summaryrefslogtreecommitdiffstats
path: root/fs/bcachefs/sb-clean.c
blob: c76ad8ea5e4a51c1fb82ea7ac5daf0a1e80a73e6 (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
// SPDX-License-Identifier: GPL-2.0

#include "bcachefs.h"
#include "btree_update_interior.h"
#include "buckets.h"
#include "error.h"
#include "journal_io.h"
#include "replicas.h"
#include "sb-clean.h"
#include "super-io.h"

/*
 * BCH_SB_FIELD_clean:
 *
 * Btree roots, and a few other things, are recovered from the journal after an
 * unclean shutdown - but after a clean shutdown, to avoid having to read the
 * journal, we can store them in the superblock.
 *
 * bch_sb_field_clean simply contains a list of journal entries, stored exactly
 * as they would be in the journal:
 */

int bch2_sb_clean_validate_late(struct bch_fs *c, struct bch_sb_field_clean *clean,
				int write)
{
	struct jset_entry *entry;
	int ret;

	for (entry = clean->start;
	     entry < (struct jset_entry *) vstruct_end(&clean->field);
	     entry = vstruct_next(entry)) {
		ret = bch2_journal_entry_validate(c, NULL, entry,
						  le16_to_cpu(c->disk_sb.sb->version),
						  BCH_SB_BIG_ENDIAN(c->disk_sb.sb),
						  write);
		if (ret)
			return ret;
	}

	return 0;
}

static struct bkey_i *btree_root_find(struct bch_fs *c,
				      struct bch_sb_field_clean *clean,
				      struct jset *j,
				      enum btree_id id, unsigned *level)
{
	struct bkey_i *k;
	struct jset_entry *entry, *start, *end;

	if (clean) {
		start = clean->start;
		end = vstruct_end(&clean->field);
	} else {
		start = j->start;
		end = vstruct_last(j);
	}

	for (entry = start; entry < end; entry = vstruct_next(entry))
		if (entry->type == BCH_JSET_ENTRY_btree_root &&
		    entry->btree_id == id)
			goto found;

	return NULL;
found:
	if (!entry->u64s)
		return ERR_PTR(-EINVAL);

	k = entry->start;
	*level = entry->level;
	return k;
}

int bch2_verify_superblock_clean(struct bch_fs *c,
				 struct bch_sb_field_clean **cleanp,
				 struct jset *j)
{
	unsigned i;
	struct bch_sb_field_clean *clean = *cleanp;
	struct printbuf buf1 = PRINTBUF;
	struct printbuf buf2 = PRINTBUF;
	int ret = 0;

	if (mustfix_fsck_err_on(j->seq != clean->journal_seq, c,
			sb_clean_journal_seq_mismatch,
			"superblock journal seq (%llu) doesn't match journal (%llu) after clean shutdown",
			le64_to_cpu(clean->journal_seq),
			le64_to_cpu(j->seq))) {
		kfree(clean);
		*cleanp = NULL;
		return 0;
	}

	for (i = 0; i < BTREE_ID_NR; i++) {
		struct bkey_i *k1, *k2;
		unsigned l1 = 0, l2 = 0;

		k1 = btree_root_find(c, clean, NULL, i, &l1);
		k2 = btree_root_find(c, NULL, j, i, &l2);

		if (!k1 && !k2)
			continue;

		printbuf_reset(&buf1);
		printbuf_reset(&buf2);

		if (k1)
			bch2_bkey_val_to_text(&buf1, c, bkey_i_to_s_c(k1));
		else
			prt_printf(&buf1, "(none)");

		if (k2)
			bch2_bkey_val_to_text(&buf2, c, bkey_i_to_s_c(k2));
		else
			prt_printf(&buf2, "(none)");

		mustfix_fsck_err_on(!k1 || !k2 ||
				    IS_ERR(k1) ||
				    IS_ERR(k2) ||
				    k1->k.u64s != k2->k.u64s ||
				    memcmp(k1, k2, bkey_bytes(&k1->k)) ||
				    l1 != l2, c,
			sb_clean_btree_root_mismatch,
			"superblock btree root %u doesn't match journal after clean shutdown\n"
			"sb:      l=%u %s\n"
			"journal: l=%u %s\n", i,
			l1, buf1.buf,
			l2, buf2.buf);
	}
fsck_err:
	printbuf_exit(&buf2);
	printbuf_exit(&buf1);
	return ret;
}

struct bch_sb_field_clean *bch2_read_superblock_clean(struct bch_fs *c)
{
	struct bch_sb_field_clean *clean, *sb_clean;
	int ret;

	mutex_lock(&c->sb_lock);
	sb_clean = bch2_sb_field_get(c->disk_sb.sb, clean);

