-
Notifications
You must be signed in to change notification settings - Fork 1.7k
/
btr0btr.cc
5330 lines (4403 loc) · 149 KB
/
btr0btr.cc
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
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*****************************************************************************
Copyright (c) 1994, 2016, Oracle and/or its affiliates. All Rights Reserved.
Copyright (c) 2012, Facebook Inc.
Copyright (c) 2014, 2017, MariaDB Corporation.
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; version 2 of the License.
This program is distributed in the hope that it will 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 to the Free Software Foundation, Inc.,
51 Franklin Street, Suite 500, Boston, MA 02110-1335 USA
*****************************************************************************/
/**************************************************//**
@file btr/btr0btr.cc
The B-tree
Created 6/2/1994 Heikki Tuuri
*******************************************************/
#include "btr0btr.h"
#ifdef UNIV_NONINL
#include "btr0btr.ic"
#endif
#include "fsp0fsp.h"
#include "page0page.h"
#include "page0zip.h"
#ifndef UNIV_HOTBACKUP
#include "btr0cur.h"
#include "btr0sea.h"
#include "btr0pcur.h"
#include "btr0defragment.h"
#include "rem0cmp.h"
#include "lock0lock.h"
#include "ibuf0ibuf.h"
#include "trx0trx.h"
#include "srv0mon.h"
/**************************************************************//**
Checks if the page in the cursor can be merged with given page.
If necessary, re-organize the merge_page.
@return TRUE if possible to merge. */
UNIV_INTERN
ibool
btr_can_merge_with_page(
/*====================*/
btr_cur_t* cursor, /*!< in: cursor on the page to merge */
ulint page_no, /*!< in: a sibling page */
buf_block_t** merge_block, /*!< out: the merge block */
mtr_t* mtr); /*!< in: mini-transaction */
#endif /* UNIV_HOTBACKUP */
/**************************************************************//**
Report that an index page is corrupted. */
UNIV_INTERN
void
btr_corruption_report(
/*==================*/
const buf_block_t* block, /*!< in: corrupted block */
const dict_index_t* index) /*!< in: index tree */
{
fprintf(stderr, "InnoDB: flag mismatch in space %u page %u"
" index %s of table %s\n",
(unsigned) buf_block_get_space(block),
(unsigned) buf_block_get_page_no(block),
index->name, index->table_name);
if (block->page.zip.data) {
buf_page_print(block->page.zip.data,
buf_block_get_zip_size(block),
BUF_PAGE_PRINT_NO_CRASH);
}
buf_page_print(buf_nonnull_block_get_frame(block), 0, 0);
}
#ifndef UNIV_HOTBACKUP
#ifdef UNIV_BLOB_DEBUG
# include "srv0srv.h"
# include "ut0rbt.h"
/** TRUE when messages about index->blobs modification are enabled. */
static ibool btr_blob_dbg_msg;
/** Issue a message about an operation on index->blobs.
@param op operation
@param b the entry being subjected to the operation
@param ctx the context of the operation */
#define btr_blob_dbg_msg_issue(op, b, ctx) \
fprintf(stderr, op " %u:%u:%u->%u %s(%u,%u,%u)\n", \
(b)->ref_page_no, (b)->ref_heap_no, \
(b)->ref_field_no, (b)->blob_page_no, ctx, \
(b)->owner, (b)->always_owner, (b)->del)
/** Insert to index->blobs a reference to an off-page column.
@param index the index tree
@param b the reference
@param ctx context (for logging) */
UNIV_INTERN
void
btr_blob_dbg_rbt_insert(
/*====================*/
dict_index_t* index, /*!< in/out: index tree */
const btr_blob_dbg_t* b, /*!< in: the reference */
const char* ctx) /*!< in: context (for logging) */
{
if (btr_blob_dbg_msg) {
btr_blob_dbg_msg_issue("insert", b, ctx);
}
mutex_enter(&index->blobs_mutex);
rbt_insert(index->blobs, b, b);
mutex_exit(&index->blobs_mutex);
}
/** Remove from index->blobs a reference to an off-page column.
