Line data Source code
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 : * https://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 : #include <linux/cache.h>
39 : #include <linux/uaccess.h>
40 : #include <linux/bitops.h>
41 : #include <linux/types.h>
42 : #include <linux/kernel.h>
43 : #include <linux/mm.h>
44 : #include <linux/string.h>
45 : #include <linux/socket.h>
46 : #include <linux/sockios.h>
47 : #include <linux/errno.h>
48 : #include <linux/in.h>
49 : #include <linux/inet.h>
50 : #include <linux/inetdevice.h>
51 : #include <linux/netdevice.h>
52 : #include <linux/if_arp.h>
53 : #include <linux/proc_fs.h>
54 : #include <linux/rcupdate.h>
55 : #include <linux/skbuff.h>
56 : #include <linux/netlink.h>
57 : #include <linux/init.h>
58 : #include <linux/list.h>
59 : #include <linux/slab.h>
60 : #include <linux/export.h>
61 : #include <linux/vmalloc.h>
62 : #include <linux/notifier.h>
63 : #include <net/net_namespace.h>
64 : #include <net/ip.h>
65 : #include <net/protocol.h>
66 : #include <net/route.h>
67 : #include <net/tcp.h>
68 : #include <net/sock.h>
69 : #include <net/ip_fib.h>
70 : #include <net/fib_notifier.h>
71 : #include <trace/events/fib.h>
72 : #include "fib_lookup.h"
73 :
74 0 : static int call_fib_entry_notifier(struct notifier_block *nb,
75 : enum fib_event_type event_type, u32 dst,
76 : int dst_len, struct fib_alias *fa,
77 : struct netlink_ext_ack *extack)
78 : {
79 0 : struct fib_entry_notifier_info info = {
80 : .info.extack = extack,
81 : .dst = dst,
82 : .dst_len = dst_len,
83 0 : .fi = fa->fa_info,
84 0 : .tos = fa->fa_tos,
85 0 : .type = fa->fa_type,
86 0 : .tb_id = fa->tb_id,
87 : };
88 0 : return call_fib4_notifier(nb, event_type, &info.info);
89 : }
90 :
91 10 : static int call_fib_entry_notifiers(struct net *net,
92 : enum fib_event_type event_type, u32 dst,
93 : int dst_len, struct fib_alias *fa,
94 : struct netlink_ext_ack *extack)
95 : {
96 10 : struct fib_entry_notifier_info info = {
97 : .info.extack = extack,
98 : .dst = dst,
99 : .dst_len = dst_len,
100 10 : .fi = fa->fa_info,
101 10 : .tos = fa->fa_tos,
102 10 : .type = fa->fa_type,
103 10 : .tb_id = fa->tb_id,
104 : };
105 10 : return call_fib4_notifiers(net, event_type, &info.info);
106 : }
107 :
108 : #define MAX_STAT_DEPTH 32
109 :
110 : #define KEYLENGTH (8*sizeof(t_key))
111 : #define KEY_MAX ((t_key)~0)
112 :
113 : typedef unsigned int t_key;
114 :
115 : #define IS_TRIE(n) ((n)->pos >= KEYLENGTH)
116 : #define IS_TNODE(n) ((n)->bits)
117 : #define IS_LEAF(n) (!(n)->bits)
118 :
119 : struct key_vector {
120 : t_key key;
121 : unsigned char pos; /* 2log(KEYLENGTH) bits needed */
122 : unsigned char bits; /* 2log(KEYLENGTH) bits needed */
123 : unsigned char slen;
124 : union {
125 : /* This list pointer if valid if (pos | bits) == 0 (LEAF) */
126 : struct hlist_head leaf;
127 : /* This array is valid if (pos | bits) > 0 (TNODE) */
128 : struct key_vector __rcu *tnode[0];
129 : };
130 : };
131 :
132 : struct tnode {
133 : struct rcu_head rcu;
134 : t_key empty_children; /* KEYLENGTH bits needed */
135 : t_key full_children; /* KEYLENGTH bits needed */
136 : struct key_vector __rcu *parent;
137 : struct key_vector kv[1];
138 : #define tn_bits kv[0].bits
139 : };
140 :
141 : #define TNODE_SIZE(n) offsetof(struct tnode, kv[0].tnode[n])
142 : #define LEAF_SIZE TNODE_SIZE(1)
143 :
144 : #ifdef CONFIG_IP_FIB_TRIE_STATS
145 : struct trie_use_stats {
146 : unsigned int gets;
147 : unsigned int backtrack;
148 : unsigned int semantic_match_passed;
149 : unsigned int semantic_match_miss;
150 : unsigned int null_node_hit;
151 : unsigned int resize_node_skipped;
152 : };
153 : #endif
154 :
155 : struct trie_stat {
156 : unsigned int totdepth;
157 : unsigned int maxdepth;
158 : unsigned int tnodes;
159 : unsigned int leaves;
160 : unsigned int nullpointers;
161 : unsigned int prefixes;
162 : unsigned int nodesizes[MAX_STAT_DEPTH];
163 : };
164 :
165 : struct trie {
166 : struct key_vector kv[1];
167 : #ifdef CONFIG_IP_FIB_TRIE_STATS
168 : struct trie_use_stats __percpu *stats;
169 : #endif
170 : };
171 :
172 : static struct key_vector *resize(struct trie *t, struct key_vector *tn);
173 : static unsigned int tnode_free_size;
174 :
175 : /*
176 : * synchronize_rcu after call_rcu for outstanding dirty memory; it should be
177 : * especially useful before resizing the root node with PREEMPT_NONE configs;
178 : * the value was obtained experimentally, aiming to avoid visible slowdown.
179 : */
180 : unsigned int sysctl_fib_sync_mem = 512 * 1024;
181 : unsigned int sysctl_fib_sync_mem_min = 64 * 1024;
182 : unsigned int sysctl_fib_sync_mem_max = 64 * 1024 * 1024;
183 :
184 : static struct kmem_cache *fn_alias_kmem __ro_after_init;
185 : static struct kmem_cache *trie_leaf_kmem __ro_after_init;
186 :
187 246 : static inline struct tnode *tn_info(struct key_vector *kv)
188 : {
189 246 : return container_of(kv, struct tnode, kv[0]);
190 : }
191 :
192 : /* caller must hold RTNL */
193 : #define node_parent(tn) rtnl_dereference(tn_info(tn)->parent)
194 : #define get_child(tn, i) rtnl_dereference((tn)->tnode[i])
195 :
196 : /* caller must hold RCU read lock or RTNL */
197 : #define node_parent_rcu(tn) rcu_dereference_rtnl(tn_info(tn)->parent)
198 : #define get_child_rcu(tn, i) rcu_dereference_rtnl((tn)->tnode[i])
199 :
200 : /* wrapper for rcu_assign_pointer */
201 25 : static inline void node_set_parent(struct key_vector *n, struct key_vector *tp)
202 : {
203 25 : if (n)
204 2 : rcu_assign_pointer(tn_info(n)->parent, tp);
205 16 : }
206 :
207 : #define NODE_INIT_PARENT(n, p) RCU_INIT_POINTER(tn_info(n)->parent, p)
208 :
209 : /* This provides us with the number of children in this node, in the case of a
210 : * leaf this will return 0 meaning none of the children are accessible.
211 : */
212 102 : static inline unsigned long child_length(const struct key_vector *tn)
213 : {
214 102 : return (1ul << tn->bits) & ~(1ul);
215 : }
216 :
217 : #define get_cindex(key, kv) (((key) ^ (kv)->key) >> (kv)->pos)
218 :
219 83 : static inline unsigned long get_index(t_key key, struct key_vector *kv)
220 : {
221 83 : unsigned long index = key ^ kv->key;
222 :
223 83 : if ((BITS_PER_LONG <= KEYLENGTH) && (KEYLENGTH == kv->pos))
224 : return 0;
225 :
226 83 : return index >> kv->pos;
227 : }
228 :
229 : /* To understand this stuff, an understanding of keys and all their bits is
230 : * necessary. Every node in the trie has a key associated with it, but not
231 : * all of the bits in that key are significant.
232 : *
233 : * Consider a node 'n' and its parent 'tp'.
234 : *
235 : * If n is a leaf, every bit in its key is significant. Its presence is
236 : * necessitated by path compression, since during a tree traversal (when
237 : * searching for a leaf - unless we are doing an insertion) we will completely
238 : * ignore all skipped bits we encounter. Thus we need to verify, at the end of
239 : * a potentially successful search, that we have indeed been walking the
240 : * correct key path.
241 : *
242 : * Note that we can never "miss" the correct key in the tree if present by
243 : * following the wrong path. Path compression ensures that segments of the key
244 : * that are the same for all keys with a given prefix are skipped, but the
245 : * skipped part *is* identical for each node in the subtrie below the skipped
246 : * bit! trie_insert() in this implementation takes care of that.
247 : *
248 : * if n is an internal node - a 'tnode' here, the various parts of its key
249 : * have many different meanings.
250 : *
251 : * Example:
252 : * _________________________________________________________________
253 : * | i | i | i | i | i | i | i | N | N | N | S | S | S | S | S | C |
254 : * -----------------------------------------------------------------
255 : * 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
256 : *
257 : * _________________________________________________________________
258 : * | C | C | C | u | u | u | u | u | u | u | u | u | u | u | u | u |
259 : * -----------------------------------------------------------------
260 : * 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
261 : *
262 : * tp->pos = 22
263 : * tp->bits = 3
264 : * n->pos = 13
265 : * n->bits = 4
266 : *
267 : * First, let's just ignore the bits that come before the parent tp, that is
268 : * the bits from (tp->pos + tp->bits) to 31. They are *known* but at this
269 : * point we do not use them for anything.
270 : *
271 : * The bits from (tp->pos) to (tp->pos + tp->bits - 1) - "N", above - are the
272 : * index into the parent's child array. That is, they will be used to find
273 : * 'n' among tp's children.
274 : *
275 : * The bits from (n->pos + n->bits) to (tp->pos - 1) - "S" - are skipped bits
276 : * for the node n.
277 : *
278 : * All the bits we have seen so far are significant to the node n. The rest
279 : * of the bits are really not needed or indeed known in n->key.
280 : *
281 : * The bits from (n->pos) to (n->pos + n->bits - 1) - "C" - are the index into
282 : * n's child array, and will of course be different for each child.
283 : *
284 : * The rest of the bits, from 0 to (n->pos -1) - "u" - are completely unknown
285 : * at this point.
286 : */
287 :
288 : static const int halve_threshold = 25;
289 : static const int inflate_threshold = 50;
290 : static const int halve_threshold_root = 15;
291 : static const int inflate_threshold_root = 30;
292 :
293 0 : static void __alias_free_mem(struct rcu_head *head)
294 : {
295 0 : struct fib_alias *fa = container_of(head, struct fib_alias, rcu);
296 0 : kmem_cache_free(fn_alias_kmem, fa);
297 0 : }
298 :
299 0 : static inline void alias_free_mem_rcu(struct fib_alias *fa)
300 : {
301 0 : call_rcu(&fa->rcu, __alias_free_mem);
302 : }
303 :
304 : #define TNODE_VMALLOC_MAX \
305 : ilog2((SIZE_MAX - TNODE_SIZE(0)) / sizeof(struct key_vector *))
306 :
307 10 : static void __node_free_rcu(struct rcu_head *head)
308 : {
309 10 : struct tnode *n = container_of(head, struct tnode, rcu);
310 :
311 10 : if (!n->tn_bits)
312 0 : kmem_cache_free(trie_leaf_kmem, n);
313 : else
314 10 : kvfree(n);
315 10 : }
316 :
317 : #define node_free(n) call_rcu(&tn_info(n)->rcu, __node_free_rcu)
318 :
319 16 : static struct tnode *tnode_alloc(int bits)
320 : {
321 16 : size_t size;
322 :
323 : /* verify bits is within bounds */
324 16 : if (bits > TNODE_VMALLOC_MAX)
325 : return NULL;
326 :
327 : /* determine size and verify it is non-zero and didn't overflow */
328 16 : size = TNODE_SIZE(1ul << bits);
329 :
330 16 : if (size <= PAGE_SIZE)
331 16 : return kzalloc(size, GFP_KERNEL);
332 : else
333 0 : return vzalloc(size);
334 : }
335 :
336 0 : static inline void empty_child_inc(struct key_vector *n)
337 : {
338 0 : tn_info(n)->empty_children++;
339 :
340 0 : if (!tn_info(n)->empty_children)
341 0 : tn_info(n)->full_children++;
342 : }
343 :
344 32 : static inline void empty_child_dec(struct key_vector *n)
345 : {
346 32 : if (!tn_info(n)->empty_children)
347 0 : tn_info(n)->full_children--;
348 :
349 32 : tn_info(n)->empty_children--;
350 32 : }
351 :
352 8 : static struct key_vector *leaf_new(t_key key, struct fib_alias *fa)
353 : {
354 8 : struct key_vector *l;
355 8 : struct tnode *kv;
356 :
357 8 : kv = kmem_cache_alloc(trie_leaf_kmem, GFP_KERNEL);
358 8 : if (!kv)
359 : return NULL;
360 :
361 : /* initialize key vector */
362 8 : l = kv->kv;
363 8 : l->key = key;
364 8 : l->pos = 0;
365 8 : l->bits = 0;
366 8 : l->slen = fa->fa_slen;
367 :
368 : /* link leaf to fib alias */
369 8 : INIT_HLIST_HEAD(&l->leaf);
370 8 : hlist_add_head(&fa->fa_list, &l->leaf);
371 :
372 8 : return l;
373 : }
374 :
375 16 : static struct key_vector *tnode_new(t_key key, int pos, int bits)
376 : {
377 16 : unsigned int shift = pos + bits;
378 16 : struct key_vector *tn;
379 16 : struct tnode *tnode;
380 :
381 : /* verify bits and pos their msb bits clear and values are valid */
382 16 : BUG_ON(!bits || (shift > KEYLENGTH));
383 :
384 16 : tnode = tnode_alloc(bits);
385 16 : if (!tnode)
386 : return NULL;
387 :
388 16 : pr_debug("AT %p s=%zu %zu\n", tnode, TNODE_SIZE(0),
389 : sizeof(struct key_vector *) << bits);
390 :
391 16 : if (bits == KEYLENGTH)
392 0 : tnode->full_children = 1;
393 : else
394 16 : tnode->empty_children = 1ul << bits;
395 :
396 16 : tn = tnode->kv;
397 16 : tn->key = (shift < KEYLENGTH) ? (key >> shift) << shift : 0;
398 16 : tn->pos = pos;
399 16 : tn->bits = bits;
400 16 : tn->slen = pos;
401 :
402 16 : return tn;
403 : }
404 :
405 : /* Check whether a tnode 'n' is "full", i.e. it is an internal node
406 : * and no bits are skipped. See discussion in dyntree paper p. 6
407 : */
408 131 : static inline int tnode_full(struct key_vector *tn, struct key_vector *n)
409 : {
410 65 : return n && ((n->pos + n->bits) == tn->pos) && IS_TNODE(n);
411 : }
412 :
413 : /* Add a child at position i overwriting the old value.
