summaryrefslogtreecommitdiff
path: root/free-space-cache.c
blob: 4bf4a6cb20722dffee5d79af5121fd375d61bd75 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
/*
 * Copyright (C) 2008 Red Hat.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include "kerncompat.h"
#include "ctree.h"
#include "free-space-cache.h"
#include "transaction.h"
#include "disk-io.h"
#include "extent_io.h"
#include "crc32c.h"
#include "bitops.h"
#include "internal.h"
#include "utils.h"

/*
 * Kernel always uses PAGE_CACHE_SIZE for sectorsize, but we don't have
 * anything like that in userspace and have to get the value from the
 * filesystem
 */
#define BITS_PER_BITMAP(sectorsize)		((sectorsize) * 8)
#define MAX_CACHE_BYTES_PER_GIG	SZ_32K

static int link_free_space(struct btrfs_free_space_ctl *ctl,
			   struct btrfs_free_space *info);
static void merge_space_tree(struct btrfs_free_space_ctl *ctl);

struct io_ctl {
	void *cur, *orig;
	void *buffer;
	struct btrfs_root *root;
	unsigned long size;
	u64 total_size;
	int index;
	int num_pages;
	unsigned check_crcs:1;
};

static int io_ctl_init(struct io_ctl *io_ctl, u64 size, u64 ino,
		       struct btrfs_root *root)
{
	memset(io_ctl, 0, sizeof(struct io_ctl));
	io_ctl->num_pages = (size + root->fs_info->sectorsize - 1) /
				root->fs_info->sectorsize;
	io_ctl->buffer = kzalloc(size, GFP_NOFS);
	if (!io_ctl->buffer)
		return -ENOMEM;
	io_ctl->total_size = size;
	io_ctl->root = root;
	if (ino != BTRFS_FREE_INO_OBJECTID)
		io_ctl->check_crcs = 1;
	return 0;
}

static void io_ctl_free(struct io_ctl *io_ctl)
{
	kfree(io_ctl->buffer);
}

static void io_ctl_unmap_page(struct io_ctl *io_ctl)
{
	if (io_ctl->cur) {
		io_ctl->cur = NULL;
		io_ctl->orig = NULL;
	}
}

static void io_ctl_map_page(struct io_ctl *io_ctl, int clear)
{
	BUG_ON(io_ctl->index >= io_ctl->num_pages);
	io_ctl->cur = io_ctl->buffer + (io_ctl->index++ *
					io_ctl->root->fs_info->sectorsize);
	io_ctl->orig = io_ctl->cur;
	io_ctl->size = io_ctl->root->fs_info->sectorsize;
	if (clear)
		memset(io_ctl->cur, 0, io_ctl->root->fs_info->sectorsize);
}

static void io_ctl_drop_pages(struct io_ctl *io_ctl)
{
	io_ctl_unmap_page(io_ctl);
}

static int io_ctl_prepare_pages(struct io_ctl *io_ctl, struct btrfs_root *root,
				struct btrfs_path *path, u64 ino)
{
	struct extent_buffer *leaf;
	struct btrfs_file_extent_item *fi;
	struct btrfs_key key;
	u64 bytenr, len;
	u64 total_read = 0;
	int ret = 0;

	key.objectid = ino;
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = 0;

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret) {
		fprintf(stderr,
		       "Couldn't find file extent item for free space inode"
		       " %Lu\n", ino);
		btrfs_release_path(path);
		return -EINVAL;
	}

	while (total_read < io_ctl->total_size) {
		if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
			ret = btrfs_next_leaf(root, path);
			if (ret) {
				ret = -EINVAL;
				break;
			}
		}
		leaf = path->nodes[0];

		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
		if (key.objectid != ino) {
			ret = -EINVAL;
			break;
		}

		if (key.type != BTRFS_EXTENT_DATA_KEY) {
			ret = -EINVAL;
			break;
		}

		fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
				    struct btrfs_file_extent_item);
		if (btrfs_file_extent_type(path->nodes[0], fi) !=
		    BTRFS_FILE_EXTENT_REG) {
			fprintf(stderr, "Not the file extent type we wanted\n");
			ret = -EINVAL;
			break;
		}

		bytenr = btrfs_file_extent_disk_bytenr(leaf, fi) +
			btrfs_file_extent_offset(leaf, fi);
		len = btrfs_file_extent_num_bytes(leaf, fi);
		ret = read_data_from_disk(root->fs_info,
					  io_ctl->buffer + key.offset, bytenr,
					  len, 0);
		if (ret)
			break;
		total_read += len;
		path->slots[0]++;
	}

	btrfs_release_path(path);
	return ret;
}

static int io_ctl_check_generation(struct io_ctl *io_ctl, u64 generation)
{
	__le64 *gen;