@param index the index tree
@param b the reference
@param ctx context (for logging) */
UNIV_INTERN
void
btr_blob_dbg_rbt_delete(
/*====================*/
dict_index_t* index, /*!< in/out: index tree */
const btr_blob_dbg_t* b, /*!< in: the reference */
const char* ctx) /*!< in: context (for logging) */
{
if (btr_blob_dbg_msg) {
btr_blob_dbg_msg_issue("delete", b, ctx);
}
mutex_enter(&index->blobs_mutex);
ut_a(rbt_delete(index->blobs, b));
mutex_exit(&index->blobs_mutex);
}
/**************************************************************//**
Comparator for items (btr_blob_dbg_t) in index->blobs.
The key in index->blobs is (ref_page_no, ref_heap_no, ref_field_no).
@return negative, 0 or positive if *a<*b, *a=*b, *a>*b */
static
int
btr_blob_dbg_cmp(
/*=============*/
const void* a, /*!< in: first btr_blob_dbg_t to compare */
const void* b) /*!< in: second btr_blob_dbg_t to compare */
{
const btr_blob_dbg_t* aa = static_cast<const btr_blob_dbg_t*>(a);
const btr_blob_dbg_t* bb = static_cast<const btr_blob_dbg_t*>(b);
ut_ad(aa != NULL);
ut_ad(bb != NULL);
if (aa->ref_page_no != bb->ref_page_no) {
return(aa->ref_page_no < bb->ref_page_no ? -1 : 1);
}
if (aa->ref_heap_no != bb->ref_heap_no) {
return(aa->ref_heap_no < bb->ref_heap_no ? -1 : 1);
}
if (aa->ref_field_no != bb->ref_field_no) {
return(aa->ref_field_no < bb->ref_field_no ? -1 : 1);
}
return(0);
}
/**************************************************************//**
Add a reference to an off-page column to the index->blobs map. */
UNIV_INTERN
void
btr_blob_dbg_add_blob(
/*==================*/
const rec_t* rec, /*!< in: clustered index record */
ulint field_no, /*!< in: off-page column number */
ulint page_no, /*!< in: start page of the column */
dict_index_t* index, /*!< in/out: index tree */
const char* ctx) /*!< in: context (for logging) */
{
btr_blob_dbg_t b;
const page_t* page = page_align(rec);
ut_a(index->blobs);
b.blob_page_no = page_no;
b.ref_page_no = page_get_page_no(page);
b.ref_heap_no = page_rec_get_heap_no(rec);
b.ref_field_no = field_no;
ut_a(b.ref_field_no >= index->n_uniq);
b.always_owner = b.owner = TRUE;
b.del = FALSE;
ut_a(!rec_get_deleted_flag(rec, page_is_comp(page)));
btr_blob_dbg_rbt_insert(index, &b, ctx);
}
/**************************************************************//**
Add to index->blobs any references to off-page columns from a record.
@return number of references added */
UNIV_INTERN
ulint
btr_blob_dbg_add_rec(
/*=================*/
const rec_t* rec, /*!< in: record */
dict_index_t* index, /*!< in/out: index */
const ulint* offsets,/*!< in: offsets */
const char* ctx) /*!< in: context (for logging) */
{
ulint count = 0;
ulint i;
btr_blob_dbg_t b;
ibool del;
ut_ad(rec_offs_validate(rec, index, offsets));
if (!rec_offs_any_extern(offsets)) {
return(0);
}
b.ref_page_no = page_get_page_no(page_align(rec));
b.ref_heap_no = page_rec_get_heap_no(rec);
del = (rec_get_deleted_flag(rec, rec_offs_comp(offsets)) != 0);
for (i = 0; i < rec_offs_n_fields(offsets); i++) {
if (rec_offs_nth_extern(offsets, i)) {
ulint len;
const byte* field_ref = rec_get_nth_field(
rec, offsets, i, &len);
ut_a(len != UNIV_SQL_NULL);
ut_a(len >= BTR_EXTERN_FIELD_REF_SIZE);
field_ref += len - BTR_EXTERN_FIELD_REF_SIZE;
if (!memcmp(field_ref, field_ref_zero,
BTR_EXTERN_FIELD_REF_SIZE)) {
/* the column has not been stored yet */
continue;
}
b.ref_field_no = i;
b.blob_page_no = mach_read_from_4(
field_ref + BTR_EXTERN_PAGE_NO);
ut_a(b.ref_field_no >= index->n_uniq);
b.always_owner = b.owner
= !(field_ref[BTR_EXTERN_LEN]
& BTR_EXTERN_OWNER_FLAG);
b.del = del;
btr_blob_dbg_rbt_insert(index, &b, ctx);
count++;
}
}
return(count);
}
/**************************************************************//**
Display the references to off-page columns.