414 : * Update the value of full_children and empty_children.
415 : */
416 42 : static void put_child(struct key_vector *tn, unsigned long i,
417 : struct key_vector *n)
418 : {
419 42 : struct key_vector *chi = get_child(tn, i);
420 42 : int isfull, wasfull;
421 :
422 42 : BUG_ON(i >= child_length(tn));
423 :
424 : /* update emptyChildren, overflow into fullChildren */
425 42 : if (!n && chi)
426 0 : empty_child_inc(tn);
427 42 : if (n && !chi)
428 32 : empty_child_dec(tn);
429 :
430 : /* update fullChildren */
431 42 : wasfull = tnode_full(tn, chi);
432 42 : isfull = tnode_full(tn, n);
433 :
434 42 : if (wasfull && !isfull)
435 2 : tn_info(tn)->full_children--;
436 40 : else if (!wasfull && isfull)
437 3 : tn_info(tn)->full_children++;
438 :
439 42 : if (n && (tn->slen < n->slen))
440 8 : tn->slen = n->slen;
441 :
442 42 : rcu_assign_pointer(tn->tnode[i], n);
443 42 : }
444 :
445 9 : static void update_children(struct key_vector *tn)
446 : {
447 9 : unsigned long i;
448 :
449 : /* update all of the child parent pointers */
450 9 : for (i = child_length(tn); i;) {
451 36 : struct key_vector *inode = get_child(tn, --i);
452 :
453 36 : if (!inode)
454 18 : continue;
455 :
456 : /* Either update the children of a tnode that
457 : * already belongs to us or update the child
458 : * to point to ourselves.
459 : */
460 18 : if (node_parent(inode) == tn)
461 2 : update_children(inode);
462 : else
463 61 : node_set_parent(inode, tn);
464 : }
465 9 : }
466 :
467 24 : static inline void put_child_root(struct key_vector *tp, t_key key,
468 : struct key_vector *n)
469 : {
470 24 : if (IS_TRIE(tp))
471 9 : rcu_assign_pointer(tp->tnode[0], n);
472 : else
473 15 : put_child(tp, get_index(key, tp), n);
474 24 : }
475 :
476 7 : static inline void tnode_free_init(struct key_vector *tn)
477 : {
478 7 : tn_info(tn)->rcu.next = NULL;
479 : }
480 :
481 3 : static inline void tnode_free_append(struct key_vector *tn,
482 : struct key_vector *n)
483 : {
484 3 : tn_info(n)->rcu.next = tn_info(tn)->rcu.next;
485 3 : tn_info(tn)->rcu.next = &tn_info(n)->rcu;
486 : }
487 :
488 7 : static void tnode_free(struct key_vector *tn)
489 : {
490 7 : struct callback_head *head = &tn_info(tn)->rcu;
491 :
492 15 : while (head) {
493 8 : head = head->next;
494 8 : tnode_free_size += TNODE_SIZE(1ul << tn->bits);
495 8 : node_free(tn);
496 :
497 8 : tn = container_of(head, struct tnode, rcu)->kv;
498 : }
499 :
500 7 : if (tnode_free_size >= sysctl_fib_sync_mem) {
501 0 : tnode_free_size = 0;
502 0 : synchronize_rcu();
503 : }
504 7 : }
505 :
506 7 : static struct key_vector *replace(struct trie *t,
507 : struct key_vector *oldtnode,
508 : struct key_vector *tn)
509 : {
510 7 : struct key_vector *tp = node_parent(oldtnode);
511 7 : unsigned long i;
512 :
513 : /* setup the parent pointer out of and back into this node */
514 7 : NODE_INIT_PARENT(tn, tp);
515 7 : put_child_root(tp, tn->key, tn);
516 :
517 : /* update all of the child parent pointers */
518 7 : update_children(tn);
519 :
520 : /* all pointers should be clean so we are done */
521 7 : tnode_free(oldtnode);
522 :
523 : /* resize children now that oldtnode is freed */
524 7 : for (i = child_length(tn); i;) {
525 32 : struct key_vector *inode = get_child(tn, --i);
526 :
527 : /* resize child node */
528 71 : if (tnode_full(tn, inode))
529 2 : tn = resize(t, inode);
530 : }
531 :
532 7 : return tp;
533 : }
534 :
535 7 : static struct key_vector *inflate(struct trie *t,
536 : struct key_vector *oldtnode)
537 : {
538 7 : struct key_vector *tn;
539 7 : unsigned long i;
540 7 : t_key m;
541 :
542 7 : pr_debug("In inflate\n");
543 :
544 7 : tn = tnode_new(oldtnode->key, oldtnode->pos - 1, oldtnode->bits + 1);
545 7 : if (!tn)
546 0 : goto notnode;
547 :
548 : /* prepare oldtnode to be freed */
549 7 : tnode_free_init(oldtnode);
550 :
551 : /* Assemble all of the pointers in our cluster, in this case that
552 : * represents all of the pointers out of our allocated nodes that
553 : * point to existing tnodes and the links between our allocated
554 : * nodes.
555 : */
556 23 : for (i = child_length(oldtnode), m = 1u << tn->pos; i;) {
557 16 : struct key_vector *inode = get_child(oldtnode, --i);
558 16 : struct key_vector *node0, *node1;
559 16 : unsigned long j, k;
560 :
561 : /* An empty child */
562 16 : if (!inode)
563 1 : continue;
564 :
565 : /* A leaf or an internal node with skipped bits */
566 15 : if (!tnode_full(oldtnode, inode)) {
567 14 : put_child(tn, get_index(inode->key, tn), inode);
568 14 : continue;
569 : }
570 :
571 : /* drop the node in the old tnode free list */
572 1 : tnode_free_append(oldtnode, inode);
573 :
574 : /* An internal node with two children */
575 1 : if (inode->bits == 1) {
576 0 : put_child(tn, 2 * i + 1, get_child(inode, 1));
577 0 : put_child(tn, 2 * i, get_child(inode, 0));
578 0 : continue;
579 : }
580 :
581 : /* We will replace this node 'inode' with two new
582 : * ones, 'node0' and 'node1', each with half of the
583 : * original children. The two new nodes will have
584 : * a position one bit further down the key and this
585 : * means that the "significant" part of their keys
586 : * (see the discussion near the top of this file)
587 : * will differ by one bit, which will be "0" in
588 : * node0's key and "1" in node1's key. Since we are
589 : * moving the key position by one step, the bit that
590 : * we are moving away from - the bit at position
591 : * (tn->pos) - is the one that will differ between
592 : * node0 and node1. So... we synthesize that bit in the
593 : * two new keys.
594 : */
595 1 : node1 = tnode_new(inode->key | m, inode->pos, inode->bits - 1);
596 1 : if (!node1)
597 0 : goto nomem;
598 1 : node0 = tnode_new(inode->key, inode->pos, inode->bits - 1);
599 :
600 1 : tnode_free_append(tn, node1);
601 1 : if (!node0)
602 0 : goto nomem;
603 1 : tnode_free_append(tn, node0);
604 :
605 : /* populate child pointers in new nodes */
606 2 : for (k = child_length(inode), j = k / 2; j;) {
607 1 : put_child(node1, --j, get_child(inode, --k));
608 1 : put_child(node0, j, get_child(inode, j));
609 1 : put_child(node1, --j, get_child(inode, --k));
610 1 : put_child(node0, j, get_child(inode, j));
611 : }
612 :
613 : /* link new nodes to parent */
614 1 : NODE_INIT_PARENT(node1, tn);
615 1 : NODE_INIT_PARENT(node0, tn);
616 :
617 : /* link parent to nodes */
618 1 : put_child(tn, 2 * i + 1, node1);
619 1 : put_child(tn, 2 * i, node0);
620 : }
621 :
622 : /* setup the parent pointers into and out of this node */
623 7 : return replace(t, oldtnode, tn);
624 0 : nomem:
625 : /* all pointers should be clean so we are done */
626 0 : tnode_free(tn);
627 : notnode:
628 : return NULL;
629 : }
630 :
631 0 : static struct key_vector *halve(struct trie *t,
632 : struct key_vector *oldtnode)
633 : {
634 0 : struct key_vector *tn;
635 0 : unsigned long i;
636 :
637 0 : pr_debug("In halve\n");
638 :
639 0 : tn = tnode_new(oldtnode->key, oldtnode->pos + 1, oldtnode->bits - 1);
640 0 : if (!tn)
641 0 : goto notnode;
642 :
643 : /* prepare oldtnode to be freed */
644 0 : tnode_free_init(oldtnode);
645 :
646 : /* Assemble all of the pointers in our cluster, in this case that
647 : * represents all of the pointers out of our allocated nodes that
648 : * point to existing tnodes and the links between our allocated
649 : * nodes.
650 : */
651 0 : for (i = child_length(oldtnode); i;) {
652 0 : struct key_vector *node1 = get_child(oldtnode, --i);
653 0 : struct key_vector *node0 = get_child(oldtnode, --i);
654 0 : struct key_vector *inode;
655 :
656 : /* At least one of the children is empty */
657 0 : if (!node1 || !node0) {
658 0 : put_child(tn, i / 2, node1 ? : node0);
659 0 : continue;
660 : }
661 :
662 : /* Two nonempty children */
663 0 : inode = tnode_new(node0->key, oldtnode->pos, 1);
664 0 : if (!inode)
665 0 : goto nomem;
666 0 : tnode_free_append(tn, inode);
667 :
668 : /* initialize pointers out of node */
669 0 : put_child(inode, 1, node1);
670 0 : put_child(inode, 0, node0);
671 0 : NODE_INIT_PARENT(inode, tn);
672 :
673 : /* link parent to node */
674 0 : put_child(tn, i / 2, inode);
675 : }
676 :
677 : /* setup the parent pointers into and out of this node */
678 0 : return replace(t, oldtnode, tn);
679 0 : nomem:
680 : /* all pointers should be clean so we are done */
681 0 : tnode_free(tn);
682 : notnode:
683 : return NULL;
684 : }
685 :
686 2 : static struct key_vector *collapse(struct trie *t,
687 : struct key_vector *oldtnode)
688 : {
689 2 : struct key_vector *n, *tp;
690 2 : unsigned long i;
691 :
692 : /* scan the tnode looking for that one child that might still exist */
693 5 : for (n = NULL, i = child_length(oldtnode); !n && i;)
694 3 : n = get_child(oldtnode, --i);
695 :
696 : /* compress one level */
697 2 : tp = node_parent(oldtnode);
698 2 : put_child_root(tp, oldtnode->key, n);
699 2 : node_set_parent(n, tp);
700 :
701 : /* drop dead node */
702 2 : node_free(oldtnode);
703 :
704 2 : return tp;
705 : }
706 :
707 0 : static unsigned char update_suffix(struct key_vector *tn)
708 : {
709 0 : unsigned char slen = tn->pos;
710 0 : unsigned long stride, i;
711 0 : unsigned char slen_max;
712 :
713 : /* only vector 0 can have a suffix length greater than or equal to
714 : * tn->pos + tn->bits, the second highest node will have a suffix
715 : * length at most of tn->pos + tn->bits - 1
716 : */
717 0 : slen_max = min_t(unsigned char, tn->pos + tn->bits - 1, tn->slen);
718 :
719 : /* search though the list of children looking for nodes that might
720 : * have a suffix greater than the one we currently have. This is
721 : * why we start with a stride of 2 since a stride of 1 would
722 : * represent the nodes with suffix length equal to tn->pos
723 : */
724 0 : for (i = 0, stride = 0x2ul ; i < child_length(tn); i += stride) {
725 0 : struct key_vector *n = get_child(tn, i);
726 :
727 0 : if (!n || (n->slen <= slen))
728 0 : continue;
729 :
730 : /* update stride and slen based on new value */
731 0 : stride <<= (n->slen - slen);
732 0 : slen = n->slen;
733 0 : i &= ~(stride - 1);
734 :
735 : /* stop searching if we have hit the maximum possible value */
736 0 : if (slen >= slen_max)
737 : break;
738 : }
739 :
740 0 : tn->slen = slen;
741 :
742 0 : return slen;
743 : }
744 :
745 : /* From "Implementing a dynamic compressed trie" by Stefan Nilsson of
746 : * the Helsinki University of Technology and Matti Tikkanen of Nokia
747 : * Telecommunications, page 6:
748 : * "A node is doubled if the ratio of non-empty children to all
749 : * children in the *doubled* node is at least 'high'."
750 : *
751 : * 'high' in this instance is the variable 'inflate_threshold'. It
752 : * is expressed as a percentage, so we multiply it with
753 : * child_length() and instead of multiplying by 2 (since the
754 : * child array will be doubled by inflate()) and multiplying
755 : * the left-hand side by 100 (to handle the percentage thing) we
756 : * multiply the left-hand side by 50.
757 : *
758 : * The left-hand side may look a bit weird: child_length(tn)
759 : * - tn->empty_children is of course the number of non-null children
760 : * in the current node. tn->full_children is the number of "full"
761 : * children, that is non-null tnodes with a skip value of 0.
762 : * All of those will be doubled in the resulting inflated tnode, so
763 : * we just count them one extra time here.
764 : *
765 : * A clearer way to write this would be:
766 : *
767 : * to_be_doubled = tn->full_children;
768 : * not_to_be_doubled = child_length(tn) - tn->empty_children -
769 : * tn->full_children;
770 : *
771 : * new_child_length = child_length(tn) * 2;
772 : *
773 : * new_fill_factor = 100 * (not_to_be_doubled + 2*to_be_doubled) /
774 : * new_child_length;
775 : * if (new_fill_factor >= inflate_threshold)
776 : *
777 : * ...and so on, tho it would mess up the while () loop.
