1 // SPDX-License-Identifier: GPL-2.0
2 /*
3 * Copyright (C) 2015 Facebook. All rights reserved.
4 */
5
6 #include <linux/kernel.h>
7 #include <linux/sched/mm.h>
8 #include "messages.h"
9 #include "ctree.h"
10 #include "disk-io.h"
11 #include "locking.h"
12 #include "free-space-tree.h"
13 #include "transaction.h"
14 #include "block-group.h"
15 #include "fs.h"
16 #include "accessors.h"
17 #include "extent-tree.h"
18 #include "root-tree.h"
19
20 static int __add_block_group_free_space(struct btrfs_trans_handle *trans,
21 struct btrfs_block_group *block_group,
22 struct btrfs_path *path);
23
btrfs_free_space_root(struct btrfs_block_group * block_group)24 struct btrfs_root *btrfs_free_space_root(struct btrfs_block_group *block_group)
25 {
26 struct btrfs_key key = {
27 .objectid = BTRFS_FREE_SPACE_TREE_OBJECTID,
28 .type = BTRFS_ROOT_ITEM_KEY,
29 .offset = 0,
30 };
31
32 if (btrfs_fs_incompat(block_group->fs_info, EXTENT_TREE_V2))
33 key.offset = block_group->global_root_id;
34 return btrfs_global_root(block_group->fs_info, &key);
35 }
36
btrfs_set_free_space_tree_thresholds(struct btrfs_block_group * cache)37 void btrfs_set_free_space_tree_thresholds(struct btrfs_block_group *cache)
38 {
39 u32 bitmap_range;
40 size_t bitmap_size;
41 u64 num_bitmaps, total_bitmap_size;
42
43 if (WARN_ON(cache->length == 0))
44 btrfs_warn(cache->fs_info, "block group %llu length is zero",
45 cache->start);
46
47 /*
48 * We convert to bitmaps when the disk space required for using extents
49 * exceeds that required for using bitmaps.
50 */
51 bitmap_range = cache->fs_info->sectorsize * BTRFS_FREE_SPACE_BITMAP_BITS;
52 num_bitmaps = div_u64(cache->length + bitmap_range - 1, bitmap_range);
53 bitmap_size = sizeof(struct btrfs_item) + BTRFS_FREE_SPACE_BITMAP_SIZE;
54 total_bitmap_size = num_bitmaps * bitmap_size;
55 cache->bitmap_high_thresh = div_u64(total_bitmap_size,
56 sizeof(struct btrfs_item));
57
58 /*
59 * We allow for a small buffer between the high threshold and low
60 * threshold to avoid thrashing back and forth between the two formats.
61 */
62 if (cache->bitmap_high_thresh > 100)
63 cache->bitmap_low_thresh = cache->bitmap_high_thresh - 100;
64 else
65 cache->bitmap_low_thresh = 0;
66 }
67
add_new_free_space_info(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path)68 static int add_new_free_space_info(struct btrfs_trans_handle *trans,
69 struct btrfs_block_group *block_group,
70 struct btrfs_path *path)
71 {
72 struct btrfs_root *root = btrfs_free_space_root(block_group);
73 struct btrfs_free_space_info *info;
74 struct btrfs_key key;
75 struct extent_buffer *leaf;
76 int ret;
77
78 key.objectid = block_group->start;
79 key.type = BTRFS_FREE_SPACE_INFO_KEY;
80 key.offset = block_group->length;
81
82 ret = btrfs_insert_empty_item(trans, root, path, &key, sizeof(*info));
83 if (ret)
84 return ret;
85
86 leaf = path->nodes[0];
87 info = btrfs_item_ptr(leaf, path->slots[0],
88 struct btrfs_free_space_info);
89 btrfs_set_free_space_extent_count(leaf, info, 0);
90 btrfs_set_free_space_flags(leaf, info, 0);
91 btrfs_release_path(path);
92 return 0;
93 }
94
btrfs_search_free_space_info(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,int cow)95 struct btrfs_free_space_info *btrfs_search_free_space_info(
96 struct btrfs_trans_handle *trans,
97 struct btrfs_block_group *block_group,
98 struct btrfs_path *path, int cow)
99 {
100 struct btrfs_fs_info *fs_info = block_group->fs_info;
101 struct btrfs_root *root = btrfs_free_space_root(block_group);
102 struct btrfs_key key;
103 int ret;
104
105 key.objectid = block_group->start;
106 key.type = BTRFS_FREE_SPACE_INFO_KEY;
107 key.offset = block_group->length;
108
109 ret = btrfs_search_slot(trans, root, &key, path, 0, cow);
110 if (ret < 0)
111 return ERR_PTR(ret);
112 if (unlikely(ret != 0)) {
113 btrfs_warn(fs_info, "missing free space info for %llu",
114 block_group->start);
115 DEBUG_WARN();
116 return ERR_PTR(-ENOENT);
117 }
118
119 return btrfs_item_ptr(path->nodes[0], path->slots[0],
120 struct btrfs_free_space_info);
121 }
122
123 /*
124 * btrfs_search_slot() but we're looking for the greatest key less than the
125 * passed key.
126 */
btrfs_search_prev_slot(struct btrfs_trans_handle * trans,struct btrfs_root * root,struct btrfs_key * key,struct btrfs_path * p,int ins_len,int cow)127 static int btrfs_search_prev_slot(struct btrfs_trans_handle *trans,
128 struct btrfs_root *root,
129 struct btrfs_key *key, struct btrfs_path *p,
130 int ins_len, int cow)
131 {
132 int ret;
133
134 ret = btrfs_search_slot(trans, root, key, p, ins_len, cow);
135 if (ret < 0)
136 return ret;
137
138 if (unlikely(ret == 0)) {
139 DEBUG_WARN();
140 return -EIO;
141 }
142
143 if (unlikely(p->slots[0] == 0)) {
144 DEBUG_WARN("no previous slot found");
145 return -EIO;
146 }
147 p->slots[0]--;
148
149 return 0;
150 }
151
free_space_bitmap_size(const struct btrfs_fs_info * fs_info,u64 size)152 static inline u32 free_space_bitmap_size(const struct btrfs_fs_info *fs_info,
153 u64 size)
154 {
155 return DIV_ROUND_UP(size >> fs_info->sectorsize_bits, BITS_PER_BYTE);
156 }
157
alloc_bitmap(u32 bitmap_size)158 static unsigned long *alloc_bitmap(u32 bitmap_size)
159 {
160 unsigned long *ret;
161 unsigned int nofs_flag;
162 u32 bitmap_rounded_size = round_up(bitmap_size, sizeof(unsigned long));
163
164 /*
165 * GFP_NOFS doesn't work with kvmalloc(), but we really can't recurse
166 * into the filesystem here. All callers hold a transaction handle
167 * open, so if a GFP_KERNEL allocation recurses into the filesystem
168 * and triggers a transaction commit, we would deadlock.
