xref: /linux/mm/list_lru.c (revision 088e88be5a380cc4e81963a9a02815da465d144f)
1 // SPDX-License-Identifier: GPL-2.0-only
2 /*
3  * Copyright (c) 2013 Red Hat, Inc. and Parallels Inc. All rights reserved.
4  * Authors: David Chinner and Glauber Costa
5  *
6  * Generic LRU infrastructure
7  */
8 #include <linux/kernel.h>
9 #include <linux/module.h>
10 #include <linux/mm.h>
11 #include <linux/list_lru.h>
12 #include <linux/slab.h>
13 #include <linux/mutex.h>
14 #include <linux/memcontrol.h>
15 #include "slab.h"
16 
17 #ifdef CONFIG_MEMCG_KMEM
18 static LIST_HEAD(list_lrus);
19 static DEFINE_MUTEX(list_lrus_mutex);
20 
21 static void list_lru_register(struct list_lru *lru)
22 {
23 	mutex_lock(&list_lrus_mutex);
24 	list_add(&lru->list, &list_lrus);
25 	mutex_unlock(&list_lrus_mutex);
26 }
27 
28 static void list_lru_unregister(struct list_lru *lru)
29 {
30 	mutex_lock(&list_lrus_mutex);
31 	list_del(&lru->list);
32 	mutex_unlock(&list_lrus_mutex);
33 }
34 
35 static int lru_shrinker_id(struct list_lru *lru)
36 {
37 	return lru->shrinker_id;
38 }
39 
40 static inline bool list_lru_memcg_aware(struct list_lru *lru)
41 {
42 	return lru->memcg_aware;
43 }
44 
45 static inline struct list_lru_one *
46 list_lru_from_memcg_idx(struct list_lru_node *nlru, int idx)
47 {
48 	struct list_lru_memcg *memcg_lrus;
49 	/*
50 	 * Either lock or RCU protects the array of per cgroup lists
51 	 * from relocation (see memcg_update_list_lru_node).
52 	 */
53 	memcg_lrus = rcu_dereference_check(nlru->memcg_lrus,
54 					   lockdep_is_held(&nlru->lock));
55 	if (memcg_lrus && idx >= 0)
56 		return memcg_lrus->lru[idx];
57 	return &nlru->lru;
58 }
59 
60 static __always_inline struct mem_cgroup *mem_cgroup_from_kmem(void *ptr)
61 {
62 	struct page *page;
63 
64 	if (!memcg_kmem_enabled())
65 		return NULL;
66 	page = virt_to_head_page(ptr);
67 	return memcg_from_slab_page(page);
68 }
69 
70 static inline struct list_lru_one *
71 list_lru_from_kmem(struct list_lru_node *nlru, void *ptr,
72 		   struct mem_cgroup **memcg_ptr)
73 {
74 	struct list_lru_one *l = &nlru->lru;
75 	struct mem_cgroup *memcg = NULL;
76 
77 	if (!nlru->memcg_lrus)
78 		goto out;
79 
80 	memcg = mem_cgroup_from_kmem(ptr);
81 	if (!memcg)
82 		goto out;
83 
84 	l = list_lru_from_memcg_idx(nlru, memcg_cache_id(memcg));
85 out:
86 	if (memcg_ptr)
87 		*memcg_ptr = memcg;
88 	return l;
89 }
90 #else
91 static void list_lru_register(struct list_lru *lru)
92 {
93 }
94 
95 static void list_lru_unregister(struct list_lru *lru)
96 {
97 }
98 
99 static int lru_shrinker_id(struct list_lru *lru)
100 {
101 	return -1;
102 }
103 
104 static inline bool list_lru_memcg_aware(struct list_lru *lru)
105 {
106 	return false;
107 }
108 
109 static inline struct list_lru_one *
110 list_lru_from_memcg_idx(struct list_lru_node *nlru, int idx)
111 {
112 	return &nlru->lru;
113 }
114 
115 static inline struct list_lru_one *
116 list_lru_from_kmem(struct list_lru_node *nlru, void *ptr,
117 		   struct mem_cgroup **memcg_ptr)
118 {
119 	if (memcg_ptr)
120 		*memcg_ptr = NULL;
121 	return &nlru->lru;
122 }
123 #endif /* CONFIG_MEMCG_KMEM */
124 
125 bool list_lru_add(struct list_lru *lru, struct list_head *item)
126 {
127 	int nid = page_to_nid(virt_to_page(item));