778 : *
779 : * anyway,
780 : * 100 * (not_to_be_doubled + 2*to_be_doubled) / new_child_length >=
781 : * inflate_threshold
782 : *
783 : * avoid a division:
784 : * 100 * (not_to_be_doubled + 2*to_be_doubled) >=
785 : * inflate_threshold * new_child_length
786 : *
787 : * expand not_to_be_doubled and to_be_doubled, and shorten:
788 : * 100 * (child_length(tn) - tn->empty_children +
789 : * tn->full_children) >= inflate_threshold * new_child_length
790 : *
791 : * expand new_child_length:
792 : * 100 * (child_length(tn) - tn->empty_children +
793 : * tn->full_children) >=
794 : * inflate_threshold * child_length(tn) * 2
795 : *
796 : * shorten again:
797 : * 50 * (tn->full_children + child_length(tn) -
798 : * tn->empty_children) >= inflate_threshold *
799 : * child_length(tn)
800 : *
801 : */
802 20 : static inline bool should_inflate(struct key_vector *tp, struct key_vector *tn)
803 : {
804 20 : unsigned long used = child_length(tn);
805 20 : unsigned long threshold = used;
806 :
807 : /* Keep root node larger */
808 20 : threshold *= IS_TRIE(tp) ? inflate_threshold_root : inflate_threshold;
809 20 : used -= tn_info(tn)->empty_children;
810 20 : used += tn_info(tn)->full_children;
811 :
812 : /* if bits == KEYLENGTH then pos = 0, and will fail below */
813 :
814 20 : return (used > 1) && tn->pos && ((50 * used) >= threshold);
815 : }
816 :
817 7 : static inline bool should_halve(struct key_vector *tp, struct key_vector *tn)
818 : {
819 7 : unsigned long used = child_length(tn);
820 7 : unsigned long threshold = used;
821 :
822 : /* Keep root node larger */
823 7 : threshold *= IS_TRIE(tp) ? halve_threshold_root : halve_threshold;
824 7 : used -= tn_info(tn)->empty_children;
825 :
826 : /* if bits == KEYLENGTH then used = 100% on wrap, and will fail below */
827 :
828 7 : return (used > 1) && (tn->bits > 1) && ((100 * used) < threshold);
829 : }
830 :
831 7 : static inline bool should_collapse(struct key_vector *tn)
832 : {
833 7 : unsigned long used = child_length(tn);
834 :
835 7 : used -= tn_info(tn)->empty_children;
836 :
837 : /* account for bits == KEYLENGTH case */
838 7 : if ((tn->bits == KEYLENGTH) && tn_info(tn)->full_children)
839 0 : used -= KEY_MAX;
840 :
841 : /* One child or none, time to drop us from the trie */
842 7 : return used < 2;
843 : }
844 :
845 : #define MAX_WORK 10
846 13 : static struct key_vector *resize(struct trie *t, struct key_vector *tn)
847 : {
848 : #ifdef CONFIG_IP_FIB_TRIE_STATS
849 : struct trie_use_stats __percpu *stats = t->stats;
850 : #endif
851 13 : struct key_vector *tp = node_parent(tn);
852 13 : unsigned long cindex = get_index(tn->key, tp);
853 13 : int max_work = MAX_WORK;
854 :
855 13 : pr_debug("In tnode_resize %p inflate_threshold=%d threshold=%d\n",
856 : tn, inflate_threshold, halve_threshold);
857 :
858 : /* track the tnode via the pointer from the parent instead of
859 : * doing it ourselves. This way we can let RCU fully do its
860 : * thing without us interfering
861 : */
862 13 : BUG_ON(tn != get_child(tp, cindex));
863 :
864 : /* Double as long as the resulting node has a number of
865 : * nonempty nodes that are above the threshold.
866 : */
867 20 : while (should_inflate(tp, tn) && max_work) {
868 7 : tp = inflate(t, tn);
869 7 : if (!tp) {
870 : #ifdef CONFIG_IP_FIB_TRIE_STATS
871 : this_cpu_inc(stats->resize_node_skipped);
872 : #endif
873 : break;
874 : }
875 :
876 7 : max_work--;
877 7 : tn = get_child(tp, cindex);
878 : }
879 :
880 : /* update parent in case inflate failed */
881 13 : tp = node_parent(tn);
882 :
883 : /* Return if at least one inflate is run */
884 13 : if (max_work != MAX_WORK)
885 : return tp;
886 :
887 : /* Halve as long as the number of empty children in this
888 : * node is above threshold.
889 : */
890 7 : while (should_halve(tp, tn) && max_work) {
891 0 : tp = halve(t, tn);
892 0 : if (!tp) {
893 : #ifdef CONFIG_IP_FIB_TRIE_STATS
894 : this_cpu_inc(stats->resize_node_skipped);
895 : #endif
896 : break;
897 : }
898 :
899 0 : max_work--;
900 0 : tn = get_child(tp, cindex);
901 : }
902 :
903 : /* Only one child remains */
904 7 : if (should_collapse(tn))
905 2 : return collapse(t, tn);
906 :
907 : /* update parent in case halve failed */
908 5 : return node_parent(tn);
909 : }
910 :
911 0 : static void node_pull_suffix(struct key_vector *tn, unsigned char slen)
912 : {
913 0 : unsigned char node_slen = tn->slen;
914 :
915 0 : while ((node_slen > tn->pos) && (node_slen > slen)) {
916 0 : slen = update_suffix(tn);
917 0 : if (node_slen == slen)
918 : break;
919 :
920 0 : tn = node_parent(tn);
921 0 : node_slen = tn->slen;
922 : }
923 0 : }
924 :
925 8 : static void node_push_suffix(struct key_vector *tn, unsigned char slen)
926 : {
927 13 : while (tn->slen < slen) {
928 5 : tn->slen = slen;
929 5 : tn = node_parent(tn);
930 : }
931 8 : }
932 :
933 : /* rcu_read_lock needs to be hold by caller from readside */
934 20 : static struct key_vector *fib_find_node(struct trie *t,
935 : struct key_vector **tp, u32 key)
936 : {
937 20 : struct key_vector *pn, *n = t->kv;
938 20 : unsigned long index = 0;
939 :
940 41 : do {
941 41 : pn = n;
942 41 : n = get_child_rcu(n, index);
943 :
944 41 : if (!n)
945 : break;
946 :
947 40 : index = get_cindex(key, n);
948 :
949 : /* This bit of code is a bit tricky but it combines multiple
950 : * checks into a single check. The prefix consists of the
951 : * prefix plus zeros for the bits in the cindex. The index
952 : * is the difference between the key and this value. From
953 : * this we can actually derive several pieces of data.
954 : * if (index >= (1ul << bits))
955 : * we have a mismatch in skip bits and failed
956 : * else
957 : * we know the value is cindex
958 : *
959 : * This check is safe even if bits == KEYLENGTH due to the
960 : * fact that we can only allocate a node with 32 bits if a
961 : * long is greater than 32 bits.
962 : */
963 40 : if (index >= (1ul << n->bits)) {
964 : n = NULL;
965 : break;
966 : }
967 :
968 : /* keep searching until we find a perfect match leaf or NULL */
969 33 : } while (IS_TNODE(n));
970 :
971 20 : *tp = pn;
972 :
973 20 : return n;
974 : }
975 :
976 : /* Return the first fib alias matching TOS with
977 : * priority less than or equal to PRIO.
978 : * If 'find_first' is set, return the first matching
979 : * fib alias, regardless of TOS and priority.
980 : */
981 14 : static struct fib_alias *fib_find_alias(struct hlist_head *fah, u8 slen,
982 : u8 tos, u32 prio, u32 tb_id,
983 : bool find_first)
984 : {
985 14 : struct fib_alias *fa;
986 :
987 14 : if (!fah)
988 : return NULL;
989 :
990 28 : hlist_for_each_entry(fa, fah, fa_list) {
991 14 : if (fa->fa_slen < slen)
992 0 : continue;
993 14 : if (fa->fa_slen != slen)
994 : break;
995 12 : if (fa->tb_id > tb_id)
996 0 : continue;
997 12 : if (fa->tb_id != tb_id)
998 : break;
999 12 : if (find_first)
1000 10 : return fa;
1001 2 : if (fa->fa_tos > tos)
1002 0 : continue;
1003 2 : if (fa->fa_info->fib_priority >= prio || fa->fa_tos < tos)
1004 2 : return fa;
1005 : }
1006 :
1007 : return NULL;
1008 : }
1009 :
1010 : static struct fib_alias *
1011 0 : fib_find_matching_alias(struct net *net, const struct fib_rt_info *fri)
1012 : {
1013 0 : u8 slen = KEYLENGTH - fri->dst_len;
1014 0 : struct key_vector *l, *tp;
1015 0 : struct fib_table *tb;
1016 0 : struct fib_alias *fa;
1017 0 : struct trie *t;
1018 :
1019 0 : tb = fib_get_table(net, fri->tb_id);
1020 0 : if (!tb)
1021 : return NULL;
1022 :
1023 0 : t = (struct trie *)tb->tb_data;
1024 0 : l = fib_find_node(t, &tp, be32_to_cpu(fri->dst));
1025 0 : if (!l)
1026 : return NULL;
1027 :
1028 0 : hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
1029 0 : if (fa->fa_slen == slen && fa->tb_id == fri->tb_id &&
1030 0 : fa->fa_tos == fri->tos && fa->fa_info == fri->fi &&
1031 0 : fa->fa_type == fri->type)
1032 0 : return fa;
1033 : }
1034 :
1035 : return NULL;
1036 : }
1037 :
1038 0 : void fib_alias_hw_flags_set(struct net *net, const struct fib_rt_info *fri)
1039 : {
1040 0 : struct fib_alias *fa_match;
1041 0 : struct sk_buff *skb;
1042 0 : int err;
1043 :
1044 0 : rcu_read_lock();
1045 :
1046 0 : fa_match = fib_find_matching_alias(net, fri);
1047 0 : if (!fa_match)
1048 0 : goto out;
1049 :
1050 0 : if (fa_match->offload == fri->offload && fa_match->trap == fri->trap &&
1051 : fa_match->offload_failed == fri->offload_failed)
1052 0 : goto out;
1053 :
1054 0 : fa_match->offload = fri->offload;
1055 0 : fa_match->trap = fri->trap;
1056 :
1057 : /* 2 means send notifications only if offload_failed was changed. */
1058 0 : if (net->ipv4.sysctl_fib_notify_on_flag_change == 2 &&
1059 0 : fa_match->offload_failed == fri->offload_failed)
1060 0 : goto out;
1061 :
1062 0 : fa_match->offload_failed = fri->offload_failed;
1063 :
1064 0 : if (!net->ipv4.sysctl_fib_notify_on_flag_change)
1065 0 : goto out;
1066 :
1067 0 : skb = nlmsg_new(fib_nlmsg_size(fa_match->fa_info), GFP_ATOMIC);
1068 0 : if (!skb) {
1069 0 : err = -ENOBUFS;
1070 0 : goto errout;
1071 : }
1072 :
1073 0 : err = fib_dump_info(skb, 0, 0, RTM_NEWROUTE, fri, 0);
1074 0 : if (err < 0) {
1075 : /* -EMSGSIZE implies BUG in fib_nlmsg_size() */
1076 0 : WARN_ON(err == -EMSGSIZE);
1077 0 : kfree_skb(skb);
1078 0 : goto errout;
1079 : }
1080 :
1081 0 : rtnl_notify(skb, net, 0, RTNLGRP_IPV4_ROUTE, NULL, GFP_ATOMIC);
1082 0 : goto out;
1083 :
1084 0 : errout:
1085 0 : rtnl_set_sk_err(net, RTNLGRP_IPV4_ROUTE, err);
1086 0 : out:
1087 0 : rcu_read_unlock();
1088 0 : }
1089 : EXPORT_SYMBOL_GPL(fib_alias_hw_flags_set);
1090 :
1091 8 : static void trie_rebalance(struct trie *t, struct key_vector *tn)
1092 : {
1093 19 : while (!IS_TRIE(tn))
1094 11 : tn = resize(t, tn);
1095 8 : }
1096 :
1097 8 : static int fib_insert_node(struct trie *t, struct key_vector *tp,
1098 : struct fib_alias *new, t_key key)
1099 : {
1100 8 : struct key_vector *n, *l;
1101 :
1102 8 : l = leaf_new(key, new);
1103 8 : if (!l)
1104 0 : goto noleaf;
1105 :
1106 : /* retrieve child from parent node */
1107 8 : n = get_child(tp, get_index(key, tp));
1108 :
1109 : /* Case 2: n is a LEAF or a TNODE and the key doesn't match.
1110 : *
1111 : * Add a new tnode here
1112 : * first tnode need some special handling
1113 : * leaves us in position for handling as case 3
1114 : */
1115 8 : if (n) {
1116 7 : struct key_vector *tn;
1117 :
1118 7 : tn = tnode_new(key, __fls(key ^ n->key), 1);
1119 7 : if (!tn)
1120 0 : goto notnode;
1121 :
1122 : /* initialize routes out of node */
1123 7 : NODE_INIT_PARENT(tn, tp);
1124 7 : put_child(tn, get_index(key, tn) ^ 1, n);
1125 :
1126 : /* start adding routes into the node */
1127 7 : put_child_root(tp, key, tn);
1128 7 : node_set_parent(n, tn);
1129 :
1130 : /* parent now has a NULL spot where the leaf can go */
1131 7 : tp = tn;
1132 : }
1133 :
1134 : /* Case 3: n is NULL, and will just insert a new leaf */
1135 8 : node_push_suffix(tp, new->fa_slen);
1136 8 : NODE_INIT_PARENT(l, tp);
1137 8 : put_child_root(tp, key, l);
1138 8 : trie_rebalance(t, tp);
1139 :
1140 8 : return 0;
1141 0 : notnode:
1142 0 : node_free(l);
1143 : noleaf:
1144 : return -ENOMEM;
1145 : }
1146 :
1147 10 : static int fib_insert_alias(struct trie *t, struct key_vector *tp,
1148 : struct key_vector *l, struct fib_alias *new,
1149 : struct fib_alias *fa, t_key key)
1150 : {
1151 10 : if (!l)
1152 8 : return fib_insert_node(t, tp, new, key);
1153 :
1154 2 : if (fa) {
1155 0 : hlist_add_before_rcu(&new->fa_list, &fa->fa_list);
1156 : } else {
1157 2 : struct fib_alias *last;
1158 :
1159 4 : hlist_for_each_entry(last, &l->leaf, fa_list) {
1160 2 : if (new->fa_slen < last->fa_slen)
1161 : break;
1162 0 : if ((new->fa_slen == last->fa_slen) &&
1163 0 : (new->tb_id > last->tb_id))
1164 : break;
1165 0 : fa = last;
1166 : }
1167 :
1168 2 : if (fa)
1169 0 : hlist_add_behind_rcu(&new->fa_list, &fa->fa_list);
1170 : else
1171 2 : hlist_add_head_rcu(&new->fa_list, &l->leaf);
1172 : }
1173 :
1174 : /* if we added to the tail node then we need to update slen */
1175 2 : if (l->slen < new->fa_slen) {
1176 0 : l->slen = new->fa_slen;
1177 0 : node_push_suffix(tp, new->fa_slen);
1178 : }
1179 :
1180 : return 0;
1181 : }
1182 :
1183 12 : static bool fib_valid_key_len(u32 key, u8 plen, struct netlink_ext_ack *extack)
1184 : {
1185 12 : if (plen > KEYLENGTH) {
1186 0 : NL_SET_ERR_MSG(extack, "Invalid prefix length");
1187 0 : return false;
1188 : }
1189 :
1190 12 : if ((plen < KEYLENGTH) && (key << plen)) {
1191 0 : NL_SET_ERR_MSG(extack,
1192 : "Invalid prefix for given prefix length");
1193 0 : return false;
1194 : }
1195 :
1196 : return true;
1197 : }
1198 :
1199 : static void fib_remove_alias(struct trie *t, struct key_vector *tp,
1200 : struct key_vector *l, struct fib_alias *old);
1201 :
1202 : /* Caller must hold RTNL. */
1203 12 : int fib_table_insert(struct net *net, struct fib_table *tb,
1204 : struct fib_config *cfg, struct netlink_ext_ack *extack)
1205 : {
1206 12 : struct trie *t = (struct trie *)tb->tb_data;
1207 12 : struct fib_alias *fa, *new_fa;
1208 12 : struct key_vector *l, *tp;
1209 12 : u16 nlflags = NLM_F_EXCL;
1210 12 : struct fib_info *fi;
1211 12 : u8 plen = cfg->fc_dst_len;
1212 12 : u8 slen = KEYLENGTH - plen;
1213 12 : u8 tos = cfg->fc_tos;
1214 12 : u32 key;
1215 12 : int err;
1216 :
1217 12 : key = ntohl(cfg->fc_dst);
1218 :
1219 12 : if (!fib_valid_key_len(key, plen, extack))
1220 : return -EINVAL;
1221 :
1222 12 : pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen);
1223 :
1224 12 : fi = fib_create_info(cfg, extack);
1225 12 : if (IS_ERR(fi)) {
1226 0 : err = PTR_ERR(fi);
1227 0 : goto err;
1228 : }
1229 :
1230 12 : l = fib_find_node(t, &tp, key);
1231 4 : fa = l ? fib_find_alias(&l->leaf, slen, tos, fi->fib_priority,
1232 12 : tb->tb_id, false) : NULL;
1233 :
1234 : /* Now fa, if non-NULL, points to the first fib alias
1235 : * with the same keys [prefix,tos,priority], if such key already
1236 : * exists or to the node before which we will insert new one.