	/*
	 * Skip the crc area.  If we don't check crcs then we just have a 64bit
	 * chunk at the front of the first page.
	 */
	if (io_ctl->check_crcs) {
		io_ctl->cur += sizeof(u32) * io_ctl->num_pages;
		io_ctl->size -= sizeof(u64) +
			(sizeof(u32) * io_ctl->num_pages);
	} else {
		io_ctl->cur += sizeof(u64);
		io_ctl->size -= sizeof(u64) * 2;
	}

	gen = io_ctl->cur;
	if (le64_to_cpu(*gen) != generation) {
		printk("btrfs: space cache generation "
		       "(%Lu) does not match inode (%Lu)\n", *gen,
		       generation);
		io_ctl_unmap_page(io_ctl);
		return -EIO;
	}
	io_ctl->cur += sizeof(u64);
	return 0;
}

static int io_ctl_check_crc(struct io_ctl *io_ctl, int index)
{
	u32 *tmp, val;
	u32 crc = ~(u32)0;
	unsigned offset = 0;

	if (!io_ctl->check_crcs) {
		io_ctl_map_page(io_ctl, 0);
		return 0;
	}

	if (index == 0)
		offset = sizeof(u32) * io_ctl->num_pages;

	tmp = io_ctl->buffer;
	tmp += index;
	val = *tmp;

	io_ctl_map_page(io_ctl, 0);
	crc = crc32c(crc, io_ctl->orig + offset,
			io_ctl->root->fs_info->sectorsize - offset);
	btrfs_csum_final(crc, (u8 *)&crc);
	if (val != crc) {
		printk("btrfs: csum mismatch on free space cache\n");
		io_ctl_unmap_page(io_ctl);
		return -EIO;
	}

	return 0;
}

static int io_ctl_read_entry(struct io_ctl *io_ctl,
			    struct btrfs_free_space *entry, u8 *type)
{
	struct btrfs_free_space_entry *e;
	int ret;

	if (!io_ctl->cur) {
		ret = io_ctl_check_crc(io_ctl, io_ctl->index);
		if (ret)
			return ret;
	}

	e = io_ctl->cur;
	entry->offset = le64_to_cpu(e->offset);
	entry->bytes = le64_to_cpu(e->bytes);
	*type = e->type;
	io_ctl->cur += sizeof(struct btrfs_free_space_entry);
	io_ctl->size -= sizeof(struct btrfs_free_space_entry);

	if (io_ctl->size >= sizeof(struct btrfs_free_space_entry))
		return 0;

	io_ctl_unmap_page(io_ctl);

	return 0;
}

static int io_ctl_read_bitmap(struct io_ctl *io_ctl,
			      struct btrfs_free_space *entry)
{
	int ret;

	ret = io_ctl_check_crc(io_ctl, io_ctl->index);
	if (ret)
		return ret;

	memcpy(entry->bitmap, io_ctl->cur, io_ctl->root->fs_info->sectorsize);
	io_ctl_unmap_page(io_ctl);

	return 0;
}


static int __load_free_space_cache(struct btrfs_root *root,
			    struct btrfs_free_space_ctl *ctl,
			    struct btrfs_path *path, u64 offset)
{
	struct btrfs_free_space_header *header;
	struct btrfs_inode_item *inode_item;
	struct extent_buffer *leaf;
	struct io_ctl io_ctl;
	struct btrfs_key key;
	struct btrfs_key inode_location;
	struct btrfs_disk_key disk_key;
	struct btrfs_free_space *e, *n;
	struct list_head bitmaps;
	u64 num_entries;
	u64 num_bitmaps;
	u64 generation;
	u64 inode_size;
	u8 type;
	int ret = 0;