169 */
170 nofs_flag = memalloc_nofs_save();
171 ret = kvzalloc(bitmap_rounded_size, GFP_KERNEL);
172 memalloc_nofs_restore(nofs_flag);
173 return ret;
174 }
175
le_bitmap_set(unsigned long * map,unsigned int start,int len)176 static void le_bitmap_set(unsigned long *map, unsigned int start, int len)
177 {
178 u8 *p = ((u8 *)map) + BIT_BYTE(start);
179 const unsigned int size = start + len;
180 int bits_to_set = BITS_PER_BYTE - (start % BITS_PER_BYTE);
181 u8 mask_to_set = BITMAP_FIRST_BYTE_MASK(start);
182
183 while (len - bits_to_set >= 0) {
184 *p |= mask_to_set;
185 len -= bits_to_set;
186 bits_to_set = BITS_PER_BYTE;
187 mask_to_set = ~0;
188 p++;
189 }
190 if (len) {
191 mask_to_set &= BITMAP_LAST_BYTE_MASK(size);
192 *p |= mask_to_set;
193 }
194 }
195
196 EXPORT_FOR_TESTS
btrfs_convert_free_space_to_bitmaps(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path)197 int btrfs_convert_free_space_to_bitmaps(struct btrfs_trans_handle *trans,
198 struct btrfs_block_group *block_group,
199 struct btrfs_path *path)
200 {
201 struct btrfs_fs_info *fs_info = trans->fs_info;
202 struct btrfs_root *root = btrfs_free_space_root(block_group);
203 struct btrfs_free_space_info *info;
204 struct btrfs_key key, found_key;
205 struct extent_buffer *leaf;
206 unsigned long *bitmap;
207 char *bitmap_cursor;
208 u64 start, end;
209 u64 bitmap_range, i;
210 u32 bitmap_size, flags, expected_extent_count;
211 u32 extent_count = 0;
212 bool done = false;
213 int nr;
214 int ret;
215
216 bitmap_size = free_space_bitmap_size(fs_info, block_group->length);
217 bitmap = alloc_bitmap(bitmap_size);
218 if (unlikely(!bitmap))
219 return 0;
220
221 start = block_group->start;
222 end = btrfs_block_group_end(block_group);
223
224 key.objectid = end - 1;
225 key.type = (u8)-1;
226 key.offset = (u64)-1;
227
228 while (!done) {
229 ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1);
230 if (unlikely(ret)) {
231 btrfs_abort_transaction(trans, ret);
232 goto out;
233 }
234
235 leaf = path->nodes[0];
236 nr = 0;
237 path->slots[0]++;
238 while (path->slots[0] > 0) {
239 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0] - 1);
240
241 if (found_key.type == BTRFS_FREE_SPACE_INFO_KEY) {
242 ASSERT(found_key.objectid == block_group->start);
243 ASSERT(found_key.offset == block_group->length);
244 done = true;
245 break;
246 } else if (found_key.type == BTRFS_FREE_SPACE_EXTENT_KEY) {
247 u64 first, last;
248
249 ASSERT(found_key.objectid >= start);
250 ASSERT(found_key.objectid < end);
251 ASSERT(found_key.objectid + found_key.offset <= end);
252
253 first = div_u64(found_key.objectid - start,
254 fs_info->sectorsize);
255 last = div_u64(found_key.objectid + found_key.offset - start,
256 fs_info->sectorsize);
257 le_bitmap_set(bitmap, first, last - first);
258
259 extent_count++;
260 nr++;
261 path->slots[0]--;
262 } else {
263 btrfs_err(fs_info, "unexpected free space tree key type %u",
264 found_key.type);
265 ret = -EUCLEAN;
266 btrfs_abort_transaction(trans, ret);
267 goto out;
268 }
269 }
270
271 ret = btrfs_del_items(trans, root, path, path->slots[0], nr);
272 if (unlikely(ret)) {
273 btrfs_abort_transaction(trans, ret);
274 goto out;
275 }
276 btrfs_release_path(path);
277 }
278
279 info = btrfs_search_free_space_info(trans, block_group, path, 1);
280 if (IS_ERR(info)) {
281 ret = PTR_ERR(info);
282 btrfs_abort_transaction(trans, ret);
283 goto out;
284 }
285 leaf = path->nodes[0];
286 flags = btrfs_free_space_flags(leaf, info);
287 flags |= BTRFS_FREE_SPACE_USING_BITMAPS;
288 block_group->using_free_space_bitmaps = true;
289 block_group->using_free_space_bitmaps_cached = true;
290 btrfs_set_free_space_flags(leaf, info, flags);
291 expected_extent_count = btrfs_free_space_extent_count(leaf, info);
292 btrfs_release_path(path);
293
294 if (unlikely(extent_count != expected_extent_count)) {
295 btrfs_err(fs_info,
296 "incorrect extent count for %llu; counted %u, expected %u",
297 block_group->start, extent_count,
298 expected_extent_count);
299 ret = -EIO;
300 btrfs_abort_transaction(trans, ret);
301 goto out;
302 }
303
304 bitmap_cursor = (char *)bitmap;
305 bitmap_range = fs_info->sectorsize * BTRFS_FREE_SPACE_BITMAP_BITS;
306 i = start;
307 while (i < end) {
308 unsigned long ptr;
309 u64 extent_size;
310 u32 data_size;
311
312 extent_size = min(end - i, bitmap_range);
313 data_size = free_space_bitmap_size(fs_info, extent_size);
314
315 key.objectid = i;
316 key.type = BTRFS_FREE_SPACE_BITMAP_KEY;
317 key.offset = extent_size;
318
319 ret = btrfs_insert_empty_item(trans, root, path, &key,
320 data_size);
321 if (unlikely(ret)) {
322 btrfs_abort_transaction(trans, ret);
323 goto out;
324 }
325
326 leaf = path->nodes[0];
327 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
328 write_extent_buffer(leaf, bitmap_cursor, ptr,
329 data_size);
330 btrfs_release_path(path);
331
332 i += extent_size;
333 bitmap_cursor += data_size;
334 }
335
336 ret = 0;
337 out:
338 kvfree(bitmap);
339 return ret;
340 }
341
342 EXPORT_FOR_TESTS
btrfs_convert_free_space_to_extents(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path)343 int btrfs_convert_free_space_to_extents(struct btrfs_trans_handle *trans,
344 struct btrfs_block_group *block_group,
345 struct btrfs_path *path)
346 {
347 struct btrfs_fs_info *fs_info = trans->fs_info;
348 struct btrfs_root *root = btrfs_free_space_root(block_group);
349 struct btrfs_free_space_info *info;
350 struct btrfs_key key, found_key;
351 struct extent_buffer *leaf;
352 unsigned long *bitmap;
353 u64 start, end;
354 u32 bitmap_size, flags, expected_extent_count;
355 unsigned long nrbits, start_bit, end_bit;
356 u32 extent_count = 0;
357 bool done = false;
358 int nr;
359 int ret;
360
361 bitmap_size = free_space_bitmap_size(fs_info, block_group->length);
362 bitmap = alloc_bitmap(bitmap_size);
363 if (unlikely(!bitmap))
364 return 0;
365
366 start = block_group->start;
367 end = btrfs_block_group_end(block_group);
368
369 key.objectid = end - 1;
370 key.type = (u8)-1;
371 key.offset = (u64)-1;
372
373 while (!done) {
374 ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1);
375 if (unlikely(ret)) {
376 btrfs_abort_transaction(trans, ret);
377 goto out;
378 }
379
380 leaf = path->nodes[0];
381 nr = 0;
382 path->slots[0]++;
383 while (path->slots[0] > 0) {
384 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0] - 1);
385
386 if (found_key.type == BTRFS_FREE_SPACE_INFO_KEY) {
387 ASSERT(found_key.objectid == block_group->start);
388 ASSERT(found_key.offset == block_group->length);
389 done = true;
390 break;
391 } else if (found_key.type == BTRFS_FREE_SPACE_BITMAP_KEY) {
392 unsigned long ptr;
393 char *bitmap_cursor;
394 u32 bitmap_pos, data_size;
395
396 ASSERT(found_key.objectid >= start);
397 ASSERT(found_key.objectid < end);
398 ASSERT(found_key.objectid + found_key.offset <= end);
399
400 bitmap_pos = div_u64(found_key.objectid - start,
401 fs_info->sectorsize *
402 BITS_PER_BYTE);
403 bitmap_cursor = ((char *)bitmap) + bitmap_pos;
404 data_size = free_space_bitmap_size(fs_info,
405 found_key.offset);
406
407 path->slots[0]--;
408 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
409 read_extent_buffer(leaf, bitmap_cursor, ptr,
410 data_size);
411
412 nr++;
413 } else {
414 btrfs_err(fs_info, "unexpected free space tree key type %u",
415 found_key.type);
416 ret = -EUCLEAN;
417 btrfs_abort_transaction(trans, ret);
418 goto out;
419 }
420 }
421
422 ret = btrfs_del_items(trans, root, path, path->slots[0], nr);
423 if (unlikely(ret)) {
424 btrfs_abort_transaction(trans, ret);
425 goto out;
426 }
427 btrfs_release_path(path);
428 }
429
430 info = btrfs_search_free_space_info(trans, block_group, path, 1);
431 if (IS_ERR(info)) {
432 ret = PTR_ERR(info);
433 btrfs_abort_transaction(trans, ret);
434 goto out;
435 }
436 leaf = path->nodes[0];
437 flags = btrfs_free_space_flags(leaf, info);
