1 // SPDX-License-Identifier: GPL-2.0-only
2 /* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
3 * Copyright (c) 2016 Facebook
4 */
5 #include <linux/bpf.h>
6 #include <linux/btf.h>
7 #include <linux/jhash.h>
8 #include <linux/filter.h>
9 #include <linux/rculist_nulls.h>
10 #include <linux/random.h>
11 #include <uapi/linux/btf.h>
12 #include <linux/rcupdate_trace.h>
13 #include <linux/btf_ids.h>
14 #include "percpu_freelist.h"
15 #include "bpf_lru_list.h"
16 #include "map_in_map.h"
17 #include <linux/bpf_mem_alloc.h>
18
19 #define HTAB_CREATE_FLAG_MASK \
20 (BPF_F_NO_PREALLOC | BPF_F_NO_COMMON_LRU | BPF_F_NUMA_NODE | \
21 BPF_F_ACCESS_MASK | BPF_F_ZERO_SEED)
22
23 #define BATCH_OPS(_name) \
24 .map_lookup_batch = \
25 _name##_map_lookup_batch, \
26 .map_lookup_and_delete_batch = \
27 _name##_map_lookup_and_delete_batch, \
28 .map_update_batch = \
29 generic_map_update_batch, \
30 .map_delete_batch = \
31 generic_map_delete_batch
32
33 /*
34 * The bucket lock has two protection scopes:
35 *
36 * 1) Serializing concurrent operations from BPF programs on different
37 * CPUs
38 *
39 * 2) Serializing concurrent operations from BPF programs and sys_bpf()
40 *
41 * BPF programs can execute in any context including perf, kprobes and
42 * tracing. As there are almost no limits where perf, kprobes and tracing
43 * can be invoked from the lock operations need to be protected against
44 * deadlocks. Deadlocks can be caused by recursion and by an invocation in
45 * the lock held section when functions which acquire this lock are invoked
46 * from sys_bpf(). BPF recursion is prevented by incrementing the per CPU
47 * variable bpf_prog_active, which prevents BPF programs attached to perf
48 * events, kprobes and tracing to be invoked before the prior invocation
49 * from one of these contexts completed. sys_bpf() uses the same mechanism
50 * by pinning the task to the current CPU and incrementing the recursion
51 * protection across the map operation.
52 *
53 * This has subtle implications on PREEMPT_RT. PREEMPT_RT forbids certain
54 * operations like memory allocations (even with GFP_ATOMIC) from atomic
55 * contexts. This is required because even with GFP_ATOMIC the memory
56 * allocator calls into code paths which acquire locks with long held lock
57 * sections. To ensure the deterministic behaviour these locks are regular
58 * spinlocks, which are converted to 'sleepable' spinlocks on RT. The only
59 * true atomic contexts on an RT kernel are the low level hardware
60 * handling, scheduling, low level interrupt handling, NMIs etc. None of
61 * these contexts should ever do memory allocations.
62 *
63 * As regular device interrupt handlers and soft interrupts are forced into
64 * thread context, the existing code which does
65 * spin_lock*(); alloc(GFP_ATOMIC); spin_unlock*();
66 * just works.
67 *
68 * In theory the BPF locks could be converted to regular spinlocks as well,
69 * but the bucket locks and percpu_freelist locks can be taken from
70 * arbitrary contexts (perf, kprobes, tracepoints) which are required to be
71 * atomic contexts even on RT. Before the introduction of bpf_mem_alloc,
72 * it is only safe to use raw spinlock for preallocated hash map on a RT kernel,
73 * because there is no memory allocation within the lock held sections. However
74 * after hash map was fully converted to use bpf_mem_alloc, there will be
75 * non-synchronous memory allocation for non-preallocated hash map, so it is
76 * safe to always use raw spinlock for bucket lock.
77 */
78 struct bucket {
79 struct hlist_nulls_head head;
80 raw_spinlock_t raw_lock;
81 };
82
83 #define HASHTAB_MAP_LOCK_COUNT 8
84 #define HASHTAB_MAP_LOCK_MASK (HASHTAB_MAP_LOCK_COUNT - 1)
85
86 struct bpf_htab {
87 struct bpf_map map;
88 struct bpf_mem_alloc ma;
89 struct bpf_mem_alloc pcpu_ma;
90 struct bucket *buckets;
91 void *elems;
92 union {
93 struct pcpu_freelist freelist;
94 struct bpf_lru lru;
95 };
96 struct htab_elem *__percpu *extra_elems;
97 /* number of elements in non-preallocated hashtable are kept
98 * in either pcount or count
99 */
100 struct percpu_counter pcount;
101 atomic_t count;
102 bool use_percpu_counter;
103 u32 n_buckets; /* number of hash buckets */
104 u32 elem_size; /* size of each element in bytes */
105 u32 hashrnd;
106 struct lock_class_key lockdep_key;
107 int __percpu *map_locked[HASHTAB_MAP_LOCK_COUNT];
108 };
109
110 /* each htab element is struct htab_elem + key + value */
111 struct htab_elem {
112 union {
113 struct hlist_nulls_node hash_node;
114 struct {
115 void *padding;
116 union {
117 struct pcpu_freelist_node fnode;
118 struct htab_elem *batch_flink;
119 };
120 };
121 };
122 union {
123 /* pointer to per-cpu pointer */
124 void *ptr_to_pptr;
125 struct bpf_lru_node lru_node;
126 };
127 u32 hash;
128 char key[] __aligned(8);
129 };
130
htab_is_prealloc(const struct bpf_htab * htab)131 static inline bool htab_is_prealloc(const struct bpf_htab *htab)
132 {
133 return !(htab->map.map_flags & BPF_F_NO_PREALLOC);
134 }
135
htab_init_buckets(struct bpf_htab * htab)136 static void htab_init_buckets(struct bpf_htab *htab)
137 {
138 unsigned int i;
139
140 for (i = 0; i < htab->n_buckets; i++) {
141 INIT_HLIST_NULLS_HEAD(&htab->buckets[i].head, i);
142 raw_spin_lock_init(&htab->buckets[i].raw_lock);
143 lockdep_set_class(&htab->buckets[i].raw_lock,
144 &htab->lockdep_key);
145 cond_resched();
146 }
147 }
148
htab_lock_bucket(const struct bpf_htab * htab,struct bucket * b,u32 hash,unsigned long * pflags)149 static inline int htab_lock_bucket(const struct bpf_htab *htab,
150 struct bucket *b, u32 hash,
151 unsigned long *pflags)
152 {
153 unsigned long flags;
154
155 hash = hash & HASHTAB_MAP_LOCK_MASK;
156
157 preempt_disable();
158 if (unlikely(__this_cpu_inc_return(*(htab->map_locked[hash])) != 1)) {
159 __this_cpu_dec(*(htab->map_locked[hash]));
160 preempt_enable();
161 return -EBUSY;
162 }
163
164 raw_spin_lock_irqsave(&b->raw_lock, flags);
165 *pflags = flags;
166
167 return 0;
168 }
169
htab_unlock_bucket(const struct bpf_htab * htab,struct bucket * b,u32 hash,unsigned long flags)170 static inline void htab_unlock_bucket(const struct bpf_htab *htab,
171 struct bucket *b, u32 hash,
172 unsigned long flags)
173 {
174 hash = hash & HASHTAB_MAP_LOCK_MASK;
175 raw_spin_unlock_irqrestore(&b->raw_lock, flags);
176 __this_cpu_dec(*(htab->map_locked[hash]));
177 preempt_enable();
178 }
179
180 static bool htab_lru_map_delete_node(void *arg, struct bpf_lru_node *node);
181
htab_is_lru(const struct bpf_htab * htab)182 static bool htab_is_lru(const struct bpf_htab *htab)
183 {
184 return htab->map.map_type == BPF_MAP_TYPE_LRU_HASH ||
185 htab->map.map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH;
186 }
187
htab_is_percpu(const struct bpf_htab * htab)188 static bool htab_is_percpu(const struct bpf_htab *htab)
189 {
190 return htab->map.map_type == BPF_MAP_TYPE_PERCPU_HASH ||
191 htab->map.map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH;
192 }
193
htab_elem_set_ptr(struct htab_elem * l,u32 key_size,void __percpu * pptr)194 static inline void htab_elem_set_ptr(struct htab_elem *l, u32 key_size,
195 void __percpu *pptr)
196 {
197 *(void __percpu **)(l->key + key_size) = pptr;
198 }
199
htab_elem_get_ptr(struct htab_elem * l,u32 key_size)200 static inline void __percpu *htab_elem_get_ptr(struct htab_elem *l, u32 key_size)
201 {
202 return *(void __percpu **)(l->key + key_size);
203 }
204
fd_htab_map_get_ptr(const struct bpf_map * map,struct htab_elem * l)205 static void *fd_htab_map_get_ptr(const struct bpf_map *map, struct htab_elem *l)
206 {
207 return *(void **)(l->key + roundup(map->key_size, 8));
208 }
209
get_htab_elem(struct bpf_htab * htab,int i)210 static struct htab_elem *get_htab_elem(struct bpf_htab *htab, int i)
211 {
212 return (struct htab_elem *) (htab->elems + i * (u64)htab->elem_size);
213 }
214
htab_has_extra_elems(struct bpf_htab * htab)215 static bool htab_has_extra_elems(struct bpf_htab *htab)
216 {
217 return !htab_is_percpu(htab) && !htab_is_lru(htab);
218 }
219
htab_free_prealloced_timers(struct bpf_htab * htab)220 static void htab_free_prealloced_timers(struct bpf_htab *htab)
221 {
222 u32 num_entries = htab->map.max_entries;
223 int i;
224
225 if (!map_value_has_timer(&htab->map))
226 return;
227 if (htab_has_extra_elems(htab))
228 num_entries += num_possible_cpus();
229
230 for (i = 0; i < num_entries; i++) {
231 struct htab_elem *elem;
232
233 elem = get_htab_elem(htab, i);
234 bpf_timer_cancel_and_free(elem->key +
235 round_up(htab->map.key_size, 8) +
236 htab->map.timer_off);
237 cond_resched();
238 }
239 }
240
htab_free_prealloced_kptrs(struct bpf_htab * htab)241 static void htab_free_prealloced_kptrs(struct bpf_htab *htab)
242 {
243 u32 num_entries = htab->map.max_entries;
244 int i;
245
246 if (!map_value_has_kptrs(&htab->map))
247 return;
248 if (htab_has_extra_elems(htab))
249 num_entries += num_possible_cpus();
250
251 for (i = 0; i < num_entries; i++) {
252 struct htab_elem *elem;
253
254 elem = get_htab_elem(htab, i);
255 bpf_map_free_kptrs(&htab->map, elem->key + round_up(htab->map.key_size, 8));
256 cond_resched();
257 }
258 }
259
htab_free_elems(struct bpf_htab * htab)260 static void htab_free_elems(struct bpf_htab *htab)
261 {
262 int i;
263
264 if (!htab_is_percpu(htab))
265 goto free_elems;
266
267 for (i = 0; i < htab->map.max_entries; i++) {
268 void __percpu *pptr;
269
270 pptr = htab_elem_get_ptr(get_htab_elem(htab, i),
271 htab->map.key_size);
272 free_percpu(pptr);
273 cond_resched();
274 }
275 free_elems:
276 bpf_map_area_free(htab->elems);
277 }
278
279 /* The LRU list has a lock (lru_lock). Each htab bucket has a lock
280 * (bucket_lock). If both locks need to be acquired together, the lock
281 * order is always lru_lock -> bucket_lock and this only happens in
282 * bpf_lru_list.c logic. For example, certain code path of
283 * bpf_lru_pop_free(), which is called by function prealloc_lru_pop(),
284 * will acquire lru_lock first followed by acquiring bucket_lock.
285 *
286 * In hashtab.c, to avoid deadlock, lock acquisition of
287 * bucket_lock followed by lru_lock is not allowed. In such cases,
288 * bucket_lock needs to be released first before acquiring lru_lock.