	INIT_LIST_HEAD(&bitmaps);

	key.objectid = BTRFS_FREE_SPACE_OBJECTID;
	key.offset = offset;
	key.type = 0;

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0) {
		return 0;
	} else if (ret > 0) {
		btrfs_release_path(path);
		return 0;
	}

	leaf = path->nodes[0];
	header = btrfs_item_ptr(leaf, path->slots[0],
				struct btrfs_free_space_header);
	num_entries = btrfs_free_space_entries(leaf, header);
	num_bitmaps = btrfs_free_space_bitmaps(leaf, header);
	generation = btrfs_free_space_generation(leaf, header);
	btrfs_free_space_key(leaf, header, &disk_key);
	btrfs_disk_key_to_cpu(&inode_location, &disk_key);
	btrfs_release_path(path);

	ret = btrfs_search_slot(NULL, root, &inode_location, path, 0, 0);
	if (ret) {
		fprintf(stderr, "Couldn't find free space inode %d\n", ret);
		return 0;
	}

	leaf = path->nodes[0];
	inode_item = btrfs_item_ptr(leaf, path->slots[0],
				    struct btrfs_inode_item);

	inode_size = btrfs_inode_size(leaf, inode_item);
	if (!inode_size || !btrfs_inode_generation(leaf, inode_item)) {
		btrfs_release_path(path);
		return 0;
	}

	if (btrfs_inode_generation(leaf, inode_item) != generation) {
		fprintf(stderr,
		       "free space inode generation (%llu) did not match "
		       "free space cache generation (%llu)\n",
		       (unsigned long long)btrfs_inode_generation(leaf,
								  inode_item),
		       (unsigned long long)generation);
		btrfs_release_path(path);
		return 0;
	}

	btrfs_release_path(path);

	if (!num_entries)
		return 0;

	ret = io_ctl_init(&io_ctl, inode_size, inode_location.objectid, root);
	if (ret)
		return ret;

	ret = io_ctl_prepare_pages(&io_ctl, root, path,
				   inode_location.objectid);
	if (ret)
		goto out;

	ret = io_ctl_check_crc(&io_ctl, 0);
	if (ret)
		goto free_cache;

	ret = io_ctl_check_generation(&io_ctl, generation);
	if (ret)
		goto free_cache;

	while (num_entries) {
		e = calloc(1, sizeof(*e));
		if (!e)
			goto free_cache;

		ret = io_ctl_read_entry(&io_ctl, e, &type);
		if (ret) {
			free(e);
			goto free_cache;
		}

		if (!e->bytes) {
			free(e);
			goto free_cache;
		}

		if (type == BTRFS_FREE_SPACE_EXTENT) {
			ret = link_free_space(ctl, e);
			if (ret) {
				fprintf(stderr,
				       "Duplicate entries in free space cache\n");
				free(e);
				goto free_cache;
			}
		} else {
			BUG_ON(!num_bitmaps);
			num_bitmaps--;
			e->bitmap = kzalloc(ctl->sectorsize, GFP_NOFS);
			if (!e->bitmap) {
				free(e);
				goto free_cache;
			}
			ret = link_free_space(ctl, e);
			ctl->total_bitmaps++;
			if (ret) {
				fprintf(stderr,
				       "Duplicate entries in free space cache\n");
				free(e->bitmap);
				free(e);
				goto free_cache;
			}
			list_add_tail(&e->list, &bitmaps);
		}

		num_entries--;
	}

	io_ctl_unmap_page(&io_ctl);

	/*
	 * We add the bitmaps at the end of the entries in order that
	 * the bitmap entries are added to the cache.
	 */
	list_for_each_entry_safe(e, n, &bitmaps, list) {
		list_del_init(&e->list);
		ret = io_ctl_read_bitmap(&io_ctl, e);
		if (ret)
			goto free_cache;
	}

	io_ctl_drop_pages(&io_ctl);
	merge_space_tree(ctl);
	ret = 1;
out:
	io_ctl_free(&io_ctl);
	return ret;
free_cache:
	io_ctl_drop_pages(&io_ctl);
	__btrfs_remove_free_space_cache(ctl);
	goto out;
}

int load_free_space_cache(struct btrfs_fs_info *fs_info,
			  struct btrfs_block_group_cache *block_group)
{
	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
	struct btrfs_path *path;
	u64 used = btrfs_block_group_used(&block_group->item);
	int ret = 0;
	int matched;

