writeback: split writeback_inodes_wb
[linux-2.6/next.git] / net / ipv4 / fib_trie.c
blob79d057a939ba6404d4e7553008a2613815a9b049
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.
7 * Robert Olsson <robert.olsson@its.uu.se> Uppsala Universitet
8 * & Swedish University of Agricultural Sciences.
10 * Jens Laas <jens.laas@data.slu.se> Swedish University of
11 * Agricultural Sciences.
13 * Hans Liss <hans.liss@its.uu.se> Uppsala Universitet
15 * This work is based on the LPC-trie which is originally descibed in:
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/
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
26 * Code from fib_hash has been reused which includes the following header:
29 * INET An implementation of the TCP/IP protocol suite for the LINUX
30 * operating system. INET is implemented using the BSD Socket
31 * interface as the means of communication with the user level.
33 * IPv4 FIB: lookup engine and maintenance routines.
36 * Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
38 * This program is free software; you can redistribute it and/or
39 * modify it under the terms of the GNU General Public License
40 * as published by the Free Software Foundation; either version
41 * 2 of the License, or (at your option) any later version.
43 * Substantial contributions to this work comes from:
45 * David S. Miller, <davem@davemloft.net>
46 * Stephen Hemminger <shemminger@osdl.org>
47 * Paul E. McKenney <paulmck@us.ibm.com>
48 * Patrick McHardy <kaber@trash.net>
51 #define VERSION "0.409"
53 #include <asm/uaccess.h>
54 #include <asm/system.h>
55 #include <linux/bitops.h>
56 #include <linux/types.h>
57 #include <linux/kernel.h>
58 #include <linux/mm.h>
59 #include <linux/string.h>
60 #include <linux/socket.h>
61 #include <linux/sockios.h>
62 #include <linux/errno.h>
63 #include <linux/in.h>
64 #include <linux/inet.h>
65 #include <linux/inetdevice.h>
66 #include <linux/netdevice.h>
67 #include <linux/if_arp.h>
68 #include <linux/proc_fs.h>
69 #include <linux/rcupdate.h>
70 #include <linux/skbuff.h>
71 #include <linux/netlink.h>
72 #include <linux/init.h>
73 #include <linux/list.h>
74 #include <linux/slab.h>
75 #include <net/net_namespace.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"
84 #define MAX_STAT_DEPTH 32
86 #define KEYLENGTH (8*sizeof(t_key))
88 typedef unsigned int t_key;
90 #define T_TNODE 0
91 #define T_LEAF 1
92 #define NODE_TYPE_MASK 0x1UL
93 #define NODE_TYPE(node) ((node)->parent & NODE_TYPE_MASK)
95 #define IS_TNODE(n) (!(n->parent & T_LEAF))
96 #define IS_LEAF(n) (n->parent & T_LEAF)
98 struct node {
99 unsigned long parent;
100 t_key key;
103 struct leaf {
104 unsigned long parent;
105 t_key key;
106 struct hlist_head list;
107 struct rcu_head rcu;
110 struct leaf_info {
111 struct hlist_node hlist;
112 struct rcu_head rcu;
113 int plen;
114 struct list_head falh;
117 struct tnode {
118 unsigned long parent;
119 t_key key;
120 unsigned char pos; /* 2log(KEYLENGTH) bits needed */
121 unsigned char bits; /* 2log(KEYLENGTH) bits needed */
122 unsigned int full_children; /* KEYLENGTH bits needed */
123 unsigned int empty_children; /* KEYLENGTH bits needed */
124 union {
125 struct rcu_head rcu;
126 struct work_struct work;
127 struct tnode *tnode_free;
129 struct node *child[0];
132 #ifdef CONFIG_IP_FIB_TRIE_STATS
133 struct trie_use_stats {
134 unsigned int gets;
135 unsigned int backtrack;
136 unsigned int semantic_match_passed;
137 unsigned int semantic_match_miss;
138 unsigned int null_node_hit;
139 unsigned int resize_node_skipped;
141 #endif
143 struct trie_stat {
144 unsigned int totdepth;
145 unsigned int maxdepth;
146 unsigned int tnodes;
147 unsigned int leaves;
148 unsigned int nullpointers;
149 unsigned int prefixes;
150 unsigned int nodesizes[MAX_STAT_DEPTH];
153 struct trie {
154 struct node *trie;
155 #ifdef CONFIG_IP_FIB_TRIE_STATS
156 struct trie_use_stats stats;
157 #endif
160 static void put_child(struct trie *t, struct tnode *tn, int i, struct node *n);
161 static void tnode_put_child_reorg(struct tnode *tn, int i, struct node *n,
162 int wasfull);
163 static struct node *resize(struct trie *t, struct tnode *tn);
164 static struct tnode *inflate(struct trie *t, struct tnode *tn);
165 static struct tnode *halve(struct trie *t, struct tnode *tn);
166 /* tnodes to free after resize(); protected by RTNL */
167 static struct tnode *tnode_free_head;
168 static size_t tnode_free_size;
171 * synchronize_rcu after call_rcu for that many pages; it should be especially
172 * useful before resizing the root node with PREEMPT_NONE configs; the value was
173 * obtained experimentally, aiming to avoid visible slowdown.
175 static const int sync_pages = 128;
177 static struct kmem_cache *fn_alias_kmem __read_mostly;
178 static struct kmem_cache *trie_leaf_kmem __read_mostly;
180 static inline struct tnode *node_parent(struct node *node)
182 return (struct tnode *)(node->parent & ~NODE_TYPE_MASK);
185 static inline struct tnode *node_parent_rcu(struct node *node)
187 struct tnode *ret = node_parent(node);
189 return rcu_dereference(ret);
192 /* Same as rcu_assign_pointer
193 * but that macro() assumes that value is a pointer.
195 static inline void node_set_parent(struct node *node, struct tnode *ptr)
197 smp_wmb();
198 node->parent = (unsigned long)ptr | NODE_TYPE(node);
201 static inline struct node *tnode_get_child(struct tnode *tn, unsigned int i)
203 BUG_ON(i >= 1U << tn->bits);
205 return tn->child[i];
208 static inline struct node *tnode_get_child_rcu(struct tnode *tn, unsigned int i)
210 struct node *ret = tnode_get_child(tn, i);
212 return rcu_dereference_check(ret,
213 rcu_read_lock_held() ||
214 lockdep_rtnl_is_held());
217 static inline int tnode_child_length(const struct tnode *tn)
219 return 1 << tn->bits;
222 static inline t_key mask_pfx(t_key k, unsigned short l)
224 return (l == 0) ? 0 : k >> (KEYLENGTH-l) << (KEYLENGTH-l);
227 static inline t_key tkey_extract_bits(t_key a, int offset, int bits)
229 if (offset < KEYLENGTH)
230 return ((t_key)(a << offset)) >> (KEYLENGTH - bits);
231 else
232 return 0;
235 static inline int tkey_equals(t_key a, t_key b)
237 return a == b;
240 static inline int tkey_sub_equals(t_key a, int offset, int bits, t_key b)
242 if (bits == 0 || offset >= KEYLENGTH)
243 return 1;
244 bits = bits > KEYLENGTH ? KEYLENGTH : bits;
245 return ((a ^ b) << offset) >> (KEYLENGTH - bits) == 0;
248 static inline int tkey_mismatch(t_key a, int offset, t_key b)
250 t_key diff = a ^ b;
251 int i = offset;
253 if (!diff)
254 return 0;
255 while ((diff << i) >> (KEYLENGTH-1) == 0)
256 i++;
257 return i;
261 To understand this stuff, an understanding of keys and all their bits is
262 necessary. Every node in the trie has a key associated with it, but not
263 all of the bits in that key are significant.
265 Consider a node 'n' and its parent 'tp'.
267 If n is a leaf, every bit in its key is significant. Its presence is
268 necessitated by path compression, since during a tree traversal (when
269 searching for a leaf - unless we are doing an insertion) we will completely
270 ignore all skipped bits we encounter. Thus we need to verify, at the end of
271 a potentially successful search, that we have indeed been walking the
272 correct key path.
274 Note that we can never "miss" the correct key in the tree if present by
275 following the wrong path. Path compression ensures that segments of the key
276 that are the same for all keys with a given prefix are skipped, but the
277 skipped part *is* identical for each node in the subtrie below the skipped
278 bit! trie_insert() in this implementation takes care of that - note the
279 call to tkey_sub_equals() in trie_insert().
281 if n is an internal node - a 'tnode' here, the various parts of its key
282 have many different meanings.
284 Example:
285 _________________________________________________________________
286 | i | i | i | i | i | i | i | N | N | N | S | S | S | S | S | C |
287 -----------------------------------------------------------------
288 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
290 _________________________________________________________________
291 | C | C | C | u | u | u | u | u | u | u | u | u | u | u | u | u |
292 -----------------------------------------------------------------
293 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
295 tp->pos = 7
296 tp->bits = 3
297 n->pos = 15
298 n->bits = 4
300 First, let's just ignore the bits that come before the parent tp, that is
301 the bits from 0 to (tp->pos-1). They are *known* but at this point we do
302 not use them for anything.
304 The bits from (tp->pos) to (tp->pos + tp->bits - 1) - "N", above - are the
305 index into the parent's child array. That is, they will be used to find
306 'n' among tp's children.
308 The bits from (tp->pos + tp->bits) to (n->pos - 1) - "S" - are skipped bits
309 for the node n.
311 All the bits we have seen so far are significant to the node n. The rest
312 of the bits are really not needed or indeed known in n->key.
314 The bits from (n->pos) to (n->pos + n->bits - 1) - "C" - are the index into
315 n's child array, and will of course be different for each child.
318 The rest of the bits, from (n->pos + n->bits) onward, are completely unknown
319 at this point.
