xref: /openbmc/linux/fs/f2fs/extent_cache.c (revision a90fa0ad)
1 // SPDX-License-Identifier: GPL-2.0
2 /*
3  * f2fs extent cache support
4  *
5  * Copyright (c) 2015 Motorola Mobility
6  * Copyright (c) 2015 Samsung Electronics
7  * Authors: Jaegeuk Kim <jaegeuk@kernel.org>
8  *          Chao Yu <chao2.yu@samsung.com>
9  *
10  * block_age-based extent cache added by:
11  * Copyright (c) 2022 xiaomi Co., Ltd.
12  *             http://www.xiaomi.com/
13  */
14 
15 #include <linux/fs.h>
16 #include <linux/f2fs_fs.h>
17 
18 #include "f2fs.h"
19 #include "node.h"
20 #include <trace/events/f2fs.h>
21 
22 static void __set_extent_info(struct extent_info *ei,
23 				unsigned int fofs, unsigned int len,
24 				block_t blk, bool keep_clen,
25 				unsigned long age, unsigned long last_blocks,
26 				enum extent_type type)
27 {
28 	ei->fofs = fofs;
29 	ei->len = len;
30 
31 	if (type == EX_READ) {
32 		ei->blk = blk;
33 		if (keep_clen)
34 			return;
35 #ifdef CONFIG_F2FS_FS_COMPRESSION
36 		ei->c_len = 0;
37 #endif
38 	} else if (type == EX_BLOCK_AGE) {
39 		ei->age = age;
40 		ei->last_blocks = last_blocks;
41 	}
42 }
43 
44 static bool __may_read_extent_tree(struct inode *inode)
45 {
46 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
47 
48 	if (!test_opt(sbi, READ_EXTENT_CACHE))
49 		return false;
50 	if (is_inode_flag_set(inode, FI_NO_EXTENT))
51 		return false;
52 	if (is_inode_flag_set(inode, FI_COMPRESSED_FILE) &&
53 			 !f2fs_sb_has_readonly(sbi))
54 		return false;
55 	return S_ISREG(inode->i_mode);
56 }
57 
58 static bool __may_age_extent_tree(struct inode *inode)
59 {
60 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
61 
62 	if (!test_opt(sbi, AGE_EXTENT_CACHE))
63 		return false;
64 	/* don't cache block age info for cold file */
65 	if (is_inode_flag_set(inode, FI_COMPRESSED_FILE))
66 		return false;
67 	if (file_is_cold(inode))
68 		return false;
69 
70 	return S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode);
71 }
72 
73 static bool __init_may_extent_tree(struct inode *inode, enum extent_type type)
74 {
75 	if (type == EX_READ)
76 		return __may_read_extent_tree(inode);
77 	else if (type == EX_BLOCK_AGE)
78 		return __may_age_extent_tree(inode);
79 	return false;
80 }
81 
82 static bool __may_extent_tree(struct inode *inode, enum extent_type type)
83 {
84 	/*
85 	 * for recovered files during mount do not create extents
86 	 * if shrinker is not registered.
87 	 */
88 	if (list_empty(&F2FS_I_SB(inode)->s_list))
89 		return false;
90 
91 	return __init_may_extent_tree(inode, type);
92 }
93 
94 static void __try_update_largest_extent(struct extent_tree *et,
95 						struct extent_node *en)
96 {
97 	if (et->type != EX_READ)
98 		return;
99 	if (en->ei.len <= et->largest.len)
100 		return;
101 
102 	et->largest = en->ei;
103 	et->largest_updated = true;
104 }
105 
106 static bool __is_extent_mergeable(struct extent_info *back,
107 		struct extent_info *front, enum extent_type type)
108 {
109 	if (type == EX_READ) {
110 #ifdef CONFIG_F2FS_FS_COMPRESSION
111 		if (back->c_len && back->len != back->c_len)
112 			return false;
113 		if (front->c_len && front->len != front->c_len)
114 			return false;
115 #endif
116 		return (back->fofs + back->len == front->fofs &&
117 				back->blk + back->len == front->blk);
118 	} else if (type == EX_BLOCK_AGE) {
119 		return (back->fofs + back->len == front->fofs &&
120 			abs(back->age - front->age) <= SAME_AGE_REGION &&
121 			abs(back->last_blocks - front->last_blocks) <=
122 							SAME_AGE_REGION);
123 	}
124 	return false;
125 }
126 
127 static bool __is_back_mergeable(struct extent_info *cur,
128 		struct extent_info *back, enum extent_type type)
129 {
130 	return __is_extent_mergeable(back, cur, type);
131 }
132 
133 static bool __is_front_mergeable(struct extent_info *cur,
134 		struct extent_info *front, enum extent_type type)
135 {
136 	return __is_extent_mergeable(cur, front, type);
137 }
138 
139 static struct rb_entry *__lookup_rb_tree_fast(struct rb_entry *cached_re,
140 							unsigned int ofs)
141 {
142 	if (cached_re) {
143 		if (cached_re->ofs <= ofs &&
144 				cached_re->ofs + cached_re->len > ofs) {
145 			return cached_re;
146 		}
147 	}
148 	return NULL;
149 }
150 
151 static struct rb_entry *__lookup_rb_tree_slow(struct rb_root_cached *root,
152 							unsigned int ofs)
153 {
154 	struct rb_node *node = root->rb_root.rb_node;
155 	struct rb_entry *re;
156 
157 	while (node) {
