summaryrefslogtreecommitdiff
path: root/drivers/md/dm-clone-metadata.c
blob: 6bc8c1d1c351c8441769739b84d6d89a701a982c (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
// SPDX-License-Identifier: GPL-2.0-only
/*
 * Copyright (C) 2019 Arrikto, Inc. All Rights Reserved.
 */

#include <linux/mm.h>
#include <linux/err.h>
#include <linux/slab.h>
#include <linux/rwsem.h>
#include <linux/bitops.h>
#include <linux/bitmap.h>
#include <linux/device-mapper.h>

#include "persistent-data/dm-bitset.h"
#include "persistent-data/dm-space-map.h"
#include "persistent-data/dm-block-manager.h"
#include "persistent-data/dm-transaction-manager.h"

#include "dm-clone-metadata.h"

#define DM_MSG_PREFIX "clone metadata"

#define SUPERBLOCK_LOCATION 0
#define SUPERBLOCK_MAGIC 0x8af27f64
#define SUPERBLOCK_CSUM_XOR 257649492

#define DM_CLONE_MAX_CONCURRENT_LOCKS 5

#define UUID_LEN 16

/* Min and max dm-clone metadata versions supported */
#define DM_CLONE_MIN_METADATA_VERSION 1
#define DM_CLONE_MAX_METADATA_VERSION 1

/*
 * On-disk metadata layout
 */
struct superblock_disk {
	__le32 csum;
	__le32 flags;
	__le64 blocknr;

	__u8 uuid[UUID_LEN];
	__le64 magic;
	__le32 version;

	__u8 metadata_space_map_root[SPACE_MAP_ROOT_SIZE];

	__le64 region_size;
	__le64 target_size;

	__le64 bitset_root;
} __packed;

/*
 * Region and Dirty bitmaps.
 *
 * dm-clone logically splits the source and destination devices in regions of
 * fixed size. The destination device's regions are gradually hydrated, i.e.,
 * we copy (clone) the source's regions to the destination device. Eventually,
 * all regions will get hydrated and all I/O will be served from the
 * destination device.
 *
 * We maintain an on-disk bitmap which tracks the state of each of the
 * destination device's regions, i.e., whether they are hydrated or not.
 *
 * To save constantly doing look ups on disk we keep an in core copy of the
 * on-disk bitmap, the region_map.
 *
 * To further reduce metadata I/O overhead we use a second bitmap, the dmap
 * (dirty bitmap), which tracks the dirty words, i.e. longs, of the region_map.
 *
 * When a region finishes hydrating dm-clone calls
 * dm_clone_set_region_hydrated(), or for discard requests
 * dm_clone_cond_set_range(), which sets the corresponding bits in region_map
 * and dmap.
 *
 * During a metadata commit we scan the dmap for dirty region_map words (longs)
 * and update accordingly the on-disk metadata. Thus, we don't have to flush to
 * disk the whole region_map. We can just flush the dirty region_map words.
 *
 * We use a dirty bitmap, which is smaller than the original region_map, to
 * reduce the amount of memory accesses during a metadata commit. As dm-bitset
 * accesses the on-disk bitmap in 64-bit word granularity, there is no
 * significant benefit in tracking the dirty region_map bits with a smaller
 * granularity.
 *
 * We could update directly the on-disk bitmap, when dm-clone calls either
 * dm_clone_set_region_hydrated() or dm_clone_cond_set_range(), buts this
 * inserts significant metadata I/O overhead in dm-clone's I/O path. Also, as
 * these two functions don't block, we can call them in interrupt context,
 * e.g., in a hooked overwrite bio's completion routine, and further reduce the
 * I/O completion latency.
 *
 * We maintain two dirty bitmaps. During a metadata commit we atomically swap
 * the currently used dmap with the unused one. This allows the metadata update
 * functions to run concurrently with an ongoing commit.
 */
struct dirty_map {
	unsigned long *dirty_words;
	unsigned int changed;
};

struct dm_clone_metadata {
	/* The metadata block device */
	struct block_device *bdev;

	sector_t target_size;
	sector_t region_size;
	unsigned long nr_regions;
	unsigned long nr_words;