323 static inline void check_tnode(const struct tnode *tn)
325 WARN_ON(tn && tn->pos+tn->bits > 32);
328 static const int halve_threshold = 25;
329 static const int inflate_threshold = 50;
330 static const int halve_threshold_root = 15;
331 static const int inflate_threshold_root = 30;
333 static void __alias_free_mem(struct rcu_head *head)
335 struct fib_alias *fa = container_of(head, struct fib_alias, rcu);
336 kmem_cache_free(fn_alias_kmem, fa);
339 static inline void alias_free_mem_rcu(struct fib_alias *fa)
341 call_rcu(&fa->rcu, __alias_free_mem);
344 static void __leaf_free_rcu(struct rcu_head *head)
346 struct leaf *l = container_of(head, struct leaf, rcu);
347 kmem_cache_free(trie_leaf_kmem, l);
350 static inline void free_leaf(struct leaf *l)
352 call_rcu_bh(&l->rcu, __leaf_free_rcu);
355 static void __leaf_info_free_rcu(struct rcu_head *head)
357 kfree(container_of(head, struct leaf_info, rcu));
360 static inline void free_leaf_info(struct leaf_info *leaf)
362 call_rcu(&leaf->rcu, __leaf_info_free_rcu);
365 static struct tnode *tnode_alloc(size_t size)
367 if (size <= PAGE_SIZE)
368 return kzalloc(size, GFP_KERNEL);
369 else
370 return __vmalloc(size, GFP_KERNEL | __GFP_ZERO, PAGE_KERNEL);
373 static void __tnode_vfree(struct work_struct *arg)
375 struct tnode *tn = container_of(arg, struct tnode, work);
376 vfree(tn);
379 static void __tnode_free_rcu(struct rcu_head *head)
381 struct tnode *tn = container_of(head, struct tnode, rcu);
382 size_t size = sizeof(struct tnode) +
383 (sizeof(struct node *) << tn->bits);
385 if (size <= PAGE_SIZE)
386 kfree(tn);
387 else {
388 INIT_WORK(&tn->work, __tnode_vfree);
389 schedule_work(&tn->work);
393 static inline void tnode_free(struct tnode *tn)
395 if (IS_LEAF(tn))
396 free_leaf((struct leaf *) tn);
397 else
398 call_rcu(&tn->rcu, __tnode_free_rcu);
401 static void tnode_free_safe(struct tnode *tn)
403 BUG_ON(IS_LEAF(tn));
404 tn->tnode_free = tnode_free_head;
405 tnode_free_head = tn;
406 tnode_free_size += sizeof(struct tnode) +
407 (sizeof(struct node *) << tn->bits);
410 static void tnode_free_flush(void)
412 struct tnode *tn;
414 while ((tn = tnode_free_head)) {
415 tnode_free_head = tn->tnode_free;
416 tn->tnode_free = NULL;
417 tnode_free(tn);
420 if (tnode_free_size >= PAGE_SIZE * sync_pages) {
421 tnode_free_size = 0;
422 synchronize_rcu();
426 static struct leaf *leaf_new(void)
428 struct leaf *l = kmem_cache_alloc(trie_leaf_kmem, GFP_KERNEL);
429 if (l) {
430 l->parent = T_LEAF;
431 INIT_HLIST_HEAD(&l->list);
433 return l;
436 static struct leaf_info *leaf_info_new(int plen)
438 struct leaf_info *li = kmalloc(sizeof(struct leaf_info), GFP_KERNEL);
439 if (li) {
440 li->plen = plen;
441 INIT_LIST_HEAD(&li->falh);
443 return li;
446 static struct tnode *tnode_new(t_key key, int pos, int bits)
448 size_t sz = sizeof(struct tnode) + (sizeof(struct node *) << bits);
449 struct tnode *tn = tnode_alloc(sz);
451 if (tn) {
452 tn->parent = T_TNODE;
453 tn->pos = pos;
454 tn->bits = bits;
455 tn->key = key;
456 tn->full_children = 0;
457 tn->empty_children = 1<<bits;
460 pr_debug("AT %p s=%u %lu\n", tn, (unsigned int) sizeof(struct tnode),
461 (unsigned long) (sizeof(struct node) << bits));
462 return tn;
466 * Check whether a tnode 'n' is "full", i.e. it is an internal node
467 * and no bits are skipped. See discussion in dyntree paper p. 6
470 static inline int tnode_full(const struct tnode *tn, const struct node *n)
472 if (n == NULL || IS_LEAF(n))
473 return 0;
475 return ((struct tnode *) n)->pos == tn->pos + tn->bits;
478 static inline void put_child(struct trie *t, struct tnode *tn, int i,
479 struct node *n)
481 tnode_put_child_reorg(tn, i, n, -1);
485 * Add a child at position i overwriting the old value.
486 * Update the value of full_children and empty_children.
489 static void tnode_put_child_reorg(struct tnode *tn, int i, struct node *n,
490 int wasfull)
492 struct node *chi = tn->child[i];
493 int isfull;
495 BUG_ON(i >= 1<<tn->bits);
497 /* update emptyChildren */
498 if (n == NULL && chi != NULL)
499 tn->empty_children++;
500 else if (n != NULL && chi == NULL)
501 tn->empty_children--;
503 /* update fullChildren */
504 if (wasfull == -1)
505 wasfull = tnode_full(tn, chi);
507 isfull = tnode_full(tn, n);
508 if (wasfull && !isfull)
509 tn->full_children--;
510 else if (!wasfull && isfull)
511 tn->full_children++;
513 if (n)
514 node_set_parent(n, tn);
516 rcu_assign_pointer(tn->child[i], n);
519 #define MAX_WORK 10
520 static struct node *resize(struct trie *t, struct tnode *tn)
522 int i;
523 struct tnode *old_tn;
524 int inflate_threshold_use;
525 int halve_threshold_use;
526 int max_work;
528 if (!tn)
529 return NULL;
531 pr_debug("In tnode_resize %p inflate_threshold=%d threshold=%d\n",
532 tn, inflate_threshold, halve_threshold);
534 /* No children */
535 if (tn->empty_children == tnode_child_length(tn)) {
536 tnode_free_safe(tn);
537 return NULL;
539 /* One child */
540 if (tn->empty_children == tnode_child_length(tn) - 1)
541 goto one_child;
543 * Double as long as the resulting node has a number of
544 * nonempty nodes that are above the threshold.
548 * From "Implementing a dynamic compressed trie" by Stefan Nilsson of
549 * the Helsinki University of Technology and Matti Tikkanen of Nokia
550 * Telecommunications, page 6:
551 * "A node is doubled if the ratio of non-empty children to all
552 * children in the *doubled* node is at least 'high'."
554 * 'high' in this instance is the variable 'inflate_threshold'. It
555 * is expressed as a percentage, so we multiply it with
556 * tnode_child_length() and instead of multiplying by 2 (since the
557 * child array will be doubled by inflate()) and multiplying
558 * the left-hand side by 100 (to handle the percentage thing) we
559 * multiply the left-hand side by 50.
561 * The left-hand side may look a bit weird: tnode_child_length(tn)
562 * - tn->empty_children is of course the number of non-null children
563 * in the current node. tn->full_children is the number of "full"
564 * children, that is non-null tnodes with a skip value of 0.
565 * All of those will be doubled in the resulting inflated tnode, so
566 * we just count them one extra time here.
568 * A clearer way to write this would be:
570 * to_be_doubled = tn->full_children;
571 * not_to_be_doubled = tnode_child_length(tn) - tn->empty_children -
572 * tn->full_children;
574 * new_child_length = tnode_child_length(tn) * 2;
576 * new_fill_factor = 100 * (not_to_be_doubled + 2*to_be_doubled) /
577 * new_child_length;
578 * if (new_fill_factor >= inflate_threshold)
580 * ...and so on, tho it would mess up the while () loop.
582 * anyway,
583 * 100 * (not_to_be_doubled + 2*to_be_doubled) / new_child_length >=
584 * inflate_threshold
586 * avoid a division:
587 * 100 * (not_to_be_doubled + 2*to_be_doubled) >=
588 * inflate_threshold * new_child_length
590 * expand not_to_be_doubled and to_be_doubled, and shorten:
591 * 100 * (tnode_child_length(tn) - tn->empty_children +
592 * tn->full_children) >= inflate_threshold * new_child_length
594 * expand new_child_length:
595 * 100 * (tnode_child_length(tn) - tn->empty_children +
596 * tn->full_children) >=
597 * inflate_threshold * tnode_child_length(tn) * 2
599 * shorten again:
600 * 50 * (tn->full_children + tnode_child_length(tn) -
601 * tn->empty_children) >= inflate_threshold *
602 * tnode_child_length(tn)
606 check_tnode(tn);
608 /* Keep root node larger */
610 if (!node_parent((struct node*) tn)) {
611 inflate_threshold_use = inflate_threshold_root;
612 halve_threshold_use = halve_threshold_root;
614 else {
615 inflate_threshold_use = inflate_threshold;
616 halve_threshold_use = halve_threshold;
619 max_work = MAX_WORK;
620 while ((tn->full_children > 0 && max_work-- &&
621 50 * (tn->full_children + tnode_child_length(tn)
622 - tn->empty_children)
623 >= inflate_threshold_use * tnode_child_length(tn))) {
625 old_tn = tn;
626 tn = inflate(t, tn);
628 if (IS_ERR(tn)) {
629 tn = old_tn;
630 #ifdef CONFIG_IP_FIB_TRIE_STATS
631 t->stats.resize_node_skipped++;
632 #endif
633 break;
637 check_tnode(tn);
639 /* Return if at least one inflate is run */
640 if( max_work != MAX_WORK)
641 return (struct node *) tn;
644 * Halve as long as the number of empty children in this
645 * node is above threshold.
648 max_work = MAX_WORK;
649 while (tn->bits > 1 && max_work-- &&
650 100 * (tnode_child_length(tn) - tn->empty_children) <
651 halve_threshold_use * tnode_child_length(tn)) {
653 old_tn = tn;
654 tn = halve(t, tn);
655 if (IS_ERR(tn)) {
656 tn = old_tn;
657 #ifdef CONFIG_IP_FIB_TRIE_STATS
658 t->stats.resize_node_skipped++;
659 #endif
660 break;
665 /* Only one child remains */
666 if (tn->empty_children == tnode_child_length(tn) - 1) {
667 one_child:
668 for (i = 0; i < tnode_child_length(tn); i++) {
669 struct node *n;
671 n = tn->child[i];
672 if (!n)
673 continue;
675 /* compress one level */
677 node_set_parent(n, NULL);
678 tnode_free_safe(tn);
679 return n;
682 return (struct node *) tn;
685 static struct tnode *inflate(struct trie *t, struct tnode *tn)
687 struct tnode *oldtnode = tn;
688 int olen = tnode_child_length(tn);
689 int i;
691 pr_debug("In inflate\n");
693 tn = tnode_new(oldtnode->key, oldtnode->pos, oldtnode->bits + 1);
695 if (!tn)
696 return ERR_PTR(-ENOMEM);
699 * Preallocate and store tnodes before the actual work so we
700 * don't get into an inconsistent state if memory allocation
701 * fails. In case of failure we return the oldnode and inflate
702 * of tnode is ignored.
