483 } 2484 2485 static const struct seq_operations fib_route_seq_ops = { 2486 .start = fib_trie_seq_start, 2487 .next = fib_trie_seq_next, 2488 .stop = fib_trie_seq_stop, 2489 .show = fib_route_seq_show, 2490 }; 2491 2492 static int fib_route_seq_open(struct inode *inode, struct file *file) 2493 { 2494 struct seq_file *seq; 2495 int rc = -ENOMEM; 2496 struct fib_trie_iter *s = kmalloc(sizeof(*s), GFP_KERNEL); 2497 2498 if (!s) 2499 goto out; 2500 2501 rc = seq_open(file, &fib_route_seq_ops); 2502 if (rc) 2503 goto out_kfree; 2504 2505 seq = file->private_data; 2506 seq->private = s; 2507 memset(s, 0, sizeof(*s)); 2508 out: 2509 return rc; 2510 out_kfree: 2511 kfree(s); 2512 goto out; 2513 } 2514 2515 static const struct file_operations fib_route_fops = { 2516 .owner = THIS_MODULE, 2517 .open = fib_route_seq_open, 2518 .read = seq_read, 2519 .llseek = seq_lseek, 2520 .release = seq_release_private, 2521 }; 2522 2523 int __init fib_proc_init(void) 2524 { 2525 if (!proc_net_fops_create("fib_trie", S_IRUGO, &fib_trie_fops)) 2526 goto out1; 2527 2528 if (!proc_net_fops_create("fib_triestat", S_IRUGO, &fib_triestat_fops)) 2529 goto out2; 2530 2531 if (!proc_net_fops_create("route", S_IRUGO, &fib_route_fops)) 2532 goto out3; 2533 2534 return 0; 2535 2536 out3: 2537 proc_net_remove("fib_triestat"); 2538 out2: 2539 proc_net_remove("fib_trie"); 2540 out1: 2541 return -ENOMEM; 2542 } 2543 2544 void __init fib_proc_exit(void) 2545 { 2546 proc_net_remove("fib_trie"); 2547 proc_net_remove("fib_triestat"); 2548 proc_net_remove("route"); 2549 } 2550 2551 #endif /* CONFIG_PROC_FS */ 2552