	/* Spinlock protecting the region and dirty bitmaps. */
	spinlock_t bitmap_lock;
	struct dirty_map dmap[2];
	struct dirty_map *current_dmap;

	/*
	 * In core copy of the on-disk bitmap to save constantly doing look ups
	 * on disk.
	 */
	unsigned long *region_map;

	/* Protected by bitmap_lock */
	unsigned int read_only;

	struct dm_block_manager *bm;
	struct dm_space_map *sm;
	struct dm_transaction_manager *tm;

	struct rw_semaphore lock;

	struct dm_disk_bitset bitset_info;
	dm_block_t bitset_root;

	/*
	 * Reading the space map root can fail, so we read it into this
	 * buffer before the superblock is locked and updated.
	 */
	__u8 metadata_space_map_root[SPACE_MAP_ROOT_SIZE];

	bool hydration_done:1;
	bool fail_io:1;
};

/*---------------------------------------------------------------------------*/

/*
 * Superblock validation.
 */
static void sb_prepare_for_write(struct dm_block_validator *v,
				 struct dm_block *b, size_t sb_block_size)
{
	struct superblock_disk *sb;
	u32 csum;

	sb = dm_block_data(b);
	sb->blocknr = cpu_to_le64(dm_block_location(b));

	csum = dm_bm_checksum(&sb->flags, sb_block_size - sizeof(__le32),
			      SUPERBLOCK_CSUM_XOR);
	sb->csum = cpu_to_le32(csum);
}

static int sb_check(struct dm_block_validator *v, struct dm_block *b,
		    size_t sb_block_size)
{
	struct superblock_disk *sb;
	u32 csum, metadata_version;

	sb = dm_block_data(b);

	if (dm_block_location(b) != le64_to_cpu(sb->blocknr)) {
		DMERR("Superblock check failed: blocknr %llu, expected %llu",
		      le64_to_cpu(sb->blocknr),
		      (unsigned long long)dm_block_location(b));
		return -ENOTBLK;
	}

	if (le64_to_cpu(sb->magic) != SUPERBLOCK_MAGIC) {
		DMERR("Superblock check failed: magic %llu, expected %llu",
		      le64_to_cpu(sb->magic),
		      (unsigned long long)SUPERBLOCK_MAGIC);
		return -EILSEQ;
	}

	csum = dm_bm_checksum(&sb->flags, sb_block_size - sizeof(__le32),
			      SUPERBLOCK_CSUM_XOR);
	if (sb->csum != cpu_to_le32(csum)) {
		DMERR("Superblock check failed: checksum %u, expected %u",
		      csum, le32_to_cpu(sb->csum));
		return -EILSEQ;
	}

	/* Check metadata version */
	metadata_version = le32_to_cpu(sb->version);
	if (metadata_version < DM_CLONE_MIN_METADATA_VERSION ||
	    metadata_version > DM_CLONE_MAX_METADATA_VERSION) {
		DMERR("Clone metadata version %u found, but only versions between %u and %u supported.",
		      metadata_version, DM_CLONE_MIN_METADATA_VERSION,
		      DM_CLONE_MAX_METADATA_VERSION);
		return -EINVAL;
	}

	return 0;
}

static struct dm_block_validator sb_validator = {
	.name = "superblock",
	.prepare_for_write = sb_prepare_for_write,
	.check = sb_check
};

/*
 * Check if the superblock is formatted or not. We consider the superblock to
 * be formatted in case we find non-zero bytes in it.
 */
static int __superblock_all_zeroes(struct dm_block_manager *bm, bool *formatted)
{
	int r;
	unsigned int i, nr_words;
	struct dm_block *sblock;
	__le64 *data_le, zero = cpu_to_le64(0);

	/*
	 * We don't use a validator here because the superblock could be all
	 * zeroes.
	 */
	r = dm_bm_read_lock(bm, SUPERBLOCK_LOCATION, NULL, &sblock);
	if (r) {
		DMERR("Failed to read_lock superblock");
		return r;
	}

	data_le = dm_block_data(sblock);
	*formatted = false;