438 flags &= ~BTRFS_FREE_SPACE_USING_BITMAPS;
439 block_group->using_free_space_bitmaps = false;
440 block_group->using_free_space_bitmaps_cached = true;
441 btrfs_set_free_space_flags(leaf, info, flags);
442 expected_extent_count = btrfs_free_space_extent_count(leaf, info);
443 btrfs_release_path(path);
444
445 nrbits = block_group->length >> fs_info->sectorsize_bits;
446 start_bit = find_next_bit_le(bitmap, nrbits, 0);
447
448 while (start_bit < nrbits) {
449 end_bit = find_next_zero_bit_le(bitmap, nrbits, start_bit);
450 ASSERT(start_bit < end_bit);
451
452 key.objectid = start + start_bit * fs_info->sectorsize;
453 key.type = BTRFS_FREE_SPACE_EXTENT_KEY;
454 key.offset = (end_bit - start_bit) * fs_info->sectorsize;
455
456 ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
457 if (unlikely(ret)) {
458 btrfs_abort_transaction(trans, ret);
459 goto out;
460 }
461 btrfs_release_path(path);
462
463 extent_count++;
464
465 start_bit = find_next_bit_le(bitmap, nrbits, end_bit);
466 }
467
468 if (unlikely(extent_count != expected_extent_count)) {
469 btrfs_err(fs_info,
470 "incorrect extent count for %llu; counted %u, expected %u",
471 block_group->start, extent_count,
472 expected_extent_count);
473 ret = -EIO;
474 btrfs_abort_transaction(trans, ret);
475 goto out;
476 }
477
478 ret = 0;
479 out:
480 kvfree(bitmap);
481 return ret;
482 }
483
update_free_space_extent_count(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,int new_extents)484 static int update_free_space_extent_count(struct btrfs_trans_handle *trans,
485 struct btrfs_block_group *block_group,
486 struct btrfs_path *path,
487 int new_extents)
488 {
489 struct btrfs_free_space_info *info;
490 u32 flags;
491 u32 extent_count;
492 int ret = 0;
493
494 if (new_extents == 0)
495 return 0;
496
497 info = btrfs_search_free_space_info(trans, block_group, path, 1);
498 if (IS_ERR(info))
499 return PTR_ERR(info);
500
501 flags = btrfs_free_space_flags(path->nodes[0], info);
502 extent_count = btrfs_free_space_extent_count(path->nodes[0], info);
503
504 extent_count += new_extents;
505 btrfs_set_free_space_extent_count(path->nodes[0], info, extent_count);
506 btrfs_release_path(path);
507
508 if (!(flags & BTRFS_FREE_SPACE_USING_BITMAPS) &&
509 extent_count > block_group->bitmap_high_thresh) {
510 ret = btrfs_convert_free_space_to_bitmaps(trans, block_group, path);
511 } else if ((flags & BTRFS_FREE_SPACE_USING_BITMAPS) &&
512 extent_count < block_group->bitmap_low_thresh) {
513 ret = btrfs_convert_free_space_to_extents(trans, block_group, path);
514 }
515
516 return ret;
517 }
518
519 EXPORT_FOR_TESTS
btrfs_free_space_test_bit(struct btrfs_block_group * block_group,struct btrfs_path * path,u64 offset)520 bool btrfs_free_space_test_bit(struct btrfs_block_group *block_group,
521 struct btrfs_path *path, u64 offset)
522 {
523 struct extent_buffer *leaf;
524 struct btrfs_key key;
525 u64 found_start, found_end;
526 unsigned long ptr, i;
527
528 leaf = path->nodes[0];
529 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
530 ASSERT(key.type == BTRFS_FREE_SPACE_BITMAP_KEY);
531
532 found_start = key.objectid;
533 found_end = key.objectid + key.offset;
534 ASSERT(offset >= found_start && offset < found_end);
535
536 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
537 i = div_u64(offset - found_start,
538 block_group->fs_info->sectorsize);
539 return extent_buffer_test_bit(leaf, ptr, i);
540 }
541
free_space_modify_bits(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,u64 * start,u64 * size,bool set_bits)542 static void free_space_modify_bits(struct btrfs_trans_handle *trans,
543 struct btrfs_block_group *block_group,
544 struct btrfs_path *path, u64 *start, u64 *size,
545 bool set_bits)
546 {
547 struct btrfs_fs_info *fs_info = block_group->fs_info;
548 struct extent_buffer *leaf;
549 struct btrfs_key key;
550 u64 end = *start + *size;
551 u64 found_start, found_end;
552 unsigned long ptr, first, last;
553
554 leaf = path->nodes[0];
555 btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
556 ASSERT(key.type == BTRFS_FREE_SPACE_BITMAP_KEY);
557
558 found_start = key.objectid;
559 found_end = key.objectid + key.offset;
560 ASSERT(*start >= found_start && *start < found_end);
561 ASSERT(end > found_start);
562
563 if (end > found_end)
564 end = found_end;
565
566 ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
567 first = (*start - found_start) >> fs_info->sectorsize_bits;
568 last = (end - found_start) >> fs_info->sectorsize_bits;
569 if (set_bits)
570 extent_buffer_bitmap_set(leaf, ptr, first, last - first);
571 else
572 extent_buffer_bitmap_clear(leaf, ptr, first, last - first);
573 btrfs_mark_buffer_dirty(trans, leaf);
574
575 *size -= end - *start;
576 *start = end;
577 }
578
579 /*
580 * We can't use btrfs_next_item() in modify_free_space_bitmap() because
581 * btrfs_next_leaf() doesn't get the path for writing. We can forgo the fancy
582 * tree walking in btrfs_next_leaf() anyways because we know exactly what we're
583 * looking for.
584 */
free_space_next_bitmap(struct btrfs_trans_handle * trans,struct btrfs_root * root,struct btrfs_path * p)585 static int free_space_next_bitmap(struct btrfs_trans_handle *trans,
586 struct btrfs_root *root, struct btrfs_path *p)
587 {
588 struct btrfs_key key;
589
590 if (p->slots[0] + 1 < btrfs_header_nritems(p->nodes[0])) {
591 p->slots[0]++;
592 return 0;
593 }
594
595 btrfs_item_key_to_cpu(p->nodes[0], &key, p->slots[0]);
596 btrfs_release_path(p);
597
598 key.objectid += key.offset;
599 key.type = (u8)-1;
600 key.offset = (u64)-1;
601
602 return btrfs_search_prev_slot(trans, root, &key, p, 0, 1);
603 }
604
605 /*
606 * If remove is 1, then we are removing free space, thus clearing bits in the
607 * bitmap. If remove is 0, then we are adding free space, thus setting bits in
608 * the bitmap.
609 */
modify_free_space_bitmap(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,u64 start,u64 size,bool remove)610 static int modify_free_space_bitmap(struct btrfs_trans_handle *trans,
611 struct btrfs_block_group *block_group,
612 struct btrfs_path *path,
613 u64 start, u64 size, bool remove)
614 {
615 struct btrfs_root *root = btrfs_free_space_root(block_group);
616 struct btrfs_key key;
617 u64 end = start + size;
618 u64 cur_start, cur_size;
619 bool prev_bit_set = false;
620 bool next_bit_set = false;
621 int new_extents;
622 int ret;
623
624 /*
625 * Read the bit for the block immediately before the extent of space if
626 * that block is within the block group.
627 */
628 if (start > block_group->start) {
629 u64 prev_block = start - block_group->fs_info->sectorsize;
630
631 key.objectid = prev_block;
632 key.type = (u8)-1;
633 key.offset = (u64)-1;
634
635 ret = btrfs_search_prev_slot(trans, root, &key, path, 0, 1);
636 if (ret)
637 return ret;
638
639 prev_bit_set = btrfs_free_space_test_bit(block_group, path, prev_block);
640
641 /* The previous block may have been in the previous bitmap. */
642 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
643 if (start >= key.objectid + key.offset) {
644 ret = free_space_next_bitmap(trans, root, path);
645 if (ret)
646 return ret;
647 }
648 } else {
649 key.objectid = start;
650 key.type = (u8)-1;
651 key.offset = (u64)-1;
652
653 ret = btrfs_search_prev_slot(trans, root, &key, path, 0, 1);
654 if (ret)
655 return ret;
656 }
657
658 /*
659 * Iterate over all of the bitmaps overlapped by the extent of space,
660 * clearing/setting bits as required.
661 */
662 cur_start = start;
663 cur_size = size;
664 while (1) {
665 free_space_modify_bits(trans, block_group, path, &cur_start,
666 &cur_size, !remove);
667 if (cur_size == 0)
668 break;
669 ret = free_space_next_bitmap(trans, root, path);
670 if (ret)
671 return ret;
672 }
673
674 /*
675 * Read the bit for the block immediately after the extent of space if
676 * that block is within the block group.
