xref: /openbmc/linux/kernel/bpf/lpm_trie.c (revision 232b0b08)
1 /*
2  * Longest prefix match list implementation
3  *
4  * Copyright (c) 2016,2017 Daniel Mack
5  * Copyright (c) 2016 David Herrmann
6  *
7  * This file is subject to the terms and conditions of version 2 of the GNU
8  * General Public License.  See the file COPYING in the main directory of the
9  * Linux distribution for more details.
10  */
11 
12 #include <linux/bpf.h>
13 #include <linux/err.h>
14 #include <linux/slab.h>
15 #include <linux/spinlock.h>
16 #include <linux/vmalloc.h>
17 #include <net/ipv6.h>
18 
19 /* Intermediate node */
20 #define LPM_TREE_NODE_FLAG_IM BIT(0)
21 
22 struct lpm_trie_node;
23 
24 struct lpm_trie_node {
25 	struct rcu_head rcu;
26 	struct lpm_trie_node __rcu	*child[2];
27 	u32				prefixlen;
28 	u32				flags;
29 	u8				data[0];
30 };
31 
32 struct lpm_trie {
33 	struct bpf_map			map;
34 	struct lpm_trie_node __rcu	*root;
35 	size_t				n_entries;
36 	size_t				max_prefixlen;
37 	size_t				data_size;
38 	raw_spinlock_t			lock;
39 };
40 
41 /* This trie implements a longest prefix match algorithm that can be used to
42  * match IP addresses to a stored set of ranges.
43  *
44  * Data stored in @data of struct bpf_lpm_key and struct lpm_trie_node is
45  * interpreted as big endian, so data[0] stores the most significant byte.
46  *
47  * Match ranges are internally stored in instances of struct lpm_trie_node
48  * which each contain their prefix length as well as two pointers that may
49  * lead to more nodes containing more specific matches. Each node also stores
50  * a value that is defined by and returned to userspace via the update_elem
51  * and lookup functions.
52  *
53  * For instance, let's start with a trie that was created with a prefix length
54  * of 32, so it can be used for IPv4 addresses, and one single element that
55  * matches 192.168.0.0/16. The data array would hence contain
56  * [0xc0, 0xa8, 0x00, 0x00] in big-endian notation. This documentation will
57  * stick to IP-address notation for readability though.
58  *
59  * As the trie is empty initially, the new node (1) will be places as root
60  * node, denoted as (R) in the example below. As there are no other node, both
61  * child pointers are %NULL.
62  *
63  *              +----------------+
64  *              |       (1)  (R) |
65  *              | 192.168.0.0/16 |
66  *              |    value: 1    |
67  *              |   [0]    [1]   |
68  *              +----------------+
69  *
70  * Next, let's add a new node (2) matching 192.168.0.0/24. As there is already
71  * a node with the same data and a smaller prefix (ie, a less specific one),
72  * node (2) will become a child of (1). In child index depends on the next bit
73  * that is outside of what (1) matches, and that bit is 0, so (2) will be
74  * child[0] of (1):
75  *
76  *              +----------------+
77  *              |       (1)  (R) |
78  *              | 192.168.0.0/16 |
79  *              |    value: 1    |
80  *              |   [0]    [1]   |
81  *              +----------------+
82  *                   |
83  *    +----------------+
84  *    |       (2)      |
85  *    | 192.168.0.0/24 |
86  *    |    value: 2    |
87  *    |   [0]    [1]   |
88  *    +----------------+
89  *
90  * The child[1] slot of (1) could be filled with another node which has bit #17
91  * (the next bit after the ones that (1) matches on) set to 1. For instance,
92  * 192.168.128.0/24:
93  *
94  *              +----------------+
95  *              |       (1)  (R) |
96  *              | 192.168.0.0/16 |
97  *              |    value: 1    |
98  *              |   [0]    [1]   |
99  *              +----------------+
100  *                   |      |
101  *    +----------------+  +------------------+
102  *    |       (2)      |  |        (3)       |
103  *    | 192.168.0.0/24 |  | 192.168.128.0/24 |
104  *    |    value: 2    |  |     value: 3     |
105  *    |   [0]    [1]   |  |    [0]    [1]    |
106  *    +----------------+  +------------------+
107  *
108  * Let's add another node (4) to the game for 192.168.1.0/24. In order to place
109  * it, node (1) is looked at first, and because (4) of the semantics laid out
110  * above (bit #17 is 0), it would normally be attached to (1) as child[0].
