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
|
/*
* Copyright (C) 2017 Oracle. All Rights Reserved.
*
* Author: Darrick J. Wong <darrick.wong@oracle.com>
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it would 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 the Free Software Foundation,
* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA.
*/
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_shared.h"
#include "xfs_format.h"
#include "xfs_trans_resv.h"
#include "xfs_mount.h"
#include "xfs_defer.h"
#include "xfs_btree.h"
#include "xfs_bit.h"
#include "xfs_log_format.h"
#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_inode.h"
#include "xfs_alloc.h"
#include "scrub/xfs_scrub.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/trace.h"
/*
* Set up scrub to check all the static metadata in each AG.
* This means the SB, AGF, AGI, and AGFL headers.
*/
int
xfs_scrub_setup_ag_header(
struct xfs_scrub_context *sc,
struct xfs_inode *ip)
{
struct xfs_mount *mp = sc->mp;
if (sc->sm->sm_agno >= mp->m_sb.sb_agcount ||
sc->sm->sm_ino || sc->sm->sm_gen)
return -EINVAL;
return xfs_scrub_setup_fs(sc, ip);
}
/* Walk all the blocks in the AGFL. */
int
xfs_scrub_walk_agfl(
struct xfs_scrub_context *sc,
int (*fn)(struct xfs_scrub_context *,
xfs_agblock_t bno, void *),
void *priv)
{
struct xfs_agf *agf;
__be32 *agfl_bno;
struct xfs_mount *mp = sc->mp;
unsigned int flfirst;
unsigned int fllast;
int i;
int error;
agf = XFS_BUF_TO_AGF(sc->sa.agf_bp);
agfl_bno = XFS_BUF_TO_AGFL_BNO(mp, sc->sa.agfl_bp);
flfirst = be32_to_cpu(agf->agf_flfirst);
fllast = be32_to_cpu(agf->agf_fllast);
/* Nothing to walk in an empty AGFL. */
if (agf->agf_flcount == cpu_to_be32(0))
return 0;
/* first to last is a consecutive list. */
if (fllast >= flfirst) {
for (i = flfirst; i <= fllast; i++) {
error = fn(sc, be32_to_cpu(agfl_bno[i]), priv);
if (error)
return error;
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
return error;
}
return 0;
}
/* first to the end */
for (i = flfirst; i < XFS_AGFL_SIZE(mp); i++) {
error = fn(sc, be32_to_cpu(agfl_bno[i]), priv);
if (error)
return error;
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
return error;
}
/* the start to last. */
for (i = 0; i <= fllast; i++) {
error = fn(sc, be32_to_cpu(agfl_bno[i]), priv);
if (error)
return error;
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
return error;
}
return 0;
}
/* Superblock */
/*
* Scrub the filesystem superblock.
*
* Note: We do /not/ attempt to check AG 0's superblock. Mount is
* responsible for validating all the geometry information in sb 0, so
* if the filesystem is capable of initiating online scrub, then clearly
* sb 0 is ok and we can use its information to check everything else.
*/
int
xfs_scrub_superblock(
struct xfs_scrub_context *sc)
{
struct xfs_mount *mp = sc->mp;
struct xfs_buf *bp;
struct xfs_dsb *sb;
xfs_agnumber_t agno;
uint32_t v2_ok;
__be32 features_mask;
int error;
__be16 vernum_mask;
agno = sc->sm->sm_agno;
if (agno == 0)
return 0;
error = xfs_trans_read_buf(mp, sc->tp, mp->m_ddev_targp,
XFS_AGB_TO_DADDR(mp, agno, XFS_SB_BLOCK(mp)),
XFS_FSS_TO_BB(mp, 1), 0, &bp, &xfs_sb_buf_ops);
if (!xfs_scrub_process_error(sc, agno, XFS_SB_BLOCK(mp), &error))
return error;
sb = XFS_BUF_TO_SBP(bp);
/*
* Verify the geometries match. Fields that are permanently
* set by mkfs are checked; fields that can be updated later
* (and are not propagated to backup superblocks) are preen
* checked.