	/* This assumes that the block size is a multiple of 8 bytes */
	BUG_ON(dm_bm_block_size(bm) % sizeof(__le64));
	nr_words = dm_bm_block_size(bm) / sizeof(__le64);
	for (i = 0; i < nr_words; i++) {
		if (data_le[i] != zero) {
			*formatted = true;
			break;
		}
	}

	dm_bm_unlock(sblock);

	return 0;
}

/*---------------------------------------------------------------------------*/

/*
 * Low-level metadata handling.
 */
static inline int superblock_read_lock(struct dm_clone_metadata *cmd,
				       struct dm_block **sblock)
{
	return dm_bm_read_lock(cmd->bm, SUPERBLOCK_LOCATION, &sb_validator, sblock);
}

static inline int superblock_write_lock(struct dm_clone_metadata *cmd,
					struct dm_block **sblock)
{
	return dm_bm_write_lock(cmd->bm, SUPERBLOCK_LOCATION, &sb_validator, sblock);
}

static inline int superblock_write_lock_zero(struct dm_clone_metadata *cmd,
					     struct dm_block **sblock)
{
	return dm_bm_write_lock_zero(cmd->bm, SUPERBLOCK_LOCATION, &sb_validator, sblock);
}

static int __copy_sm_root(struct dm_clone_metadata *cmd)
{
	int r;
	size_t root_size;

	r = dm_sm_root_size(cmd->sm, &root_size);
	if (r)
		return r;

	return dm_sm_copy_root(cmd->sm, &cmd->metadata_space_map_root, root_size);
}

/* Save dm-clone metadata in superblock */
static void __prepare_superblock(struct dm_clone_metadata *cmd,
				 struct superblock_disk *sb)
{
	sb->flags = cpu_to_le32(0UL);

	/* FIXME: UUID is currently unused */
	memset(sb->uuid, 0, sizeof(sb->uuid));

	sb->magic = cpu_to_le64(SUPERBLOCK_MAGIC);
	sb->version = cpu_to_le32(DM_CLONE_MAX_METADATA_VERSION);

	/* Save the metadata space_map root */
	memcpy(&sb->metadata_space_map_root, &cmd->metadata_space_map_root,
	       sizeof(cmd->metadata_space_map_root));

	sb->region_size = cpu_to_le64(cmd->region_size);
	sb->target_size = cpu_to_le64(cmd->target_size);
	sb->bitset_root = cpu_to_le64(cmd->bitset_root);
}

static int __open_metadata(struct dm_clone_metadata *cmd)
{
	int r;
	struct dm_block *sblock;
	struct superblock_disk *sb;

	r = superblock_read_lock(cmd, &sblock);

	if (r) {
		DMERR("Failed to read_lock superblock");
		return r;
	}

	sb = dm_block_data(sblock);

	/* Verify that target_size and region_size haven't changed. */
	if (cmd->region_size != le64_to_cpu(sb->region_size) ||
	    cmd->target_size != le64_to_cpu(sb->target_size)) {
		DMERR("Region and/or target size don't match the ones in metadata");
		r = -EINVAL;
		goto out_with_lock;
	}

	r = dm_tm_open_with_sm(cmd->bm, SUPERBLOCK_LOCATION,
			       sb->metadata_space_map_root,
			       sizeof(sb->metadata_space_map_root),
			       &cmd->tm, &cmd->sm);

	if (r) {
		DMERR("dm_tm_open_with_sm failed");
		goto out_with_lock;
	}

	dm_disk_bitset_init(cmd->tm, &cmd->bitset_info);
	cmd->bitset_root = le64_to_cpu(sb->bitset_root);

out_with_lock:
	dm_bm_unlock(sblock);

	return r;
}

static int __format_metadata(struct dm_clone_metadata *cmd)
{
	int r;
	struct dm_block *sblock;
	struct superblock_disk *sb;

	r = dm_tm_create_with_sm(cmd->bm, SUPERBLOCK_LOCATION, &cmd->tm, &cmd->sm);
	if (r) {
		DMERR("Failed to create transaction manager");
		return r;
	}

	dm_disk_bitset_init(cmd->tm, &cmd->bitset_info);

	r = dm_bitset_empty(&cmd->bitset_info, &cmd->bitset_root);
	if (r) {
		DMERR("Failed to create empty on-disk bitset");
		goto err_with_tm;
	}

	r = dm_bitset_resize(&cmd->bitset_info, cmd->bitset_root, 0,
			     cmd->nr_regions, false, &cmd->bitset_root);
	if (r) {
		DMERR("Failed to resize on-disk bitset to %lu entries", cmd->nr_regions);
		goto err_with_tm;
	}