158 		re = rb_entry(node, struct rb_entry, rb_node);
159 
160 		if (ofs < re->ofs)
161 			node = node->rb_left;
162 		else if (ofs >= re->ofs + re->len)
163 			node = node->rb_right;
164 		else
165 			return re;
166 	}
167 	return NULL;
168 }
169 
170 struct rb_entry *f2fs_lookup_rb_tree(struct rb_root_cached *root,
171 				struct rb_entry *cached_re, unsigned int ofs)
172 {
173 	struct rb_entry *re;
174 
175 	re = __lookup_rb_tree_fast(cached_re, ofs);
176 	if (!re)
177 		return __lookup_rb_tree_slow(root, ofs);
178 
179 	return re;
180 }
181 
182 struct rb_node **f2fs_lookup_rb_tree_ext(struct f2fs_sb_info *sbi,
183 					struct rb_root_cached *root,
184 					struct rb_node **parent,
185 					unsigned long long key, bool *leftmost)
186 {
187 	struct rb_node **p = &root->rb_root.rb_node;
188 	struct rb_entry *re;
189 
190 	while (*p) {
191 		*parent = *p;
192 		re = rb_entry(*parent, struct rb_entry, rb_node);
193 
194 		if (key < re->key) {
195 			p = &(*p)->rb_left;
196 		} else {
197 			p = &(*p)->rb_right;
198 			*leftmost = false;
199 		}
200 	}
201 
202 	return p;
203 }
204 
205 struct rb_node **f2fs_lookup_rb_tree_for_insert(struct f2fs_sb_info *sbi,
206 				struct rb_root_cached *root,
207 				struct rb_node **parent,
208 				unsigned int ofs, bool *leftmost)
209 {
210 	struct rb_node **p = &root->rb_root.rb_node;
211 	struct rb_entry *re;
212 
213 	while (*p) {
214 		*parent = *p;
215 		re = rb_entry(*parent, struct rb_entry, rb_node);
216 
217 		if (ofs < re->ofs) {
218 			p = &(*p)->rb_left;
219 		} else if (ofs >= re->ofs + re->len) {
220 			p = &(*p)->rb_right;
221 			*leftmost = false;
222 		} else {
223 			f2fs_bug_on(sbi, 1);
224 		}
225 	}
226 
227 	return p;
228 }
229 
230 /*
231  * lookup rb entry in position of @ofs in rb-tree,
232  * if hit, return the entry, otherwise, return NULL
233  * @prev_ex: extent before ofs
234  * @next_ex: extent after ofs
235  * @insert_p: insert point for new extent at ofs
236  * in order to simpfy the insertion after.
237  * tree must stay unchanged between lookup and insertion.
238  */
239 struct rb_entry *f2fs_lookup_rb_tree_ret(struct rb_root_cached *root,
240 				struct rb_entry *cached_re,
241 				unsigned int ofs,
242 				struct rb_entry **prev_entry,
243 				struct rb_entry **next_entry,
244 				struct rb_node ***insert_p,
245 				struct rb_node **insert_parent,
246 				bool force, bool *leftmost)
247 {
248 	struct rb_node **pnode = &root->rb_root.rb_node;
249 	struct rb_node *parent = NULL, *tmp_node;
250 	struct rb_entry *re = cached_re;
251 
252 	*insert_p = NULL;
253 	*insert_parent = NULL;
254 	*prev_entry = NULL;
255 	*next_entry = NULL;
256 
257 	if (RB_EMPTY_ROOT(&root->rb_root))
258 		return NULL;
259 
260 	if (re) {
261 		if (re->ofs <= ofs && re->ofs + re->len > ofs)
262 			goto lookup_neighbors;
263 	}
264 
265 	if (leftmost)
266 		*leftmost = true;
267 
268 	while (*pnode) {
269 		parent = *pnode;
270 		re = rb_entry(*pnode, struct rb_entry, rb_node);
271 
272 		if (ofs < re->ofs) {
273 			pnode = &(*pnode)->rb_left;
274 		} else if (ofs >= re->ofs + re->len) {
275 			pnode = &(*pnode)->rb_right;
276 			if (leftmost)
277 				*leftmost = false;
278 		} else {
279 			goto lookup_neighbors;
280 		}
281 	}
282 
283 	*insert_p = pnode;
284 	*insert_parent = parent;
285 
286 	re = rb_entry(parent, struct rb_entry, rb_node);
287 	tmp_node = parent;
288 	if (parent && ofs > re->ofs)
289 		tmp_node = rb_next(parent);
290 	*next_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node);
291 
292 	tmp_node = parent;
293 	if (parent && ofs < re->ofs)
294 		tmp_node = rb_prev(parent);
295 	*prev_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node);
296 	return NULL;
297 
298 lookup_neighbors:
299 	if (ofs == re->ofs || force) {
300 		/* lookup prev node for merging backward later */
301 		tmp_node = rb_prev(&re->rb_node);
302 		*prev_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node);
303 	}
304 	if (ofs == re->ofs + re->len - 1 || force) {
305 		/* lookup next node for merging frontward later */
306 		tmp_node = rb_next(&re->rb_node);
307 		*next_entry = rb_entry_safe(tmp_node, struct rb_entry, rb_node);
308 	}
309 	return re;
310 }
311 
312 bool f2fs_check_rb_tree_consistence(struct f2fs_sb_info *sbi,
313 				struct rb_root_cached *root, bool check_key)
314 {
315 #ifdef CONFIG_F2FS_CHECK_FS
316 	struct rb_node *cur = rb_first_cached(root), *next;
317 	struct rb_entry *cur_re, *next_re;