705 for (i = 0; i < olen; i++) {
706 struct tnode *inode;
708 inode = (struct tnode *) tnode_get_child(oldtnode, i);
709 if (inode &&
710 IS_TNODE(inode) &&
711 inode->pos == oldtnode->pos + oldtnode->bits &&
712 inode->bits > 1) {
713 struct tnode *left, *right;
714 t_key m = ~0U << (KEYLENGTH - 1) >> inode->pos;
716 left = tnode_new(inode->key&(~m), inode->pos + 1,
717 inode->bits - 1);
718 if (!left)
719 goto nomem;
721 right = tnode_new(inode->key|m, inode->pos + 1,
722 inode->bits - 1);
724 if (!right) {
725 tnode_free(left);
726 goto nomem;
729 put_child(t, tn, 2*i, (struct node *) left);
730 put_child(t, tn, 2*i+1, (struct node *) right);
734 for (i = 0; i < olen; i++) {
735 struct tnode *inode;
736 struct node *node = tnode_get_child(oldtnode, i);
737 struct tnode *left, *right;
738 int size, j;
740 /* An empty child */
741 if (node == NULL)
742 continue;
744 /* A leaf or an internal node with skipped bits */
746 if (IS_LEAF(node) || ((struct tnode *) node)->pos >
747 tn->pos + tn->bits - 1) {
748 if (tkey_extract_bits(node->key,
749 oldtnode->pos + oldtnode->bits,
750 1) == 0)
751 put_child(t, tn, 2*i, node);
752 else
753 put_child(t, tn, 2*i+1, node);
754 continue;
757 /* An internal node with two children */
758 inode = (struct tnode *) node;
760 if (inode->bits == 1) {
761 put_child(t, tn, 2*i, inode->child[0]);
762 put_child(t, tn, 2*i+1, inode->child[1]);
764 tnode_free_safe(inode);
765 continue;
768 /* An internal node with more than two children */
770 /* We will replace this node 'inode' with two new
771 * ones, 'left' and 'right', each with half of the
772 * original children. The two new nodes will have
773 * a position one bit further down the key and this
774 * means that the "significant" part of their keys
775 * (see the discussion near the top of this file)
776 * will differ by one bit, which will be "0" in
777 * left's key and "1" in right's key. Since we are
778 * moving the key position by one step, the bit that
779 * we are moving away from - the bit at position
780 * (inode->pos) - is the one that will differ between
781 * left and right. So... we synthesize that bit in the
782 * two new keys.
783 * The mask 'm' below will be a single "one" bit at
784 * the position (inode->pos)
787 /* Use the old key, but set the new significant
788 * bit to zero.
791 left = (struct tnode *) tnode_get_child(tn, 2*i);
792 put_child(t, tn, 2*i, NULL);
794 BUG_ON(!left);
796 right = (struct tnode *) tnode_get_child(tn, 2*i+1);
797 put_child(t, tn, 2*i+1, NULL);
799 BUG_ON(!right);
801 size = tnode_child_length(left);
802 for (j = 0; j < size; j++) {
803 put_child(t, left, j, inode->child[j]);
804 put_child(t, right, j, inode->child[j + size]);
806 put_child(t, tn, 2*i, resize(t, left));
807 put_child(t, tn, 2*i+1, resize(t, right));
809 tnode_free_safe(inode);
811 tnode_free_safe(oldtnode);
812 return tn;
813 nomem:
815 int size = tnode_child_length(tn);
816 int j;
818 for (j = 0; j < size; j++)
819 if (tn->child[j])
820 tnode_free((struct tnode *)tn->child[j]);
822 tnode_free(tn);
824 return ERR_PTR(-ENOMEM);
828 static struct tnode *halve(struct trie *t, struct tnode *tn)
830 struct tnode *oldtnode = tn;
831 struct node *left, *right;
832 int i;
833 int olen = tnode_child_length(tn);
835 pr_debug("In halve\n");
837 tn = tnode_new(oldtnode->key, oldtnode->pos, oldtnode->bits - 1);
839 if (!tn)
840 return ERR_PTR(-ENOMEM);
843 * Preallocate and store tnodes before the actual work so we
844 * don't get into an inconsistent state if memory allocation
845 * fails. In case of failure we return the oldnode and halve
846 * of tnode is ignored.
849 for (i = 0; i < olen; i += 2) {
850 left = tnode_get_child(oldtnode, i);
851 right = tnode_get_child(oldtnode, i+1);
853 /* Two nonempty children */
854 if (left && right) {
855 struct tnode *newn;
857 newn = tnode_new(left->key, tn->pos + tn->bits, 1);
859 if (!newn)
860 goto nomem;
862 put_child(t, tn, i/2, (struct node *)newn);
867 for (i = 0; i < olen; i += 2) {
868 struct tnode *newBinNode;
870 left = tnode_get_child(oldtnode, i);
871 right = tnode_get_child(oldtnode, i+1);
873 /* At least one of the children is empty */
874 if (left == NULL) {
875 if (right == NULL) /* Both are empty */
876 continue;
877 put_child(t, tn, i/2, right);
878 continue;
881 if (right == NULL) {
882 put_child(t, tn, i/2, left);
883 continue;
886 /* Two nonempty children */
887 newBinNode = (struct tnode *) tnode_get_child(tn, i/2);
888 put_child(t, tn, i/2, NULL);
889 put_child(t, newBinNode, 0, left);
890 put_child(t, newBinNode, 1, right);
891 put_child(t, tn, i/2, resize(t, newBinNode));
893 tnode_free_safe(oldtnode);
894 return tn;
895 nomem:
897 int size = tnode_child_length(tn);
898 int j;
900 for (j = 0; j < size; j++)
901 if (tn->child[j])
902 tnode_free((struct tnode *)tn->child[j]);
904 tnode_free(tn);
906 return ERR_PTR(-ENOMEM);
910 /* readside must use rcu_read_lock currently dump routines
911 via get_fa_head and dump */
913 static struct leaf_info *find_leaf_info(struct leaf *l, int plen)
915 struct hlist_head *head = &l->list;
916 struct hlist_node *node;
917 struct leaf_info *li;
919 hlist_for_each_entry_rcu(li, node, head, hlist)
920 if (li->plen == plen)
921 return li;
923 return NULL;
926 static inline struct list_head *get_fa_head(struct leaf *l, int plen)
928 struct leaf_info *li = find_leaf_info(l, plen);
930 if (!li)
931 return NULL;
933 return &li->falh;
936 static void insert_leaf_info(struct hlist_head *head, struct leaf_info *new)
938 struct leaf_info *li = NULL, *last = NULL;
939 struct hlist_node *node;
941 if (hlist_empty(head)) {
942 hlist_add_head_rcu(&new->hlist, head);
943 } else {
944 hlist_for_each_entry(li, node, head, hlist) {
945 if (new->plen > li->plen)
946 break;
948 last = li;
950 if (last)
951 hlist_add_after_rcu(&last->hlist, &new->hlist);
952 else
953 hlist_add_before_rcu(&new->hlist, &li->hlist);
957 /* rcu_read_lock needs to be hold by caller from readside */
959 static struct leaf *
960 fib_find_node(struct trie *t, u32 key)
962 int pos;
963 struct tnode *tn;
964 struct node *n;
966 pos = 0;
967 n = rcu_dereference_check(t->trie,
968 rcu_read_lock_held() ||
969 lockdep_rtnl_is_held());
971 while (n != NULL && NODE_TYPE(n) == T_TNODE) {
972 tn = (struct tnode *) n;
974 check_tnode(tn);
976 if (tkey_sub_equals(tn->key, pos, tn->pos-pos, key)) {
977 pos = tn->pos + tn->bits;
978 n = tnode_get_child_rcu(tn,
979 tkey_extract_bits(key,
980 tn->pos,
981 tn->bits));
982 } else
983 break;
985 /* Case we have found a leaf. Compare prefixes */
987 if (n != NULL && IS_LEAF(n) && tkey_equals(key, n->key))
988 return (struct leaf *)n;
990 return NULL;
993 static void trie_rebalance(struct trie *t, struct tnode *tn)
995 int wasfull;
996 t_key cindex, key;
997 struct tnode *tp;
999 key = tn->key;
1001 while (tn != NULL && (tp = node_parent((struct node *)tn)) != NULL) {
1002 cindex = tkey_extract_bits(key, tp->pos, tp->bits);
1003 wasfull = tnode_full(tp, tnode_get_child(tp, cindex));
1004 tn = (struct tnode *) resize(t, (struct tnode *)tn);
1006 tnode_put_child_reorg((struct tnode *)tp, cindex,
1007 (struct node *)tn, wasfull);
1009 tp = node_parent((struct node *) tn);
1010 if (!tp)
1011 rcu_assign_pointer(t->trie, (struct node *)tn);
1013 tnode_free_flush();
1014 if (!tp)
1015 break;
1016 tn = tp;
1019 /* Handle last (top) tnode */
1020 if (IS_TNODE(tn))
1021 tn = (struct tnode *)resize(t, (struct tnode *)tn);
1023 rcu_assign_pointer(t->trie, (struct node *)tn);
1024 tnode_free_flush();
1027 /* only used from updater-side */
1029 static struct list_head *fib_insert_node(struct trie *t, u32 key, int plen)
1031 int pos, newpos;
1032 struct tnode *tp = NULL, *tn = NULL;
1033 struct node *n;
1034 struct leaf *l;
1035 int missbit;
1036 struct list_head *fa_head = NULL;
1037 struct leaf_info *li;
1038 t_key cindex;
1040 pos = 0;
1041 n = t->trie;
1043 /* If we point to NULL, stop. Either the tree is empty and we should
1044 * just put a new leaf in if, or we have reached an empty child slot,
1045 * and we should just put our new leaf in that.
