summaryrefslogtreecommitdiff
path: root/fs/bcachefs/replicas.c
blob: a7a4e280840b90c2ce755f34758b615b0c64cfe0 (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

#include "bcachefs.h"
#include "replicas.h"
#include "super-io.h"

struct bch_replicas_entry_padded {
	struct bch_replicas_entry	e;
	u8				pad[BCH_SB_MEMBERS_MAX];
};

static int bch2_cpu_replicas_to_sb_replicas(struct bch_fs *,
					    struct bch_replicas_cpu *);

/* Replicas tracking - in memory: */

static inline int u8_cmp(u8 l, u8 r)
{
	return (l > r) - (l < r);
}

static void replicas_entry_sort(struct bch_replicas_entry *e)
{
	bubble_sort(e->devs, e->nr_devs, u8_cmp);
}

#define for_each_cpu_replicas_entry(_r, _i)				\
	for (_i = (_r)->entries;					\
	     (void *) (_i) < (void *) (_r)->entries + (_r)->nr * (_r)->entry_size;\
	     _i = (void *) (_i) + (_r)->entry_size)

static inline struct bch_replicas_entry *
cpu_replicas_entry(struct bch_replicas_cpu *r, unsigned i)
{
	return (void *) r->entries + r->entry_size * i;
}

static void bch2_cpu_replicas_sort(struct bch_replicas_cpu *r)
{
	eytzinger0_sort(r->entries, r->nr, r->entry_size, memcmp, NULL);
}

static void replicas_entry_to_text(struct printbuf *out,
				  struct bch_replicas_entry *e)
{
	unsigned i;

	pr_buf(out, "%s: %u/%u [",
	       bch2_data_types[e->data_type],
	       e->nr_required,
	       e->nr_devs);

	for (i = 0; i < e->nr_devs; i++)
		pr_buf(out, i ? " %u" : "%u", e->devs[i]);
	pr_buf(out, "]");
}

void bch2_cpu_replicas_to_text(struct printbuf *out,
			      struct bch_replicas_cpu *r)
{
	struct bch_replicas_entry *e;
	bool first = true;

	for_each_cpu_replicas_entry(r, e) {
		if (!first)
			pr_buf(out, " ");
		first = false;

		replicas_entry_to_text(out, e);
	}
}

static void extent_to_replicas(struct bkey_s_c k,
			       struct bch_replicas_entry *r)
{
	if (bkey_extent_is_data(k.k)) {
		struct bkey_s_c_extent e = bkey_s_c_to_extent(k);
		const union bch_extent_entry *entry;
		struct extent_ptr_decoded p;

		r->nr_required	= 1;

		extent_for_each_ptr_decode(e, p, entry)
			if (!p.ptr.cached)
				r->devs[r->nr_devs++] = p.ptr.dev;
	}
}

static void bkey_to_replicas(enum bkey_type type,
			     struct bkey_s_c k,
			     struct bch_replicas_entry *e)
{
	e->nr_devs = 0;

	switch (type) {
	case BKEY_TYPE_BTREE:
		e->data_type = BCH_DATA_BTREE;
		extent_to_replicas(k, e);
		break;
	case BKEY_TYPE_EXTENTS:
		e->data_type = BCH_DATA_USER;
		extent_to_replicas(k, e);
		break;
	default:
		break;
	}

	replicas_entry_sort(e);
}

static inline void devlist_to_replicas(struct bch_devs_list devs,
				       enum bch_data_type data_type,
				       struct bch_replicas_entry *e)
{
	unsigned i;

	BUG_ON(!data_type ||
	       data_type == BCH_DATA_SB ||
	       data_type >= BCH_DATA_NR);

	e->data_type	= data_type;
	e->nr_devs	= 0;
	e->nr_required	= 1;

	for (i = 0; i < devs.nr; i++)
		e->devs[e->nr_devs++] = devs.devs[i];

	replicas_entry_sort(e);
}

static struct bch_replicas_cpu *
cpu_replicas_add_entry(struct bch_replicas_cpu *old,
		       struct bch_replicas_entry *new_entry)
{
	struct bch_replicas_cpu *new;
	unsigned i, nr, entry_size;

	entry_size = max_t(unsigned, old->entry_size,
			   replicas_entry_bytes(new_entry));
	nr = old->nr + 1;

	new = kzalloc(sizeof(struct bch_replicas_cpu) +
		      nr * entry_size, GFP_NOIO);
	if (!new)
		return NULL;

