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
|
// SPDX-License-Identifier: GPL-2.0
#include "bcachefs.h"
#include "bkey_methods.h"
#include "btree_types.h"
#include "alloc_background.h"
#include "dirent.h"
#include "ec.h"
#include "error.h"
#include "extents.h"
#include "inode.h"
#include "lru.h"
#include "quota.h"
#include "reflink.h"
#include "subvolume.h"
#include "xattr.h"
const char * const bch2_bkey_types[] = {
#define x(name, nr) #name,
BCH_BKEY_TYPES()
#undef x
NULL
};
static int deleted_key_invalid(const struct bch_fs *c, struct bkey_s_c k,
int rw, struct printbuf *err)
{
return 0;
}
#define bch2_bkey_ops_deleted ((struct bkey_ops) { \
.key_invalid = deleted_key_invalid, \
})
#define bch2_bkey_ops_whiteout ((struct bkey_ops) { \
.key_invalid = deleted_key_invalid, \
})
static int empty_val_key_invalid(const struct bch_fs *c, struct bkey_s_c k,
int rw, struct printbuf *err)
{
if (bkey_val_bytes(k.k)) {
prt_printf(err, "incorrect value size (%zu != 0)",
bkey_val_bytes(k.k));
return -EINVAL;
}
return 0;
}
#define bch2_bkey_ops_error ((struct bkey_ops) { \
.key_invalid = empty_val_key_invalid, \
})
static int key_type_cookie_invalid(const struct bch_fs *c, struct bkey_s_c k,
int rw, struct printbuf *err)
{
if (bkey_val_bytes(k.k) != sizeof(struct bch_cookie)) {
prt_printf(err, "incorrect value size (%zu != %zu)",
bkey_val_bytes(k.k), sizeof(struct bch_cookie));
return -EINVAL;
}
return 0;
}
#define bch2_bkey_ops_cookie ((struct bkey_ops) { \
.key_invalid = key_type_cookie_invalid, \
})
#define bch2_bkey_ops_hash_whiteout ((struct bkey_ops) {\
.key_invalid = empty_val_key_invalid, \
})
static int key_type_inline_data_invalid(const struct bch_fs *c, struct bkey_s_c k,
int rw, struct printbuf *err)
{
return 0;
}
static void key_type_inline_data_to_text(struct printbuf *out, struct bch_fs *c,
struct bkey_s_c k)
{
struct bkey_s_c_inline_data d = bkey_s_c_to_inline_data(k);
unsigned datalen = bkey_inline_data_bytes(k.k);
prt_printf(out, "datalen %u: %*phN",
datalen, min(datalen, 32U), d.v->data);
}
#define bch2_bkey_ops_inline_data ((struct bkey_ops) { \
.key_invalid = key_type_inline_data_invalid, \
.val_to_text = key_type_inline_data_to_text, \
})
static int key_type_set_invalid(const struct bch_fs *c, struct bkey_s_c k,
int rw, struct printbuf *err)
{
if (bkey_val_bytes(k.k)) {
prt_printf(err, "incorrect value size (%zu != %zu)",
bkey_val_bytes(k.k), sizeof(struct bch_cookie));
return -EINVAL;
}
return 0;
}
static bool key_type_set_merge(struct bch_fs *c, struct bkey_s l, struct bkey_s_c r)
{
bch2_key_resize(l.k, l.k->size + r.k->size);
return true;
}
#define bch2_bkey_ops_set ((struct bkey_ops) { \
.key_invalid = key_type_set_invalid, \
.key_merge = key_type_set_merge, \
})
const struct bkey_ops bch2_bkey_ops[] = {
#define x(name, nr) [KEY_TYPE_##name] = bch2_bkey_ops_##name,
BCH_BKEY_TYPES()
#undef x
};
int bch2_bkey_val_invalid(struct bch_fs *c, struct bkey_s_c k,