	/* Flush to disk all blocks, except the superblock */
	r = dm_tm_pre_commit(cmd->tm);
	if (r) {
		DMERR("dm_tm_pre_commit failed");
		goto err_with_tm;
	}

	r = __copy_sm_root(cmd);
	if (r) {
		DMERR("__copy_sm_root failed");
		goto err_with_tm;
	}

	r = superblock_write_lock_zero(cmd, &sblock);
	if (r) {
		DMERR("Failed to write_lock superblock");
		goto err_with_tm;
	}

	sb = dm_block_data(sblock);
	__prepare_superblock(cmd, sb);
	r = dm_tm_commit(cmd->tm, sblock);
	if (r) {
		DMERR("Failed to commit superblock");
		goto err_with_tm;
	}

	return 0;

err_with_tm:
	dm_sm_destroy(cmd->sm);
	dm_tm_destroy(cmd->tm);

	return r;
}

static int __open_or_format_metadata(struct dm_clone_metadata *cmd, bool may_format_device)
{
	int r;
	bool formatted = false;

	r = __superblock_all_zeroes(cmd->bm, &formatted);
	if (r)
		return r;

	if (!formatted)
		return may_format_device ? __format_metadata(cmd) : -EPERM;

	return __open_metadata(cmd);
}

static int __create_persistent_data_structures(struct dm_clone_metadata *cmd,
					       bool may_format_device)
{
	int r;

	/* Create block manager */
	cmd->bm = dm_block_manager_create(cmd->bdev,
					 DM_CLONE_METADATA_BLOCK_SIZE << SECTOR_SHIFT,
					 DM_CLONE_MAX_CONCURRENT_LOCKS);
	if (IS_ERR(cmd->bm)) {
		DMERR("Failed to create block manager");
		return PTR_ERR(cmd->bm);
	}

	r = __open_or_format_metadata(cmd, may_format_device);
	if (r)
		dm_block_manager_destroy(cmd->bm);

	return r;
}

static void __destroy_persistent_data_structures(struct dm_clone_metadata *cmd)
{
	dm_sm_destroy(cmd->sm);
	dm_tm_destroy(cmd->tm);
	dm_block_manager_destroy(cmd->bm);
}

/*---------------------------------------------------------------------------*/

static size_t bitmap_size(unsigned long nr_bits)
{
	return BITS_TO_LONGS(nr_bits) * sizeof(long);
}

static int dirty_map_init(struct dm_clone_metadata *cmd)
{
	cmd->dmap[0].changed = 0;
	cmd->dmap[0].dirty_words = kvzalloc(bitmap_size(cmd->nr_words), GFP_KERNEL);

	if (!cmd->dmap[0].dirty_words) {
		DMERR("Failed to allocate dirty bitmap");
		return -ENOMEM;
	}

	cmd->dmap[1].changed = 0;
	cmd->dmap[1].dirty_words = kvzalloc(bitmap_size(cmd->nr_words), GFP_KERNEL);

	if (!cmd->dmap[1].dirty_words) {
		DMERR("Failed to allocate dirty bitmap");
		kvfree(cmd->dmap[0].dirty_words);
		return -ENOMEM;
	}

	cmd->current_dmap = &cmd->dmap[0];

	return 0;
}

static void dirty_map_exit(struct dm_clone_metadata *cmd)
{
	kvfree(cmd->dmap[0].dirty_words);
	kvfree(cmd->dmap[1].dirty_words);
}

static int __load_bitset_in_core(struct dm_clone_metadata *cmd)
{
	int r;
	unsigned long i;
	struct dm_bitset_cursor c;