111  * However, that slot is already allocated, so a new node is needed in between.
112  * That node does not have a value attached to it and it will never be
113  * returned to users as result of a lookup. It is only there to differentiate
114  * the traversal further. It will get a prefix as wide as necessary to
115  * distinguish its two children:
116  *
117  *                      +----------------+
118  *                      |       (1)  (R) |
119  *                      | 192.168.0.0/16 |
120  *                      |    value: 1    |
121  *                      |   [0]    [1]   |
122  *                      +----------------+
123  *                           |      |
124  *            +----------------+  +------------------+
125  *            |       (4)  (I) |  |        (3)       |
126  *            | 192.168.0.0/23 |  | 192.168.128.0/24 |
127  *            |    value: ---  |  |     value: 3     |
128  *            |   [0]    [1]   |  |    [0]    [1]    |
129  *            +----------------+  +------------------+
130  *                 |      |
131  *  +----------------+  +----------------+
132  *  |       (2)      |  |       (5)      |
133  *  | 192.168.0.0/24 |  | 192.168.1.0/24 |
134  *  |    value: 2    |  |     value: 5   |
135  *  |   [0]    [1]   |  |   [0]    [1]   |
136  *  +----------------+  +----------------+
137  *
138  * 192.168.1.1/32 would be a child of (5) etc.
139  *
140  * An intermediate node will be turned into a 'real' node on demand. In the
141  * example above, (4) would be re-used if 192.168.0.0/23 is added to the trie.
142  *
143  * A fully populated trie would have a height of 32 nodes, as the trie was
144  * created with a prefix length of 32.
145  *
146  * The lookup starts at the root node. If the current node matches and if there
147  * is a child that can be used to become more specific, the trie is traversed
148  * downwards. The last node in the traversal that is a non-intermediate one is
149  * returned.
150  */
151 
152 static inline int extract_bit(const u8 *data, size_t index)
153 {
154 	return !!(data[index / 8] & (1 << (7 - (index % 8))));
155 }
156 
157 /**
158  * longest_prefix_match() - determine the longest prefix
159  * @trie:	The trie to get internal sizes from
160  * @node:	The node to operate on
161  * @key:	The key to compare to @node
162  *
163  * Determine the longest prefix of @node that matches the bits in @key.
164  */
165 static size_t longest_prefix_match(const struct lpm_trie *trie,
166 				   const struct lpm_trie_node *node,
167 				   const struct bpf_lpm_trie_key *key)
168 {
169 	size_t prefixlen = 0;
170 	size_t i;
171 
172 	for (i = 0; i < trie->data_size; i++) {
173 		size_t b;
174 
175 		b = 8 - fls(node->data[i] ^ key->data[i]);
176 		prefixlen += b;
177 
178 		if (prefixlen >= node->prefixlen || prefixlen >= key->prefixlen)
179 			return min(node->prefixlen, key->prefixlen);
180 
181 		if (b < 8)
182 			break;
183 	}
184 
185 	return prefixlen;
186 }
187 
188 /* Called from syscall or from eBPF program */
189 static void *trie_lookup_elem(struct bpf_map *map, void *_key)
190 {
191 	struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
192 	struct lpm_trie_node *node, *found = NULL;
193 	struct bpf_lpm_trie_key *key = _key;
194 
195 	/* Start walking the trie from the root node ... */
196 
197 	for (node = rcu_dereference(trie->root); node;) {
198 		unsigned int next_bit;
199 		size_t matchlen;
200 
201 		/* Determine the longest prefix of @node that matches @key.
