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
|
/*-
* See the file LICENSE for redistribution information.
*
* Copyright (c) 1996-2009 Oracle. All rights reserved.
*
* $Id$
*/
/*
* Copyright (c) 1995, 1996
* Margo Seltzer. All rights reserved.
*/
/*
* Copyright (c) 1995, 1996
* The President and Fellows of Harvard University. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* Margo Seltzer.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
DBPRIVATE
PREFIX __ham
INCLUDE #include "db_int.h"
INCLUDE #include "dbinc/crypto.h"
INCLUDE #include "dbinc/db_page.h"
INCLUDE #include "dbinc/db_dispatch.h"
INCLUDE #include "dbinc/db_am.h"
INCLUDE #include "dbinc/hash.h"
INCLUDE #include "dbinc/log.h"
INCLUDE #include "dbinc/txn.h"
INCLUDE
/*
* HASH-insdel: used for hash to insert/delete a pair of entries onto a master
* page. The pair might be regular key/data pairs or they might be the
* structures that refer to off page items, duplicates or offpage duplicates.
* opcode - PUTPAIR/DELPAIR + big masks
* fileid - identifies the file referenced
* pgno - page within file
* ndx - index on the page of the item being added (item index)
* pagelsn - lsn on the page before the update
* key - the key being inserted
* data - the data being inserted
*/
BEGIN insdel 42 21
ARG opcode u_int32_t lu
DB fileid int32_t ld
ARG pgno db_pgno_t lu
ARG ndx u_int32_t lu
POINTER pagelsn DB_LSN * lu
DBT key DBT s
DBT data DBT s
END
/*
* Used to add and remove overflow pages.
* prev_pgno is the previous page that is going to get modified to
* point to this one. If this is the first page in a chain
* then prev_pgno should be PGNO_INVALID.
* new_pgno is the page being allocated.
* next_pgno is the page that follows this one. On allocation,
* this should be PGNO_INVALID. For deletes, it may exist.
* pagelsn is the old lsn on the page.
*/
BEGIN newpage 42 22
ARG opcode u_int32_t lu
DB fileid int32_t ld
ARG prev_pgno db_pgno_t lu
POINTER prevlsn DB_LSN * lu
ARG new_pgno db_pgno_t lu
POINTER pagelsn DB_LSN * lu
ARG next_pgno db_pgno_t lu
POINTER nextlsn DB_LSN * lu
END
/*
* Splitting requires two types of log messages. The second logs the
* data on the original page. To redo the split, we have to visit the
* new page (pages) and add the items back on the page if they are not
* yet there.
*/
BEGIN splitdata 42 24
DB fileid int32_t ld
ARG opcode u_int32_t lu
ARG pgno db_pgno_t lu
PGDBT pageimage DBT s
POINTER pagelsn DB_LSN * lu
END
/*
* HASH-replace: is used for hash to handle partial puts that only
* affect a single master page.
* fileid - identifies the file referenced
* pgno - page within file
* ndx - index on the page of the item being modified (item index)
* pagelsn - lsn on the page before the update
* off - offset in the old item where the new item is going.
* olditem - DBT that describes the part of the item being replaced.
* newitem - DBT of the new item.
* makedup - this was a replacement that made an item a duplicate.
*/
BEGIN replace 42 25
DB fileid int32_t ld
ARG pgno db_pgno_t lu
ARG ndx u_int32_t lu
POINTER pagelsn DB_LSN * lu
ARG off int32_t ld
DBT olditem DBT s
DBT newitem DBT s
ARG makedup u_int32_t lu
END
/*
* Used when we empty the first page in a bucket and there are pages after
* it. The page after it gets copied into the bucket page (since bucket
* pages have to be in fixed locations).
* pgno: the bucket page
* pagelsn: the old LSN on the bucket page
* next_pgno: the page number of the next page
* nnext_pgno: page after next_pgno (may need to change its prev)
* nnextlsn: the LSN of nnext_pgno.
*/
BEGIN copypage 42 28
DB fileid int32_t ld
ARG pgno db_pgno_t lu
POINTER pagelsn DB_LSN * lu
ARG next_pgno db_pgno_t lu
POINTER nextlsn DB_LSN * lu
ARG nnext_pgno db_pgno_t lu
POINTER nnextlsn DB_LSN * lu
PGDBT page DBT s
END
/*
* This record logs the meta-data aspects of a split operation. It has enough
* information so that we can record both an individual page allocation as well
* as a group allocation which we do because in sub databases, the pages in
* a hash doubling, must be contiguous. If we do a group allocation, the
* number of pages allocated is bucket + 1, pgno is the page number of the
* first newly allocated bucket.