This function is to be called from a debugger,
for example when a breakpoint on ut_dbg_assertion_failed is hit. */
UNIV_INTERN
void
btr_blob_dbg_print(
/*===============*/
const dict_index_t* index) /*!< in: index tree */
{
const ib_rbt_node_t* node;
if (!index->blobs) {
return;
}
/* We intentionally do not acquire index->blobs_mutex here.
This function is to be called from a debugger, and the caller
should make sure that the index->blobs_mutex is held. */
for (node = rbt_first(index->blobs);
node != NULL; node = rbt_next(index->blobs, node)) {
const btr_blob_dbg_t* b
= rbt_value(btr_blob_dbg_t, node);
fprintf(stderr, "%u:%u:%u->%u%s%s%s\n",
b->ref_page_no, b->ref_heap_no, b->ref_field_no,
b->blob_page_no,
b->owner ? "" : "(disowned)",
b->always_owner ? "" : "(has disowned)",
b->del ? "(deleted)" : "");
}
}
/**************************************************************//**
Remove from index->blobs any references to off-page columns from a record.
@return number of references removed */
UNIV_INTERN
ulint
btr_blob_dbg_remove_rec(
/*====================*/
const rec_t* rec, /*!< in: record */
dict_index_t* index, /*!< in/out: index */
const ulint* offsets,/*!< in: offsets */
const char* ctx) /*!< in: context (for logging) */
{
ulint i;
ulint count = 0;
btr_blob_dbg_t b;
ut_ad(rec_offs_validate(rec, index, offsets));
if (!rec_offs_any_extern(offsets)) {
return(0);
}
b.ref_page_no = page_get_page_no(page_align(rec));
b.ref_heap_no = page_rec_get_heap_no(rec);
for (i = 0; i < rec_offs_n_fields(offsets); i++) {
if (rec_offs_nth_extern(offsets, i)) {
ulint len;
const byte* field_ref = rec_get_nth_field(
rec, offsets, i, &len);
ut_a(len != UNIV_SQL_NULL);
ut_a(len >= BTR_EXTERN_FIELD_REF_SIZE);
field_ref += len - BTR_EXTERN_FIELD_REF_SIZE;
b.ref_field_no = i;
b.blob_page_no = mach_read_from_4(
field_ref + BTR_EXTERN_PAGE_NO);
switch (b.blob_page_no) {
case 0:
/* The column has not been stored yet.
The BLOB pointer must be all zero.
There cannot be a BLOB starting at
page 0, because page 0 is reserved for
the tablespace header. */
ut_a(!memcmp(field_ref, field_ref_zero,
BTR_EXTERN_FIELD_REF_SIZE));
/* fall through */
case FIL_NULL:
/* the column has been freed already */
continue;
}
btr_blob_dbg_rbt_delete(index, &b, ctx);
count++;
}
}
return(count);
}
/**************************************************************//**
Check that there are no references to off-page columns from or to
the given page. Invoked when freeing or clearing a page.