*/
if (sb->sb_blocksize != cpu_to_be32(mp->m_sb.sb_blocksize))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_dblocks != cpu_to_be64(mp->m_sb.sb_dblocks))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_rblocks != cpu_to_be64(mp->m_sb.sb_rblocks))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_rextents != cpu_to_be64(mp->m_sb.sb_rextents))
xfs_scrub_block_set_corrupt(sc, bp);
if (!uuid_equal(&sb->sb_uuid, &mp->m_sb.sb_uuid))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_logstart != cpu_to_be64(mp->m_sb.sb_logstart))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_rootino != cpu_to_be64(mp->m_sb.sb_rootino))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_rbmino != cpu_to_be64(mp->m_sb.sb_rbmino))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_rsumino != cpu_to_be64(mp->m_sb.sb_rsumino))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_rextsize != cpu_to_be32(mp->m_sb.sb_rextsize))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_agblocks != cpu_to_be32(mp->m_sb.sb_agblocks))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_agcount != cpu_to_be32(mp->m_sb.sb_agcount))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_rbmblocks != cpu_to_be32(mp->m_sb.sb_rbmblocks))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_logblocks != cpu_to_be32(mp->m_sb.sb_logblocks))
xfs_scrub_block_set_corrupt(sc, bp);
/* Check sb_versionnum bits that are set at mkfs time. */
vernum_mask = cpu_to_be16(~XFS_SB_VERSION_OKBITS |
XFS_SB_VERSION_NUMBITS |
XFS_SB_VERSION_ALIGNBIT |
XFS_SB_VERSION_DALIGNBIT |
XFS_SB_VERSION_SHAREDBIT |
XFS_SB_VERSION_LOGV2BIT |
XFS_SB_VERSION_SECTORBIT |
XFS_SB_VERSION_EXTFLGBIT |
XFS_SB_VERSION_DIRV2BIT);
if ((sb->sb_versionnum & vernum_mask) !=
(cpu_to_be16(mp->m_sb.sb_versionnum) & vernum_mask))
xfs_scrub_block_set_corrupt(sc, bp);
/* Check sb_versionnum bits that can be set after mkfs time. */
vernum_mask = cpu_to_be16(XFS_SB_VERSION_ATTRBIT |
XFS_SB_VERSION_NLINKBIT |
XFS_SB_VERSION_QUOTABIT);
if ((sb->sb_versionnum & vernum_mask) !=
(cpu_to_be16(mp->m_sb.sb_versionnum) & vernum_mask))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_sectsize != cpu_to_be16(mp->m_sb.sb_sectsize))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_inodesize != cpu_to_be16(mp->m_sb.sb_inodesize))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_inopblock != cpu_to_be16(mp->m_sb.sb_inopblock))
xfs_scrub_block_set_corrupt(sc, bp);
if (memcmp(sb->sb_fname, mp->m_sb.sb_fname, sizeof(sb->sb_fname)))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_blocklog != mp->m_sb.sb_blocklog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_sectlog != mp->m_sb.sb_sectlog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_inodelog != mp->m_sb.sb_inodelog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_inopblog != mp->m_sb.sb_inopblog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_agblklog != mp->m_sb.sb_agblklog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_rextslog != mp->m_sb.sb_rextslog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_imax_pct != mp->m_sb.sb_imax_pct)
xfs_scrub_block_set_preen(sc, bp);
/*
* Skip the summary counters since we track them in memory anyway.
* sb_icount, sb_ifree, sb_fdblocks, sb_frexents
*/
if (sb->sb_uquotino != cpu_to_be64(mp->m_sb.sb_uquotino))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_gquotino != cpu_to_be64(mp->m_sb.sb_gquotino))
xfs_scrub_block_set_preen(sc, bp);
/*
* Skip the quota flags since repair will force quotacheck.
