xref: /openbmc/linux/net/ipv4/fib_trie.c (revision 0e96cf7f)
1 // SPDX-License-Identifier: GPL-2.0-or-later
2 /*
3  *
4  *   Robert Olsson <robert.olsson@its.uu.se> Uppsala Universitet
5  *     & Swedish University of Agricultural Sciences.
6  *
7  *   Jens Laas <jens.laas@data.slu.se> Swedish University of
8  *     Agricultural Sciences.
9  *
10  *   Hans Liss <hans.liss@its.uu.se>  Uppsala Universitet
11  *
12  * This work is based on the LPC-trie which is originally described in:
13  *
14  * An experimental study of compression methods for dynamic tries
15  * Stefan Nilsson and Matti Tikkanen. Algorithmica, 33(1):19-33, 2002.
16  * http://www.csc.kth.se/~snilsson/software/dyntrie2/
17  *
18  * IP-address lookup using LC-tries. Stefan Nilsson and Gunnar Karlsson
19  * IEEE Journal on Selected Areas in Communications, 17(6):1083-1092, June 1999
20  *
21  * Code from fib_hash has been reused which includes the following header:
22  *
23  * INET		An implementation of the TCP/IP protocol suite for the LINUX
24  *		operating system.  INET is implemented using the  BSD Socket
25  *		interface as the means of communication with the user level.
26  *
27  *		IPv4 FIB: lookup engine and maintenance routines.
28  *
29  * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
30  *
31  * Substantial contributions to this work comes from:
32  *
33  *		David S. Miller, <davem@davemloft.net>
34  *		Stephen Hemminger <shemminger@osdl.org>
35  *		Paul E. McKenney <paulmck@us.ibm.com>
36  *		Patrick McHardy <kaber@trash.net>
37  */
38 
39 #define VERSION "0.409"
40 
41 #include <linux/cache.h>
42 #include <linux/uaccess.h>
43 #include <linux/bitops.h>
44 #include <linux/types.h>
45 #include <linux/kernel.h>
46 #include <linux/mm.h>
47 #include <linux/string.h>
48 #include <linux/socket.h>
49 #include <linux/sockios.h>
50 #include <linux/errno.h>
51 #include <linux/in.h>
52 #include <linux/inet.h>
53 #include <linux/inetdevice.h>
54 #include <linux/netdevice.h>
55 #include <linux/if_arp.h>
56 #include <linux/proc_fs.h>
57 #include <linux/rcupdate.h>
58 #include <linux/skbuff.h>
59 #include <linux/netlink.h>
60 #include <linux/init.h>
61 #include <linux/list.h>
62 #include <linux/slab.h>
63 #include <linux/export.h>
64 #include <linux/vmalloc.h>
65 #include <linux/notifier.h>
66 #include <net/net_namespace.h>
67 #include <net/ip.h>
68 #include <net/protocol.h>
69 #include <net/route.h>
70 #include <net/tcp.h>
71 #include <net/sock.h>
72 #include <net/ip_fib.h>
73 #include <net/fib_notifier.h>
74 #include <trace/events/fib.h>
75 #include "fib_lookup.h"
76 
77 static int call_fib_entry_notifier(struct notifier_block *nb, struct net *net,
78 				   enum fib_event_type event_type, u32 dst,
79 				   int dst_len, struct fib_alias *fa)
80 {
81 	struct fib_entry_notifier_info info = {
82 		.dst = dst,
83 		.dst_len = dst_len,
84 		.fi = fa->fa_info,
85 		.tos = fa->fa_tos,
86 		.type = fa->fa_type,
87 		.tb_id = fa->tb_id,
88 	};
89 	return call_fib4_notifier(nb, net, event_type, &info.info);
90 }
91 
92 static int call_fib_entry_notifiers(struct net *net,
93 				    enum fib_event_type event_type, u32 dst,
94 				    int dst_len, struct fib_alias *fa,
95 				    struct netlink_ext_ack *extack)
96 {
97 	struct fib_entry_notifier_info info = {
98 		.info.extack = extack,
99 		.dst = dst,
100 		.dst_len = dst_len,
101 		.fi = fa->fa_info,
102 		.tos = fa->fa_tos,
103 		.type = fa->fa_type,
104 		.tb_id = fa->tb_id,
105 	};
106 	return call_fib4_notifiers(net, event_type, &info.info);
107 }
108 
109 #define MAX_STAT_DEPTH 32
110 
111 #define KEYLENGTH	(8*sizeof(t_key))
112 #define KEY_MAX		((t_key)~0)
113 
114 typedef unsigned int t_key;
115 
116 #define IS_TRIE(n)	((n)->pos >= KEYLENGTH)
117 #define IS_TNODE(n)	((n)->bits)
118 #define IS_LEAF(n)	(!(n)->bits)
119 
120 struct key_vector {
121 	t_key key;
122 	unsigned char pos;		/* 2log(KEYLENGTH) bits needed */
123 	unsigned char bits;		/* 2log(KEYLENGTH) bits needed */
124 	unsigned char slen;
125 	union {
126 		/* This list pointer if valid if (pos | bits) == 0 (LEAF) */
127 		struct hlist_head leaf;
128 		/* This array is valid if (pos | bits) > 0 (TNODE) */
129 		struct key_vector __rcu *tnode[0];
130 	};
131 };
132 
133 struct tnode {
134 	struct rcu_head rcu;
135 	t_key empty_children;		/* KEYLENGTH bits needed */
136 	t_key full_children;		/* KEYLENGTH bits needed */
137 	struct key_vector __rcu *parent;
138 	struct key_vector kv[1];
139 #define tn_bits kv[0].bits
140 };
141 
142 #define TNODE_SIZE(n)	offsetof(struct tnode, kv[0].tnode[n])
143 #define LEAF_SIZE	TNODE_SIZE(1)
144 
145 #ifdef CONFIG_IP_FIB_TRIE_STATS
146 struct trie_use_stats {
147 	unsigned int gets;
148 	unsigned int backtrack;
149 	unsigned int semantic_match_passed;
150 	unsigned int semantic_match_miss;
151 	unsigned int null_node_hit;
152 	unsigned int resize_node_skipped;
153 };
154 #endif
155 
156 struct trie_stat {
157 	unsigned int totdepth;
158 	unsigned int maxdepth;
159 	unsigned int tnodes;
160 	unsigned int leaves;
161 	unsigned int nullpointers;
162 	unsigned int prefixes;
163 	unsigned int nodesizes[MAX_STAT_DEPTH];
164 };
165 
166 struct trie {
167 	struct key_vector kv[1];
168 #ifdef CONFIG_IP_FIB_TRIE_STATS
169 	struct trie_use_stats __percpu *stats;
170 #endif
171 };
172 
173 static struct key_vector *resize(struct trie *t, struct key_vector *tn);
174 static unsigned int tnode_free_size;
175 
176 /*
177  * synchronize_rcu after call_rcu for outstanding dirty memory; it should be
178  * especially useful before resizing the root node with PREEMPT_NONE configs;
179  * the value was obtained experimentally, aiming to avoid visible slowdown.
180  */
181 unsigned int sysctl_fib_sync_mem = 512 * 1024;
182 unsigned int sysctl_fib_sync_mem_min = 64 * 1024;
183 unsigned int sysctl_fib_sync_mem_max = 64 * 1024 * 1024;
184 
185 static struct kmem_cache *fn_alias_kmem __ro_after_init;
186 static struct kmem_cache *trie_leaf_kmem __ro_after_init;
187 
188 static inline struct tnode *tn_info(struct key_vector *kv)
189 {
190 	return container_of(kv, struct tnode, kv[0]);
191 }
192 
193 /* caller must hold RTNL */
194 #define node_parent(tn) rtnl_dereference(tn_info(tn)->parent)
195 #define get_child(tn, i) rtnl_dereference((tn)->tnode[i])
196 
197 /* caller must hold RCU read lock or RTNL */
198 #define node_parent_rcu(tn) rcu_dereference_rtnl(tn_info(tn)->parent)
199 #define get_child_rcu(tn, i) rcu_dereference_rtnl((tn)->tnode[i])
200 
201 /* wrapper for rcu_assign_pointer */
202 static inline void node_set_parent(struct key_vector *n, struct key_vector *tp)
203 {
204 	if (n)
205 		rcu_assign_pointer(tn_info(n)->parent, tp);
206 }
207 
208 #define NODE_INIT_PARENT(n, p) RCU_INIT_POINTER(tn_info(n)->parent, p)
209 
210 /* This provides us with the number of children in this node, in the case of a
211  * leaf this will return 0 meaning none of the children are accessible.
212  */
213 static inline unsigned long child_length(const struct key_vector *tn)
214 {
215 	return (1ul << tn->bits) & ~(1ul);
216 }
217 
218 #define get_cindex(key, kv) (((key) ^ (kv)->key) >> (kv)->pos)
219 
220 static inline unsigned long get_index(t_key key, struct key_vector *kv)
221 {
222 	unsigned long index = key ^ kv->key;
223 
224 	if ((BITS_PER_LONG <= KEYLENGTH) && (KEYLENGTH == kv->pos))
225 		return 0;
226 
227 	return index >> kv->pos;
228 }
229 
230 /* To understand this stuff, an understanding of keys and all their bits is
231  * necessary. Every node in the trie has a key associated with it, but not
232  * all of the bits in that key are significant.
233  *
234  * Consider a node 'n' and its parent 'tp'.
235  *
236  * If n is a leaf, every bit in its key is significant. Its presence is
237  * necessitated by path compression, since during a tree traversal (when
238  * searching for a leaf - unless we are doing an insertion) we will completely
239  * ignore all skipped bits we encounter. Thus we need to verify, at the end of
240  * a potentially successful search, that we have indeed been walking the
241  * correct key path.
242  *
243  * Note that we can never "miss" the correct key in the tree if present by
244  * following the wrong path. Path compression ensures that segments of the key
245  * that are the same for all keys with a given prefix are skipped, but the
246  * skipped part *is* identical for each node in the subtrie below the skipped
247  * bit! trie_insert() in this implementation takes care of that.
248  *
249  * if n is an internal node - a 'tnode' here, the various parts of its key
250  * have many different meanings.
251  *
252  * Example:
253  * _________________________________________________________________
254  * | i | i | i | i | i | i | i | N | N | N | S | S | S | S | S | C |
255  * -----------------------------------------------------------------
256  *  31  30  29  28  27  26  25  24  23  22  21  20  19  18  17  16
257  *
258  * _________________________________________________________________
259  * | C | C | C | u | u | u | u | u | u | u | u | u | u | u | u | u |
260  * -----------------------------------------------------------------
261  *  15  14  13  12  11  10   9   8   7   6   5   4   3   2   1   0
262  *
263  * tp->pos = 22
264  * tp->bits = 3
265  * n->pos = 13
266  * n->bits = 4
267  *
268  * First, let's just ignore the bits that come before the parent tp, that is
269  * the bits from (tp->pos + tp->bits) to 31. They are *known* but at this
270  * point we do not use them for anything.
271  *
272  * The bits from (tp->pos) to (tp->pos + tp->bits - 1) - "N", above - are the
273  * index into the parent's child array. That is, they will be used to find
274  * 'n' among tp's children.
275  *
276  * The bits from (n->pos + n->bits) to (tp->pos - 1) - "S" - are skipped bits
277  * for the node n.
278  *
279  * All the bits we have seen so far are significant to the node n. The rest
280  * of the bits are really not needed or indeed known in n->key.
281  *
282  * The bits from (n->pos) to (n->pos + n->bits - 1) - "C" - are the index into
283  * n's child array, and will of course be different for each child.
284  *
285  * The rest of the bits, from 0 to (n->pos -1) - "u" - are completely unknown
286  * at this point.