128 	struct list_lru_node *nlru = &lru->node[nid];
129 	struct mem_cgroup *memcg;
130 	struct list_lru_one *l;
131 
132 	spin_lock(&nlru->lock);
133 	if (list_empty(item)) {
134 		l = list_lru_from_kmem(nlru, item, &memcg);
135 		list_add_tail(item, &l->list);
136 		/* Set shrinker bit if the first element was added */
137 		if (!l->nr_items++)
138 			memcg_set_shrinker_bit(memcg, nid,
139 					       lru_shrinker_id(lru));
140 		nlru->nr_items++;
141 		spin_unlock(&nlru->lock);
142 		return true;
143 	}
144 	spin_unlock(&nlru->lock);
145 	return false;
146 }
147 EXPORT_SYMBOL_GPL(list_lru_add);
148 
149 bool list_lru_del(struct list_lru *lru, struct list_head *item)
150 {
151 	int nid = page_to_nid(virt_to_page(item));
152 	struct list_lru_node *nlru = &lru->node[nid];
153 	struct list_lru_one *l;
154 
155 	spin_lock(&nlru->lock);
156 	if (!list_empty(item)) {
157 		l = list_lru_from_kmem(nlru, item, NULL);
158 		list_del_init(item);
159 		l->nr_items--;
160 		nlru->nr_items--;
161 		spin_unlock(&nlru->lock);
162 		return true;
163 	}
164 	spin_unlock(&nlru->lock);
165 	return false;
166 }
167 EXPORT_SYMBOL_GPL(list_lru_del);
168 
169 void list_lru_isolate(struct list_lru_one *list, struct list_head *item)
170 {
171 	list_del_init(item);
172 	list->nr_items--;
173 }
174 EXPORT_SYMBOL_GPL(list_lru_isolate);
175 
176 void list_lru_isolate_move(struct list_lru_one *list, struct list_head *item,
177 			   struct list_head *head)
178 {
179 	list_move(item, head);
180 	list->nr_items--;
181 }
182 EXPORT_SYMBOL_GPL(list_lru_isolate_move);
183 
184 unsigned long list_lru_count_one(struct list_lru *lru,
185 				 int nid, struct mem_cgroup *memcg)
186 {
187 	struct list_lru_node *nlru = &lru->node[nid];
188 	struct list_lru_one *l;
189 	unsigned long count;
190 
191 	rcu_read_lock();
192 	l = list_lru_from_memcg_idx(nlru, memcg_cache_id(memcg));
193 	count = l->nr_items;
194 	rcu_read_unlock();
195 
196 	return count;
197 }
198 EXPORT_SYMBOL_GPL(list_lru_count_one);
199 
200 unsigned long list_lru_count_node(struct list_lru *lru, int nid)
201 {
202 	struct list_lru_node *nlru;
203 
204 	nlru = &lru->node[nid];
205 	return nlru->nr_items;
206 }
207 EXPORT_SYMBOL_GPL(list_lru_count_node);
208 
209 static unsigned long
210 __list_lru_walk_one(struct list_lru_node *nlru, int memcg_idx,
211 		    list_lru_walk_cb isolate, void *cb_arg,
212 		    unsigned long *nr_to_walk)
213 {
214 
215 	struct list_lru_one *l;
216 	struct list_head *item, *n;
217 	unsigned long isolated = 0;
218 
219 	l = list_lru_from_memcg_idx(nlru, memcg_idx);
220 restart:
221 	list_for_each_safe(item, n, &l->list) {
222 		enum lru_status ret;
223 
224 		/*
225 		 * decrement nr_to_walk first so that we don't livelock if we
226 		 * get stuck on large numbesr of LRU_RETRY items
227 		 */
228 		if (!*nr_to_walk)
229 			break;
230 		--*nr_to_walk;
231 
232 		ret = isolate(item, l, &nlru->lock, cb_arg);
233 		switch (ret) {
234 		case LRU_REMOVED_RETRY:
235 			assert_spin_locked(&nlru->lock);
236 			/* fall through */
237 		case LRU_REMOVED:
238 			isolated++;
239 			nlru->nr_items--;
240 			/*
241 			 * If the lru lock has been dropped, our list
242 			 * traversal is now invalid and so we have to
243 			 * restart from scratch.