318 
319 	if (!cur)
320 		return true;
321 
322 	while (cur) {
323 		next = rb_next(cur);
324 		if (!next)
325 			return true;
326 
327 		cur_re = rb_entry(cur, struct rb_entry, rb_node);
328 		next_re = rb_entry(next, struct rb_entry, rb_node);
329 
330 		if (check_key) {
331 			if (cur_re->key > next_re->key) {
332 				f2fs_info(sbi, "inconsistent rbtree, "
333 					"cur(%llu) next(%llu)",
334 					cur_re->key, next_re->key);
335 				return false;
336 			}
337 			goto next;
338 		}
339 
340 		if (cur_re->ofs + cur_re->len > next_re->ofs) {
341 			f2fs_info(sbi, "inconsistent rbtree, cur(%u, %u) next(%u, %u)",
342 				  cur_re->ofs, cur_re->len,
343 				  next_re->ofs, next_re->len);
344 			return false;
345 		}
346 next:
347 		cur = next;
348 	}
349 #endif
350 	return true;
351 }
352 
353 static struct kmem_cache *extent_tree_slab;
354 static struct kmem_cache *extent_node_slab;
355 
356 static struct extent_node *__attach_extent_node(struct f2fs_sb_info *sbi,
357 				struct extent_tree *et, struct extent_info *ei,
358 				struct rb_node *parent, struct rb_node **p,
359 				bool leftmost)
360 {
361 	struct extent_tree_info *eti = &sbi->extent_tree[et->type];
362 	struct extent_node *en;
363 
364 	en = f2fs_kmem_cache_alloc(extent_node_slab, GFP_ATOMIC, false, sbi);
365 	if (!en)
366 		return NULL;
367 
368 	en->ei = *ei;
369 	INIT_LIST_HEAD(&en->list);
370 	en->et = et;
371 
372 	rb_link_node(&en->rb_node, parent, p);
373 	rb_insert_color_cached(&en->rb_node, &et->root, leftmost);
374 	atomic_inc(&et->node_cnt);
375 	atomic_inc(&eti->total_ext_node);
376 	return en;
377 }
378 
379 static void __detach_extent_node(struct f2fs_sb_info *sbi,
380 				struct extent_tree *et, struct extent_node *en)
381 {
382 	struct extent_tree_info *eti = &sbi->extent_tree[et->type];
383 
384 	rb_erase_cached(&en->rb_node, &et->root);
385 	atomic_dec(&et->node_cnt);
386 	atomic_dec(&eti->total_ext_node);
387 
388 	if (et->cached_en == en)
389 		et->cached_en = NULL;
390 	kmem_cache_free(extent_node_slab, en);
391 }
392 
393 /*
394  * Flow to release an extent_node:
395  * 1. list_del_init
396  * 2. __detach_extent_node
397  * 3. kmem_cache_free.
398  */
399 static void __release_extent_node(struct f2fs_sb_info *sbi,
400 			struct extent_tree *et, struct extent_node *en)
401 {
402 	struct extent_tree_info *eti = &sbi->extent_tree[et->type];
403 
404 	spin_lock(&eti->extent_lock);
405 	f2fs_bug_on(sbi, list_empty(&en->list));
406 	list_del_init(&en->list);
407 	spin_unlock(&eti->extent_lock);
408 
409 	__detach_extent_node(sbi, et, en);
410 }
411 
412 static struct extent_tree *__grab_extent_tree(struct inode *inode,
413 						enum extent_type type)
414 {
415 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
416 	struct extent_tree_info *eti = &sbi->extent_tree[type];
417 	struct extent_tree *et;
418 	nid_t ino = inode->i_ino;
419 
420 	mutex_lock(&eti->extent_tree_lock);
421 	et = radix_tree_lookup(&eti->extent_tree_root, ino);
422 	if (!et) {
423 		et = f2fs_kmem_cache_alloc(extent_tree_slab,
424 					GFP_NOFS, true, NULL);
425 		f2fs_radix_tree_insert(&eti->extent_tree_root, ino, et);
426 		memset(et, 0, sizeof(struct extent_tree));
427 		et->ino = ino;
428 		et->type = type;
429 		et->root = RB_ROOT_CACHED;
430 		et->cached_en = NULL;
431 		rwlock_init(&et->lock);
432 		INIT_LIST_HEAD(&et->list);
433 		atomic_set(&et->node_cnt, 0);
434 		atomic_inc(&eti->total_ext_tree);
435 	} else {
436 		atomic_dec(&eti->total_zombie_tree);
437 		list_del_init(&et->list);
438 	}
439 	mutex_unlock(&eti->extent_tree_lock);
440 
441 	/* never died until evict_inode */
442 	F2FS_I(inode)->extent_tree[type] = et;
443 
444 	return et;
445 }
446 
447 static unsigned int __free_extent_tree(struct f2fs_sb_info *sbi,
448 					struct extent_tree *et)
449 {
450 	struct rb_node *node, *next;
451 	struct extent_node *en;
452 	unsigned int count = atomic_read(&et->node_cnt);
453 
454 	node = rb_first_cached(&et->root);
455 	while (node) {
456 		next = rb_next(node);
457 		en = rb_entry(node, struct extent_node, rb_node);
458 		__release_extent_node(sbi, et, en);
459 		node = next;
460 	}
461 
462 	return count - atomic_read(&et->node_cnt);
463 }
464 
465 static void __drop_largest_extent(struct extent_tree *et,
466 					pgoff_t fofs, unsigned int len)
467 {
468 	if (fofs < et->largest.fofs + et->largest.len &&
469 			fofs + len > et->largest.fofs) {
470 		et->largest.len = 0;