677 */
678 if (end < btrfs_block_group_end(block_group)) {
679 /* The next block may be in the next bitmap. */
680 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
681 if (end >= key.objectid + key.offset) {
682 ret = free_space_next_bitmap(trans, root, path);
683 if (ret)
684 return ret;
685 }
686
687 next_bit_set = btrfs_free_space_test_bit(block_group, path, end);
688 }
689
690 if (remove) {
691 new_extents = -1;
692 if (prev_bit_set) {
693 /* Leftover on the left. */
694 new_extents++;
695 }
696 if (next_bit_set) {
697 /* Leftover on the right. */
698 new_extents++;
699 }
700 } else {
701 new_extents = 1;
702 if (prev_bit_set) {
703 /* Merging with neighbor on the left. */
704 new_extents--;
705 }
706 if (next_bit_set) {
707 /* Merging with neighbor on the right. */
708 new_extents--;
709 }
710 }
711
712 btrfs_release_path(path);
713 return update_free_space_extent_count(trans, block_group, path, new_extents);
714 }
715
remove_free_space_extent(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,u64 start,u64 size)716 static int remove_free_space_extent(struct btrfs_trans_handle *trans,
717 struct btrfs_block_group *block_group,
718 struct btrfs_path *path,
719 u64 start, u64 size)
720 {
721 struct btrfs_root *root = btrfs_free_space_root(block_group);
722 struct btrfs_key key;
723 u64 found_start, found_end;
724 u64 end = start + size;
725 int new_extents = -1;
726 int ret;
727
728 key.objectid = start;
729 key.type = (u8)-1;
730 key.offset = (u64)-1;
731
732 ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1);
733 if (ret)
734 return ret;
735
736 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
737
738 ASSERT(key.type == BTRFS_FREE_SPACE_EXTENT_KEY);
739
740 found_start = key.objectid;
741 found_end = key.objectid + key.offset;
742 ASSERT(start >= found_start && end <= found_end);
743
744 /*
745 * Okay, now that we've found the free space extent which contains the
746 * free space that we are removing, there are four cases:
747 *
748 * 1. We're using the whole extent: delete the key we found and
749 * decrement the free space extent count.
750 * 2. We are using part of the extent starting at the beginning: delete
751 * the key we found and insert a new key representing the leftover at
752 * the end. There is no net change in the number of extents.
753 * 3. We are using part of the extent ending at the end: delete the key
754 * we found and insert a new key representing the leftover at the
755 * beginning. There is no net change in the number of extents.
756 * 4. We are using part of the extent in the middle: delete the key we
757 * found and insert two new keys representing the leftovers on each
758 * side. Where we used to have one extent, we now have two, so increment
759 * the extent count. We may need to convert the block group to bitmaps
760 * as a result.
761 */
762
763 /* Delete the existing key (cases 1-4). */
764 ret = btrfs_del_item(trans, root, path);
765 if (ret)
766 return ret;
767
768 /* Add a key for leftovers at the beginning (cases 3 and 4). */
769 if (start > found_start) {
770 key.objectid = found_start;
771 key.type = BTRFS_FREE_SPACE_EXTENT_KEY;
772 key.offset = start - found_start;
773
774 btrfs_release_path(path);
775 ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
776 if (ret)
777 return ret;
778 new_extents++;
779 }
780
781 /* Add a key for leftovers at the end (cases 2 and 4). */
782 if (end < found_end) {
783 key.objectid = end;
784 key.type = BTRFS_FREE_SPACE_EXTENT_KEY;
785 key.offset = found_end - end;
786
787 btrfs_release_path(path);
788 ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
789 if (ret)
790 return ret;
791 new_extents++;
792 }
793
794 btrfs_release_path(path);
795 return update_free_space_extent_count(trans, block_group, path, new_extents);
796 }
797
using_bitmaps(struct btrfs_block_group * bg,struct btrfs_path * path)798 static int using_bitmaps(struct btrfs_block_group *bg, struct btrfs_path *path)
799 {
800 struct btrfs_free_space_info *info;
801 u32 flags;
802
803 if (bg->using_free_space_bitmaps_cached)
804 return bg->using_free_space_bitmaps;
805
806 info = btrfs_search_free_space_info(NULL, bg, path, 0);
807 if (IS_ERR(info))
808 return PTR_ERR(info);
809 flags = btrfs_free_space_flags(path->nodes[0], info);
810 btrfs_release_path(path);
811
812 bg->using_free_space_bitmaps = (flags & BTRFS_FREE_SPACE_USING_BITMAPS);
813 bg->using_free_space_bitmaps_cached = true;
814
815 return bg->using_free_space_bitmaps;
816 }
817
818 EXPORT_FOR_TESTS
__btrfs_remove_from_free_space_tree(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,u64 start,u64 size)819 int __btrfs_remove_from_free_space_tree(struct btrfs_trans_handle *trans,
820 struct btrfs_block_group *block_group,
821 struct btrfs_path *path, u64 start, u64 size)
822 {
823 int ret;
824
825 ret = __add_block_group_free_space(trans, block_group, path);
826 if (ret)
827 return ret;
828
829 ret = using_bitmaps(block_group, path);
830 if (ret < 0)
831 return ret;
832
833 if (ret)
834 return modify_free_space_bitmap(trans, block_group, path,
835 start, size, true);
836
837 return remove_free_space_extent(trans, block_group, path, start, size);
838 }
839
btrfs_remove_from_free_space_tree(struct btrfs_trans_handle * trans,u64 start,u64 size)840 int btrfs_remove_from_free_space_tree(struct btrfs_trans_handle *trans,
841 u64 start, u64 size)
842 {
843 struct btrfs_block_group *block_group;
844 BTRFS_PATH_AUTO_FREE(path);
845 int ret;
846
847 if (!btrfs_fs_compat_ro(trans->fs_info, FREE_SPACE_TREE))
848 return 0;
849
850 path = btrfs_alloc_path();
851 if (unlikely(!path)) {
852 ret = -ENOMEM;
853 btrfs_abort_transaction(trans, ret);
854 return ret;
855 }
856
857 block_group = btrfs_lookup_block_group(trans->fs_info, start);
858 if (unlikely(!block_group)) {
859 DEBUG_WARN("no block group found for start=%llu", start);
860 ret = -ENOENT;
861 btrfs_abort_transaction(trans, ret);
862 return ret;
863 }
864
865 mutex_lock(&block_group->free_space_lock);
866 ret = __btrfs_remove_from_free_space_tree(trans, block_group, path, start, size);
867 mutex_unlock(&block_group->free_space_lock);
868 if (ret)
869 btrfs_abort_transaction(trans, ret);
870
871 btrfs_put_block_group(block_group);
872
873 return ret;
874 }
875
add_free_space_extent(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,u64 start,u64 size)876 static int add_free_space_extent(struct btrfs_trans_handle *trans,
877 struct btrfs_block_group *block_group,
878 struct btrfs_path *path,
879 u64 start, u64 size)
880 {
881 struct btrfs_root *root = btrfs_free_space_root(block_group);
882 struct btrfs_key key, new_key;
883 u64 found_start, found_end;
884 u64 end = start + size;
885 int new_extents = 1;
886 int ret;
887
888 /*
889 * We are adding a new extent of free space, but we need to merge
890 * extents. There are four cases here:
891 *
892 * 1. The new extent does not have any immediate neighbors to merge
893 * with: add the new key and increment the free space extent count. We
894 * may need to convert the block group to bitmaps as a result.
895 * 2. The new extent has an immediate neighbor before it: remove the
896 * previous key and insert a new key combining both of them. There is no
897 * net change in the number of extents.
898 * 3. The new extent has an immediate neighbor after it: remove the next
899 * key and insert a new key combining both of them. There is no net
900 * change in the number of extents.
901 * 4. The new extent has immediate neighbors on both sides: remove both
902 * of the keys and insert a new key combining all of them. Where we used
903 * to have two extents, we now have one, so decrement the extent count.
904 */
905
906 new_key.objectid = start;
907 new_key.type = BTRFS_FREE_SPACE_EXTENT_KEY;
908 new_key.offset = size;
909
910 /* Search for a neighbor on the left. */
911 if (start == block_group->start)
912 goto right;
913 key.objectid = start - 1;
914 key.type = (u8)-1;
915 key.offset = (u64)-1;
916
917 ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1);
918 if (ret)
919 return ret;
920
921 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
922
923 if (key.type != BTRFS_FREE_SPACE_EXTENT_KEY) {
924 ASSERT(key.type == BTRFS_FREE_SPACE_INFO_KEY);
925 btrfs_release_path(path);
926 goto right;
927 }
928
929 found_start = key.objectid;
930 found_end = key.objectid + key.offset;
931 ASSERT(found_start >= block_group->start &&
932 found_end > block_group->start);
933 ASSERT(found_start < start && found_end <= start);
934
935 /*
936 * Delete the neighbor on the left and absorb it into the new key (cases
937 * 2 and 4).