244 			 */
245 			if (ret == LRU_REMOVED_RETRY)
246 				goto restart;
247 			break;
248 		case LRU_ROTATE:
249 			list_move_tail(item, &l->list);
250 			break;
251 		case LRU_SKIP:
252 			break;
253 		case LRU_RETRY:
254 			/*
255 			 * The lru lock has been dropped, our list traversal is
256 			 * now invalid and so we have to restart from scratch.
257 			 */
258 			assert_spin_locked(&nlru->lock);
259 			goto restart;
260 		default:
261 			BUG();
262 		}
263 	}
264 	return isolated;
265 }
266 
267 unsigned long
268 list_lru_walk_one(struct list_lru *lru, int nid, struct mem_cgroup *memcg,
269 		  list_lru_walk_cb isolate, void *cb_arg,
270 		  unsigned long *nr_to_walk)
271 {
272 	struct list_lru_node *nlru = &lru->node[nid];
273 	unsigned long ret;
274 
275 	spin_lock(&nlru->lock);
276 	ret = __list_lru_walk_one(nlru, memcg_cache_id(memcg), isolate, cb_arg,
277 				  nr_to_walk);
278 	spin_unlock(&nlru->lock);
279 	return ret;
280 }
281 EXPORT_SYMBOL_GPL(list_lru_walk_one);
282 
283 unsigned long
284 list_lru_walk_one_irq(struct list_lru *lru, int nid, struct mem_cgroup *memcg,
285 		      list_lru_walk_cb isolate, void *cb_arg,
286 		      unsigned long *nr_to_walk)
287 {
288 	struct list_lru_node *nlru = &lru->node[nid];
289 	unsigned long ret;
290 
291 	spin_lock_irq(&nlru->lock);
292 	ret = __list_lru_walk_one(nlru, memcg_cache_id(memcg), isolate, cb_arg,
293 				  nr_to_walk);
294 	spin_unlock_irq(&nlru->lock);
295 	return ret;
296 }
297 
298 unsigned long list_lru_walk_node(struct list_lru *lru, int nid,
299 				 list_lru_walk_cb isolate, void *cb_arg,
300 				 unsigned long *nr_to_walk)
301 {
302 	long isolated = 0;
303 	int memcg_idx;
304 
305 	isolated += list_lru_walk_one(lru, nid, NULL, isolate, cb_arg,
306 				      nr_to_walk);
307 	if (*nr_to_walk > 0 && list_lru_memcg_aware(lru)) {
308 		for_each_memcg_cache_index(memcg_idx) {
309 			struct list_lru_node *nlru = &lru->node[nid];
310 
311 			spin_lock(&nlru->lock);
312 			isolated += __list_lru_walk_one(nlru, memcg_idx,
313 							isolate, cb_arg,
314 							nr_to_walk);
315 			spin_unlock(&nlru->lock);
316 
317 			if (*nr_to_walk <= 0)
318 				break;
319 		}
320 	}
321 	return isolated;
322 }
323 EXPORT_SYMBOL_GPL(list_lru_walk_node);
324 
325 static void init_one_lru(struct list_lru_one *l)
326 {
327 	INIT_LIST_HEAD(&l->list);
328 	l->nr_items = 0;
329 }
330 
331 #ifdef CONFIG_MEMCG_KMEM
332 static void __memcg_destroy_list_lru_node(struct list_lru_memcg *memcg_lrus,
333 					  int begin, int end)
334 {
335 	int i;
336 
337 	for (i = begin; i < end; i++)