	if (fsck_err_on(!sb_clean, c,
			sb_clean_missing,
			"superblock marked clean but clean section not present")) {
		SET_BCH_SB_CLEAN(c->disk_sb.sb, false);
		c->sb.clean = false;
		mutex_unlock(&c->sb_lock);
		return NULL;
	}

	clean = kmemdup(sb_clean, vstruct_bytes(&sb_clean->field),
			GFP_KERNEL);
	if (!clean) {
		mutex_unlock(&c->sb_lock);
		return ERR_PTR(-BCH_ERR_ENOMEM_read_superblock_clean);
	}

	ret = bch2_sb_clean_validate_late(c, clean, READ);
	if (ret) {
		mutex_unlock(&c->sb_lock);
		return ERR_PTR(ret);
	}

	mutex_unlock(&c->sb_lock);

	return clean;
fsck_err:
	mutex_unlock(&c->sb_lock);
	return ERR_PTR(ret);
}

static struct jset_entry *jset_entry_init(struct jset_entry **end, size_t size)
{
	struct jset_entry *entry = *end;
	unsigned u64s = DIV_ROUND_UP(size, sizeof(u64));

	memset(entry, 0, u64s * sizeof(u64));
	/*
	 * The u64s field counts from the start of data, ignoring the shared
	 * fields.
	 */
	entry->u64s = cpu_to_le16(u64s - 1);

	*end = vstruct_next(*end);
	return entry;
}

void bch2_journal_super_entries_add_common(struct bch_fs *c,
					   struct jset_entry **end,
					   u64 journal_seq)
{
	struct bch_dev *ca;
	unsigned i, dev;

	percpu_down_read(&c->mark_lock);

	if (!journal_seq) {
		for (i = 0; i < ARRAY_SIZE(c->usage); i++)
			bch2_fs_usage_acc_to_base(c, i);
	} else {
		bch2_fs_usage_acc_to_base(c, journal_seq & JOURNAL_BUF_MASK);
	}

	{
		struct jset_entry_usage *u =
			container_of(jset_entry_init(end, sizeof(*u)),
				     struct jset_entry_usage, entry);

		u->entry.type	= BCH_JSET_ENTRY_usage;
		u->entry.btree_id = BCH_FS_USAGE_inodes;
		u->v		= cpu_to_le64(c->usage_base->nr_inodes);
	}

	{
		struct jset_entry_usage *u =
			container_of(jset_entry_init(end, sizeof(*u)),
				     struct jset_entry_usage, entry);

		u->entry.type	= BCH_JSET_ENTRY_usage;
		u->entry.btree_id = BCH_FS_USAGE_key_version;
		u->v		= cpu_to_le64(atomic64_read(&c->key_version));
	}

	for (i = 0; i < BCH_REPLICAS_MAX; i++) {
		struct jset_entry_usage *u =
			container_of(jset_entry_init(end, sizeof(*u)),
				     struct jset_entry_usage, entry);

		u->entry.type	= BCH_JSET_ENTRY_usage;
		u->entry.btree_id = BCH_FS_USAGE_reserved;
		u->entry.level	= i;
		u->v		= cpu_to_le64(c->usage_base->persistent_reserved[i]);
	}

	for (i = 0; i < c->replicas.nr; i++) {
		struct bch_replicas_entry *e =
			cpu_replicas_entry(&c->replicas, i);
		struct jset_entry_data_usage *u =
			container_of(jset_entry_init(end, sizeof(*u) + e->nr_devs),
				     struct jset_entry_data_usage, entry);

		u->entry.type	= BCH_JSET_ENTRY_data_usage;
		u->v		= cpu_to_le64(c->usage_base->replicas[i]);
		unsafe_memcpy(&u->r, e, replicas_entry_bytes(e),
			      "embedded variable length struct");
	}

	for_each_member_device(ca, c, dev) {
		unsigned b = sizeof(struct jset_entry_dev_usage) +
			sizeof(struct jset_entry_dev_usage_type) * BCH_DATA_NR;
		struct jset_entry_dev_usage *u =
			container_of(jset_entry_init(end, b),
				     struct jset_entry_dev_usage, entry);

		u->entry.type = BCH_JSET_ENTRY_dev_usage;
		u->dev = cpu_to_le32(dev);
		u->buckets_ec		= cpu_to_le64(ca->usage_base->buckets_ec);

		for (i = 0; i < BCH_DATA_NR; i++) {
			u->d[i].buckets = cpu_to_le64(ca->usage_base->d[i].buckets);
			u->d[i].sectors	= cpu_to_le64(ca->usage_base->d[i].sectors);
			u->d[i].fragmented = cpu_to_le64(ca->usage_base->d[i].fragmented);
		}
	}

	percpu_up_read(&c->mark_lock);