1046 * If we point to a T_TNODE, check if it matches our key. Note that
1047 * a T_TNODE might be skipping any number of bits - its 'pos' need
1048 * not be the parent's 'pos'+'bits'!
1050 * If it does match the current key, get pos/bits from it, extract
1051 * the index from our key, push the T_TNODE and walk the tree.
1053 * If it doesn't, we have to replace it with a new T_TNODE.
1055 * If we point to a T_LEAF, it might or might not have the same key
1056 * as we do. If it does, just change the value, update the T_LEAF's
1057 * value, and return it.
1058 * If it doesn't, we need to replace it with a T_TNODE.
1061 while (n != NULL && NODE_TYPE(n) == T_TNODE) {
1062 tn = (struct tnode *) n;
1064 check_tnode(tn);
1066 if (tkey_sub_equals(tn->key, pos, tn->pos-pos, key)) {
1067 tp = tn;
1068 pos = tn->pos + tn->bits;
1069 n = tnode_get_child(tn,
1070 tkey_extract_bits(key,
1071 tn->pos,
1072 tn->bits));
1074 BUG_ON(n && node_parent(n) != tn);
1075 } else
1076 break;
1080 * n ----> NULL, LEAF or TNODE
1082 * tp is n's (parent) ----> NULL or TNODE
1085 BUG_ON(tp && IS_LEAF(tp));
1087 /* Case 1: n is a leaf. Compare prefixes */
1089 if (n != NULL && IS_LEAF(n) && tkey_equals(key, n->key)) {
1090 l = (struct leaf *) n;
1091 li = leaf_info_new(plen);
1093 if (!li)
1094 return NULL;
1096 fa_head = &li->falh;
1097 insert_leaf_info(&l->list, li);
1098 goto done;
1100 l = leaf_new();
1102 if (!l)
1103 return NULL;
1105 l->key = key;
1106 li = leaf_info_new(plen);
1108 if (!li) {
1109 free_leaf(l);
1110 return NULL;
1113 fa_head = &li->falh;
1114 insert_leaf_info(&l->list, li);
1116 if (t->trie && n == NULL) {
1117 /* Case 2: n is NULL, and will just insert a new leaf */
1119 node_set_parent((struct node *)l, tp);
1121 cindex = tkey_extract_bits(key, tp->pos, tp->bits);
1122 put_child(t, (struct tnode *)tp, cindex, (struct node *)l);
1123 } else {
1124 /* Case 3: n is a LEAF or a TNODE and the key doesn't match. */
1126 * Add a new tnode here
1127 * first tnode need some special handling
1130 if (tp)
1131 pos = tp->pos+tp->bits;
1132 else
1133 pos = 0;
1135 if (n) {
1136 newpos = tkey_mismatch(key, pos, n->key);
1137 tn = tnode_new(n->key, newpos, 1);
1138 } else {
1139 newpos = 0;
1140 tn = tnode_new(key, newpos, 1); /* First tnode */
1143 if (!tn) {
1144 free_leaf_info(li);
1145 free_leaf(l);
1146 return NULL;
1149 node_set_parent((struct node *)tn, tp);
1151 missbit = tkey_extract_bits(key, newpos, 1);
1152 put_child(t, tn, missbit, (struct node *)l);
1153 put_child(t, tn, 1-missbit, n);
1155 if (tp) {
1156 cindex = tkey_extract_bits(key, tp->pos, tp->bits);
1157 put_child(t, (struct tnode *)tp, cindex,
1158 (struct node *)tn);
1159 } else {
1160 rcu_assign_pointer(t->trie, (struct node *)tn);
1161 tp = tn;
1165 if (tp && tp->pos + tp->bits > 32)
1166 pr_warning("fib_trie"
1167 " tp=%p pos=%d, bits=%d, key=%0x plen=%d\n",
1168 tp, tp->pos, tp->bits, key, plen);
1170 /* Rebalance the trie */
1172 trie_rebalance(t, tp);
1173 done:
1174 return fa_head;
1178 * Caller must hold RTNL.
1180 int fib_table_insert(struct fib_table *tb, struct fib_config *cfg)
1182 struct trie *t = (struct trie *) tb->tb_data;
1183 struct fib_alias *fa, *new_fa;
1184 struct list_head *fa_head = NULL;
1185 struct fib_info *fi;
1186 int plen = cfg->fc_dst_len;
1187 u8 tos = cfg->fc_tos;
1188 u32 key, mask;
1189 int err;
1190 struct leaf *l;
1192 if (plen > 32)
1193 return -EINVAL;
1195 key = ntohl(cfg->fc_dst);
1197 pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen);
1199 mask = ntohl(inet_make_mask(plen));
1201 if (key & ~mask)
1202 return -EINVAL;
1204 key = key & mask;
1206 fi = fib_create_info(cfg);
1207 if (IS_ERR(fi)) {
1208 err = PTR_ERR(fi);
1209 goto err;
1212 l = fib_find_node(t, key);
1213 fa = NULL;
1215 if (l) {
1216 fa_head = get_fa_head(l, plen);
1217 fa = fib_find_alias(fa_head, tos, fi->fib_priority);
1220 /* Now fa, if non-NULL, points to the first fib alias
1221 * with the same keys [prefix,tos,priority], if such key already
1222 * exists or to the node before which we will insert new one.
1224 * If fa is NULL, we will need to allocate a new one and
1225 * insert to the head of f.
1227 * If f is NULL, no fib node matched the destination key
1228 * and we need to allocate a new one of those as well.
1231 if (fa && fa->fa_tos == tos &&
1232 fa->fa_info->fib_priority == fi->fib_priority) {
1233 struct fib_alias *fa_first, *fa_match;
1235 err = -EEXIST;
1236 if (cfg->fc_nlflags & NLM_F_EXCL)
1237 goto out;
1239 /* We have 2 goals:
1240 * 1. Find exact match for type, scope, fib_info to avoid
1241 * duplicate routes
1242 * 2. Find next 'fa' (or head), NLM_F_APPEND inserts before it
1244 fa_match = NULL;
1245 fa_first = fa;
1246 fa = list_entry(fa->fa_list.prev, struct fib_alias, fa_list);
1247 list_for_each_entry_continue(fa, fa_head, fa_list) {
1248 if (fa->fa_tos != tos)
1249 break;
1250 if (fa->fa_info->fib_priority != fi->fib_priority)
1251 break;
1252 if (fa->fa_type == cfg->fc_type &&
1253 fa->fa_scope == cfg->fc_scope &&
1254 fa->fa_info == fi) {
1255 fa_match = fa;
1256 break;
1260 if (cfg->fc_nlflags & NLM_F_REPLACE) {
1261 struct fib_info *fi_drop;
1262 u8 state;
1264 fa = fa_first;
1265 if (fa_match) {
1266 if (fa == fa_match)
1267 err = 0;
1268 goto out;
1270 err = -ENOBUFS;
1271 new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1272 if (new_fa == NULL)
1273 goto out;
1275 fi_drop = fa->fa_info;
1276 new_fa->fa_tos = fa->fa_tos;
1277 new_fa->fa_info = fi;
1278 new_fa->fa_type = cfg->fc_type;
1279 new_fa->fa_scope = cfg->fc_scope;
1280 state = fa->fa_state;
1281 new_fa->fa_state = state & ~FA_S_ACCESSED;
1283 list_replace_rcu(&fa->fa_list, &new_fa->fa_list);
1284 alias_free_mem_rcu(fa);
1286 fib_release_info(fi_drop);
1287 if (state & FA_S_ACCESSED)
1288 rt_cache_flush(cfg->fc_nlinfo.nl_net, -1);
1289 rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen,
1290 tb->tb_id, &cfg->fc_nlinfo, NLM_F_REPLACE);
1292 goto succeeded;
1294 /* Error if we find a perfect match which
1295 * uses the same scope, type, and nexthop
1296 * information.
1298 if (fa_match)
1299 goto out;
1301 if (!(cfg->fc_nlflags & NLM_F_APPEND))
1302 fa = fa_first;
1304 err = -ENOENT;
1305 if (!(cfg->fc_nlflags & NLM_F_CREATE))
1306 goto out;
1308 err = -ENOBUFS;
1309 new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1310 if (new_fa == NULL)
1311 goto out;
1313 new_fa->fa_info = fi;
1314 new_fa->fa_tos = tos;
1315 new_fa->fa_type = cfg->fc_type;
1316 new_fa->fa_scope = cfg->fc_scope;
1317 new_fa->fa_state = 0;
1319 * Insert new entry to the list.