202 		 * If it's the maximum possible prefix for this trie, we have
203 		 * an exact match and can return it directly.
204 		 */
205 		matchlen = longest_prefix_match(trie, node, key);
206 		if (matchlen == trie->max_prefixlen) {
207 			found = node;
208 			break;
209 		}
210 
211 		/* If the number of bits that match is smaller than the prefix
212 		 * length of @node, bail out and return the node we have seen
213 		 * last in the traversal (ie, the parent).
214 		 */
215 		if (matchlen < node->prefixlen)
216 			break;
217 
218 		/* Consider this node as return candidate unless it is an
219 		 * artificially added intermediate one.
220 		 */
221 		if (!(node->flags & LPM_TREE_NODE_FLAG_IM))
222 			found = node;
223 
224 		/* If the node match is fully satisfied, let's see if we can
225 		 * become more specific. Determine the next bit in the key and
226 		 * traverse down.
227 		 */
228 		next_bit = extract_bit(key->data, node->prefixlen);
229 		node = rcu_dereference(node->child[next_bit]);
230 	}
231 
232 	if (!found)
233 		return NULL;
234 
235 	return found->data + trie->data_size;
236 }
237 
238 static struct lpm_trie_node *lpm_trie_node_alloc(const struct lpm_trie *trie,
239 						 const void *value)
240 {
241 	struct lpm_trie_node *node;
242 	size_t size = sizeof(struct lpm_trie_node) + trie->data_size;
243 
244 	if (value)
245 		size += trie->map.value_size;
246 
247 	node = kmalloc(size, GFP_ATOMIC | __GFP_NOWARN);
248 	if (!node)
249 		return NULL;
250 
251 	node->flags = 0;
252 
253 	if (value)
254 		memcpy(node->data + trie->data_size, value,
255 		       trie->map.value_size);
256 
257 	return node;
258 }
259 
260 /* Called from syscall or from eBPF program */
261 static int trie_update_elem(struct bpf_map *map,
262 			    void *_key, void *value, u64 flags)
263 {
264 	struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
265 	struct lpm_trie_node *node, *im_node = NULL, *new_node = NULL;
266 	struct lpm_trie_node __rcu **slot;
267 	struct bpf_lpm_trie_key *key = _key;
268 	unsigned long irq_flags;
269 	unsigned int next_bit;
270 	size_t matchlen = 0;
271 	int ret = 0;
272 
273 	if (unlikely(flags > BPF_EXIST))
274 		return -EINVAL;
275 
276 	if (key->prefixlen > trie->max_prefixlen)
277 		return -EINVAL;
278 
279 	raw_spin_lock_irqsave(&trie->lock, irq_flags);
280 
281 	/* Allocate and fill a new node */
282 
283 	if (trie->n_entries == trie->map.max_entries) {
284 		ret = -ENOSPC;
285 		goto out;
286 	}
287 
288 	new_node = lpm_trie_node_alloc(trie, value);
289 	if (!new_node) {
290 		ret = -ENOMEM;
291 		goto out;
292 	}
293 
294 	trie->n_entries++;
295 
296 	new_node->prefixlen = key->prefixlen;
297 	RCU_INIT_POINTER(new_node->child[0], NULL);
298 	RCU_INIT_POINTER(new_node->child[1], NULL);
299 	memcpy(new_node->data, key->data, trie->data_size);
300 
301 	/* Now find a slot to attach the new node. To do that, walk the tree
302 	 * from the root and match as many bits as possible for each node until
303 	 * we either find an empty slot or a slot that needs to be replaced by
304 	 * an intermediate node.