287  */
288 
289 static const int halve_threshold = 25;
290 static const int inflate_threshold = 50;
291 static const int halve_threshold_root = 15;
292 static const int inflate_threshold_root = 30;
293 
294 static void __alias_free_mem(struct rcu_head *head)
295 {
296 	struct fib_alias *fa = container_of(head, struct fib_alias, rcu);
297 	kmem_cache_free(fn_alias_kmem, fa);
298 }
299 
300 static inline void alias_free_mem_rcu(struct fib_alias *fa)
301 {
302 	call_rcu(&fa->rcu, __alias_free_mem);
303 }
304 
305 #define TNODE_KMALLOC_MAX \
306 	ilog2((PAGE_SIZE - TNODE_SIZE(0)) / sizeof(struct key_vector *))
307 #define TNODE_VMALLOC_MAX \
308 	ilog2((SIZE_MAX - TNODE_SIZE(0)) / sizeof(struct key_vector *))
309 
310 static void __node_free_rcu(struct rcu_head *head)
311 {
312 	struct tnode *n = container_of(head, struct tnode, rcu);
313 
314 	if (!n->tn_bits)
315 		kmem_cache_free(trie_leaf_kmem, n);
316 	else
317 		kvfree(n);
318 }
319 
320 #define node_free(n) call_rcu(&tn_info(n)->rcu, __node_free_rcu)
321 
322 static struct tnode *tnode_alloc(int bits)
323 {
324 	size_t size;
325 
326 	/* verify bits is within bounds */
327 	if (bits > TNODE_VMALLOC_MAX)
328 		return NULL;
329 
330 	/* determine size and verify it is non-zero and didn't overflow */
331 	size = TNODE_SIZE(1ul << bits);
332 
333 	if (size <= PAGE_SIZE)
334 		return kzalloc(size, GFP_KERNEL);
335 	else
336 		return vzalloc(size);
337 }
338 
339 static inline void empty_child_inc(struct key_vector *n)
340 {
341 	++tn_info(n)->empty_children ? : ++tn_info(n)->full_children;
342 }
343 
344 static inline void empty_child_dec(struct key_vector *n)
345 {
346 	tn_info(n)->empty_children-- ? : tn_info(n)->full_children--;
347 }
348 
349 static struct key_vector *leaf_new(t_key key, struct fib_alias *fa)
350 {
351 	struct key_vector *l;
352 	struct tnode *kv;
353 
354 	kv = kmem_cache_alloc(trie_leaf_kmem, GFP_KERNEL);
355 	if (!kv)
356 		return NULL;
357 
358 	/* initialize key vector */
359 	l = kv->kv;
360 	l->key = key;
361 	l->pos = 0;
362 	l->bits = 0;
363 	l->slen = fa->fa_slen;
364 
365 	/* link leaf to fib alias */
366 	INIT_HLIST_HEAD(&l->leaf);
367 	hlist_add_head(&fa->fa_list, &l->leaf);
368 
369 	return l;
370 }
371 
372 static struct key_vector *tnode_new(t_key key, int pos, int bits)
373 {
374 	unsigned int shift = pos + bits;
375 	struct key_vector *tn;
376 	struct tnode *tnode;
377 
378 	/* verify bits and pos their msb bits clear and values are valid */
379 	BUG_ON(!bits || (shift > KEYLENGTH));
380 
381 	tnode = tnode_alloc(bits);
382 	if (!tnode)
383 		return NULL;
384 
385 	pr_debug("AT %p s=%zu %zu\n", tnode, TNODE_SIZE(0),
386 		 sizeof(struct key_vector *) << bits);
387 
388 	if (bits == KEYLENGTH)
389 		tnode->full_children = 1;
390 	else
391 		tnode->empty_children = 1ul << bits;
392 
393 	tn = tnode->kv;
394 	tn->key = (shift < KEYLENGTH) ? (key >> shift) << shift : 0;
395 	tn->pos = pos;
396 	tn->bits = bits;
397 	tn->slen = pos;
398 
399 	return tn;
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 static inline int tnode_full(struct key_vector *tn, struct key_vector *n)
406 {
407 	return n && ((n->pos + n->bits) == tn->pos) && IS_TNODE(n);
408 }
409 
410 /* Add a child at position i overwriting the old value.
411  * Update the value of full_children and empty_children.
412  */
413 static void put_child(struct key_vector *tn, unsigned long i,
414 		      struct key_vector *n)
415 {
416 	struct key_vector *chi = get_child(tn, i);
417 	int isfull, wasfull;
418 
419 	BUG_ON(i >= child_length(tn));
420 
421 	/* update emptyChildren, overflow into fullChildren */
422 	if (!n && chi)
423 		empty_child_inc(tn);
424 	if (n && !chi)
425 		empty_child_dec(tn);
426 
427 	/* update fullChildren */
428 	wasfull = tnode_full(tn, chi);
429 	isfull = tnode_full(tn, n);
430 
431 	if (wasfull && !isfull)
432 		tn_info(tn)->full_children--;
433 	else if (!wasfull && isfull)
434 		tn_info(tn)->full_children++;
435 
436 	if (n && (tn->slen < n->slen))
437 		tn->slen = n->slen;
438 
439 	rcu_assign_pointer(tn->tnode[i], n);
440 }
441 
442 static void update_children(struct key_vector *tn)
443 {
444 	unsigned long i;
445 
446 	/* update all of the child parent pointers */
447 	for (i = child_length(tn); i;) {
448 		struct key_vector *inode = get_child(tn, --i);
449 
450 		if (!inode)
451 			continue;
452 
453 		/* Either update the children of a tnode that
454 		 * already belongs to us or update the child
455 		 * to point to ourselves.
456 		 */
457 		if (node_parent(inode) == tn)
458 			update_children(inode);
459 		else
460 			node_set_parent(inode, tn);
461 	}
462 }
463 
464 static inline void put_child_root(struct key_vector *tp, t_key key,
465 				  struct key_vector *n)
466 {
467 	if (IS_TRIE(tp))
468 		rcu_assign_pointer(tp->tnode[0], n);
469 	else
470 		put_child(tp, get_index(key, tp), n);
471 }
472 
473 static inline void tnode_free_init(struct key_vector *tn)
474 {
475 	tn_info(tn)->rcu.next = NULL;
476 }
477 
478 static inline void tnode_free_append(struct key_vector *tn,
479 				     struct key_vector *n)
480 {
481 	tn_info(n)->rcu.next = tn_info(tn)->rcu.next;
482 	tn_info(tn)->rcu.next = &tn_info(n)->rcu;
483 }
484 
485 static void tnode_free(struct key_vector *tn)
486 {
487 	struct callback_head *head = &tn_info(tn)->rcu;
488 
489 	while (head) {
490 		head = head->next;
491 		tnode_free_size += TNODE_SIZE(1ul << tn->bits);
492 		node_free(tn);
493 
494 		tn = container_of(head, struct tnode, rcu)->kv;
495 	}
496 
497 	if (tnode_free_size >= sysctl_fib_sync_mem) {
498 		tnode_free_size = 0;
499 		synchronize_rcu();
500 	}
501 }
502 
503 static struct key_vector *replace(struct trie *t,
504 				  struct key_vector *oldtnode,
505 				  struct key_vector *tn)
506 {
507 	struct key_vector *tp = node_parent(oldtnode);
508 	unsigned long i;
509 
510 	/* setup the parent pointer out of and back into this node */
511 	NODE_INIT_PARENT(tn, tp);
512 	put_child_root(tp, tn->key, tn);
513 
514 	/* update all of the child parent pointers */
515 	update_children(tn);
516 
517 	/* all pointers should be clean so we are done */
518 	tnode_free(oldtnode);
519 
520 	/* resize children now that oldtnode is freed */
521 	for (i = child_length(tn); i;) {
522 		struct key_vector *inode = get_child(tn, --i);
523 
524 		/* resize child node */
525 		if (tnode_full(tn, inode))
526 			tn = resize(t, inode);
527 	}
528 
529 	return tp;
530 }
531 
532 static struct key_vector *inflate(struct trie *t,
533 				  struct key_vector *oldtnode)
534 {
535 	struct key_vector *tn;
536 	unsigned long i;
537 	t_key m;
538 
539 	pr_debug("In inflate\n");
540 
541 	tn = tnode_new(oldtnode->key, oldtnode->pos - 1, oldtnode->bits + 1);
542 	if (!tn)
543 		goto notnode;
544 
545 	/* prepare oldtnode to be freed */
546 	tnode_free_init(oldtnode);
547 
548 	/* Assemble all of the pointers in our cluster, in this case that
549 	 * represents all of the pointers out of our allocated nodes that
550 	 * point to existing tnodes and the links between our allocated
551 	 * nodes.
552 	 */
553 	for (i = child_length(oldtnode), m = 1u << tn->pos; i;) {
554 		struct key_vector *inode = get_child(oldtnode, --i);
555 		struct key_vector *node0, *node1;
556 		unsigned long j, k;
557 
558 		/* An empty child */
559 		if (!inode)
560 			continue;
561 
562 		/* A leaf or an internal node with skipped bits */
563 		if (!tnode_full(oldtnode, inode)) {
564 			put_child(tn, get_index(inode->key, tn), inode);
565 			continue;
566 		}
567 
568 		/* drop the node in the old tnode free list */
569 		tnode_free_append(oldtnode, inode);
570 
571 		/* An internal node with two children */
572 		if (inode->bits == 1) {
573 			put_child(tn, 2 * i + 1, get_child(inode, 1));
574 			put_child(tn, 2 * i, get_child(inode, 0));
575 			continue;
576 		}
577 
578 		/* We will replace this node 'inode' with two new
579 		 * ones, 'node0' and 'node1', each with half of the
580 		 * original children. The two new nodes will have
581 		 * a position one bit further down the key and this
582 		 * means that the "significant" part of their keys
583 		 * (see the discussion near the top of this file)
584 		 * will differ by one bit, which will be "0" in
585 		 * node0's key and "1" in node1's key. Since we are
586 		 * moving the key position by one step, the bit that
587 		 * we are moving away from - the bit at position
588 		 * (tn->pos) - is the one that will differ between
589 		 * node0 and node1. So... we synthesize that bit in the
590 		 * two new keys.
591 		 */
592 		node1 = tnode_new(inode->key | m, inode->pos, inode->bits - 1);
593 		if (!node1)
594 			goto nomem;
595 		node0 = tnode_new(inode->key, inode->pos, inode->bits - 1);
596 
597 		tnode_free_append(tn, node1);
598 		if (!node0)
599 			goto nomem;
600 		tnode_free_append(tn, node0);
601 
602 		/* populate child pointers in new nodes */
603 		for (k = child_length(inode), j = k / 2; j;) {
604 			put_child(node1, --j, get_child(inode, --k));
605 			put_child(node0, j, get_child(inode, j));
606 			put_child(node1, --j, get_child(inode, --k));
607 			put_child(node0, j, get_child(inode, j));
608 		}
609 
610 		/* link new nodes to parent */
611 		NODE_INIT_PARENT(node1, tn);
612 		NODE_INIT_PARENT(node0, tn);
613 
614 		/* link parent to nodes */
615 		put_child(tn, 2 * i + 1, node1);
616 		put_child(tn, 2 * i, node0);
617 	}
618 
619 	/* setup the parent pointers into and out of this node */
620 	return replace(t, oldtnode, tn);
621 nomem:
622 	/* all pointers should be clean so we are done */
623 	tnode_free(tn);
624 notnode:
625 	return NULL;
626 }
627 
628 static struct key_vector *halve(struct trie *t,
629 				struct key_vector *oldtnode)
630 {
631 	struct key_vector *tn;
632 	unsigned long i;
633 
634 	pr_debug("In halve\n");
635 
636 	tn = tnode_new(oldtnode->key, oldtnode->pos + 1, oldtnode->bits - 1);
637 	if (!tn)
638 		goto notnode;
639 
640 	/* prepare oldtnode to be freed */
641 	tnode_free_init(oldtnode);
642 
643 	/* Assemble all of the pointers in our cluster, in this case that
644 	 * represents all of the pointers out of our allocated nodes that
645 	 * point to existing tnodes and the links between our allocated
646 	 * nodes.