1322 if (!fa_head) {
1323 fa_head = fib_insert_node(t, key, plen);
1324 if (unlikely(!fa_head)) {
1325 err = -ENOMEM;
1326 goto out_free_new_fa;
1330 list_add_tail_rcu(&new_fa->fa_list,
1331 (fa ? &fa->fa_list : fa_head));
1333 rt_cache_flush(cfg->fc_nlinfo.nl_net, -1);
1334 rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id,
1335 &cfg->fc_nlinfo, 0);
1336 succeeded:
1337 return 0;
1339 out_free_new_fa:
1340 kmem_cache_free(fn_alias_kmem, new_fa);
1341 out:
1342 fib_release_info(fi);
1343 err:
1344 return err;
1347 /* should be called with rcu_read_lock */
1348 static int check_leaf(struct trie *t, struct leaf *l,
1349 t_key key, const struct flowi *flp,
1350 struct fib_result *res)
1352 struct leaf_info *li;
1353 struct hlist_head *hhead = &l->list;
1354 struct hlist_node *node;
1356 hlist_for_each_entry_rcu(li, node, hhead, hlist) {
1357 int err;
1358 int plen = li->plen;
1359 __be32 mask = inet_make_mask(plen);
1361 if (l->key != (key & ntohl(mask)))
1362 continue;
1364 err = fib_semantic_match(&li->falh, flp, res, plen);
1366 #ifdef CONFIG_IP_FIB_TRIE_STATS
1367 if (err <= 0)
1368 t->stats.semantic_match_passed++;
1369 else
1370 t->stats.semantic_match_miss++;
1371 #endif
1372 if (err <= 0)
1373 return err;
1376 return 1;
1379 int fib_table_lookup(struct fib_table *tb, const struct flowi *flp,
1380 struct fib_result *res)
1382 struct trie *t = (struct trie *) tb->tb_data;
1383 int ret;
1384 struct node *n;
1385 struct tnode *pn;
1386 int pos, bits;
1387 t_key key = ntohl(flp->fl4_dst);
1388 int chopped_off;
1389 t_key cindex = 0;
1390 int current_prefix_length = KEYLENGTH;
1391 struct tnode *cn;
1392 t_key node_prefix, key_prefix, pref_mismatch;
1393 int mp;
1395 rcu_read_lock();
1397 n = rcu_dereference(t->trie);
1398 if (!n)
1399 goto failed;
1401 #ifdef CONFIG_IP_FIB_TRIE_STATS
1402 t->stats.gets++;
1403 #endif
1405 /* Just a leaf? */
1406 if (IS_LEAF(n)) {
1407 ret = check_leaf(t, (struct leaf *)n, key, flp, res);
1408 goto found;
1411 pn = (struct tnode *) n;
1412 chopped_off = 0;
1414 while (pn) {
1415 pos = pn->pos;
1416 bits = pn->bits;
1418 if (!chopped_off)
1419 cindex = tkey_extract_bits(mask_pfx(key, current_prefix_length),
1420 pos, bits);
1422 n = tnode_get_child_rcu(pn, cindex);
1424 if (n == NULL) {
1425 #ifdef CONFIG_IP_FIB_TRIE_STATS
1426 t->stats.null_node_hit++;
1427 #endif
1428 goto backtrace;
1431 if (IS_LEAF(n)) {
1432 ret = check_leaf(t, (struct leaf *)n, key, flp, res);
1433 if (ret > 0)
1434 goto backtrace;
1435 goto found;
1438 cn = (struct tnode *)n;
1441 * It's a tnode, and we can do some extra checks here if we
1442 * like, to avoid descending into a dead-end branch.
1443 * This tnode is in the parent's child array at index
1444 * key[p_pos..p_pos+p_bits] but potentially with some bits
1445 * chopped off, so in reality the index may be just a
1446 * subprefix, padded with zero at the end.
1447 * We can also take a look at any skipped bits in this
1448 * tnode - everything up to p_pos is supposed to be ok,
1449 * and the non-chopped bits of the index (se previous
1450 * paragraph) are also guaranteed ok, but the rest is
1451 * considered unknown.
1453 * The skipped bits are key[pos+bits..cn->pos].
1456 /* If current_prefix_length < pos+bits, we are already doing
1457 * actual prefix matching, which means everything from
1458 * pos+(bits-chopped_off) onward must be zero along some
1459 * branch of this subtree - otherwise there is *no* valid
1460 * prefix present. Here we can only check the skipped
1461 * bits. Remember, since we have already indexed into the
1462 * parent's child array, we know that the bits we chopped of
1463 * *are* zero.
1466 /* NOTA BENE: Checking only skipped bits
1467 for the new node here */
1469 if (current_prefix_length < pos+bits) {
1470 if (tkey_extract_bits(cn->key, current_prefix_length,
1471 cn->pos - current_prefix_length)
1472 || !(cn->child[0]))
1473 goto backtrace;
1477 * If chopped_off=0, the index is fully validated and we
1478 * only need to look at the skipped bits for this, the new,
1479 * tnode. What we actually want to do is to find out if
1480 * these skipped bits match our key perfectly, or if we will
1481 * have to count on finding a matching prefix further down,
1482 * because if we do, we would like to have some way of
1483 * verifying the existence of such a prefix at this point.
1486 /* The only thing we can do at this point is to verify that
1487 * any such matching prefix can indeed be a prefix to our
1488 * key, and if the bits in the node we are inspecting that
1489 * do not match our key are not ZERO, this cannot be true.
1490 * Thus, find out where there is a mismatch (before cn->pos)
1491 * and verify that all the mismatching bits are zero in the
1492 * new tnode's key.
1496 * Note: We aren't very concerned about the piece of
1497 * the key that precede pn->pos+pn->bits, since these
1498 * have already been checked. The bits after cn->pos
1499 * aren't checked since these are by definition
1500 * "unknown" at this point. Thus, what we want to see
1501 * is if we are about to enter the "prefix matching"
1502 * state, and in that case verify that the skipped
1503 * bits that will prevail throughout this subtree are
1504 * zero, as they have to be if we are to find a
1505 * matching prefix.
1508 node_prefix = mask_pfx(cn->key, cn->pos);
1509 key_prefix = mask_pfx(key, cn->pos);
1510 pref_mismatch = key_prefix^node_prefix;
1511 mp = 0;
1514 * In short: If skipped bits in this node do not match
1515 * the search key, enter the "prefix matching"
1516 * state.directly.
1518 if (pref_mismatch) {
1519 while (!(pref_mismatch & (1<<(KEYLENGTH-1)))) {
1520 mp++;
1521 pref_mismatch = pref_mismatch << 1;
1523 key_prefix = tkey_extract_bits(cn->key, mp, cn->pos-mp);
1525 if (key_prefix != 0)
1526 goto backtrace;
1528 if (current_prefix_length >= cn->pos)
1529 current_prefix_length = mp;
1532 pn = (struct tnode *)n; /* Descend */
1533 chopped_off = 0;
1534 continue;
1536 backtrace:
1537 chopped_off++;
1539 /* As zero don't change the child key (cindex) */
1540 while ((chopped_off <= pn->bits)
1541 && !(cindex & (1<<(chopped_off-1))))
1542 chopped_off++;
1544 /* Decrease current_... with bits chopped off */
1545 if (current_prefix_length > pn->pos + pn->bits - chopped_off)
1546 current_prefix_length = pn->pos + pn->bits
1547 - chopped_off;
1550 * Either we do the actual chop off according or if we have
1551 * chopped off all bits in this tnode walk up to our parent.
1554 if (chopped_off <= pn->bits) {
1555 cindex &= ~(1 << (chopped_off-1));
1556 } else {
1557 struct tnode *parent = node_parent_rcu((struct node *) pn);
1558 if (!parent)
1559 goto failed;
1561 /* Get Child's index */
1562 cindex = tkey_extract_bits(pn->key, parent->pos, parent->bits);
1563 pn = parent;
1564 chopped_off = 0;
1566 #ifdef CONFIG_IP_FIB_TRIE_STATS
1567 t->stats.backtrack++;
1568 #endif
1569 goto backtrace;
1572 failed:
1573 ret = 1;
1574 found:
1575 rcu_read_unlock();
1576 return ret;
1580 * Remove the leaf and return parent.
1582 static void trie_leaf_remove(struct trie *t, struct leaf *l)
1584 struct tnode *tp = node_parent((struct node *) l);
1586 pr_debug("entering trie_leaf_remove(%p)\n", l);
1588 if (tp) {
1589 t_key cindex = tkey_extract_bits(l->key, tp->pos, tp->bits);
1590 put_child(t, (struct tnode *)tp, cindex, NULL);
1591 trie_rebalance(t, tp);
1592 } else
1593 rcu_assign_pointer(t->trie, NULL);
1595 free_leaf(l);
1599 * Caller must hold RTNL.
1601 int fib_table_delete(struct fib_table *tb, struct fib_config *cfg)
1603 struct trie *t = (struct trie *) tb->tb_data;
1604 u32 key, mask;
1605 int plen = cfg->fc_dst_len;
1606 u8 tos = cfg->fc_tos;
1607 struct fib_alias *fa, *fa_to_delete;
1608 struct list_head *fa_head;
1609 struct leaf *l;
1610 struct leaf_info *li;
1612 if (plen > 32)
1613 return -EINVAL;
1615 key = ntohl(cfg->fc_dst);
1616 mask = ntohl(inet_make_mask(plen));
1618 if (key & ~mask)
1619 return -EINVAL;
1621 key = key & mask;
1622 l = fib_find_node(t, key);
1624 if (!l)
1625 return -ESRCH;
1627 fa_head = get_fa_head(l, plen);
1628 fa = fib_find_alias(fa_head, tos, 0);
1630 if (!fa)
1631 return -ESRCH;
1633 pr_debug("Deleting %08x/%d tos=%d t=%p\n", key, plen, tos, t);
1635 fa_to_delete = NULL;
1636 fa = list_entry(fa->fa_list.prev, struct fib_alias, fa_list);
1637 list_for_each_entry_continue(fa, fa_head, fa_list) {
1638 struct fib_info *fi = fa->fa_info;
1640 if (fa->fa_tos != tos)
1641 break;
1643 if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) &&
1644 (cfg->fc_scope == RT_SCOPE_NOWHERE ||
1645 fa->fa_scope == cfg->fc_scope) &&
1646 (!cfg->fc_protocol ||
1647 fi->fib_protocol == cfg->fc_protocol) &&
1648 fib_nh_match(cfg, fi) == 0) {
1649 fa_to_delete = fa;
1650 break;
1654 if (!fa_to_delete)
1655 return -ESRCH;
1657 fa = fa_to_delete;
1658 rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id,
1659 &cfg->fc_nlinfo, 0);
1661 l = fib_find_node(t, key);
1662 li = find_leaf_info(l, plen);
1664 list_del_rcu(&fa->fa_list);
1666 if (list_empty(fa_head)) {
1667 hlist_del_rcu(&li->hlist);
1668 free_leaf_info(li);
1671 if (hlist_empty(&l->list))
1672 trie_leaf_remove(t, l);
1674 if (fa->fa_state & FA_S_ACCESSED)
1675 rt_cache_flush(cfg->fc_nlinfo.nl_net, -1);
1677 fib_release_info(fa->fa_info);
1678 alias_free_mem_rcu(fa);
1679 return 0;
1682 static int trie_flush_list(struct list_head *head)
1684 struct fib_alias *fa, *fa_node;
1685 int found = 0;
1687 list_for_each_entry_safe(fa, fa_node, head, fa_list) {
1688 struct fib_info *fi = fa->fa_info;
1690 if (fi && (fi->fib_flags & RTNH_F_DEAD)) {
1691 list_del_rcu(&fa->fa_list);
1692 fib_release_info(fa->fa_info);
1693 alias_free_mem_rcu(fa);
1694 found++;
1697 return found;
1700 static int trie_flush_leaf(struct leaf *l)
1702 int found = 0;
1703 struct hlist_head *lih = &l->list;
1704 struct hlist_node *node, *tmp;
1705 struct leaf_info *li = NULL;
1707 hlist_for_each_entry_safe(li, node, tmp, lih, hlist) {
1708 found += trie_flush_list(&li->falh);
1710 if (list_empty(&li->falh)) {
1711 hlist_del_rcu(&li->hlist);
1712 free_leaf_info(li);
1715 return found;
1719 * Scan for the next right leaf starting at node p->child[idx]
1720 * Since we have back pointer, no recursion necessary.