338 		kfree(memcg_lrus->lru[i]);
339 }
340 
341 static int __memcg_init_list_lru_node(struct list_lru_memcg *memcg_lrus,
342 				      int begin, int end)
343 {
344 	int i;
345 
346 	for (i = begin; i < end; i++) {
347 		struct list_lru_one *l;
348 
349 		l = kmalloc(sizeof(struct list_lru_one), GFP_KERNEL);
350 		if (!l)
351 			goto fail;
352 
353 		init_one_lru(l);
354 		memcg_lrus->lru[i] = l;
355 	}
356 	return 0;
357 fail:
358 	__memcg_destroy_list_lru_node(memcg_lrus, begin, i);
359 	return -ENOMEM;
360 }
361 
362 static int memcg_init_list_lru_node(struct list_lru_node *nlru)
363 {
364 	struct list_lru_memcg *memcg_lrus;
365 	int size = memcg_nr_cache_ids;
366 
367 	memcg_lrus = kvmalloc(sizeof(*memcg_lrus) +
368 			      size * sizeof(void *), GFP_KERNEL);
369 	if (!memcg_lrus)
370 		return -ENOMEM;
371 
372 	if (__memcg_init_list_lru_node(memcg_lrus, 0, size)) {
373 		kvfree(memcg_lrus);
374 		return -ENOMEM;
375 	}
376 	RCU_INIT_POINTER(nlru->memcg_lrus, memcg_lrus);
377 
378 	return 0;
379 }
380 
381 static void memcg_destroy_list_lru_node(struct list_lru_node *nlru)
382 {
383 	struct list_lru_memcg *memcg_lrus;
384 	/*
385 	 * This is called when shrinker has already been unregistered,
386 	 * and nobody can use it. So, there is no need to use kvfree_rcu().
387 	 */
388 	memcg_lrus = rcu_dereference_protected(nlru->memcg_lrus, true);
389 	__memcg_destroy_list_lru_node(memcg_lrus, 0, memcg_nr_cache_ids);
390 	kvfree(memcg_lrus);
391 }
392 
393 static void kvfree_rcu(struct rcu_head *head)
394 {
395 	struct list_lru_memcg *mlru;
396 
397 	mlru = container_of(head, struct list_lru_memcg, rcu);
398 	kvfree(mlru);
399 }
400 
401 static int memcg_update_list_lru_node(struct list_lru_node *nlru,
402 				      int old_size, int new_size)
403 {
404 	struct list_lru_memcg *old, *new;
405 
406 	BUG_ON(old_size > new_size);
407 
408 	old = rcu_dereference_protected(nlru->memcg_lrus,
409 					lockdep_is_held(&list_lrus_mutex));
410 	new = kvmalloc(sizeof(*new) + new_size * sizeof(void *), GFP_KERNEL);
411 	if (!new)
412 		return -ENOMEM;
413 
414 	if (__memcg_init_list_lru_node(new, old_size, new_size)) {
415 		kvfree(new);
416 		return -ENOMEM;
417 	}
418 
419 	memcpy(&new->lru, &old->lru, old_size * sizeof(void *));
420 
421 	/*
422 	 * The locking below allows readers that hold nlru->lock avoid taking
423 	 * rcu_read_lock (see list_lru_from_memcg_idx).
424 	 *
425 	 * Since list_lru_{add,del} may be called under an IRQ-safe lock,
426 	 * we have to use IRQ-safe primitives here to avoid deadlock.
