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
|
/* verify.c --- verification of FSX filesystems
*
* ====================================================================
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing,
* software distributed under the License is distributed on an
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
* KIND, either express or implied. See the License for the
* specific language governing permissions and limitations
* under the License.
* ====================================================================
*/
#include "verify.h"
#include "fs_x.h"
#include "svn_time.h"
#include "private/svn_subr_private.h"
#include "cached_data.h"
#include "rep-cache.h"
#include "util.h"
#include "index.h"
#include "../libsvn_fs/fs-loader.h"
#include "svn_private_config.h"
/** Verifying. **/
/* Baton type expected by verify_walker(). The purpose is to limit the
* number of notifications sent.
*/
typedef struct verify_walker_baton_t
{
/* number of calls to verify_walker() since the last clean */
int iteration_count;
/* progress notification callback to invoke periodically (may be NULL) */
svn_fs_progress_notify_func_t notify_func;
/* baton to use with NOTIFY_FUNC */
void *notify_baton;
/* remember the last revision for which we called notify_func */
svn_revnum_t last_notified_revision;
} verify_walker_baton_t;
/* Used by svn_fs_x__verify().
Implements svn_fs_x__walk_rep_reference().walker. */
static svn_error_t *
verify_walker(svn_fs_x__representation_t *rep,
void *baton,
svn_fs_t *fs,
apr_pool_t *scratch_pool)
{
verify_walker_baton_t *walker_baton = baton;
/* notify and free resources periodically */
if (walker_baton->iteration_count > 1000)
{
svn_revnum_t revision = svn_fs_x__get_revnum(rep->id.change_set);
if ( walker_baton->notify_func
&& revision != walker_baton->last_notified_revision)
{
walker_baton->notify_func(revision,
walker_baton->notify_baton,
scratch_pool);
walker_baton->last_notified_revision = revision;
}
walker_baton->iteration_count = 0;
}
/* access the repo data */
SVN_ERR(svn_fs_x__check_rep(rep, fs, scratch_pool));
/* update resource usage counters */
walker_baton->iteration_count++;
return SVN_NO_ERROR;
}
/* Verify the rep cache DB's consistency with our rev / pack data.
* The function signature is similar to svn_fs_x__verify.
* The values of START and END have already been auto-selected and
* verified.
*/
static svn_error_t *
verify_rep_cache(svn_fs_t *fs,
svn_revnum_t start,
svn_revnum_t end,
svn_fs_progress_notify_func_t notify_func,
void *notify_baton,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_boolean_t exists;
/* rep-cache verification. */
SVN_ERR(svn_fs_x__exists_rep_cache(&exists, fs, scratch_pool));
if (exists)
{
/* provide a baton to allow the reuse of open file handles between
iterations (saves 2/3 of OS level file operations). */
verify_walker_baton_t *baton
= apr_pcalloc(scratch_pool, sizeof(*baton));
baton->last_notified_revision = SVN_INVALID_REVNUM;
baton->notify_func = notify_func;
baton->notify_baton = notify_baton;
/* tell the user that we are now ready to do *something* */
if (notify_func)
