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
|
/*****************************************************************************
Copyright (c) 2014, 2015, Oracle and/or its affiliates. All Rights Reserved.
Copyright (c) 2019, 2020, 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, Fifth Floor, Boston, MA 02110-1335 USA
*****************************************************************************/
/********************************************************************//**
@file include/btr0bulk.h
The B-tree bulk load
Created 03/11/2014 Shaohua Wang
*************************************************************************/
#ifndef btr0bulk_h
#define btr0bulk_h
#include "dict0dict.h"
#include "rem0types.h"
#include "page0cur.h"
#include <vector>
/** Innodb B-tree index fill factor for bulk load. */
extern uint innobase_fill_factor;
/*
The proper function call sequence of PageBulk is as below:
-- PageBulk::init
-- PageBulk::insert
-- PageBulk::finish
-- PageBulk::compress(COMPRESSED table only)
-- PageBulk::pageSplit(COMPRESSED table only)
-- PageBulk::commit
*/
class PageBulk
{
public:
/** Constructor
@param[in] index B-tree index
@param[in] page_no page number
@param[in] level page level
@param[in] trx_id transaction id */
PageBulk(
dict_index_t* index,
trx_id_t trx_id,
uint32_t page_no,
ulint level)
:
m_heap(NULL),
m_index(index),
m_mtr(),
m_trx_id(trx_id),
m_block(NULL),
m_page(NULL),
m_page_zip(NULL),
m_cur_rec(NULL),
m_page_no(page_no),
m_level(level),
m_is_comp(dict_table_is_comp(index->table)),
m_heap_top(NULL),
m_rec_no(0),
m_free_space(0),
m_reserved_space(0),
#ifdef UNIV_DEBUG
m_total_data(0),
#endif /* UNIV_DEBUG */
m_modify_clock(0),
m_err(DB_SUCCESS)
{
ut_ad(!dict_index_is_spatial(m_index));
ut_ad(!m_index->table->is_temporary());
}
/** Deconstructor */
~PageBulk()
{
mem_heap_free(m_heap);
}
/** Initialize members and allocate page if needed and start mtr.
Note: must be called and only once right after constructor.
@return error code */
dberr_t init();
/** Insert a record in the page.
@param[in] rec record
@param[in] offsets record offsets */
inline void insert(const rec_t* rec, rec_offs* offsets);
private:
/** Page format */
enum format { REDUNDANT, DYNAMIC, COMPRESSED };
/** Mark end of insertion to the page. Scan all records to set page
dirs, and set page header members.
@tparam format the page format */
template<format> inline void finishPage();
/** Insert a record in the page.
@tparam format the page format
@param[in,out] rec record
@param[in] offsets record offsets */
template<format> inline void insertPage(rec_t* rec, rec_offs* offsets);
public:
/** Mark end of insertion to the page. Scan all records to set page
dirs, and set page header members. */
inline void finish();
/** @return whether finish() actually needs to do something */
inline bool needs_finish() const;
/** Commit mtr for a page
@param[in] success Flag whether all inserts succeed. */
void commit(bool success);
/** Compress if it is compressed table
@return true compress successfully or no need to compress
@return false compress failed. */
bool compress();
/** Check whether the record needs to be stored externally.
@return true
@return false */
bool needExt(const dtuple_t* tuple, ulint rec_size);
/** Store external record
@param[in] big_rec external recrod
@param[in] offsets record offsets
@return error code */
dberr_t storeExt(const big_rec_t* big_rec, rec_offs* offsets);
/** Get node pointer
@return node pointer */
dtuple_t* getNodePtr();
/** Get split rec in the page. We split a page in half when compresssion
fails, and the split rec should be copied to the new page.
@return split rec */
rec_t* getSplitRec();
/** Copy all records after split rec including itself.
@param[in] rec split rec */
void copyIn(rec_t* split_rec);
/** Remove all records after split rec including itself.
@param[in] rec split rec */
void copyOut(rec_t* split_rec);
/** Set next page
@param[in] next_page_no next page no */
inline void setNext(ulint next_page_no);
/** Set previous page
@param[in] prev_page_no previous page no */
inline void setPrev(ulint prev_page_no);
/** Release block by commiting mtr */
inline void release();
/** Start mtr and latch block */
inline dberr_t latch();
/** Check if required space is available in the page for the rec
to be inserted. We check fill factor & padding here.