	path = btrfs_alloc_path();
	if (!path)
		return 0;

	ret = __load_free_space_cache(fs_info->tree_root, ctl, path,
				      block_group->key.objectid);
	btrfs_free_path(path);

	matched = (ctl->free_space == (block_group->key.offset - used -
				       block_group->bytes_super));
	if (ret == 1 && !matched) {
		__btrfs_remove_free_space_cache(ctl);
		fprintf(stderr,
		       "block group %llu has wrong amount of free space\n",
		       block_group->key.objectid);
		ret = -1;
	}

	if (ret < 0) {
		ret = 0;

		fprintf(stderr,
		       "failed to load free space cache for block group %llu\n",
		       block_group->key.objectid);
	}

	return ret;
}

static inline unsigned long offset_to_bit(u64 bitmap_start, u32 unit,
					  u64 offset)
{
	BUG_ON(offset < bitmap_start);
	offset -= bitmap_start;
	return (unsigned long)(offset / unit);
}

static inline unsigned long bytes_to_bits(u64 bytes, u32 unit)
{
	return (unsigned long)(bytes / unit);
}

static int tree_insert_offset(struct rb_root *root, u64 offset,
			      struct rb_node *node, int bitmap)
{
	struct rb_node **p = &root->rb_node;
	struct rb_node *parent = NULL;
	struct btrfs_free_space *info;

	while (*p) {
		parent = *p;
		info = rb_entry(parent, struct btrfs_free_space, offset_index);

		if (offset < info->offset) {
			p = &(*p)->rb_left;
		} else if (offset > info->offset) {
			p = &(*p)->rb_right;
		} else {
			/*
			 * we could have a bitmap entry and an extent entry
			 * share the same offset.  If this is the case, we want
			 * the extent entry to always be found first if we do a
			 * linear search through the tree, since we want to have
			 * the quickest allocation time, and allocating from an
			 * extent is faster than allocating from a bitmap.  So
			 * if we're inserting a bitmap and we find an entry at
			 * this offset, we want to go right, or after this entry
			 * logically.  If we are inserting an extent and we've
			 * found a bitmap, we want to go left, or before
			 * logically.
			 */
			if (bitmap) {
				if (info->bitmap)
					return -EEXIST;
				p = &(*p)->rb_right;
			} else {
				if (!info->bitmap)
					return -EEXIST;
				p = &(*p)->rb_left;
			}
		}
	}

	rb_link_node(node, parent, p);
	rb_insert_color(node, root);

	return 0;
}

/*
 * searches the tree for the given offset.
 *
 * fuzzy - If this is set, then we are trying to make an allocation, and we just
 * want a section that has at least bytes size and comes at or after the given
 * offset.
 */
static struct btrfs_free_space *
tree_search_offset(struct btrfs_free_space_ctl *ctl,
		   u64 offset, int bitmap_only, int fuzzy)
{
	struct rb_node *n = ctl->free_space_offset.rb_node;
	struct btrfs_free_space *entry, *prev = NULL;
	u32 sectorsize = ctl->sectorsize;

	/* find entry that is closest to the 'offset' */
	while (1) {
		if (!n) {
			entry = NULL;
			break;
		}

		entry = rb_entry(n, struct btrfs_free_space, offset_index);
		prev = entry;

		if (offset < entry->offset)
			n = n->rb_left;
		else if (offset > entry->offset)
			n = n->rb_right;
		else
			break;
	}

	if (bitmap_only) {
		if (!entry)
			return NULL;
		if (entry->bitmap)
			return entry;

		/*
		 * bitmap entry and extent entry may share same offset,
		 * in that case, bitmap entry comes after extent entry.
		 */
		n = rb_next(n);
		if (!n)
			return NULL;
		entry = rb_entry(n, struct btrfs_free_space, offset_index);
		if (entry->offset != offset)
			return NULL;

		WARN_ON(!entry->bitmap);
		return entry;
	} else if (entry) {
		if (entry->bitmap) {
			/*
			 * if previous extent entry covers the offset,
			 * we should return it instead of the bitmap entry
			 */
			n = rb_prev(&entry->offset_index);
			if (n) {
				prev = rb_entry(n, struct btrfs_free_space,
						offset_index);
				if (!prev->bitmap &&
				    prev->offset + prev->bytes > offset)
					entry = prev;
			}
		}
		return entry;
	}

	if (!prev)
		return NULL;