notify_func(SVN_INVALID_REVNUM, notify_baton, scratch_pool);
/* Do not attempt to walk the rep-cache database if its file does
not exist, since doing so would create it --- which may confuse
the administrator. Don't take any lock. */
SVN_ERR(svn_fs_x__walk_rep_reference(fs, start, end,
verify_walker, baton,
cancel_func, cancel_baton,
scratch_pool));
}
return SVN_NO_ERROR;
}
/* Verify that the MD5 checksum of the data between offsets START and END
* in FILE matches the EXPECTED checksum. If there is a mismatch use the
* indedx NAME in the error message. Supports cancellation with CANCEL_FUNC
* and CANCEL_BATON. SCRATCH_POOL is for temporary allocations. */
static svn_error_t *
verify_index_checksum(apr_file_t *file,
const char *name,
apr_off_t start,
apr_off_t end,
svn_checksum_t *expected,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
unsigned char buffer[SVN__STREAM_CHUNK_SIZE];
apr_off_t size = end - start;
svn_checksum_t *actual;
svn_checksum_ctx_t *context
= svn_checksum_ctx_create(svn_checksum_md5, scratch_pool);
/* Calculate the index checksum. */
SVN_ERR(svn_io_file_seek(file, APR_SET, &start, scratch_pool));
while (size > 0)
{
apr_size_t to_read = size > sizeof(buffer)
? sizeof(buffer)
: (apr_size_t)size;
SVN_ERR(svn_io_file_read_full2(file, buffer, to_read, NULL, NULL,
scratch_pool));
SVN_ERR(svn_checksum_update(context, buffer, to_read));
size -= to_read;
if (cancel_func)
SVN_ERR(cancel_func(cancel_baton));
}
SVN_ERR(svn_checksum_final(&actual, context, scratch_pool));
/* Verify that it matches the expected checksum. */
if (!svn_checksum_match(expected, actual))
{
const char *file_name;
SVN_ERR(svn_io_file_name_get(&file_name, file, scratch_pool));
SVN_ERR(svn_checksum_mismatch_err(expected, actual, scratch_pool,
_("%s checksum mismatch in file %s"),
name, file_name));
}
return SVN_NO_ERROR;
}
/* Verify the MD5 checksums of the index data in the rev / pack file
* containing revision START in FS. If given, invoke CANCEL_FUNC with
* CANCEL_BATON at regular intervals. Use SCRATCH_POOL for temporary
* allocations.
*/
static svn_error_t *
verify_index_checksums(svn_fs_t *fs,
svn_revnum_t start,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_fs_x__revision_file_t *rev_file;
/* Open the rev / pack file and read the footer */
SVN_ERR(svn_fs_x__open_pack_or_rev_file(&rev_file, fs, start,
scratch_pool, scratch_pool));
SVN_ERR(svn_fs_x__auto_read_footer(rev_file));
/* Verify the index contents against the checksum from the footer. */
SVN_ERR(verify_index_checksum(rev_file->file, "L2P index",
rev_file->l2p_offset, rev_file->p2l_offset,
rev_file->l2p_checksum,
cancel_func, cancel_baton, scratch_pool));
SVN_ERR(verify_index_checksum(rev_file->file, "P2L index",
rev_file->p2l_offset, rev_file->footer_offset,
rev_file->p2l_checksum,
cancel_func, cancel_baton, scratch_pool));
/* Done. */
SVN_ERR(svn_fs_x__close_revision_file(rev_file));
return SVN_NO_ERROR;
}
/* Verify that for all log-to-phys index entries for revisions START to
* START + COUNT-1 in FS there is a consistent entry in the phys-to-log
* index. If given, invoke CANCEL_FUNC with CANCEL_BATON at regular
* intervals. Use SCRATCH_POOL for temporary allocations.