@param[in] length required length
@return true if space is available */
inline bool isSpaceAvailable(ulint rec_size);
/** Get page no */
uint32_t getPageNo() const { return m_page_no; }
/** Get page level */
ulint getLevel()
{
return(m_level);
}
/** Get record no */
ulint getRecNo()
{
return(m_rec_no);
}
/** Get page */
page_t* getPage()
{
return(m_page);
}
/** Get page zip */
page_zip_des_t* getPageZip()
{
return(m_page_zip);
}
dberr_t getError()
{
return(m_err);
}
void set_modified() { m_mtr.set_modified(*m_block); }
/* Memory heap for internal allocation */
mem_heap_t* m_heap;
private:
/** The index B-tree */
dict_index_t* m_index;
/** The mini-transaction */
mtr_t m_mtr;
/** The transaction id */
trx_id_t m_trx_id;
/** The buffer block */
buf_block_t* m_block;
/** The page */
page_t* m_page;
/** The page zip descriptor */
page_zip_des_t* m_page_zip;
/** The current rec, just before the next insert rec */
rec_t* m_cur_rec;
/** The page no */
uint32_t m_page_no;
/** The page level in B-tree */
ulint m_level;
/** Flag: is page in compact format */
const bool m_is_comp;
/** The heap top in page for next insert */
byte* m_heap_top;
/** User record no */
ulint m_rec_no;
/** The free space left in the page */
ulint m_free_space;
/** The reserved space for fill factor */
ulint m_reserved_space;
/** The padding space for compressed page */
ulint m_padding_space;
#ifdef UNIV_DEBUG
/** Total data in the page */
ulint m_total_data;
#endif /* UNIV_DEBUG */
/** The modify clock value of the buffer block
when the block is re-pinned */
ib_uint64_t m_modify_clock;
/** Operation result DB_SUCCESS or error code */
dberr_t m_err;
};
typedef std::vector<PageBulk*, ut_allocator<PageBulk*> >
page_bulk_vector;
class BtrBulk
{
public:
/** Constructor
@param[in] index B-tree index
@param[in] trx transaction */
BtrBulk(
dict_index_t* index,
const trx_t* trx)
:
m_index(index),
m_trx(trx)
{
ut_ad(!dict_index_is_spatial(index));
}
/** Insert a tuple
@param[in] tuple tuple to insert.
@return error code */
dberr_t insert(dtuple_t* tuple)
{
return(insert(tuple, 0));
}
/** Btree bulk load finish. We commit the last page in each level
and copy the last page in top level to the root page of the index
if no error occurs.
@param[in] err whether bulk load was successful until now
@return error code */
dberr_t finish(dberr_t err);
/** Release all latches */
void release();
/** Re-latch all latches */
void latch();
table_name_t table_name() { return m_index->table->name; }
private:
/** Insert a tuple to a page in a level
@param[in] tuple tuple to insert
@param[in] level B-tree level
@return error code */
dberr_t insert(dtuple_t* tuple, ulint level);
/** Split a page
@param[in] page_bulk page to split
@param[in] next_page_bulk next page
@return error code */
dberr_t pageSplit(PageBulk* page_bulk,
PageBulk* next_page_bulk);
/** Commit(finish) a page. We set next/prev page no, compress a page of
compressed table and split the page if compression fails, insert a node
pointer to father page if needed, and commit mini-transaction.
@param[in] page_bulk page to commit
@param[in] next_page_bulk next page
@param[in] insert_father flag whether need to insert node ptr
@return error code */
dberr_t pageCommit(PageBulk* page_bulk,
PageBulk* next_page_bulk,
bool insert_father);
/** Abort a page when an error occurs
@param[in] page_bulk page bulk object
Note: we should call pageAbort for a PageBulk object, which is not in
m_page_bulks after pageCommit, and we will commit or abort PageBulk
objects in function "finish". */
void pageAbort(PageBulk* page_bulk)
{
page_bulk->commit(false);
}
/** Log free check */
inline void logFreeCheck();
private:
/** B-tree index */
dict_index_t*const m_index;
/** Transaction */
const trx_t*const m_trx;
/** Root page level */
ulint m_root_level;
/** Page cursor vector for all level */
page_bulk_vector m_page_bulks;
};
#endif
|