	/* find last entry before the 'offset' */
	entry = prev;
	if (entry->offset > offset) {
		n = rb_prev(&entry->offset_index);
		if (n) {
			entry = rb_entry(n, struct btrfs_free_space,
					offset_index);
			BUG_ON(entry->offset > offset);
		} else {
			if (fuzzy)
				return entry;
			else
				return NULL;
		}
	}

	if (entry->bitmap) {
		n = rb_prev(&entry->offset_index);
		if (n) {
			prev = rb_entry(n, struct btrfs_free_space,
					offset_index);
			if (!prev->bitmap &&
			    prev->offset + prev->bytes > offset)
				return prev;
		}
		if (entry->offset + BITS_PER_BITMAP(sectorsize) * ctl->unit > offset)
			return entry;
	} else if (entry->offset + entry->bytes > offset)
		return entry;

	if (!fuzzy)
		return NULL;

	while (1) {
		if (entry->bitmap) {
			if (entry->offset + BITS_PER_BITMAP(sectorsize) *
			    ctl->unit > offset)
				break;
		} else {
			if (entry->offset + entry->bytes > offset)
				break;
		}

		n = rb_next(&entry->offset_index);
		if (!n)
			return NULL;
		entry = rb_entry(n, struct btrfs_free_space, offset_index);
	}
	return entry;
}

void unlink_free_space(struct btrfs_free_space_ctl *ctl,
		       struct btrfs_free_space *info)
{
	rb_erase(&info->offset_index, &ctl->free_space_offset);
	ctl->free_extents--;
	ctl->free_space -= info->bytes;
}

static int link_free_space(struct btrfs_free_space_ctl *ctl,
			   struct btrfs_free_space *info)
{
	int ret = 0;

	BUG_ON(!info->bitmap && !info->bytes);
	ret = tree_insert_offset(&ctl->free_space_offset, info->offset,
				 &info->offset_index, (info->bitmap != NULL));
	if (ret)
		return ret;

	ctl->free_space += info->bytes;
	ctl->free_extents++;
	return ret;
}

static int search_bitmap(struct btrfs_free_space_ctl *ctl,
			 struct btrfs_free_space *bitmap_info, u64 *offset,
			 u64 *bytes)
{
	unsigned long found_bits = 0;
	unsigned long bits, i;
	unsigned long next_zero;
	u32 sectorsize = ctl->sectorsize;

	i = offset_to_bit(bitmap_info->offset, ctl->unit,
			  max_t(u64, *offset, bitmap_info->offset));
	bits = bytes_to_bits(*bytes, ctl->unit);

	for_each_set_bit_from(i, bitmap_info->bitmap, BITS_PER_BITMAP(sectorsize)) {
		next_zero = find_next_zero_bit(bitmap_info->bitmap,
					       BITS_PER_BITMAP(sectorsize), i);
		if ((next_zero - i) >= bits) {
			found_bits = next_zero - i;
			break;
		}
		i = next_zero;
	}

	if (found_bits) {
		*offset = (u64)(i * ctl->unit) + bitmap_info->offset;
		*bytes = (u64)(found_bits) * ctl->unit;
		return 0;
	}

	return -1;
}

struct btrfs_free_space *
btrfs_find_free_space(struct btrfs_free_space_ctl *ctl, u64 offset, u64 bytes)
{
	return tree_search_offset(ctl, offset, 0, 0);
}

static void try_merge_free_space(struct btrfs_free_space_ctl *ctl,
				struct btrfs_free_space *info)
{
	struct btrfs_free_space *left_info;
	struct btrfs_free_space *right_info;
	u64 offset = info->offset;
	u64 bytes = info->bytes;