*/
static svn_error_t *
compare_l2p_to_p2l_index(svn_fs_t *fs,
svn_revnum_t start,
svn_revnum_t count,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_revnum_t i;
apr_pool_t *iterpool = svn_pool_create(scratch_pool);
apr_array_header_t *max_ids;
/* common file access structure */
svn_fs_x__revision_file_t *rev_file;
SVN_ERR(svn_fs_x__open_pack_or_rev_file(&rev_file, fs, start, scratch_pool,
iterpool));
/* determine the range of items to check for each revision */
SVN_ERR(svn_fs_x__l2p_get_max_ids(&max_ids, fs, start, count, scratch_pool,
iterpool));
/* check all items in all revisions if the given range */
for (i = 0; i < max_ids->nelts; ++i)
{
apr_uint64_t k;
apr_uint64_t max_id = APR_ARRAY_IDX(max_ids, i, apr_uint64_t);
svn_revnum_t revision = start + i;
for (k = 0; k < max_id; ++k)
{
apr_off_t offset;
apr_uint32_t sub_item;
svn_fs_x__id_t l2p_item;
svn_fs_x__id_t *p2l_item;
l2p_item.change_set = svn_fs_x__change_set_by_rev(revision);
l2p_item.number = k;
/* get L2P entry. Ignore unused entries. */
SVN_ERR(svn_fs_x__item_offset(&offset, &sub_item, fs, rev_file,
&l2p_item, iterpool));
if (offset == -1)
continue;
/* find the corresponding P2L entry */
SVN_ERR(svn_fs_x__p2l_item_lookup(&p2l_item, fs, rev_file,
revision, offset, sub_item,
iterpool, iterpool));
if (p2l_item == NULL)
return svn_error_createf(SVN_ERR_FS_INDEX_INCONSISTENT,
NULL,
_("p2l index entry not found for "
"PHYS o%s:s%ld returned by "
"l2p index for LOG r%ld:i%ld"),
apr_off_t_toa(scratch_pool, offset),
(long)sub_item, revision, (long)k);
if (!svn_fs_x__id_eq(&l2p_item, p2l_item))
return svn_error_createf(SVN_ERR_FS_INDEX_INCONSISTENT,
NULL,
_("p2l index info LOG r%ld:i%ld"
" does not match "
"l2p index for LOG r%ld:i%ld"),
svn_fs_x__get_revnum(p2l_item->change_set),
(long)p2l_item->number, revision,
(long)k);
svn_pool_clear(iterpool);
}
if (cancel_func)
SVN_ERR(cancel_func(cancel_baton));
}
svn_pool_destroy(iterpool);
SVN_ERR(svn_fs_x__close_revision_file(rev_file));
return SVN_NO_ERROR;
}
/* Verify that for all phys-to-log index entries for revisions START to
* START + COUNT-1 in FS there is a consistent entry in the log-to-phys
* index. If given, invoke CANCEL_FUNC with CANCEL_BATON at regular
* intervals. Use SCRATCH_POOL for temporary allocations.
*
* Please note that we can only check on pack / rev file granularity and
* must only be called for a single rev / pack file.
*/
static svn_error_t *
compare_p2l_to_l2p_index(svn_fs_t *fs,
svn_revnum_t start,
svn_revnum_t count,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_fs_x__data_t *ffd = fs->fsap_data;
apr_pool_t *iterpool = svn_pool_create(scratch_pool);
apr_pool_t *iterpool2 = svn_pool_create(scratch_pool);
apr_off_t max_offset;
apr_off_t offset = 0;
/* common file access structure */
svn_fs_x__revision_file_t *rev_file;
SVN_ERR(svn_fs_x__open_pack_or_rev_file(&rev_file, fs, start, scratch_pool,
iterpool));
/* get the size of the rev / pack file as covered by the P2L index */
SVN_ERR(svn_fs_x__p2l_get_max_offset(&max_offset, fs, rev_file, start,
scratch_pool));
/* for all offsets in the file, get the P2L index entries and check
them against the L2P index */
for (offset = 0; offset < max_offset; )
{
apr_array_header_t *entries;
svn_fs_x__p2l_entry_t *last_entry;
int i;
svn_pool_clear(iterpool);
/* get all entries for the current block */
SVN_ERR(svn_fs_x__p2l_index_lookup(&entries, fs, rev_file, start,
offset, ffd->p2l_page_size,
iterpool, iterpool));