471 		et->largest_updated = true;
472 	}
473 }
474 
475 void f2fs_init_read_extent_tree(struct inode *inode, struct page *ipage)
476 {
477 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
478 	struct extent_tree_info *eti = &sbi->extent_tree[EX_READ];
479 	struct f2fs_extent *i_ext = &F2FS_INODE(ipage)->i_ext;
480 	struct extent_tree *et;
481 	struct extent_node *en;
482 	struct extent_info ei;
483 
484 	if (!__may_extent_tree(inode, EX_READ)) {
485 		/* drop largest read extent */
486 		if (i_ext && i_ext->len) {
487 			f2fs_wait_on_page_writeback(ipage, NODE, true, true);
488 			i_ext->len = 0;
489 			set_page_dirty(ipage);
490 		}
491 		goto out;
492 	}
493 
494 	et = __grab_extent_tree(inode, EX_READ);
495 
496 	if (!i_ext || !i_ext->len)
497 		goto out;
498 
499 	get_read_extent_info(&ei, i_ext);
500 
501 	write_lock(&et->lock);
502 	if (atomic_read(&et->node_cnt))
503 		goto unlock_out;
504 
505 	en = __attach_extent_node(sbi, et, &ei, NULL,
506 				&et->root.rb_root.rb_node, true);
507 	if (en) {
508 		et->largest = en->ei;
509 		et->cached_en = en;
510 
511 		spin_lock(&eti->extent_lock);
512 		list_add_tail(&en->list, &eti->extent_list);
513 		spin_unlock(&eti->extent_lock);
514 	}
515 unlock_out:
516 	write_unlock(&et->lock);
517 out:
518 	if (!F2FS_I(inode)->extent_tree[EX_READ])
519 		set_inode_flag(inode, FI_NO_EXTENT);
520 }
521 
522 void f2fs_init_age_extent_tree(struct inode *inode)
523 {
524 	if (!__init_may_extent_tree(inode, EX_BLOCK_AGE))
525 		return;
526 	__grab_extent_tree(inode, EX_BLOCK_AGE);
527 }
528 
529 void f2fs_init_extent_tree(struct inode *inode)
530 {
531 	/* initialize read cache */
532 	if (__init_may_extent_tree(inode, EX_READ))
533 		__grab_extent_tree(inode, EX_READ);
534 
535 	/* initialize block age cache */
536 	if (__init_may_extent_tree(inode, EX_BLOCK_AGE))
537 		__grab_extent_tree(inode, EX_BLOCK_AGE);
538 }
539 
540 static bool __lookup_extent_tree(struct inode *inode, pgoff_t pgofs,
541 			struct extent_info *ei, enum extent_type type)
542 {
543 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
544 	struct extent_tree_info *eti = &sbi->extent_tree[type];
545 	struct extent_tree *et = F2FS_I(inode)->extent_tree[type];
546 	struct extent_node *en;
547 	bool ret = false;
548 
549 	if (!et)
550 		return false;
551 
552 	trace_f2fs_lookup_extent_tree_start(inode, pgofs, type);
553 
554 	read_lock(&et->lock);
555 
556 	if (type == EX_READ &&
557 			et->largest.fofs <= pgofs &&
558 			et->largest.fofs + et->largest.len > pgofs) {
559 		*ei = et->largest;
560 		ret = true;
561 		stat_inc_largest_node_hit(sbi);
562 		goto out;
563 	}
564 
565 	en = (struct extent_node *)f2fs_lookup_rb_tree(&et->root,
566 				(struct rb_entry *)et->cached_en, pgofs);
567 	if (!en)
568 		goto out;
569 
570 	if (en == et->cached_en)
571 		stat_inc_cached_node_hit(sbi, type);
572 	else
573 		stat_inc_rbtree_node_hit(sbi, type);
574 
575 	*ei = en->ei;
576 	spin_lock(&eti->extent_lock);
577 	if (!list_empty(&en->list)) {
578 		list_move_tail(&en->list, &eti->extent_list);
579 		et->cached_en = en;
580 	}
581 	spin_unlock(&eti->extent_lock);
582 	ret = true;
583 out:
584 	stat_inc_total_hit(sbi, type);
585 	read_unlock(&et->lock);
586 
587 	if (type == EX_READ)
588 		trace_f2fs_lookup_read_extent_tree_end(inode, pgofs, ei);
589 	else if (type == EX_BLOCK_AGE)
590 		trace_f2fs_lookup_age_extent_tree_end(inode, pgofs, ei);
591 	return ret;
592 }
593 
594 static struct extent_node *__try_merge_extent_node(struct f2fs_sb_info *sbi,
595 				struct extent_tree *et, struct extent_info *ei,
596 				struct extent_node *prev_ex,
597 				struct extent_node *next_ex)
598 {
599 	struct extent_tree_info *eti = &sbi->extent_tree[et->type];
600 	struct extent_node *en = NULL;
601 
602 	if (prev_ex && __is_back_mergeable(ei, &prev_ex->ei, et->type)) {
603 		prev_ex->ei.len += ei->len;
604 		ei = &prev_ex->ei;
605 		en = prev_ex;
606 	}
607 
608 	if (next_ex && __is_front_mergeable(ei, &next_ex->ei, et->type)) {
609 		next_ex->ei.fofs = ei->fofs;
610 		next_ex->ei.len += ei->len;
611 		if (et->type == EX_READ)
612 			next_ex->ei.blk = ei->blk;
613 		if (en)
614 			__release_extent_node(sbi, et, prev_ex);
615 
616 		en = next_ex;
617 	}
618 
619 	if (!en)
620 		return NULL;
621 
622 	__try_update_largest_extent(et, en);
623 
624 	spin_lock(&eti->extent_lock);
625 	if (!list_empty(&en->list)) {
626 		list_move_tail(&en->list, &eti->extent_list);