	/*
	 * first we want to see if there is free space adjacent to the range we
	 * are adding, if there is remove that struct and add a new one to
	 * cover the entire range
	 */
	right_info = tree_search_offset(ctl, offset + bytes, 0, 0);
	if (right_info && rb_prev(&right_info->offset_index))
		left_info = rb_entry(rb_prev(&right_info->offset_index),
				     struct btrfs_free_space, offset_index);
	else
		left_info = tree_search_offset(ctl, offset - 1, 0, 0);

	if (right_info && !right_info->bitmap) {
		unlink_free_space(ctl, right_info);
		info->bytes += right_info->bytes;
		free(right_info);
	}

	if (left_info && !left_info->bitmap &&
	    left_info->offset + left_info->bytes == offset) {
		unlink_free_space(ctl, left_info);
		info->offset = left_info->offset;
		info->bytes += left_info->bytes;
		free(left_info);
	}
}

void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
			   u64 bytes)
{
	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
	struct btrfs_free_space *info;
	struct rb_node *n;
	int count = 0;

	for (n = rb_first(&ctl->free_space_offset); n; n = rb_next(n)) {
		info = rb_entry(n, struct btrfs_free_space, offset_index);
		if (info->bytes >= bytes && !block_group->ro)
			count++;
		printk("entry offset %llu, bytes %llu, bitmap %s\n",
		       (unsigned long long)info->offset,
		       (unsigned long long)info->bytes,
		       (info->bitmap) ? "yes" : "no");
	}
	printk("%d blocks of free space at or bigger than bytes is \n", count);
}

int btrfs_init_free_space_ctl(struct btrfs_block_group_cache *block_group,
			      int sectorsize)
{
	struct btrfs_free_space_ctl *ctl;

	ctl = calloc(1, sizeof(*ctl));
	if (!ctl)
		return -ENOMEM;

	ctl->sectorsize = sectorsize;
	ctl->unit = sectorsize;
	ctl->start = block_group->key.objectid;
	ctl->private = block_group;
	block_group->free_space_ctl = ctl;

	return 0;
}

void __btrfs_remove_free_space_cache(struct btrfs_free_space_ctl *ctl)
{
	struct btrfs_free_space *info;
	struct rb_node *node;

	while ((node = rb_last(&ctl->free_space_offset)) != NULL) {
		info = rb_entry(node, struct btrfs_free_space, offset_index);
		unlink_free_space(ctl, info);
		free(info->bitmap);
		free(info);
	}
}

void btrfs_remove_free_space_cache(struct btrfs_block_group_cache *block_group)
{
	__btrfs_remove_free_space_cache(block_group->free_space_ctl);
}

int btrfs_add_free_space(struct btrfs_free_space_ctl *ctl, u64 offset,
			 u64 bytes)
{
	struct btrfs_free_space *info;
	int ret = 0;

	info = calloc(1, sizeof(*info));
	if (!info)
		return -ENOMEM;

	info->offset = offset;
	info->bytes = bytes;

	try_merge_free_space(ctl, info);

	ret = link_free_space(ctl, info);
	if (ret) {
		printk(KERN_CRIT "btrfs: unable to add free space :%d\n", ret);
		BUG_ON(ret == -EEXIST);
	}

	return ret;
}

/*
 * Merges all the free space cache and kills the bitmap entries since we just
 * want to use the free space cache to verify it's correct, no reason to keep
 * the bitmaps around to confuse things.
 */
static void merge_space_tree(struct btrfs_free_space_ctl *ctl)
{
	struct btrfs_free_space *e, *prev = NULL;
	struct rb_node *n;
	int ret;
	u32 sectorsize = ctl->sectorsize;

again:
	prev = NULL;
	for (n = rb_first(&ctl->free_space_offset); n; n = rb_next(n)) {
		e = rb_entry(n, struct btrfs_free_space, offset_index);
		if (e->bitmap) {
			u64 offset = e->offset, bytes = ctl->unit;
			u64 end;

			end = e->offset + (u64)(BITS_PER_BITMAP(sectorsize) * ctl->unit);

			unlink_free_space(ctl, e);
			while (!(search_bitmap(ctl, e, &offset, &bytes))) {
				ret = btrfs_add_free_space(ctl, offset,
							   bytes);
				BUG_ON(ret);
				offset += bytes;
				if (offset >= end)
					break;
				bytes = ctl->unit;
			}
			free(e->bitmap);
			free(e);
			goto again;
		}
		if (!prev)
			goto next;
		if (prev->offset + prev->bytes == e->offset) {
			unlink_free_space(ctl, prev);
			unlink_free_space(ctl, e);
			prev->bytes += e->bytes;
			free(e);
			link_free_space(ctl, prev);
			goto again;
		}
next:
		prev = e;
	}
}