int rw, struct printbuf *err)
{
if (k.k->type >= KEY_TYPE_MAX) {
prt_printf(err, "invalid type (%u >= %u)", k.k->type, KEY_TYPE_MAX);
return -EINVAL;
}
return bch2_bkey_ops[k.k->type].key_invalid(c, k, rw, err);
}
static unsigned bch2_key_types_allowed[] = {
[BKEY_TYPE_extents] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_whiteout)|
(1U << KEY_TYPE_error)|
(1U << KEY_TYPE_cookie)|
(1U << KEY_TYPE_extent)|
(1U << KEY_TYPE_reservation)|
(1U << KEY_TYPE_reflink_p)|
(1U << KEY_TYPE_inline_data),
[BKEY_TYPE_inodes] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_whiteout)|
(1U << KEY_TYPE_inode)|
(1U << KEY_TYPE_inode_v2)|
(1U << KEY_TYPE_inode_generation),
[BKEY_TYPE_dirents] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_whiteout)|
(1U << KEY_TYPE_hash_whiteout)|
(1U << KEY_TYPE_dirent),
[BKEY_TYPE_xattrs] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_whiteout)|
(1U << KEY_TYPE_cookie)|
(1U << KEY_TYPE_hash_whiteout)|
(1U << KEY_TYPE_xattr),
[BKEY_TYPE_alloc] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_alloc)|
(1U << KEY_TYPE_alloc_v2)|
(1U << KEY_TYPE_alloc_v3)|
(1U << KEY_TYPE_alloc_v4),
[BKEY_TYPE_quotas] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_quota),
[BKEY_TYPE_stripes] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_stripe),
[BKEY_TYPE_reflink] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_reflink_v)|
(1U << KEY_TYPE_indirect_inline_data),
[BKEY_TYPE_subvolumes] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_subvolume),
[BKEY_TYPE_snapshots] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_snapshot),
[BKEY_TYPE_lru] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_lru),
[BKEY_TYPE_freespace] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_set),
[BKEY_TYPE_need_discard] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_set),
[BKEY_TYPE_btree] =
(1U << KEY_TYPE_deleted)|
(1U << KEY_TYPE_btree_ptr)|
(1U << KEY_TYPE_btree_ptr_v2),
};
int __bch2_bkey_invalid(struct bch_fs *c, struct bkey_s_c k,
enum btree_node_type type,
int rw, struct printbuf *err)
{
if (k.k->u64s < BKEY_U64s) {
prt_printf(err, "u64s too small (%u < %zu)", k.k->u64s, BKEY_U64s);
return -EINVAL;
}
if (!(bch2_key_types_allowed[type] & (1U << k.k->type))) {
prt_printf(err, "invalid key type for btree %s (%s)",
bch2_btree_ids[type], bch2_bkey_types[k.k->type]);
return -EINVAL;
}
if (btree_node_type_is_extents(type) && !bkey_whiteout(k.k)) {
if (k.k->size == 0) {
prt_printf(err, "size == 0");
return -EINVAL;
}
if (k.k->size > k.k->p.offset) {
prt_printf(err, "size greater than offset (%u > %llu)",
k.k->size, k.k->p.offset);
return -EINVAL;
}
} else {
if (k.k->size) {
prt_printf(err, "size != 0");
return -EINVAL;
}
}
if (type != BKEY_TYPE_btree &&
!btree_type_has_snapshots(type) &&
k.k->p.snapshot) {
prt_printf(err, "nonzero snapshot");
return -EINVAL;
}
if (type != BKEY_TYPE_btree &&
btree_type_has_snapshots(type) &&
!k.k->p.snapshot) {