	new->nr		= nr;
	new->entry_size	= entry_size;

	for (i = 0; i < old->nr; i++)
		memcpy(cpu_replicas_entry(new, i),
		       cpu_replicas_entry(old, i),
		       old->entry_size);

	memcpy(cpu_replicas_entry(new, old->nr),
	       new_entry,
	       replicas_entry_bytes(new_entry));

	bch2_cpu_replicas_sort(new);
	return new;
}

static bool __replicas_has_entry(struct bch_replicas_cpu *r,
				 struct bch_replicas_entry *search)
{
	return replicas_entry_bytes(search) <= r->entry_size &&
		eytzinger0_find(r->entries, r->nr,
				r->entry_size,
				memcmp, search) < r->nr;
}

static bool replicas_has_entry(struct bch_fs *c,
			       struct bch_replicas_entry *search,
			       bool check_gc_replicas)
{
	struct bch_replicas_cpu *r, *gc_r;
	bool marked;

	rcu_read_lock();
	r = rcu_dereference(c->replicas);
	marked = __replicas_has_entry(r, search) &&
		(!check_gc_replicas ||
		 likely(!(gc_r = rcu_dereference(c->replicas_gc))) ||
		 __replicas_has_entry(gc_r, search));
	rcu_read_unlock();

	return marked;
}

noinline
static int bch2_mark_replicas_slowpath(struct bch_fs *c,
				struct bch_replicas_entry *new_entry)
{
	struct bch_replicas_cpu *old_gc, *new_gc = NULL, *old_r, *new_r = NULL;
	int ret = -ENOMEM;

	mutex_lock(&c->sb_lock);

	old_gc = rcu_dereference_protected(c->replicas_gc,
					   lockdep_is_held(&c->sb_lock));
	if (old_gc && !__replicas_has_entry(old_gc, new_entry)) {
		new_gc = cpu_replicas_add_entry(old_gc, new_entry);
		if (!new_gc)
			goto err;
	}

	old_r = rcu_dereference_protected(c->replicas,
					  lockdep_is_held(&c->sb_lock));
	if (!__replicas_has_entry(old_r, new_entry)) {
		new_r = cpu_replicas_add_entry(old_r, new_entry);
		if (!new_r)
			goto err;

		ret = bch2_cpu_replicas_to_sb_replicas(c, new_r);
		if (ret)
			goto err;
	}

	/* allocations done, now commit: */

	if (new_r)
		bch2_write_super(c);

	/* don't update in memory replicas until changes are persistent */

	if (new_gc) {
		rcu_assign_pointer(c->replicas_gc, new_gc);
		kfree_rcu(old_gc, rcu);
	}

	if (new_r) {
		rcu_assign_pointer(c->replicas, new_r);
		kfree_rcu(old_r, rcu);
	}

	mutex_unlock(&c->sb_lock);
	return 0;
err:
	mutex_unlock(&c->sb_lock);
	kfree(new_gc);
	kfree(new_r);
	return ret;
}

static int __bch2_mark_replicas(struct bch_fs *c,
				struct bch_replicas_entry *devs)
{
	return likely(replicas_has_entry(c, devs, true))
		? 0
		: bch2_mark_replicas_slowpath(c, devs);
}

int bch2_mark_replicas(struct bch_fs *c,
		       enum bch_data_type data_type,
		       struct bch_devs_list devs)
{
	struct bch_replicas_entry_padded search;

	if (!devs.nr)
		return 0;

	memset(&search, 0, sizeof(search));

	BUG_ON(devs.nr >= BCH_REPLICAS_MAX);

	devlist_to_replicas(devs, data_type, &search.e);

	return __bch2_mark_replicas(c, &search.e);
}

int bch2_mark_bkey_replicas(struct bch_fs *c,
			    enum bkey_type type,
			    struct bkey_s_c k)
{
	struct bch_replicas_entry_padded search;
	int ret;

	memset(&search, 0, sizeof(search));

	if (type == BKEY_TYPE_EXTENTS) {
		struct bch_devs_list cached = bch2_bkey_cached_devs(k);
		unsigned i;

		for (i = 0; i < cached.nr; i++)
			if ((ret = bch2_mark_replicas(c, BCH_DATA_CACHED,
						bch2_dev_list_single(cached.devs[i]))))
				return ret;
	}

	bkey_to_replicas(type, k, &search.e);