289 */
prealloc_lru_pop(struct bpf_htab * htab,void * key,u32 hash)290 static struct htab_elem *prealloc_lru_pop(struct bpf_htab *htab, void *key,
291 u32 hash)
292 {
293 struct bpf_lru_node *node = bpf_lru_pop_free(&htab->lru, hash);
294 struct htab_elem *l;
295
296 if (node) {
297 l = container_of(node, struct htab_elem, lru_node);
298 memcpy(l->key, key, htab->map.key_size);
299 return l;
300 }
301
302 return NULL;
303 }
304
prealloc_init(struct bpf_htab * htab)305 static int prealloc_init(struct bpf_htab *htab)
306 {
307 u32 num_entries = htab->map.max_entries;
308 int err = -ENOMEM, i;
309
310 if (htab_has_extra_elems(htab))
311 num_entries += num_possible_cpus();
312
313 htab->elems = bpf_map_area_alloc((u64)htab->elem_size * num_entries,
314 htab->map.numa_node);
315 if (!htab->elems)
316 return -ENOMEM;
317
318 if (!htab_is_percpu(htab))
319 goto skip_percpu_elems;
320
321 for (i = 0; i < num_entries; i++) {
322 u32 size = round_up(htab->map.value_size, 8);
323 void __percpu *pptr;
324
325 pptr = bpf_map_alloc_percpu(&htab->map, size, 8,
326 GFP_USER | __GFP_NOWARN);
327 if (!pptr)
328 goto free_elems;
329 htab_elem_set_ptr(get_htab_elem(htab, i), htab->map.key_size,
330 pptr);
331 cond_resched();
332 }
333
334 skip_percpu_elems:
335 if (htab_is_lru(htab))
336 err = bpf_lru_init(&htab->lru,
337 htab->map.map_flags & BPF_F_NO_COMMON_LRU,
338 offsetof(struct htab_elem, hash) -
339 offsetof(struct htab_elem, lru_node),
340 htab_lru_map_delete_node,
341 htab);
342 else
343 err = pcpu_freelist_init(&htab->freelist);
344
345 if (err)
346 goto free_elems;
347
348 if (htab_is_lru(htab))
349 bpf_lru_populate(&htab->lru, htab->elems,
350 offsetof(struct htab_elem, lru_node),
351 htab->elem_size, num_entries);
352 else
353 pcpu_freelist_populate(&htab->freelist,
354 htab->elems + offsetof(struct htab_elem, fnode),
355 htab->elem_size, num_entries);
356
357 return 0;
358
359 free_elems:
360 htab_free_elems(htab);
361 return err;
362 }
363
prealloc_destroy(struct bpf_htab * htab)364 static void prealloc_destroy(struct bpf_htab *htab)
365 {
366 htab_free_elems(htab);
367
368 if (htab_is_lru(htab))
369 bpf_lru_destroy(&htab->lru);
370 else
371 pcpu_freelist_destroy(&htab->freelist);
372 }
373
alloc_extra_elems(struct bpf_htab * htab)374 static int alloc_extra_elems(struct bpf_htab *htab)
375 {
376 struct htab_elem *__percpu *pptr, *l_new;
377 struct pcpu_freelist_node *l;
378 int cpu;
379
380 pptr = bpf_map_alloc_percpu(&htab->map, sizeof(struct htab_elem *), 8,
381 GFP_USER | __GFP_NOWARN);
382 if (!pptr)
383 return -ENOMEM;
384
385 for_each_possible_cpu(cpu) {
386 l = pcpu_freelist_pop(&htab->freelist);
387 /* pop will succeed, since prealloc_init()
388 * preallocated extra num_possible_cpus elements
389 */
390 l_new = container_of(l, struct htab_elem, fnode);
391 *per_cpu_ptr(pptr, cpu) = l_new;
392 }
393 htab->extra_elems = pptr;
394 return 0;
395 }
396
397 /* Called from syscall */
htab_map_alloc_check(union bpf_attr * attr)398 static int htab_map_alloc_check(union bpf_attr *attr)
399 {
400 bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
401 attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
402 bool lru = (attr->map_type == BPF_MAP_TYPE_LRU_HASH ||
403 attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
404 /* percpu_lru means each cpu has its own LRU list.
405 * it is different from BPF_MAP_TYPE_PERCPU_HASH where
406 * the map's value itself is percpu. percpu_lru has
407 * nothing to do with the map's value.
408 */
409 bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU);
410 bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC);
411 bool zero_seed = (attr->map_flags & BPF_F_ZERO_SEED);
412 int numa_node = bpf_map_attr_numa_node(attr);
413
414 BUILD_BUG_ON(offsetof(struct htab_elem, fnode.next) !=
415 offsetof(struct htab_elem, hash_node.pprev));
416
417 if (lru && !bpf_capable())
418 /* LRU implementation is much complicated than other
419 * maps. Hence, limit to CAP_BPF.
420 */
421 return -EPERM;
422
423 if (zero_seed && !capable(CAP_SYS_ADMIN))
424 /* Guard against local DoS, and discourage production use. */
425 return -EPERM;
426
427 if (attr->map_flags & ~HTAB_CREATE_FLAG_MASK ||
428 !bpf_map_flags_access_ok(attr->map_flags))
429 return -EINVAL;
430
431 if (!lru && percpu_lru)
432 return -EINVAL;
433
434 if (lru && !prealloc)
435 return -ENOTSUPP;
436
437 if (numa_node != NUMA_NO_NODE && (percpu || percpu_lru))
438 return -EINVAL;
439
440 /* check sanity of attributes.
441 * value_size == 0 may be allowed in the future to use map as a set
442 */
443 if (attr->max_entries == 0 || attr->key_size == 0 ||
444 attr->value_size == 0)
445 return -EINVAL;
446
447 if ((u64)attr->key_size + attr->value_size >= KMALLOC_MAX_SIZE -
448 sizeof(struct htab_elem))
449 /* if key_size + value_size is bigger, the user space won't be
450 * able to access the elements via bpf syscall. This check
451 * also makes sure that the elem_size doesn't overflow and it's
452 * kmalloc-able later in htab_map_update_elem()
453 */
454 return -E2BIG;
455
456 return 0;
457 }
458
htab_map_alloc(union bpf_attr * attr)459 static struct bpf_map *htab_map_alloc(union bpf_attr *attr)
460 {
461 bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
462 attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
463 bool lru = (attr->map_type == BPF_MAP_TYPE_LRU_HASH ||
464 attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
465 /* percpu_lru means each cpu has its own LRU list.
466 * it is different from BPF_MAP_TYPE_PERCPU_HASH where
467 * the map's value itself is percpu. percpu_lru has
468 * nothing to do with the map's value.
469 */
470 bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU);
471 bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC);
472 struct bpf_htab *htab;
473 int err, i;
474
475 htab = bpf_map_area_alloc(sizeof(*htab), NUMA_NO_NODE);
476 if (!htab)
477 return ERR_PTR(-ENOMEM);
478
479 lockdep_register_key(&htab->lockdep_key);
480
481 bpf_map_init_from_attr(&htab->map, attr);
482
483 if (percpu_lru) {
484 /* ensure each CPU's lru list has >=1 elements.
485 * since we are at it, make each lru list has the same
486 * number of elements.
487 */
488 htab->map.max_entries = roundup(attr->max_entries,
489 num_possible_cpus());
490 if (htab->map.max_entries < attr->max_entries)
491 htab->map.max_entries = rounddown(attr->max_entries,
492 num_possible_cpus());
493 }
494
495 /* hash table size must be power of 2 */
496 htab->n_buckets = roundup_pow_of_two(htab->map.max_entries);
497
498 htab->elem_size = sizeof(struct htab_elem) +
499 round_up(htab->map.key_size, 8);
500 if (percpu)
501 htab->elem_size += sizeof(void *);
502 else
503 htab->elem_size += round_up(htab->map.value_size, 8);
504
505 err = -E2BIG;
506 /* prevent zero size kmalloc and check for u32 overflow */
507 if (htab->n_buckets == 0 ||
508 htab->n_buckets > U32_MAX / sizeof(struct bucket))
509 goto free_htab;
510
511 err = -ENOMEM;
512 htab->buckets = bpf_map_area_alloc(htab->n_buckets *
513 sizeof(struct bucket),
514 htab->map.numa_node);
515 if (!htab->buckets)
516 goto free_htab;
517
518 for (i = 0; i < HASHTAB_MAP_LOCK_COUNT; i++) {
519 htab->map_locked[i] = bpf_map_alloc_percpu(&htab->map,
520 sizeof(int),
521 sizeof(int),
522 GFP_USER);
523 if (!htab->map_locked[i])
524 goto free_map_locked;
525 }
526
527 if (htab->map.map_flags & BPF_F_ZERO_SEED)
528 htab->hashrnd = 0;
529 else
530 htab->hashrnd = get_random_u32();
531
532 htab_init_buckets(htab);
533
534 /* compute_batch_value() computes batch value as num_online_cpus() * 2
535 * and __percpu_counter_compare() needs
536 * htab->max_entries - cur_number_of_elems to be more than batch * num_online_cpus()
537 * for percpu_counter to be faster than atomic_t. In practice the average bpf
538 * hash map size is 10k, which means that a system with 64 cpus will fill
539 * hashmap to 20% of 10k before percpu_counter becomes ineffective. Therefore
540 * define our own batch count as 32 then 10k hash map can be filled up to 80%:
541 * 10k - 8k > 32 _batch_ * 64 _cpus_
542 * and __percpu_counter_compare() will still be fast. At that point hash map
543 * collisions will dominate its performance anyway. Assume that hash map filled
544 * to 50+% isn't going to be O(1) and use the following formula to choose
545 * between percpu_counter and atomic_t.
546 */
547 #define PERCPU_COUNTER_BATCH 32
548 if (attr->max_entries / 2 > num_online_cpus() * PERCPU_COUNTER_BATCH)
549 htab->use_percpu_counter = true;
550
551 if (htab->use_percpu_counter) {
552 err = percpu_counter_init(&htab->pcount, 0, GFP_KERNEL);
553 if (err)
554 goto free_map_locked;
555 }
556
557 if (prealloc) {
558 err = prealloc_init(htab);
559 if (err)
560 goto free_map_locked;
561
562 if (!percpu && !lru) {
563 /* lru itself can remove the least used element, so
564 * there is no need for an extra elem during map_update.
565 */
566 err = alloc_extra_elems(htab);
567 if (err)
568 goto free_prealloc;
569 }
570 } else {
571 err = bpf_mem_alloc_init(&htab->ma, htab->elem_size, false);
572 if (err)
573 goto free_map_locked;
574 if (percpu) {
575 err = bpf_mem_alloc_init(&htab->pcpu_ma,
576 round_up(htab->map.value_size, 8), true);
577 if (err)
578 goto free_map_locked;
579 }
580 }
581
582 return &htab->map;
583
584 free_prealloc:
585 prealloc_destroy(htab);
586 free_map_locked:
587 if (htab->use_percpu_counter)
588 percpu_counter_destroy(&htab->pcount);
589 for (i = 0; i < HASHTAB_MAP_LOCK_COUNT; i++)
590 free_percpu(htab->map_locked[i]);
591 bpf_map_area_free(htab->buckets);
592 bpf_mem_alloc_destroy(&htab->pcpu_ma);
593 bpf_mem_alloc_destroy(&htab->ma);
594 free_htab:
595 lockdep_unregister_key(&htab->lockdep_key);
596 bpf_map_area_free(htab);
597 return ERR_PTR(err);
598 }
599
htab_map_hash(const void * key,u32 key_len,u32 hashrnd)600 static inline u32 htab_map_hash(const void *key, u32 key_len, u32 hashrnd)
601 {
602 return jhash(key, key_len, hashrnd);
603 }
604
__select_bucket(struct bpf_htab * htab,u32 hash)605 static inline struct bucket *__select_bucket(struct bpf_htab *htab, u32 hash)
606 {
607 return &htab->buckets[hash & (htab->n_buckets - 1)];
608 }
609
select_bucket(struct bpf_htab * htab,u32 hash)610 static inline struct hlist_nulls_head *select_bucket(struct bpf_htab *htab, u32 hash)
611 {
612 return &__select_bucket(htab, hash)->head;
613 }
614
615 /* this lookup function can only be called with bucket lock taken */
lookup_elem_raw(struct hlist_nulls_head * head,u32 hash,void * key,u32 key_size)616 static struct htab_elem *lookup_elem_raw(struct hlist_nulls_head *head, u32 hash,
617 void *key, u32 key_size)
618 {
619 struct hlist_nulls_node *n;
620 struct htab_elem *l;
621
622 hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
623 if (l->hash == hash && !memcmp(&l->key, key, key_size))
624 return l;
625
626 return NULL;
627 }
628
629 /* can be called without bucket lock. it will repeat the loop in
630 * the unlikely event when elements moved from one bucket into another
631 * while link list is being walked
632 */
lookup_nulls_elem_raw(struct hlist_nulls_head * head,u32 hash,void * key,u32 key_size,u32 n_buckets)633 static struct htab_elem *lookup_nulls_elem_raw(struct hlist_nulls_head *head,
634 u32 hash, void *key,
635 u32 key_size, u32 n_buckets)
636 {
637 struct hlist_nulls_node *n;
638 struct htab_elem *l;
639
640 again:
641 hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
642 if (l->hash == hash && !memcmp(&l->key, key, key_size))
643 return l;
644
645 if (unlikely(get_nulls_value(n) != (hash & (n_buckets - 1))))
646 goto again;
647
648 return NULL;
649 }
650
651 /* Called from syscall or from eBPF program directly, so
652 * arguments have to match bpf_map_lookup_elem() exactly.