@return TRUE when no orphan references exist */
UNIV_INTERN
ibool
btr_blob_dbg_is_empty(
/*==================*/
dict_index_t* index, /*!< in: index */
ulint page_no) /*!< in: page number */
{
const ib_rbt_node_t* node;
ibool success = TRUE;
if (!index->blobs) {
return(success);
}
mutex_enter(&index->blobs_mutex);
for (node = rbt_first(index->blobs);
node != NULL; node = rbt_next(index->blobs, node)) {
const btr_blob_dbg_t* b
= rbt_value(btr_blob_dbg_t, node);
if (b->ref_page_no != page_no && b->blob_page_no != page_no) {
continue;
}
fprintf(stderr,
"InnoDB: orphan BLOB ref%s%s%s %u:%u:%u->%u\n",
b->owner ? "" : "(disowned)",
b->always_owner ? "" : "(has disowned)",
b->del ? "(deleted)" : "",
b->ref_page_no, b->ref_heap_no, b->ref_field_no,
b->blob_page_no);
if (b->blob_page_no != page_no || b->owner || !b->del) {
success = FALSE;
}
}
mutex_exit(&index->blobs_mutex);
return(success);
}
/**************************************************************//**
Count and process all references to off-page columns on a page.
@return number of references processed */
UNIV_INTERN
ulint
btr_blob_dbg_op(
/*============*/
const page_t* page, /*!< in: B-tree leaf page */
const rec_t* rec, /*!< in: record to start from
(NULL to process the whole page) */
dict_index_t* index, /*!< in/out: index */
const char* ctx, /*!< in: context (for logging) */
const btr_blob_dbg_op_f op) /*!< in: operation on records */
{
ulint count = 0;
mem_heap_t* heap = NULL;
ulint offsets_[REC_OFFS_NORMAL_SIZE];
ulint* offsets = offsets_;
rec_offs_init(offsets_);
ut_a(fil_page_get_type(page) == FIL_PAGE_INDEX);
ut_a(!rec || page_align(rec) == page);
if (!index->blobs || !page_is_leaf(page)
|| !dict_index_is_clust(index)) {
return(0);
}
if (rec == NULL) {
rec = page_get_infimum_rec(page);
}
do {
offsets = rec_get_offsets(rec, index, offsets,
ULINT_UNDEFINED, &heap);
count += op(rec, index, offsets, ctx);
rec = page_rec_get_next_const(rec);
} while (!page_rec_is_supremum(rec));
if (heap) {
mem_heap_free(heap);
}
return(count);
}
/**************************************************************//**
Count and add to index->blobs any references to off-page columns
from records on a page.
@return number of references added */
UNIV_INTERN
ulint
btr_blob_dbg_add(
/*=============*/
const page_t* page, /*!< in: rewritten page */
dict_index_t* index, /*!< in/out: index */
const char* ctx) /*!< in: context (for logging) */
{
btr_blob_dbg_assert_empty(index, page_get_page_no(page));
return(btr_blob_dbg_op(page, NULL, index, ctx, btr_blob_dbg_add_rec));
}
/**************************************************************//**
Count and remove from index->blobs any references to off-page columns
from records on a page.
Used when reorganizing a page, before copying the records.