	return search.e.nr_devs
		? __bch2_mark_replicas(c, &search.e)
		: 0;
}

int bch2_replicas_gc_end(struct bch_fs *c, int ret)
{
	struct bch_replicas_cpu *new_r, *old_r;

	lockdep_assert_held(&c->replicas_gc_lock);

	mutex_lock(&c->sb_lock);

	new_r = rcu_dereference_protected(c->replicas_gc,
					  lockdep_is_held(&c->sb_lock));
	rcu_assign_pointer(c->replicas_gc, NULL);

	if (ret)
		goto err;

	if (bch2_cpu_replicas_to_sb_replicas(c, new_r)) {
		ret = -ENOSPC;
		goto err;
	}

	bch2_write_super(c);

	/* don't update in memory replicas until changes are persistent */

	old_r = rcu_dereference_protected(c->replicas,
					  lockdep_is_held(&c->sb_lock));

	rcu_assign_pointer(c->replicas, new_r);
	kfree_rcu(old_r, rcu);
out:
	mutex_unlock(&c->sb_lock);
	return ret;
err:
	kfree_rcu(new_r, rcu);
	goto out;
}

int bch2_replicas_gc_start(struct bch_fs *c, unsigned typemask)
{
	struct bch_replicas_cpu *dst, *src;
	struct bch_replicas_entry *e;

	lockdep_assert_held(&c->replicas_gc_lock);

	mutex_lock(&c->sb_lock);
	BUG_ON(c->replicas_gc);

	src = rcu_dereference_protected(c->replicas,
					lockdep_is_held(&c->sb_lock));

	dst = kzalloc(sizeof(struct bch_replicas_cpu) +
		      src->nr * src->entry_size, GFP_NOIO);
	if (!dst) {
		mutex_unlock(&c->sb_lock);
		return -ENOMEM;
	}

	dst->nr		= 0;
	dst->entry_size	= src->entry_size;

	for_each_cpu_replicas_entry(src, e)
		if (!((1 << e->data_type) & typemask))
			memcpy(cpu_replicas_entry(dst, dst->nr++),
			       e, dst->entry_size);

	bch2_cpu_replicas_sort(dst);

	rcu_assign_pointer(c->replicas_gc, dst);
	mutex_unlock(&c->sb_lock);

	return 0;
}

/* Replicas tracking - superblock: */

static struct bch_replicas_cpu *
__bch2_sb_replicas_to_cpu_replicas(struct bch_sb_field_replicas *sb_r)
{
	struct bch_replicas_entry *e, *dst;
	struct bch_replicas_cpu *cpu_r;
	unsigned nr = 0, entry_size = 0, idx = 0;

	for_each_replicas_entry(sb_r, e) {
		entry_size = max_t(unsigned, entry_size,
				   replicas_entry_bytes(e));
		nr++;
	}

	cpu_r = kzalloc(sizeof(struct bch_replicas_cpu) +
			nr * entry_size, GFP_NOIO);
	if (!cpu_r)
		return NULL;

	cpu_r->nr		= nr;
	cpu_r->entry_size	= entry_size;

	for_each_replicas_entry(sb_r, e) {
		dst = cpu_replicas_entry(cpu_r, idx++);
		memcpy(dst, e, replicas_entry_bytes(e));
		replicas_entry_sort(dst);
	}

	return cpu_r;
}

static struct bch_replicas_cpu *
__bch2_sb_replicas_v0_to_cpu_replicas(struct bch_sb_field_replicas_v0 *sb_r)
{
	struct bch_replicas_entry_v0 *e;
	struct bch_replicas_cpu *cpu_r;
	unsigned nr = 0, entry_size = 0, idx = 0;

	for_each_replicas_entry(sb_r, e) {
		entry_size = max_t(unsigned, entry_size,
				   replicas_entry_bytes(e));
		nr++;
	}

	entry_size += sizeof(struct bch_replicas_entry) -
		sizeof(struct bch_replicas_entry_v0);

	cpu_r = kzalloc(sizeof(struct bch_replicas_cpu) +
			nr * entry_size, GFP_NOIO);
	if (!cpu_r)
		return NULL;

	cpu_r->nr		= nr;
	cpu_r->entry_size	= entry_size;

	for_each_replicas_entry(sb_r, e) {
		struct bch_replicas_entry *dst =
			cpu_replicas_entry(cpu_r, idx++);