1237 : *
1238 : * If fa is NULL, we will need to allocate a new one and
1239 : * insert to the tail of the section matching the suffix length
1240 : * of the new alias.
1241 : */
1242 :
1243 4 : if (fa && fa->fa_tos == tos &&
1244 2 : fa->fa_info->fib_priority == fi->fib_priority) {
1245 2 : struct fib_alias *fa_first, *fa_match;
1246 :
1247 2 : err = -EEXIST;
1248 2 : if (cfg->fc_nlflags & NLM_F_EXCL)
1249 0 : goto out;
1250 :
1251 4 : nlflags &= ~NLM_F_EXCL;
1252 :
1253 : /* We have 2 goals:
1254 : * 1. Find exact match for type, scope, fib_info to avoid
1255 : * duplicate routes
1256 : * 2. Find next 'fa' (or head), NLM_F_APPEND inserts before it
1257 : */
1258 2 : fa_match = NULL;
1259 4 : fa_first = fa;
1260 2 : hlist_for_each_entry_from(fa, fa_list) {
1261 2 : if ((fa->fa_slen != slen) ||
1262 2 : (fa->tb_id != tb->tb_id) ||
1263 2 : (fa->fa_tos != tos))
1264 : break;
1265 2 : if (fa->fa_info->fib_priority != fi->fib_priority)
1266 : break;
1267 2 : if (fa->fa_type == cfg->fc_type &&
1268 : fa->fa_info == fi) {
1269 : fa_match = fa;
1270 : break;
1271 : }
1272 : }
1273 :
1274 2 : if (cfg->fc_nlflags & NLM_F_REPLACE) {
1275 0 : struct fib_info *fi_drop;
1276 0 : u8 state;
1277 :
1278 0 : nlflags |= NLM_F_REPLACE;
1279 0 : fa = fa_first;
1280 0 : if (fa_match) {
1281 0 : if (fa == fa_match)
1282 0 : err = 0;
1283 0 : goto out;
1284 : }
1285 0 : err = -ENOBUFS;
1286 0 : new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1287 0 : if (!new_fa)
1288 0 : goto out;
1289 :
1290 0 : fi_drop = fa->fa_info;
1291 0 : new_fa->fa_tos = fa->fa_tos;
1292 0 : new_fa->fa_info = fi;
1293 0 : new_fa->fa_type = cfg->fc_type;
1294 0 : state = fa->fa_state;
1295 0 : new_fa->fa_state = state & ~FA_S_ACCESSED;
1296 0 : new_fa->fa_slen = fa->fa_slen;
1297 0 : new_fa->tb_id = tb->tb_id;
1298 0 : new_fa->fa_default = -1;
1299 0 : new_fa->offload = 0;
1300 0 : new_fa->trap = 0;
1301 0 : new_fa->offload_failed = 0;
1302 :
1303 0 : hlist_replace_rcu(&fa->fa_list, &new_fa->fa_list);
1304 :
1305 0 : if (fib_find_alias(&l->leaf, fa->fa_slen, 0, 0,
1306 : tb->tb_id, true) == new_fa) {
1307 0 : enum fib_event_type fib_event;
1308 :
1309 0 : fib_event = FIB_EVENT_ENTRY_REPLACE;
1310 0 : err = call_fib_entry_notifiers(net, fib_event,
1311 : key, plen,
1312 : new_fa, extack);
1313 0 : if (err) {
1314 0 : hlist_replace_rcu(&new_fa->fa_list,
1315 : &fa->fa_list);
1316 0 : goto out_free_new_fa;
1317 : }
1318 : }
1319 :
1320 0 : rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen,
1321 0 : tb->tb_id, &cfg->fc_nlinfo, nlflags);
1322 :
1323 0 : alias_free_mem_rcu(fa);
1324 :
1325 0 : fib_release_info(fi_drop);
1326 0 : if (state & FA_S_ACCESSED)
1327 0 : rt_cache_flush(cfg->fc_nlinfo.nl_net);
1328 :
1329 0 : goto succeeded;
1330 : }
1331 : /* Error if we find a perfect match which
1332 : * uses the same scope, type, and nexthop
1333 : * information.
1334 : */
1335 2 : if (fa_match)
1336 2 : goto out;
1337 :
1338 0 : if (cfg->fc_nlflags & NLM_F_APPEND)
1339 : nlflags |= NLM_F_APPEND;
1340 : else
1341 0 : fa = fa_first;
1342 : }
1343 10 : err = -ENOENT;
1344 10 : if (!(cfg->fc_nlflags & NLM_F_CREATE))
1345 0 : goto out;
1346 :
1347 10 : nlflags |= NLM_F_CREATE;
1348 10 : err = -ENOBUFS;
1349 10 : new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1350 10 : if (!new_fa)
1351 0 : goto out;
1352 :
1353 10 : new_fa->fa_info = fi;
1354 10 : new_fa->fa_tos = tos;
1355 10 : new_fa->fa_type = cfg->fc_type;
1356 10 : new_fa->fa_state = 0;
1357 10 : new_fa->fa_slen = slen;
1358 10 : new_fa->tb_id = tb->tb_id;
1359 10 : new_fa->fa_default = -1;
1360 10 : new_fa->offload = 0;
1361 10 : new_fa->trap = 0;
1362 10 : new_fa->offload_failed = 0;
1363 :
1364 : /* Insert new entry to the list. */
1365 10 : err = fib_insert_alias(t, tp, l, new_fa, fa, key);
1366 10 : if (err)
1367 0 : goto out_free_new_fa;
1368 :
1369 : /* The alias was already inserted, so the node must exist. */
1370 10 : l = l ? l : fib_find_node(t, &tp, key);
1371 10 : if (WARN_ON_ONCE(!l))
1372 0 : goto out_free_new_fa;
1373 :
1374 10 : if (fib_find_alias(&l->leaf, new_fa->fa_slen, 0, 0, tb->tb_id, true) ==
1375 : new_fa) {
1376 10 : enum fib_event_type fib_event;
1377 :
1378 10 : fib_event = FIB_EVENT_ENTRY_REPLACE;
1379 10 : err = call_fib_entry_notifiers(net, fib_event, key, plen,
1380 : new_fa, extack);
1381 10 : if (err)
1382 0 : goto out_remove_new_fa;
1383 : }
1384 :
1385 10 : if (!plen)
1386 1 : tb->tb_num_default++;
1387 :
1388 10 : rt_cache_flush(cfg->fc_nlinfo.nl_net);
1389 10 : rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, new_fa->tb_id,
1390 10 : &cfg->fc_nlinfo, nlflags);
1391 : succeeded:
1392 : return 0;
1393 :
1394 0 : out_remove_new_fa:
1395 0 : fib_remove_alias(t, tp, l, new_fa);
1396 0 : out_free_new_fa:
1397 0 : kmem_cache_free(fn_alias_kmem, new_fa);
1398 2 : out:
1399 2 : fib_release_info(fi);
1400 : err:
1401 : return err;
1402 : }
1403 :
1404 49 : static inline t_key prefix_mismatch(t_key key, struct key_vector *n)
1405 : {
1406 49 : t_key prefix = n->key;
1407 :
1408 49 : return (key ^ prefix) & (prefix | -prefix);
1409 : }
1410 :
1411 67 : bool fib_lookup_good_nhc(const struct fib_nh_common *nhc, int fib_flags,
1412 : const struct flowi4 *flp)
1413 : {
1414 67 : if (nhc->nhc_flags & RTNH_F_DEAD)
1415 : return false;
1416 :
1417 67 : if (ip_ignore_linkdown(nhc->nhc_dev) &&
1418 0 : nhc->nhc_flags & RTNH_F_LINKDOWN &&
1419 0 : !(fib_flags & FIB_LOOKUP_IGNORE_LINKSTATE))
1420 : return false;
1421 :
1422 67 : if (!(flp->flowi4_flags & FLOWI_FLAG_SKIP_NH_OIF)) {
1423 67 : if (flp->flowi4_oif &&
1424 1 : flp->flowi4_oif != nhc->nhc_oif)
1425 0 : return false;
1426 : }
1427 :
1428 : return true;
1429 : }
1430 :
1431 : /* should be called with rcu_read_lock */
1432 87 : int fib_table_lookup(struct fib_table *tb, const struct flowi4 *flp,
1433 : struct fib_result *res, int fib_flags)
1434 : {
1435 87 : struct trie *t = (struct trie *) tb->tb_data;
1436 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1437 : struct trie_use_stats __percpu *stats = t->stats;
1438 : #endif
1439 87 : const t_key key = ntohl(flp->daddr);
1440 87 : struct key_vector *n, *pn;
1441 87 : struct fib_alias *fa;
1442 87 : unsigned long index;
1443 87 : t_key cindex;
1444 :
1445 87 : pn = t->kv;
1446 87 : cindex = 0;
1447 :
1448 87 : n = get_child_rcu(pn, cindex);
1449 87 : if (!n) {
1450 0 : trace_fib_table_lookup(tb->tb_id, flp, NULL, -EAGAIN);
1451 0 : return -EAGAIN;
1452 : }
1453 :
1454 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1455 : this_cpu_inc(stats->gets);
1456 : #endif
1457 :
1458 : /* Step 1: Travel to the longest prefix match in the trie */
1459 330 : for (;;) {
1460 330 : index = get_cindex(key, n);
1461 :
1462 : /* This bit of code is a bit tricky but it combines multiple
1463 : * checks into a single check. The prefix consists of the
1464 : * prefix plus zeros for the "bits" in the prefix. The index
1465 : * is the difference between the key and this value. From
1466 : * this we can actually derive several pieces of data.
1467 : * if (index >= (1ul << bits))
1468 : * we have a mismatch in skip bits and failed
1469 : * else
1470 : * we know the value is cindex
1471 : *
1472 : * This check is safe even if bits == KEYLENGTH due to the
1473 : * fact that we can only allocate a node with 32 bits if a
1474 : * long is greater than 32 bits.
1475 : */
1476 330 : if (index >= (1ul << n->bits))
1477 : break;
1478 :
1479 : /* we have found a leaf. Prefixes have already been compared */
1480 281 : if (IS_LEAF(n))
1481 38 : goto found;
1482 :
1483 : /* only record pn and cindex if we are going to be chopping
1484 : * bits later. Otherwise we are just wasting cycles.
1485 : */
1486 243 : if (n->slen > n->pos) {
1487 237 : pn = n;
1488 237 : cindex = index;
1489 : }
1490 :
1491 243 : n = get_child_rcu(n, index);
1492 243 : if (unlikely(!n))
1493 0 : goto backtrace;
1494 : }
1495 :
1496 : /* Step 2: Sort out leaves and begin backtracing for longest prefix */
1497 49 : for (;;) {
1498 : /* record the pointer where our next node pointer is stored */
1499 49 : struct key_vector __rcu **cptr = n->tnode;
1500 :
1501 : /* This test verifies that none of the bits that differ
1502 : * between the key and the prefix exist in the region of
1503 : * the lsb and higher in the prefix.
1504 : */
1505 49 : if (unlikely(prefix_mismatch(key, n)) || (n->slen == n->pos))
1506 20 : goto backtrace;
1507 :
1508 : /* exit out and process leaf */
1509 29 : if (unlikely(IS_LEAF(n)))
1510 : break;
1511 :
1512 : /* Don't bother recording parent info. Since we are in
1513 : * prefix match mode we will have to come back to wherever
1514 : * we started this traversal anyway
1515 : */
1516 :
1517 0 : while ((n = rcu_dereference(*cptr)) == NULL) {
1518 0 : backtrace:
1519 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1520 : if (!n)
1521 : this_cpu_inc(stats->null_node_hit);
1522 : #endif
1523 : /* If we are at cindex 0 there are no more bits for
1524 : * us to strip at this level so we must ascend back
1525 : * up one level to see if there are any more bits to
1526 : * be stripped there.
1527 : */
1528 20 : while (!cindex) {
1529 20 : t_key pkey = pn->key;
1530 :
1531 : /* If we don't have a parent then there is
1532 : * nothing for us to do as we do not have any
1533 : * further nodes to parse.