@return number of references removed */
UNIV_INTERN
ulint
btr_blob_dbg_remove(
/*================*/
const page_t* page, /*!< in: b-tree page */
dict_index_t* index, /*!< in/out: index */
const char* ctx) /*!< in: context (for logging) */
{
ulint count;
count = btr_blob_dbg_op(page, NULL, index, ctx,
btr_blob_dbg_remove_rec);
/* Check that no references exist. */
btr_blob_dbg_assert_empty(index, page_get_page_no(page));
return(count);
}
/**************************************************************//**
Restore in index->blobs any references to off-page columns
Used when page reorganize fails due to compressed page overflow. */
UNIV_INTERN
void
btr_blob_dbg_restore(
/*=================*/
const page_t* npage, /*!< in: page that failed to compress */
const page_t* page, /*!< in: copy of original page */
dict_index_t* index, /*!< in/out: index */
const char* ctx) /*!< in: context (for logging) */
{
ulint removed;
ulint added;
ut_a(page_get_page_no(npage) == page_get_page_no(page));
ut_a(page_get_space_id(npage) == page_get_space_id(page));
removed = btr_blob_dbg_remove(npage, index, ctx);
added = btr_blob_dbg_add(page, index, ctx);
ut_a(added == removed);
}
/**************************************************************//**
Modify the 'deleted' flag of a record. */
UNIV_INTERN
void
btr_blob_dbg_set_deleted_flag(
/*==========================*/
const rec_t* rec, /*!< in: record */
dict_index_t* index, /*!< in/out: index */
const ulint* offsets,/*!< in: rec_get_offs(rec, index) */
ibool del) /*!< in: TRUE=deleted, FALSE=exists */
{
const ib_rbt_node_t* node;
btr_blob_dbg_t b;
btr_blob_dbg_t* c;
ulint i;
ut_ad(rec_offs_validate(rec, index, offsets));
ut_a(dict_index_is_clust(index));
ut_a(del == !!del);/* must be FALSE==0 or TRUE==1 */
if (!rec_offs_any_extern(offsets) || !index->blobs) {
return;
}
b.ref_page_no = page_get_page_no(page_align(rec));
b.ref_heap_no = page_rec_get_heap_no(rec);
for (i = 0; i < rec_offs_n_fields(offsets); i++) {
if (rec_offs_nth_extern(offsets, i)) {
ulint len;
const byte* field_ref = rec_get_nth_field(
rec, offsets, i, &len);
ut_a(len != UNIV_SQL_NULL);
ut_a(len >= BTR_EXTERN_FIELD_REF_SIZE);
field_ref += len - BTR_EXTERN_FIELD_REF_SIZE;
b.ref_field_no = i;
b.blob_page_no = mach_read_from_4(
field_ref + BTR_EXTERN_PAGE_NO);
switch (b.blob_page_no) {
case 0:
ut_a(memcmp(field_ref, field_ref_zero,
BTR_EXTERN_FIELD_REF_SIZE));
/* page number 0 is for the
page allocation bitmap */
case FIL_NULL:
/* the column has been freed already */
ut_error;
}
mutex_enter(&index->blobs_mutex);
node = rbt_lookup(index->blobs, &b);
ut_a(node);
c = rbt_value(btr_blob_dbg_t, node);
/* The flag should be modified. */
c->del = del;
if (btr_blob_dbg_msg) {
b = *c;
mutex_exit(&index->blobs_mutex);
btr_blob_dbg_msg_issue("del_mk", &b, "");
} else {
mutex_exit(&index->blobs_mutex);
}
}
}
}
/**************************************************************//**
Change the ownership of an off-page column. */
UNIV_INTERN
void
btr_blob_dbg_owner(
/*===============*/
const rec_t* rec, /*!< in: record */
dict_index_t* index, /*!< in/out: index */
const ulint* offsets,/*!< in: rec_get_offs(rec, index) */
ulint i, /*!< in: ith field in rec */
ibool own) /*!< in: TRUE=owned, FALSE=disowned */
{
const ib_rbt_node_t* node;
btr_blob_dbg_t b;
const byte* field_ref;
ulint len;
ut_ad(rec_offs_validate(rec, index, offsets));
ut_a(rec_offs_nth_extern(offsets, i));
field_ref = rec_get_nth_field(rec, offsets, i, &len);
ut_a(len != UNIV_SQL_NULL);
ut_a(len >= BTR_EXTERN_FIELD_REF_SIZE);
field_ref += len - BTR_EXTERN_FIELD_REF_SIZE;
b.ref_page_no = page_get_page_no(page_align(rec));
b.ref_heap_no = page_rec_get_heap_no(rec);
b.ref_field_no = i;
b.owner = !(field_ref[BTR_EXTERN_LEN] & BTR_EXTERN_OWNER_FLAG);
b.blob_page_no = mach_read_from_4(field_ref + BTR_EXTERN_PAGE_NO);
ut_a(b.owner == own);
mutex_enter(&index->blobs_mutex);
node = rbt_lookup(index->blobs, &b);
/* row_ins_clust_index_entry_by_modify() invokes
btr_cur_unmark_extern_fields() also for the newly inserted
references, which are all zero bytes until the columns are stored.