		dst->data_type	= e->data_type;
		dst->nr_devs	= e->nr_devs;
		dst->nr_required = 1;
		memcpy(dst->devs, e->devs, e->nr_devs);
		replicas_entry_sort(dst);
	}

	return cpu_r;
}

int bch2_sb_replicas_to_cpu_replicas(struct bch_fs *c)
{
	struct bch_sb_field_replicas *sb_v1;
	struct bch_sb_field_replicas_v0 *sb_v0;
	struct bch_replicas_cpu *cpu_r, *old_r;

	if ((sb_v1 = bch2_sb_get_replicas(c->disk_sb.sb)))
		cpu_r = __bch2_sb_replicas_to_cpu_replicas(sb_v1);
	else if ((sb_v0 = bch2_sb_get_replicas_v0(c->disk_sb.sb)))
		cpu_r = __bch2_sb_replicas_v0_to_cpu_replicas(sb_v0);
	else
		cpu_r = kzalloc(sizeof(struct bch_replicas_cpu), GFP_NOIO);

	if (!cpu_r)
		return -ENOMEM;

	bch2_cpu_replicas_sort(cpu_r);

	old_r = rcu_dereference_check(c->replicas, lockdep_is_held(&c->sb_lock));
	rcu_assign_pointer(c->replicas, cpu_r);
	if (old_r)
		kfree_rcu(old_r, rcu);

	return 0;
}

static int bch2_cpu_replicas_to_sb_replicas_v0(struct bch_fs *c,
					       struct bch_replicas_cpu *r)
{
	struct bch_sb_field_replicas_v0 *sb_r;
	struct bch_replicas_entry_v0 *dst;
	struct bch_replicas_entry *src;
	size_t bytes;

	bytes = sizeof(struct bch_sb_field_replicas);

	for_each_cpu_replicas_entry(r, src)
		bytes += replicas_entry_bytes(src) - 1;

	sb_r = bch2_sb_resize_replicas_v0(&c->disk_sb,
			DIV_ROUND_UP(bytes, sizeof(u64)));
	if (!sb_r)
		return -ENOSPC;

	bch2_sb_field_delete(&c->disk_sb, BCH_SB_FIELD_replicas);
	sb_r = bch2_sb_get_replicas_v0(c->disk_sb.sb);

	memset(&sb_r->entries, 0,
	       vstruct_end(&sb_r->field) -
	       (void *) &sb_r->entries);

	dst = sb_r->entries;
	for_each_cpu_replicas_entry(r, src) {
		dst->data_type	= src->data_type;
		dst->nr_devs	= src->nr_devs;
		memcpy(dst->devs, src->devs, src->nr_devs);

		dst = replicas_entry_next(dst);

		BUG_ON((void *) dst > vstruct_end(&sb_r->field));
	}

	return 0;
}

static int bch2_cpu_replicas_to_sb_replicas(struct bch_fs *c,
					    struct bch_replicas_cpu *r)
{
	struct bch_sb_field_replicas *sb_r;
	struct bch_replicas_entry *dst, *src;
	bool need_v1 = false;
	size_t bytes;

	bytes = sizeof(struct bch_sb_field_replicas);

	for_each_cpu_replicas_entry(r, src) {
		bytes += replicas_entry_bytes(src);
		if (src->nr_required != 1)
			need_v1 = true;
	}

	if (!need_v1)
		return bch2_cpu_replicas_to_sb_replicas_v0(c, r);

	sb_r = bch2_sb_resize_replicas(&c->disk_sb,
			DIV_ROUND_UP(bytes, sizeof(u64)));
	if (!sb_r)
		return -ENOSPC;

	bch2_sb_field_delete(&c->disk_sb, BCH_SB_FIELD_replicas_v0);
	sb_r = bch2_sb_get_replicas(c->disk_sb.sb);

	memset(&sb_r->entries, 0,
	       vstruct_end(&sb_r->field) -
	       (void *) &sb_r->entries);

	dst = sb_r->entries;
	for_each_cpu_replicas_entry(r, src) {
		memcpy(dst, src, replicas_entry_bytes(src));

		dst = replicas_entry_next(dst);

		BUG_ON((void *) dst > vstruct_end(&sb_r->field));
	}

	return 0;
}

static const char *check_dup_replicas_entries(struct bch_replicas_cpu *cpu_r)
{
	unsigned i;

	sort_cmp_size(cpu_r->entries,
		      cpu_r->nr,
		      cpu_r->entry_size,
		      memcmp, NULL);

	for (i = 0; i + 1 < cpu_r->nr; i++) {
		struct bch_replicas_entry *l =
			cpu_replicas_entry(cpu_r, i);
		struct bch_replicas_entry *r =
			cpu_replicas_entry(cpu_r, i + 1);