if (entries->nelts == 0)
return svn_error_createf(SVN_ERR_FS_INDEX_CORRUPTION,
NULL,
_("p2l does not cover offset %s"
" for revision %ld"),
apr_off_t_toa(scratch_pool, offset), start);
/* process all entries (and later continue with the next block) */
last_entry
= &APR_ARRAY_IDX(entries, entries->nelts-1, svn_fs_x__p2l_entry_t);
offset = last_entry->offset + last_entry->size;
for (i = 0; i < entries->nelts; ++i)
{
apr_uint32_t k;
svn_fs_x__p2l_entry_t *entry
= &APR_ARRAY_IDX(entries, i, svn_fs_x__p2l_entry_t);
/* check all sub-items for consist entries in the L2P index */
for (k = 0; k < entry->item_count; ++k)
{
apr_off_t l2p_offset;
apr_uint32_t sub_item;
svn_fs_x__id_t *p2l_item = &entry->items[k];
svn_revnum_t revision
= svn_fs_x__get_revnum(p2l_item->change_set);
svn_pool_clear(iterpool2);
SVN_ERR(svn_fs_x__item_offset(&l2p_offset, &sub_item, fs,
rev_file, p2l_item, iterpool2));
if (sub_item != k || l2p_offset != entry->offset)
return svn_error_createf(SVN_ERR_FS_INDEX_INCONSISTENT,
NULL,
_("l2p index entry PHYS o%s:s%ld "
"does not match p2l index value "
"LOG r%ld:i%ld for PHYS o%s:s%ld"),
apr_off_t_toa(scratch_pool,
l2p_offset),
(long)sub_item,
revision,
(long)p2l_item->number,
apr_off_t_toa(scratch_pool,
entry->offset),
(long)k);
}
}
if (cancel_func)
SVN_ERR(cancel_func(cancel_baton));
}
svn_pool_destroy(iterpool2);
svn_pool_destroy(iterpool);
SVN_ERR(svn_fs_x__close_revision_file(rev_file));
return SVN_NO_ERROR;
}
/* Items smaller than this can be read at once into a buffer and directly
* be checksummed. Larger items require stream processing.
* Must be a multiple of 8. */
#define STREAM_THRESHOLD 4096
/* Verify that the next SIZE bytes read from FILE are NUL. SIZE must not
* exceed STREAM_THRESHOLD. Use SCRATCH_POOL for temporary allocations.
*/
static svn_error_t *
expect_buffer_nul(apr_file_t *file,
apr_off_t size,
apr_pool_t *scratch_pool)
{
union
{
unsigned char buffer[STREAM_THRESHOLD];
apr_uint64_t chunks[STREAM_THRESHOLD / sizeof(apr_uint64_t)];
} data;
apr_size_t i;
SVN_ERR_ASSERT(size <= STREAM_THRESHOLD);
/* read the whole data block; error out on failure */
data.chunks[(size - 1)/ sizeof(apr_uint64_t)] = 0;
SVN_ERR(svn_io_file_read_full2(file, data.buffer, size, NULL, NULL,
scratch_pool));
/* chunky check */
for (i = 0; i < size / sizeof(apr_uint64_t); ++i)
if (data.chunks[i] != 0)
break;
/* byte-wise check upon mismatch or at the end of the block */
for (i *= sizeof(apr_uint64_t); i < size; ++i)
if (data.buffer[i] != 0)
{
const char *file_name;
apr_off_t offset;
SVN_ERR(svn_io_file_name_get(&file_name, file, scratch_pool));
SVN_ERR(svn_fs_x__get_file_offset(&offset, file, scratch_pool));
offset -= size - i;
return svn_error_createf(SVN_ERR_FS_CORRUPT, NULL,
_("Empty section in file %s contains "
"non-NUL data at offset %s"),
file_name,
apr_off_t_toa(scratch_pool, offset));
}
return SVN_NO_ERROR;
}
/* Verify that the next SIZE bytes read from FILE are NUL.
* Use SCRATCH_POOL for temporary allocations.
*/
static svn_error_t *
read_all_nul(apr_file_t *file,
apr_off_t size,
apr_pool_t *scratch_pool)
{
for (; size >= STREAM_THRESHOLD; size -= STREAM_THRESHOLD)
SVN_ERR(expect_buffer_nul(file, STREAM_THRESHOLD, scratch_pool));
if (size)
SVN_ERR(expect_buffer_nul(file, size, scratch_pool));
return SVN_NO_ERROR;
}
/* Compare the ACTUAL checksum with the one expected by ENTRY.
* Return an error in case of mismatch. Use the name of FILE
* in error message. Allocate temporary data in SCRATCH_POOL.