627 		et->cached_en = en;
628 	}
629 	spin_unlock(&eti->extent_lock);
630 	return en;
631 }
632 
633 static struct extent_node *__insert_extent_tree(struct f2fs_sb_info *sbi,
634 				struct extent_tree *et, struct extent_info *ei,
635 				struct rb_node **insert_p,
636 				struct rb_node *insert_parent,
637 				bool leftmost)
638 {
639 	struct extent_tree_info *eti = &sbi->extent_tree[et->type];
640 	struct rb_node **p;
641 	struct rb_node *parent = NULL;
642 	struct extent_node *en = NULL;
643 
644 	if (insert_p && insert_parent) {
645 		parent = insert_parent;
646 		p = insert_p;
647 		goto do_insert;
648 	}
649 
650 	leftmost = true;
651 
652 	p = f2fs_lookup_rb_tree_for_insert(sbi, &et->root, &parent,
653 						ei->fofs, &leftmost);
654 do_insert:
655 	en = __attach_extent_node(sbi, et, ei, parent, p, leftmost);
656 	if (!en)
657 		return NULL;
658 
659 	__try_update_largest_extent(et, en);
660 
661 	/* update in global extent list */
662 	spin_lock(&eti->extent_lock);
663 	list_add_tail(&en->list, &eti->extent_list);
664 	et->cached_en = en;
665 	spin_unlock(&eti->extent_lock);
666 	return en;
667 }
668 
669 static void __update_extent_tree_range(struct inode *inode,
670 			struct extent_info *tei, enum extent_type type)
671 {
672 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
673 	struct extent_tree *et = F2FS_I(inode)->extent_tree[type];
674 	struct extent_node *en = NULL, *en1 = NULL;
675 	struct extent_node *prev_en = NULL, *next_en = NULL;
676 	struct extent_info ei, dei, prev;
677 	struct rb_node **insert_p = NULL, *insert_parent = NULL;
678 	unsigned int fofs = tei->fofs, len = tei->len;
679 	unsigned int end = fofs + len;
680 	bool updated = false;
681 	bool leftmost = false;
682 
683 	if (!et)
684 		return;
685 
686 	if (type == EX_READ)
687 		trace_f2fs_update_read_extent_tree_range(inode, fofs, len,
688 						tei->blk, 0);
689 	else if (type == EX_BLOCK_AGE)
690 		trace_f2fs_update_age_extent_tree_range(inode, fofs, len,
691 						tei->age, tei->last_blocks);
692 
693 	write_lock(&et->lock);
694 
695 	if (type == EX_READ) {
696 		if (is_inode_flag_set(inode, FI_NO_EXTENT)) {
697 			write_unlock(&et->lock);
698 			return;
699 		}
700 
701 		prev = et->largest;
702 		dei.len = 0;
703 
704 		/*
705 		 * drop largest extent before lookup, in case it's already
706 		 * been shrunk from extent tree
707 		 */
708 		__drop_largest_extent(et, fofs, len);
709 	}
710 
711 	/* 1. lookup first extent node in range [fofs, fofs + len - 1] */
712 	en = (struct extent_node *)f2fs_lookup_rb_tree_ret(&et->root,
713 					(struct rb_entry *)et->cached_en, fofs,
714 					(struct rb_entry **)&prev_en,
715 					(struct rb_entry **)&next_en,
716 					&insert_p, &insert_parent, false,
717 					&leftmost);
718 	if (!en)
719 		en = next_en;
720 
721 	/* 2. invlidate all extent nodes in range [fofs, fofs + len - 1] */
722 	while (en && en->ei.fofs < end) {
723 		unsigned int org_end;
724 		int parts = 0;	/* # of parts current extent split into */
725 
726 		next_en = en1 = NULL;
727 
728 		dei = en->ei;
729 		org_end = dei.fofs + dei.len;
730 		f2fs_bug_on(sbi, fofs >= org_end);
731 
732 		if (fofs > dei.fofs && (type != EX_READ ||
733 				fofs - dei.fofs >= F2FS_MIN_EXTENT_LEN)) {
734 			en->ei.len = fofs - en->ei.fofs;
735 			prev_en = en;
736 			parts = 1;
737 		}
738 
739 		if (end < org_end && (type != EX_READ ||
740 				org_end - end >= F2FS_MIN_EXTENT_LEN)) {
741 			if (parts) {
742 				__set_extent_info(&ei,
743 					end, org_end - end,
744 					end - dei.fofs + dei.blk, false,
745 					dei.age, dei.last_blocks,
746 					type);
747 				en1 = __insert_extent_tree(sbi, et, &ei,
748 							NULL, NULL, true);
749 				next_en = en1;
750 			} else {
751 				__set_extent_info(&en->ei,
752 					end, en->ei.len - (end - dei.fofs),
753 					en->ei.blk + (end - dei.fofs), true,
754 					dei.age, dei.last_blocks,
755 					type);
756 				next_en = en;
757 			}
758 			parts++;
759 		}
760 
761 		if (!next_en) {
762 			struct rb_node *node = rb_next(&en->rb_node);
763 
764 			next_en = rb_entry_safe(node, struct extent_node,
765 						rb_node);
766 		}
767 
768 		if (parts)
769 			__try_update_largest_extent(et, en);
770 		else
771 			__release_extent_node(sbi, et, en);
772 
773 		/*
774 		 * if original extent is split into zero or two parts, extent
775 		 * tree has been altered by deletion or insertion, therefore
776 		 * invalidate pointers regard to tree.