1534 : */
1535 20 : if (IS_TRIE(pn)) {
1536 20 : trace_fib_table_lookup(tb->tb_id, flp,
1537 : NULL, -EAGAIN);
1538 20 : return -EAGAIN;
1539 : }
1540 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1541 : this_cpu_inc(stats->backtrack);
1542 : #endif
1543 : /* Get Child's index */
1544 0 : pn = node_parent_rcu(pn);
1545 0 : cindex = get_index(pkey, pn);
1546 : }
1547 :
1548 : /* strip the least significant bit from the cindex */
1549 0 : cindex &= cindex - 1;
1550 :
1551 : /* grab pointer for next child node */
1552 0 : cptr = &pn->tnode[cindex];
1553 : }
1554 : }
1555 :
1556 29 : found:
1557 : /* this line carries forward the xor from earlier in the function */
1558 67 : index = key ^ n->key;
1559 :
1560 : /* Step 3: Process the leaf, if that fails fall back to backtracing */
1561 163 : hlist_for_each_entry_rcu(fa, &n->leaf, fa_list) {
1562 96 : struct fib_info *fi = fa->fa_info;
1563 96 : struct fib_nh_common *nhc;
1564 96 : int nhsel, err;
1565 :
1566 96 : if ((BITS_PER_LONG > KEYLENGTH) || (fa->fa_slen < KEYLENGTH)) {
1567 96 : if (index >= (1ul << fa->fa_slen))
1568 29 : continue;
1569 : }
1570 67 : if (fa->fa_tos && fa->fa_tos != flp->flowi4_tos)
1571 0 : continue;
1572 67 : if (fi->fib_dead)
1573 0 : continue;
1574 67 : if (fa->fa_info->fib_scope < flp->flowi4_scope)
1575 0 : continue;
1576 67 : fib_alias_accessed(fa);
1577 67 : err = fib_props[fa->fa_type].error;
1578 67 : if (unlikely(err < 0)) {
1579 0 : out_reject:
1580 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1581 : this_cpu_inc(stats->semantic_match_passed);
1582 : #endif
1583 0 : trace_fib_table_lookup(tb->tb_id, flp, NULL, err);
1584 67 : return err;
1585 : }
1586 67 : if (fi->fib_flags & RTNH_F_DEAD)
1587 0 : continue;
1588 :
1589 67 : if (unlikely(fi->nh)) {
1590 0 : if (nexthop_is_blackhole(fi->nh)) {
1591 0 : err = fib_props[RTN_BLACKHOLE].error;
1592 0 : goto out_reject;
1593 : }
1594 :
1595 0 : nhc = nexthop_get_nhc_lookup(fi->nh, fib_flags, flp,
1596 : &nhsel);
1597 0 : if (nhc)
1598 0 : goto set_result;
1599 0 : goto miss;
1600 : }
1601 :
1602 134 : for (nhsel = 0; nhsel < fib_info_num_path(fi); nhsel++) {
1603 67 : nhc = fib_info_nhc(fi, nhsel);
1604 :
1605 67 : if (!fib_lookup_good_nhc(nhc, fib_flags, flp))
1606 0 : continue;
1607 67 : set_result:
1608 67 : if (!(fib_flags & FIB_LOOKUP_NOREF))
1609 0 : refcount_inc(&fi->fib_clntref);
1610 :
1611 67 : res->prefix = htonl(n->key);
1612 67 : res->prefixlen = KEYLENGTH - fa->fa_slen;
1613 67 : res->nh_sel = nhsel;
1614 67 : res->nhc = nhc;
1615 67 : res->type = fa->fa_type;
1616 67 : res->scope = fi->fib_scope;
1617 67 : res->fi = fi;
1618 67 : res->table = tb;
1619 67 : res->fa_head = &n->leaf;
1620 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1621 : this_cpu_inc(stats->semantic_match_passed);
1622 : #endif
1623 67 : trace_fib_table_lookup(tb->tb_id, flp, nhc, err);
1624 :
1625 67 : return err;
1626 : }
1627 : }
1628 0 : miss:
1629 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1630 : this_cpu_inc(stats->semantic_match_miss);
1631 : #endif
1632 0 : goto backtrace;
1633 : }
1634 : EXPORT_SYMBOL_GPL(fib_table_lookup);
1635 :
1636 0 : static void fib_remove_alias(struct trie *t, struct key_vector *tp,
1637 : struct key_vector *l, struct fib_alias *old)
1638 : {
1639 : /* record the location of the previous list_info entry */
1640 0 : struct hlist_node **pprev = old->fa_list.pprev;
1641 0 : struct fib_alias *fa = hlist_entry(pprev, typeof(*fa), fa_list.next);
1642 :
1643 : /* remove the fib_alias from the list */
1644 0 : hlist_del_rcu(&old->fa_list);
1645 :
1646 : /* if we emptied the list this leaf will be freed and we can sort
1647 : * out parent suffix lengths as a part of trie_rebalance
1648 : */
1649 0 : if (hlist_empty(&l->leaf)) {
1650 0 : if (tp->slen == l->slen)
1651 0 : node_pull_suffix(tp, tp->pos);
1652 0 : put_child_root(tp, l->key, NULL);
1653 0 : node_free(l);
1654 0 : trie_rebalance(t, tp);
1655 0 : return;
1656 : }
1657 :
1658 : /* only access fa if it is pointing at the last valid hlist_node */
1659 0 : if (*pprev)
1660 : return;
1661 :
1662 : /* update the trie with the latest suffix length */
1663 0 : l->slen = fa->fa_slen;
1664 0 : node_pull_suffix(tp, fa->fa_slen);
1665 : }
1666 :
1667 0 : static void fib_notify_alias_delete(struct net *net, u32 key,
1668 : struct hlist_head *fah,
1669 : struct fib_alias *fa_to_delete,
1670 : struct netlink_ext_ack *extack)
1671 : {
1672 0 : struct fib_alias *fa_next, *fa_to_notify;
1673 0 : u32 tb_id = fa_to_delete->tb_id;
1674 0 : u8 slen = fa_to_delete->fa_slen;
1675 0 : enum fib_event_type fib_event;
1676 :
1677 : /* Do not notify if we do not care about the route. */
1678 0 : if (fib_find_alias(fah, slen, 0, 0, tb_id, true) != fa_to_delete)
1679 : return;
1680 :
1681 : /* Determine if the route should be replaced by the next route in the
1682 : * list.
1683 : */
1684 0 : fa_next = hlist_entry_safe(fa_to_delete->fa_list.next,
1685 : struct fib_alias, fa_list);
1686 0 : if (fa_next && fa_next->fa_slen == slen && fa_next->tb_id == tb_id) {
1687 : fib_event = FIB_EVENT_ENTRY_REPLACE;
1688 : fa_to_notify = fa_next;
1689 : } else {
1690 0 : fib_event = FIB_EVENT_ENTRY_DEL;
1691 0 : fa_to_notify = fa_to_delete;
1692 : }
1693 0 : call_fib_entry_notifiers(net, fib_event, key, KEYLENGTH - slen,
1694 : fa_to_notify, extack);
1695 : }
1696 :
1697 : /* Caller must hold RTNL. */
1698 0 : int fib_table_delete(struct net *net, struct fib_table *tb,
1699 : struct fib_config *cfg, struct netlink_ext_ack *extack)
1700 : {
1701 0 : struct trie *t = (struct trie *) tb->tb_data;
1702 0 : struct fib_alias *fa, *fa_to_delete;
1703 0 : struct key_vector *l, *tp;
1704 0 : u8 plen = cfg->fc_dst_len;
1705 0 : u8 slen = KEYLENGTH - plen;
1706 0 : u8 tos = cfg->fc_tos;
1707 0 : u32 key;
1708 :
1709 0 : key = ntohl(cfg->fc_dst);
1710 :
1711 0 : if (!fib_valid_key_len(key, plen, extack))
1712 : return -EINVAL;
1713 :
1714 0 : l = fib_find_node(t, &tp, key);
1715 0 : if (!l)
1716 : return -ESRCH;
1717 :
1718 0 : fa = fib_find_alias(&l->leaf, slen, tos, 0, tb->tb_id, false);
1719 0 : if (!fa)
1720 : return -ESRCH;
1721 :
1722 : pr_debug("Deleting %08x/%d tos=%d t=%p\n", key, plen, tos, t);
1723 :
1724 0 : fa_to_delete = NULL;
1725 0 : hlist_for_each_entry_from(fa, fa_list) {
1726 0 : struct fib_info *fi = fa->fa_info;
1727 :
1728 0 : if ((fa->fa_slen != slen) ||
1729 0 : (fa->tb_id != tb->tb_id) ||
1730 0 : (fa->fa_tos != tos))
1731 : break;
1732 :
1733 0 : if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) &&
1734 0 : (cfg->fc_scope == RT_SCOPE_NOWHERE ||
1735 0 : fa->fa_info->fib_scope == cfg->fc_scope) &&
1736 0 : (!cfg->fc_prefsrc ||
1737 0 : fi->fib_prefsrc == cfg->fc_prefsrc) &&
1738 0 : (!cfg->fc_protocol ||
1739 0 : fi->fib_protocol == cfg->fc_protocol) &&
1740 0 : fib_nh_match(net, cfg, fi, extack) == 0 &&
1741 0 : fib_metrics_match(cfg, fi)) {
1742 : fa_to_delete = fa;
1743 : break;
1744 : }
1745 : }
1746 :
1747 0 : if (!fa_to_delete)
1748 : return -ESRCH;
1749 :
1750 0 : fib_notify_alias_delete(net, key, &l->leaf, fa_to_delete, extack);
1751 0 : rtmsg_fib(RTM_DELROUTE, htonl(key), fa_to_delete, plen, tb->tb_id,
1752 0 : &cfg->fc_nlinfo, 0);
1753 :
1754 0 : if (!plen)
1755 0 : tb->tb_num_default--;
1756 :
1757 0 : fib_remove_alias(t, tp, l, fa_to_delete);
1758 :
1759 0 : if (fa_to_delete->fa_state & FA_S_ACCESSED)
1760 0 : rt_cache_flush(cfg->fc_nlinfo.nl_net);
1761 :
1762 0 : fib_release_info(fa_to_delete->fa_info);
1763 0 : alias_free_mem_rcu(fa_to_delete);
1764 0 : return 0;
1765 : }
1766 :
1767 : /* Scan for the next leaf starting at the provided key value */
1768 16 : static struct key_vector *leaf_walk_rcu(struct key_vector **tn, t_key key)
1769 : {
1770 16 : struct key_vector *pn, *n = *tn;
1771 20 : unsigned long cindex;
1772 :
1773 : /* this loop is meant to try and find the key in the trie */
1774 20 : do {
1775 : /* record parent and next child index */
1776 20 : pn = n;
1777 20 : cindex = (key > pn->key) ? get_index(key, pn) : 0;
1778 :
1779 20 : if (cindex >> pn->bits)
1780 : break;
1781 :
1782 : /* descend into the next child */
1783 12 : n = get_child_rcu(pn, cindex++);
1784 12 : if (!n)
1785 : break;
1786 :
1787 : /* guarantee forward progress on the keys */
1788 12 : if (IS_LEAF(n) && (n->key >= key))
1789 4 : goto found;
1790 8 : } while (IS_TNODE(n));
1791 :
1792 : /* this loop will search for the next leaf with a greater key */
1793 52 : while (!IS_TRIE(pn)) {
1794 : /* if we exhausted the parent node we will need to climb */
1795 50 : if (cindex >= (1ul << pn->bits)) {
1796 12 : t_key pkey = pn->key;
1797 :
1798 12 : pn = node_parent_rcu(pn);
1799 12 : cindex = get_index(pkey, pn) + 1;
1800 12 : continue;
1801 : }
1802 :
1803 : /* grab the next available node */
1804 38 : n = get_child_rcu(pn, cindex++);
1805 38 : if (!n)
1806 20 : continue;
1807 :
1808 : /* no need to compare keys since we bumped the index */
1809 18 : if (IS_LEAF(n))
1810 10 : goto found;
1811 :
1812 : /* Rescan start scanning in new node */
1813 : pn = n;
1814 : cindex = 0;
1815 : }
1816 :
1817 2 : *tn = pn;
1818 2 : return NULL; /* Root of trie */
1819 14 : found:
1820 : /* if we are at the limit for keys just return NULL for the tnode */
1821 14 : *tn = pn;
1822 14 : return n;
1823 : }
1824 :
1825 0 : static void fib_trie_free(struct fib_table *tb)
1826 : {
1827 0 : struct trie *t = (struct trie *)tb->tb_data;
1828 0 : struct key_vector *pn = t->kv;
1829 0 : unsigned long cindex = 1;
1830 0 : struct hlist_node *tmp;
1831 0 : struct fib_alias *fa;
1832 :
1833 : /* walk trie in reverse order and free everything */
1834 0 : for (;;) {
1835 0 : struct key_vector *n;
1836 :
1837 0 : if (!(cindex--)) {
1838 0 : t_key pkey = pn->key;
1839 :
1840 0 : if (IS_TRIE(pn))
1841 : break;
1842 :
1843 0 : n = pn;
1844 0 : pn = node_parent(pn);
1845 :
1846 : /* drop emptied tnode */
1847 0 : put_child_root(pn, n->key, NULL);
1848 0 : node_free(n);
1849 :
1850 0 : cindex = get_index(pkey, pn);
1851 :
1852 0 : continue;
1853 : }
1854 :
1855 : /* grab the next available node */
1856 0 : n = get_child(pn, cindex);
1857 0 : if (!n)
1858 0 : continue;
1859 :
1860 0 : if (IS_TNODE(n)) {
1861 : /* record pn and cindex for leaf walking */
1862 0 : pn = n;
1863 0 : cindex = 1ul << n->bits;
1864 :
1865 0 : continue;
1866 : }
1867 :
1868 0 : hlist_for_each_entry_safe(fa, tmp, &n->leaf, fa_list) {
1869 0 : hlist_del_rcu(&fa->fa_list);
1870 0 : alias_free_mem_rcu(fa);
1871 : }
1872 :
1873 0 : put_child_root(pn, n->key, NULL);
1874 0 : node_free(n);
1875 : }
1876 :
1877 : #ifdef CONFIG_IP_FIB_TRIE_STATS
1878 : free_percpu(t->stats);
1879 : #endif
1880 0 : kfree(tb);
1881 0 : }
1882 :
1883 0 : struct fib_table *fib_trie_unmerge(struct fib_table *oldtb)
1884 : {
1885 0 : struct trie *ot = (struct trie *)oldtb->tb_data;
1886 0 : struct key_vector *l, *tp = ot->kv;
1887 0 : struct fib_table *local_tb;
1888 0 : struct fib_alias *fa;
1889 0 : struct trie *lt;
1890 0 : t_key key = 0;
1891 :
1892 0 : if (oldtb->tb_data == oldtb->__data)
1893 : return oldtb;
1894 :
1895 0 : local_tb = fib_trie_table(RT_TABLE_LOCAL, NULL);
1896 0 : if (!local_tb)
1897 : return NULL;
1898 :
1899 0 : lt = (struct trie *)local_tb->tb_data;
1900 :
1901 0 : while ((l = leaf_walk_rcu(&tp, key)) != NULL) {
1902 0 : struct key_vector *local_l = NULL, *local_tp;
1903 :
1904 0 : hlist_for_each_entry(fa, &l->leaf, fa_list) {
1905 0 : struct fib_alias *new_fa;
1906 :
1907 0 : if (local_tb->tb_id != fa->tb_id)
1908 0 : continue;
1909 :
1910 : /* clone fa for new local table */