*/
static svn_error_t *
expected_checksum(apr_file_t *file,
svn_fs_x__p2l_entry_t *entry,
apr_uint32_t actual,
apr_pool_t *scratch_pool)
{
if (actual != entry->fnv1_checksum)
{
const char *file_name;
SVN_ERR(svn_io_file_name_get(&file_name, file, scratch_pool));
SVN_ERR(svn_io_file_name_get(&file_name, file, scratch_pool));
return svn_error_createf(SVN_ERR_FS_CORRUPT, NULL,
_("Checksum mismatch in item at offset %s of "
"length %s bytes in file %s"),
apr_off_t_toa(scratch_pool, entry->offset),
apr_off_t_toa(scratch_pool, entry->size),
file_name);
}
return SVN_NO_ERROR;
}
/* Verify that the FNV checksum over the next ENTRY->SIZE bytes read
* from FILE will match ENTRY's expected checksum. SIZE must not
* exceed STREAM_THRESHOLD. Use SCRATCH_POOL for temporary allocations.
*/
static svn_error_t *
expected_buffered_checksum(apr_file_t *file,
svn_fs_x__p2l_entry_t *entry,
apr_pool_t *scratch_pool)
{
unsigned char buffer[STREAM_THRESHOLD];
SVN_ERR_ASSERT(entry->size <= STREAM_THRESHOLD);
SVN_ERR(svn_io_file_read_full2(file, buffer, (apr_size_t)entry->size,
NULL, NULL, scratch_pool));
SVN_ERR(expected_checksum(file, entry,
svn__fnv1a_32x4(buffer, (apr_size_t)entry->size),
scratch_pool));
return SVN_NO_ERROR;
}
/* Verify that the FNV checksum over the next ENTRY->SIZE bytes read from
* FILE will match ENTRY's expected checksum.
* Use SCRATCH_POOL for temporary allocations.
*/
static svn_error_t *
expected_streamed_checksum(apr_file_t *file,
svn_fs_x__p2l_entry_t *entry,
apr_pool_t *scratch_pool)
{
unsigned char buffer[STREAM_THRESHOLD];
svn_checksum_t *checksum;
svn_checksum_ctx_t *context
= svn_checksum_ctx_create(svn_checksum_fnv1a_32x4, scratch_pool);
apr_off_t size = entry->size;
while (size > 0)
{
apr_size_t to_read = size > sizeof(buffer)
? sizeof(buffer)
: (apr_size_t)size;
SVN_ERR(svn_io_file_read_full2(file, buffer, to_read, NULL, NULL,
scratch_pool));
SVN_ERR(svn_checksum_update(context, buffer, to_read));
size -= to_read;
}
SVN_ERR(svn_checksum_final(&checksum, context, scratch_pool));
SVN_ERR(expected_checksum(file, entry,
ntohl(*(const apr_uint32_t *)checksum->digest),
scratch_pool));
return SVN_NO_ERROR;
}
/* Verify that for all phys-to-log index entries for revisions START to
* START + COUNT-1 in FS match the actual pack / rev file contents.
* If given, invoke CANCEL_FUNC with CANCEL_BATON at regular intervals.
* Use SCRATCH_POOL for temporary allocations.
*
* Please note that we can only check on pack / rev file granularity and
* must only be called for a single rev / pack file.