938 */
939 if (found_end == start) {
940 ret = btrfs_del_item(trans, root, path);
941 if (ret)
942 return ret;
943 new_key.objectid = found_start;
944 new_key.offset += key.offset;
945 new_extents--;
946 }
947 btrfs_release_path(path);
948
949 right:
950 /* Search for a neighbor on the right. */
951 if (end == btrfs_block_group_end(block_group))
952 goto insert;
953 key.objectid = end;
954 key.type = (u8)-1;
955 key.offset = (u64)-1;
956
957 ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1);
958 if (ret)
959 return ret;
960
961 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
962
963 if (key.type != BTRFS_FREE_SPACE_EXTENT_KEY) {
964 ASSERT(key.type == BTRFS_FREE_SPACE_INFO_KEY);
965 btrfs_release_path(path);
966 goto insert;
967 }
968
969 found_start = key.objectid;
970 found_end = key.objectid + key.offset;
971 ASSERT(found_start >= block_group->start &&
972 found_end > block_group->start);
973 ASSERT((found_start < start && found_end <= start) ||
974 (found_start >= end && found_end > end));
975
976 /*
977 * Delete the neighbor on the right and absorb it into the new key
978 * (cases 3 and 4).
979 */
980 if (found_start == end) {
981 ret = btrfs_del_item(trans, root, path);
982 if (ret)
983 return ret;
984 new_key.offset += key.offset;
985 new_extents--;
986 }
987 btrfs_release_path(path);
988
989 insert:
990 /* Insert the new key (cases 1-4). */
991 ret = btrfs_insert_empty_item(trans, root, path, &new_key, 0);
992 if (ret)
993 return ret;
994
995 btrfs_release_path(path);
996 return update_free_space_extent_count(trans, block_group, path, new_extents);
997 }
998
999 EXPORT_FOR_TESTS
__btrfs_add_to_free_space_tree(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path,u64 start,u64 size)1000 int __btrfs_add_to_free_space_tree(struct btrfs_trans_handle *trans,
1001 struct btrfs_block_group *block_group,
1002 struct btrfs_path *path, u64 start, u64 size)
1003 {
1004 int ret;
1005
1006 ret = __add_block_group_free_space(trans, block_group, path);
1007 if (ret)
1008 return ret;
1009
1010 ret = using_bitmaps(block_group, path);
1011 if (ret < 0)
1012 return ret;
1013
1014 if (ret)
1015 return modify_free_space_bitmap(trans, block_group, path,
1016 start, size, false);
1017
1018 return add_free_space_extent(trans, block_group, path, start, size);
1019 }
1020
btrfs_add_to_free_space_tree(struct btrfs_trans_handle * trans,u64 start,u64 size)1021 int btrfs_add_to_free_space_tree(struct btrfs_trans_handle *trans,
1022 u64 start, u64 size)
1023 {
1024 struct btrfs_block_group *block_group;
1025 BTRFS_PATH_AUTO_FREE(path);
1026 int ret;
1027
1028 if (!btrfs_fs_compat_ro(trans->fs_info, FREE_SPACE_TREE))
1029 return 0;
1030
1031 path = btrfs_alloc_path();
1032 if (unlikely(!path)) {
1033 ret = -ENOMEM;
1034 btrfs_abort_transaction(trans, ret);
1035 return ret;
1036 }
1037
1038 block_group = btrfs_lookup_block_group(trans->fs_info, start);
1039 if (unlikely(!block_group)) {
1040 DEBUG_WARN("no block group found for start=%llu", start);
1041 ret = -ENOENT;
1042 btrfs_abort_transaction(trans, ret);
1043 return ret;
1044 }
1045
1046 mutex_lock(&block_group->free_space_lock);
1047 ret = __btrfs_add_to_free_space_tree(trans, block_group, path, start, size);
1048 mutex_unlock(&block_group->free_space_lock);
1049 if (ret)
1050 btrfs_abort_transaction(trans, ret);
1051
1052 btrfs_put_block_group(block_group);
1053
1054 return ret;
1055 }
1056
1057 /*
1058 * Populate the free space tree by walking the extent tree. Operations on the
1059 * extent tree that happen as a result of writes to the free space tree will go
1060 * through the normal add/remove hooks.
1061 */
populate_free_space_tree(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group)1062 static int populate_free_space_tree(struct btrfs_trans_handle *trans,
1063 struct btrfs_block_group *block_group)
1064 {
1065 struct btrfs_root *extent_root;
1066 BTRFS_PATH_AUTO_FREE(path);
1067 BTRFS_PATH_AUTO_FREE(path2);
1068 struct btrfs_key key;
1069 u64 start, end;
1070 int ret;
1071
1072 path = btrfs_alloc_path();
1073 if (!path)
1074 return -ENOMEM;
1075
1076 path2 = btrfs_alloc_path();
1077 if (!path2)
1078 return -ENOMEM;
1079
1080 path->reada = READA_FORWARD;
1081
1082 ret = add_new_free_space_info(trans, block_group, path2);
1083 if (ret)
1084 return ret;
1085
1086 extent_root = btrfs_extent_root(trans->fs_info, block_group->start);
1087 if (unlikely(!extent_root)) {
1088 btrfs_err(trans->fs_info,
1089 "missing extent root for block group at offset %llu",
1090 block_group->start);
1091 return -EUCLEAN;
1092 }
1093
1094 mutex_lock(&block_group->free_space_lock);
1095
1096 /*
1097 * Iterate through all of the extent and metadata items in this block
1098 * group, adding the free space between them and the free space at the
1099 * end. Note that EXTENT_ITEM and METADATA_ITEM are less than
1100 * BLOCK_GROUP_ITEM, so an extent may precede the block group that it's
1101 * contained in.
1102 */
1103 key.objectid = block_group->start;
1104 key.type = BTRFS_EXTENT_ITEM_KEY;
1105 key.offset = 0;
1106
1107 ret = btrfs_search_slot_for_read(extent_root, &key, path, 1, 0);
1108 if (ret < 0)
1109 goto out_locked;
1110 /*
1111 * If ret is 1 (no key found), it means this is an empty block group,
1112 * without any extents allocated from it and there's no block group
1113 * item (key BTRFS_BLOCK_GROUP_ITEM_KEY) located in the extent tree
1114 * because we are using the block group tree feature (so block group
1115 * items are stored in the block group tree) or this is a new block
1116 * group created in the current transaction and its block group item
1117 * was not yet inserted in the extent tree (that happens in
1118 * btrfs_create_pending_block_groups() -> insert_block_group_item()).
1119 * It also means there are no extents allocated for block groups with a
1120 * start offset beyond this block group's end offset (this is the last,
1121 * highest, block group).