1722 static struct leaf *leaf_walk_rcu(struct tnode *p, struct node *c)
1724 do {
1725 t_key idx;
1727 if (c)
1728 idx = tkey_extract_bits(c->key, p->pos, p->bits) + 1;
1729 else
1730 idx = 0;
1732 while (idx < 1u << p->bits) {
1733 c = tnode_get_child_rcu(p, idx++);
1734 if (!c)
1735 continue;
1737 if (IS_LEAF(c)) {
1738 prefetch(p->child[idx]);
1739 return (struct leaf *) c;
1742 /* Rescan start scanning in new node */
1743 p = (struct tnode *) c;
1744 idx = 0;
1747 /* Node empty, walk back up to parent */
1748 c = (struct node *) p;
1749 } while ( (p = node_parent_rcu(c)) != NULL);
1751 return NULL; /* Root of trie */
1754 static struct leaf *trie_firstleaf(struct trie *t)
1756 struct tnode *n = (struct tnode *) rcu_dereference(t->trie);
1758 if (!n)
1759 return NULL;
1761 if (IS_LEAF(n)) /* trie is just a leaf */
1762 return (struct leaf *) n;
1764 return leaf_walk_rcu(n, NULL);
1767 static struct leaf *trie_nextleaf(struct leaf *l)
1769 struct node *c = (struct node *) l;
1770 struct tnode *p = node_parent_rcu(c);
1772 if (!p)
1773 return NULL; /* trie with just one leaf */
1775 return leaf_walk_rcu(p, c);
1778 static struct leaf *trie_leafindex(struct trie *t, int index)
1780 struct leaf *l = trie_firstleaf(t);
1782 while (l && index-- > 0)
1783 l = trie_nextleaf(l);
1785 return l;
1790 * Caller must hold RTNL.
1792 int fib_table_flush(struct fib_table *tb)
1794 struct trie *t = (struct trie *) tb->tb_data;
1795 struct leaf *l, *ll = NULL;
1796 int found = 0;
1798 for (l = trie_firstleaf(t); l; l = trie_nextleaf(l)) {
1799 found += trie_flush_leaf(l);
1801 if (ll && hlist_empty(&ll->list))
1802 trie_leaf_remove(t, ll);
1803 ll = l;
1806 if (ll && hlist_empty(&ll->list))
1807 trie_leaf_remove(t, ll);
1809 pr_debug("trie_flush found=%d\n", found);
1810 return found;
1813 void fib_table_select_default(struct fib_table *tb,
1814 const struct flowi *flp,
1815 struct fib_result *res)
1817 struct trie *t = (struct trie *) tb->tb_data;
1818 int order, last_idx;
1819 struct fib_info *fi = NULL;
1820 struct fib_info *last_resort;
1821 struct fib_alias *fa = NULL;
1822 struct list_head *fa_head;
1823 struct leaf *l;
1825 last_idx = -1;
1826 last_resort = NULL;
1827 order = -1;
1829 rcu_read_lock();
1831 l = fib_find_node(t, 0);
1832 if (!l)
1833 goto out;
1835 fa_head = get_fa_head(l, 0);
1836 if (!fa_head)
1837 goto out;
1839 if (list_empty(fa_head))
1840 goto out;
1842 list_for_each_entry_rcu(fa, fa_head, fa_list) {
1843 struct fib_info *next_fi = fa->fa_info;
1845 if (fa->fa_scope != res->scope ||
1846 fa->fa_type != RTN_UNICAST)
1847 continue;
1849 if (next_fi->fib_priority > res->fi->fib_priority)
1850 break;
1851 if (!next_fi->fib_nh[0].nh_gw ||
1852 next_fi->fib_nh[0].nh_scope != RT_SCOPE_LINK)
1853 continue;
1854 fa->fa_state |= FA_S_ACCESSED;
1856 if (fi == NULL) {
1857 if (next_fi != res->fi)
1858 break;
1859 } else if (!fib_detect_death(fi, order, &last_resort,
1860 &last_idx, tb->tb_default)) {
1861 fib_result_assign(res, fi);
1862 tb->tb_default = order;
1863 goto out;
1865 fi = next_fi;
1866 order++;
1868 if (order <= 0 || fi == NULL) {
1869 tb->tb_default = -1;
1870 goto out;
1873 if (!fib_detect_death(fi, order, &last_resort, &last_idx,
1874 tb->tb_default)) {
1875 fib_result_assign(res, fi);
1876 tb->tb_default = order;
1877 goto out;
1879 if (last_idx >= 0)
1880 fib_result_assign(res, last_resort);
1881 tb->tb_default = last_idx;
1882 out:
1883 rcu_read_unlock();
1886 static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah,
1887 struct fib_table *tb,
1888 struct sk_buff *skb, struct netlink_callback *cb)
1890 int i, s_i;
1891 struct fib_alias *fa;
1892 __be32 xkey = htonl(key);
1894 s_i = cb->args[5];
1895 i = 0;
1897 /* rcu_read_lock is hold by caller */
1899 list_for_each_entry_rcu(fa, fah, fa_list) {
1900 if (i < s_i) {
1901 i++;
1902 continue;
1905 if (fib_dump_info(skb, NETLINK_CB(cb->skb).pid,
1906 cb->nlh->nlmsg_seq,
1907 RTM_NEWROUTE,
1908 tb->tb_id,
1909 fa->fa_type,
1910 fa->fa_scope,
1911 xkey,
1912 plen,
1913 fa->fa_tos,
1914 fa->fa_info, NLM_F_MULTI) < 0) {
1915 cb->args[5] = i;
1916 return -1;
1918 i++;
1920 cb->args[5] = i;
1921 return skb->len;
1924 static int fn_trie_dump_leaf(struct leaf *l, struct fib_table *tb,
1925 struct sk_buff *skb, struct netlink_callback *cb)
1927 struct leaf_info *li;
1928 struct hlist_node *node;
1929 int i, s_i;
1931 s_i = cb->args[4];
1932 i = 0;
1934 /* rcu_read_lock is hold by caller */
1935 hlist_for_each_entry_rcu(li, node, &l->list, hlist) {
1936 if (i < s_i) {
1937 i++;
1938 continue;
1941 if (i > s_i)
1942 cb->args[5] = 0;
1944 if (list_empty(&li->falh))
1945 continue;
1947 if (fn_trie_dump_fa(l->key, li->plen, &li->falh, tb, skb, cb) < 0) {
1948 cb->args[4] = i;
1949 return -1;
1951 i++;
1954 cb->args[4] = i;
1955 return skb->len;
1958 int fib_table_dump(struct fib_table *tb, struct sk_buff *skb,
1959 struct netlink_callback *cb)
1961 struct leaf *l;
1962 struct trie *t = (struct trie *) tb->tb_data;
1963 t_key key = cb->args[2];
1964 int count = cb->args[3];
1966 rcu_read_lock();
1967 /* Dump starting at last key.
1968 * Note: 0.0.0.0/0 (ie default) is first key.