653 * The return value is adjusted by BPF instructions
654 * in htab_map_gen_lookup().
655 */
__htab_map_lookup_elem(struct bpf_map * map,void * key)656 static void *__htab_map_lookup_elem(struct bpf_map *map, void *key)
657 {
658 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
659 struct hlist_nulls_head *head;
660 struct htab_elem *l;
661 u32 hash, key_size;
662
663 WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
664 !rcu_read_lock_bh_held());
665
666 key_size = map->key_size;
667
668 hash = htab_map_hash(key, key_size, htab->hashrnd);
669
670 head = select_bucket(htab, hash);
671
672 l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
673
674 return l;
675 }
676
htab_map_lookup_elem(struct bpf_map * map,void * key)677 static void *htab_map_lookup_elem(struct bpf_map *map, void *key)
678 {
679 struct htab_elem *l = __htab_map_lookup_elem(map, key);
680
681 if (l)
682 return l->key + round_up(map->key_size, 8);
683
684 return NULL;
685 }
686
687 /* inline bpf_map_lookup_elem() call.
688 * Instead of:
689 * bpf_prog
690 * bpf_map_lookup_elem
691 * map->ops->map_lookup_elem
692 * htab_map_lookup_elem
693 * __htab_map_lookup_elem
694 * do:
695 * bpf_prog
696 * __htab_map_lookup_elem
697 */
htab_map_gen_lookup(struct bpf_map * map,struct bpf_insn * insn_buf)698 static int htab_map_gen_lookup(struct bpf_map *map, struct bpf_insn *insn_buf)
699 {
700 struct bpf_insn *insn = insn_buf;
701 const int ret = BPF_REG_0;
702
703 BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
704 (void *(*)(struct bpf_map *map, void *key))NULL));
705 *insn++ = BPF_EMIT_CALL(__htab_map_lookup_elem);
706 *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 1);
707 *insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
708 offsetof(struct htab_elem, key) +
709 round_up(map->key_size, 8));
710 return insn - insn_buf;
711 }
712
__htab_lru_map_lookup_elem(struct bpf_map * map,void * key,const bool mark)713 static __always_inline void *__htab_lru_map_lookup_elem(struct bpf_map *map,
714 void *key, const bool mark)
715 {
716 struct htab_elem *l = __htab_map_lookup_elem(map, key);
717
718 if (l) {
719 if (mark)
720 bpf_lru_node_set_ref(&l->lru_node);
721 return l->key + round_up(map->key_size, 8);
722 }
723
724 return NULL;
725 }
726
htab_lru_map_lookup_elem(struct bpf_map * map,void * key)727 static void *htab_lru_map_lookup_elem(struct bpf_map *map, void *key)
728 {
729 return __htab_lru_map_lookup_elem(map, key, true);
730 }
731
htab_lru_map_lookup_elem_sys(struct bpf_map * map,void * key)732 static void *htab_lru_map_lookup_elem_sys(struct bpf_map *map, void *key)
733 {
734 return __htab_lru_map_lookup_elem(map, key, false);
735 }
736
htab_lru_map_gen_lookup(struct bpf_map * map,struct bpf_insn * insn_buf)737 static int htab_lru_map_gen_lookup(struct bpf_map *map,
738 struct bpf_insn *insn_buf)
739 {
740 struct bpf_insn *insn = insn_buf;
741 const int ret = BPF_REG_0;
742 const int ref_reg = BPF_REG_1;
743
744 BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
745 (void *(*)(struct bpf_map *map, void *key))NULL));
746 *insn++ = BPF_EMIT_CALL(__htab_map_lookup_elem);
747 *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 4);
748 *insn++ = BPF_LDX_MEM(BPF_B, ref_reg, ret,
749 offsetof(struct htab_elem, lru_node) +
750 offsetof(struct bpf_lru_node, ref));
751 *insn++ = BPF_JMP_IMM(BPF_JNE, ref_reg, 0, 1);
752 *insn++ = BPF_ST_MEM(BPF_B, ret,
753 offsetof(struct htab_elem, lru_node) +
754 offsetof(struct bpf_lru_node, ref),
755 1);
756 *insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
757 offsetof(struct htab_elem, key) +
758 round_up(map->key_size, 8));
759 return insn - insn_buf;
760 }
761
check_and_free_fields(struct bpf_htab * htab,struct htab_elem * elem)762 static void check_and_free_fields(struct bpf_htab *htab,
763 struct htab_elem *elem)
764 {
765 void *map_value = elem->key + round_up(htab->map.key_size, 8);
766
767 if (map_value_has_timer(&htab->map))
768 bpf_timer_cancel_and_free(map_value + htab->map.timer_off);
769 if (map_value_has_kptrs(&htab->map))
770 bpf_map_free_kptrs(&htab->map, map_value);
771 }
772
773 /* It is called from the bpf_lru_list when the LRU needs to delete
774 * older elements from the htab.
775 */
htab_lru_map_delete_node(void * arg,struct bpf_lru_node * node)776 static bool htab_lru_map_delete_node(void *arg, struct bpf_lru_node *node)
777 {
778 struct bpf_htab *htab = arg;
779 struct htab_elem *l = NULL, *tgt_l;
780 struct hlist_nulls_head *head;
781 struct hlist_nulls_node *n;
782 unsigned long flags;
783 struct bucket *b;
784 int ret;
785
786 tgt_l = container_of(node, struct htab_elem, lru_node);
787 b = __select_bucket(htab, tgt_l->hash);
788 head = &b->head;
789
790 ret = htab_lock_bucket(htab, b, tgt_l->hash, &flags);
791 if (ret)
792 return false;
793
794 hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
795 if (l == tgt_l) {
796 hlist_nulls_del_rcu(&l->hash_node);
797 check_and_free_fields(htab, l);
798 break;
799 }
800
801 htab_unlock_bucket(htab, b, tgt_l->hash, flags);
802
803 return l == tgt_l;
804 }
805
806 /* Called from syscall */
htab_map_get_next_key(struct bpf_map * map,void * key,void * next_key)807 static int htab_map_get_next_key(struct bpf_map *map, void *key, void *next_key)
808 {
809 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
810 struct hlist_nulls_head *head;
811 struct htab_elem *l, *next_l;
812 u32 hash, key_size;
813 int i = 0;
814
815 WARN_ON_ONCE(!rcu_read_lock_held());
816
817 key_size = map->key_size;
818
819 if (!key)
820 goto find_first_elem;
821
822 hash = htab_map_hash(key, key_size, htab->hashrnd);
823
824 head = select_bucket(htab, hash);
825
826 /* lookup the key */
827 l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
828
829 if (!l)
830 goto find_first_elem;
831
832 /* key was found, get next key in the same bucket */
833 next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_next_rcu(&l->hash_node)),
834 struct htab_elem, hash_node);
835
836 if (next_l) {
837 /* if next elem in this hash list is non-zero, just return it */
838 memcpy(next_key, next_l->key, key_size);
839 return 0;
840 }
841
842 /* no more elements in this hash list, go to the next bucket */
843 i = hash & (htab->n_buckets - 1);
844 i++;
845
846 find_first_elem:
847 /* iterate over buckets */
848 for (; i < htab->n_buckets; i++) {
849 head = select_bucket(htab, i);
850
851 /* pick first element in the bucket */
852 next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_first_rcu(head)),
853 struct htab_elem, hash_node);
854 if (next_l) {
855 /* if it's not empty, just return it */
856 memcpy(next_key, next_l->key, key_size);
857 return 0;
858 }
859 }
860
861 /* iterated over all buckets and all elements */
862 return -ENOENT;
863 }
864
htab_elem_free(struct bpf_htab * htab,struct htab_elem * l)865 static void htab_elem_free(struct bpf_htab *htab, struct htab_elem *l)
866 {
867 if (htab->map.map_type == BPF_MAP_TYPE_PERCPU_HASH)
868 bpf_mem_cache_free(&htab->pcpu_ma, l->ptr_to_pptr);
869 check_and_free_fields(htab, l);
870 bpf_mem_cache_free(&htab->ma, l);
871 }
872
htab_put_fd_value(struct bpf_htab * htab,struct htab_elem * l)873 static void htab_put_fd_value(struct bpf_htab *htab, struct htab_elem *l)
874 {
875 struct bpf_map *map = &htab->map;
876 void *ptr;
877
878 if (map->ops->map_fd_put_ptr) {
879 ptr = fd_htab_map_get_ptr(map, l);
880 map->ops->map_fd_put_ptr(ptr);
881 }
882 }
883
is_map_full(struct bpf_htab * htab)884 static bool is_map_full(struct bpf_htab *htab)
885 {
886 if (htab->use_percpu_counter)
887 return __percpu_counter_compare(&htab->pcount, htab->map.max_entries,
888 PERCPU_COUNTER_BATCH) >= 0;
889 return atomic_read(&htab->count) >= htab->map.max_entries;
890 }
891
inc_elem_count(struct bpf_htab * htab)892 static void inc_elem_count(struct bpf_htab *htab)
893 {
894 if (htab->use_percpu_counter)
895 percpu_counter_add_batch(&htab->pcount, 1, PERCPU_COUNTER_BATCH);
896 else
897 atomic_inc(&htab->count);
898 }
899
dec_elem_count(struct bpf_htab * htab)900 static void dec_elem_count(struct bpf_htab *htab)
901 {
902 if (htab->use_percpu_counter)
903 percpu_counter_add_batch(&htab->pcount, -1, PERCPU_COUNTER_BATCH);
904 else
905 atomic_dec(&htab->count);
906 }
907
908
free_htab_elem(struct bpf_htab * htab,struct htab_elem * l)909 static void free_htab_elem(struct bpf_htab *htab, struct htab_elem *l)
910 {
911 htab_put_fd_value(htab, l);
912
913 if (htab_is_prealloc(htab)) {
914 check_and_free_fields(htab, l);
915 __pcpu_freelist_push(&htab->freelist, &l->fnode);
916 } else {
917 dec_elem_count(htab);
918 htab_elem_free(htab, l);
919 }
920 }
921
pcpu_copy_value(struct bpf_htab * htab,void __percpu * pptr,void * value,bool onallcpus)922 static void pcpu_copy_value(struct bpf_htab *htab, void __percpu *pptr,
923 void *value, bool onallcpus)
924 {
925 if (!onallcpus) {
926 /* copy true value_size bytes */
927 memcpy(this_cpu_ptr(pptr), value, htab->map.value_size);
928 } else {
929 u32 size = round_up(htab->map.value_size, 8);
930 int off = 0, cpu;
931
932 for_each_possible_cpu(cpu) {
933 bpf_long_memcpy(per_cpu_ptr(pptr, cpu),
934 value + off, size);
935 off += size;
936 }
937 }
938 }
939
pcpu_init_value(struct bpf_htab * htab,void __percpu * pptr,void * value,bool onallcpus)940 static void pcpu_init_value(struct bpf_htab *htab, void __percpu *pptr,
941 void *value, bool onallcpus)
942 {
943 /* When not setting the initial value on all cpus, zero-fill element
944 * values for other cpus. Otherwise, bpf program has no way to ensure
945 * known initial values for cpus other than current one
946 * (onallcpus=false always when coming from bpf prog).