1122 */
1123 start = block_group->start;
1124 end = btrfs_block_group_end(block_group);
1125 while (ret == 0) {
1126 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1127
1128 if (key.type == BTRFS_EXTENT_ITEM_KEY ||
1129 key.type == BTRFS_METADATA_ITEM_KEY) {
1130 if (key.objectid >= end)
1131 break;
1132
1133 if (start < key.objectid) {
1134 ret = __btrfs_add_to_free_space_tree(trans,
1135 block_group,
1136 path2, start,
1137 key.objectid -
1138 start);
1139 if (ret)
1140 goto out_locked;
1141 }
1142 start = key.objectid;
1143 if (key.type == BTRFS_METADATA_ITEM_KEY)
1144 start += trans->fs_info->nodesize;
1145 else
1146 start += key.offset;
1147 } else if (key.type == BTRFS_BLOCK_GROUP_ITEM_KEY) {
1148 if (key.objectid != block_group->start)
1149 break;
1150 }
1151
1152 ret = btrfs_next_item(extent_root, path);
1153 if (ret < 0)
1154 goto out_locked;
1155 }
1156 if (start < end) {
1157 ret = __btrfs_add_to_free_space_tree(trans, block_group, path2,
1158 start, end - start);
1159 if (ret)
1160 goto out_locked;
1161 }
1162
1163 ret = 0;
1164 out_locked:
1165 mutex_unlock(&block_group->free_space_lock);
1166
1167 return ret;
1168 }
1169
btrfs_create_free_space_tree(struct btrfs_fs_info * fs_info)1170 int btrfs_create_free_space_tree(struct btrfs_fs_info *fs_info)
1171 {
1172 struct btrfs_trans_handle *trans;
1173 struct btrfs_root *tree_root = fs_info->tree_root;
1174 struct btrfs_root *free_space_root;
1175 struct btrfs_block_group *block_group;
1176 struct rb_node *node;
1177 int ret;
1178
1179 trans = btrfs_start_transaction(tree_root, 0);
1180 if (IS_ERR(trans))
1181 return PTR_ERR(trans);
1182
1183 set_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
1184 set_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
1185 free_space_root = btrfs_create_tree(trans,
1186 BTRFS_FREE_SPACE_TREE_OBJECTID);
1187 if (IS_ERR(free_space_root)) {
1188 ret = PTR_ERR(free_space_root);
1189 btrfs_abort_transaction(trans, ret);
1190 btrfs_end_transaction(trans);
1191 goto out_clear;
1192 }
1193 ret = btrfs_global_root_insert(free_space_root);
1194 if (unlikely(ret)) {
1195 btrfs_put_root(free_space_root);
1196 btrfs_abort_transaction(trans, ret);
1197 btrfs_end_transaction(trans);
1198 goto out_clear;
1199 }
1200
1201 node = rb_first_cached(&fs_info->block_group_cache_tree);
1202 while (node) {
1203 block_group = rb_entry(node, struct btrfs_block_group,
1204 cache_node);
1205 ret = populate_free_space_tree(trans, block_group);
1206 if (unlikely(ret)) {
1207 btrfs_abort_transaction(trans, ret);
1208 btrfs_end_transaction(trans);
1209 goto out_clear;
1210 }
1211 node = rb_next(node);
1212 }
1213
1214 btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE);
1215 btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE_VALID);
1216 clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
1217 ret = btrfs_commit_transaction(trans);
1218
1219 /*
1220 * Now that we've committed the transaction any reading of our commit
1221 * root will be safe, so we can cache from the free space tree now.
1222 */
1223 clear_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
1224 return ret;
1225
1226 out_clear:
1227 clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
1228 clear_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
1229 return ret;
1230 }
1231
clear_free_space_tree(struct btrfs_trans_handle * trans,struct btrfs_root * root)1232 static int clear_free_space_tree(struct btrfs_trans_handle *trans,
1233 struct btrfs_root *root)
1234 {
1235 BTRFS_PATH_AUTO_FREE(path);
1236 struct btrfs_key key;
1237 struct rb_node *node;
1238 int nr;
1239 int ret;
1240
1241 path = btrfs_alloc_path();
1242 if (!path)
1243 return -ENOMEM;
1244
1245 key.objectid = 0;
1246 key.type = 0;
1247 key.offset = 0;
1248
1249 while (1) {
1250 ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1251 if (ret < 0)
1252 return ret;
1253
1254 nr = btrfs_header_nritems(path->nodes[0]);
1255 if (!nr)
1256 break;
1257
1258 path->slots[0] = 0;
1259 ret = btrfs_del_items(trans, root, path, 0, nr);
1260 if (ret)
1261 return ret;
1262
1263 btrfs_release_path(path);
1264 }
1265
1266 node = rb_first_cached(&trans->fs_info->block_group_cache_tree);
1267 while (node) {
1268 struct btrfs_block_group *bg;
1269
1270 bg = rb_entry(node, struct btrfs_block_group, cache_node);
1271 clear_bit(BLOCK_GROUP_FLAG_FREE_SPACE_ADDED, &bg->runtime_flags);
1272 node = rb_next(node);
1273 cond_resched();
1274 }
1275
1276 return 0;
1277 }
1278
btrfs_delete_free_space_tree(struct btrfs_fs_info * fs_info)1279 int btrfs_delete_free_space_tree(struct btrfs_fs_info *fs_info)
1280 {
1281 struct btrfs_trans_handle *trans;
1282 struct btrfs_root *tree_root = fs_info->tree_root;
1283 struct btrfs_key key = {
1284 .objectid = BTRFS_FREE_SPACE_TREE_OBJECTID,
1285 .type = BTRFS_ROOT_ITEM_KEY,
1286 .offset = 0,
1287 };
1288 struct btrfs_root *free_space_root = btrfs_global_root(fs_info, &key);
1289 int ret;
1290
1291 trans = btrfs_start_transaction(tree_root, 0);
1292 if (IS_ERR(trans))
1293 return PTR_ERR(trans);
1294
1295 btrfs_clear_fs_compat_ro(fs_info, FREE_SPACE_TREE);
1296 btrfs_clear_fs_compat_ro(fs_info, FREE_SPACE_TREE_VALID);
1297
1298 ret = clear_free_space_tree(trans, free_space_root);
1299 if (unlikely(ret)) {
1300 btrfs_abort_transaction(trans, ret);
1301 btrfs_end_transaction(trans);
1302 return ret;
1303 }
1304
1305 ret = btrfs_del_root(trans, &free_space_root->root_key);
1306 if (unlikely(ret)) {
1307 btrfs_abort_transaction(trans, ret);
1308 btrfs_end_transaction(trans);
1309 return ret;
1310 }
1311
1312 btrfs_global_root_delete(free_space_root);
1313
1314 spin_lock(&fs_info->trans_lock);
1315 list_del(&free_space_root->dirty_list);
1316 spin_unlock(&fs_info->trans_lock);
1317
1318 btrfs_tree_lock(free_space_root->node);
1319 btrfs_clear_buffer_dirty(trans, free_space_root->node);
1320 btrfs_tree_unlock(free_space_root->node);
1321 ret = btrfs_free_tree_block(trans, btrfs_root_id(free_space_root),
1322 free_space_root->node, 0, 1);
1323 btrfs_put_root(free_space_root);
1324 if (unlikely(ret < 0)) {
1325 btrfs_abort_transaction(trans, ret);
1326 btrfs_end_transaction(trans);
1327 return ret;
1328 }
1329
1330 return btrfs_commit_transaction(trans);
1331 }
1332
btrfs_rebuild_free_space_tree(struct btrfs_fs_info * fs_info)1333 int btrfs_rebuild_free_space_tree(struct btrfs_fs_info *fs_info)
1334 {
1335 struct btrfs_trans_handle *trans;
1336 struct btrfs_key key = {
1337 .objectid = BTRFS_FREE_SPACE_TREE_OBJECTID,
1338 .type = BTRFS_ROOT_ITEM_KEY,
1339 .offset = 0,
1340 };
1341 struct btrfs_root *free_space_root = btrfs_global_root(fs_info, &key);
1342 struct rb_node *node;
1343 int ret;
1344
1345 trans = btrfs_start_transaction(free_space_root, 1);
1346 if (IS_ERR(trans))
1347 return PTR_ERR(trans);
1348
1349 set_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
1350 set_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
1351
1352 ret = clear_free_space_tree(trans, free_space_root);
1353 if (unlikely(ret)) {
1354 btrfs_abort_transaction(trans, ret);
1355 btrfs_end_transaction(trans);
1356 goto out_clear;
1357 }
1358
1359 node = rb_first_cached(&fs_info->block_group_cache_tree);
1360 while (node) {
1361 struct btrfs_block_group *block_group;
1362
1363 block_group = rb_entry(node, struct btrfs_block_group,
1364 cache_node);
1365
1366 if (test_bit(BLOCK_GROUP_FLAG_FREE_SPACE_ADDED,
1367 &block_group->runtime_flags))
1368 goto next;
1369
1370 ret = populate_free_space_tree(trans, block_group);
1371 if (unlikely(ret)) {
1372 btrfs_abort_transaction(trans, ret);
1373 btrfs_end_transaction(trans);
1374 goto out_clear;
1375 }
1376 next:
1377 if (btrfs_should_end_transaction(trans)) {
1378 btrfs_end_transaction(trans);
1379 trans = btrfs_start_transaction(free_space_root, 1);
1380 if (IS_ERR(trans)) {
1381 ret = PTR_ERR(trans);
1382 goto out_clear;
1383 }
1384 }
1385 node = rb_next(node);
1386 }
1387
1388 btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE);
1389 btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE_VALID);
1390 clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
1391
1392 ret = btrfs_commit_transaction(trans);
1393 clear_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
1394 return ret;
1395
1396 out_clear:
1397 clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
1398 return ret;
1399 }
1400
__add_block_group_free_space(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group,struct btrfs_path * path)1401 static int __add_block_group_free_space(struct btrfs_trans_handle *trans,
1402 struct btrfs_block_group *block_group,
1403 struct btrfs_path *path)
1404 {
1405 bool own_path = false;
1406 int ret;
1407
1408 if (!test_and_clear_bit(BLOCK_GROUP_FLAG_NEEDS_FREE_SPACE,
1409 &block_group->runtime_flags))
1410 return 0;
1411
1412 /*
1413 * While rebuilding the free space tree we may allocate new metadata
1414 * block groups while modifying the free space tree.