1970 if (count == 0)
1971 l = trie_firstleaf(t);
1972 else {
1973 /* Normally, continue from last key, but if that is missing
1974 * fallback to using slow rescan
1976 l = fib_find_node(t, key);
1977 if (!l)
1978 l = trie_leafindex(t, count);
1981 while (l) {
1982 cb->args[2] = l->key;
1983 if (fn_trie_dump_leaf(l, tb, skb, cb) < 0) {
1984 cb->args[3] = count;
1985 rcu_read_unlock();
1986 return -1;
1989 ++count;
1990 l = trie_nextleaf(l);
1991 memset(&cb->args[4], 0,
1992 sizeof(cb->args) - 4*sizeof(cb->args[0]));
1994 cb->args[3] = count;
1995 rcu_read_unlock();
1997 return skb->len;
2000 void __init fib_hash_init(void)
2002 fn_alias_kmem = kmem_cache_create("ip_fib_alias",
2003 sizeof(struct fib_alias),
2004 0, SLAB_PANIC, NULL);
2006 trie_leaf_kmem = kmem_cache_create("ip_fib_trie",
2007 max(sizeof(struct leaf),
2008 sizeof(struct leaf_info)),
2009 0, SLAB_PANIC, NULL);
2013 /* Fix more generic FIB names for init later */
2014 struct fib_table *fib_hash_table(u32 id)
2016 struct fib_table *tb;
2017 struct trie *t;
2019 tb = kmalloc(sizeof(struct fib_table) + sizeof(struct trie),
2020 GFP_KERNEL);
2021 if (tb == NULL)
2022 return NULL;
2024 tb->tb_id = id;
2025 tb->tb_default = -1;
2027 t = (struct trie *) tb->tb_data;
2028 memset(t, 0, sizeof(*t));
2030 if (id == RT_TABLE_LOCAL)
2031 pr_info("IPv4 FIB: Using LC-trie version %s\n", VERSION);
2033 return tb;
2036 #ifdef CONFIG_PROC_FS
2037 /* Depth first Trie walk iterator */
2038 struct fib_trie_iter {
2039 struct seq_net_private p;
2040 struct fib_table *tb;
2041 struct tnode *tnode;
2042 unsigned index;
2043 unsigned depth;
2046 static struct node *fib_trie_get_next(struct fib_trie_iter *iter)
2048 struct tnode *tn = iter->tnode;
2049 unsigned cindex = iter->index;
2050 struct tnode *p;
2052 /* A single entry routing table */
2053 if (!tn)
2054 return NULL;
2056 pr_debug("get_next iter={node=%p index=%d depth=%d}\n",
2057 iter->tnode, iter->index, iter->depth);
2058 rescan:
2059 while (cindex < (1<<tn->bits)) {
2060 struct node *n = tnode_get_child_rcu(tn, cindex);
2062 if (n) {
2063 if (IS_LEAF(n)) {
2064 iter->tnode = tn;
2065 iter->index = cindex + 1;
2066 } else {
2067 /* push down one level */
2068 iter->tnode = (struct tnode *) n;
2069 iter->index = 0;
2070 ++iter->depth;
2072 return n;
2075 ++cindex;
2078 /* Current node exhausted, pop back up */
2079 p = node_parent_rcu((struct node *)tn);
2080 if (p) {
2081 cindex = tkey_extract_bits(tn->key, p->pos, p->bits)+1;
2082 tn = p;
2083 --iter->depth;
2084 goto rescan;
2087 /* got root? */
2088 return NULL;
2091 static struct node *fib_trie_get_first(struct fib_trie_iter *iter,
2092 struct trie *t)
2094 struct node *n;
2096 if (!t)
2097 return NULL;
2099 n = rcu_dereference(t->trie);
2100 if (!n)
2101 return NULL;
2103 if (IS_TNODE(n)) {
2104 iter->tnode = (struct tnode *) n;
2105 iter->index = 0;
2106 iter->depth = 1;
2107 } else {
2108 iter->tnode = NULL;
2109 iter->index = 0;
2110 iter->depth = 0;
2113 return n;
2116 static void trie_collect_stats(struct trie *t, struct trie_stat *s)
2118 struct node *n;
2119 struct fib_trie_iter iter;
2121 memset(s, 0, sizeof(*s));
2123 rcu_read_lock();
2124 for (n = fib_trie_get_first(&iter, t); n; n = fib_trie_get_next(&iter)) {
2125 if (IS_LEAF(n)) {
2126 struct leaf *l = (struct leaf *)n;
2127 struct leaf_info *li;
2128 struct hlist_node *tmp;
2130 s->leaves++;
2131 s->totdepth += iter.depth;
2132 if (iter.depth > s->maxdepth)
2133 s->maxdepth = iter.depth;
2135 hlist_for_each_entry_rcu(li, tmp, &l->list, hlist)
2136 ++s->prefixes;
2137 } else {
2138 const struct tnode *tn = (const struct tnode *) n;
2139 int i;
2141 s->tnodes++;
2142 if (tn->bits < MAX_STAT_DEPTH)
2143 s->nodesizes[tn->bits]++;
2145 for (i = 0; i < (1<<tn->bits); i++)
2146 if (!tn->child[i])
2147 s->nullpointers++;
2150 rcu_read_unlock();
2154 * This outputs /proc/net/fib_triestats
2156 static void trie_show_stats(struct seq_file *seq, struct trie_stat *stat)
2158 unsigned i, max, pointers, bytes, avdepth;
2160 if (stat->leaves)
2161 avdepth = stat->totdepth*100 / stat->leaves;
2162 else
2163 avdepth = 0;
2165 seq_printf(seq, "\tAver depth: %u.%02d\n",
2166 avdepth / 100, avdepth % 100);
2167 seq_printf(seq, "\tMax depth: %u\n", stat->maxdepth);
2169 seq_printf(seq, "\tLeaves: %u\n", stat->leaves);
2170 bytes = sizeof(struct leaf) * stat->leaves;
2172 seq_printf(seq, "\tPrefixes: %u\n", stat->prefixes);
2173 bytes += sizeof(struct leaf_info) * stat->prefixes;
2175 seq_printf(seq, "\tInternal nodes: %u\n\t", stat->tnodes);
2176 bytes += sizeof(struct tnode) * stat->tnodes;
2178 max = MAX_STAT_DEPTH;
2179 while (max > 0 && stat->nodesizes[max-1] == 0)
2180 max--;
2182 pointers = 0;
2183 for (i = 1; i <= max; i++)
2184 if (stat->nodesizes[i] != 0) {
2185 seq_printf(seq, " %u: %u", i, stat->nodesizes[i]);
2186 pointers += (1<<i) * stat->nodesizes[i];
2188 seq_putc(seq, '\n');
2189 seq_printf(seq, "\tPointers: %u\n", pointers);
2191 bytes += sizeof(struct node *) * pointers;
2192 seq_printf(seq, "Null ptrs: %u\n", stat->nullpointers);
2193 seq_printf(seq, "Total size: %u kB\n", (bytes + 1023) / 1024);
2196 #ifdef CONFIG_IP_FIB_TRIE_STATS
2197 static void trie_show_usage(struct seq_file *seq,
2198 const struct trie_use_stats *stats)
2200 seq_printf(seq, "\nCounters:\n---------\n");
2201 seq_printf(seq, "gets = %u\n", stats->gets);
2202 seq_printf(seq, "backtracks = %u\n", stats->backtrack);
2203 seq_printf(seq, "semantic match passed = %u\n",
2204 stats->semantic_match_passed);
2205 seq_printf(seq, "semantic match miss = %u\n",
2206 stats->semantic_match_miss);
2207 seq_printf(seq, "null node hit= %u\n", stats->null_node_hit);
2208 seq_printf(seq, "skipped node resize = %u\n\n",
2209 stats->resize_node_skipped);
2211 #endif /* CONFIG_IP_FIB_TRIE_STATS */
2213 static void fib_table_print(struct seq_file *seq, struct fib_table *tb)
2215 if (tb->tb_id == RT_TABLE_LOCAL)
2216 seq_puts(seq, "Local:\n");
2217 else if (tb->tb_id == RT_TABLE_MAIN)
2218 seq_puts(seq, "Main:\n");
2219 else
2220 seq_printf(seq, "Id %d:\n", tb->tb_id);
2224 static int fib_triestat_seq_show(struct seq_file *seq, void *v)
2226 struct net *net = (struct net *)seq->private;
2227 unsigned int h;
2229 seq_printf(seq,
2230 "Basic info: size of leaf:"
2231 " %Zd bytes, size of tnode: %Zd bytes.\n",
2232 sizeof(struct leaf), sizeof(struct tnode));
2234 for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2235 struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2236 struct hlist_node *node;
2237 struct fib_table *tb;
2239 hlist_for_each_entry_rcu(tb, node, head, tb_hlist) {
2240 struct trie *t = (struct trie *) tb->tb_data;
2241 struct trie_stat stat;
2243 if (!t)
2244 continue;
2246 fib_table_print(seq, tb);
2248 trie_collect_stats(t, &stat);
2249 trie_show_stats(seq, &stat);
2250 #ifdef CONFIG_IP_FIB_TRIE_STATS
2251 trie_show_usage(seq, &t->stats);
2252 #endif
2256 return 0;
2259 static int fib_triestat_seq_open(struct inode *inode, struct file *file)
2261 return single_open_net(inode, file, fib_triestat_seq_show);
2264 static const struct file_operations fib_triestat_fops = {
2265 .owner = THIS_MODULE,
2266 .open = fib_triestat_seq_open,
2267 .read = seq_read,
2268 .llseek = seq_lseek,
2269 .release = single_release_net,
2272 static struct node *fib_trie_get_idx(struct seq_file *seq, loff_t pos)
2274 struct fib_trie_iter *iter = seq->private;
2275 struct net *net = seq_file_net(seq);
2276 loff_t idx = 0;
2277 unsigned int h;
2279 for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2280 struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2281 struct hlist_node *node;
2282 struct fib_table *tb;
2284 hlist_for_each_entry_rcu(tb, node, head, tb_hlist) {
2285 struct node *n;
2287 for (n = fib_trie_get_first(iter,
2288 (struct trie *) tb->tb_data);
2289 n; n = fib_trie_get_next(iter))
2290 if (pos == idx++) {
2291 iter->tb = tb;
2292 return n;
2297 return NULL;
2300 static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos)
2301 __acquires(RCU)
2303 rcu_read_lock();
2304 return fib_trie_get_idx(seq, *pos);
2307 static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos)
2309 struct fib_trie_iter *iter = seq->private;
2310 struct net *net = seq_file_net(seq);
2311 struct fib_table *tb = iter->tb;
2312 struct hlist_node *tb_node;
2313 unsigned int h;
2314 struct node *n;
2316 ++*pos;
2317 /* next node in same table */
2318 n = fib_trie_get_next(iter);
2319 if (n)
2320 return n;