427 	 */
428 	spin_lock_irq(&nlru->lock);
429 	rcu_assign_pointer(nlru->memcg_lrus, new);
430 	spin_unlock_irq(&nlru->lock);
431 
432 	call_rcu(&old->rcu, kvfree_rcu);
433 	return 0;
434 }
435 
436 static void memcg_cancel_update_list_lru_node(struct list_lru_node *nlru,
437 					      int old_size, int new_size)
438 {
439 	struct list_lru_memcg *memcg_lrus;
440 
441 	memcg_lrus = rcu_dereference_protected(nlru->memcg_lrus,
442 					       lockdep_is_held(&list_lrus_mutex));
443 	/* do not bother shrinking the array back to the old size, because we
444 	 * cannot handle allocation failures here */
445 	__memcg_destroy_list_lru_node(memcg_lrus, old_size, new_size);
446 }
447 
448 static int memcg_init_list_lru(struct list_lru *lru, bool memcg_aware)
449 {
450 	int i;
451 
452 	lru->memcg_aware = memcg_aware;
453 
454 	if (!memcg_aware)
455 		return 0;
456 
457 	for_each_node(i) {
458 		if (memcg_init_list_lru_node(&lru->node[i]))
459 			goto fail;
460 	}
461 	return 0;
462 fail:
463 	for (i = i - 1; i >= 0; i--) {
464 		if (!lru->node[i].memcg_lrus)
465 			continue;
466 		memcg_destroy_list_lru_node(&lru->node[i]);
467 	}
468 	return -ENOMEM;
469 }
470 
471 static void memcg_destroy_list_lru(struct list_lru *lru)
472 {
473 	int i;
474 
475 	if (!list_lru_memcg_aware(lru))
476 		return;
477 
478 	for_each_node(i)
479 		memcg_destroy_list_lru_node(&lru->node[i]);
480 }
481 
482 static int memcg_update_list_lru(struct list_lru *lru,
483 				 int old_size, int new_size)
484 {
485 	int i;
486 
487 	if (!list_lru_memcg_aware(lru))
488 		return 0;
489 
490 	for_each_node(i) {
491 		if (memcg_update_list_lru_node(&lru->node[i],
492 					       old_size, new_size))
493 			goto fail;
494 	}
495 	return 0;
496 fail:
497 	for (i = i - 1; i >= 0; i--) {
498 		if (!lru->node[i].memcg_lrus)
499 			continue;
500 
501 		memcg_cancel_update_list_lru_node(&lru->node[i],
502 						  old_size, new_size);
503 	}
504 	return -ENOMEM;
505 }
506 
507 static void memcg_cancel_update_list_lru(struct list_lru *lru,
508 					 int old_size, int new_size)
509 {
510 	int i;
511 
512 	if (!list_lru_memcg_aware(lru))
513 		return;
514 
515 	for_each_node(i)
516 		memcg_cancel_update_list_lru_node(&lru->node[i],
517 						  old_size, new_size);
518 }
519 
520 int memcg_update_all_list_lrus(int new_size)
521 {
522 	int ret = 0;
523 	struct list_lru *lru;
524 	int old_size = memcg_nr_cache_ids;
525 
526 	mutex_lock(&list_lrus_mutex);
527 	list_for_each_entry(lru, &list_lrus, list) {
528 		ret = memcg_update_list_lru(lru, old_size, new_size);
529 		if (ret)
530 			goto fail;
531 	}
532 out:
533 	mutex_unlock(&list_lrus_mutex);
534 	return ret;
535 fail:
536 	list_for_each_entry_continue_reverse(lru, &list_lrus, list)
537 		memcg_cancel_update_list_lru(lru, old_size, new_size);
538 	goto out;
539 }
540 
541 static void memcg_drain_list_lru_node(struct list_lru *lru, int nid,
542 				      int src_idx, struct mem_cgroup *dst_memcg)
543 {
544 	struct list_lru_node *nlru = &lru->node[nid];
545 	int dst_idx = dst_memcg->kmemcg_id;
546 	struct list_lru_one *src, *dst;
547 	bool set;
548 
549 	/*
550 	 * Since list_lru_{add,del} may be called under an IRQ-safe lock,
551 	 * we have to use IRQ-safe primitives here to avoid deadlock.