1911 0 : new_fa = kmem_cache_alloc(fn_alias_kmem, GFP_KERNEL);
1912 0 : if (!new_fa)
1913 0 : goto out;
1914 :
1915 0 : memcpy(new_fa, fa, sizeof(*fa));
1916 :
1917 : /* insert clone into table */
1918 0 : if (!local_l)
1919 0 : local_l = fib_find_node(lt, &local_tp, l->key);
1920 :
1921 0 : if (fib_insert_alias(lt, local_tp, local_l, new_fa,
1922 : NULL, l->key)) {
1923 0 : kmem_cache_free(fn_alias_kmem, new_fa);
1924 0 : goto out;
1925 : }
1926 : }
1927 :
1928 : /* stop loop if key wrapped back to 0 */
1929 0 : key = l->key + 1;
1930 0 : if (key < l->key)
1931 : break;
1932 : }
1933 :
1934 : return local_tb;
1935 0 : out:
1936 0 : fib_trie_free(local_tb);
1937 :
1938 0 : return NULL;
1939 : }
1940 :
1941 : /* Caller must hold RTNL */
1942 0 : void fib_table_flush_external(struct fib_table *tb)
1943 : {
1944 0 : struct trie *t = (struct trie *)tb->tb_data;
1945 0 : struct key_vector *pn = t->kv;
1946 0 : unsigned long cindex = 1;
1947 0 : struct hlist_node *tmp;
1948 0 : struct fib_alias *fa;
1949 :
1950 : /* walk trie in reverse order */
1951 0 : for (;;) {
1952 0 : unsigned char slen = 0;
1953 0 : struct key_vector *n;
1954 :
1955 0 : if (!(cindex--)) {
1956 0 : t_key pkey = pn->key;
1957 :
1958 : /* cannot resize the trie vector */
1959 0 : if (IS_TRIE(pn))
1960 : break;
1961 :
1962 : /* update the suffix to address pulled leaves */
1963 0 : if (pn->slen > pn->pos)
1964 0 : update_suffix(pn);
1965 :
1966 : /* resize completed node */
1967 0 : pn = resize(t, pn);
1968 0 : cindex = get_index(pkey, pn);
1969 :
1970 0 : continue;
1971 : }
1972 :
1973 : /* grab the next available node */
1974 0 : n = get_child(pn, cindex);
1975 0 : if (!n)
1976 0 : continue;
1977 :
1978 0 : if (IS_TNODE(n)) {
1979 : /* record pn and cindex for leaf walking */
1980 0 : pn = n;
1981 0 : cindex = 1ul << n->bits;
1982 :
1983 0 : continue;
1984 : }
1985 :
1986 0 : hlist_for_each_entry_safe(fa, tmp, &n->leaf, fa_list) {
1987 : /* if alias was cloned to local then we just
1988 : * need to remove the local copy from main
1989 : */
1990 0 : if (tb->tb_id != fa->tb_id) {
1991 0 : hlist_del_rcu(&fa->fa_list);
1992 0 : alias_free_mem_rcu(fa);
1993 0 : continue;
1994 : }
1995 :
1996 : /* record local slen */
1997 0 : slen = fa->fa_slen;
1998 : }
1999 :
2000 : /* update leaf slen */
2001 0 : n->slen = slen;
2002 :
2003 0 : if (hlist_empty(&n->leaf)) {
2004 0 : put_child_root(pn, n->key, NULL);
2005 0 : node_free(n);
2006 : }
2007 : }
2008 0 : }
2009 :
2010 : /* Caller must hold RTNL. */
2011 0 : int fib_table_flush(struct net *net, struct fib_table *tb, bool flush_all)
2012 : {
2013 0 : struct trie *t = (struct trie *)tb->tb_data;
2014 0 : struct key_vector *pn = t->kv;
2015 0 : unsigned long cindex = 1;
2016 0 : struct hlist_node *tmp;
2017 0 : struct fib_alias *fa;
2018 0 : int found = 0;
2019 :
2020 : /* walk trie in reverse order */
2021 0 : for (;;) {
2022 0 : unsigned char slen = 0;
2023 0 : struct key_vector *n;
2024 :
2025 0 : if (!(cindex--)) {
2026 0 : t_key pkey = pn->key;
2027 :
2028 : /* cannot resize the trie vector */
2029 0 : if (IS_TRIE(pn))
2030 : break;
2031 :
2032 : /* update the suffix to address pulled leaves */
2033 0 : if (pn->slen > pn->pos)
2034 0 : update_suffix(pn);
2035 :
2036 : /* resize completed node */
2037 0 : pn = resize(t, pn);
2038 0 : cindex = get_index(pkey, pn);
2039 :
2040 0 : continue;
2041 : }
2042 :
2043 : /* grab the next available node */
2044 0 : n = get_child(pn, cindex);
2045 0 : if (!n)
2046 0 : continue;
2047 :
2048 0 : if (IS_TNODE(n)) {
2049 : /* record pn and cindex for leaf walking */
2050 0 : pn = n;
2051 0 : cindex = 1ul << n->bits;
2052 :
2053 0 : continue;
2054 : }
2055 :
2056 0 : hlist_for_each_entry_safe(fa, tmp, &n->leaf, fa_list) {
2057 0 : struct fib_info *fi = fa->fa_info;
2058 :
2059 0 : if (!fi || tb->tb_id != fa->tb_id ||
2060 0 : (!(fi->fib_flags & RTNH_F_DEAD) &&
2061 0 : !fib_props[fa->fa_type].error)) {
2062 0 : slen = fa->fa_slen;
2063 0 : continue;
2064 : }
2065 :
2066 : /* Do not flush error routes if network namespace is
2067 : * not being dismantled
2068 : */
2069 0 : if (!flush_all && fib_props[fa->fa_type].error) {
2070 0 : slen = fa->fa_slen;
2071 0 : continue;
2072 : }
2073 :
2074 0 : fib_notify_alias_delete(net, n->key, &n->leaf, fa,
2075 : NULL);
2076 0 : hlist_del_rcu(&fa->fa_list);
2077 0 : fib_release_info(fa->fa_info);
2078 0 : alias_free_mem_rcu(fa);
2079 0 : found++;
2080 : }
2081 :
2082 : /* update leaf slen */
2083 0 : n->slen = slen;
2084 :
2085 0 : if (hlist_empty(&n->leaf)) {
2086 0 : put_child_root(pn, n->key, NULL);
2087 0 : node_free(n);
2088 : }
2089 : }
2090 :
2091 0 : pr_debug("trie_flush found=%d\n", found);
2092 0 : return found;
2093 : }
2094 :
2095 : /* derived from fib_trie_free */
2096 0 : static void __fib_info_notify_update(struct net *net, struct fib_table *tb,
2097 : struct nl_info *info)
2098 : {
2099 0 : struct trie *t = (struct trie *)tb->tb_data;
2100 0 : struct key_vector *pn = t->kv;
2101 0 : unsigned long cindex = 1;
2102 0 : struct fib_alias *fa;
2103 :
2104 0 : for (;;) {
2105 0 : struct key_vector *n;
2106 :
2107 0 : if (!(cindex--)) {
2108 0 : t_key pkey = pn->key;
2109 :
2110 0 : if (IS_TRIE(pn))
2111 : break;
2112 :
2113 0 : pn = node_parent(pn);
2114 0 : cindex = get_index(pkey, pn);
2115 0 : continue;
2116 : }
2117 :
2118 : /* grab the next available node */
2119 0 : n = get_child(pn, cindex);
2120 0 : if (!n)
2121 0 : continue;
2122 :
2123 0 : if (IS_TNODE(n)) {
2124 : /* record pn and cindex for leaf walking */
2125 0 : pn = n;
2126 0 : cindex = 1ul << n->bits;
2127 :
2128 0 : continue;
2129 : }
2130 :
2131 0 : hlist_for_each_entry(fa, &n->leaf, fa_list) {
2132 0 : struct fib_info *fi = fa->fa_info;
2133 :
2134 0 : if (!fi || !fi->nh_updated || fa->tb_id != tb->tb_id)
2135 0 : continue;
2136 :
2137 0 : rtmsg_fib(RTM_NEWROUTE, htonl(n->key), fa,
2138 0 : KEYLENGTH - fa->fa_slen, tb->tb_id,
2139 : info, NLM_F_REPLACE);
2140 : }
2141 : }
2142 0 : }
2143 :
2144 0 : void fib_info_notify_update(struct net *net, struct nl_info *info)
2145 : {
2146 0 : unsigned int h;
2147 :
2148 0 : for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2149 0 : struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2150 0 : struct fib_table *tb;
2151 :
2152 0 : hlist_for_each_entry_rcu(tb, head, tb_hlist,
2153 : lockdep_rtnl_is_held())
2154 0 : __fib_info_notify_update(net, tb, info);
2155 : }
2156 0 : }
2157 :
2158 0 : static int fib_leaf_notify(struct key_vector *l, struct fib_table *tb,
2159 : struct notifier_block *nb,
2160 : struct netlink_ext_ack *extack)
2161 : {
2162 0 : struct fib_alias *fa;
2163 0 : int last_slen = -1;
2164 0 : int err;
2165 :
2166 0 : hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
2167 0 : struct fib_info *fi = fa->fa_info;
2168 :
2169 0 : if (!fi)
2170 0 : continue;
2171 :
2172 : /* local and main table can share the same trie,
2173 : * so don't notify twice for the same entry.
2174 : */
2175 0 : if (tb->tb_id != fa->tb_id)
2176 0 : continue;
2177 :
2178 0 : if (fa->fa_slen == last_slen)
2179 0 : continue;
2180 :
2181 0 : last_slen = fa->fa_slen;
2182 0 : err = call_fib_entry_notifier(nb, FIB_EVENT_ENTRY_REPLACE,
2183 0 : l->key, KEYLENGTH - fa->fa_slen,
2184 : fa, extack);
2185 0 : if (err)
2186 0 : return err;
2187 : }
2188 : return 0;
2189 : }
2190 :
2191 0 : static int fib_table_notify(struct fib_table *tb, struct notifier_block *nb,
2192 : struct netlink_ext_ack *extack)
2193 : {
2194 0 : struct trie *t = (struct trie *)tb->tb_data;
2195 0 : struct key_vector *l, *tp = t->kv;
2196 0 : t_key key = 0;
2197 0 : int err;
2198 :
2199 0 : while ((l = leaf_walk_rcu(&tp, key)) != NULL) {
2200 0 : err = fib_leaf_notify(l, tb, nb, extack);
2201 0 : if (err)
2202 0 : return err;
2203 :
2204 0 : key = l->key + 1;
2205 : /* stop in case of wrap around */
2206 0 : if (key < l->key)
2207 : break;
2208 : }
2209 : return 0;
2210 : }
2211 :
2212 0 : int fib_notify(struct net *net, struct notifier_block *nb,
2213 : struct netlink_ext_ack *extack)
2214 : {
2215 0 : unsigned int h;
2216 0 : int err;
2217 :
2218 0 : for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2219 0 : struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2220 0 : struct fib_table *tb;
2221 :
2222 0 : hlist_for_each_entry_rcu(tb, head, tb_hlist) {
2223 0 : err = fib_table_notify(tb, nb, extack);
2224 0 : if (err)
2225 0 : return err;
2226 : }
2227 : }
2228 : return 0;
2229 : }
2230 :
2231 0 : static void __trie_free_rcu(struct rcu_head *head)
2232 : {
2233 0 : struct fib_table *tb = container_of(head, struct fib_table, rcu);
2234 : #ifdef CONFIG_IP_FIB_TRIE_STATS
2235 : struct trie *t = (struct trie *)tb->tb_data;
2236 :
2237 : if (tb->tb_data == tb->__data)
2238 : free_percpu(t->stats);
2239 : #endif /* CONFIG_IP_FIB_TRIE_STATS */
2240 0 : kfree(tb);
2241 0 : }
2242 :
2243 0 : void fib_free_table(struct fib_table *tb)
2244 : {
2245 0 : call_rcu(&tb->rcu, __trie_free_rcu);
2246 0 : }
2247 :
2248 14 : static int fn_trie_dump_leaf(struct key_vector *l, struct fib_table *tb,
2249 : struct sk_buff *skb, struct netlink_callback *cb,
2250 : struct fib_dump_filter *filter)
2251 : {
2252 14 : unsigned int flags = NLM_F_MULTI;
2253 14 : __be32 xkey = htonl(l->key);
2254 14 : int i, s_i, i_fa, s_fa, err;
2255 14 : struct fib_alias *fa;
2256 :
2257 14 : if (filter->filter_set ||
2258 14 : !filter->dump_exceptions || !filter->dump_routes)
2259 0 : flags |= NLM_F_DUMP_FILTERED;
2260 :
2261 14 : s_i = cb->args[4];
2262 14 : s_fa = cb->args[5];
2263 14 : i = 0;
2264 :
2265 : /* rcu_read_lock is hold by caller */
2266 46 : hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
2267 18 : struct fib_info *fi = fa->fa_info;
2268 :
2269 18 : if (i < s_i)
2270 0 : goto next;
2271 :
2272 18 : i_fa = 0;
2273 :
2274 18 : if (tb->tb_id != fa->tb_id)
2275 9 : goto next;
2276 :
2277 9 : if (filter->filter_set) {
2278 0 : if (filter->rt_type && fa->fa_type != filter->rt_type)
2279 0 : goto next;
2280 :
2281 0 : if ((filter->protocol &&
2282 0 : fi->fib_protocol != filter->protocol))
2283 0 : goto next;
2284 :
2285 0 : if (filter->dev &&
2286 0 : !fib_info_nh_uses_dev(fi, filter->dev))
2287 0 : goto next;
2288 : }
2289 :
2290 9 : if (filter->dump_routes) {
2291 9 : if (!s_fa) {
2292 9 : struct fib_rt_info fri;
2293 :
2294 9 : fri.fi = fi;
2295 9 : fri.tb_id = tb->tb_id;
2296 9 : fri.dst = xkey;
2297 9 : fri.dst_len = KEYLENGTH - fa->fa_slen;
2298 9 : fri.tos = fa->fa_tos;
2299 9 : fri.type = fa->fa_type;
2300 9 : fri.offload = fa->offload;
2301 9 : fri.trap = fa->trap;
2302 9 : fri.offload_failed = fa->offload_failed;
2303 18 : err = fib_dump_info(skb,
2304 9 : NETLINK_CB(cb->skb).portid,
2305 9 : cb->nlh->nlmsg_seq,
2306 : RTM_NEWROUTE, &fri, flags);
2307 9 : if (err < 0)
2308 0 : goto stop;
2309 : }
2310 :
2311 9 : i_fa++;
2312 : }
2313 :
2314 9 : if (filter->dump_exceptions) {
2315 9 : err = fib_dump_info_fnhe(skb, cb, tb->tb_id, fi,
2316 : &i_fa, s_fa, flags);
2317 9 : if (err < 0)
2318 0 : goto stop;
2319 : }
2320 :
2321 9 : next:
2322 18 : i++;
2323 : }
2324 :
2325 14 : cb->args[4] = i;
2326 14 : return skb->len;
2327 :
2328 0 : stop:
2329 0 : cb->args[4] = i;
2330 0 : cb->args[5] = i_fa;
2331 0 : return err;
2332 : }
2333 :
2334 : /* rcu_read_lock needs to be hold by caller from readside */
2335 2 : int fib_table_dump(struct fib_table *tb, struct sk_buff *skb,
2336 : struct netlink_callback *cb, struct fib_dump_filter *filter)
2337 : {
2338 2 : struct trie *t = (struct trie *)tb->tb_data;
2339 2 : struct key_vector *l, *tp = t->kv;
2340 : /* Dump starting at last key.