1415 *
1416 * Because during the rebuild (at btrfs_rebuild_free_space_tree()) we
1417 * can use multiple transactions, every time btrfs_end_transaction() is
1418 * called at btrfs_rebuild_free_space_tree() we finish the creation of
1419 * new block groups by calling btrfs_create_pending_block_groups(), and
1420 * that in turn calls us, through btrfs_add_block_group_free_space(),
1421 * to add a free space info item and a free space extent item for the
1422 * block group.
1423 *
1424 * Then later btrfs_rebuild_free_space_tree() may find such new block
1425 * groups and processes them with populate_free_space_tree(), which can
1426 * fail with EEXIST since there are already items for the block group in
1427 * the free space tree. Notice that we say "may find" because a new
1428 * block group may be added to the block groups rbtree in a node before
1429 * or after the block group currently being processed by the rebuild
1430 * process. So signal the rebuild process to skip such new block groups
1431 * if it finds them.
1432 */
1433 set_bit(BLOCK_GROUP_FLAG_FREE_SPACE_ADDED, &block_group->runtime_flags);
1434
1435 if (!path) {
1436 path = btrfs_alloc_path();
1437 if (unlikely(!path)) {
1438 btrfs_abort_transaction(trans, -ENOMEM);
1439 return -ENOMEM;
1440 }
1441 own_path = true;
1442 }
1443
1444 ret = add_new_free_space_info(trans, block_group, path);
1445 if (unlikely(ret)) {
1446 btrfs_abort_transaction(trans, ret);
1447 goto out;
1448 }
1449
1450 ret = __btrfs_add_to_free_space_tree(trans, block_group, path,
1451 block_group->start, block_group->length);
1452 if (ret)
1453 btrfs_abort_transaction(trans, ret);
1454
1455 out:
1456 if (own_path)
1457 btrfs_free_path(path);
1458
1459 return ret;
1460 }
1461
btrfs_add_block_group_free_space(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group)1462 int btrfs_add_block_group_free_space(struct btrfs_trans_handle *trans,
1463 struct btrfs_block_group *block_group)
1464 {
1465 int ret;
1466
1467 if (!btrfs_fs_compat_ro(trans->fs_info, FREE_SPACE_TREE))
1468 return 0;
1469
1470 mutex_lock(&block_group->free_space_lock);
1471 ret = __add_block_group_free_space(trans, block_group, NULL);
1472 mutex_unlock(&block_group->free_space_lock);
1473 return ret;
1474 }
1475
btrfs_remove_block_group_free_space(struct btrfs_trans_handle * trans,struct btrfs_block_group * block_group)1476 int btrfs_remove_block_group_free_space(struct btrfs_trans_handle *trans,
1477 struct btrfs_block_group *block_group)
1478 {
1479 struct btrfs_root *root = btrfs_free_space_root(block_group);
1480 BTRFS_PATH_AUTO_FREE(path);
1481 struct btrfs_key key, found_key;
1482 struct extent_buffer *leaf;
1483 u64 start, end;
1484 bool done = false;
1485 int nr;
1486 int ret;
1487
1488 if (!btrfs_fs_compat_ro(trans->fs_info, FREE_SPACE_TREE))
1489 return 0;
1490
1491 if (test_bit(BLOCK_GROUP_FLAG_NEEDS_FREE_SPACE, &block_group->runtime_flags)) {
1492 /* We never added this block group to the free space tree. */
1493 return 0;
1494 }
1495
1496 path = btrfs_alloc_path();
1497 if (unlikely(!path)) {
1498 ret = -ENOMEM;
1499 btrfs_abort_transaction(trans, ret);
1500 return ret;
1501 }
1502
1503 start = block_group->start;
1504 end = btrfs_block_group_end(block_group);
1505
1506 key.objectid = end - 1;
1507 key.type = (u8)-1;
1508 key.offset = (u64)-1;
1509
1510 while (!done) {
1511 ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1);
1512 if (unlikely(ret)) {
1513 btrfs_abort_transaction(trans, ret);
1514 return ret;
1515 }
1516
1517 leaf = path->nodes[0];
1518 nr = 0;
1519 path->slots[0]++;
1520 while (path->slots[0] > 0) {
1521 btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0] - 1);
1522
1523 if (found_key.type == BTRFS_FREE_SPACE_INFO_KEY) {
1524 ASSERT(found_key.objectid == block_group->start);
1525 ASSERT(found_key.offset == block_group->length);
1526 done = true;
1527 nr++;
1528 path->slots[0]--;
1529 break;
1530 } else if (found_key.type == BTRFS_FREE_SPACE_EXTENT_KEY ||
1531 found_key.type == BTRFS_FREE_SPACE_BITMAP_KEY) {
1532 ASSERT(found_key.objectid >= start);
1533 ASSERT(found_key.objectid < end);
1534 ASSERT(found_key.objectid + found_key.offset <= end);
1535 nr++;
1536 path->slots[0]--;
1537 } else {
1538 btrfs_err(trans->fs_info, "unexpected free space tree key type %u",
1539 found_key.type);
1540 ret = -EUCLEAN;
1541 btrfs_abort_transaction(trans, ret);
1542 return ret;
1543 }
1544 }
1545
1546 ret = btrfs_del_items(trans, root, path, path->slots[0], nr);
1547 if (unlikely(ret)) {
1548 btrfs_abort_transaction(trans, ret);
1549 return ret;
1550 }
1551 btrfs_release_path(path);
1552 }
1553
1554 return 0;
1555 }
1556
validate_free_space_key(struct btrfs_block_group * block_group,const struct btrfs_key * key,u8 expected_type)1557 static int validate_free_space_key(struct btrfs_block_group *block_group,
1558 const struct btrfs_key *key, u8 expected_type)
1559 {
1560 const u64 end = btrfs_block_group_end(block_group);
1561
1562 if (unlikely(key->type != expected_type)) {
1563 btrfs_err(block_group->fs_info,
1564 "block group %llu has unexpected free space key type %u, expected %u",
1565 block_group->start, key->type, expected_type);
1566 return -EUCLEAN;
1567 }
1568
1569 if (unlikely(key->objectid + key->offset > end)) {
1570 btrfs_err(block_group->fs_info,
1571 "block group %llu has invalid free space key (%llu %u %llu)",
1572 block_group->start, key->objectid, key->type,
1573 key->offset);
1574 return -EUCLEAN;
1575 }
1576
1577 return 0;
1578 }
1579
load_free_space_bitmaps(struct btrfs_caching_control * caching_ctl,struct btrfs_path * path,u32 expected_extent_count)1580 static int load_free_space_bitmaps(struct btrfs_caching_control *caching_ctl,
1581 struct btrfs_path *path,
1582 u32 expected_extent_count)
1583 {
1584 struct btrfs_block_group *block_group = caching_ctl->block_group;
1585 struct btrfs_fs_info *fs_info = block_group->fs_info;
1586 struct btrfs_root *root;
1587 struct btrfs_key key;
1588 bool prev_bit_set = false;
1589 /* Initialize to silence GCC. */
1590 u64 extent_start = 0;
1591 const u64 end = btrfs_block_group_end(block_group);
1592 u64 offset;
1593 u64 total_found = 0;
1594 u32 extent_count = 0;
1595 int ret;
1596
1597 root = btrfs_free_space_root(block_group);
1598
1599 while (1) {
1600 ret = btrfs_next_item(root, path);
1601 if (ret < 0)
1602 return ret;
1603 if (ret)
1604 break;
1605
1606 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1607
1608 if (key.type == BTRFS_FREE_SPACE_INFO_KEY)
1609 break;
1610
1611 ret = validate_free_space_key(block_group, &key, BTRFS_FREE_SPACE_BITMAP_KEY);
1612 if (unlikely(ret))
1613 return ret;
1614
1615 offset = key.objectid;
1616 while (offset < key.objectid + key.offset) {
1617 bool bit_set;
1618
1619 bit_set = btrfs_free_space_test_bit(block_group, path, offset);
1620 if (!prev_bit_set && bit_set) {
1621 extent_start = offset;
1622 } else if (prev_bit_set && !bit_set) {
1623 u64 space_added;
1624