947 */
948 if (!onallcpus) {
949 u32 size = round_up(htab->map.value_size, 8);
950 int current_cpu = raw_smp_processor_id();
951 int cpu;
952
953 for_each_possible_cpu(cpu) {
954 if (cpu == current_cpu)
955 bpf_long_memcpy(per_cpu_ptr(pptr, cpu), value,
956 size);
957 else
958 memset(per_cpu_ptr(pptr, cpu), 0, size);
959 }
960 } else {
961 pcpu_copy_value(htab, pptr, value, onallcpus);
962 }
963 }
964
fd_htab_map_needs_adjust(const struct bpf_htab * htab)965 static bool fd_htab_map_needs_adjust(const struct bpf_htab *htab)
966 {
967 return htab->map.map_type == BPF_MAP_TYPE_HASH_OF_MAPS &&
968 BITS_PER_LONG == 64;
969 }
970
alloc_htab_elem(struct bpf_htab * htab,void * key,void * value,u32 key_size,u32 hash,bool percpu,bool onallcpus,struct htab_elem * old_elem)971 static struct htab_elem *alloc_htab_elem(struct bpf_htab *htab, void *key,
972 void *value, u32 key_size, u32 hash,
973 bool percpu, bool onallcpus,
974 struct htab_elem *old_elem)
975 {
976 u32 size = htab->map.value_size;
977 bool prealloc = htab_is_prealloc(htab);
978 struct htab_elem *l_new, **pl_new;
979 void __percpu *pptr;
980
981 if (prealloc) {
982 if (old_elem) {
983 /* if we're updating the existing element,
984 * use per-cpu extra elems to avoid freelist_pop/push
985 */
986 pl_new = this_cpu_ptr(htab->extra_elems);
987 l_new = *pl_new;
988 htab_put_fd_value(htab, old_elem);
989 *pl_new = old_elem;
990 } else {
991 struct pcpu_freelist_node *l;
992
993 l = __pcpu_freelist_pop(&htab->freelist);
994 if (!l)
995 return ERR_PTR(-E2BIG);
996 l_new = container_of(l, struct htab_elem, fnode);
997 }
998 } else {
999 if (is_map_full(htab))
1000 if (!old_elem)
1001 /* when map is full and update() is replacing
1002 * old element, it's ok to allocate, since
1003 * old element will be freed immediately.
1004 * Otherwise return an error
1005 */
1006 return ERR_PTR(-E2BIG);
1007 inc_elem_count(htab);
1008 l_new = bpf_mem_cache_alloc(&htab->ma);
1009 if (!l_new) {
1010 l_new = ERR_PTR(-ENOMEM);
1011 goto dec_count;
1012 }
1013 check_and_init_map_value(&htab->map,
1014 l_new->key + round_up(key_size, 8));
1015 }
1016
1017 memcpy(l_new->key, key, key_size);
1018 if (percpu) {
1019 if (prealloc) {
1020 pptr = htab_elem_get_ptr(l_new, key_size);
1021 } else {
1022 /* alloc_percpu zero-fills */
1023 pptr = bpf_mem_cache_alloc(&htab->pcpu_ma);
1024 if (!pptr) {
1025 bpf_mem_cache_free(&htab->ma, l_new);
1026 l_new = ERR_PTR(-ENOMEM);
1027 goto dec_count;
1028 }
1029 l_new->ptr_to_pptr = pptr;
1030 pptr = *(void **)pptr;
1031 }
1032
1033 pcpu_init_value(htab, pptr, value, onallcpus);
1034
1035 if (!prealloc)
1036 htab_elem_set_ptr(l_new, key_size, pptr);
1037 } else if (fd_htab_map_needs_adjust(htab)) {
1038 size = round_up(size, 8);
1039 memcpy(l_new->key + round_up(key_size, 8), value, size);
1040 } else {
1041 copy_map_value(&htab->map,
1042 l_new->key + round_up(key_size, 8),
1043 value);
1044 }
1045
1046 l_new->hash = hash;
1047 return l_new;
1048 dec_count:
1049 dec_elem_count(htab);
1050 return l_new;
1051 }
1052
check_flags(struct bpf_htab * htab,struct htab_elem * l_old,u64 map_flags)1053 static int check_flags(struct bpf_htab *htab, struct htab_elem *l_old,
1054 u64 map_flags)
1055 {
1056 if (l_old && (map_flags & ~BPF_F_LOCK) == BPF_NOEXIST)
1057 /* elem already exists */
1058 return -EEXIST;
1059
1060 if (!l_old && (map_flags & ~BPF_F_LOCK) == BPF_EXIST)
1061 /* elem doesn't exist, cannot update it */
1062 return -ENOENT;
1063
1064 return 0;
1065 }
1066
1067 /* Called from syscall or from eBPF program */
htab_map_update_elem(struct bpf_map * map,void * key,void * value,u64 map_flags)1068 static int htab_map_update_elem(struct bpf_map *map, void *key, void *value,
1069 u64 map_flags)
1070 {
1071 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1072 struct htab_elem *l_new = NULL, *l_old;
1073 struct hlist_nulls_head *head;
1074 unsigned long flags;
1075 struct bucket *b;
1076 u32 key_size, hash;
1077 int ret;
1078
1079 if (unlikely((map_flags & ~BPF_F_LOCK) > BPF_EXIST))
1080 /* unknown flags */
1081 return -EINVAL;
1082
1083 WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
1084 !rcu_read_lock_bh_held());
1085
1086 key_size = map->key_size;
1087
1088 hash = htab_map_hash(key, key_size, htab->hashrnd);
1089
1090 b = __select_bucket(htab, hash);
1091 head = &b->head;
1092
1093 if (unlikely(map_flags & BPF_F_LOCK)) {
1094 if (unlikely(!map_value_has_spin_lock(map)))
1095 return -EINVAL;
1096 /* find an element without taking the bucket lock */
1097 l_old = lookup_nulls_elem_raw(head, hash, key, key_size,
1098 htab->n_buckets);
1099 ret = check_flags(htab, l_old, map_flags);
1100 if (ret)
1101 return ret;
1102 if (l_old) {
1103 /* grab the element lock and update value in place */
1104 copy_map_value_locked(map,
1105 l_old->key + round_up(key_size, 8),
1106 value, false);
1107 return 0;
1108 }
1109 /* fall through, grab the bucket lock and lookup again.
1110 * 99.9% chance that the element won't be found,
1111 * but second lookup under lock has to be done.
1112 */
1113 }
1114
1115 ret = htab_lock_bucket(htab, b, hash, &flags);
1116 if (ret)
1117 return ret;
1118
1119 l_old = lookup_elem_raw(head, hash, key, key_size);
1120
1121 ret = check_flags(htab, l_old, map_flags);
1122 if (ret)
1123 goto err;
1124
1125 if (unlikely(l_old && (map_flags & BPF_F_LOCK))) {
1126 /* first lookup without the bucket lock didn't find the element,
1127 * but second lookup with the bucket lock found it.
1128 * This case is highly unlikely, but has to be dealt with:
1129 * grab the element lock in addition to the bucket lock
1130 * and update element in place
1131 */
1132 copy_map_value_locked(map,
1133 l_old->key + round_up(key_size, 8),
1134 value, false);
1135 ret = 0;
1136 goto err;
1137 }
1138
1139 l_new = alloc_htab_elem(htab, key, value, key_size, hash, false, false,
1140 l_old);
1141 if (IS_ERR(l_new)) {
1142 /* all pre-allocated elements are in use or memory exhausted */
1143 ret = PTR_ERR(l_new);
1144 goto err;
1145 }
1146
1147 /* add new element to the head of the list, so that
1148 * concurrent search will find it before old elem
1149 */
1150 hlist_nulls_add_head_rcu(&l_new->hash_node, head);
1151 if (l_old) {
1152 hlist_nulls_del_rcu(&l_old->hash_node);
1153 if (!htab_is_prealloc(htab))
1154 free_htab_elem(htab, l_old);
1155 else
1156 check_and_free_fields(htab, l_old);
1157 }
1158 ret = 0;
1159 err:
1160 htab_unlock_bucket(htab, b, hash, flags);
1161 return ret;
1162 }
1163
htab_lru_push_free(struct bpf_htab * htab,struct htab_elem * elem)1164 static void htab_lru_push_free(struct bpf_htab *htab, struct htab_elem *elem)
1165 {
1166 check_and_free_fields(htab, elem);
1167 bpf_lru_push_free(&htab->lru, &elem->lru_node);
1168 }
1169
htab_lru_map_update_elem(struct bpf_map * map,void * key,void * value,u64 map_flags)1170 static int htab_lru_map_update_elem(struct bpf_map *map, void *key, void *value,
1171 u64 map_flags)
1172 {
1173 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1174 struct htab_elem *l_new, *l_old = NULL;
1175 struct hlist_nulls_head *head;
1176 unsigned long flags;
1177 struct bucket *b;
1178 u32 key_size, hash;
1179 int ret;
1180
1181 if (unlikely(map_flags > BPF_EXIST))
1182 /* unknown flags */
1183 return -EINVAL;
1184
1185 WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
1186 !rcu_read_lock_bh_held());
1187
1188 key_size = map->key_size;
1189
1190 hash = htab_map_hash(key, key_size, htab->hashrnd);
1191
1192 b = __select_bucket(htab, hash);
1193 head = &b->head;
1194
1195 /* For LRU, we need to alloc before taking bucket's
1196 * spinlock because getting free nodes from LRU may need
1197 * to remove older elements from htab and this removal
1198 * operation will need a bucket lock.
1199 */
1200 l_new = prealloc_lru_pop(htab, key, hash);
1201 if (!l_new)
1202 return -ENOMEM;
1203 copy_map_value(&htab->map,
1204 l_new->key + round_up(map->key_size, 8), value);
1205
1206 ret = htab_lock_bucket(htab, b, hash, &flags);
1207 if (ret)
1208 return ret;
1209
1210 l_old = lookup_elem_raw(head, hash, key, key_size);
1211
1212 ret = check_flags(htab, l_old, map_flags);
1213 if (ret)
1214 goto err;
1215
1216 /* add new element to the head of the list, so that
1217 * concurrent search will find it before old elem
1218 */
1219 hlist_nulls_add_head_rcu(&l_new->hash_node, head);
1220 if (l_old) {
1221 bpf_lru_node_set_ref(&l_new->lru_node);
1222 hlist_nulls_del_rcu(&l_old->hash_node);
1223 }
1224 ret = 0;
1225
1226 err:
1227 htab_unlock_bucket(htab, b, hash, flags);
1228
1229 if (ret)
1230 htab_lru_push_free(htab, l_new);
1231 else if (l_old)
1232 htab_lru_push_free(htab, l_old);
1233
1234 return ret;
1235 }
1236
__htab_percpu_map_update_elem(struct bpf_map * map,void * key,void * value,u64 map_flags,bool onallcpus)1237 static int __htab_percpu_map_update_elem(struct bpf_map *map, void *key,
1238 void *value, u64 map_flags,
1239 bool onallcpus)
1240 {
1241 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1242 struct htab_elem *l_new = NULL, *l_old;
1243 struct hlist_nulls_head *head;
1244 unsigned long flags;
1245 struct bucket *b;
1246 u32 key_size, hash;
1247 int ret;
1248
1249 if (unlikely(map_flags > BPF_EXIST))
1250 /* unknown flags */
1251 return -EINVAL;
1252
1253 WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
1254 !rcu_read_lock_bh_held());
1255
1256 key_size = map->key_size;
1257
1258 hash = htab_map_hash(key, key_size, htab->hashrnd);
1259
1260 b = __select_bucket(htab, hash);
1261 head = &b->head;
1262
1263 ret = htab_lock_bucket(htab, b, hash, &flags);
1264 if (ret)
1265 return ret;
1266
1267 l_old = lookup_elem_raw(head, hash, key, key_size);
1268
1269 ret = check_flags(htab, l_old, map_flags);
1270 if (ret)
1271 goto err;
1272
1273 if (l_old) {
1274 /* per-cpu hash map can update value in-place */
1275 pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size),
1276 value, onallcpus);
1277 } else {
1278 l_new = alloc_htab_elem(htab, key, value, key_size,
1279 hash, true, onallcpus, NULL);
1280 if (IS_ERR(l_new)) {
1281 ret = PTR_ERR(l_new);
1282 goto err;
1283 }
1284 hlist_nulls_add_head_rcu(&l_new->hash_node, head);
1285 }
1286 ret = 0;
1287 err:
1288 htab_unlock_bucket(htab, b, hash, flags);
1289 return ret;
1290 }
1291
__htab_lru_percpu_map_update_elem(struct bpf_map * map,void * key,void * value,u64 map_flags,bool onallcpus)1292 static int __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
1293 void *value, u64 map_flags,
1294 bool onallcpus)
1295 {
1296 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1297 struct htab_elem *l_new = NULL, *l_old;
1298 struct hlist_nulls_head *head;
1299 unsigned long flags;
1300 struct bucket *b;
1301 u32 key_size, hash;
1302 int ret;
1303
1304 if (unlikely(map_flags > BPF_EXIST))
1305 /* unknown flags */
1306 return -EINVAL;
1307
1308 WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
1309 !rcu_read_lock_bh_held());
1310
1311 key_size = map->key_size;
1312
1313 hash = htab_map_hash(key, key_size, htab->hashrnd);
1314
1315 b = __select_bucket(htab, hash);
1316 head = &b->head;
1317
1318 /* For LRU, we need to alloc before taking bucket's
1319 * spinlock because LRU's elem alloc may need
1320 * to remove older elem from htab and this removal
1321 * operation will need a bucket lock.