prt_printf(err, "snapshot == 0");
return -EINVAL;
}
if (type != BKEY_TYPE_btree &&
!bkey_cmp(k.k->p, POS_MAX)) {
prt_printf(err, "key at POS_MAX");
return -EINVAL;
}
return 0;
}
int bch2_bkey_invalid(struct bch_fs *c, struct bkey_s_c k,
enum btree_node_type type,
int rw, struct printbuf *err)
{
return __bch2_bkey_invalid(c, k, type, rw, err) ?:
bch2_bkey_val_invalid(c, k, rw, err);
}
int bch2_bkey_in_btree_node(struct btree *b, struct bkey_s_c k,
struct printbuf *err)
{
if (bpos_cmp(k.k->p, b->data->min_key) < 0) {
prt_printf(err, "key before start of btree node");
return -EINVAL;
}
if (bpos_cmp(k.k->p, b->data->max_key) > 0) {
prt_printf(err, "key past end of btree node");
return -EINVAL;
}
return 0;
}
void bch2_bpos_to_text(struct printbuf *out, struct bpos pos)
{
if (!bpos_cmp(pos, POS_MIN))
prt_printf(out, "POS_MIN");
else if (!bpos_cmp(pos, POS_MAX))
prt_printf(out, "POS_MAX");
else if (!bpos_cmp(pos, SPOS_MAX))
prt_printf(out, "SPOS_MAX");
else {
if (pos.inode == U64_MAX)
prt_printf(out, "U64_MAX");
else
prt_printf(out, "%llu", pos.inode);
prt_printf(out, ":");
if (pos.offset == U64_MAX)
prt_printf(out, "U64_MAX");
else
prt_printf(out, "%llu", pos.offset);
prt_printf(out, ":");
if (pos.snapshot == U32_MAX)
prt_printf(out, "U32_MAX");
else
prt_printf(out, "%u", pos.snapshot);
}
}
void bch2_bkey_to_text(struct printbuf *out, const struct bkey *k)
{
if (k) {
prt_printf(out, "u64s %u type ", k->u64s);
if (k->type < KEY_TYPE_MAX)
prt_printf(out, "%s ", bch2_bkey_types[k->type]);
else
prt_printf(out, "%u ", k->type);
bch2_bpos_to_text(out, k->p);
prt_printf(out, " len %u ver %llu", k->size, k->version.lo);
} else {
prt_printf(out, "(null)");
}
}
void bch2_val_to_text(struct printbuf *out, struct bch_fs *c,
struct bkey_s_c k)
{
if (k.k->type < KEY_TYPE_MAX) {
const struct bkey_ops *ops = &bch2_bkey_ops[k.k->type];
if (likely(ops->val_to_text))
ops->val_to_text(out, c, k);
} else {
prt_printf(out, "(invalid type %u)", k.k->type);
}
}
void bch2_bkey_val_to_text(struct printbuf *out, struct bch_fs *c,
struct bkey_s_c k)
{
bch2_bkey_to_text(out, k.k);
if (bkey_val_bytes(k.k)) {
prt_printf(out, ": ");
bch2_val_to_text(out, c, k);
}
}
void bch2_bkey_swab_val(struct bkey_s k)
{
const struct bkey_ops *ops = &bch2_bkey_ops[k.k->type];
if (ops->swab)
ops->swab(k);
}
bool bch2_bkey_normalize(struct bch_fs *c, struct bkey_s k)
{
const struct bkey_ops *ops = &bch2_bkey_ops[k.k->type];
return ops->key_normalize
? ops->key_normalize(c, k)
: false;
}
bool bch2_bkey_merge(struct bch_fs *c, struct bkey_s l, struct bkey_s_c r)
{
const struct bkey_ops *ops = &bch2_bkey_ops[l.k->type];
return bch2_bkey_maybe_mergable(l.k, r.k) &&
(u64) l.k->size + r.k->size <= KEY_SIZE_MAX &&
bch2_bkey_ops[l.k->type].key_merge &&
!bch2_key_merging_disabled &&
ops->key_merge(c, l, r);
}
static const struct old_bkey_type {
u8 btree_node_type;
u8 old;
u8 new;
} bkey_renumber_table[] = {
{BKEY_TYPE_btree, 128, KEY_TYPE_btree_ptr },