777 		 */
778 		if (parts != 1) {
779 			insert_p = NULL;
780 			insert_parent = NULL;
781 		}
782 		en = next_en;
783 	}
784 
785 	if (type == EX_BLOCK_AGE)
786 		goto update_age_extent_cache;
787 
788 	/* 3. update extent in read extent cache */
789 	BUG_ON(type != EX_READ);
790 
791 	if (tei->blk) {
792 		__set_extent_info(&ei, fofs, len, tei->blk, false,
793 				  0, 0, EX_READ);
794 		if (!__try_merge_extent_node(sbi, et, &ei, prev_en, next_en))
795 			__insert_extent_tree(sbi, et, &ei,
796 					insert_p, insert_parent, leftmost);
797 
798 		/* give up extent_cache, if split and small updates happen */
799 		if (dei.len >= 1 &&
800 				prev.len < F2FS_MIN_EXTENT_LEN &&
801 				et->largest.len < F2FS_MIN_EXTENT_LEN) {
802 			et->largest.len = 0;
803 			et->largest_updated = true;
804 			set_inode_flag(inode, FI_NO_EXTENT);
805 		}
806 	}
807 
808 	if (is_inode_flag_set(inode, FI_NO_EXTENT))
809 		__free_extent_tree(sbi, et);
810 
811 	if (et->largest_updated) {
812 		et->largest_updated = false;
813 		updated = true;
814 	}
815 	goto out_read_extent_cache;
816 update_age_extent_cache:
817 	if (!tei->last_blocks)
818 		goto out_read_extent_cache;
819 
820 	__set_extent_info(&ei, fofs, len, 0, false,
821 			tei->age, tei->last_blocks, EX_BLOCK_AGE);
822 	if (!__try_merge_extent_node(sbi, et, &ei, prev_en, next_en))
823 		__insert_extent_tree(sbi, et, &ei,
824 					insert_p, insert_parent, leftmost);
825 out_read_extent_cache:
826 	write_unlock(&et->lock);
827 
828 	if (updated)
829 		f2fs_mark_inode_dirty_sync(inode, true);
830 }
831 
832 #ifdef CONFIG_F2FS_FS_COMPRESSION
833 void f2fs_update_read_extent_tree_range_compressed(struct inode *inode,
834 				pgoff_t fofs, block_t blkaddr, unsigned int llen,
835 				unsigned int c_len)
836 {
837 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
838 	struct extent_tree *et = F2FS_I(inode)->extent_tree[EX_READ];
839 	struct extent_node *en = NULL;
840 	struct extent_node *prev_en = NULL, *next_en = NULL;
841 	struct extent_info ei;
842 	struct rb_node **insert_p = NULL, *insert_parent = NULL;
843 	bool leftmost = false;
844 
845 	trace_f2fs_update_read_extent_tree_range(inode, fofs, llen,
846 						blkaddr, c_len);
847 
848 	/* it is safe here to check FI_NO_EXTENT w/o et->lock in ro image */
849 	if (is_inode_flag_set(inode, FI_NO_EXTENT))
850 		return;
851 
852 	write_lock(&et->lock);
853 
854 	en = (struct extent_node *)f2fs_lookup_rb_tree_ret(&et->root,
855 				(struct rb_entry *)et->cached_en, fofs,
856 				(struct rb_entry **)&prev_en,
857 				(struct rb_entry **)&next_en,
858 				&insert_p, &insert_parent, false,
859 				&leftmost);
860 	if (en)
861 		goto unlock_out;
862 
863 	__set_extent_info(&ei, fofs, llen, blkaddr, true, 0, 0, EX_READ);
864 	ei.c_len = c_len;
865 
866 	if (!__try_merge_extent_node(sbi, et, &ei, prev_en, next_en))
867 		__insert_extent_tree(sbi, et, &ei,
868 				insert_p, insert_parent, leftmost);
869 unlock_out:
870 	write_unlock(&et->lock);
871 }
872 #endif
873 
874 static unsigned long long __calculate_block_age(unsigned long long new,
875 						unsigned long long old)
876 {
877 	unsigned long long diff;
878 
879 	diff = (new >= old) ? new - (new - old) : new + (old - new);
880 
881 	return div_u64(diff * LAST_AGE_WEIGHT, 100);
882 }
883 
884 /* This returns a new age and allocated blocks in ei */
885 static int __get_new_block_age(struct inode *inode, struct extent_info *ei,
886 						block_t blkaddr)
887 {
888 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
889 	loff_t f_size = i_size_read(inode);
890 	unsigned long long cur_blocks =
891 				atomic64_read(&sbi->allocated_data_blocks);
892 	struct extent_info tei = *ei;	/* only fofs and len are valid */
893 
894 	/*
895 	 * When I/O is not aligned to a PAGE_SIZE, update will happen to the last
896 	 * file block even in seq write. So don't record age for newly last file
897 	 * block here.
898 	 */
899 	if ((f_size >> PAGE_SHIFT) == ei->fofs && f_size & (PAGE_SIZE - 1) &&
900 			blkaddr == NEW_ADDR)
901 		return -EINVAL;
902 
903 	if (__lookup_extent_tree(inode, ei->fofs, &tei, EX_BLOCK_AGE)) {
904 		unsigned long long cur_age;
905 
906 		if (cur_blocks >= tei.last_blocks)
907 			cur_age = cur_blocks - tei.last_blocks;
908 		else
909 			/* allocated_data_blocks overflow */
910 			cur_age = ULLONG_MAX - tei.last_blocks + cur_blocks;
911 
912 		if (tei.age)
913 			ei->age = __calculate_block_age(cur_age, tei.age);
914 		else
915 			ei->age = cur_age;
916 		ei->last_blocks = cur_blocks;
917 		WARN_ON(ei->age > cur_blocks);
918 		return 0;
919 	}
920 
921 	f2fs_bug_on(sbi, blkaddr == NULL_ADDR);
922 
923 	/* the data block was allocated for the first time */
924 	if (blkaddr == NEW_ADDR)
925 		goto out;
926 
927 	if (__is_valid_data_blkaddr(blkaddr) &&
928 	    !f2fs_is_valid_blkaddr(sbi, blkaddr, DATA_GENERIC_ENHANCE)) {