1322 */
1323 if (map_flags != BPF_EXIST) {
1324 l_new = prealloc_lru_pop(htab, key, hash);
1325 if (!l_new)
1326 return -ENOMEM;
1327 }
1328
1329 ret = htab_lock_bucket(htab, b, hash, &flags);
1330 if (ret)
1331 return ret;
1332
1333 l_old = lookup_elem_raw(head, hash, key, key_size);
1334
1335 ret = check_flags(htab, l_old, map_flags);
1336 if (ret)
1337 goto err;
1338
1339 if (l_old) {
1340 bpf_lru_node_set_ref(&l_old->lru_node);
1341
1342 /* per-cpu hash map can update value in-place */
1343 pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size),
1344 value, onallcpus);
1345 } else {
1346 pcpu_init_value(htab, htab_elem_get_ptr(l_new, key_size),
1347 value, onallcpus);
1348 hlist_nulls_add_head_rcu(&l_new->hash_node, head);
1349 l_new = NULL;
1350 }
1351 ret = 0;
1352 err:
1353 htab_unlock_bucket(htab, b, hash, flags);
1354 if (l_new)
1355 bpf_lru_push_free(&htab->lru, &l_new->lru_node);
1356 return ret;
1357 }
1358
htab_percpu_map_update_elem(struct bpf_map * map,void * key,void * value,u64 map_flags)1359 static int htab_percpu_map_update_elem(struct bpf_map *map, void *key,
1360 void *value, u64 map_flags)
1361 {
1362 return __htab_percpu_map_update_elem(map, key, value, map_flags, false);
1363 }
1364
htab_lru_percpu_map_update_elem(struct bpf_map * map,void * key,void * value,u64 map_flags)1365 static int htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
1366 void *value, u64 map_flags)
1367 {
1368 return __htab_lru_percpu_map_update_elem(map, key, value, map_flags,
1369 false);
1370 }
1371
1372 /* Called from syscall or from eBPF program */
htab_map_delete_elem(struct bpf_map * map,void * key)1373 static int htab_map_delete_elem(struct bpf_map *map, void *key)
1374 {
1375 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1376 struct hlist_nulls_head *head;
1377 struct bucket *b;
1378 struct htab_elem *l;
1379 unsigned long flags;
1380 u32 hash, key_size;
1381 int ret;
1382
1383 WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
1384 !rcu_read_lock_bh_held());
1385
1386 key_size = map->key_size;
1387
1388 hash = htab_map_hash(key, key_size, htab->hashrnd);
1389 b = __select_bucket(htab, hash);
1390 head = &b->head;
1391
1392 ret = htab_lock_bucket(htab, b, hash, &flags);
1393 if (ret)
1394 return ret;
1395
1396 l = lookup_elem_raw(head, hash, key, key_size);
1397
1398 if (l) {
1399 hlist_nulls_del_rcu(&l->hash_node);
1400 free_htab_elem(htab, l);
1401 } else {
1402 ret = -ENOENT;
1403 }
1404
1405 htab_unlock_bucket(htab, b, hash, flags);
1406 return ret;
1407 }
1408
htab_lru_map_delete_elem(struct bpf_map * map,void * key)1409 static int htab_lru_map_delete_elem(struct bpf_map *map, void *key)
1410 {
1411 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1412 struct hlist_nulls_head *head;
1413 struct bucket *b;
1414 struct htab_elem *l;
1415 unsigned long flags;
1416 u32 hash, key_size;
1417 int ret;
1418
1419 WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held() &&
1420 !rcu_read_lock_bh_held());
1421
1422 key_size = map->key_size;
1423
1424 hash = htab_map_hash(key, key_size, htab->hashrnd);
1425 b = __select_bucket(htab, hash);
1426 head = &b->head;
1427
1428 ret = htab_lock_bucket(htab, b, hash, &flags);
1429 if (ret)
1430 return ret;
1431
1432 l = lookup_elem_raw(head, hash, key, key_size);
1433
1434 if (l)
1435 hlist_nulls_del_rcu(&l->hash_node);
1436 else
1437 ret = -ENOENT;
1438
1439 htab_unlock_bucket(htab, b, hash, flags);
1440 if (l)
1441 htab_lru_push_free(htab, l);
1442 return ret;
1443 }
1444
delete_all_elements(struct bpf_htab * htab)1445 static void delete_all_elements(struct bpf_htab *htab)
1446 {
1447 int i;
1448
1449 /* It's called from a worker thread, so disable migration here,
1450 * since bpf_mem_cache_free() relies on that.
1451 */
1452 migrate_disable();
1453 for (i = 0; i < htab->n_buckets; i++) {
1454 struct hlist_nulls_head *head = select_bucket(htab, i);
1455 struct hlist_nulls_node *n;
1456 struct htab_elem *l;
1457
1458 hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
1459 hlist_nulls_del_rcu(&l->hash_node);
1460 htab_elem_free(htab, l);
1461 }
1462 }
1463 migrate_enable();
1464 }
1465
htab_free_malloced_timers(struct bpf_htab * htab)1466 static void htab_free_malloced_timers(struct bpf_htab *htab)
1467 {
1468 int i;
1469
1470 rcu_read_lock();
1471 for (i = 0; i < htab->n_buckets; i++) {
1472 struct hlist_nulls_head *head = select_bucket(htab, i);
1473 struct hlist_nulls_node *n;
1474 struct htab_elem *l;
1475
1476 hlist_nulls_for_each_entry(l, n, head, hash_node) {
1477 /* We don't reset or free kptr on uref dropping to zero,
1478 * hence just free timer.
1479 */
1480 bpf_timer_cancel_and_free(l->key +
1481 round_up(htab->map.key_size, 8) +
1482 htab->map.timer_off);
1483 }
1484 cond_resched_rcu();
1485 }
1486 rcu_read_unlock();
1487 }
1488
htab_map_free_timers(struct bpf_map * map)1489 static void htab_map_free_timers(struct bpf_map *map)
1490 {
1491 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1492
1493 /* We don't reset or free kptr on uref dropping to zero. */
1494 if (!map_value_has_timer(&htab->map))
1495 return;
1496 if (!htab_is_prealloc(htab))
1497 htab_free_malloced_timers(htab);
1498 else
1499 htab_free_prealloced_timers(htab);
1500 }
1501
1502 /* Called when map->refcnt goes to zero, either from workqueue or from syscall */
htab_map_free(struct bpf_map * map)1503 static void htab_map_free(struct bpf_map *map)
1504 {
1505 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1506 int i;
1507
1508 /* bpf_free_used_maps() or close(map_fd) will trigger this map_free callback.
1509 * bpf_free_used_maps() is called after bpf prog is no longer executing.
1510 * There is no need to synchronize_rcu() here to protect map elements.
1511 */
1512
1513 /* htab no longer uses call_rcu() directly. bpf_mem_alloc does it
1514 * underneath and is reponsible for waiting for callbacks to finish
1515 * during bpf_mem_alloc_destroy().
1516 */
1517 if (!htab_is_prealloc(htab)) {
1518 delete_all_elements(htab);
1519 } else {
1520 htab_free_prealloced_kptrs(htab);
1521 prealloc_destroy(htab);
1522 }
1523
1524 bpf_map_free_kptr_off_tab(map);
1525 free_percpu(htab->extra_elems);
1526 bpf_map_area_free(htab->buckets);
1527 bpf_mem_alloc_destroy(&htab->pcpu_ma);
1528 bpf_mem_alloc_destroy(&htab->ma);
1529 if (htab->use_percpu_counter)
1530 percpu_counter_destroy(&htab->pcount);
1531 for (i = 0; i < HASHTAB_MAP_LOCK_COUNT; i++)
1532 free_percpu(htab->map_locked[i]);
1533 lockdep_unregister_key(&htab->lockdep_key);
1534 bpf_map_area_free(htab);
1535 }
1536
htab_map_seq_show_elem(struct bpf_map * map,void * key,struct seq_file * m)1537 static void htab_map_seq_show_elem(struct bpf_map *map, void *key,
1538 struct seq_file *m)
1539 {
1540 void *value;
1541
1542 rcu_read_lock();
1543
1544 value = htab_map_lookup_elem(map, key);
1545 if (!value) {
1546 rcu_read_unlock();
1547 return;
1548 }
1549
1550 btf_type_seq_show(map->btf, map->btf_key_type_id, key, m);
1551 seq_puts(m, ": ");
1552 btf_type_seq_show(map->btf, map->btf_value_type_id, value, m);
1553 seq_puts(m, "\n");
1554
1555 rcu_read_unlock();
1556 }
1557
__htab_map_lookup_and_delete_elem(struct bpf_map * map,void * key,void * value,bool is_lru_map,bool is_percpu,u64 flags)1558 static int __htab_map_lookup_and_delete_elem(struct bpf_map *map, void *key,
1559 void *value, bool is_lru_map,
1560 bool is_percpu, u64 flags)
1561 {
1562 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1563 struct hlist_nulls_head *head;
1564 unsigned long bflags;
1565 struct htab_elem *l;
1566 u32 hash, key_size;
1567 struct bucket *b;
1568 int ret;
1569
1570 key_size = map->key_size;
1571
1572 hash = htab_map_hash(key, key_size, htab->hashrnd);
1573 b = __select_bucket(htab, hash);
1574 head = &b->head;
1575
1576 ret = htab_lock_bucket(htab, b, hash, &bflags);
1577 if (ret)
1578 return ret;
1579
1580 l = lookup_elem_raw(head, hash, key, key_size);
1581 if (!l) {
1582 ret = -ENOENT;
1583 } else {
1584 if (is_percpu) {
1585 u32 roundup_value_size = round_up(map->value_size, 8);
1586 void __percpu *pptr;
1587 int off = 0, cpu;
1588
1589 pptr = htab_elem_get_ptr(l, key_size);
1590 for_each_possible_cpu(cpu) {
1591 bpf_long_memcpy(value + off,
1592 per_cpu_ptr(pptr, cpu),
1593 roundup_value_size);
1594 off += roundup_value_size;
1595 }
1596 } else {
1597 u32 roundup_key_size = round_up(map->key_size, 8);
1598
1599 if (flags & BPF_F_LOCK)
1600 copy_map_value_locked(map, value, l->key +
1601 roundup_key_size,
1602 true);
1603 else
1604 copy_map_value(map, value, l->key +
1605 roundup_key_size);
1606 check_and_init_map_value(map, value);
1607 }
1608
1609 hlist_nulls_del_rcu(&l->hash_node);
1610 if (!is_lru_map)
1611 free_htab_elem(htab, l);
1612 }
1613
1614 htab_unlock_bucket(htab, b, hash, bflags);
1615
1616 if (is_lru_map && l)
1617 htab_lru_push_free(htab, l);
1618
1619 return ret;
1620 }
1621
htab_map_lookup_and_delete_elem(struct bpf_map * map,void * key,void * value,u64 flags)1622 static int htab_map_lookup_and_delete_elem(struct bpf_map *map, void *key,
1623 void *value, u64 flags)
1624 {
1625 return __htab_map_lookup_and_delete_elem(map, key, value, false, false,
1626 flags);
1627 }
1628
htab_percpu_map_lookup_and_delete_elem(struct bpf_map * map,void * key,void * value,u64 flags)1629 static int htab_percpu_map_lookup_and_delete_elem(struct bpf_map *map,
1630 void *key, void *value,
1631 u64 flags)
1632 {
1633 return __htab_map_lookup_and_delete_elem(map, key, value, false, true,
1634 flags);
1635 }
1636
htab_lru_map_lookup_and_delete_elem(struct bpf_map * map,void * key,void * value,u64 flags)1637 static int htab_lru_map_lookup_and_delete_elem(struct bpf_map *map, void *key,
1638 void *value, u64 flags)
1639 {
1640 return __htab_map_lookup_and_delete_elem(map, key, value, true, false,
1641 flags);
1642 }
1643
htab_lru_percpu_map_lookup_and_delete_elem(struct bpf_map * map,void * key,void * value,u64 flags)1644 static int htab_lru_percpu_map_lookup_and_delete_elem(struct bpf_map *map,
1645 void *key, void *value,
1646 u64 flags)
1647 {
1648 return __htab_map_lookup_and_delete_elem(map, key, value, true, true,
1649 flags);
1650 }