	for (i = 0; i < 2; i++) {
		struct jset_entry_clock *clock =
			container_of(jset_entry_init(end, sizeof(*clock)),
				     struct jset_entry_clock, entry);

		clock->entry.type = BCH_JSET_ENTRY_clock;
		clock->rw	= i;
		clock->time	= cpu_to_le64(atomic64_read(&c->io_clock[i].now));
	}
}

static int bch2_sb_clean_validate(struct bch_sb *sb,
				  struct bch_sb_field *f,
				  struct printbuf *err)
{
	struct bch_sb_field_clean *clean = field_to_type(f, clean);

	if (vstruct_bytes(&clean->field) < sizeof(*clean)) {
		prt_printf(err, "wrong size (got %zu should be %zu)",
		       vstruct_bytes(&clean->field), sizeof(*clean));
		return -BCH_ERR_invalid_sb_clean;
	}

	return 0;
}

static void bch2_sb_clean_to_text(struct printbuf *out, struct bch_sb *sb,
				  struct bch_sb_field *f)
{
	struct bch_sb_field_clean *clean = field_to_type(f, clean);
	struct jset_entry *entry;

	prt_printf(out, "flags:          %x",	le32_to_cpu(clean->flags));
	prt_newline(out);
	prt_printf(out, "journal_seq:    %llu",	le64_to_cpu(clean->journal_seq));
	prt_newline(out);

	for (entry = clean->start;
	     entry != vstruct_end(&clean->field);
	     entry = vstruct_next(entry)) {
		if (entry->type == BCH_JSET_ENTRY_btree_keys &&
		    !entry->u64s)
			continue;

		bch2_journal_entry_to_text(out, NULL, entry);
		prt_newline(out);
	}
}

const struct bch_sb_field_ops bch_sb_field_ops_clean = {
	.validate	= bch2_sb_clean_validate,
	.to_text	= bch2_sb_clean_to_text,
};

int bch2_fs_mark_dirty(struct bch_fs *c)
{
	int ret;

	/*
	 * Unconditionally write superblock, to verify it hasn't changed before
	 * we go rw:
	 */

	mutex_lock(&c->sb_lock);
	SET_BCH_SB_CLEAN(c->disk_sb.sb, false);
	c->disk_sb.sb->features[0] |= cpu_to_le64(BCH_SB_FEATURES_ALWAYS);

	ret = bch2_write_super(c);
	mutex_unlock(&c->sb_lock);

	return ret;
}

void bch2_fs_mark_clean(struct bch_fs *c)
{
	struct bch_sb_field_clean *sb_clean;
	struct jset_entry *entry;
	unsigned u64s;
	int ret;

	mutex_lock(&c->sb_lock);
	if (BCH_SB_CLEAN(c->disk_sb.sb))
		goto out;

	SET_BCH_SB_CLEAN(c->disk_sb.sb, true);

	c->disk_sb.sb->compat[0] |= cpu_to_le64(1ULL << BCH_COMPAT_alloc_info);
	c->disk_sb.sb->compat[0] |= cpu_to_le64(1ULL << BCH_COMPAT_alloc_metadata);
	c->disk_sb.sb->features[0] &= cpu_to_le64(~(1ULL << BCH_FEATURE_extents_above_btree_updates));
	c->disk_sb.sb->features[0] &= cpu_to_le64(~(1ULL << BCH_FEATURE_btree_updates_journalled));

	u64s = sizeof(*sb_clean) / sizeof(u64) + c->journal.entry_u64s_reserved;

	sb_clean = bch2_sb_field_resize(&c->disk_sb, clean, u64s);
	if (!sb_clean) {
		bch_err(c, "error resizing superblock while setting filesystem clean");
		goto out;
	}

	sb_clean->flags		= 0;
	sb_clean->journal_seq	= cpu_to_le64(atomic64_read(&c->journal.seq));

	/* Trying to catch outstanding bug: */
	BUG_ON(le64_to_cpu(sb_clean->journal_seq) > S64_MAX);

	entry = sb_clean->start;
	bch2_journal_super_entries_add_common(c, &entry, 0);
	entry = bch2_btree_roots_to_journal_entries(c, entry, 0);
	BUG_ON((void *) entry > vstruct_end(&sb_clean->field));

	memset(entry, 0,
	       vstruct_end(&sb_clean->field) - (void *) entry);

	/*
	 * this should be in the write path, and we should be validating every
	 * superblock section:
	 */
	ret = bch2_sb_clean_validate_late(c, sb_clean, WRITE);
	if (ret) {
		bch_err(c, "error writing marking filesystem clean: validate error");
		goto out;
	}

	bch2_write_super(c);
out:
	mutex_unlock(&c->sb_lock);
}