305 	 */
306 	slot = &trie->root;
307 
308 	while ((node = rcu_dereference_protected(*slot,
309 					lockdep_is_held(&trie->lock)))) {
310 		matchlen = longest_prefix_match(trie, node, key);
311 
312 		if (node->prefixlen != matchlen ||
313 		    node->prefixlen == key->prefixlen ||
314 		    node->prefixlen == trie->max_prefixlen)
315 			break;
316 
317 		next_bit = extract_bit(key->data, node->prefixlen);
318 		slot = &node->child[next_bit];
319 	}
320 
321 	/* If the slot is empty (a free child pointer or an empty root),
322 	 * simply assign the @new_node to that slot and be done.
323 	 */
324 	if (!node) {
325 		rcu_assign_pointer(*slot, new_node);
326 		goto out;
327 	}
328 
329 	/* If the slot we picked already exists, replace it with @new_node
330 	 * which already has the correct data array set.
331 	 */
332 	if (node->prefixlen == matchlen) {
333 		new_node->child[0] = node->child[0];
334 		new_node->child[1] = node->child[1];
335 
336 		if (!(node->flags & LPM_TREE_NODE_FLAG_IM))
337 			trie->n_entries--;
338 
339 		rcu_assign_pointer(*slot, new_node);
340 		kfree_rcu(node, rcu);
341 
342 		goto out;
343 	}
344 
345 	/* If the new node matches the prefix completely, it must be inserted
346 	 * as an ancestor. Simply insert it between @node and *@slot.
347 	 */
348 	if (matchlen == key->prefixlen) {
349 		next_bit = extract_bit(node->data, matchlen);
350 		rcu_assign_pointer(new_node->child[next_bit], node);
351 		rcu_assign_pointer(*slot, new_node);
352 		goto out;
353 	}
354 
355 	im_node = lpm_trie_node_alloc(trie, NULL);
356 	if (!im_node) {
357 		ret = -ENOMEM;
358 		goto out;
359 	}
360 
361 	im_node->prefixlen = matchlen;
362 	im_node->flags |= LPM_TREE_NODE_FLAG_IM;
363 	memcpy(im_node->data, node->data, trie->data_size);
364 
365 	/* Now determine which child to install in which slot */
366 	if (extract_bit(key->data, matchlen)) {
367 		rcu_assign_pointer(im_node->child[0], node);
368 		rcu_assign_pointer(im_node->child[1], new_node);
369 	} else {
370 		rcu_assign_pointer(im_node->child[0], new_node);
371 		rcu_assign_pointer(im_node->child[1], node);
372 	}
373 
374 	/* Finally, assign the intermediate node to the determined spot */
375 	rcu_assign_pointer(*slot, im_node);
376 
377 out:
378 	if (ret) {
379 		if (new_node)
380 			trie->n_entries--;
381 
382 		kfree(new_node);
383 		kfree(im_node);
384 	}
385 
386 	raw_spin_unlock_irqrestore(&trie->lock, irq_flags);
387 
388 	return ret;
389 }
390 
391 static int trie_delete_elem(struct bpf_map *map, void *key)
392 {
393 	/* TODO */
394 	return -ENOSYS;
395 }
396 
397 #define LPM_DATA_SIZE_MAX	256
398 #define LPM_DATA_SIZE_MIN	1
399 
400 #define LPM_VAL_SIZE_MAX	(KMALLOC_MAX_SIZE - LPM_DATA_SIZE_MAX - \
401 				 sizeof(struct lpm_trie_node))
402 #define LPM_VAL_SIZE_MIN	1
403 
404 #define LPM_KEY_SIZE(X)		(sizeof(struct bpf_lpm_trie_key) + (X))
405 #define LPM_KEY_SIZE_MAX	LPM_KEY_SIZE(LPM_DATA_SIZE_MAX)
406 #define LPM_KEY_SIZE_MIN	LPM_KEY_SIZE(LPM_DATA_SIZE_MIN)
407 
408 static struct bpf_map *trie_alloc(union bpf_attr *attr)
409 {
410 	struct lpm_trie *trie;
411 	u64 cost = sizeof(*trie), cost_per_node;