552 	 */
553 	spin_lock_irq(&nlru->lock);
554 
555 	src = list_lru_from_memcg_idx(nlru, src_idx);
556 	dst = list_lru_from_memcg_idx(nlru, dst_idx);
557 
558 	list_splice_init(&src->list, &dst->list);
559 	set = (!dst->nr_items && src->nr_items);
560 	dst->nr_items += src->nr_items;
561 	if (set)
562 		memcg_set_shrinker_bit(dst_memcg, nid, lru_shrinker_id(lru));
563 	src->nr_items = 0;
564 
565 	spin_unlock_irq(&nlru->lock);
566 }
567 
568 static void memcg_drain_list_lru(struct list_lru *lru,
569 				 int src_idx, struct mem_cgroup *dst_memcg)
570 {
571 	int i;
572 
573 	if (!list_lru_memcg_aware(lru))
574 		return;
575 
576 	for_each_node(i)
577 		memcg_drain_list_lru_node(lru, i, src_idx, dst_memcg);
578 }
579 
580 void memcg_drain_all_list_lrus(int src_idx, struct mem_cgroup *dst_memcg)
581 {
582 	struct list_lru *lru;
583 
584 	mutex_lock(&list_lrus_mutex);
585 	list_for_each_entry(lru, &list_lrus, list)
586 		memcg_drain_list_lru(lru, src_idx, dst_memcg);
587 	mutex_unlock(&list_lrus_mutex);
588 }
589 #else
590 static int memcg_init_list_lru(struct list_lru *lru, bool memcg_aware)
591 {
592 	return 0;
593 }
594 
595 static void memcg_destroy_list_lru(struct list_lru *lru)
596 {
597 }
598 #endif /* CONFIG_MEMCG_KMEM */
599 
600 int __list_lru_init(struct list_lru *lru, bool memcg_aware,
601 		    struct lock_class_key *key, struct shrinker *shrinker)
602 {
603 	int i;
604 	int err = -ENOMEM;
605 
606 #ifdef CONFIG_MEMCG_KMEM
607 	if (shrinker)
608 		lru->shrinker_id = shrinker->id;
609 	else
610 		lru->shrinker_id = -1;
611 #endif
612 	memcg_get_cache_ids();
613 
614 	lru->node = kcalloc(nr_node_ids, sizeof(*lru->node), GFP_KERNEL);
615 	if (!lru->node)
616 		goto out;
617 
618 	for_each_node(i) {
619 		spin_lock_init(&lru->node[i].lock);
620 		if (key)
621 			lockdep_set_class(&lru->node[i].lock, key);
622 		init_one_lru(&lru->node[i].lru);
623 	}
624 
625 	err = memcg_init_list_lru(lru, memcg_aware);
626 	if (err) {
627 		kfree(lru->node);
628 		/* Do this so a list_lru_destroy() doesn't crash: */
629 		lru->node = NULL;
630 		goto out;
631 	}
632 
633 	list_lru_register(lru);
634 out:
635 	memcg_put_cache_ids();
636 	return err;
637 }
638 EXPORT_SYMBOL_GPL(__list_lru_init);
639 
640 void list_lru_destroy(struct list_lru *lru)
641 {
642 	/* Already destroyed or not yet initialized? */
643 	if (!lru->node)
644 		return;
645 
646 	memcg_get_cache_ids();
647 
648 	list_lru_unregister(lru);
649 
650 	memcg_destroy_list_lru(lru);
651 	kfree(lru->node);
652 	lru->node = NULL;
653 
654 #ifdef CONFIG_MEMCG_KMEM
655 	lru->shrinker_id = -1;
656 #endif
657 	memcg_put_cache_ids();
658 }
659 EXPORT_SYMBOL_GPL(list_lru_destroy);
660