* sb_qflags
*/
if (sb->sb_flags != mp->m_sb.sb_flags)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_shared_vn != mp->m_sb.sb_shared_vn)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_inoalignmt != cpu_to_be32(mp->m_sb.sb_inoalignmt))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_unit != cpu_to_be32(mp->m_sb.sb_unit))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_width != cpu_to_be32(mp->m_sb.sb_width))
xfs_scrub_block_set_preen(sc, bp);
if (sb->sb_dirblklog != mp->m_sb.sb_dirblklog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_logsectlog != mp->m_sb.sb_logsectlog)
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_logsectsize != cpu_to_be16(mp->m_sb.sb_logsectsize))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_logsunit != cpu_to_be32(mp->m_sb.sb_logsunit))
xfs_scrub_block_set_corrupt(sc, bp);
/* Do we see any invalid bits in sb_features2? */
if (!xfs_sb_version_hasmorebits(&mp->m_sb)) {
if (sb->sb_features2 != 0)
xfs_scrub_block_set_corrupt(sc, bp);
} else {
v2_ok = XFS_SB_VERSION2_OKBITS;
if (XFS_SB_VERSION_NUM(&mp->m_sb) >= XFS_SB_VERSION_5)
v2_ok |= XFS_SB_VERSION2_CRCBIT;
if (!!(sb->sb_features2 & cpu_to_be32(~v2_ok)))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_features2 != sb->sb_bad_features2)
xfs_scrub_block_set_preen(sc, bp);
}
/* Check sb_features2 flags that are set at mkfs time. */
features_mask = cpu_to_be32(XFS_SB_VERSION2_LAZYSBCOUNTBIT |
XFS_SB_VERSION2_PROJID32BIT |
XFS_SB_VERSION2_CRCBIT |
XFS_SB_VERSION2_FTYPE);
if ((sb->sb_features2 & features_mask) !=
(cpu_to_be32(mp->m_sb.sb_features2) & features_mask))
xfs_scrub_block_set_corrupt(sc, bp);
/* Check sb_features2 flags that can be set after mkfs time. */
features_mask = cpu_to_be32(XFS_SB_VERSION2_ATTR2BIT);
if ((sb->sb_features2 & features_mask) !=
(cpu_to_be32(mp->m_sb.sb_features2) & features_mask))
xfs_scrub_block_set_corrupt(sc, bp);
if (!xfs_sb_version_hascrc(&mp->m_sb)) {
/* all v5 fields must be zero */
if (memchr_inv(&sb->sb_features_compat, 0,
sizeof(struct xfs_dsb) -
offsetof(struct xfs_dsb, sb_features_compat)))
xfs_scrub_block_set_corrupt(sc, bp);
} else {
/* Check compat flags; all are set at mkfs time. */
features_mask = cpu_to_be32(XFS_SB_FEAT_COMPAT_UNKNOWN);
if ((sb->sb_features_compat & features_mask) !=
(cpu_to_be32(mp->m_sb.sb_features_compat) & features_mask))
xfs_scrub_block_set_corrupt(sc, bp);
/* Check ro compat flags; all are set at mkfs time. */
features_mask = cpu_to_be32(XFS_SB_FEAT_RO_COMPAT_UNKNOWN |
XFS_SB_FEAT_RO_COMPAT_FINOBT |
XFS_SB_FEAT_RO_COMPAT_RMAPBT |
XFS_SB_FEAT_RO_COMPAT_REFLINK);
if ((sb->sb_features_ro_compat & features_mask) !=
(cpu_to_be32(mp->m_sb.sb_features_ro_compat) &
features_mask))
xfs_scrub_block_set_corrupt(sc, bp);
/* Check incompat flags; all are set at mkfs time. */
features_mask = cpu_to_be32(XFS_SB_FEAT_INCOMPAT_UNKNOWN |
XFS_SB_FEAT_INCOMPAT_FTYPE |
XFS_SB_FEAT_INCOMPAT_SPINODES |
XFS_SB_FEAT_INCOMPAT_META_UUID);
if ((sb->sb_features_incompat & features_mask) !=
(cpu_to_be32(mp->m_sb.sb_features_incompat) &
features_mask))
xfs_scrub_block_set_corrupt(sc, bp);
/* Check log incompat flags; all are set at mkfs time. */
features_mask = cpu_to_be32(XFS_SB_FEAT_INCOMPAT_LOG_UNKNOWN);
if ((sb->sb_features_log_incompat & features_mask) !=
(cpu_to_be32(mp->m_sb.sb_features_log_incompat) &
features_mask))
xfs_scrub_block_set_corrupt(sc, bp);
/* Don't care about sb_crc */
if (sb->sb_spino_align != cpu_to_be32(mp->m_sb.sb_spino_align))
xfs_scrub_block_set_corrupt(sc, bp);
if (sb->sb_pquotino != cpu_to_be64(mp->m_sb.sb_pquotino))
xfs_scrub_block_set_preen(sc, bp);
/* Don't care about sb_lsn */
}