{BKEY_TYPE_extents, 128, KEY_TYPE_extent },
{BKEY_TYPE_extents, 129, KEY_TYPE_extent },
{BKEY_TYPE_extents, 130, KEY_TYPE_reservation },
{BKEY_TYPE_inodes, 128, KEY_TYPE_inode },
{BKEY_TYPE_inodes, 130, KEY_TYPE_inode_generation },
{BKEY_TYPE_dirents, 128, KEY_TYPE_dirent },
{BKEY_TYPE_dirents, 129, KEY_TYPE_hash_whiteout },
{BKEY_TYPE_xattrs, 128, KEY_TYPE_xattr },
{BKEY_TYPE_xattrs, 129, KEY_TYPE_hash_whiteout },
{BKEY_TYPE_alloc, 128, KEY_TYPE_alloc },
{BKEY_TYPE_quotas, 128, KEY_TYPE_quota },
};
void bch2_bkey_renumber(enum btree_node_type btree_node_type,
struct bkey_packed *k,
int write)
{
const struct old_bkey_type *i;
for (i = bkey_renumber_table;
i < bkey_renumber_table + ARRAY_SIZE(bkey_renumber_table);
i++)
if (btree_node_type == i->btree_node_type &&
k->type == (write ? i->new : i->old)) {
k->type = write ? i->old : i->new;
break;
}
}
void __bch2_bkey_compat(unsigned level, enum btree_id btree_id,
unsigned version, unsigned big_endian,
int write,
struct bkey_format *f,
struct bkey_packed *k)
{
const struct bkey_ops *ops;
struct bkey uk;
struct bkey_s u;
unsigned nr_compat = 5;
int i;
/*
* Do these operations in reverse order in the write path:
*/
for (i = 0; i < nr_compat; i++)
switch (!write ? i : nr_compat - 1 - i) {
case 0:
if (big_endian != CPU_BIG_ENDIAN)
bch2_bkey_swab_key(f, k);
break;
case 1:
if (version < bcachefs_metadata_version_bkey_renumber)
bch2_bkey_renumber(__btree_node_type(level, btree_id), k, write);
break;
case 2:
if (version < bcachefs_metadata_version_inode_btree_change &&
btree_id == BTREE_ID_inodes) {
if (!bkey_packed(k)) {
struct bkey_i *u = packed_to_bkey(k);
swap(u->k.p.inode, u->k.p.offset);
} else if (f->bits_per_field[BKEY_FIELD_INODE] &&
f->bits_per_field[BKEY_FIELD_OFFSET]) {
struct bkey_format tmp = *f, *in = f, *out = &tmp;
swap(tmp.bits_per_field[BKEY_FIELD_INODE],
tmp.bits_per_field[BKEY_FIELD_OFFSET]);
swap(tmp.field_offset[BKEY_FIELD_INODE],
tmp.field_offset[BKEY_FIELD_OFFSET]);
if (!write)
swap(in, out);
uk = __bch2_bkey_unpack_key(in, k);
swap(uk.p.inode, uk.p.offset);
BUG_ON(!bch2_bkey_pack_key(k, &uk, out));
}
}
break;
case 3:
if (version < bcachefs_metadata_version_snapshot &&
(level || btree_type_has_snapshots(btree_id))) {
struct bkey_i *u = packed_to_bkey(k);
if (u) {
u->k.p.snapshot = write
? 0 : U32_MAX;
} else {
u64 min_packed = f->field_offset[BKEY_FIELD_SNAPSHOT];
u64 max_packed = min_packed +
~(~0ULL << f->bits_per_field[BKEY_FIELD_SNAPSHOT]);
uk = __bch2_bkey_unpack_key(f, k);
uk.p.snapshot = write
? min_packed : min_t(u64, U32_MAX, max_packed);
BUG_ON(!bch2_bkey_pack_key(k, &uk, f));
}
}
break;
case 4:
if (!bkey_packed(k)) {
u = bkey_i_to_s(packed_to_bkey(k));
} else {
uk = __bch2_bkey_unpack_key(f, k);
u.k = &uk;
u.v = bkeyp_val(f, k);
}
if (big_endian != CPU_BIG_ENDIAN)
bch2_bkey_swab_val(u);
ops = &bch2_bkey_ops[k->type];
if (ops->compat)
ops->compat(btree_id, version, big_endian, write, u);
break;
default:
BUG();
}
}
|