647 	 */
648 	for (i = child_length(oldtnode); i;) {
649 		struct key_vector *node1 = get_child(oldtnode, --i);
650 		struct key_vector *node0 = get_child(oldtnode, --i);
651 		struct key_vector *inode;
652 
653 		/* At least one of the children is empty */
654 		if (!node1 || !node0) {
655 			put_child(tn, i / 2, node1 ? : node0);
656 			continue;
657 		}
658 
659 		/* Two nonempty children */
660 		inode = tnode_new(node0->key, oldtnode->pos, 1);
661 		if (!inode)
662 			goto nomem;
663 		tnode_free_append(tn, inode);
664 
665 		/* initialize pointers out of node */
666 		put_child(inode, 1, node1);
667 		put_child(inode, 0, node0);
668 		NODE_INIT_PARENT(inode, tn);
669 
670 		/* link parent to node */
671 		put_child(tn, i / 2, inode);
672 	}
673 
674 	/* setup the parent pointers into and out of this node */
675 	return replace(t, oldtnode, tn);
676 nomem:
677 	/* all pointers should be clean so we are done */
678 	tnode_free(tn);
679 notnode:
680 	return NULL;
681 }
682 
683 static struct key_vector *collapse(struct trie *t,
684 				   struct key_vector *oldtnode)
685 {
686 	struct key_vector *n, *tp;
687 	unsigned long i;
688 
689 	/* scan the tnode looking for that one child that might still exist */
690 	for (n = NULL, i = child_length(oldtnode); !n && i;)
691 		n = get_child(oldtnode, --i);
692 
693 	/* compress one level */
694 	tp = node_parent(oldtnode);
695 	put_child_root(tp, oldtnode->key, n);
696 	node_set_parent(n, tp);
697 
698 	/* drop dead node */
699 	node_free(oldtnode);
700 
701 	return tp;
702 }
703 
704 static unsigned char update_suffix(struct key_vector *tn)
705 {
706 	unsigned char slen = tn->pos;
707 	unsigned long stride, i;
708 	unsigned char slen_max;
709 
710 	/* only vector 0 can have a suffix length greater than or equal to
711 	 * tn->pos + tn->bits, the second highest node will have a suffix
712 	 * length at most of tn->pos + tn->bits - 1
713 	 */
714 	slen_max = min_t(unsigned char, tn->pos + tn->bits - 1, tn->slen);
715 
716 	/* search though the list of children looking for nodes that might
717 	 * have a suffix greater than the one we currently have.  This is
718 	 * why we start with a stride of 2 since a stride of 1 would
719 	 * represent the nodes with suffix length equal to tn->pos
720 	 */
721 	for (i = 0, stride = 0x2ul ; i < child_length(tn); i += stride) {
722 		struct key_vector *n = get_child(tn, i);
723 
724 		if (!n || (n->slen <= slen))
725 			continue;
726 
727 		/* update stride and slen based on new value */
728 		stride <<= (n->slen - slen);
729 		slen = n->slen;
730 		i &= ~(stride - 1);
731 
732 		/* stop searching if we have hit the maximum possible value */
733 		if (slen >= slen_max)
734 			break;
735 	}
736 
737 	tn->slen = slen;
738 
739 	return slen;
740 }
741 
742 /* From "Implementing a dynamic compressed trie" by Stefan Nilsson of
743  * the Helsinki University of Technology and Matti Tikkanen of Nokia
744  * Telecommunications, page 6:
745  * "A node is doubled if the ratio of non-empty children to all
746  * children in the *doubled* node is at least 'high'."
747  *
748  * 'high' in this instance is the variable 'inflate_threshold'. It
749  * is expressed as a percentage, so we multiply it with
750  * child_length() and instead of multiplying by 2 (since the
751  * child array will be doubled by inflate()) and multiplying
752  * the left-hand side by 100 (to handle the percentage thing) we
753  * multiply the left-hand side by 50.
754  *
755  * The left-hand side may look a bit weird: child_length(tn)
756  * - tn->empty_children is of course the number of non-null children
757  * in the current node. tn->full_children is the number of "full"
758  * children, that is non-null tnodes with a skip value of 0.
759  * All of those will be doubled in the resulting inflated tnode, so
760  * we just count them one extra time here.
761  *
762  * A clearer way to write this would be:
763  *
764  * to_be_doubled = tn->full_children;
765  * not_to_be_doubled = child_length(tn) - tn->empty_children -
766  *     tn->full_children;
767  *
768  * new_child_length = child_length(tn) * 2;
769  *
770  * new_fill_factor = 100 * (not_to_be_doubled + 2*to_be_doubled) /
771  *      new_child_length;
772  * if (new_fill_factor >= inflate_threshold)
773  *
774  * ...and so on, tho it would mess up the while () loop.
775  *
776  * anyway,
777  * 100 * (not_to_be_doubled + 2*to_be_doubled) / new_child_length >=
778  *      inflate_threshold
779  *
780  * avoid a division:
781  * 100 * (not_to_be_doubled + 2*to_be_doubled) >=
782  *      inflate_threshold * new_child_length
783  *
784  * expand not_to_be_doubled and to_be_doubled, and shorten:
785  * 100 * (child_length(tn) - tn->empty_children +
786  *    tn->full_children) >= inflate_threshold * new_child_length
787  *
788  * expand new_child_length:
789  * 100 * (child_length(tn) - tn->empty_children +
790  *    tn->full_children) >=
791  *      inflate_threshold * child_length(tn) * 2
792  *
793  * shorten again:
794  * 50 * (tn->full_children + child_length(tn) -
795  *    tn->empty_children) >= inflate_threshold *
796  *    child_length(tn)
797  *
798  */
799 static inline bool should_inflate(struct key_vector *tp, struct key_vector *tn)
800 {
801 	unsigned long used = child_length(tn);
802 	unsigned long threshold = used;
803 
804 	/* Keep root node larger */
805 	threshold *= IS_TRIE(tp) ? inflate_threshold_root : inflate_threshold;
806 	used -= tn_info(tn)->empty_children;
807 	used += tn_info(tn)->full_children;
808 
809 	/* if bits == KEYLENGTH then pos = 0, and will fail below */
810 
811 	return (used > 1) && tn->pos && ((50 * used) >= threshold);
812 }
813 
814 static inline bool should_halve(struct key_vector *tp, struct key_vector *tn)
815 {
816 	unsigned long used = child_length(tn);
817 	unsigned long threshold = used;
818 
819 	/* Keep root node larger */
820 	threshold *= IS_TRIE(tp) ? halve_threshold_root : halve_threshold;
821 	used -= tn_info(tn)->empty_children;
822 
823 	/* if bits == KEYLENGTH then used = 100% on wrap, and will fail below */
824 
825 	return (used > 1) && (tn->bits > 1) && ((100 * used) < threshold);
826 }
827 
828 static inline bool should_collapse(struct key_vector *tn)
829 {
830 	unsigned long used = child_length(tn);
831 
832 	used -= tn_info(tn)->empty_children;
833 
834 	/* account for bits == KEYLENGTH case */
835 	if ((tn->bits == KEYLENGTH) && tn_info(tn)->full_children)
836 		used -= KEY_MAX;
837 
838 	/* One child or none, time to drop us from the trie */
839 	return used < 2;
840 }
841 
842 #define MAX_WORK 10
843 static struct key_vector *resize(struct trie *t, struct key_vector *tn)
844 {
845 #ifdef CONFIG_IP_FIB_TRIE_STATS
846 	struct trie_use_stats __percpu *stats = t->stats;
847 #endif
848 	struct key_vector *tp = node_parent(tn);
849 	unsigned long cindex = get_index(tn->key, tp);
850 	int max_work = MAX_WORK;
851 
852 	pr_debug("In tnode_resize %p inflate_threshold=%d threshold=%d\n",
853 		 tn, inflate_threshold, halve_threshold);
854 
855 	/* track the tnode via the pointer from the parent instead of
856 	 * doing it ourselves.  This way we can let RCU fully do its
857 	 * thing without us interfering
858 	 */
859 	BUG_ON(tn != get_child(tp, cindex));
860 
861 	/* Double as long as the resulting node has a number of
862 	 * nonempty nodes that are above the threshold.
863 	 */
864 	while (should_inflate(tp, tn) && max_work) {
865 		tp = inflate(t, tn);
866 		if (!tp) {
867 #ifdef CONFIG_IP_FIB_TRIE_STATS
868 			this_cpu_inc(stats->resize_node_skipped);
869 #endif
870 			break;
871 		}
872 
873 		max_work--;
874 		tn = get_child(tp, cindex);
875 	}
876 
877 	/* update parent in case inflate failed */
878 	tp = node_parent(tn);
879 
880 	/* Return if at least one inflate is run */
881 	if (max_work != MAX_WORK)
882 		return tp;
883 
884 	/* Halve as long as the number of empty children in this
885 	 * node is above threshold.
886 	 */
887 	while (should_halve(tp, tn) && max_work) {
888 		tp = halve(t, tn);
889 		if (!tp) {
890 #ifdef CONFIG_IP_FIB_TRIE_STATS
891 			this_cpu_inc(stats->resize_node_skipped);
892 #endif
893 			break;
894 		}
895 
896 		max_work--;
897 		tn = get_child(tp, cindex);
898 	}
899 
900 	/* Only one child remains */
901 	if (should_collapse(tn))
902 		return collapse(t, tn);
903 
904 	/* update parent in case halve failed */
905 	return node_parent(tn);
906 }
907 
908 static void node_pull_suffix(struct key_vector *tn, unsigned char slen)
909 {
910 	unsigned char node_slen = tn->slen;
911 
912 	while ((node_slen > tn->pos) && (node_slen > slen)) {
913 		slen = update_suffix(tn);
914 		if (node_slen == slen)
915 			break;
916 
917 		tn = node_parent(tn);
918 		node_slen = tn->slen;
919 	}
920 }
921 
922 static void node_push_suffix(struct key_vector *tn, unsigned char slen)
923 {
924 	while (tn->slen < slen) {
925 		tn->slen = slen;
926 		tn = node_parent(tn);
927 	}
928 }
929 
930 /* rcu_read_lock needs to be hold by caller from readside */
931 static struct key_vector *fib_find_node(struct trie *t,
932 					struct key_vector **tp, u32 key)
933 {
934 	struct key_vector *pn, *n = t->kv;
935 	unsigned long index = 0;
936 
937 	do {
938 		pn = n;
939 		n = get_child_rcu(n, index);
940 
941 		if (!n)
942 			break;
943 
944 		index = get_cindex(key, n);
945 
946 		/* This bit of code is a bit tricky but it combines multiple
947 		 * checks into a single check.  The prefix consists of the
948 		 * prefix plus zeros for the bits in the cindex. The index
949 		 * is the difference between the key and this value.  From
950 		 * this we can actually derive several pieces of data.
951 		 *   if (index >= (1ul << bits))
952 		 *     we have a mismatch in skip bits and failed
953 		 *   else
954 		 *     we know the value is cindex
955 		 *
956 		 * This check is safe even if bits == KEYLENGTH due to the
957 		 * fact that we can only allocate a node with 32 bits if a
958 		 * long is greater than 32 bits.
959 		 */
960 		if (index >= (1ul << n->bits)) {
961 			n = NULL;
962 			break;
963 		}
964 
965 		/* keep searching until we find a perfect match leaf or NULL */
966 	} while (IS_TNODE(n));
967 
968 	*tp = pn;
969 
970 	return n;
971 }
972 
973 /* Return the first fib alias matching TOS with
974  * priority less than or equal to PRIO.
975  */
976 static struct fib_alias *fib_find_alias(struct hlist_head *fah, u8 slen,
977 					u8 tos, u32 prio, u32 tb_id)
978 {
979 	struct fib_alias *fa;
980 
981 	if (!fah)
982 		return NULL;
983 
984 	hlist_for_each_entry(fa, fah, fa_list) {
985 		if (fa->fa_slen < slen)
986 			continue;
987 		if (fa->fa_slen != slen)
988 			break;
989 		if (fa->tb_id > tb_id)
990 			continue;
991 		if (fa->tb_id != tb_id)
992 			break;
993 		if (fa->fa_tos > tos)
994 			continue;
995 		if (fa->fa_info->fib_priority >= prio || fa->fa_tos < tos)
996 			return fa;
997 	}
998 
999 	return NULL;
1000 }
1001 
1002 static void trie_rebalance(struct trie *t, struct key_vector *tn)
1003 {
1004 	while (!IS_TRIE(tn))
1005 		tn = resize(t, tn);
1006 }
1007 
1008 static int fib_insert_node(struct trie *t, struct key_vector *tp,
1009 			   struct fib_alias *new, t_key key)
1010 {
1011 	struct key_vector *n, *l;
1012 
1013 	l = leaf_new(key, new);
1014 	if (!l)
1015 		goto noleaf;
1016 
1017 	/* retrieve child from parent node */
1018 	n = get_child(tp, get_index(key, tp));
1019 
1020 	/* Case 2: n is a LEAF or a TNODE and the key doesn't match.