929 		f2fs_bug_on(sbi, 1);
930 		return -EINVAL;
931 	}
932 out:
933 	/*
934 	 * init block age with zero, this can happen when the block age extent
935 	 * was reclaimed due to memory constraint or system reboot
936 	 */
937 	ei->age = 0;
938 	ei->last_blocks = cur_blocks;
939 	return 0;
940 }
941 
942 static void __update_extent_cache(struct dnode_of_data *dn, enum extent_type type)
943 {
944 	struct extent_info ei = {};
945 
946 	if (!__may_extent_tree(dn->inode, type))
947 		return;
948 
949 	ei.fofs = f2fs_start_bidx_of_node(ofs_of_node(dn->node_page), dn->inode) +
950 								dn->ofs_in_node;
951 	ei.len = 1;
952 
953 	if (type == EX_READ) {
954 		if (dn->data_blkaddr == NEW_ADDR)
955 			ei.blk = NULL_ADDR;
956 		else
957 			ei.blk = dn->data_blkaddr;
958 	} else if (type == EX_BLOCK_AGE) {
959 		if (__get_new_block_age(dn->inode, &ei, dn->data_blkaddr))
960 			return;
961 	}
962 	__update_extent_tree_range(dn->inode, &ei, type);
963 }
964 
965 static unsigned int __shrink_extent_tree(struct f2fs_sb_info *sbi, int nr_shrink,
966 					enum extent_type type)
967 {
968 	struct extent_tree_info *eti = &sbi->extent_tree[type];
969 	struct extent_tree *et, *next;
970 	struct extent_node *en;
971 	unsigned int node_cnt = 0, tree_cnt = 0;
972 	int remained;
973 
974 	if (!atomic_read(&eti->total_zombie_tree))
975 		goto free_node;
976 
977 	if (!mutex_trylock(&eti->extent_tree_lock))
978 		goto out;
979 
980 	/* 1. remove unreferenced extent tree */
981 	list_for_each_entry_safe(et, next, &eti->zombie_list, list) {
982 		if (atomic_read(&et->node_cnt)) {
983 			write_lock(&et->lock);
984 			node_cnt += __free_extent_tree(sbi, et);
985 			write_unlock(&et->lock);
986 		}
987 		f2fs_bug_on(sbi, atomic_read(&et->node_cnt));
988 		list_del_init(&et->list);
989 		radix_tree_delete(&eti->extent_tree_root, et->ino);
990 		kmem_cache_free(extent_tree_slab, et);
991 		atomic_dec(&eti->total_ext_tree);
992 		atomic_dec(&eti->total_zombie_tree);
993 		tree_cnt++;
994 
995 		if (node_cnt + tree_cnt >= nr_shrink)
996 			goto unlock_out;
997 		cond_resched();
998 	}
999 	mutex_unlock(&eti->extent_tree_lock);
1000 
1001 free_node:
1002 	/* 2. remove LRU extent entries */
1003 	if (!mutex_trylock(&eti->extent_tree_lock))
1004 		goto out;
1005 
1006 	remained = nr_shrink - (node_cnt + tree_cnt);
1007 
1008 	spin_lock(&eti->extent_lock);
1009 	for (; remained > 0; remained--) {
1010 		if (list_empty(&eti->extent_list))
1011 			break;
1012 		en = list_first_entry(&eti->extent_list,
1013 					struct extent_node, list);
1014 		et = en->et;
1015 		if (!write_trylock(&et->lock)) {
1016 			/* refresh this extent node's position in extent list */
1017 			list_move_tail(&en->list, &eti->extent_list);
1018 			continue;
1019 		}
1020 
1021 		list_del_init(&en->list);
1022 		spin_unlock(&eti->extent_lock);
1023 
1024 		__detach_extent_node(sbi, et, en);
1025 
1026 		write_unlock(&et->lock);
1027 		node_cnt++;
1028 		spin_lock(&eti->extent_lock);
1029 	}
1030 	spin_unlock(&eti->extent_lock);
1031 
1032 unlock_out:
1033 	mutex_unlock(&eti->extent_tree_lock);
1034 out:
1035 	trace_f2fs_shrink_extent_tree(sbi, node_cnt, tree_cnt, type);
1036 
1037 	return node_cnt + tree_cnt;
1038 }
1039 
1040 /* read extent cache operations */
1041 bool f2fs_lookup_read_extent_cache(struct inode *inode, pgoff_t pgofs,
1042 				struct extent_info *ei)
1043 {
1044 	if (!__may_extent_tree(inode, EX_READ))
1045 		return false;
1046 
1047 	return __lookup_extent_tree(inode, pgofs, ei, EX_READ);
1048 }
1049 
1050 void f2fs_update_read_extent_cache(struct dnode_of_data *dn)
1051 {
1052 	return __update_extent_cache(dn, EX_READ);
1053 }
1054 
1055 void f2fs_update_read_extent_cache_range(struct dnode_of_data *dn,
1056 				pgoff_t fofs, block_t blkaddr, unsigned int len)
1057 {
1058 	struct extent_info ei = {
1059 		.fofs = fofs,
1060 		.len = len,
1061 		.blk = blkaddr,
1062 	};
1063 
1064 	if (!__may_extent_tree(dn->inode, EX_READ))
1065 		return;
1066 
1067 	__update_extent_tree_range(dn->inode, &ei, EX_READ);
1068 }
1069 
1070 unsigned int f2fs_shrink_read_extent_tree(struct f2fs_sb_info *sbi, int nr_shrink)
1071 {
1072 	if (!test_opt(sbi, READ_EXTENT_CACHE))
1073 		return 0;
1074 
1075 	return __shrink_extent_tree(sbi, nr_shrink, EX_READ);
1076 }
1077 
1078 /* block age extent cache operations */
1079 bool f2fs_lookup_age_extent_cache(struct inode *inode, pgoff_t pgofs,
1080 				struct extent_info *ei)
1081 {
1082 	if (!__may_extent_tree(inode, EX_BLOCK_AGE))
1083 		return false;
1084 
1085 	return __lookup_extent_tree(inode, pgofs, ei, EX_BLOCK_AGE);
1086 }
1087 
1088 void f2fs_update_age_extent_cache(struct dnode_of_data *dn)
1089 {
1090 	return __update_extent_cache(dn, EX_BLOCK_AGE);
1091 }
1092 
1093 void f2fs_update_age_extent_cache_range(struct dnode_of_data *dn,
1094 				pgoff_t fofs, unsigned int len)
1095 {