412 	int ret;
413 
414 	if (!capable(CAP_SYS_ADMIN))
415 		return ERR_PTR(-EPERM);
416 
417 	/* check sanity of attributes */
418 	if (attr->max_entries == 0 ||
419 	    attr->map_flags != BPF_F_NO_PREALLOC ||
420 	    attr->key_size < LPM_KEY_SIZE_MIN ||
421 	    attr->key_size > LPM_KEY_SIZE_MAX ||
422 	    attr->value_size < LPM_VAL_SIZE_MIN ||
423 	    attr->value_size > LPM_VAL_SIZE_MAX)
424 		return ERR_PTR(-EINVAL);
425 
426 	trie = kzalloc(sizeof(*trie), GFP_USER | __GFP_NOWARN);
427 	if (!trie)
428 		return ERR_PTR(-ENOMEM);
429 
430 	/* copy mandatory map attributes */
431 	trie->map.map_type = attr->map_type;
432 	trie->map.key_size = attr->key_size;
433 	trie->map.value_size = attr->value_size;
434 	trie->map.max_entries = attr->max_entries;
435 	trie->data_size = attr->key_size -
436 			  offsetof(struct bpf_lpm_trie_key, data);
437 	trie->max_prefixlen = trie->data_size * 8;
438 
439 	cost_per_node = sizeof(struct lpm_trie_node) +
440 			attr->value_size + trie->data_size;
441 	cost += (u64) attr->max_entries * cost_per_node;
442 	if (cost >= U32_MAX - PAGE_SIZE) {
443 		ret = -E2BIG;
444 		goto out_err;
445 	}
446 
447 	trie->map.pages = round_up(cost, PAGE_SIZE) >> PAGE_SHIFT;
448 
449 	ret = bpf_map_precharge_memlock(trie->map.pages);
450 	if (ret)
451 		goto out_err;
452 
453 	raw_spin_lock_init(&trie->lock);
454 
455 	return &trie->map;
456 out_err:
457 	kfree(trie);
458 	return ERR_PTR(ret);
459 }
460 
461 static void trie_free(struct bpf_map *map)
462 {
463 	struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
464 	struct lpm_trie_node __rcu **slot;
465 	struct lpm_trie_node *node;
466 
467 	raw_spin_lock(&trie->lock);
468 
469 	/* Always start at the root and walk down to a node that has no
470 	 * children. Then free that node, nullify its reference in the parent
471 	 * and start over.
472 	 */
473 
474 	for (;;) {
475 		slot = &trie->root;
476 
477 		for (;;) {
478 			node = rcu_dereference_protected(*slot,
479 					lockdep_is_held(&trie->lock));
480 			if (!node)
481 				goto unlock;
482 
483 			if (rcu_access_pointer(node->child[0])) {
484 				slot = &node->child[0];
485 				continue;
486 			}
487 
488 			if (rcu_access_pointer(node->child[1])) {
489 				slot = &node->child[1];
490 				continue;
491 			}
492 
493 			kfree(node);
494 			RCU_INIT_POINTER(*slot, NULL);
495 			break;
496 		}
497 	}
498 
499 unlock:
500 	raw_spin_unlock(&trie->lock);
501 }
502 
503 static int trie_get_next_key(struct bpf_map *map, void *key, void *next_key)
504 {
505 	return -ENOTSUPP;
506 }
507 
508 static const struct bpf_map_ops trie_ops = {
509 	.map_alloc = trie_alloc,
510 	.map_free = trie_free,
511 	.map_get_next_key = trie_get_next_key,
512 	.map_lookup_elem = trie_lookup_elem,
513 	.map_update_elem = trie_update_elem,
514 	.map_delete_elem = trie_delete_elem,
515 };
516 
517 static struct bpf_map_type_list trie_type __ro_after_init = {
518 	.ops = &trie_ops,
519 	.type = BPF_MAP_TYPE_LPM_TRIE,
520 };
521 
522 static int __init register_trie_map(void)
523 {
524 	bpf_register_map_type(&trie_type);
525 	return 0;
526 }
527 late_initcall(register_trie_map);
528