	/* Flush bitset cache */
	r = dm_bitset_flush(&cmd->bitset_info, cmd->bitset_root, &cmd->bitset_root);
	if (r)
		return r;

	r = dm_bitset_cursor_begin(&cmd->bitset_info, cmd->bitset_root, cmd->nr_regions, &c);
	if (r)
		return r;

	for (i = 0; ; i++) {
		if (dm_bitset_cursor_get_value(&c))
			__set_bit(i, cmd->region_map);
		else
			__clear_bit(i, cmd->region_map);

		if (i >= (cmd->nr_regions - 1))
			break;

		r = dm_bitset_cursor_next(&c);

		if (r)
			break;
	}

	dm_bitset_cursor_end(&c);

	return r;
}

struct dm_clone_metadata *dm_clone_metadata_open(struct block_device *bdev,
						 sector_t target_size,
						 sector_t region_size)
{
	int r;
	struct dm_clone_metadata *cmd;

	cmd = kzalloc(sizeof(*cmd), GFP_KERNEL);
	if (!cmd) {
		DMERR("Failed to allocate memory for dm-clone metadata");
		return ERR_PTR(-ENOMEM);
	}

	cmd->bdev = bdev;
	cmd->target_size = target_size;
	cmd->region_size = region_size;
	cmd->nr_regions = dm_sector_div_up(cmd->target_size, cmd->region_size);
	cmd->nr_words = BITS_TO_LONGS(cmd->nr_regions);

	init_rwsem(&cmd->lock);
	spin_lock_init(&cmd->bitmap_lock);
	cmd->read_only = 0;
	cmd->fail_io = false;
	cmd->hydration_done = false;

	cmd->region_map = kvmalloc(bitmap_size(cmd->nr_regions), GFP_KERNEL);
	if (!cmd->region_map) {
		DMERR("Failed to allocate memory for region bitmap");
		r = -ENOMEM;
		goto out_with_md;
	}

	r = __create_persistent_data_structures(cmd, true);
	if (r)
		goto out_with_region_map;

	r = __load_bitset_in_core(cmd);
	if (r) {
		DMERR("Failed to load on-disk region map");
		goto out_with_pds;
	}

	r = dirty_map_init(cmd);
	if (r)
		goto out_with_pds;

	if (bitmap_full(cmd->region_map, cmd->nr_regions))
		cmd->hydration_done = true;

	return cmd;

out_with_pds:
	__destroy_persistent_data_structures(cmd);

out_with_region_map:
	kvfree(cmd->region_map);

out_with_md:
	kfree(cmd);

	return ERR_PTR(r);
}

void dm_clone_metadata_close(struct dm_clone_metadata *cmd)
{
	if (!cmd->fail_io)
		__destroy_persistent_data_structures(cmd);

	dirty_map_exit(cmd);
	kvfree(cmd->region_map);
	kfree(cmd);
}

bool dm_clone_is_hydration_done(struct dm_clone_metadata *cmd)
{
	return cmd->hydration_done;
}

bool dm_clone_is_region_hydrated(struct dm_clone_metadata *cmd, unsigned long region_nr)
{
	return dm_clone_is_hydration_done(cmd) || test_bit(region_nr, cmd->region_map);
}

bool dm_clone_is_range_hydrated(struct dm_clone_metadata *cmd,
				unsigned long start, unsigned long nr_regions)
{
	unsigned long bit;

	if (dm_clone_is_hydration_done(cmd))
		return true;

	bit = find_next_zero_bit(cmd->region_map, cmd->nr_regions, start);

	return (bit >= (start + nr_regions));
}

unsigned long dm_clone_nr_of_hydrated_regions(struct dm_clone_metadata *cmd)
{
	return bitmap_weight(cmd->region_map, cmd->nr_regions);
}

unsigned long dm_clone_find_next_unhydrated_region(struct dm_clone_metadata *cmd,
						   unsigned long start)
{
	return find_next_zero_bit(cmd->region_map, cmd->nr_regions, start);
}

static int __update_metadata_word(struct dm_clone_metadata *cmd, unsigned long word)
{
	int r;
	unsigned long index = word * BITS_PER_LONG;
	unsigned long max_index = min(cmd->nr_regions, (word + 1) * BITS_PER_LONG);

	while (index < max_index) {
		if (test_bit(index, cmd->region_map)) {
			r = dm_bitset_set_bit(&cmd->bitset_info, cmd->bitset_root,
					      index, &cmd->bitset_root);

			if (r) {
				DMERR("dm_bitset_set_bit failed");
				return r;
			}
		}
		index++;
	}

	return 0;
}

static int __metadata_commit(struct dm_clone_metadata *cmd)
{
	int r;
	struct dm_block *sblock;
	struct superblock_disk *sb;