1625 ret = btrfs_add_new_free_space(block_group,
1626 extent_start,
1627 offset,
1628 &space_added);
1629 if (ret)
1630 return ret;
1631 total_found += space_added;
1632 if (total_found > CACHING_CTL_WAKE_UP) {
1633 total_found = 0;
1634 wake_up(&caching_ctl->wait);
1635 }
1636 extent_count++;
1637 }
1638 prev_bit_set = bit_set;
1639 offset += fs_info->sectorsize;
1640 }
1641 }
1642 if (prev_bit_set) {
1643 ret = btrfs_add_new_free_space(block_group, extent_start, end, NULL);
1644 if (ret)
1645 return ret;
1646 extent_count++;
1647 }
1648
1649 if (unlikely(extent_count != expected_extent_count)) {
1650 btrfs_err(fs_info,
1651 "incorrect extent count for %llu; counted %u, expected %u",
1652 block_group->start, extent_count,
1653 expected_extent_count);
1654 DEBUG_WARN();
1655 return -EIO;
1656 }
1657
1658 return 0;
1659 }
1660
load_free_space_extents(struct btrfs_caching_control * caching_ctl,struct btrfs_path * path,u32 expected_extent_count)1661 static int load_free_space_extents(struct btrfs_caching_control *caching_ctl,
1662 struct btrfs_path *path,
1663 u32 expected_extent_count)
1664 {
1665 struct btrfs_block_group *block_group = caching_ctl->block_group;
1666 struct btrfs_fs_info *fs_info = block_group->fs_info;
1667 struct btrfs_root *root;
1668 struct btrfs_key key;
1669 u64 total_found = 0;
1670 u32 extent_count = 0;
1671 int ret;
1672
1673 root = btrfs_free_space_root(block_group);
1674
1675 while (1) {
1676 u64 space_added;
1677
1678 ret = btrfs_next_item(root, path);
1679 if (ret < 0)
1680 return ret;
1681 if (ret)
1682 break;
1683
1684 btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1685
1686 if (key.type == BTRFS_FREE_SPACE_INFO_KEY)
1687 break;
1688
1689 ret = validate_free_space_key(block_group, &key, BTRFS_FREE_SPACE_EXTENT_KEY);
1690 if (unlikely(ret))
1691 return ret;
1692
1693 ret = btrfs_add_new_free_space(block_group, key.objectid,
1694 key.objectid + key.offset,
1695 &space_added);
1696 if (ret)
1697 return ret;
1698 total_found += space_added;
1699 if (total_found > CACHING_CTL_WAKE_UP) {
1700 total_found = 0;
1701 wake_up(&caching_ctl->wait);
1702 }
1703 extent_count++;
1704 }
1705
1706 if (unlikely(extent_count != expected_extent_count)) {
1707 btrfs_err(fs_info,
1708 "incorrect extent count for %llu; counted %u, expected %u",
1709 block_group->start, extent_count,
1710 expected_extent_count);
1711 DEBUG_WARN();
1712 return -EIO;
1713 }
1714
1715 return 0;
1716 }
1717
btrfs_load_free_space_tree(struct btrfs_caching_control * caching_ctl)1718 int btrfs_load_free_space_tree(struct btrfs_caching_control *caching_ctl)
1719 {
1720 struct btrfs_block_group *block_group;
1721 struct btrfs_free_space_info *info;
1722 BTRFS_PATH_AUTO_FREE(path);
1723 u32 extent_count, flags;
1724
1725 block_group = caching_ctl->block_group;
1726
1727 path = btrfs_alloc_path();
1728 if (!path)
1729 return -ENOMEM;
1730
1731 /*
1732 * Just like caching_thread() doesn't want to deadlock on the extent
1733 * tree, we don't want to deadlock on the free space tree.
1734 */
1735 path->skip_locking = true;
1736 path->search_commit_root = true;
1737 path->reada = READA_FORWARD;
1738
1739 info = btrfs_search_free_space_info(NULL, block_group, path, 0);
1740 if (IS_ERR(info))
1741 return PTR_ERR(info);
1742
1743 extent_count = btrfs_free_space_extent_count(path->nodes[0], info);
1744 flags = btrfs_free_space_flags(path->nodes[0], info);
1745
1746 /*
1747 * We left path pointing to the free space info item, so now
1748 * load_free_space_foo can just iterate through the free space tree from
1749 * there.
1750 */
1751 if (flags & BTRFS_FREE_SPACE_USING_BITMAPS)
1752 return load_free_space_bitmaps(caching_ctl, path, extent_count);
1753 else
1754 return load_free_space_extents(caching_ctl, path, extent_count);
1755 }
1756
delete_orphan_free_space_entries(struct btrfs_root * fst_root,struct btrfs_path * path,u64 first_bg_bytenr)1757 static int delete_orphan_free_space_entries(struct btrfs_root *fst_root,
1758 struct btrfs_path *path,
1759 u64 first_bg_bytenr)
1760 {
1761 struct btrfs_trans_handle *trans;
1762 int ret;
1763
1764 trans = btrfs_start_transaction(fst_root, 1);
1765 if (IS_ERR(trans))
1766 return PTR_ERR(trans);
1767
1768 while (true) {
1769 struct btrfs_key key = { 0 };
1770 int i;
1771
1772 ret = btrfs_search_slot(trans, fst_root, &key, path, -1, 1);
1773 if (ret < 0)
1774 break;
1775 ASSERT(ret > 0);
1776 ret = 0;
1777 for (i = 0; i < btrfs_header_nritems(path->nodes[0]); i++) {
1778 btrfs_item_key_to_cpu(path->nodes[0], &key, i);
1779 if (key.objectid >= first_bg_bytenr) {
1780 /*
1781 * Only break the for() loop and continue to
1782 * delete items.
1783 */
1784 break;
1785 }
1786 }
1787 /* No items to delete, finished. */
1788 if (i == 0)
1789 break;
1790
1791 ret = btrfs_del_items(trans, fst_root, path, 0, i);
1792 if (ret < 0)
1793 break;
1794 btrfs_release_path(path);
1795 }
1796 btrfs_release_path(path);
1797 btrfs_end_transaction(trans);
1798 if (ret == 0)
1799 btrfs_info(fst_root->fs_info, "deleted orphan free space tree entries");
1800 return ret;
1801 }
1802
1803 /* Remove any free space entry before the first block group. */
btrfs_delete_orphan_free_space_entries(struct btrfs_fs_info * fs_info)1804 int btrfs_delete_orphan_free_space_entries(struct btrfs_fs_info *fs_info)
1805 {
1806 BTRFS_PATH_AUTO_RELEASE(path);
1807 struct btrfs_key key = {
1808 .objectid = BTRFS_FREE_SPACE_TREE_OBJECTID,
1809 .type = BTRFS_ROOT_ITEM_KEY,
1810 .offset = 0,
1811 };
1812 struct btrfs_root *root;
1813 struct btrfs_block_group *bg;
1814 u64 first_bg_bytenr;
1815 int ret;
1816
1817 /*
1818 * Extent tree v2 has multiple global roots based on the block group.
1819 * This means we cannot easily grab the global free space tree and locate
1820 * orphan items. Furthermore this is still experimental, all users
1821 * should use the latest btrfs-progs anyway.
1822 */
1823 if (btrfs_fs_incompat(fs_info, EXTENT_TREE_V2))
1824 return 0;
1825 if (!btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE))
1826 return 0;
1827 root = btrfs_global_root(fs_info, &key);
1828 if (!root)
1829 return 0;
1830
1831 key.objectid = 0;
1832 key.type = 0;
1833 key.offset = 0;
1834
1835 bg = btrfs_lookup_first_block_group(fs_info, 0);
1836 if (unlikely(!bg)) {
1837 btrfs_err(fs_info, "no block group found");
1838 return -EUCLEAN;
1839 }
1840 first_bg_bytenr = bg->start;
1841 btrfs_put_block_group(bg);
1842
1843 ret = btrfs_search_slot(NULL, root, &key, &path, 0, 0);
1844 if (ret < 0)
1845 return ret;
1846 /* There should not be an all-zero key in fst. */
1847 ASSERT(ret > 0);
1848
1849 /* Empty free space tree. */
1850 if (path.slots[0] >= btrfs_header_nritems(path.nodes[0]))
1851 return 0;
1852
1853 btrfs_item_key_to_cpu(path.nodes[0], &key, path.slots[0]);
1854 if (key.objectid >= first_bg_bytenr)
1855 return 0;
1856 btrfs_release_path(&path);
1857 return delete_orphan_free_space_entries(root, &path, first_bg_bytenr);
1858 }
1859