*
* bucket: Old maximum bucket number.
* mmpgno: Master meta-data page number (0 if same as mpgno).
* mmetalsn: Lsn of the master meta-data page.
* mpgno: Meta-data page number.
* metalsn: Lsn of the meta-data page.
* pgno: Page allocated to bucket + 1 (first newly allocated page)
* pagelsn: Lsn of either the first page allocated (if newalloc == 0) or
* the last page allocated (if newalloc == 1).
* newalloc: 1 indicates that this record did the actual allocation;
* 0 indicates that the pages were already allocated from a
* previous (failed) allocation.
* last_pgno: the last page in the file before this op (4.3+).
*/
BEGIN_COMPAT metagroup 42 29
DB fileid int32_t ld
ARG bucket u_int32_t lu
ARG mmpgno db_pgno_t lu
POINTER mmetalsn DB_LSN * lu
ARG mpgno db_pgno_t lu
POINTER metalsn DB_LSN * lu
ARG pgno db_pgno_t lu
POINTER pagelsn DB_LSN * lu
ARG newalloc u_int32_t lu
END
BEGIN metagroup 43 29
DB fileid int32_t ld
ARG bucket u_int32_t lu
ARG mmpgno db_pgno_t lu
POINTER mmetalsn DB_LSN * lu
ARG mpgno db_pgno_t lu
POINTER metalsn DB_LSN * lu
ARG pgno db_pgno_t lu
POINTER pagelsn DB_LSN * lu
ARG newalloc u_int32_t lu
ARG last_pgno db_pgno_t lu
END
/*
* groupalloc
*
* This is used in conjunction with MPOOL_NEW_GROUP when we are creating
* a new database to make sure that we recreate or reclaim free pages
* when we allocate a chunk of contiguous ones during database creation.
*
* meta_lsn: meta-data lsn
* start_pgno: starting page number
* num: number of allocated pages
* unused: unused, historically the meta-data free list page number
* last_pgno: the last page in the file before this op (4.3+).
*/
BEGIN_COMPAT groupalloc 42 32
DB fileid int32_t ld
POINTER meta_lsn DB_LSN * lu
ARG start_pgno db_pgno_t lu
ARG num u_int32_t lu
ARG free db_pgno_t lu
END
BEGIN groupalloc 43 32
DB fileid int32_t ld
POINTER meta_lsn DB_LSN * lu
ARG start_pgno db_pgno_t lu
ARG num u_int32_t lu
ARG unused db_pgno_t lu
ARG last_pgno db_pgno_t lu
END
/*
* Records for backing out cursor adjustment.
* curadj - added or deleted a record or a dup
* within a record.
* pgno - page that was effected
* indx - indx of recrod effected.
* len - if a dup its length.
* dup_off - if a dup its offset
* add - 1 if add 0 if delete
* is_dup - 1 if dup 0 otherwise.
* order - order assigned to this deleted record or dup.
*
* chgpg - rmoved a page, move the records to a new page
* mode - CHGPG page was deleted or records move to new page.
* - SPLIT we split a bucket
* - DUP we convered to off page duplicates.
* old_pgno, new_pgno - old and new page numbers.
* old_index, new_index - old and new index numbers, NDX_INVALID if
* it effects all records on the page.
* For three opcodes new in 3.3 (DB_HAM_DELFIRSTPG, DELMIDPG,
* and DELLASTPG), we overload old_indx and new_indx to avoid
* needing a new log record type: old_indx stores the only
* indx of interest to these records, and new_indx stores the
* order that's assigned to the lowest deleted record we're
* moving.
*/
BEGIN curadj 42 33
DB fileid int32_t ld
ARG pgno db_pgno_t lu
ARG indx u_int32_t lu
ARG len u_int32_t lu
ARG dup_off u_int32_t lu
ARG add int ld
ARG is_dup int ld
ARG order u_int32_t lu
END
BEGIN chgpg 42 34
DB fileid int32_t ld
ARG mode db_ham_mode ld
ARG old_pgno db_pgno_t lu
ARG new_pgno db_pgno_t lu
ARG old_indx u_int32_t lu
ARG new_indx u_int32_t lu
END
|