1021 	 *
1022 	 *  Add a new tnode here
1023 	 *  first tnode need some special handling
1024 	 *  leaves us in position for handling as case 3
1025 	 */
1026 	if (n) {
1027 		struct key_vector *tn;
1028 
1029 		tn = tnode_new(key, __fls(key ^ n->key), 1);
1030 		if (!tn)
1031 			goto notnode;
1032 
1033 		/* initialize routes out of node */
1034 		NODE_INIT_PARENT(tn, tp);
1035 		put_child(tn, get_index(key, tn) ^ 1, n);
1036 
1037 		/* start adding routes into the node */
1038 		put_child_root(tp, key, tn);
1039 		node_set_parent(n, tn);
1040 
1041 		/* parent now has a NULL spot where the leaf can go */
1042 		tp = tn;
1043 	}
1044 
1045 	/* Case 3: n is NULL, and will just insert a new leaf */
1046 	node_push_suffix(tp, new->fa_slen);
1047 	NODE_INIT_PARENT(l, tp);
1048 	put_child_root(tp, key, l);
1049 	trie_rebalance(t, tp);
1050 
1051 	return 0;
1052 notnode:
1053 	node_free(l);
1054 noleaf:
1055 	return -ENOMEM;
1056 }
1057 
1058 /* fib notifier for ADD is sent before calling fib_insert_alias with
1059  * the expectation that the only possible failure ENOMEM
1060  */
1061 static int fib_insert_alias(struct trie *t, struct key_vector *tp,
1062 			    struct key_vector *l, struct fib_alias *new,
1063 			    struct fib_alias *fa, t_key key)
1064 {
1065 	if (!l)
1066 		return fib_insert_node(t, tp, new, key);
1067 
1068 	if (fa) {
1069 		hlist_add_before_rcu(&new->fa_list, &fa->fa_list);
1070 	} else {
1071 		struct fib_alias *last;
1072 
1073 		hlist_for_each_entry(last, &l->leaf, fa_list) {
1074 			if (new->fa_slen < last->fa_slen)
1075 				break;
1076 			if ((new->fa_slen == last->fa_slen) &&
1077 			    (new->tb_id > last->tb_id))
1078 				break;
1079 			fa = last;
1080 		}
1081 
1082 		if (fa)
1083 			hlist_add_behind_rcu(&new->fa_list, &fa->fa_list);
1084 		else
1085 			hlist_add_head_rcu(&new->fa_list, &l->leaf);
1086 	}
1087 
1088 	/* if we added to the tail node then we need to update slen */
1089 	if (l->slen < new->fa_slen) {
1090 		l->slen = new->fa_slen;
1091 		node_push_suffix(tp, new->fa_slen);
1092 	}
1093 
1094 	return 0;
1095 }
1096 
1097 static bool fib_valid_key_len(u32 key, u8 plen, struct netlink_ext_ack *extack)
1098 {
1099 	if (plen > KEYLENGTH) {
1100 		NL_SET_ERR_MSG(extack, "Invalid prefix length");
1101 		return false;
1102 	}
1103 
1104 	if ((plen < KEYLENGTH) && (key << plen)) {
1105 		NL_SET_ERR_MSG(extack,
1106 			       "Invalid prefix for given prefix length");
1107 		return false;
1108 	}
1109 
1110 	return true;
1111 }
1112 
1113 /* Caller must hold RTNL. */
1114 int fib_table_insert(struct net *net, struct fib_table *tb,
1115 		     struct fib_config *cfg, struct netlink_ext_ack *extack)
1116 {
1117 	enum fib_event_type event = FIB_EVENT_ENTRY_ADD;
1118 	struct trie *t = (struct trie *)tb->tb_data;
1119 	struct fib_alias *fa, *new_fa;
1120 	struct key_vector *l, *tp;
1121 	u16 nlflags = NLM_F_EXCL;
1122 	struct fib_info *fi;
1123 	u8 plen = cfg->fc_dst_len;
1124 	u8 slen = KEYLENGTH - plen;
1125 	u8 tos = cfg->fc_tos;
1126 	u32 key;
1127 	int err;
1128 
1129 	key = ntohl(cfg->fc_dst);
1130 
1131 	if (!fib_valid_key_len(key, plen, extack))
1132 		return -EINVAL;
1133 
1134 	pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen);
1135 
1136 	fi = fib_create_info(cfg, extack);
1137 	if (IS_ERR(fi)) {
1138 		err = PTR_ERR(fi);
1139 		goto err;
1140 	}
1141 
1142 	l = fib_find_node(t, &tp, key);
1143 	fa = l ? fib_find_alias(&l->leaf, slen, tos, fi->fib_priority,
1144 				tb->tb_id) : NULL;
1145 
1146 	/* Now fa, if non-NULL, points to the first fib alias
1147 	 * with the same keys [prefix,tos,priority], if such key already
1148 	 * exists or to the node before which we will insert new one.
1149 	 *
1150 	 * If fa is NULL, we will need to allocate a new one and
1151 	 * insert to the tail of the section matching the suffix length
1152 	 * of the new alias.
1153 	 */
1154 
1155 	if (fa && fa->fa_tos == tos &&
1156 	    fa->fa_info->fib_priority == fi->fib_priority) {
1157 		struct fib_alias *fa_first, *fa_match;
1158 
1159 		err = -EEXIST;
1160 		if (cfg->fc_nlflags & NLM_F_EXCL)
1161 			goto out;
1162 
1163 		nlflags &= ~NLM_F_EXCL;
1164 
1165 		/* We have 2 goals:
1166 		 * 1. Find exact match for type, scope, fib_info to avoid
1167 		 * duplicate routes
1168 		 * 2. Find next 'fa' (or head), NLM_F_APPEND inserts before it
1169 		 */
1170 		fa_match = NULL;
1171 		fa_first = fa;
1172 		hlist_for_each_entry_from(fa, fa_list) {
1173 			if ((fa->fa_slen != slen) ||
1174 			    (fa->tb_id != tb->tb_id) ||
1175 			    (fa->fa_tos != tos))
1176 				break;
1177 			if (fa->fa_info->fib_priority != fi->fib_priority)
1178 				break;
1179 			if (fa->fa_type == cfg->fc_type &&
1180 			    fa->fa_info == fi) {
1181 				fa_match = fa;
1182 				break;
1183 			}
1184 		}
1185 
1186 		if (cfg->fc_nlflags & NLM_F_REPLACE) {
1187 			struct fib_info *fi_drop;
1188 			u8 state;
1189 
1190 			nlflags |= NLM_F_REPLACE;
1191 			fa = fa_first;
1192 			if (fa_match) {
1193 				if (fa == fa_match)
1194 					err = 0;
1195 				goto out;
1196 			}
1197 			err = -ENOBUFS;
1198 			new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1199 			if (!new_fa)
1200 				goto out;
1201 
1202 			fi_drop = fa->fa_info;
1203 			new_fa->fa_tos = fa->fa_tos;
1204 			new_fa->fa_info = fi;
1205 			new_fa->fa_type = cfg->fc_type;
1206 			state = fa->fa_state;
1207 			new_fa->fa_state = state & ~FA_S_ACCESSED;
1208 			new_fa->fa_slen = fa->fa_slen;
1209 			new_fa->tb_id = tb->tb_id;
1210 			new_fa->fa_default = -1;
1211 
1212 			err = call_fib_entry_notifiers(net,
1213 						       FIB_EVENT_ENTRY_REPLACE,
1214 						       key, plen, new_fa,
1215 						       extack);
1216 			if (err)
1217 				goto out_free_new_fa;
1218 
1219 			rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen,
1220 				  tb->tb_id, &cfg->fc_nlinfo, nlflags);
1221 
1222 			hlist_replace_rcu(&fa->fa_list, &new_fa->fa_list);
1223 
1224 			alias_free_mem_rcu(fa);
1225 
1226 			fib_release_info(fi_drop);
1227 			if (state & FA_S_ACCESSED)
1228 				rt_cache_flush(cfg->fc_nlinfo.nl_net);
1229 
1230 			goto succeeded;
1231 		}
1232 		/* Error if we find a perfect match which
1233 		 * uses the same scope, type, and nexthop
1234 		 * information.
1235 		 */
1236 		if (fa_match)
1237 			goto out;
1238 
1239 		if (cfg->fc_nlflags & NLM_F_APPEND) {
1240 			event = FIB_EVENT_ENTRY_APPEND;
1241 			nlflags |= NLM_F_APPEND;
1242 		} else {
1243 			fa = fa_first;
1244 		}
1245 	}
1246 	err = -ENOENT;
1247 	if (!(cfg->fc_nlflags & NLM_F_CREATE))
1248 		goto out;
1249 
1250 	nlflags |= NLM_F_CREATE;
1251 	err = -ENOBUFS;
1252 	new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1253 	if (!new_fa)
1254 		goto out;
1255 
1256 	new_fa->fa_info = fi;
1257 	new_fa->fa_tos = tos;
1258 	new_fa->fa_type = cfg->fc_type;
1259 	new_fa->fa_state = 0;
1260 	new_fa->fa_slen = slen;
1261 	new_fa->tb_id = tb->tb_id;
1262 	new_fa->fa_default = -1;
1263 
1264 	err = call_fib_entry_notifiers(net, event, key, plen, new_fa, extack);
1265 	if (err)
1266 		goto out_free_new_fa;
1267 
1268 	/* Insert new entry to the list. */
1269 	err = fib_insert_alias(t, tp, l, new_fa, fa, key);
1270 	if (err)
1271 		goto out_fib_notif;
1272 
1273 	if (!plen)
1274 		tb->tb_num_default++;
1275 
1276 	rt_cache_flush(cfg->fc_nlinfo.nl_net);
1277 	rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, new_fa->tb_id,
1278 		  &cfg->fc_nlinfo, nlflags);
1279 succeeded:
1280 	return 0;
1281 
1282 out_fib_notif:
1283 	/* notifier was sent that entry would be added to trie, but
1284 	 * the add failed and need to recover. Only failure for
1285 	 * fib_insert_alias is ENOMEM.
1286 	 */
1287 	NL_SET_ERR_MSG(extack, "Failed to insert route into trie");
1288 	call_fib_entry_notifiers(net, FIB_EVENT_ENTRY_DEL, key,
1289 				 plen, new_fa, NULL);
1290 out_free_new_fa:
1291 	kmem_cache_free(fn_alias_kmem, new_fa);
1292 out:
1293 	fib_release_info(fi);
1294 err:
1295 	return err;
1296 }
1297 
1298 static inline t_key prefix_mismatch(t_key key, struct key_vector *n)
1299 {
1300 	t_key prefix = n->key;
1301 
1302 	return (key ^ prefix) & (prefix | -prefix);
1303 }
1304 
1305 /* should be called with rcu_read_lock */
1306 int fib_table_lookup(struct fib_table *tb, const struct flowi4 *flp,
1307 		     struct fib_result *res, int fib_flags)
1308 {
1309 	struct trie *t = (struct trie *) tb->tb_data;
1310 #ifdef CONFIG_IP_FIB_TRIE_STATS
1311 	struct trie_use_stats __percpu *stats = t->stats;
1312 #endif
1313 	const t_key key = ntohl(flp->daddr);
1314 	struct key_vector *n, *pn;
1315 	struct fib_alias *fa;
1316 	unsigned long index;
1317 	t_key cindex;
1318 
1319 	pn = t->kv;
1320 	cindex = 0;
1321 
1322 	n = get_child_rcu(pn, cindex);
1323 	if (!n) {
1324 		trace_fib_table_lookup(tb->tb_id, flp, NULL, -EAGAIN);
1325 		return -EAGAIN;
1326 	}
1327 
1328 #ifdef CONFIG_IP_FIB_TRIE_STATS
1329 	this_cpu_inc(stats->gets);
1330 #endif
1331 
1332 	/* Step 1: Travel to the longest prefix match in the trie */
1333 	for (;;) {
1334 		index = get_cindex(key, n);
1335 
1336 		/* This bit of code is a bit tricky but it combines multiple
1337 		 * checks into a single check.  The prefix consists of the
1338 		 * prefix plus zeros for the "bits" in the prefix. The index
1339 		 * is the difference between the key and this value.  From
1340 		 * this we can actually derive several pieces of data.
1341 		 *   if (index >= (1ul << bits))
1342 		 *     we have a mismatch in skip bits and failed
1343 		 *   else
1344 		 *     we know the value is cindex
1345 		 *
1346 		 * This check is safe even if bits == KEYLENGTH due to the
1347 		 * fact that we can only allocate a node with 32 bits if a
1348 		 * long is greater than 32 bits.
1349 		 */
1350 		if (index >= (1ul << n->bits))
1351 			break;
1352 
1353 		/* we have found a leaf. Prefixes have already been compared */
1354 		if (IS_LEAF(n))
1355 			goto found;
1356 
1357 		/* only record pn and cindex if we are going to be chopping
1358 		 * bits later.  Otherwise we are just wasting cycles.
1359 		 */
1360 		if (n->slen > n->pos) {
1361 			pn = n;
1362 			cindex = index;
1363 		}
1364 
1365 		n = get_child_rcu(n, index);
1366 		if (unlikely(!n))
1367 			goto backtrace;
1368 	}
1369 
1370 	/* Step 2: Sort out leaves and begin backtracing for longest prefix */
1371 	for (;;) {
1372 		/* record the pointer where our next node pointer is stored */
1373 		struct key_vector __rcu **cptr = n->tnode;
1374 
1375 		/* This test verifies that none of the bits that differ
1376 		 * between the key and the prefix exist in the region of
1377 		 * the lsb and higher in the prefix.