2341 : * Note: 0.0.0.0/0 (ie default) is first key.
2342 : */
2343 2 : int count = cb->args[2];
2344 2 : t_key key = cb->args[3];
2345 :
2346 : /* First time here, count and key are both always 0. Count > 0
2347 : * and key == 0 means the dump has wrapped around and we are done.
2348 : */
2349 2 : if (count && !key)
2350 0 : return skb->len;
2351 :
2352 16 : while ((l = leaf_walk_rcu(&tp, key)) != NULL) {
2353 14 : int err;
2354 :
2355 14 : err = fn_trie_dump_leaf(l, tb, skb, cb, filter);
2356 14 : if (err < 0) {
2357 0 : cb->args[3] = key;
2358 0 : cb->args[2] = count;
2359 0 : return err;
2360 : }
2361 :
2362 14 : ++count;
2363 14 : key = l->key + 1;
2364 :
2365 14 : memset(&cb->args[4], 0,
2366 : sizeof(cb->args) - 4*sizeof(cb->args[0]));
2367 :
2368 : /* stop loop if key wrapped back to 0 */
2369 14 : if (key < l->key)
2370 : break;
2371 : }
2372 :
2373 2 : cb->args[3] = key;
2374 2 : cb->args[2] = count;
2375 :
2376 2 : return skb->len;
2377 : }
2378 :
2379 1 : void __init fib_trie_init(void)
2380 : {
2381 1 : fn_alias_kmem = kmem_cache_create("ip_fib_alias",
2382 : sizeof(struct fib_alias),
2383 : 0, SLAB_PANIC, NULL);
2384 :
2385 1 : trie_leaf_kmem = kmem_cache_create("ip_fib_trie",
2386 : LEAF_SIZE,
2387 : 0, SLAB_PANIC, NULL);
2388 1 : }
2389 :
2390 2 : struct fib_table *fib_trie_table(u32 id, struct fib_table *alias)
2391 : {
2392 2 : struct fib_table *tb;
2393 2 : struct trie *t;
2394 2 : size_t sz = sizeof(*tb);
2395 :
2396 2 : if (!alias)
2397 1 : sz += sizeof(struct trie);
2398 :
2399 2 : tb = kzalloc(sz, GFP_KERNEL);
2400 2 : if (!tb)
2401 : return NULL;
2402 :
2403 2 : tb->tb_id = id;
2404 2 : tb->tb_num_default = 0;
2405 2 : tb->tb_data = (alias ? alias->__data : tb->__data);
2406 :
2407 2 : if (alias)
2408 : return tb;
2409 :
2410 1 : t = (struct trie *) tb->tb_data;
2411 1 : t->kv[0].pos = KEYLENGTH;
2412 1 : t->kv[0].slen = KEYLENGTH;
2413 : #ifdef CONFIG_IP_FIB_TRIE_STATS
2414 : t->stats = alloc_percpu(struct trie_use_stats);
2415 : if (!t->stats) {
2416 : kfree(tb);
2417 : tb = NULL;
2418 : }
2419 : #endif
2420 :
2421 1 : return tb;
2422 : }
2423 :
2424 : #ifdef CONFIG_PROC_FS
2425 : /* Depth first Trie walk iterator */
2426 : struct fib_trie_iter {
2427 : struct seq_net_private p;
2428 : struct fib_table *tb;
2429 : struct key_vector *tnode;
2430 : unsigned int index;
2431 : unsigned int depth;
2432 : };
2433 :
2434 0 : static struct key_vector *fib_trie_get_next(struct fib_trie_iter *iter)
2435 : {
2436 0 : unsigned long cindex = iter->index;
2437 0 : struct key_vector *pn = iter->tnode;
2438 0 : t_key pkey;
2439 :
2440 0 : pr_debug("get_next iter={node=%p index=%d depth=%d}\n",
2441 : iter->tnode, iter->index, iter->depth);
2442 :
2443 0 : while (!IS_TRIE(pn)) {
2444 0 : while (cindex < child_length(pn)) {
2445 0 : struct key_vector *n = get_child_rcu(pn, cindex++);
2446 :
2447 0 : if (!n)
2448 0 : continue;
2449 :
2450 0 : if (IS_LEAF(n)) {
2451 0 : iter->tnode = pn;
2452 0 : iter->index = cindex;
2453 : } else {
2454 : /* push down one level */
2455 0 : iter->tnode = n;
2456 0 : iter->index = 0;
2457 0 : ++iter->depth;
2458 : }
2459 :
2460 : return n;
2461 : }
2462 :
2463 : /* Current node exhausted, pop back up */
2464 0 : pkey = pn->key;
2465 0 : pn = node_parent_rcu(pn);
2466 0 : cindex = get_index(pkey, pn) + 1;
2467 0 : --iter->depth;
2468 : }
2469 :
2470 : /* record root node so further searches know we are done */
2471 0 : iter->tnode = pn;
2472 0 : iter->index = 0;
2473 :
2474 0 : return NULL;
2475 : }
2476 :
2477 0 : static struct key_vector *fib_trie_get_first(struct fib_trie_iter *iter,
2478 : struct trie *t)
2479 : {
2480 0 : struct key_vector *n, *pn;
2481 :
2482 0 : if (!t)
2483 : return NULL;
2484 :
2485 0 : pn = t->kv;
2486 0 : n = rcu_dereference(pn->tnode[0]);
2487 0 : if (!n)
2488 : return NULL;
2489 :
2490 0 : if (IS_TNODE(n)) {
2491 0 : iter->tnode = n;
2492 0 : iter->index = 0;
2493 0 : iter->depth = 1;
2494 : } else {
2495 0 : iter->tnode = pn;
2496 0 : iter->index = 0;
2497 0 : iter->depth = 0;
2498 : }
2499 :
2500 : return n;
2501 : }
2502 :
2503 0 : static void trie_collect_stats(struct trie *t, struct trie_stat *s)
2504 : {
2505 0 : struct key_vector *n;
2506 0 : struct fib_trie_iter iter;
2507 :
2508 0 : memset(s, 0, sizeof(*s));
2509 :
2510 0 : rcu_read_lock();
2511 0 : for (n = fib_trie_get_first(&iter, t); n; n = fib_trie_get_next(&iter)) {
2512 0 : if (IS_LEAF(n)) {
2513 0 : struct fib_alias *fa;
2514 :
2515 0 : s->leaves++;
2516 0 : s->totdepth += iter.depth;
2517 0 : if (iter.depth > s->maxdepth)
2518 0 : s->maxdepth = iter.depth;
2519 :
2520 0 : hlist_for_each_entry_rcu(fa, &n->leaf, fa_list)
2521 0 : ++s->prefixes;
2522 : } else {
2523 0 : s->tnodes++;
2524 0 : if (n->bits < MAX_STAT_DEPTH)
2525 0 : s->nodesizes[n->bits]++;
2526 0 : s->nullpointers += tn_info(n)->empty_children;
2527 : }
2528 : }
2529 0 : rcu_read_unlock();
2530 0 : }
2531 :
2532 : /*
2533 : * This outputs /proc/net/fib_triestats
2534 : */
2535 0 : static void trie_show_stats(struct seq_file *seq, struct trie_stat *stat)
2536 : {
2537 0 : unsigned int i, max, pointers, bytes, avdepth;
2538 :
2539 0 : if (stat->leaves)
2540 0 : avdepth = stat->totdepth*100 / stat->leaves;
2541 : else
2542 : avdepth = 0;
2543 :
2544 0 : seq_printf(seq, "\tAver depth: %u.%02d\n",
2545 : avdepth / 100, avdepth % 100);
2546 0 : seq_printf(seq, "\tMax depth: %u\n", stat->maxdepth);
2547 :
2548 0 : seq_printf(seq, "\tLeaves: %u\n", stat->leaves);
2549 0 : bytes = LEAF_SIZE * stat->leaves;
2550 :
2551 0 : seq_printf(seq, "\tPrefixes: %u\n", stat->prefixes);
2552 0 : bytes += sizeof(struct fib_alias) * stat->prefixes;
2553 :
2554 0 : seq_printf(seq, "\tInternal nodes: %u\n\t", stat->tnodes);
2555 0 : bytes += TNODE_SIZE(0) * stat->tnodes;
2556 :
2557 0 : max = MAX_STAT_DEPTH;
2558 0 : while (max > 0 && stat->nodesizes[max-1] == 0)
2559 : max--;
2560 :
2561 : pointers = 0;
2562 0 : for (i = 1; i < max; i++)
2563 0 : if (stat->nodesizes[i] != 0) {
2564 0 : seq_printf(seq, " %u: %u", i, stat->nodesizes[i]);
2565 0 : pointers += (1<<i) * stat->nodesizes[i];
2566 : }
2567 0 : seq_putc(seq, '\n');
2568 0 : seq_printf(seq, "\tPointers: %u\n", pointers);
2569 :
2570 0 : bytes += sizeof(struct key_vector *) * pointers;
2571 0 : seq_printf(seq, "Null ptrs: %u\n", stat->nullpointers);
2572 0 : seq_printf(seq, "Total size: %u kB\n", (bytes + 1023) / 1024);
2573 0 : }
2574 :
2575 : #ifdef CONFIG_IP_FIB_TRIE_STATS
2576 : static void trie_show_usage(struct seq_file *seq,
2577 : const struct trie_use_stats __percpu *stats)
2578 : {
2579 : struct trie_use_stats s = { 0 };
2580 : int cpu;
2581 :
2582 : /* loop through all of the CPUs and gather up the stats */
2583 : for_each_possible_cpu(cpu) {
2584 : const struct trie_use_stats *pcpu = per_cpu_ptr(stats, cpu);
2585 :
2586 : s.gets += pcpu->gets;
2587 : s.backtrack += pcpu->backtrack;
2588 : s.semantic_match_passed += pcpu->semantic_match_passed;
2589 : s.semantic_match_miss += pcpu->semantic_match_miss;
2590 : s.null_node_hit += pcpu->null_node_hit;
2591 : s.resize_node_skipped += pcpu->resize_node_skipped;
2592 : }
2593 :
2594 : seq_printf(seq, "\nCounters:\n---------\n");
2595 : seq_printf(seq, "gets = %u\n", s.gets);
2596 : seq_printf(seq, "backtracks = %u\n", s.backtrack);
2597 : seq_printf(seq, "semantic match passed = %u\n",
2598 : s.semantic_match_passed);
2599 : seq_printf(seq, "semantic match miss = %u\n", s.semantic_match_miss);
2600 : seq_printf(seq, "null node hit= %u\n", s.null_node_hit);
2601 : seq_printf(seq, "skipped node resize = %u\n\n", s.resize_node_skipped);
2602 : }
2603 : #endif /* CONFIG_IP_FIB_TRIE_STATS */
2604 :
2605 0 : static void fib_table_print(struct seq_file *seq, struct fib_table *tb)
2606 : {
2607 0 : if (tb->tb_id == RT_TABLE_LOCAL)
2608 0 : seq_puts(seq, "Local:\n");
2609 0 : else if (tb->tb_id == RT_TABLE_MAIN)
2610 0 : seq_puts(seq, "Main:\n");
2611 : else
2612 0 : seq_printf(seq, "Id %d:\n", tb->tb_id);
2613 0 : }
2614 :
2615 :
2616 0 : static int fib_triestat_seq_show(struct seq_file *seq, void *v)
2617 : {
2618 0 : struct net *net = (struct net *)seq->private;
2619 0 : unsigned int h;
2620 :
2621 0 : seq_printf(seq,
2622 : "Basic info: size of leaf:"
2623 : " %zd bytes, size of tnode: %zd bytes.\n",
2624 : LEAF_SIZE, TNODE_SIZE(0));
2625 :
2626 0 : rcu_read_lock();
2627 0 : for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2628 0 : struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2629 0 : struct fib_table *tb;
2630 :
2631 0 : hlist_for_each_entry_rcu(tb, head, tb_hlist) {
2632 0 : struct trie *t = (struct trie *) tb->tb_data;
2633 0 : struct trie_stat stat;
2634 :
2635 0 : if (!t)
2636 0 : continue;
2637 :
2638 0 : fib_table_print(seq, tb);
2639 :
2640 0 : trie_collect_stats(t, &stat);
2641 0 : trie_show_stats(seq, &stat);
2642 : #ifdef CONFIG_IP_FIB_TRIE_STATS
2643 : trie_show_usage(seq, t->stats);
2644 : #endif
2645 : }
2646 0 : cond_resched_rcu();
2647 : }
2648 0 : rcu_read_unlock();
2649 :
2650 0 : return 0;
2651 : }
2652 :
2653 0 : static struct key_vector *fib_trie_get_idx(struct seq_file *seq, loff_t pos)
2654 : {
2655 0 : struct fib_trie_iter *iter = seq->private;
2656 0 : struct net *net = seq_file_net(seq);
2657 0 : loff_t idx = 0;
2658 0 : unsigned int h;
2659 :
2660 0 : for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
2661 0 : struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2662 0 : struct fib_table *tb;
2663 :
2664 0 : hlist_for_each_entry_rcu(tb, head, tb_hlist) {
2665 0 : struct key_vector *n;
2666 :
2667 0 : for (n = fib_trie_get_first(iter,
2668 0 : (struct trie *) tb->tb_data);
2669 0 : n; n = fib_trie_get_next(iter))
2670 0 : if (pos == idx++) {
2671 0 : iter->tb = tb;
2672 0 : return n;
2673 : }
2674 : }
2675 : }
2676 :
2677 : return NULL;
2678 : }
2679 :
2680 0 : static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos)
2681 : __acquires(RCU)
2682 : {
2683 0 : rcu_read_lock();
2684 0 : return fib_trie_get_idx(seq, *pos);
2685 : }
2686 :
2687 0 : static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos)
2688 : {
2689 0 : struct fib_trie_iter *iter = seq->private;
2690 0 : struct net *net = seq_file_net(seq);