The node lookup must fail if and only if that is the case. */
ut_a(!memcmp(field_ref, field_ref_zero, BTR_EXTERN_FIELD_REF_SIZE)
== !node);
if (node) {
btr_blob_dbg_t* c = rbt_value(btr_blob_dbg_t, node);
/* Some code sets ownership from TRUE to TRUE.
We do not allow changing ownership from FALSE to FALSE. */
ut_a(own || c->owner);
c->owner = own;
if (!own) {
c->always_owner = FALSE;
}
}
mutex_exit(&index->blobs_mutex);
}
#endif /* UNIV_BLOB_DEBUG */
/*
Latching strategy of the InnoDB B-tree
--------------------------------------
A tree latch protects all non-leaf nodes of the tree. Each node of a tree
also has a latch of its own.
A B-tree operation normally first acquires an S-latch on the tree. It
searches down the tree and releases the tree latch when it has the
leaf node latch. To save CPU time we do not acquire any latch on
non-leaf nodes of the tree during a search, those pages are only bufferfixed.
If an operation needs to restructure the tree, it acquires an X-latch on
the tree before searching to a leaf node. If it needs, for example, to
split a leaf,
(1) InnoDB decides the split point in the leaf,
(2) allocates a new page,
(3) inserts the appropriate node pointer to the first non-leaf level,
(4) releases the tree X-latch,
(5) and then moves records from the leaf to the new allocated page.
Node pointers
-------------
Leaf pages of a B-tree contain the index records stored in the
tree. On levels n > 0 we store 'node pointers' to pages on level
n - 1. For each page there is exactly one node pointer stored:
thus the our tree is an ordinary B-tree, not a B-link tree.
A node pointer contains a prefix P of an index record. The prefix
is long enough so that it determines an index record uniquely.
The file page number of the child page is added as the last
field. To the child page we can store node pointers or index records
which are >= P in the alphabetical order, but < P1 if there is
a next node pointer on the level, and P1 is its prefix.
If a node pointer with a prefix P points to a non-leaf child,
then the leftmost record in the child must have the same
prefix P. If it points to a leaf node, the child is not required
to contain any record with a prefix equal to P. The leaf case
is decided this way to allow arbitrary deletions in a leaf node
without touching upper levels of the tree.
We have predefined a special minimum record which we
define as the smallest record in any alphabetical order.
A minimum record is denoted by setting a bit in the record
header. A minimum record acts as the prefix of a node pointer
which points to a leftmost node on any level of the tree.
File page allocation
--------------------
In the root node of a B-tree there are two file segment headers.
The leaf pages of a tree are allocated from one file segment, to
make them consecutive on disk if possible. From the other file segment
we allocate pages for the non-leaf levels of the tree.
*/
#ifdef UNIV_BTR_DEBUG
/**************************************************************//**
Checks a file segment header within a B-tree root page.
@return TRUE if valid */
static
ibool
btr_root_fseg_validate(
/*===================*/
const fseg_header_t* seg_header, /*!< in: segment header */
ulint space) /*!< in: tablespace identifier */
{
ulint offset = mach_read_from_2(seg_header + FSEG_HDR_OFFSET);
if (UNIV_UNLIKELY(srv_pass_corrupt_table != 0)) {
return (mach_read_from_4(seg_header + FSEG_HDR_SPACE) == space)
&& (offset >= FIL_PAGE_DATA)
&& (offset <= UNIV_PAGE_SIZE - FIL_PAGE_DATA_END);
}
ut_a(mach_read_from_4(seg_header + FSEG_HDR_SPACE) == space);
ut_a(offset >= FIL_PAGE_DATA);
ut_a(offset <= UNIV_PAGE_SIZE - FIL_PAGE_DATA_END);
return(TRUE);
}
#endif /* UNIV_BTR_DEBUG */
/**************************************************************//**
Gets the root node of a tree and x- or s-latches it.