		BUG_ON(memcmp(l, r, cpu_r->entry_size) > 0);

		if (!memcmp(l, r, cpu_r->entry_size))
			return "duplicate replicas entry";
	}

	return NULL;
}

static const char *bch2_sb_validate_replicas(struct bch_sb *sb, struct bch_sb_field *f)
{
	struct bch_sb_field_replicas *sb_r = field_to_type(f, replicas);
	struct bch_sb_field_members *mi = bch2_sb_get_members(sb);
	struct bch_replicas_cpu *cpu_r = NULL;
	struct bch_replicas_entry *e;
	const char *err;
	unsigned i;

	for_each_replicas_entry(sb_r, e) {
		err = "invalid replicas entry: invalid data type";
		if (e->data_type >= BCH_DATA_NR)
			goto err;

		err = "invalid replicas entry: no devices";
		if (!e->nr_devs)
			goto err;

		err = "invalid replicas entry: bad nr_required";
		if (!e->nr_required ||
		    (e->nr_required > 1 &&
		     e->nr_required >= e->nr_devs))
			goto err;

		err = "invalid replicas entry: invalid device";
		for (i = 0; i < e->nr_devs; i++)
			if (!bch2_dev_exists(sb, mi, e->devs[i]))
				goto err;
	}

	err = "cannot allocate memory";
	cpu_r = __bch2_sb_replicas_to_cpu_replicas(sb_r);
	if (!cpu_r)
		goto err;

	err = check_dup_replicas_entries(cpu_r);
err:
	kfree(cpu_r);
	return err;
}

static void bch2_sb_replicas_to_text(struct printbuf *out,
				     struct bch_sb *sb,
				     struct bch_sb_field *f)
{
	struct bch_sb_field_replicas *r = field_to_type(f, replicas);
	struct bch_replicas_entry *e;
	bool first = true;

	for_each_replicas_entry(r, e) {
		if (!first)
			pr_buf(out, " ");
		first = false;

		replicas_entry_to_text(out, e);
	}
}

const struct bch_sb_field_ops bch_sb_field_ops_replicas = {
	.validate	= bch2_sb_validate_replicas,
	.to_text	= bch2_sb_replicas_to_text,
};

static const char *bch2_sb_validate_replicas_v0(struct bch_sb *sb, struct bch_sb_field *f)
{
	struct bch_sb_field_replicas_v0 *sb_r = field_to_type(f, replicas_v0);
	struct bch_sb_field_members *mi = bch2_sb_get_members(sb);
	struct bch_replicas_cpu *cpu_r = NULL;
	struct bch_replicas_entry_v0 *e;
	const char *err;
	unsigned i;

	for_each_replicas_entry_v0(sb_r, e) {
		err = "invalid replicas entry: invalid data type";
		if (e->data_type >= BCH_DATA_NR)
			goto err;

		err = "invalid replicas entry: no devices";
		if (!e->nr_devs)
			goto err;

		err = "invalid replicas entry: invalid device";
		for (i = 0; i < e->nr_devs; i++)
			if (!bch2_dev_exists(sb, mi, e->devs[i]))
				goto err;
	}

	err = "cannot allocate memory";
	cpu_r = __bch2_sb_replicas_v0_to_cpu_replicas(sb_r);
	if (!cpu_r)
		goto err;

	err = check_dup_replicas_entries(cpu_r);
err:
	kfree(cpu_r);
	return err;
}

const struct bch_sb_field_ops bch_sb_field_ops_replicas_v0 = {
	.validate	= bch2_sb_validate_replicas_v0,
};