	/* Flush bitset cache */
	r = dm_bitset_flush(&cmd->bitset_info, cmd->bitset_root, &cmd->bitset_root);
	if (r) {
		DMERR("dm_bitset_flush failed");
		return r;
	}

	/* Flush to disk all blocks, except the superblock */
	r = dm_tm_pre_commit(cmd->tm);
	if (r) {
		DMERR("dm_tm_pre_commit failed");
		return r;
	}

	/* Save the space map root in cmd->metadata_space_map_root */
	r = __copy_sm_root(cmd);
	if (r) {
		DMERR("__copy_sm_root failed");
		return r;
	}

	/* Lock the superblock */
	r = superblock_write_lock_zero(cmd, &sblock);
	if (r) {
		DMERR("Failed to write_lock superblock");
		return r;
	}

	/* Save the metadata in superblock */
	sb = dm_block_data(sblock);
	__prepare_superblock(cmd, sb);

	/* Unlock superblock and commit it to disk */
	r = dm_tm_commit(cmd->tm, sblock);
	if (r) {
		DMERR("Failed to commit superblock");
		return r;
	}

	/*
	 * FIXME: Find a more efficient way to check if the hydration is done.
	 */
	if (bitmap_full(cmd->region_map, cmd->nr_regions))
		cmd->hydration_done = true;

	return 0;
}

static int __flush_dmap(struct dm_clone_metadata *cmd, struct dirty_map *dmap)
{
	int r;
	unsigned long word, flags;

	word = 0;
	do {
		word = find_next_bit(dmap->dirty_words, cmd->nr_words, word);

		if (word == cmd->nr_words)
			break;

		r = __update_metadata_word(cmd, word);

		if (r)
			return r;

		__clear_bit(word, dmap->dirty_words);
		word++;
	} while (word < cmd->nr_words);

	r = __metadata_commit(cmd);

	if (r)
		return r;

	/* Update the changed flag */
	spin_lock_irqsave(&cmd->bitmap_lock, flags);
	dmap->changed = 0;
	spin_unlock_irqrestore(&cmd->bitmap_lock, flags);

	return 0;
}

int dm_clone_metadata_commit(struct dm_clone_metadata *cmd)
{
	int r = -EPERM;
	unsigned long flags;
	struct dirty_map *dmap, *next_dmap;

	down_write(&cmd->lock);

	if (cmd->fail_io || dm_bm_is_read_only(cmd->bm))
		goto out;

	/* Get current dirty bitmap */
	dmap = cmd->current_dmap;

	/* Get next dirty bitmap */
	next_dmap = (dmap == &cmd->dmap[0]) ? &cmd->dmap[1] : &cmd->dmap[0];

	/*
	 * The last commit failed, so we don't have a clean dirty-bitmap to
	 * use.
	 */
	if (WARN_ON(next_dmap->changed)) {
		r = -EINVAL;
		goto out;
	}

	/* Swap dirty bitmaps */
	spin_lock_irqsave(&cmd->bitmap_lock, flags);
	cmd->current_dmap = next_dmap;
	spin_unlock_irqrestore(&cmd->bitmap_lock, flags);

	/*
	 * No one is accessing the old dirty bitmap anymore, so we can flush
	 * it.
	 */
	r = __flush_dmap(cmd, dmap);
out:
	up_write(&cmd->lock);

	return r;
}

int dm_clone_set_region_hydrated(struct dm_clone_metadata *cmd, unsigned long region_nr)
{
	int r = 0;
	struct dirty_map *dmap;
	unsigned long word, flags;

	word = region_nr / BITS_PER_LONG;

	spin_lock_irqsave(&cmd->bitmap_lock, flags);

	if (cmd->read_only) {
		r = -EPERM;
		goto out;
	}

	dmap = cmd->current_dmap;