if (xfs_sb_version_hasmetauuid(&mp->m_sb)) {
/* The metadata UUID must be the same for all supers */
if (!uuid_equal(&sb->sb_meta_uuid, &mp->m_sb.sb_meta_uuid))
xfs_scrub_block_set_corrupt(sc, bp);
}
/* Everything else must be zero. */
if (memchr_inv(sb + 1, 0,
BBTOB(bp->b_length) - sizeof(struct xfs_dsb)))
xfs_scrub_block_set_corrupt(sc, bp);
return error;
}
/* AGF */
/* Scrub the AGF. */
int
xfs_scrub_agf(
struct xfs_scrub_context *sc)
{
struct xfs_mount *mp = sc->mp;
struct xfs_agf *agf;
xfs_agnumber_t agno;
xfs_agblock_t agbno;
xfs_agblock_t eoag;
xfs_agblock_t agfl_first;
xfs_agblock_t agfl_last;
xfs_agblock_t agfl_count;
xfs_agblock_t fl_count;
int level;
int error = 0;
agno = sc->sa.agno = sc->sm->sm_agno;
error = xfs_scrub_ag_read_headers(sc, agno, &sc->sa.agi_bp,
&sc->sa.agf_bp, &sc->sa.agfl_bp);
if (!xfs_scrub_process_error(sc, agno, XFS_AGF_BLOCK(sc->mp), &error))
goto out;
agf = XFS_BUF_TO_AGF(sc->sa.agf_bp);
/* Check the AG length */
eoag = be32_to_cpu(agf->agf_length);
if (eoag != xfs_ag_block_count(mp, agno))
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
/* Check the AGF btree roots and levels */
agbno = be32_to_cpu(agf->agf_roots[XFS_BTNUM_BNO]);
if (!xfs_verify_agbno(mp, agno, agbno))
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
agbno = be32_to_cpu(agf->agf_roots[XFS_BTNUM_CNT]);
if (!xfs_verify_agbno(mp, agno, agbno))
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
level = be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNO]);
if (level <= 0 || level > XFS_BTREE_MAXLEVELS)
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
level = be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNT]);
if (level <= 0 || level > XFS_BTREE_MAXLEVELS)
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
if (xfs_sb_version_hasrmapbt(&mp->m_sb)) {
agbno = be32_to_cpu(agf->agf_roots[XFS_BTNUM_RMAP]);
if (!xfs_verify_agbno(mp, agno, agbno))
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
level = be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]);
if (level <= 0 || level > XFS_BTREE_MAXLEVELS)
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
}
if (xfs_sb_version_hasreflink(&mp->m_sb)) {
agbno = be32_to_cpu(agf->agf_refcount_root);
if (!xfs_verify_agbno(mp, agno, agbno))
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
level = be32_to_cpu(agf->agf_refcount_level);
if (level <= 0 || level > XFS_BTREE_MAXLEVELS)
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
}
/* Check the AGFL counters */
agfl_first = be32_to_cpu(agf->agf_flfirst);
agfl_last = be32_to_cpu(agf->agf_fllast);
agfl_count = be32_to_cpu(agf->agf_flcount);
if (agfl_last > agfl_first)
fl_count = agfl_last - agfl_first + 1;
else
fl_count = XFS_AGFL_SIZE(mp) - agfl_first + agfl_last + 1;
if (agfl_count != 0 && fl_count != agfl_count)
xfs_scrub_block_set_corrupt(sc, sc->sa.agf_bp);
out:
return error;
}
/* AGFL */
/* Scrub an AGFL block. */
STATIC int
xfs_scrub_agfl_block(
struct xfs_scrub_context *sc,
xfs_agblock_t agbno,
void *priv)
{
struct xfs_mount *mp = sc->mp;
xfs_agnumber_t agno = sc->sa.agno;
if (!xfs_verify_agbno(mp, agno, agbno))
xfs_scrub_block_set_corrupt(sc, sc->sa.agfl_bp);
return 0;
}
/* Scrub the AGFL. */
int
xfs_scrub_agfl(
struct xfs_scrub_context *sc)
{
xfs_agnumber_t agno;
int error;
agno = sc->sa.agno = sc->sm->sm_agno;
error = xfs_scrub_ag_read_headers(sc, agno, &sc->sa.agi_bp,
&sc->sa.agf_bp, &sc->sa.agfl_bp);
if (!xfs_scrub_process_error(sc, agno, XFS_AGFL_BLOCK(sc->mp), &error))
goto out;
if (!sc->sa.agf_bp)
return -EFSCORRUPTED;
/* Check the blocks in the AGFL. */
return xfs_scrub_walk_agfl(sc, xfs_scrub_agfl_block, NULL);
out:
return error;
}
|