*/
static svn_error_t *
compare_p2l_to_rev(svn_fs_t *fs,
svn_revnum_t start,
svn_revnum_t count,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_fs_x__data_t *ffd = fs->fsap_data;
apr_pool_t *iterpool = svn_pool_create(scratch_pool);
apr_off_t max_offset;
apr_off_t offset = 0;
svn_fs_x__revision_file_t *rev_file;
/* open the pack / rev file that is covered by the p2l index */
SVN_ERR(svn_fs_x__open_pack_or_rev_file(&rev_file, fs, start, scratch_pool,
iterpool));
/* check file size vs. range covered by index */
SVN_ERR(svn_fs_x__auto_read_footer(rev_file));
SVN_ERR(svn_fs_x__p2l_get_max_offset(&max_offset, fs, rev_file, start,
scratch_pool));
if (rev_file->l2p_offset != max_offset)
return svn_error_createf(SVN_ERR_FS_INDEX_INCONSISTENT, NULL,
_("File size of %s for revision r%ld does "
"not match p2l index size of %s"),
apr_off_t_toa(scratch_pool,
rev_file->l2p_offset),
start,
apr_off_t_toa(scratch_pool,
max_offset));
SVN_ERR(svn_io_file_aligned_seek(rev_file->file, ffd->block_size, NULL, 0,
scratch_pool));
/* for all offsets in the file, get the P2L index entries and check
them against the L2P index */
for (offset = 0; offset < max_offset; )
{
apr_array_header_t *entries;
int i;
svn_pool_clear(iterpool);
/* get all entries for the current block */
SVN_ERR(svn_fs_x__p2l_index_lookup(&entries, fs, rev_file, start,
offset, ffd->p2l_page_size,
iterpool, iterpool));
/* The above might have moved the file pointer.
* Ensure we actually start reading at OFFSET. */
SVN_ERR(svn_io_file_aligned_seek(rev_file->file, ffd->block_size,
NULL, offset, iterpool));
/* process all entries (and later continue with the next block) */
for (i = 0; i < entries->nelts; ++i)
{
svn_fs_x__p2l_entry_t *entry
= &APR_ARRAY_IDX(entries, i, svn_fs_x__p2l_entry_t);
/* skip bits we previously checked */
if (i == 0 && entry->offset < offset)
continue;
/* skip zero-sized entries */
if (entry->size == 0)
continue;
/* p2l index must cover all rev / pack file offsets exactly once */
if (entry->offset != offset)
return svn_error_createf(SVN_ERR_FS_INDEX_INCONSISTENT,
NULL,
_("p2l index entry for revision r%ld"
" is non-contiguous between offsets "
" %s and %s"),
start,
apr_off_t_toa(scratch_pool, offset),
apr_off_t_toa(scratch_pool,
entry->offset));
/* empty sections must contain NUL bytes only */
if (entry->type == SVN_FS_X__ITEM_TYPE_UNUSED)
{
/* skip filler entry at the end of the p2l index */
if (entry->offset != max_offset)
SVN_ERR(read_all_nul(rev_file->file, entry->size, iterpool));
}
else
{
if (entry->size < STREAM_THRESHOLD)
SVN_ERR(expected_buffered_checksum(rev_file->file, entry,
iterpool));
else
SVN_ERR(expected_streamed_checksum(rev_file->file, entry,
iterpool));
}
/* advance offset */
offset += entry->size;
}
if (cancel_func)
SVN_ERR(cancel_func(cancel_baton));
}
svn_pool_destroy(iterpool);
return SVN_NO_ERROR;
}
/* Verify that the revprops of the revisions START to END in FS can be
* accessed. Invoke CANCEL_FUNC with CANCEL_BATON at regular intervals.
*
* The values of START and END have already been auto-selected and
* verified.
*/
static svn_error_t *
verify_revprops(svn_fs_t *fs,
svn_revnum_t start,
svn_revnum_t end,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_revnum_t revision;
apr_pool_t *iterpool = svn_pool_create(scratch_pool);
for (revision = start; revision < end; ++revision)
{
svn_string_t *date;
apr_time_t timetemp;
svn_pool_clear(iterpool);
/* Access the svn:date revprop.
* This implies parsing all revprops for that revision. */
SVN_ERR(svn_fs_x__revision_prop(&date, fs, revision,
SVN_PROP_REVISION_DATE,
iterpool, iterpool));
/* The time stamp is the only revprop that, if given, needs to
* have a valid content. */
if (date)
SVN_ERR(svn_time_from_cstring(&timetemp, date->data, iterpool));
if (cancel_func)
SVN_ERR(cancel_func(cancel_baton));
}
svn_pool_destroy(iterpool);
return SVN_NO_ERROR;
}
/* Verify that on-disk representation has not been tempered with (in a way
* that leaves the repository in a corrupted state). This compares log-to-
* phys with phys-to-log indexes, verifies the low-level checksums and
* checks that all revprops are available. The function signature is
* similar to svn_fs_x__verify.