	__set_bit(word, dmap->dirty_words);
	__set_bit(region_nr, cmd->region_map);
	dmap->changed = 1;

out:
	spin_unlock_irqrestore(&cmd->bitmap_lock, flags);

	return r;
}

int dm_clone_cond_set_range(struct dm_clone_metadata *cmd, unsigned long start,
			    unsigned long nr_regions)
{
	int r = 0;
	struct dirty_map *dmap;
	unsigned long word, region_nr, flags;

	spin_lock_irqsave(&cmd->bitmap_lock, flags);

	if (cmd->read_only) {
		r = -EPERM;
		goto out;
	}

	dmap = cmd->current_dmap;
	for (region_nr = start; region_nr < (start + nr_regions); region_nr++) {
		if (!test_bit(region_nr, cmd->region_map)) {
			word = region_nr / BITS_PER_LONG;
			__set_bit(word, dmap->dirty_words);
			__set_bit(region_nr, cmd->region_map);
			dmap->changed = 1;
		}
	}
out:
	spin_unlock_irqrestore(&cmd->bitmap_lock, flags);

	return r;
}

/*
 * WARNING: This must not be called concurrently with either
 * dm_clone_set_region_hydrated() or dm_clone_cond_set_range(), as it changes
 * cmd->region_map without taking the cmd->bitmap_lock spinlock. The only
 * exception is after setting the metadata to read-only mode, using
 * dm_clone_metadata_set_read_only().
 *
 * We don't take the spinlock because __load_bitset_in_core() does I/O, so it
 * may block.
 */
int dm_clone_reload_in_core_bitset(struct dm_clone_metadata *cmd)
{
	int r = -EINVAL;

	down_write(&cmd->lock);

	if (cmd->fail_io)
		goto out;

	r = __load_bitset_in_core(cmd);
out:
	up_write(&cmd->lock);

	return r;
}

bool dm_clone_changed_this_transaction(struct dm_clone_metadata *cmd)
{
	bool r;
	unsigned long flags;

	spin_lock_irqsave(&cmd->bitmap_lock, flags);
	r = cmd->dmap[0].changed || cmd->dmap[1].changed;
	spin_unlock_irqrestore(&cmd->bitmap_lock, flags);

	return r;
}

int dm_clone_metadata_abort(struct dm_clone_metadata *cmd)
{
	int r = -EPERM;

	down_write(&cmd->lock);

	if (cmd->fail_io || dm_bm_is_read_only(cmd->bm))
		goto out;

	__destroy_persistent_data_structures(cmd);

	r = __create_persistent_data_structures(cmd, false);
	if (r) {
		/* If something went wrong we can neither write nor read the metadata */
		cmd->fail_io = true;
	}
out:
	up_write(&cmd->lock);

	return r;
}

void dm_clone_metadata_set_read_only(struct dm_clone_metadata *cmd)
{
	unsigned long flags;

	down_write(&cmd->lock);

	spin_lock_irqsave(&cmd->bitmap_lock, flags);
	cmd->read_only = 1;
	spin_unlock_irqrestore(&cmd->bitmap_lock, flags);

	if (!cmd->fail_io)
		dm_bm_set_read_only(cmd->bm);

	up_write(&cmd->lock);
}

void dm_clone_metadata_set_read_write(struct dm_clone_metadata *cmd)
{
	unsigned long flags;

	down_write(&cmd->lock);

	spin_lock_irqsave(&cmd->bitmap_lock, flags);
	cmd->read_only = 0;
	spin_unlock_irqrestore(&cmd->bitmap_lock, flags);

	if (!cmd->fail_io)
		dm_bm_set_read_write(cmd->bm);

	up_write(&cmd->lock);
}

int dm_clone_get_free_metadata_block_count(struct dm_clone_metadata *cmd,
					   dm_block_t *result)
{
	int r = -EINVAL;

	down_read(&cmd->lock);

	if (!cmd->fail_io)
		r = dm_sm_get_nr_free(cmd->sm, result);

	up_read(&cmd->lock);

	return r;
}

int dm_clone_get_metadata_dev_size(struct dm_clone_metadata *cmd,
				   dm_block_t *result)
{
	int r = -EINVAL;

	down_read(&cmd->lock);

	if (!cmd->fail_io)
		r = dm_sm_get_nr_blocks(cmd->sm, result);

	up_read(&cmd->lock);

	return r;
}