1378 		 */
1379 		if (unlikely(prefix_mismatch(key, n)) || (n->slen == n->pos))
1380 			goto backtrace;
1381 
1382 		/* exit out and process leaf */
1383 		if (unlikely(IS_LEAF(n)))
1384 			break;
1385 
1386 		/* Don't bother recording parent info.  Since we are in
1387 		 * prefix match mode we will have to come back to wherever
1388 		 * we started this traversal anyway
1389 		 */
1390 
1391 		while ((n = rcu_dereference(*cptr)) == NULL) {
1392 backtrace:
1393 #ifdef CONFIG_IP_FIB_TRIE_STATS
1394 			if (!n)
1395 				this_cpu_inc(stats->null_node_hit);
1396 #endif
1397 			/* If we are at cindex 0 there are no more bits for
1398 			 * us to strip at this level so we must ascend back
1399 			 * up one level to see if there are any more bits to
1400 			 * be stripped there.
1401 			 */
1402 			while (!cindex) {
1403 				t_key pkey = pn->key;
1404 
1405 				/* If we don't have a parent then there is
1406 				 * nothing for us to do as we do not have any
1407 				 * further nodes to parse.
1408 				 */
1409 				if (IS_TRIE(pn)) {
1410 					trace_fib_table_lookup(tb->tb_id, flp,
1411 							       NULL, -EAGAIN);
1412 					return -EAGAIN;
1413 				}
1414 #ifdef CONFIG_IP_FIB_TRIE_STATS
1415 				this_cpu_inc(stats->backtrack);
1416 #endif
1417 				/* Get Child's index */
1418 				pn = node_parent_rcu(pn);
1419 				cindex = get_index(pkey, pn);
1420 			}
1421 
1422 			/* strip the least significant bit from the cindex */
1423 			cindex &= cindex - 1;
1424 
1425 			/* grab pointer for next child node */
1426 			cptr = &pn->tnode[cindex];
1427 		}
1428 	}
1429 
1430 found:
1431 	/* this line carries forward the xor from earlier in the function */
1432 	index = key ^ n->key;
1433 
1434 	/* Step 3: Process the leaf, if that fails fall back to backtracing */
1435 	hlist_for_each_entry_rcu(fa, &n->leaf, fa_list) {
1436 		struct fib_info *fi = fa->fa_info;
1437 		int nhsel, err;
1438 
1439 		if ((BITS_PER_LONG > KEYLENGTH) || (fa->fa_slen < KEYLENGTH)) {
1440 			if (index >= (1ul << fa->fa_slen))
1441 				continue;
1442 		}
1443 		if (fa->fa_tos && fa->fa_tos != flp->flowi4_tos)
1444 			continue;
1445 		if (fi->fib_dead)
1446 			continue;
1447 		if (fa->fa_info->fib_scope < flp->flowi4_scope)
1448 			continue;
1449 		fib_alias_accessed(fa);
1450 		err = fib_props[fa->fa_type].error;
1451 		if (unlikely(err < 0)) {
1452 #ifdef CONFIG_IP_FIB_TRIE_STATS
1453 			this_cpu_inc(stats->semantic_match_passed);
1454 #endif
1455 			trace_fib_table_lookup(tb->tb_id, flp, NULL, err);
1456 			return err;
1457 		}
1458 		if (fi->fib_flags & RTNH_F_DEAD)
1459 			continue;
1460 		for (nhsel = 0; nhsel < fi->fib_nhs; nhsel++) {
1461 			struct fib_nh_common *nhc = fib_info_nhc(fi, nhsel);
1462 
1463 			if (nhc->nhc_flags & RTNH_F_DEAD)
1464 				continue;
1465 			if (ip_ignore_linkdown(nhc->nhc_dev) &&
1466 			    nhc->nhc_flags & RTNH_F_LINKDOWN &&
1467 			    !(fib_flags & FIB_LOOKUP_IGNORE_LINKSTATE))
1468 				continue;
1469 			if (!(flp->flowi4_flags & FLOWI_FLAG_SKIP_NH_OIF)) {
1470 				if (flp->flowi4_oif &&
1471 				    flp->flowi4_oif != nhc->nhc_oif)
1472 					continue;
1473 			}
1474 
1475 			if (!(fib_flags & FIB_LOOKUP_NOREF))
1476 				refcount_inc(&fi->fib_clntref);
1477 
1478 			res->prefix = htonl(n->key);
1479 			res->prefixlen = KEYLENGTH - fa->fa_slen;
1480 			res->nh_sel = nhsel;
1481 			res->nhc = nhc;
1482 			res->type = fa->fa_type;
1483 			res->scope = fi->fib_scope;
1484 			res->fi = fi;
1485 			res->table = tb;
1486 			res->fa_head = &n->leaf;
1487 #ifdef CONFIG_IP_FIB_TRIE_STATS
1488 			this_cpu_inc(stats->semantic_match_passed);
1489 #endif
1490 			trace_fib_table_lookup(tb->tb_id, flp, nhc, err);
1491 
1492 			return err;
1493 		}
1494 	}
1495 #ifdef CONFIG_IP_FIB_TRIE_STATS
1496 	this_cpu_inc(stats->semantic_match_miss);
1497 #endif
1498 	goto backtrace;
1499 }
1500 EXPORT_SYMBOL_GPL(fib_table_lookup);
1501 
1502 static void fib_remove_alias(struct trie *t, struct key_vector *tp,
1503 			     struct key_vector *l, struct fib_alias *old)
1504 {
1505 	/* record the location of the previous list_info entry */
1506 	struct hlist_node **pprev = old->fa_list.pprev;
1507 	struct fib_alias *fa = hlist_entry(pprev, typeof(*fa), fa_list.next);
1508 
1509 	/* remove the fib_alias from the list */
1510 	hlist_del_rcu(&old->fa_list);
1511 
1512 	/* if we emptied the list this leaf will be freed and we can sort
1513 	 * out parent suffix lengths as a part of trie_rebalance
1514 	 */
1515 	if (hlist_empty(&l->leaf)) {
1516 		if (tp->slen == l->slen)
1517 			node_pull_suffix(tp, tp->pos);
1518 		put_child_root(tp, l->key, NULL);
1519 		node_free(l);
1520 		trie_rebalance(t, tp);
1521 		return;
1522 	}
1523 
1524 	/* only access fa if it is pointing at the last valid hlist_node */
1525 	if (*pprev)
1526 		return;
1527 
1528 	/* update the trie with the latest suffix length */
1529 	l->slen = fa->fa_slen;
1530 	node_pull_suffix(tp, fa->fa_slen);
1531 }
1532 
1533 /* Caller must hold RTNL. */
1534 int fib_table_delete(struct net *net, struct fib_table *tb,
1535 		     struct fib_config *cfg, struct netlink_ext_ack *extack)
1536 {
1537 	struct trie *t = (struct trie *) tb->tb_data;
1538 	struct fib_alias *fa, *fa_to_delete;
1539 	struct key_vector *l, *tp;
1540 	u8 plen = cfg->fc_dst_len;
1541 	u8 slen = KEYLENGTH - plen;
1542 	u8 tos = cfg->fc_tos;
1543 	u32 key;
1544 
1545 	key = ntohl(cfg->fc_dst);
1546 
1547 	if (!fib_valid_key_len(key, plen, extack))
1548 		return -EINVAL;
1549 
1550 	l = fib_find_node(t, &tp, key);
1551 	if (!l)
1552 		return -ESRCH;
1553 
1554 	fa = fib_find_alias(&l->leaf, slen, tos, 0, tb->tb_id);
1555 	if (!fa)
1556 		return -ESRCH;
1557 
1558 	pr_debug("Deleting %08x/%d tos=%d t=%p\n", key, plen, tos, t);
1559 
1560 	fa_to_delete = NULL;
1561 	hlist_for_each_entry_from(fa, fa_list) {
1562 		struct fib_info *fi = fa->fa_info;
1563 
1564 		if ((fa->fa_slen != slen) ||
1565 		    (fa->tb_id != tb->tb_id) ||
1566 		    (fa->fa_tos != tos))
1567 			break;
1568 
1569 		if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) &&
1570 		    (cfg->fc_scope == RT_SCOPE_NOWHERE ||
1571 		     fa->fa_info->fib_scope == cfg->fc_scope) &&
1572 		    (!cfg->fc_prefsrc ||
1573 		     fi->fib_prefsrc == cfg->fc_prefsrc) &&
1574 		    (!cfg->fc_protocol ||
1575 		     fi->fib_protocol == cfg->fc_protocol) &&
1576 		    fib_nh_match(cfg, fi, extack) == 0 &&
1577 		    fib_metrics_match(cfg, fi)) {
1578 			fa_to_delete = fa;
1579 			break;
1580 		}
1581 	}
1582 
1583 	if (!fa_to_delete)
1584 		return -ESRCH;
1585 
1586 	call_fib_entry_notifiers(net, FIB_EVENT_ENTRY_DEL, key, plen,
1587 				 fa_to_delete, extack);
1588 	rtmsg_fib(RTM_DELROUTE, htonl(key), fa_to_delete, plen, tb->tb_id,
1589 		  &cfg->fc_nlinfo, 0);
1590 
1591 	if (!plen)
1592 		tb->tb_num_default--;
1593 
1594 	fib_remove_alias(t, tp, l, fa_to_delete);
1595 
1596 	if (fa_to_delete->fa_state & FA_S_ACCESSED)
1597 		rt_cache_flush(cfg->fc_nlinfo.nl_net);
1598 
1599 	fib_release_info(fa_to_delete->fa_info);
1600 	alias_free_mem_rcu(fa_to_delete);
1601 	return 0;
1602 }
1603 
1604 /* Scan for the next leaf starting at the provided key value */
1605 static struct key_vector *leaf_walk_rcu(struct key_vector **tn, t_key key)
1606 {
1607 	struct key_vector *pn, *n = *tn;
1608 	unsigned long cindex;
1609 
1610 	/* this loop is meant to try and find the key in the trie */
1611 	do {
1612 		/* record parent and next child index */
1613 		pn = n;
1614 		cindex = (key > pn->key) ? get_index(key, pn) : 0;
1615 
1616 		if (cindex >> pn->bits)
1617 			break;
1618 
1619 		/* descend into the next child */
1620 		n = get_child_rcu(pn, cindex++);
1621 		if (!n)
1622 			break;
1623 
1624 		/* guarantee forward progress on the keys */
1625 		if (IS_LEAF(n) && (n->key >= key))
1626 			goto found;
1627 	} while (IS_TNODE(n));
1628 
1629 	/* this loop will search for the next leaf with a greater key */
1630 	while (!IS_TRIE(pn)) {
1631 		/* if we exhausted the parent node we will need to climb */
1632 		if (cindex >= (1ul << pn->bits)) {
1633 			t_key pkey = pn->key;
1634 
1635 			pn = node_parent_rcu(pn);
1636 			cindex = get_index(pkey, pn) + 1;
1637 			continue;
1638 		}
1639 
1640 		/* grab the next available node */
1641 		n = get_child_rcu(pn, cindex++);
1642 		if (!n)
1643 			continue;
1644 
1645 		/* no need to compare keys since we bumped the index */
1646 		if (IS_LEAF(n))
1647 			goto found;
1648 
1649 		/* Rescan start scanning in new node */
1650 		pn = n;
1651 		cindex = 0;
1652 	}
1653 
1654 	*tn = pn;
1655 	return NULL; /* Root of trie */
1656 found:
1657 	/* if we are at the limit for keys just return NULL for the tnode */
1658 	*tn = pn;
1659 	return n;
1660 }
1661 
1662 static void fib_trie_free(struct fib_table *tb)
1663 {
1664 	struct trie *t = (struct trie *)tb->tb_data;
1665 	struct key_vector *pn = t->kv;
1666 	unsigned long cindex = 1;
1667 	struct hlist_node *tmp;
1668 	struct fib_alias *fa;
1669 
1670 	/* walk trie in reverse order and free everything */
1671 	for (;;) {
1672 		struct key_vector *n;
1673 
1674 		if (!(cindex--)) {
1675 			t_key pkey = pn->key;
1676 
1677 			if (IS_TRIE(pn))
1678 				break;
1679 
1680 			n = pn;
1681 			pn = node_parent(pn);
1682 
1683 			/* drop emptied tnode */
1684 			put_child_root(pn, n->key, NULL);
1685 			node_free(n);
1686 
1687 			cindex = get_index(pkey, pn);
1688 
1689 			continue;
1690 		}
1691 
1692 		/* grab the next available node */
1693 		n = get_child(pn, cindex);
1694 		if (!n)
1695 			continue;
1696 
1697 		if (IS_TNODE(n)) {
1698 			/* record pn and cindex for leaf walking */
1699 			pn = n;
1700 			cindex = 1ul << n->bits;