2322 /* walk rest of this hash chain */
2323 h = tb->tb_id & (FIB_TABLE_HASHSZ - 1);
2324 while ( (tb_node = rcu_dereference(tb->tb_hlist.next)) ) {
2325 tb = hlist_entry(tb_node, struct fib_table, tb_hlist);
2326 n = fib_trie_get_first(iter, (struct trie *) tb->tb_data);
2327 if (n)
2328 goto found;
2331 /* new hash chain */
2332 while (++h < FIB_TABLE_HASHSZ) {
2333 struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2334 hlist_for_each_entry_rcu(tb, tb_node, head, tb_hlist) {
2335 n = fib_trie_get_first(iter, (struct trie *) tb->tb_data);
2336 if (n)
2337 goto found;
2340 return NULL;
2342 found:
2343 iter->tb = tb;
2344 return n;
2347 static void fib_trie_seq_stop(struct seq_file *seq, void *v)
2348 __releases(RCU)
2350 rcu_read_unlock();
2353 static void seq_indent(struct seq_file *seq, int n)
2355 while (n-- > 0) seq_puts(seq, " ");
2358 static inline const char *rtn_scope(char *buf, size_t len, enum rt_scope_t s)
2360 switch (s) {
2361 case RT_SCOPE_UNIVERSE: return "universe";
2362 case RT_SCOPE_SITE: return "site";
2363 case RT_SCOPE_LINK: return "link";
2364 case RT_SCOPE_HOST: return "host";
2365 case RT_SCOPE_NOWHERE: return "nowhere";
2366 default:
2367 snprintf(buf, len, "scope=%d", s);
2368 return buf;
2372 static const char *const rtn_type_names[__RTN_MAX] = {
2373 [RTN_UNSPEC] = "UNSPEC",
2374 [RTN_UNICAST] = "UNICAST",
2375 [RTN_LOCAL] = "LOCAL",
2376 [RTN_BROADCAST] = "BROADCAST",
2377 [RTN_ANYCAST] = "ANYCAST",
2378 [RTN_MULTICAST] = "MULTICAST",
2379 [RTN_BLACKHOLE] = "BLACKHOLE",
2380 [RTN_UNREACHABLE] = "UNREACHABLE",
2381 [RTN_PROHIBIT] = "PROHIBIT",
2382 [RTN_THROW] = "THROW",
2383 [RTN_NAT] = "NAT",
2384 [RTN_XRESOLVE] = "XRESOLVE",
2387 static inline const char *rtn_type(char *buf, size_t len, unsigned t)
2389 if (t < __RTN_MAX && rtn_type_names[t])
2390 return rtn_type_names[t];
2391 snprintf(buf, len, "type %u", t);
2392 return buf;
2395 /* Pretty print the trie */
2396 static int fib_trie_seq_show(struct seq_file *seq, void *v)
2398 const struct fib_trie_iter *iter = seq->private;
2399 struct node *n = v;
2401 if (!node_parent_rcu(n))
2402 fib_table_print(seq, iter->tb);
2404 if (IS_TNODE(n)) {
2405 struct tnode *tn = (struct tnode *) n;
2406 __be32 prf = htonl(mask_pfx(tn->key, tn->pos));
2408 seq_indent(seq, iter->depth-1);
2409 seq_printf(seq, " +-- %pI4/%d %d %d %d\n",
2410 &prf, tn->pos, tn->bits, tn->full_children,
2411 tn->empty_children);
2413 } else {
2414 struct leaf *l = (struct leaf *) n;
2415 struct leaf_info *li;
2416 struct hlist_node *node;
2417 __be32 val = htonl(l->key);
2419 seq_indent(seq, iter->depth);
2420 seq_printf(seq, " |-- %pI4\n", &val);
2422 hlist_for_each_entry_rcu(li, node, &l->list, hlist) {
2423 struct fib_alias *fa;
2425 list_for_each_entry_rcu(fa, &li->falh, fa_list) {
2426 char buf1[32], buf2[32];
2428 seq_indent(seq, iter->depth+1);
2429 seq_printf(seq, " /%d %s %s", li->plen,
2430 rtn_scope(buf1, sizeof(buf1),
2431 fa->fa_scope),
2432 rtn_type(buf2, sizeof(buf2),
2433 fa->fa_type));
2434 if (fa->fa_tos)
2435 seq_printf(seq, " tos=%d", fa->fa_tos);
2436 seq_putc(seq, '\n');
2441 return 0;
2444 static const struct seq_operations fib_trie_seq_ops = {
2445 .start = fib_trie_seq_start,
2446 .next = fib_trie_seq_next,
2447 .stop = fib_trie_seq_stop,
2448 .show = fib_trie_seq_show,
2451 static int fib_trie_seq_open(struct inode *inode, struct file *file)
2453 return seq_open_net(inode, file, &fib_trie_seq_ops,
2454 sizeof(struct fib_trie_iter));
2457 static const struct file_operations fib_trie_fops = {
2458 .owner = THIS_MODULE,
2459 .open = fib_trie_seq_open,
2460 .read = seq_read,
2461 .llseek = seq_lseek,
2462 .release = seq_release_net,
2465 struct fib_route_iter {
2466 struct seq_net_private p;
2467 struct trie *main_trie;
2468 loff_t pos;
2469 t_key key;
2472 static struct leaf *fib_route_get_idx(struct fib_route_iter *iter, loff_t pos)
2474 struct leaf *l = NULL;
2475 struct trie *t = iter->main_trie;
2477 /* use cache location of last found key */
2478 if (iter->pos > 0 && pos >= iter->pos && (l = fib_find_node(t, iter->key)))
2479 pos -= iter->pos;
2480 else {
2481 iter->pos = 0;
2482 l = trie_firstleaf(t);
2485 while (l && pos-- > 0) {
2486 iter->pos++;
2487 l = trie_nextleaf(l);
2490 if (l)
2491 iter->key = pos; /* remember it */
2492 else
2493 iter->pos = 0; /* forget it */
2495 return l;
2498 static void *fib_route_seq_start(struct seq_file *seq, loff_t *pos)
2499 __acquires(RCU)
2501 struct fib_route_iter *iter = seq->private;
2502 struct fib_table *tb;
2504 rcu_read_lock();
2505 tb = fib_get_table(seq_file_net(seq), RT_TABLE_MAIN);
2506 if (!tb)
2507 return NULL;
2509 iter->main_trie = (struct trie *) tb->tb_data;
2510 if (*pos == 0)
2511 return SEQ_START_TOKEN;
2512 else
2513 return fib_route_get_idx(iter, *pos - 1);
2516 static void *fib_route_seq_next(struct seq_file *seq, void *v, loff_t *pos)
2518 struct fib_route_iter *iter = seq->private;
2519 struct leaf *l = v;
2521 ++*pos;
2522 if (v == SEQ_START_TOKEN) {
2523 iter->pos = 0;
2524 l = trie_firstleaf(iter->main_trie);
2525 } else {
2526 iter->pos++;
2527 l = trie_nextleaf(l);
2530 if (l)
2531 iter->key = l->key;
2532 else
2533 iter->pos = 0;
2534 return l;
2537 static void fib_route_seq_stop(struct seq_file *seq, void *v)
2538 __releases(RCU)
2540 rcu_read_unlock();
2543 static unsigned fib_flag_trans(int type, __be32 mask, const struct fib_info *fi)
2545 static unsigned type2flags[RTN_MAX + 1] = {
2546 [7] = RTF_REJECT, [8] = RTF_REJECT,
2548 unsigned flags = type2flags[type];
2550 if (fi && fi->fib_nh->nh_gw)
2551 flags |= RTF_GATEWAY;
2552 if (mask == htonl(0xFFFFFFFF))
2553 flags |= RTF_HOST;
2554 flags |= RTF_UP;
2555 return flags;
2559 * This outputs /proc/net/route.
2560 * The format of the file is not supposed to be changed
2561 * and needs to be same as fib_hash output to avoid breaking
2562 * legacy utilities
2564 static int fib_route_seq_show(struct seq_file *seq, void *v)
2566 struct leaf *l = v;
2567 struct leaf_info *li;
2568 struct hlist_node *node;
2570 if (v == SEQ_START_TOKEN) {
2571 seq_printf(seq, "%-127s\n", "Iface\tDestination\tGateway "
2572 "\tFlags\tRefCnt\tUse\tMetric\tMask\t\tMTU"
2573 "\tWindow\tIRTT");
2574 return 0;
2577 hlist_for_each_entry_rcu(li, node, &l->list, hlist) {
2578 struct fib_alias *fa;
2579 __be32 mask, prefix;
2581 mask = inet_make_mask(li->plen);
2582 prefix = htonl(l->key);
2584 list_for_each_entry_rcu(fa, &li->falh, fa_list) {
2585 const struct fib_info *fi = fa->fa_info;
2586 unsigned flags = fib_flag_trans(fa->fa_type, mask, fi);
2587 int len;
2589 if (fa->fa_type == RTN_BROADCAST
2590 || fa->fa_type == RTN_MULTICAST)
2591 continue;
2593 if (fi)
2594 seq_printf(seq,
2595 "%s\t%08X\t%08X\t%04X\t%d\t%u\t"
2596 "%d\t%08X\t%d\t%u\t%u%n",
2597 fi->fib_dev ? fi->fib_dev->name : "*",
2598 prefix,
2599 fi->fib_nh->nh_gw, flags, 0, 0,
2600 fi->fib_priority,
2601 mask,
2602 (fi->fib_advmss ?
2603 fi->fib_advmss + 40 : 0),
2604 fi->fib_window,
2605 fi->fib_rtt >> 3, &len);
2606 else
2607 seq_printf(seq,
2608 "*\t%08X\t%08X\t%04X\t%d\t%u\t"
2609 "%d\t%08X\t%d\t%u\t%u%n",
2610 prefix, 0, flags, 0, 0, 0,
2611 mask, 0, 0, 0, &len);
2613 seq_printf(seq, "%*s\n", 127 - len, "");
2617 return 0;
2620 static const struct seq_operations fib_route_seq_ops = {
2621 .start = fib_route_seq_start,
2622 .next = fib_route_seq_next,
2623 .stop = fib_route_seq_stop,
2624 .show = fib_route_seq_show,
2627 static int fib_route_seq_open(struct inode *inode, struct file *file)
2629 return seq_open_net(inode, file, &fib_route_seq_ops,
2630 sizeof(struct fib_route_iter));
2633 static const struct file_operations fib_route_fops = {
2634 .owner = THIS_MODULE,
2635 .open = fib_route_seq_open,
2636 .read = seq_read,
2637 .llseek = seq_lseek,
2638 .release = seq_release_net,
2641 int __net_init fib_proc_init(struct net *net)
2643 if (!proc_net_fops_create(net, "fib_trie", S_IRUGO, &fib_trie_fops))
2644 goto out1;
2646 if (!proc_net_fops_create(net, "fib_triestat", S_IRUGO,
2647 &fib_triestat_fops))
2648 goto out2;
2650 if (!proc_net_fops_create(net, "route", S_IRUGO, &fib_route_fops))
2651 goto out3;
2653 return 0;
2655 out3:
2656 proc_net_remove(net, "fib_triestat");
2657 out2:
2658 proc_net_remove(net, "fib_trie");
2659 out1:
2660 return -ENOMEM;
2663 void __net_exit fib_proc_exit(struct net *net)
2665 proc_net_remove(net, "fib_trie");
2666 proc_net_remove(net, "fib_triestat");
2667 proc_net_remove(net, "route");
2670 #endif /* CONFIG_PROC_FS */