/* Query replicas: */

bool bch2_replicas_marked(struct bch_fs *c,
			  enum bch_data_type data_type,
			  struct bch_devs_list devs,
			  bool check_gc_replicas)
{
	struct bch_replicas_entry_padded search;

	if (!devs.nr)
		return true;

	memset(&search, 0, sizeof(search));

	devlist_to_replicas(devs, data_type, &search.e);

	return replicas_has_entry(c, &search.e, check_gc_replicas);
}

bool bch2_bkey_replicas_marked(struct bch_fs *c,
			       enum bkey_type type,
			       struct bkey_s_c k,
			       bool check_gc_replicas)
{
	struct bch_replicas_entry_padded search;

	memset(&search, 0, sizeof(search));

	if (type == BKEY_TYPE_EXTENTS) {
		struct bch_devs_list cached = bch2_bkey_cached_devs(k);
		unsigned i;

		for (i = 0; i < cached.nr; i++)
			if (!bch2_replicas_marked(c, BCH_DATA_CACHED,
					bch2_dev_list_single(cached.devs[i]),
					check_gc_replicas))
				return false;
	}

	bkey_to_replicas(type, k, &search.e);

	return search.e.nr_devs
		? replicas_has_entry(c, &search.e, check_gc_replicas)
		: true;
}

struct replicas_status __bch2_replicas_status(struct bch_fs *c,
					      struct bch_devs_mask online_devs)
{
	struct bch_sb_field_members *mi;
	struct bch_replicas_entry *e;
	struct bch_replicas_cpu *r;
	unsigned i, nr_online, nr_offline;
	struct replicas_status ret;

	memset(&ret, 0, sizeof(ret));

	for (i = 0; i < ARRAY_SIZE(ret.replicas); i++)
		ret.replicas[i].redundancy = INT_MAX;

	mi = bch2_sb_get_members(c->disk_sb.sb);
	rcu_read_lock();
	r = rcu_dereference(c->replicas);

	for_each_cpu_replicas_entry(r, e) {
		if (e->data_type >= ARRAY_SIZE(ret.replicas))
			panic("e %p data_type %u\n", e, e->data_type);

		nr_online = nr_offline = 0;

		for (i = 0; i < e->nr_devs; i++) {
			BUG_ON(!bch2_dev_exists(c->disk_sb.sb, mi,
						e->devs[i]));

			if (test_bit(e->devs[i], online_devs.d))
				nr_online++;
			else
				nr_offline++;
		}

		ret.replicas[e->data_type].redundancy =
			min(ret.replicas[e->data_type].redundancy,
			    (int) nr_online - (int) e->nr_required);

		ret.replicas[e->data_type].nr_offline =
			max(ret.replicas[e->data_type].nr_offline,
			    nr_offline);
	}

	rcu_read_unlock();

	for (i = 0; i < ARRAY_SIZE(ret.replicas); i++)
		if (ret.replicas[i].redundancy == INT_MAX)
			ret.replicas[i].redundancy = 0;

	return ret;
}

struct replicas_status bch2_replicas_status(struct bch_fs *c)
{
	return __bch2_replicas_status(c, bch2_online_devs(c));
}

static bool have_enough_devs(struct replicas_status s,
			     enum bch_data_type type,
			     bool force_if_degraded,
			     bool force_if_lost)
{
	return (!s.replicas[type].nr_offline || force_if_degraded) &&
		(s.replicas[type].redundancy >= 0 || force_if_lost);
}

bool bch2_have_enough_devs(struct replicas_status s, unsigned flags)
{
	return (have_enough_devs(s, BCH_DATA_JOURNAL,
				 flags & BCH_FORCE_IF_METADATA_DEGRADED,
				 flags & BCH_FORCE_IF_METADATA_LOST) &&
		have_enough_devs(s, BCH_DATA_BTREE,
				 flags & BCH_FORCE_IF_METADATA_DEGRADED,
				 flags & BCH_FORCE_IF_METADATA_LOST) &&
		have_enough_devs(s, BCH_DATA_USER,
				 flags & BCH_FORCE_IF_DATA_DEGRADED,
				 flags & BCH_FORCE_IF_DATA_LOST));
}

int bch2_replicas_online(struct bch_fs *c, bool meta)
{
	struct replicas_status s = bch2_replicas_status(c);

	return (meta
		? min(s.replicas[BCH_DATA_JOURNAL].redundancy,
		      s.replicas[BCH_DATA_BTREE].redundancy)
		: s.replicas[BCH_DATA_USER].redundancy) + 1;
}

unsigned bch2_dev_has_data(struct bch_fs *c, struct bch_dev *ca)
{
	struct bch_replicas_entry *e;
	struct bch_replicas_cpu *r;
	unsigned i, ret = 0;

	rcu_read_lock();
	r = rcu_dereference(c->replicas);

	for_each_cpu_replicas_entry(r, e)
		for (i = 0; i < e->nr_devs; i++)
			if (e->devs[i] == ca->dev_idx)
				ret |= 1 << e->data_type;

	rcu_read_unlock();

	return ret;
}