*
* The values of START and END have already been auto-selected and
* verified.
*/
static svn_error_t *
verify_metadata_consistency(svn_fs_t *fs,
svn_revnum_t start,
svn_revnum_t end,
svn_fs_progress_notify_func_t notify_func,
void *notify_baton,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_error_t *err;
svn_fs_x__data_t *ffd = fs->fsap_data;
svn_revnum_t revision, next_revision;
apr_pool_t *iterpool = svn_pool_create(scratch_pool);
for (revision = start; revision <= end; revision = next_revision)
{
svn_revnum_t count = svn_fs_x__packed_base_rev(fs, revision);
svn_revnum_t pack_start = count;
svn_revnum_t pack_end = pack_start + svn_fs_x__pack_size(fs, revision);
svn_pool_clear(iterpool);
if (notify_func && (pack_start % ffd->max_files_per_dir == 0))
notify_func(pack_start, notify_baton, iterpool);
/* Check for external corruption to the indexes. */
err = verify_index_checksums(fs, pack_start, cancel_func,
cancel_baton, iterpool);
/* two-way index check */
if (!err)
err = compare_l2p_to_p2l_index(fs, pack_start, pack_end - pack_start,
cancel_func, cancel_baton, iterpool);
if (!err)
err = compare_p2l_to_l2p_index(fs, pack_start, pack_end - pack_start,
cancel_func, cancel_baton, iterpool);
/* verify in-index checksums and types vs. actual rev / pack files */
if (!err)
err = compare_p2l_to_rev(fs, pack_start, pack_end - pack_start,
cancel_func, cancel_baton, iterpool);
/* ensure that revprops are available and accessible */
if (!err)
err = verify_revprops(fs, pack_start, pack_end,
cancel_func, cancel_baton, iterpool);
/* concurrent packing is one of the reasons why verification may fail.
Make sure, we operate on up-to-date information. */
if (err)
SVN_ERR(svn_fs_x__read_min_unpacked_rev(&ffd->min_unpacked_rev,
fs, scratch_pool));
/* retry the whole shard if it got packed in the meantime */
if (err && count != svn_fs_x__pack_size(fs, revision))
{
svn_error_clear(err);
/* We could simply assign revision here but the code below is
more intuitive to maintainers. */
next_revision = svn_fs_x__packed_base_rev(fs, revision);
}
else
{
SVN_ERR(err);
next_revision = pack_end;
}
}
svn_pool_destroy(iterpool);
return SVN_NO_ERROR;
}
svn_error_t *
svn_fs_x__verify(svn_fs_t *fs,
svn_revnum_t start,
svn_revnum_t end,
svn_fs_progress_notify_func_t notify_func,
void *notify_baton,
svn_cancel_func_t cancel_func,
void *cancel_baton,
apr_pool_t *scratch_pool)
{
svn_fs_x__data_t *ffd = fs->fsap_data;
svn_revnum_t youngest = ffd->youngest_rev_cache; /* cache is current */
/* Input validation. */
if (! SVN_IS_VALID_REVNUM(start))
start = 0;
if (! SVN_IS_VALID_REVNUM(end))
end = youngest;
SVN_ERR(svn_fs_x__ensure_revision_exists(start, fs, scratch_pool));
SVN_ERR(svn_fs_x__ensure_revision_exists(end, fs, scratch_pool));
/* log/phys index consistency. We need to check them first to make
sure we can access the rev / pack files in format7. */
SVN_ERR(verify_metadata_consistency(fs, start, end,
notify_func, notify_baton,
cancel_func, cancel_baton,
scratch_pool));
/* rep cache consistency */
SVN_ERR(verify_rep_cache(fs, start, end, notify_func, notify_baton,
cancel_func, cancel_baton, scratch_pool));
return SVN_NO_ERROR;
}
|