2691 0 : struct fib_table *tb = iter->tb;
2692 0 : struct hlist_node *tb_node;
2693 0 : unsigned int h;
2694 0 : struct key_vector *n;
2695 :
2696 0 : ++*pos;
2697 : /* next node in same table */
2698 0 : n = fib_trie_get_next(iter);
2699 0 : if (n)
2700 : return n;
2701 :
2702 : /* walk rest of this hash chain */
2703 0 : h = tb->tb_id & (FIB_TABLE_HASHSZ - 1);
2704 0 : while ((tb_node = rcu_dereference(hlist_next_rcu(&tb->tb_hlist)))) {
2705 0 : tb = hlist_entry(tb_node, struct fib_table, tb_hlist);
2706 0 : n = fib_trie_get_first(iter, (struct trie *) tb->tb_data);
2707 0 : if (n)
2708 0 : goto found;
2709 : }
2710 :
2711 : /* new hash chain */
2712 0 : while (++h < FIB_TABLE_HASHSZ) {
2713 0 : struct hlist_head *head = &net->ipv4.fib_table_hash[h];
2714 0 : hlist_for_each_entry_rcu(tb, head, tb_hlist) {
2715 0 : n = fib_trie_get_first(iter, (struct trie *) tb->tb_data);
2716 0 : if (n)
2717 0 : goto found;
2718 : }
2719 : }
2720 : return NULL;
2721 :
2722 0 : found:
2723 0 : iter->tb = tb;
2724 0 : return n;
2725 : }
2726 :
2727 0 : static void fib_trie_seq_stop(struct seq_file *seq, void *v)
2728 : __releases(RCU)
2729 : {
2730 0 : rcu_read_unlock();
2731 0 : }
2732 :
2733 0 : static void seq_indent(struct seq_file *seq, int n)
2734 : {
2735 0 : while (n-- > 0)
2736 0 : seq_puts(seq, " ");
2737 : }
2738 :
2739 0 : static inline const char *rtn_scope(char *buf, size_t len, enum rt_scope_t s)
2740 : {
2741 0 : switch (s) {
2742 : case RT_SCOPE_UNIVERSE: return "universe";
2743 0 : case RT_SCOPE_SITE: return "site";
2744 0 : case RT_SCOPE_LINK: return "link";
2745 0 : case RT_SCOPE_HOST: return "host";
2746 0 : case RT_SCOPE_NOWHERE: return "nowhere";
2747 0 : default:
2748 0 : snprintf(buf, len, "scope=%d", s);
2749 0 : return buf;
2750 : }
2751 : }
2752 :
2753 : static const char *const rtn_type_names[__RTN_MAX] = {
2754 : [RTN_UNSPEC] = "UNSPEC",
2755 : [RTN_UNICAST] = "UNICAST",
2756 : [RTN_LOCAL] = "LOCAL",
2757 : [RTN_BROADCAST] = "BROADCAST",
2758 : [RTN_ANYCAST] = "ANYCAST",
2759 : [RTN_MULTICAST] = "MULTICAST",
2760 : [RTN_BLACKHOLE] = "BLACKHOLE",
2761 : [RTN_UNREACHABLE] = "UNREACHABLE",
2762 : [RTN_PROHIBIT] = "PROHIBIT",
2763 : [RTN_THROW] = "THROW",
2764 : [RTN_NAT] = "NAT",
2765 : [RTN_XRESOLVE] = "XRESOLVE",
2766 : };
2767 :
2768 0 : static inline const char *rtn_type(char *buf, size_t len, unsigned int t)
2769 : {
2770 0 : if (t < __RTN_MAX && rtn_type_names[t])
2771 : return rtn_type_names[t];
2772 0 : snprintf(buf, len, "type %u", t);
2773 0 : return buf;
2774 : }
2775 :
2776 : /* Pretty print the trie */
2777 0 : static int fib_trie_seq_show(struct seq_file *seq, void *v)
2778 : {
2779 0 : const struct fib_trie_iter *iter = seq->private;
2780 0 : struct key_vector *n = v;
2781 :
2782 0 : if (IS_TRIE(node_parent_rcu(n)))
2783 0 : fib_table_print(seq, iter->tb);
2784 :
2785 0 : if (IS_TNODE(n)) {
2786 0 : __be32 prf = htonl(n->key);
2787 :
2788 0 : seq_indent(seq, iter->depth-1);
2789 0 : seq_printf(seq, " +-- %pI4/%zu %u %u %u\n",
2790 0 : &prf, KEYLENGTH - n->pos - n->bits, n->bits,
2791 0 : tn_info(n)->full_children,
2792 0 : tn_info(n)->empty_children);
2793 : } else {
2794 0 : __be32 val = htonl(n->key);
2795 0 : struct fib_alias *fa;
2796 :
2797 0 : seq_indent(seq, iter->depth);
2798 0 : seq_printf(seq, " |-- %pI4\n", &val);
2799 :
2800 0 : hlist_for_each_entry_rcu(fa, &n->leaf, fa_list) {
2801 0 : char buf1[32], buf2[32];
2802 :
2803 0 : seq_indent(seq, iter->depth + 1);
2804 0 : seq_printf(seq, " /%zu %s %s",
2805 0 : KEYLENGTH - fa->fa_slen,
2806 : rtn_scope(buf1, sizeof(buf1),
2807 0 : fa->fa_info->fib_scope),
2808 : rtn_type(buf2, sizeof(buf2),
2809 0 : fa->fa_type));
2810 0 : if (fa->fa_tos)
2811 0 : seq_printf(seq, " tos=%d", fa->fa_tos);
2812 0 : seq_putc(seq, '\n');
2813 : }
2814 : }
2815 :
2816 0 : return 0;
2817 : }
2818 :
2819 : static const struct seq_operations fib_trie_seq_ops = {
2820 : .start = fib_trie_seq_start,
2821 : .next = fib_trie_seq_next,
2822 : .stop = fib_trie_seq_stop,
2823 : .show = fib_trie_seq_show,
2824 : };
2825 :
2826 : struct fib_route_iter {
2827 : struct seq_net_private p;
2828 : struct fib_table *main_tb;
2829 : struct key_vector *tnode;
2830 : loff_t pos;
2831 : t_key key;
2832 : };
2833 :
2834 0 : static struct key_vector *fib_route_get_idx(struct fib_route_iter *iter,
2835 : loff_t pos)
2836 : {
2837 0 : struct key_vector *l, **tp = &iter->tnode;
2838 0 : t_key key;
2839 :
2840 : /* use cached location of previously found key */
2841 0 : if (iter->pos > 0 && pos >= iter->pos) {
2842 0 : key = iter->key;
2843 : } else {
2844 0 : iter->pos = 1;
2845 0 : key = 0;
2846 : }
2847 :
2848 0 : pos -= iter->pos;
2849 :
2850 0 : while ((l = leaf_walk_rcu(tp, key)) && (pos-- > 0)) {
2851 0 : key = l->key + 1;
2852 0 : iter->pos++;
2853 0 : l = NULL;
2854 :
2855 : /* handle unlikely case of a key wrap */
2856 0 : if (!key)
2857 : break;
2858 : }
2859 :
2860 0 : if (l)
2861 0 : iter->key = l->key; /* remember it */
2862 : else
2863 0 : iter->pos = 0; /* forget it */
2864 :
2865 0 : return l;
2866 : }
2867 :
2868 0 : static void *fib_route_seq_start(struct seq_file *seq, loff_t *pos)
2869 : __acquires(RCU)
2870 : {
2871 0 : struct fib_route_iter *iter = seq->private;
2872 0 : struct fib_table *tb;
2873 0 : struct trie *t;
2874 :
2875 0 : rcu_read_lock();
2876 :
2877 0 : tb = fib_get_table(seq_file_net(seq), RT_TABLE_MAIN);
2878 0 : if (!tb)
2879 : return NULL;
2880 :
2881 0 : iter->main_tb = tb;
2882 0 : t = (struct trie *)tb->tb_data;
2883 0 : iter->tnode = t->kv;
2884 :
2885 0 : if (*pos != 0)
2886 0 : return fib_route_get_idx(iter, *pos);
2887 :
2888 0 : iter->pos = 0;
2889 0 : iter->key = KEY_MAX;
2890 :
2891 0 : return SEQ_START_TOKEN;
2892 : }
2893 :
2894 0 : static void *fib_route_seq_next(struct seq_file *seq, void *v, loff_t *pos)
2895 : {
2896 0 : struct fib_route_iter *iter = seq->private;
2897 0 : struct key_vector *l = NULL;
2898 0 : t_key key = iter->key + 1;
2899 :
2900 0 : ++*pos;
2901 :
2902 : /* only allow key of 0 for start of sequence */
2903 0 : if ((v == SEQ_START_TOKEN) || key)
2904 0 : l = leaf_walk_rcu(&iter->tnode, key);
2905 :
2906 0 : if (l) {
2907 0 : iter->key = l->key;
2908 0 : iter->pos++;
2909 : } else {
2910 0 : iter->pos = 0;
2911 : }
2912 :
2913 0 : return l;
2914 : }
2915 :
2916 0 : static void fib_route_seq_stop(struct seq_file *seq, void *v)
2917 : __releases(RCU)
2918 : {
2919 0 : rcu_read_unlock();
2920 0 : }
2921 :
2922 0 : static unsigned int fib_flag_trans(int type, __be32 mask, struct fib_info *fi)
2923 : {
2924 0 : unsigned int flags = 0;
2925 :
2926 0 : if (type == RTN_UNREACHABLE || type == RTN_PROHIBIT)
2927 0 : flags = RTF_REJECT;
2928 0 : if (fi) {
2929 0 : const struct fib_nh_common *nhc = fib_info_nhc(fi, 0);
2930 :
2931 0 : if (nhc->nhc_gw.ipv4)
2932 0 : flags |= RTF_GATEWAY;
2933 : }
2934 0 : if (mask == htonl(0xFFFFFFFF))
2935 0 : flags |= RTF_HOST;
2936 0 : flags |= RTF_UP;
2937 0 : return flags;
2938 : }
2939 :
2940 : /*
2941 : * This outputs /proc/net/route.
2942 : * The format of the file is not supposed to be changed
2943 : * and needs to be same as fib_hash output to avoid breaking
2944 : * legacy utilities
2945 : */
2946 0 : static int fib_route_seq_show(struct seq_file *seq, void *v)
2947 : {
2948 0 : struct fib_route_iter *iter = seq->private;
2949 0 : struct fib_table *tb = iter->main_tb;
2950 0 : struct fib_alias *fa;
2951 0 : struct key_vector *l = v;
2952 0 : __be32 prefix;
2953 :
2954 0 : if (v == SEQ_START_TOKEN) {
2955 0 : seq_printf(seq, "%-127s\n", "Iface\tDestination\tGateway "
2956 : "\tFlags\tRefCnt\tUse\tMetric\tMask\t\tMTU"
2957 : "\tWindow\tIRTT");
2958 0 : return 0;
2959 : }
2960 :
2961 0 : prefix = htonl(l->key);
2962 :
2963 0 : hlist_for_each_entry_rcu(fa, &l->leaf, fa_list) {
2964 0 : struct fib_info *fi = fa->fa_info;
2965 0 : __be32 mask = inet_make_mask(KEYLENGTH - fa->fa_slen);
2966 0 : unsigned int flags = fib_flag_trans(fa->fa_type, mask, fi);
2967 :
2968 0 : if ((fa->fa_type == RTN_BROADCAST) ||
2969 : (fa->fa_type == RTN_MULTICAST))
2970 0 : continue;
2971 :
2972 0 : if (fa->tb_id != tb->tb_id)
2973 0 : continue;
2974 :
2975 0 : seq_setwidth(seq, 127);
2976 :
2977 0 : if (fi) {
2978 0 : struct fib_nh_common *nhc = fib_info_nhc(fi, 0);
2979 0 : __be32 gw = 0;
2980 :
2981 0 : if (nhc->nhc_gw_family == AF_INET)
2982 0 : gw = nhc->nhc_gw.ipv4;
2983 :
2984 0 : seq_printf(seq,
2985 : "%s\t%08X\t%08X\t%04X\t%d\t%u\t"
2986 : "%d\t%08X\t%d\t%u\t%u",
2987 0 : nhc->nhc_dev ? nhc->nhc_dev->name : "*",
2988 : prefix, gw, flags, 0, 0,
2989 : fi->fib_priority,
2990 : mask,
2991 0 : (fi->fib_advmss ?
2992 : fi->fib_advmss + 40 : 0),
2993 : fi->fib_window,
2994 0 : fi->fib_rtt >> 3);
2995 : } else {
2996 0 : seq_printf(seq,
2997 : "*\t%08X\t%08X\t%04X\t%d\t%u\t"
2998 : "%d\t%08X\t%d\t%u\t%u",
2999 : prefix, 0, flags, 0, 0, 0,
3000 : mask, 0, 0, 0);
3001 : }
3002 0 : seq_pad(seq, '\n');
3003 : }
3004 :
3005 : return 0;
3006 : }
3007 :
3008 : static const struct seq_operations fib_route_seq_ops = {
3009 : .start = fib_route_seq_start,
3010 : .next = fib_route_seq_next,
3011 : .stop = fib_route_seq_stop,
3012 : .show = fib_route_seq_show,
3013 : };
3014 :
3015 1 : int __net_init fib_proc_init(struct net *net)
3016 : {
3017 1 : if (!proc_create_net("fib_trie", 0444, net->proc_net, &fib_trie_seq_ops,
3018 : sizeof(struct fib_trie_iter)))
3019 0 : goto out1;
3020 :
3021 1 : if (!proc_create_net_single("fib_triestat", 0444, net->proc_net,
3022 : fib_triestat_seq_show, NULL))
3023 0 : goto out2;
3024 :
3025 1 : if (!proc_create_net("route", 0444, net->proc_net, &fib_route_seq_ops,
3026 : sizeof(struct fib_route_iter)))
3027 0 : goto out3;
3028 :
3029 : return 0;
3030 :
3031 0 : out3:
3032 0 : remove_proc_entry("fib_triestat", net->proc_net);
3033 0 : out2:
3034 0 : remove_proc_entry("fib_trie", net->proc_net);
3035 : out1:
3036 : return -ENOMEM;
3037 : }
3038 :
3039 0 : void __net_exit fib_proc_exit(struct net *net)
3040 : {
3041 0 : remove_proc_entry("fib_trie", net->proc_net);
3042 0 : remove_proc_entry("fib_triestat", net->proc_net);
3043 0 : remove_proc_entry("route", net->proc_net);
3044 0 : }
3045 :
3046 : #endif /* CONFIG_PROC_FS */
|