int btrfs_clear_free_space_cache(struct btrfs_fs_info *fs_info,
				 struct btrfs_block_group_cache *bg)
{
	struct btrfs_trans_handle *trans;
	struct btrfs_root *tree_root = fs_info->tree_root;
	struct btrfs_path path;
	struct btrfs_key key;
	struct btrfs_disk_key location;
	struct btrfs_free_space_header *sc_header;
	struct extent_buffer *node;
	u64 ino;
	int slot;
	int ret;

	trans = btrfs_start_transaction(tree_root, 1);
	if (IS_ERR(trans))
		return PTR_ERR(trans);

	btrfs_init_path(&path);

	key.objectid = BTRFS_FREE_SPACE_OBJECTID;
	key.type = 0;
	key.offset = bg->key.objectid;

	ret = btrfs_search_slot(trans, tree_root, &key, &path, -1, 1);
	if (ret > 0) {
		ret = 0;
		goto out;
	}
	if (ret < 0)
		goto out;

	node = path.nodes[0];
	slot = path.slots[0];
	sc_header = btrfs_item_ptr(node, slot, struct btrfs_free_space_header);
	btrfs_free_space_key(node, sc_header, &location);
	ino = location.objectid;

	/* Delete the free space header, as we have the ino to continue */
	ret = btrfs_del_item(trans, tree_root, &path);
	if (ret < 0) {
		error("failed to remove free space header for block group %llu: %d",
		      bg->key.objectid, ret);
		goto out;
	}
	btrfs_release_path(&path);

	/* Iterate from the end of the free space cache inode */
	key.objectid = ino;
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = (u64)-1;
	ret = btrfs_search_slot(trans, tree_root, &key, &path, -1, 1);
	if (ret < 0) {
		error("failed to locate free space cache extent for block group %llu: %d",
		      bg->key.objectid, ret);
		goto out;
	}
	while (1) {
		struct btrfs_file_extent_item *fi;
		u64 disk_bytenr;
		u64 disk_num_bytes;

		ret = btrfs_previous_item(tree_root, &path, ino,
					  BTRFS_EXTENT_DATA_KEY);
		if (ret > 0) {
			ret = 0;
			break;
		}
		if (ret < 0) {
			error(
	"failed to locate free space cache extent for block group %llu: %d",
				bg->key.objectid, ret);
			goto out;
		}
		node = path.nodes[0];
		slot = path.slots[0];
		btrfs_item_key_to_cpu(node, &key, slot);
		fi = btrfs_item_ptr(node, slot, struct btrfs_file_extent_item);
		disk_bytenr = btrfs_file_extent_disk_bytenr(node, fi);
		disk_num_bytes = btrfs_file_extent_disk_num_bytes(node, fi);

		ret = btrfs_free_extent(trans, tree_root, disk_bytenr,
					disk_num_bytes, 0, tree_root->objectid,
					ino, key.offset);
		if (ret < 0) {
			error("failed to remove backref for disk bytenr %llu: %d",
			      disk_bytenr, ret);
			goto out;
		}
		ret = btrfs_del_item(trans, tree_root, &path);
		if (ret < 0) {
			error(
	"failed to remove free space extent data for ino %llu offset %llu: %d",
			      ino, key.offset, ret);
			goto out;
		}
	}
	btrfs_release_path(&path);

	/* Now delete free space cache inode item */
	key.objectid = ino;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;

	ret = btrfs_search_slot(trans, tree_root, &key, &path, -1, 1);
	if (ret > 0)
		warning("free space inode %llu not found, ignore", ino);
	if (ret < 0) {
		error(
	"failed to locate free space cache inode %llu for block group %llu: %d",
		      ino, bg->key.objectid, ret);
		goto out;
	}
	ret = btrfs_del_item(trans, tree_root, &path);
	if (ret < 0) {
		error(
	"failed to delete free space cache inode %llu for block group %llu: %d",
		      ino, bg->key.objectid, ret);
	}
out:
	btrfs_release_path(&path);
	if (!ret)
		btrfs_commit_transaction(trans, tree_root);
	return ret;
}