1651
1652 static int
__htab_map_lookup_and_delete_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr,bool do_delete,bool is_lru_map,bool is_percpu)1653 __htab_map_lookup_and_delete_batch(struct bpf_map *map,
1654 const union bpf_attr *attr,
1655 union bpf_attr __user *uattr,
1656 bool do_delete, bool is_lru_map,
1657 bool is_percpu)
1658 {
1659 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
1660 u32 bucket_cnt, total, key_size, value_size, roundup_key_size;
1661 void *keys = NULL, *values = NULL, *value, *dst_key, *dst_val;
1662 void __user *uvalues = u64_to_user_ptr(attr->batch.values);
1663 void __user *ukeys = u64_to_user_ptr(attr->batch.keys);
1664 void __user *ubatch = u64_to_user_ptr(attr->batch.in_batch);
1665 u32 batch, max_count, size, bucket_size, map_id;
1666 struct htab_elem *node_to_free = NULL;
1667 u64 elem_map_flags, map_flags;
1668 struct hlist_nulls_head *head;
1669 struct hlist_nulls_node *n;
1670 unsigned long flags = 0;
1671 bool locked = false;
1672 struct htab_elem *l;
1673 struct bucket *b;
1674 int ret = 0;
1675
1676 elem_map_flags = attr->batch.elem_flags;
1677 if ((elem_map_flags & ~BPF_F_LOCK) ||
1678 ((elem_map_flags & BPF_F_LOCK) && !map_value_has_spin_lock(map)))
1679 return -EINVAL;
1680
1681 map_flags = attr->batch.flags;
1682 if (map_flags)
1683 return -EINVAL;
1684
1685 max_count = attr->batch.count;
1686 if (!max_count)
1687 return 0;
1688
1689 if (put_user(0, &uattr->batch.count))
1690 return -EFAULT;
1691
1692 batch = 0;
1693 if (ubatch && copy_from_user(&batch, ubatch, sizeof(batch)))
1694 return -EFAULT;
1695
1696 if (batch >= htab->n_buckets)
1697 return -ENOENT;
1698
1699 key_size = htab->map.key_size;
1700 roundup_key_size = round_up(htab->map.key_size, 8);
1701 value_size = htab->map.value_size;
1702 size = round_up(value_size, 8);
1703 if (is_percpu)
1704 value_size = size * num_possible_cpus();
1705 total = 0;
1706 /* while experimenting with hash tables with sizes ranging from 10 to
1707 * 1000, it was observed that a bucket can have up to 5 entries.
1708 */
1709 bucket_size = 5;
1710
1711 alloc:
1712 /* We cannot do copy_from_user or copy_to_user inside
1713 * the rcu_read_lock. Allocate enough space here.
1714 */
1715 keys = kvmalloc_array(key_size, bucket_size, GFP_USER | __GFP_NOWARN);
1716 values = kvmalloc_array(value_size, bucket_size, GFP_USER | __GFP_NOWARN);
1717 if (!keys || !values) {
1718 ret = -ENOMEM;
1719 goto after_loop;
1720 }
1721
1722 again:
1723 bpf_disable_instrumentation();
1724 rcu_read_lock();
1725 again_nocopy:
1726 dst_key = keys;
1727 dst_val = values;
1728 b = &htab->buckets[batch];
1729 head = &b->head;
1730 /* do not grab the lock unless need it (bucket_cnt > 0). */
1731 if (locked) {
1732 ret = htab_lock_bucket(htab, b, batch, &flags);
1733 if (ret) {
1734 rcu_read_unlock();
1735 bpf_enable_instrumentation();
1736 goto after_loop;
1737 }
1738 }
1739
1740 bucket_cnt = 0;
1741 hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
1742 bucket_cnt++;
1743
1744 if (bucket_cnt && !locked) {
1745 locked = true;
1746 goto again_nocopy;
1747 }
1748
1749 if (bucket_cnt > (max_count - total)) {
1750 if (total == 0)
1751 ret = -ENOSPC;
1752 /* Note that since bucket_cnt > 0 here, it is implicit
1753 * that the locked was grabbed, so release it.
1754 */
1755 htab_unlock_bucket(htab, b, batch, flags);
1756 rcu_read_unlock();
1757 bpf_enable_instrumentation();
1758 goto after_loop;
1759 }
1760
1761 if (bucket_cnt > bucket_size) {
1762 bucket_size = bucket_cnt;
1763 /* Note that since bucket_cnt > 0 here, it is implicit
1764 * that the locked was grabbed, so release it.
1765 */
1766 htab_unlock_bucket(htab, b, batch, flags);
1767 rcu_read_unlock();
1768 bpf_enable_instrumentation();
1769 kvfree(keys);
1770 kvfree(values);
1771 goto alloc;
1772 }
1773
1774 /* Next block is only safe to run if you have grabbed the lock */
1775 if (!locked)
1776 goto next_batch;
1777
1778 hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
1779 memcpy(dst_key, l->key, key_size);
1780
1781 if (is_percpu) {
1782 int off = 0, cpu;
1783 void __percpu *pptr;
1784
1785 pptr = htab_elem_get_ptr(l, map->key_size);
1786 for_each_possible_cpu(cpu) {
1787 bpf_long_memcpy(dst_val + off,
1788 per_cpu_ptr(pptr, cpu), size);
1789 off += size;
1790 }
1791 } else {
1792 value = l->key + roundup_key_size;
1793 if (map->map_type == BPF_MAP_TYPE_HASH_OF_MAPS) {
1794 struct bpf_map **inner_map = value;
1795
1796 /* Actual value is the id of the inner map */
1797 map_id = map->ops->map_fd_sys_lookup_elem(*inner_map);
1798 value = &map_id;
1799 }
1800
1801 if (elem_map_flags & BPF_F_LOCK)
1802 copy_map_value_locked(map, dst_val, value,
1803 true);
1804 else
1805 copy_map_value(map, dst_val, value);
1806 check_and_init_map_value(map, dst_val);
1807 }
1808 if (do_delete) {
1809 hlist_nulls_del_rcu(&l->hash_node);
1810
1811 /* bpf_lru_push_free() will acquire lru_lock, which
1812 * may cause deadlock. See comments in function
1813 * prealloc_lru_pop(). Let us do bpf_lru_push_free()
1814 * after releasing the bucket lock.
1815 */
1816 if (is_lru_map) {
1817 l->batch_flink = node_to_free;
1818 node_to_free = l;
1819 } else {
1820 free_htab_elem(htab, l);
1821 }
1822 }
1823 dst_key += key_size;
1824 dst_val += value_size;
1825 }
1826
1827 htab_unlock_bucket(htab, b, batch, flags);
1828 locked = false;
1829
1830 while (node_to_free) {
1831 l = node_to_free;
1832 node_to_free = node_to_free->batch_flink;
1833 htab_lru_push_free(htab, l);
1834 }
1835
1836 next_batch:
1837 /* If we are not copying data, we can go to next bucket and avoid
1838 * unlocking the rcu.
1839 */
1840 if (!bucket_cnt && (batch + 1 < htab->n_buckets)) {
1841 batch++;
1842 goto again_nocopy;
1843 }
1844
1845 rcu_read_unlock();
1846 bpf_enable_instrumentation();
1847 if (bucket_cnt && (copy_to_user(ukeys + total * key_size, keys,
1848 key_size * bucket_cnt) ||
1849 copy_to_user(uvalues + total * value_size, values,
1850 value_size * bucket_cnt))) {
1851 ret = -EFAULT;
1852 goto after_loop;
1853 }
1854
1855 total += bucket_cnt;
1856 batch++;
1857 if (batch >= htab->n_buckets) {
1858 ret = -ENOENT;
1859 goto after_loop;
1860 }
1861 goto again;
1862
1863 after_loop:
1864 if (ret == -EFAULT)
1865 goto out;
1866
1867 /* copy # of entries and next batch */
1868 ubatch = u64_to_user_ptr(attr->batch.out_batch);
1869 if (copy_to_user(ubatch, &batch, sizeof(batch)) ||
1870 put_user(total, &uattr->batch.count))
1871 ret = -EFAULT;
1872
1873 out:
1874 kvfree(keys);
1875 kvfree(values);
1876 return ret;
1877 }
1878
1879 static int
htab_percpu_map_lookup_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1880 htab_percpu_map_lookup_batch(struct bpf_map *map, const union bpf_attr *attr,
1881 union bpf_attr __user *uattr)
1882 {
1883 return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
1884 false, true);
1885 }
1886
1887 static int
htab_percpu_map_lookup_and_delete_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1888 htab_percpu_map_lookup_and_delete_batch(struct bpf_map *map,
1889 const union bpf_attr *attr,
1890 union bpf_attr __user *uattr)
1891 {
1892 return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
1893 false, true);
1894 }
1895
1896 static int
htab_map_lookup_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1897 htab_map_lookup_batch(struct bpf_map *map, const union bpf_attr *attr,
1898 union bpf_attr __user *uattr)
1899 {
1900 return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
1901 false, false);
1902 }
1903
1904 static int
htab_map_lookup_and_delete_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1905 htab_map_lookup_and_delete_batch(struct bpf_map *map,
1906 const union bpf_attr *attr,
1907 union bpf_attr __user *uattr)
1908 {
1909 return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
1910 false, false);
1911 }
1912
1913 static int
htab_lru_percpu_map_lookup_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1914 htab_lru_percpu_map_lookup_batch(struct bpf_map *map,
1915 const union bpf_attr *attr,
1916 union bpf_attr __user *uattr)
1917 {
1918 return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
1919 true, true);
1920 }
1921
1922 static int
htab_lru_percpu_map_lookup_and_delete_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1923 htab_lru_percpu_map_lookup_and_delete_batch(struct bpf_map *map,
1924 const union bpf_attr *attr,
1925 union bpf_attr __user *uattr)
1926 {
1927 return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
1928 true, true);
1929 }
1930
1931 static int
htab_lru_map_lookup_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1932 htab_lru_map_lookup_batch(struct bpf_map *map, const union bpf_attr *attr,
1933 union bpf_attr __user *uattr)
1934 {
1935 return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
1936 true, false);
1937 }
1938
1939 static int
htab_lru_map_lookup_and_delete_batch(struct bpf_map * map,const union bpf_attr * attr,union bpf_attr __user * uattr)1940 htab_lru_map_lookup_and_delete_batch(struct bpf_map *map,
1941 const union bpf_attr *attr,
1942 union bpf_attr __user *uattr)
1943 {
1944 return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
1945 true, false);
1946 }
1947
1948 struct bpf_iter_seq_hash_map_info {
1949 struct bpf_map *map;
1950 struct bpf_htab *htab;
1951 void *percpu_value_buf; // non-zero means percpu hash
1952 u32 bucket_id;
1953 u32 skip_elems;
1954 };
1955
1956 static struct htab_elem *
bpf_hash_map_seq_find_next(struct bpf_iter_seq_hash_map_info * info,struct htab_elem * prev_elem)1957 bpf_hash_map_seq_find_next(struct bpf_iter_seq_hash_map_info *info,
1958 struct htab_elem *prev_elem)
1959 {
1960 const struct bpf_htab *htab = info->htab;
1961 u32 skip_elems = info->skip_elems;
1962 u32 bucket_id = info->bucket_id;
1963 struct hlist_nulls_head *head;
1964 struct hlist_nulls_node *n;
1965 struct htab_elem *elem;
1966 struct bucket *b;
1967 u32 i, count;
1968
1969 if (bucket_id >= htab->n_buckets)
1970 return NULL;
1971
1972 /* try to find next elem in the same bucket */
1973 if (prev_elem) {
1974 /* no update/deletion on this bucket, prev_elem should be still valid
1975 * and we won't skip elements.