@return root page, x- or s-latched */
buf_block_t*
btr_root_block_get(
/*===============*/
const dict_index_t* index, /*!< in: index tree */
ulint mode, /*!< in: either RW_S_LATCH
or RW_X_LATCH */
mtr_t* mtr) /*!< in: mtr */
{
ulint space;
ulint zip_size;
ulint root_page_no;
buf_block_t* block;
space = dict_index_get_space(index);
zip_size = dict_table_zip_size(index->table);
root_page_no = dict_index_get_page(index);
block = btr_block_get(space, zip_size, root_page_no, mode, (dict_index_t*)index, mtr);
if (!block) {
if (index && index->table) {
index->table->file_unreadable = true;
ib_push_warning(
static_cast<THD*>(NULL), DB_DECRYPTION_FAILED,
"Table %s in tablespace %lu is encrypted but encryption service or"
" used key_id is not available. "
" Can't continue reading table.",
index->table->name, space);
}
return NULL;
}
SRV_CORRUPT_TABLE_CHECK(block, return(0););
btr_assert_not_corrupted(block, index);
#ifdef UNIV_BTR_DEBUG
if (!dict_index_is_ibuf(index)) {
const page_t* root = buf_block_get_frame(block);
if (UNIV_UNLIKELY(srv_pass_corrupt_table != 0)) {
if (!btr_root_fseg_validate(FIL_PAGE_DATA
+ PAGE_BTR_SEG_LEAF
+ root, space))
return(NULL);
if (!btr_root_fseg_validate(FIL_PAGE_DATA
+ PAGE_BTR_SEG_TOP
+ root, space))
return(NULL);
return(block);
}
ut_a(btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF
+ root, space));
ut_a(btr_root_fseg_validate(FIL_PAGE_DATA + PAGE_BTR_SEG_TOP
+ root, space));
}
#endif /* UNIV_BTR_DEBUG */
return(block);
}
/**************************************************************//**
Gets the root node of a tree and x-latches it.
@return root page, x-latched */
UNIV_INTERN
page_t*
btr_root_get(
/*=========*/
const dict_index_t* index, /*!< in: index tree */
mtr_t* mtr) /*!< in: mtr */
{
buf_block_t* root = btr_root_block_get(index, RW_X_LATCH,
mtr);
if (root && root->page.encrypted == true) {
root = NULL;
}
return(root ? buf_block_get_frame(root) : NULL);
}
/**************************************************************//**
Gets the height of the B-tree (the level of the root, when the leaf
level is assumed to be 0). The caller must hold an S or X latch on
the index.
@return tree height (level of the root) */
UNIV_INTERN
ulint
btr_height_get(
/*===========*/
dict_index_t* index, /*!< in: index tree */
mtr_t* mtr) /*!< in/out: mini-transaction */
{
ulint height=0;
buf_block_t* root_block;
ut_ad(mtr_memo_contains(mtr, dict_index_get_lock(index),
MTR_MEMO_S_LOCK)
|| mtr_memo_contains(mtr, dict_index_get_lock(index),
MTR_MEMO_X_LOCK));
/* S latches the page */
root_block = btr_root_block_get(index, RW_S_LATCH, mtr);
ut_ad(root_block); // The index must not be corrupted
if (root_block) {
height = btr_page_get_level(buf_nonnull_block_get_frame(root_block),
mtr);
/* Release the S latch on the root page. */
mtr_memo_release(mtr, root_block, MTR_MEMO_PAGE_S_FIX);
#ifdef UNIV_SYNC_DEBUG
sync_thread_reset_level(&root_block->lock);
#endif /* UNIV_SYNC_DEBUG */
}
return(height);
}
/**************************************************************//**
Checks a file segment header within a B-tree root page and updates
the segment header space id.