1701 
1702 			continue;
1703 		}
1704 
1705 		hlist_for_each_entry_safe(fa, tmp, &n->leaf, fa_list) {
1706 			hlist_del_rcu(&fa->fa_list);
1707 			alias_free_mem_rcu(fa);
1708 		}
1709 
1710 		put_child_root(pn, n->key, NULL);
1711 		node_free(n);
1712 	}
1713 
1714 #ifdef CONFIG_IP_FIB_TRIE_STATS
1715 	free_percpu(t->stats);
1716 #endif
1717 	kfree(tb);
1718 }
1719 
1720 struct fib_table *fib_trie_unmerge(struct fib_table *oldtb)
1721 {
1722 	struct trie *ot = (struct trie *)oldtb->tb_data;
1723 	struct key_vector *l, *tp = ot->kv;
1724 	struct fib_table *local_tb;
1725 	struct fib_alias *fa;
1726 	struct trie *lt;
1727 	t_key key = 0;
1728 
1729 	if (oldtb->tb_data == oldtb->__data)
1730 		return oldtb;
1731 
1732 	local_tb = fib_trie_table(RT_TABLE_LOCAL, NULL);
1733 	if (!local_tb)
1734 		return NULL;
1735 
1736 	lt = (struct trie *)local_tb->tb_data;
1737 
1738 	while ((l = leaf_walk_rcu(&tp, key)) != NULL) {
1739 		struct key_vector *local_l = NULL, *local_tp;
1740 
1741 		hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
1742 			struct fib_alias *new_fa;
1743 
1744 			if (local_tb->tb_id != fa->tb_id)
1745 				continue;
1746 
1747 			/* clone fa for new local table */
1748 			new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1749 			if (!new_fa)
1750 				goto out;
1751 
1752 			memcpy(new_fa, fa, sizeof(*fa));
1753 
1754 			/* insert clone into table */
1755 			if (!local_l)
1756 				local_l = fib_find_node(lt, &local_tp, l->key);
1757 
1758 			if (fib_insert_alias(lt, local_tp, local_l, new_fa,
1759 					     NULL, l->key)) {
1760 				kmem_cache_free(fn_alias_kmem, new_fa);
1761 				goto out;
1762 			}
1763 		}
1764 
1765 		/* stop loop if key wrapped back to 0 */
1766 		key = l->key + 1;
1767 		if (key < l->key)
1768 			break;
1769 	}
1770 
1771 	return local_tb;
1772 out:
1773 	fib_trie_free(local_tb);
1774 
1775 	return NULL;
1776 }
1777 
1778 /* Caller must hold RTNL */
1779 void fib_table_flush_external(struct fib_table *tb)
1780 {
1781 	struct trie *t = (struct trie *)tb->tb_data;
1782 	struct key_vector *pn = t->kv;
1783 	unsigned long cindex = 1;
1784 	struct hlist_node *tmp;
1785 	struct fib_alias *fa;
1786 
1787 	/* walk trie in reverse order */
1788 	for (;;) {
1789 		unsigned char slen = 0;
1790 		struct key_vector *n;
1791 
1792 		if (!(cindex--)) {
1793 			t_key pkey = pn->key;
1794 
1795 			/* cannot resize the trie vector */
1796 			if (IS_TRIE(pn))
1797 				break;
1798 
1799 			/* update the suffix to address pulled leaves */
1800 			if (pn->slen > pn->pos)
1801 				update_suffix(pn);
1802 
1803 			/* resize completed node */
1804 			pn = resize(t, pn);
1805 			cindex = get_index(pkey, pn);
1806 
1807 			continue;
1808 		}
1809 
1810 		/* grab the next available node */
1811 		n = get_child(pn, cindex);
1812 		if (!n)
1813 			continue;
1814 
1815 		if (IS_TNODE(n)) {
1816 			/* record pn and cindex for leaf walking */
1817 			pn = n;
1818 			cindex = 1ul << n->bits;
1819 
1820 			continue;
1821 		}
1822 
1823 		hlist_for_each_entry_safe(fa, tmp, &n->leaf, fa_list) {
1824 			/* if alias was cloned to local then we just
1825 			 * need to remove the local copy from main
1826 			 */
1827 			if (tb->tb_id != fa->tb_id) {
1828 				hlist_del_rcu(&fa->fa_list);
1829 				alias_free_mem_rcu(fa);
1830 				continue;
1831 			}
1832 
1833 			/* record local slen */
1834 			slen = fa->fa_slen;
1835 		}
1836 
1837 		/* update leaf slen */
1838 		n->slen = slen;
1839 
1840 		if (hlist_empty(&n->leaf)) {
1841 			put_child_root(pn, n->key, NULL);
1842 			node_free(n);
1843 		}
1844 	}
1845 }
1846 
1847 /* Caller must hold RTNL. */
1848 int fib_table_flush(struct net *net, struct fib_table *tb, bool flush_all)
1849 {
1850 	struct trie *t = (struct trie *)tb->tb_data;
1851 	struct key_vector *pn = t->kv;
1852 	unsigned long cindex = 1;
1853 	struct hlist_node *tmp;
1854 	struct fib_alias *fa;
1855 	int found = 0;
1856 
1857 	/* walk trie in reverse order */
1858 	for (;;) {
1859 		unsigned char slen = 0;
1860 		struct key_vector *n;
1861 
1862 		if (!(cindex--)) {
1863 			t_key pkey = pn->key;
1864 
1865 			/* cannot resize the trie vector */
1866 			if (IS_TRIE(pn))
1867 				break;
1868 
1869 			/* update the suffix to address pulled leaves */
1870 			if (pn->slen > pn->pos)
1871 				update_suffix(pn);
1872 
1873 			/* resize completed node */
1874 			pn = resize(t, pn);
1875 			cindex = get_index(pkey, pn);
1876 
1877 			continue;
1878 		}
1879 
1880 		/* grab the next available node */
1881 		n = get_child(pn, cindex);
1882 		if (!n)
1883 			continue;
1884 
1885 		if (IS_TNODE(n)) {
1886 			/* record pn and cindex for leaf walking */
1887 			pn = n;
1888 			cindex = 1ul << n->bits;
1889 
1890 			continue;
1891 		}
1892 
1893 		hlist_for_each_entry_safe(fa, tmp, &n->leaf, fa_list) {
1894 			struct fib_info *fi = fa->fa_info;
1895 
1896 			if (!fi || tb->tb_id != fa->tb_id ||
1897 			    (!(fi->fib_flags & RTNH_F_DEAD) &&
1898 			     !fib_props[fa->fa_type].error)) {
1899 				slen = fa->fa_slen;
1900 				continue;
1901 			}
1902 
1903 			/* Do not flush error routes if network namespace is
1904 			 * not being dismantled
1905 			 */
1906 			if (!flush_all && fib_props[fa->fa_type].error) {
1907 				slen = fa->fa_slen;
1908 				continue;
1909 			}
1910 
1911 			call_fib_entry_notifiers(net, FIB_EVENT_ENTRY_DEL,
1912 						 n->key,
1913 						 KEYLENGTH - fa->fa_slen, fa,
1914 						 NULL);
1915 			hlist_del_rcu(&fa->fa_list);
1916 			fib_release_info(fa->fa_info);
1917 			alias_free_mem_rcu(fa);
1918 			found++;
1919 		}
1920 
1921 		/* update leaf slen */
1922 		n->slen = slen;
1923 
1924 		if (hlist_empty(&n->leaf)) {
1925 			put_child_root(pn, n->key, NULL);
1926 			node_free(n);
1927 		}
1928 	}
1929 
1930 	pr_debug("trie_flush found=%d\n", found);
1931 	return found;
1932 }
1933 
1934 static void fib_leaf_notify(struct net *net, struct key_vector *l,
1935 			    struct fib_table *tb, struct notifier_block *nb)
1936 {
1937 	struct fib_alias *fa;
1938 
1939 	hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
1940 		struct fib_info *fi = fa->fa_info;
1941 
1942 		if (!fi)
1943 			continue;
1944 
1945 		/* local and main table can share the same trie,
1946 		 * so don't notify twice for the same entry.
1947 		 */
1948 		if (tb->tb_id != fa->tb_id)
1949 			continue;
1950 
1951 		call_fib_entry_notifier(nb, net, FIB_EVENT_ENTRY_ADD, l->key,
1952 					KEYLENGTH - fa->fa_slen, fa);
1953 	}
1954 }
1955 
1956 static void fib_table_notify(struct net *net, struct fib_table *tb,
1957 			     struct notifier_block *nb)
1958 {
1959 	struct trie *t = (struct trie *)tb->tb_data;
1960 	struct key_vector *l, *tp = t->kv;
1961 	t_key key = 0;
1962 
1963 	while ((l = leaf_walk_rcu(&tp, key)) != NULL) {
1964 		fib_leaf_notify(net, l, tb, nb);
1965 
1966 		key = l->key + 1;
1967 		/* stop in case of wrap around */
1968 		if (key < l->key)
1969 			break;
1970 	}
1971 }
1972 
1973 void fib_notify(struct net *net, struct notifier_block *nb)
1974 {
1975 	unsigned int h;
1976 
1977 	for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
1978 		struct hlist_head *head = &net->ipv4.fib_table_hash[h];
1979 		struct fib_table *tb;
1980 
1981 		hlist_for_each_entry_rcu(tb, head, tb_hlist)
1982 			fib_table_notify(net, tb, nb);
1983 	}
1984 }
1985 
1986 static void __trie_free_rcu(struct rcu_head *head)
1987 {
1988 	struct fib_table *tb = container_of(head, struct fib_table, rcu);
1989 #ifdef CONFIG_IP_FIB_TRIE_STATS
1990 	struct trie *t = (struct trie *)tb->tb_data;
1991 
1992 	if (tb->tb_data == tb->__data)
1993 		free_percpu(t->stats);
1994 #endif /* CONFIG_IP_FIB_TRIE_STATS */
1995 	kfree(tb);
1996 }
1997 
1998 void fib_free_table(struct fib_table *tb)
1999 {
2000 	call_rcu(&tb->rcu, __trie_free_rcu);
2001 }
2002 
2003 static int fn_trie_dump_leaf(struct key_vector *l, struct fib_table *tb,
2004 			     struct sk_buff *skb, struct netlink_callback *cb,
2005 			     struct fib_dump_filter *filter)
2006 {
2007 	unsigned int flags = NLM_F_MULTI;
2008 	__be32 xkey = htonl(l->key);
2009 	struct fib_alias *fa;
2010 	int i, s_i;
2011 
2012 	if (filter->filter_set)
2013 		flags |= NLM_F_DUMP_FILTERED;
2014 
2015 	s_i = cb->args[4];
2016 	i = 0;
2017 
2018 	/* rcu_read_lock is hold by caller */
2019 	hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
2020 		int err;
2021 
2022 		if (i < s_i)
2023 			goto next;
2024 
2025 		if (tb->tb_id != fa->tb_id)
2026 			goto next;
2027 
2028 		if (filter->filter_set) {
2029 			if (filter->rt_type && fa->fa_type != filter->rt_type)
2030 				goto next;
2031 
2032 			if ((filter->protocol &&
2033 			     fa->fa_info->fib_protocol != filter->protocol))
2034 				goto next;
2035 
2036 			if (filter->dev &&
2037 			    !fib_info_nh_uses_dev(fa->fa_info, filter->dev))
2038 				goto next;
2039 		}
2040 
2041 		err = fib_dump_info(skb, NETLINK_CB(cb->skb).portid,
2042 				    cb->nlh->nlmsg_seq, RTM_NEWROUTE,
2043 				    tb->tb_id, fa->fa_type,
2044 				    xkey, KEYLENGTH - fa->fa_slen,
2045 				    fa->fa_tos, fa->fa_info, flags);
2046 		if (err < 0) {
2047 			cb->args[4] = i;
2048 			return err;
2049 		}
2050 next:
2051 		i++;
2052 	}
2053 
2054 	cb->args[4] = i;
2055 	return skb->len;
2056 }
2057 
2058 /* rcu_read_lock needs to be hold by caller from readside */
2059 int fib_table_dump(struct fib_table *tb, struct sk_buff *skb,
2060 		   struct netlink_callback *cb, struct fib_dump_filter *filter)
2061 {
2062 	struct trie *t = (struct trie *)tb->tb_data;
2063 	struct key_vector *l, *tp = t->kv;
2064 	/* Dump starting at last key.
2065 	 * Note: 0.0.0.0/0 (ie default) is first key.