1976 */
1977 n = rcu_dereference_raw(hlist_nulls_next_rcu(&prev_elem->hash_node));
1978 elem = hlist_nulls_entry_safe(n, struct htab_elem, hash_node);
1979 if (elem)
1980 return elem;
1981
1982 /* not found, unlock and go to the next bucket */
1983 b = &htab->buckets[bucket_id++];
1984 rcu_read_unlock();
1985 skip_elems = 0;
1986 }
1987
1988 for (i = bucket_id; i < htab->n_buckets; i++) {
1989 b = &htab->buckets[i];
1990 rcu_read_lock();
1991
1992 count = 0;
1993 head = &b->head;
1994 hlist_nulls_for_each_entry_rcu(elem, n, head, hash_node) {
1995 if (count >= skip_elems) {
1996 info->bucket_id = i;
1997 info->skip_elems = count;
1998 return elem;
1999 }
2000 count++;
2001 }
2002
2003 rcu_read_unlock();
2004 skip_elems = 0;
2005 }
2006
2007 info->bucket_id = i;
2008 info->skip_elems = 0;
2009 return NULL;
2010 }
2011
bpf_hash_map_seq_start(struct seq_file * seq,loff_t * pos)2012 static void *bpf_hash_map_seq_start(struct seq_file *seq, loff_t *pos)
2013 {
2014 struct bpf_iter_seq_hash_map_info *info = seq->private;
2015 struct htab_elem *elem;
2016
2017 elem = bpf_hash_map_seq_find_next(info, NULL);
2018 if (!elem)
2019 return NULL;
2020
2021 if (*pos == 0)
2022 ++*pos;
2023 return elem;
2024 }
2025
bpf_hash_map_seq_next(struct seq_file * seq,void * v,loff_t * pos)2026 static void *bpf_hash_map_seq_next(struct seq_file *seq, void *v, loff_t *pos)
2027 {
2028 struct bpf_iter_seq_hash_map_info *info = seq->private;
2029
2030 ++*pos;
2031 ++info->skip_elems;
2032 return bpf_hash_map_seq_find_next(info, v);
2033 }
2034
__bpf_hash_map_seq_show(struct seq_file * seq,struct htab_elem * elem)2035 static int __bpf_hash_map_seq_show(struct seq_file *seq, struct htab_elem *elem)
2036 {
2037 struct bpf_iter_seq_hash_map_info *info = seq->private;
2038 u32 roundup_key_size, roundup_value_size;
2039 struct bpf_iter__bpf_map_elem ctx = {};
2040 struct bpf_map *map = info->map;
2041 struct bpf_iter_meta meta;
2042 int ret = 0, off = 0, cpu;
2043 struct bpf_prog *prog;
2044 void __percpu *pptr;
2045
2046 meta.seq = seq;
2047 prog = bpf_iter_get_info(&meta, elem == NULL);
2048 if (prog) {
2049 ctx.meta = &meta;
2050 ctx.map = info->map;
2051 if (elem) {
2052 roundup_key_size = round_up(map->key_size, 8);
2053 ctx.key = elem->key;
2054 if (!info->percpu_value_buf) {
2055 ctx.value = elem->key + roundup_key_size;
2056 } else {
2057 roundup_value_size = round_up(map->value_size, 8);
2058 pptr = htab_elem_get_ptr(elem, map->key_size);
2059 for_each_possible_cpu(cpu) {
2060 bpf_long_memcpy(info->percpu_value_buf + off,
2061 per_cpu_ptr(pptr, cpu),
2062 roundup_value_size);
2063 off += roundup_value_size;
2064 }
2065 ctx.value = info->percpu_value_buf;
2066 }
2067 }
2068 ret = bpf_iter_run_prog(prog, &ctx);
2069 }
2070
2071 return ret;
2072 }
2073
bpf_hash_map_seq_show(struct seq_file * seq,void * v)2074 static int bpf_hash_map_seq_show(struct seq_file *seq, void *v)
2075 {
2076 return __bpf_hash_map_seq_show(seq, v);
2077 }
2078
bpf_hash_map_seq_stop(struct seq_file * seq,void * v)2079 static void bpf_hash_map_seq_stop(struct seq_file *seq, void *v)
2080 {
2081 if (!v)
2082 (void)__bpf_hash_map_seq_show(seq, NULL);
2083 else
2084 rcu_read_unlock();
2085 }
2086
bpf_iter_init_hash_map(void * priv_data,struct bpf_iter_aux_info * aux)2087 static int bpf_iter_init_hash_map(void *priv_data,
2088 struct bpf_iter_aux_info *aux)
2089 {
2090 struct bpf_iter_seq_hash_map_info *seq_info = priv_data;
2091 struct bpf_map *map = aux->map;
2092 void *value_buf;
2093 u32 buf_size;
2094
2095 if (map->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
2096 map->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH) {
2097 buf_size = round_up(map->value_size, 8) * num_possible_cpus();
2098 value_buf = kmalloc(buf_size, GFP_USER | __GFP_NOWARN);
2099 if (!value_buf)
2100 return -ENOMEM;
2101
2102 seq_info->percpu_value_buf = value_buf;
2103 }
2104
2105 bpf_map_inc_with_uref(map);
2106 seq_info->map = map;
2107 seq_info->htab = container_of(map, struct bpf_htab, map);
2108 return 0;
2109 }
2110
bpf_iter_fini_hash_map(void * priv_data)2111 static void bpf_iter_fini_hash_map(void *priv_data)
2112 {
2113 struct bpf_iter_seq_hash_map_info *seq_info = priv_data;
2114
2115 bpf_map_put_with_uref(seq_info->map);
2116 kfree(seq_info->percpu_value_buf);
2117 }
2118
2119 static const struct seq_operations bpf_hash_map_seq_ops = {
2120 .start = bpf_hash_map_seq_start,
2121 .next = bpf_hash_map_seq_next,
2122 .stop = bpf_hash_map_seq_stop,
2123 .show = bpf_hash_map_seq_show,
2124 };
2125
2126 static const struct bpf_iter_seq_info iter_seq_info = {
2127 .seq_ops = &bpf_hash_map_seq_ops,
2128 .init_seq_private = bpf_iter_init_hash_map,
2129 .fini_seq_private = bpf_iter_fini_hash_map,
2130 .seq_priv_size = sizeof(struct bpf_iter_seq_hash_map_info),
2131 };
2132
bpf_for_each_hash_elem(struct bpf_map * map,bpf_callback_t callback_fn,void * callback_ctx,u64 flags)2133 static int bpf_for_each_hash_elem(struct bpf_map *map, bpf_callback_t callback_fn,
2134 void *callback_ctx, u64 flags)
2135 {
2136 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
2137 struct hlist_nulls_head *head;
2138 struct hlist_nulls_node *n;
2139 struct htab_elem *elem;
2140 u32 roundup_key_size;
2141 int i, num_elems = 0;
2142 void __percpu *pptr;
2143 struct bucket *b;
2144 void *key, *val;
2145 bool is_percpu;
2146 u64 ret = 0;
2147
2148 if (flags != 0)
2149 return -EINVAL;
2150
2151 is_percpu = htab_is_percpu(htab);
2152
2153 roundup_key_size = round_up(map->key_size, 8);
2154 /* disable migration so percpu value prepared here will be the
2155 * same as the one seen by the bpf program with bpf_map_lookup_elem().
2156 */
2157 if (is_percpu)
2158 migrate_disable();
2159 for (i = 0; i < htab->n_buckets; i++) {
2160 b = &htab->buckets[i];
2161 rcu_read_lock();
2162 head = &b->head;
2163 hlist_nulls_for_each_entry_rcu(elem, n, head, hash_node) {
2164 key = elem->key;
2165 if (is_percpu) {
2166 /* current cpu value for percpu map */
2167 pptr = htab_elem_get_ptr(elem, map->key_size);
2168 val = this_cpu_ptr(pptr);
2169 } else {
2170 val = elem->key + roundup_key_size;
2171 }
2172 num_elems++;
2173 ret = callback_fn((u64)(long)map, (u64)(long)key,
2174 (u64)(long)val, (u64)(long)callback_ctx, 0);
2175 /* return value: 0 - continue, 1 - stop and return */
2176 if (ret) {
2177 rcu_read_unlock();
2178 goto out;
2179 }
2180 }
2181 rcu_read_unlock();
2182 }
2183 out:
2184 if (is_percpu)
2185 migrate_enable();
2186 return num_elems;
2187 }
2188
2189 BTF_ID_LIST_SINGLE(htab_map_btf_ids, struct, bpf_htab)
2190 const struct bpf_map_ops htab_map_ops = {
2191 .map_meta_equal = bpf_map_meta_equal,
2192 .map_alloc_check = htab_map_alloc_check,
2193 .map_alloc = htab_map_alloc,
2194 .map_free = htab_map_free,
2195 .map_get_next_key = htab_map_get_next_key,
2196 .map_release_uref = htab_map_free_timers,
2197 .map_lookup_elem = htab_map_lookup_elem,
2198 .map_lookup_and_delete_elem = htab_map_lookup_and_delete_elem,
2199 .map_update_elem = htab_map_update_elem,
2200 .map_delete_elem = htab_map_delete_elem,
2201 .map_gen_lookup = htab_map_gen_lookup,
2202 .map_seq_show_elem = htab_map_seq_show_elem,
2203 .map_set_for_each_callback_args = map_set_for_each_callback_args,
2204 .map_for_each_callback = bpf_for_each_hash_elem,
2205 BATCH_OPS(htab),
2206 .map_btf_id = &htab_map_btf_ids[0],
2207 .iter_seq_info = &iter_seq_info,
2208 };
2209
2210 const struct bpf_map_ops htab_lru_map_ops = {
2211 .map_meta_equal = bpf_map_meta_equal,
2212 .map_alloc_check = htab_map_alloc_check,
2213 .map_alloc = htab_map_alloc,
2214 .map_free = htab_map_free,
2215 .map_get_next_key = htab_map_get_next_key,
2216 .map_release_uref = htab_map_free_timers,
2217 .map_lookup_elem = htab_lru_map_lookup_elem,
2218 .map_lookup_and_delete_elem = htab_lru_map_lookup_and_delete_elem,
2219 .map_lookup_elem_sys_only = htab_lru_map_lookup_elem_sys,
2220 .map_update_elem = htab_lru_map_update_elem,
2221 .map_delete_elem = htab_lru_map_delete_elem,
2222 .map_gen_lookup = htab_lru_map_gen_lookup,
2223 .map_seq_show_elem = htab_map_seq_show_elem,
2224 .map_set_for_each_callback_args = map_set_for_each_callback_args,
2225 .map_for_each_callback = bpf_for_each_hash_elem,
2226 BATCH_OPS(htab_lru),
2227 .map_btf_id = &htab_map_btf_ids[0],
2228 .iter_seq_info = &iter_seq_info,
2229 };
2230
2231 /* Called from eBPF program */
htab_percpu_map_lookup_elem(struct bpf_map * map,void * key)2232 static void *htab_percpu_map_lookup_elem(struct bpf_map *map, void *key)
2233 {
2234 struct htab_elem *l = __htab_map_lookup_elem(map, key);
2235
2236 if (l)
2237 return this_cpu_ptr(htab_elem_get_ptr(l, map->key_size));
2238 else
2239 return NULL;
2240 }
2241
htab_percpu_map_lookup_percpu_elem(struct bpf_map * map,void * key,u32 cpu)2242 static void *htab_percpu_map_lookup_percpu_elem(struct bpf_map *map, void *key, u32 cpu)
2243 {
2244 struct htab_elem *l;
2245
2246 if (cpu >= nr_cpu_ids)
2247 return NULL;
2248
2249 l = __htab_map_lookup_elem(map, key);
2250 if (l)
2251 return per_cpu_ptr(htab_elem_get_ptr(l, map->key_size), cpu);
2252 else
2253 return NULL;
2254 }
2255
htab_lru_percpu_map_lookup_elem(struct bpf_map * map,void * key)2256 static void *htab_lru_percpu_map_lookup_elem(struct bpf_map *map, void *key)
2257 {
2258 struct htab_elem *l = __htab_map_lookup_elem(map, key);
2259
2260 if (l) {
2261 bpf_lru_node_set_ref(&l->lru_node);
2262 return this_cpu_ptr(htab_elem_get_ptr(l, map->key_size));
2263 }
2264
2265 return NULL;
2266 }
2267
htab_lru_percpu_map_lookup_percpu_elem(struct bpf_map * map,void * key,u32 cpu)2268 static void *htab_lru_percpu_map_lookup_percpu_elem(struct bpf_map *map, void *key, u32 cpu)
2269 {
2270 struct htab_elem *l;
2271
2272 if (cpu >= nr_cpu_ids)
2273 return NULL;
2274
2275 l = __htab_map_lookup_elem(map, key);
2276 if (l) {
2277 bpf_lru_node_set_ref(&l->lru_node);
2278 return per_cpu_ptr(htab_elem_get_ptr(l, map->key_size), cpu);
2279 }
2280
2281 return NULL;
2282 }
2283
bpf_percpu_hash_copy(struct bpf_map * map,void * key,void * value)2284 int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value)
2285 {
2286 struct htab_elem *l;
2287 void __percpu *pptr;
2288 int ret = -ENOENT;
2289 int cpu, off = 0;
2290 u32 size;
2291
2292 /* per_cpu areas are zero-filled and bpf programs can only
2293 * access 'value_size' of them, so copying rounded areas
2294 * will not leak any kernel data
2295 */
2296 size = round_up(map->value_size, 8);
2297 rcu_read_lock();
2298 l = __htab_map_lookup_elem(map, key);
2299 if (!l)
2300 goto out;
2301 /* We do not mark LRU map element here in order to not mess up
2302 * eviction heuristics when user space does a map walk.