1096 	struct extent_info ei = {
1097 		.fofs = fofs,
1098 		.len = len,
1099 	};
1100 
1101 	if (!__may_extent_tree(dn->inode, EX_BLOCK_AGE))
1102 		return;
1103 
1104 	__update_extent_tree_range(dn->inode, &ei, EX_BLOCK_AGE);
1105 }
1106 
1107 unsigned int f2fs_shrink_age_extent_tree(struct f2fs_sb_info *sbi, int nr_shrink)
1108 {
1109 	if (!test_opt(sbi, AGE_EXTENT_CACHE))
1110 		return 0;
1111 
1112 	return __shrink_extent_tree(sbi, nr_shrink, EX_BLOCK_AGE);
1113 }
1114 
1115 static unsigned int __destroy_extent_node(struct inode *inode,
1116 					enum extent_type type)
1117 {
1118 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
1119 	struct extent_tree *et = F2FS_I(inode)->extent_tree[type];
1120 	unsigned int node_cnt = 0;
1121 
1122 	if (!et || !atomic_read(&et->node_cnt))
1123 		return 0;
1124 
1125 	write_lock(&et->lock);
1126 	node_cnt = __free_extent_tree(sbi, et);
1127 	write_unlock(&et->lock);
1128 
1129 	return node_cnt;
1130 }
1131 
1132 void f2fs_destroy_extent_node(struct inode *inode)
1133 {
1134 	__destroy_extent_node(inode, EX_READ);
1135 	__destroy_extent_node(inode, EX_BLOCK_AGE);
1136 }
1137 
1138 static void __drop_extent_tree(struct inode *inode, enum extent_type type)
1139 {
1140 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
1141 	struct extent_tree *et = F2FS_I(inode)->extent_tree[type];
1142 	bool updated = false;
1143 
1144 	if (!__may_extent_tree(inode, type))
1145 		return;
1146 
1147 	write_lock(&et->lock);
1148 	__free_extent_tree(sbi, et);
1149 	if (type == EX_READ) {
1150 		set_inode_flag(inode, FI_NO_EXTENT);
1151 		if (et->largest.len) {
1152 			et->largest.len = 0;
1153 			updated = true;
1154 		}
1155 	}
1156 	write_unlock(&et->lock);
1157 	if (updated)
1158 		f2fs_mark_inode_dirty_sync(inode, true);
1159 }
1160 
1161 void f2fs_drop_extent_tree(struct inode *inode)
1162 {
1163 	__drop_extent_tree(inode, EX_READ);
1164 	__drop_extent_tree(inode, EX_BLOCK_AGE);
1165 }
1166 
1167 static void __destroy_extent_tree(struct inode *inode, enum extent_type type)
1168 {
1169 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
1170 	struct extent_tree_info *eti = &sbi->extent_tree[type];
1171 	struct extent_tree *et = F2FS_I(inode)->extent_tree[type];
1172 	unsigned int node_cnt = 0;
1173 
1174 	if (!et)
1175 		return;
1176 
1177 	if (inode->i_nlink && !is_bad_inode(inode) &&
1178 					atomic_read(&et->node_cnt)) {
1179 		mutex_lock(&eti->extent_tree_lock);
1180 		list_add_tail(&et->list, &eti->zombie_list);
1181 		atomic_inc(&eti->total_zombie_tree);
1182 		mutex_unlock(&eti->extent_tree_lock);
1183 		return;
1184 	}
1185 
1186 	/* free all extent info belong to this extent tree */
1187 	node_cnt = __destroy_extent_node(inode, type);
1188 
1189 	/* delete extent tree entry in radix tree */
1190 	mutex_lock(&eti->extent_tree_lock);
1191 	f2fs_bug_on(sbi, atomic_read(&et->node_cnt));
1192 	radix_tree_delete(&eti->extent_tree_root, inode->i_ino);
1193 	kmem_cache_free(extent_tree_slab, et);
1194 	atomic_dec(&eti->total_ext_tree);
1195 	mutex_unlock(&eti->extent_tree_lock);
1196 
1197 	F2FS_I(inode)->extent_tree[type] = NULL;
1198 
1199 	trace_f2fs_destroy_extent_tree(inode, node_cnt, type);
1200 }
1201 
1202 void f2fs_destroy_extent_tree(struct inode *inode)
1203 {
1204 	__destroy_extent_tree(inode, EX_READ);
1205 	__destroy_extent_tree(inode, EX_BLOCK_AGE);
1206 }
1207 
1208 static void __init_extent_tree_info(struct extent_tree_info *eti)
1209 {
1210 	INIT_RADIX_TREE(&eti->extent_tree_root, GFP_NOIO);
1211 	mutex_init(&eti->extent_tree_lock);
1212 	INIT_LIST_HEAD(&eti->extent_list);
1213 	spin_lock_init(&eti->extent_lock);
1214 	atomic_set(&eti->total_ext_tree, 0);
1215 	INIT_LIST_HEAD(&eti->zombie_list);
1216 	atomic_set(&eti->total_zombie_tree, 0);
1217 	atomic_set(&eti->total_ext_node, 0);
1218 }
1219 
1220 void f2fs_init_extent_cache_info(struct f2fs_sb_info *sbi)
1221 {
1222 	__init_extent_tree_info(&sbi->extent_tree[EX_READ]);
1223 	__init_extent_tree_info(&sbi->extent_tree[EX_BLOCK_AGE]);
1224 
1225 	/* initialize for block age extents */
1226 	atomic64_set(&sbi->allocated_data_blocks, 0);
1227 	sbi->hot_data_age_threshold = DEF_HOT_DATA_AGE_THRESHOLD;
1228 	sbi->warm_data_age_threshold = DEF_WARM_DATA_AGE_THRESHOLD;
1229 }
1230 
1231 int __init f2fs_create_extent_cache(void)
1232 {
1233 	extent_tree_slab = f2fs_kmem_cache_create("f2fs_extent_tree",
1234 			sizeof(struct extent_tree));
1235 	if (!extent_tree_slab)
1236 		return -ENOMEM;
1237 	extent_node_slab = f2fs_kmem_cache_create("f2fs_extent_node",
1238 			sizeof(struct extent_node));
1239 	if (!extent_node_slab) {
1240 		kmem_cache_destroy(extent_tree_slab);
1241 		return -ENOMEM;
1242 	}
1243 	return 0;
1244 }
1245 
1246 void f2fs_destroy_extent_cache(void)
1247 {
1248 	kmem_cache_destroy(extent_node_slab);
1249 	kmem_cache_destroy(extent_tree_slab);
1250 }
1251