2066 	 */
2067 	int count = cb->args[2];
2068 	t_key key = cb->args[3];
2069 
2070 	while ((l = leaf_walk_rcu(&tp, key)) != NULL) {
2071 		int err;
2072 
2073 		err = fn_trie_dump_leaf(l, tb, skb, cb, filter);
2074 		if (err < 0) {
2075 			cb->args[3] = key;
2076 			cb->args[2] = count;
2077 			return err;
2078 		}
2079 
2080 		++count;
2081 		key = l->key + 1;
2082 
2083 		memset(&cb->args[4], 0,
2084 		       sizeof(cb->args) - 4*sizeof(cb->args[0]));
2085 
2086 		/* stop loop if key wrapped back to 0 */
2087 		if (key < l->key)
2088 			break;
2089 	}
2090 
2091 	cb->args[3] = key;
2092 	cb->args[2] = count;
2093 
2094 	return skb->len;
2095 }
2096 
2097 void __init fib_trie_init(void)
2098 {
2099 	fn_alias_kmem = kmem_cache_create("ip_fib_alias",
2100 					  sizeof(struct fib_alias),
2101 					  0, SLAB_PANIC, NULL);
2102 
2103 	trie_leaf_kmem = kmem_cache_create("ip_fib_trie",
2104 					   LEAF_SIZE,
2105 					   0, SLAB_PANIC, NULL);
2106 }
2107 
2108 struct fib_table *fib_trie_table(u32 id, struct fib_table *alias)
2109 {
2110 	struct fib_table *tb;
2111 	struct trie *t;
2112 	size_t sz = sizeof(*tb);
2113 
2114 	if (!alias)
2115 		sz += sizeof(struct trie);
2116 
2117 	tb = kzalloc(sz, GFP_KERNEL);
2118 	if (!tb)
2119 		return NULL;
2120 
2121 	tb->tb_id = id;
2122 	tb->tb_num_default = 0;
2123 	tb->tb_data = (alias ? alias->__data : tb->__data);
2124 
2125 	if (alias)
2126 		return tb;
2127 
2128 	t = (struct trie *) tb->tb_data;
2129 	t->kv[0].pos = KEYLENGTH;
2130 	t->kv[0].slen = KEYLENGTH;
2131 #ifdef CONFIG_IP_FIB_TRIE_STATS
2132 	t->stats = alloc_percpu(struct trie_use_stats);
2133 	if (!t->stats) {
2134 		kfree(tb);
2135 		tb = NULL;
2136 	}
2137 #endif
2138 
2139 	return tb;
2140 }
2141 
2142 #ifdef CONFIG_PROC_FS
2143 /* Depth first Trie walk iterator */
2144 struct fib_trie_iter {
2145 	struct seq_net_private p;
2146 	struct fib_table *tb;
2147 	struct key_vector *tnode;
2148 	unsigned int index;
2149 	unsigned int depth;
2150 };
2151 
2152 static struct key_vector *fib_trie_get_next(struct fib_trie_iter *iter)
2153 {
2154 	unsigned long cindex = iter->index;
2155 	struct key_vector *pn = iter->tnode;
2156 	t_key pkey;
2157 
2158 	pr_debug("get_next iter={node=%p index=%d depth=%d}\n",
2159 		 iter->tnode, iter->index, iter->depth);
2160 
2161 	while (!IS_TRIE(pn)) {
2162 		while (cindex < child_length(pn)) {
2163 			struct key_vector *n = get_child_rcu(pn, cindex++);
2164 
2165 			if (!n)
2166 				continue;
2167 
2168 			if (IS_LEAF(n)) {
2169 				iter->tnode = pn;
2170 				iter->index = cindex;
2171 			} else {
2172 				/* push down one level */
2173 				iter->tnode = n;
2174 				iter->index = 0;
2175 				++iter->depth;
2176 			}
2177 
2178 			return n;
2179 		}
2180 
2181 		/* Current node exhausted, pop back up */
2182 		pkey = pn->key;
2183 		pn = node_parent_rcu(pn);
2184 		cindex = get_index(pkey, pn) + 1;
2185 		--iter->depth;
2186 	}
2187 
2188 	/* record root node so further searches know we are done */
2189 	iter->tnode = pn;
2190 	iter->index = 0;
2191 
2192 	return NULL;
2193 }
2194 
2195 static struct key_vector *fib_trie_get_first(struct fib_trie_iter *iter,
2196 					     struct trie *t)
2197 {
2198 	struct key_vector *n, *pn;
2199 
2200 	if (!t)
2201 		return NULL;
2202 
2203 	pn = t->kv;
2204 	n = rcu_dereference(pn->tnode[0]);
2205 	if (!n)
2206 		return NULL;
2207 
2208 	if (IS_TNODE(n)) {
2209 		iter->tnode = n;
2210 		iter->index = 0;
2211 		iter->depth = 1;
2212 	} else {
2213 		iter->tnode = pn;
2214 		iter->index = 0;
2215 		iter->depth = 0;
2216 	}
2217 
2218 	return n;
2219 }
2220 
2221 static void trie_collect_stats(struct trie *t, struct trie_stat *s)
2222 {
2223 	struct key_vector *n;
2224 	struct fib_trie_iter iter;
2225 
2226 	memset(s, 0, sizeof(*s));
2227 
2228 	rcu_read_lock();
2229 	for (n = fib_trie_get_first(&iter, t); n; n = fib_trie_get_next(&iter)) {
2230 		if (IS_LEAF(n)) {
2231 			struct fib_alias *fa;
2232 
2233 			s->leaves++;
2234 			s->totdepth += iter.depth;
2235 			if (iter.depth > s->maxdepth)
2236 				s->maxdepth = iter.depth;
2237 
2238 			hlist_for_each_entry_rcu(fa, &n->leaf, fa_list)
2239 				++s->prefixes;
2240 		} else {
2241 			s->tnodes++;
2242 			if (n->bits < MAX_STAT_DEPTH)
2243 				s->nodesizes[n->bits]++;
2244 			s->nullpointers += tn_info(n)->empty_children;
2245 		}
2246 	}
2247 	rcu_read_unlock();
2248 }
2249 
2250 /*
2251  *	This outputs /proc/net/fib_triestats
2252  */
2253 static void trie_show_stats(struct seq_file *seq, struct trie_stat *stat)
2254 {
2255 	unsigned int i, max, pointers, bytes, avdepth;
2256 
2257 	if (stat->leaves)
2258 		avdepth = stat->totdepth*100 / stat->leaves;
2259 	else
2260 		avdepth = 0;
2261 
2262 	seq_printf(seq, "\tAver depth:     %u.%02d\n",
2263 		   avdepth / 100, avdepth % 100);
2264 	seq_printf(seq, "\tMax depth:      %u\n", stat->maxdepth);
2265 
2266 	seq_printf(seq, "\tLeaves:         %u\n", stat->leaves);
2267 	bytes = LEAF_SIZE * stat->leaves;
2268 
2269 	seq_printf(seq, "\tPrefixes:       %u\n", stat->prefixes);
2270 	bytes += sizeof(struct fib_alias) * stat->prefixes;
2271 
2272 	seq_printf(seq, "\tInternal nodes: %u\n\t", stat->tnodes);
2273 	bytes += TNODE_SIZE(0) * stat->tnodes;
2274 
2275 	max = MAX_STAT_DEPTH;
2276 	while (max > 0 && stat->nodesizes[max-1] == 0)
2277 		max--;
2278 
2279 	pointers = 0;
2280 	for (i = 1; i < max; i++)
2281 		if (stat->nodesizes[i] != 0) {
2282 			seq_printf(seq, "  %u: %u",  i, stat->nodesizes[i]);
2283 			pointers += (1<<i) * stat->nodesizes[i];
2284 		}
2285 	seq_putc(seq, '\n');
2286 	seq_printf(seq, "\tPointers: %u\n", pointers);
2287 
2288 	bytes += sizeof(struct key_vector *) * pointers;
2289 	seq_printf(seq, "Null ptrs: %u\n", stat->nullpointers);
2290 	seq_printf(seq, "Total size: %u  kB\n", (bytes + 1023) / 1024);
2291 }
2292 
2293 #ifdef CONFIG_IP_FIB_TRIE_STATS
2294 static void trie_show_usage(struct seq_file *seq,
2295 			    const struct trie_use_stats __percpu *stats)
2296 {
2297 	struct trie_use_stats s = { 0 };
2298 	int cpu;
2299 
2300 	/* loop through all of the CPUs and gather up the stats */
2301 	for_each_possible_cpu(cpu) {
2302 		const struct trie_use_stats *pcpu = per_cpu_ptr(stats, cpu);
2303 
2304 		s.gets += pcpu->gets;
2305 		s.backtrack += pcpu->backtrack;
2306 		s.semantic_match_passed += pcpu->semantic_match_passed;
2307 		s.semantic_match_miss += pcpu->semantic_match_miss;
2308 		s.null_node_hit += pcpu->null_node_hit;
2309 		s.resize_node_skipped += pcpu->resize_node_skipped;
2310 	}
2311 
2312 	seq_printf(seq, "\nCounters:\n---------\n");
2313 	seq_printf(seq, "gets = %u\n", s.gets);
2314 	seq_printf(seq, "backtracks = %u\n", s.backtrack);
2315 	seq_printf(seq, "semantic match passed = %u\n",
2316 		   s.semantic_match_passed);
2317 	seq_printf(seq, "semantic match miss = %u\n", s.semantic_match_miss);
2318 	seq_printf(seq, "null node hit= %u\n", s.null_node_hit);
2319 	seq_printf(seq, "skipped node resize = %u\n\n", s.resize_node_skipped);
2320 }
2321 #endif /*  CONFIG_IP_FIB_TRIE_STATS */
2322 
2323 static void fib_table_print(struct seq_file *seq, struct fib_table *tb)
2324 {
2325 	if (tb->tb_id == RT_TABLE_LOCAL)
2326 		seq_puts(seq, "Local:\n");
2327 	else if (tb->tb_id == RT_TABLE_MAIN)
2328 		seq_puts(seq, "Main:\n");
2329 	else
2330 		seq_printf(seq, "Id %d:\n", tb->tb_id);
2331 }
2332 
2333 
2334 static int fib_triestat_seq_show(struct seq_file *seq, void *v)
2335 {
2336 	struct net *net = (struct net *)seq->private;
2337 	unsigned int h;
2338 
2339 	seq_printf(seq,
2340 		   "Basic info: size of leaf:"
2341 		   " %zd bytes, size of tnode: %zd bytes.\n",
2342 		   LEAF_SIZE, TNODE_SIZE(0));
2343 
2344 	for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2345 		struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2346 		struct fib_table *tb;
2347 
2348 		hlist_for_each_entry_rcu(tb, head, tb_hlist) {
2349 			struct trie *t = (struct trie *) tb->tb_data;
2350 			struct trie_stat stat;
2351 
2352 			if (!t)
2353 				continue;
2354 
2355 			fib_table_print(seq, tb);
2356 
2357 			trie_collect_stats(t, &stat);
2358 			trie_show_stats(seq, &stat);
2359 #ifdef CONFIG_IP_FIB_TRIE_STATS
2360 			trie_show_usage(seq, t->stats);
2361 #endif
2362 		}
2363 	}
2364 
2365 	return 0;
2366 }
2367 
2368 static struct key_vector *fib_trie_get_idx(struct seq_file *seq, loff_t pos)
2369 {
2370 	struct fib_trie_iter *iter = seq->private;
2371 	struct net *net = seq_file_net(seq);
2372 	loff_t idx = 0;
2373 	unsigned int h;
2374 
2375 	for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2376 		struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2377 		struct fib_table *tb;
2378 
2379 		hlist_for_each_entry_rcu(tb, head, tb_hlist) {
2380 			struct key_vector *n;
2381 
2382 			for (n = fib_trie_get_first(iter,
2383 						    (struct trie *) tb->tb_data);
2384 			     n; n = fib_trie_get_next(iter))
2385 				if (pos == idx++) {
2386 					iter->tb = tb;
2387 					return n;
2388 				}
2389 		}
2390 	}
2391 
2392 	return NULL;
2393 }
2394 
2395 static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos)
2396 	__acquires(RCU)
2397 {
2398 	rcu_read_lock();
2399 	return fib_trie_get_idx(seq, *pos);
2400 }
2401 
2402 static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos)
2403 {
2404 	struct fib_trie_iter *iter = seq->private;
2405 	struct net *net = seq_file_net(seq);
2406 	struct fib_table *tb = iter->tb;
2407 	struct hlist_node *tb_node;