2303 */
2304 pptr = htab_elem_get_ptr(l, map->key_size);
2305 for_each_possible_cpu(cpu) {
2306 bpf_long_memcpy(value + off,
2307 per_cpu_ptr(pptr, cpu), size);
2308 off += size;
2309 }
2310 ret = 0;
2311 out:
2312 rcu_read_unlock();
2313 return ret;
2314 }
2315
bpf_percpu_hash_update(struct bpf_map * map,void * key,void * value,u64 map_flags)2316 int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value,
2317 u64 map_flags)
2318 {
2319 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
2320 int ret;
2321
2322 rcu_read_lock();
2323 if (htab_is_lru(htab))
2324 ret = __htab_lru_percpu_map_update_elem(map, key, value,
2325 map_flags, true);
2326 else
2327 ret = __htab_percpu_map_update_elem(map, key, value, map_flags,
2328 true);
2329 rcu_read_unlock();
2330
2331 return ret;
2332 }
2333
htab_percpu_map_seq_show_elem(struct bpf_map * map,void * key,struct seq_file * m)2334 static void htab_percpu_map_seq_show_elem(struct bpf_map *map, void *key,
2335 struct seq_file *m)
2336 {
2337 struct htab_elem *l;
2338 void __percpu *pptr;
2339 int cpu;
2340
2341 rcu_read_lock();
2342
2343 l = __htab_map_lookup_elem(map, key);
2344 if (!l) {
2345 rcu_read_unlock();
2346 return;
2347 }
2348
2349 btf_type_seq_show(map->btf, map->btf_key_type_id, key, m);
2350 seq_puts(m, ": {\n");
2351 pptr = htab_elem_get_ptr(l, map->key_size);
2352 for_each_possible_cpu(cpu) {
2353 seq_printf(m, "\tcpu%d: ", cpu);
2354 btf_type_seq_show(map->btf, map->btf_value_type_id,
2355 per_cpu_ptr(pptr, cpu), m);
2356 seq_puts(m, "\n");
2357 }
2358 seq_puts(m, "}\n");
2359
2360 rcu_read_unlock();
2361 }
2362
2363 const struct bpf_map_ops htab_percpu_map_ops = {
2364 .map_meta_equal = bpf_map_meta_equal,
2365 .map_alloc_check = htab_map_alloc_check,
2366 .map_alloc = htab_map_alloc,
2367 .map_free = htab_map_free,
2368 .map_get_next_key = htab_map_get_next_key,
2369 .map_lookup_elem = htab_percpu_map_lookup_elem,
2370 .map_lookup_and_delete_elem = htab_percpu_map_lookup_and_delete_elem,
2371 .map_update_elem = htab_percpu_map_update_elem,
2372 .map_delete_elem = htab_map_delete_elem,
2373 .map_lookup_percpu_elem = htab_percpu_map_lookup_percpu_elem,
2374 .map_seq_show_elem = htab_percpu_map_seq_show_elem,
2375 .map_set_for_each_callback_args = map_set_for_each_callback_args,
2376 .map_for_each_callback = bpf_for_each_hash_elem,
2377 BATCH_OPS(htab_percpu),
2378 .map_btf_id = &htab_map_btf_ids[0],
2379 .iter_seq_info = &iter_seq_info,
2380 };
2381
2382 const struct bpf_map_ops htab_lru_percpu_map_ops = {
2383 .map_meta_equal = bpf_map_meta_equal,
2384 .map_alloc_check = htab_map_alloc_check,
2385 .map_alloc = htab_map_alloc,
2386 .map_free = htab_map_free,
2387 .map_get_next_key = htab_map_get_next_key,
2388 .map_lookup_elem = htab_lru_percpu_map_lookup_elem,
2389 .map_lookup_and_delete_elem = htab_lru_percpu_map_lookup_and_delete_elem,
2390 .map_update_elem = htab_lru_percpu_map_update_elem,
2391 .map_delete_elem = htab_lru_map_delete_elem,
2392 .map_lookup_percpu_elem = htab_lru_percpu_map_lookup_percpu_elem,
2393 .map_seq_show_elem = htab_percpu_map_seq_show_elem,
2394 .map_set_for_each_callback_args = map_set_for_each_callback_args,
2395 .map_for_each_callback = bpf_for_each_hash_elem,
2396 BATCH_OPS(htab_lru_percpu),
2397 .map_btf_id = &htab_map_btf_ids[0],
2398 .iter_seq_info = &iter_seq_info,
2399 };
2400
fd_htab_map_alloc_check(union bpf_attr * attr)2401 static int fd_htab_map_alloc_check(union bpf_attr *attr)
2402 {
2403 if (attr->value_size != sizeof(u32))
2404 return -EINVAL;
2405 return htab_map_alloc_check(attr);
2406 }
2407
fd_htab_map_free(struct bpf_map * map)2408 static void fd_htab_map_free(struct bpf_map *map)
2409 {
2410 struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
2411 struct hlist_nulls_node *n;
2412 struct hlist_nulls_head *head;
2413 struct htab_elem *l;
2414 int i;
2415
2416 for (i = 0; i < htab->n_buckets; i++) {
2417 head = select_bucket(htab, i);
2418
2419 hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
2420 void *ptr = fd_htab_map_get_ptr(map, l);
2421
2422 map->ops->map_fd_put_ptr(ptr);
2423 }
2424 }
2425
2426 htab_map_free(map);
2427 }
2428
2429 /* only called from syscall */
bpf_fd_htab_map_lookup_elem(struct bpf_map * map,void * key,u32 * value)2430 int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value)
2431 {
2432 void **ptr;
2433 int ret = 0;
2434
2435 if (!map->ops->map_fd_sys_lookup_elem)
2436 return -ENOTSUPP;
2437
2438 rcu_read_lock();
2439 ptr = htab_map_lookup_elem(map, key);
2440 if (ptr)
2441 *value = map->ops->map_fd_sys_lookup_elem(READ_ONCE(*ptr));
2442 else
2443 ret = -ENOENT;
2444 rcu_read_unlock();
2445
2446 return ret;
2447 }
2448
2449 /* only called from syscall */
bpf_fd_htab_map_update_elem(struct bpf_map * map,struct file * map_file,void * key,void * value,u64 map_flags)2450 int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file,
2451 void *key, void *value, u64 map_flags)
2452 {
2453 void *ptr;
2454 int ret;
2455 u32 ufd = *(u32 *)value;
2456
2457 ptr = map->ops->map_fd_get_ptr(map, map_file, ufd);
2458 if (IS_ERR(ptr))
2459 return PTR_ERR(ptr);
2460
2461 ret = htab_map_update_elem(map, key, &ptr, map_flags);
2462 if (ret)
2463 map->ops->map_fd_put_ptr(ptr);
2464
2465 return ret;
2466 }
2467
htab_of_map_alloc(union bpf_attr * attr)2468 static struct bpf_map *htab_of_map_alloc(union bpf_attr *attr)
2469 {
2470 struct bpf_map *map, *inner_map_meta;
2471
2472 inner_map_meta = bpf_map_meta_alloc(attr->inner_map_fd);
2473 if (IS_ERR(inner_map_meta))
2474 return inner_map_meta;
2475
2476 map = htab_map_alloc(attr);
2477 if (IS_ERR(map)) {
2478 bpf_map_meta_free(inner_map_meta);
2479 return map;
2480 }
2481
2482 map->inner_map_meta = inner_map_meta;
2483
2484 return map;
2485 }
2486
htab_of_map_lookup_elem(struct bpf_map * map,void * key)2487 static void *htab_of_map_lookup_elem(struct bpf_map *map, void *key)
2488 {
2489 struct bpf_map **inner_map = htab_map_lookup_elem(map, key);
2490
2491 if (!inner_map)
2492 return NULL;
2493
2494 return READ_ONCE(*inner_map);
2495 }
2496
htab_of_map_gen_lookup(struct bpf_map * map,struct bpf_insn * insn_buf)2497 static int htab_of_map_gen_lookup(struct bpf_map *map,
2498 struct bpf_insn *insn_buf)
2499 {
2500 struct bpf_insn *insn = insn_buf;
2501 const int ret = BPF_REG_0;
2502
2503 BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
2504 (void *(*)(struct bpf_map *map, void *key))NULL));
2505 *insn++ = BPF_EMIT_CALL(__htab_map_lookup_elem);
2506 *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 2);
2507 *insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
2508 offsetof(struct htab_elem, key) +
2509 round_up(map->key_size, 8));
2510 *insn++ = BPF_LDX_MEM(BPF_DW, ret, ret, 0);
2511
2512 return insn - insn_buf;
2513 }
2514
htab_of_map_free(struct bpf_map * map)2515 static void htab_of_map_free(struct bpf_map *map)
2516 {
2517 bpf_map_meta_free(map->inner_map_meta);
2518 fd_htab_map_free(map);
2519 }
2520
2521 const struct bpf_map_ops htab_of_maps_map_ops = {
2522 .map_alloc_check = fd_htab_map_alloc_check,
2523 .map_alloc = htab_of_map_alloc,
2524 .map_free = htab_of_map_free,
2525 .map_get_next_key = htab_map_get_next_key,
2526 .map_lookup_elem = htab_of_map_lookup_elem,
2527 .map_delete_elem = htab_map_delete_elem,
2528 .map_fd_get_ptr = bpf_map_fd_get_ptr,
2529 .map_fd_put_ptr = bpf_map_fd_put_ptr,
2530 .map_fd_sys_lookup_elem = bpf_map_fd_sys_lookup_elem,
2531 .map_gen_lookup = htab_of_map_gen_lookup,
2532 .map_check_btf = map_check_no_btf,
2533 BATCH_OPS(htab),
2534 .map_btf_id = &htab_map_btf_ids[0],
2535 };
2536