@return TRUE if valid */
static
bool
btr_root_fseg_adjust_on_import(
/*===========================*/
fseg_header_t* seg_header, /*!< in/out: segment header */
page_zip_des_t* page_zip, /*!< in/out: compressed page,
or NULL */
ulint space, /*!< in: tablespace identifier */
mtr_t* mtr) /*!< in/out: mini-transaction */
{
ulint offset = mach_read_from_2(seg_header + FSEG_HDR_OFFSET);
if (offset < FIL_PAGE_DATA
|| offset > UNIV_PAGE_SIZE - FIL_PAGE_DATA_END) {
return(FALSE);
} else if (page_zip) {
mach_write_to_4(seg_header + FSEG_HDR_SPACE, space);
page_zip_write_header(page_zip, seg_header + FSEG_HDR_SPACE,
4, mtr);
} else {
mlog_write_ulint(seg_header + FSEG_HDR_SPACE,
space, MLOG_4BYTES, mtr);
}
return(TRUE);
}
/**************************************************************//**
Checks and adjusts the root node of a tree during IMPORT TABLESPACE.
@return error code, or DB_SUCCESS */
UNIV_INTERN
dberr_t
btr_root_adjust_on_import(
/*======================*/
const dict_index_t* index) /*!< in: index tree */
{
dberr_t err;
mtr_t mtr;
page_t* page;
buf_block_t* block;
page_zip_des_t* page_zip;
dict_table_t* table = index->table;
ulint space_id = dict_index_get_space(index);
ulint zip_size = dict_table_zip_size(table);
ulint root_page_no = dict_index_get_page(index);
mtr_start(&mtr);
mtr_set_log_mode(&mtr, MTR_LOG_NO_REDO);
DBUG_EXECUTE_IF("ib_import_trigger_corruption_3",
return(DB_CORRUPTION););
block = btr_block_get(
space_id, zip_size, root_page_no, RW_X_LATCH, (dict_index_t*)index, &mtr);
page = buf_block_get_frame(block);
page_zip = buf_block_get_page_zip(block);
/* Check that this is a B-tree page and both the PREV and NEXT
pointers are FIL_NULL, because the root page does not have any
siblings. */
if (fil_page_get_type(page) != FIL_PAGE_INDEX
|| fil_page_get_prev(page) != FIL_NULL
|| fil_page_get_next(page) != FIL_NULL) {
err = DB_CORRUPTION;
} else if (dict_index_is_clust(index)) {
bool page_is_compact_format;
page_is_compact_format = page_is_comp(page) > 0;
/* Check if the page format and table format agree. */
if (page_is_compact_format != dict_table_is_comp(table)) {
err = DB_CORRUPTION;
} else {
/* Check that the table flags and the tablespace
flags match. */
ulint flags = fil_space_get_flags(table->space);
if (flags
&& flags != dict_tf_to_fsp_flags(table->flags)) {
err = DB_CORRUPTION;
} else {
err = DB_SUCCESS;
}
}
} else {
err = DB_SUCCESS;
}
/* Check and adjust the file segment headers, if all OK so far. */
if (err == DB_SUCCESS
&& (!btr_root_fseg_adjust_on_import(
FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF
+ page, page_zip, space_id, &mtr)
|| !btr_root_fseg_adjust_on_import(
FIL_PAGE_DATA + PAGE_BTR_SEG_TOP
+ page, page_zip, space_id, &mtr))) {
err = DB_CORRUPTION;
}
mtr_commit(&mtr);
return(err);
}
/*************************************************************//**
Gets pointer to the previous user record in the tree. It is assumed that
the caller has appropriate latches on the page and its neighbor.
@return previous user record, NULL if there is none */
UNIV_INTERN
rec_t*
btr_get_prev_user_rec(
/*==================*/
rec_t* rec, /*!< in: record on leaf level */
mtr_t* mtr) /*!< in: mtr holding a latch on the page, and if
needed, also to the previous page */
{
page_t* page;
page_t* prev_page;
ulint prev_page_no;
if (!page_rec_is_infimum(rec)) {
rec_t* prev_rec = page_rec_get_prev(rec);
if (!page_rec_is_infimum(prev_rec)) {
return(prev_rec);
}
}
page = page_align(rec);
prev_page_no = btr_page_get_prev(page, mtr);
if (prev_page_no != FIL_NULL) {
ulint space;
ulint zip_size;
buf_block_t* prev_block;
space = page_get_space_id(page);
zip_size = fil_space_get_zip_size(space);