2408 	unsigned int h;
2409 	struct key_vector *n;
2410 
2411 	++*pos;
2412 	/* next node in same table */
2413 	n = fib_trie_get_next(iter);
2414 	if (n)
2415 		return n;
2416 
2417 	/* walk rest of this hash chain */
2418 	h = tb->tb_id & (FIB_TABLE_HASHSZ - 1);
2419 	while ((tb_node = rcu_dereference(hlist_next_rcu(&tb->tb_hlist)))) {
2420 		tb = hlist_entry(tb_node, struct fib_table, tb_hlist);
2421 		n = fib_trie_get_first(iter, (struct trie *) tb->tb_data);
2422 		if (n)
2423 			goto found;
2424 	}
2425 
2426 	/* new hash chain */
2427 	while (++h < FIB_TABLE_HASHSZ) {
2428 		struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2429 		hlist_for_each_entry_rcu(tb, head, tb_hlist) {
2430 			n = fib_trie_get_first(iter, (struct trie *) tb->tb_data);
2431 			if (n)
2432 				goto found;
2433 		}
2434 	}
2435 	return NULL;
2436 
2437 found:
2438 	iter->tb = tb;
2439 	return n;
2440 }
2441 
2442 static void fib_trie_seq_stop(struct seq_file *seq, void *v)
2443 	__releases(RCU)
2444 {
2445 	rcu_read_unlock();
2446 }
2447 
2448 static void seq_indent(struct seq_file *seq, int n)
2449 {
2450 	while (n-- > 0)
2451 		seq_puts(seq, "   ");
2452 }
2453 
2454 static inline const char *rtn_scope(char *buf, size_t len, enum rt_scope_t s)
2455 {
2456 	switch (s) {
2457 	case RT_SCOPE_UNIVERSE: return "universe";
2458 	case RT_SCOPE_SITE:	return "site";
2459 	case RT_SCOPE_LINK:	return "link";
2460 	case RT_SCOPE_HOST:	return "host";
2461 	case RT_SCOPE_NOWHERE:	return "nowhere";
2462 	default:
2463 		snprintf(buf, len, "scope=%d", s);
2464 		return buf;
2465 	}
2466 }
2467 
2468 static const char *const rtn_type_names[__RTN_MAX] = {
2469 	[RTN_UNSPEC] = "UNSPEC",
2470 	[RTN_UNICAST] = "UNICAST",
2471 	[RTN_LOCAL] = "LOCAL",
2472 	[RTN_BROADCAST] = "BROADCAST",
2473 	[RTN_ANYCAST] = "ANYCAST",
2474 	[RTN_MULTICAST] = "MULTICAST",
2475 	[RTN_BLACKHOLE] = "BLACKHOLE",
2476 	[RTN_UNREACHABLE] = "UNREACHABLE",
2477 	[RTN_PROHIBIT] = "PROHIBIT",
2478 	[RTN_THROW] = "THROW",
2479 	[RTN_NAT] = "NAT",
2480 	[RTN_XRESOLVE] = "XRESOLVE",
2481 };
2482 
2483 static inline const char *rtn_type(char *buf, size_t len, unsigned int t)
2484 {
2485 	if (t < __RTN_MAX && rtn_type_names[t])
2486 		return rtn_type_names[t];
2487 	snprintf(buf, len, "type %u", t);
2488 	return buf;
2489 }
2490 
2491 /* Pretty print the trie */
2492 static int fib_trie_seq_show(struct seq_file *seq, void *v)
2493 {
2494 	const struct fib_trie_iter *iter = seq->private;
2495 	struct key_vector *n = v;
2496 
2497 	if (IS_TRIE(node_parent_rcu(n)))
2498 		fib_table_print(seq, iter->tb);
2499 
2500 	if (IS_TNODE(n)) {
2501 		__be32 prf = htonl(n->key);
2502 
2503 		seq_indent(seq, iter->depth-1);
2504 		seq_printf(seq, "  +-- %pI4/%zu %u %u %u\n",
2505 			   &prf, KEYLENGTH - n->pos - n->bits, n->bits,
2506 			   tn_info(n)->full_children,
2507 			   tn_info(n)->empty_children);
2508 	} else {
2509 		__be32 val = htonl(n->key);
2510 		struct fib_alias *fa;
2511 
2512 		seq_indent(seq, iter->depth);
2513 		seq_printf(seq, "  |-- %pI4\n", &val);
2514 
2515 		hlist_for_each_entry_rcu(fa, &n->leaf, fa_list) {
2516 			char buf1[32], buf2[32];
2517 
2518 			seq_indent(seq, iter->depth + 1);
2519 			seq_printf(seq, "  /%zu %s %s",
2520 				   KEYLENGTH - fa->fa_slen,
2521 				   rtn_scope(buf1, sizeof(buf1),
2522 					     fa->fa_info->fib_scope),
2523 				   rtn_type(buf2, sizeof(buf2),
2524 					    fa->fa_type));
2525 			if (fa->fa_tos)
2526 				seq_printf(seq, " tos=%d", fa->fa_tos);
2527 			seq_putc(seq, '\n');
2528 		}
2529 	}
2530 
2531 	return 0;
2532 }
2533 
2534 static const struct seq_operations fib_trie_seq_ops = {
2535 	.start  = fib_trie_seq_start,
2536 	.next   = fib_trie_seq_next,
2537 	.stop   = fib_trie_seq_stop,
2538 	.show   = fib_trie_seq_show,
2539 };
2540 
2541 struct fib_route_iter {
2542 	struct seq_net_private p;
2543 	struct fib_table *main_tb;
2544 	struct key_vector *tnode;
2545 	loff_t	pos;
2546 	t_key	key;
2547 };
2548 
2549 static struct key_vector *fib_route_get_idx(struct fib_route_iter *iter,
2550 					    loff_t pos)
2551 {
2552 	struct key_vector *l, **tp = &iter->tnode;
2553 	t_key key;
2554 
2555 	/* use cached location of previously found key */
2556 	if (iter->pos > 0 && pos >= iter->pos) {
2557 		key = iter->key;
2558 	} else {
2559 		iter->pos = 1;
2560 		key = 0;
2561 	}
2562 
2563 	pos -= iter->pos;
2564 
2565 	while ((l = leaf_walk_rcu(tp, key)) && (pos-- > 0)) {
2566 		key = l->key + 1;
2567 		iter->pos++;
2568 		l = NULL;
2569 
2570 		/* handle unlikely case of a key wrap */
2571 		if (!key)
2572 			break;
2573 	}
2574 
2575 	if (l)
2576 		iter->key = l->key;	/* remember it */
2577 	else
2578 		iter->pos = 0;		/* forget it */
2579 
2580 	return l;
2581 }
2582 
2583 static void *fib_route_seq_start(struct seq_file *seq, loff_t *pos)
2584 	__acquires(RCU)
2585 {
2586 	struct fib_route_iter *iter = seq->private;
2587 	struct fib_table *tb;
2588 	struct trie *t;
2589 
2590 	rcu_read_lock();
2591 
2592 	tb = fib_get_table(seq_file_net(seq), RT_TABLE_MAIN);
2593 	if (!tb)
2594 		return NULL;
2595 
2596 	iter->main_tb = tb;
2597 	t = (struct trie *)tb->tb_data;
2598 	iter->tnode = t->kv;
2599 
2600 	if (*pos != 0)
2601 		return fib_route_get_idx(iter, *pos);
2602 
2603 	iter->pos = 0;
2604 	iter->key = KEY_MAX;
2605 
2606 	return SEQ_START_TOKEN;
2607 }
2608 
2609 static void *fib_route_seq_next(struct seq_file *seq, void *v, loff_t *pos)
2610 {
2611 	struct fib_route_iter *iter = seq->private;
2612 	struct key_vector *l = NULL;
2613 	t_key key = iter->key + 1;
2614 
2615 	++*pos;
2616 
2617 	/* only allow key of 0 for start of sequence */
2618 	if ((v == SEQ_START_TOKEN) || key)
2619 		l = leaf_walk_rcu(&iter->tnode, key);
2620 
2621 	if (l) {
2622 		iter->key = l->key;
2623 		iter->pos++;
2624 	} else {
2625 		iter->pos = 0;
2626 	}
2627 
2628 	return l;
2629 }
2630 
2631 static void fib_route_seq_stop(struct seq_file *seq, void *v)
2632 	__releases(RCU)
2633 {
2634 	rcu_read_unlock();
2635 }
2636 
2637 static unsigned int fib_flag_trans(int type, __be32 mask, const struct fib_info *fi)
2638 {
2639 	unsigned int flags = 0;
2640 
2641 	if (type == RTN_UNREACHABLE || type == RTN_PROHIBIT)
2642 		flags = RTF_REJECT;
2643 	if (fi && fi->fib_nh->fib_nh_gw4)
2644 		flags |= RTF_GATEWAY;
2645 	if (mask == htonl(0xFFFFFFFF))
2646 		flags |= RTF_HOST;
2647 	flags |= RTF_UP;
2648 	return flags;
2649 }
2650 
2651 /*
2652  *	This outputs /proc/net/route.
2653  *	The format of the file is not supposed to be changed
2654  *	and needs to be same as fib_hash output to avoid breaking
2655  *	legacy utilities
2656  */
2657 static int fib_route_seq_show(struct seq_file *seq, void *v)
2658 {
2659 	struct fib_route_iter *iter = seq->private;
2660 	struct fib_table *tb = iter->main_tb;
2661 	struct fib_alias *fa;
2662 	struct key_vector *l = v;
2663 	__be32 prefix;
2664 
2665 	if (v == SEQ_START_TOKEN) {
2666 		seq_printf(seq, "%-127s\n", "Iface\tDestination\tGateway "
2667 			   "\tFlags\tRefCnt\tUse\tMetric\tMask\t\tMTU"
2668 			   "\tWindow\tIRTT");
2669 		return 0;
2670 	}
2671 
2672 	prefix = htonl(l->key);
2673 
2674 	hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
2675 		const struct fib_info *fi = fa->fa_info;
2676 		__be32 mask = inet_make_mask(KEYLENGTH - fa->fa_slen);
2677 		unsigned int flags = fib_flag_trans(fa->fa_type, mask, fi);
2678 
2679 		if ((fa->fa_type == RTN_BROADCAST) ||
2680 		    (fa->fa_type == RTN_MULTICAST))
2681 			continue;
2682 
2683 		if (fa->tb_id != tb->tb_id)
2684 			continue;
2685 
2686 		seq_setwidth(seq, 127);
2687 
2688 		if (fi)
2689 			seq_printf(seq,
2690 				   "%s\t%08X\t%08X\t%04X\t%d\t%u\t"
2691 				   "%d\t%08X\t%d\t%u\t%u",
2692 				   fi->fib_dev ? fi->fib_dev->name : "*",
2693 				   prefix,
2694 				   fi->fib_nh->fib_nh_gw4, flags, 0, 0,
2695 				   fi->fib_priority,
2696 				   mask,
2697 				   (fi->fib_advmss ?
2698 				    fi->fib_advmss + 40 : 0),
2699 				   fi->fib_window,
2700 				   fi->fib_rtt >> 3);
2701 		else
2702 			seq_printf(seq,
2703 				   "*\t%08X\t%08X\t%04X\t%d\t%u\t"
2704 				   "%d\t%08X\t%d\t%u\t%u",
2705 				   prefix, 0, flags, 0, 0, 0,
2706 				   mask, 0, 0, 0);
2707 
2708 		seq_pad(seq, '\n');
2709 	}
2710 
2711 	return 0;
2712 }
2713 
2714 static const struct seq_operations fib_route_seq_ops = {
2715 	.start  = fib_route_seq_start,
2716 	.next   = fib_route_seq_next,
2717 	.stop   = fib_route_seq_stop,
2718 	.show   = fib_route_seq_show,
2719 };
2720 
2721 int __net_init fib_proc_init(struct net *net)
2722 {
2723 	if (!proc_create_net("fib_trie", 0444, net->proc_net, &fib_trie_seq_ops,
2724 			sizeof(struct fib_trie_iter)))
2725 		goto out1;
2726 
2727 	if (!proc_create_net_single("fib_triestat", 0444, net->proc_net,
2728 			fib_triestat_seq_show, NULL))
2729 		goto out2;
2730 
2731 	if (!proc_create_net("route", 0444, net->proc_net, &fib_route_seq_ops,
2732 			sizeof(struct fib_route_iter)))
2733 		goto out3;
2734 
2735 	return 0;
2736 
2737 out3:
2738 	remove_proc_entry("fib_triestat", net->proc_net);
2739 out2:
2740 	remove_proc_entry("fib_trie", net->proc_net);
2741 out1:
2742 	return -ENOMEM;
2743 }
2744 
2745 void __net_exit fib_proc_exit(struct net *net)
2746 {
2747 	remove_proc_entry("fib_trie", net->proc_net);
2748 	remove_proc_entry("fib_triestat", net->proc_net);
2749 	remove_proc_entry("route", net->proc_net);
2750 }
2751 
2752 #endif /* CONFIG_PROC_FS */
2753