mariadb/storage/innobase/include/btr0btr.ic
Jan Lindström 6dad23f04a MDEV-5834: Merge Kakao Defragmentation implementation to MariaDB 10.1
Merge https://github.com/kakao/mariadb-10.0 that contains Facebook's
    implementation for defragmentation

    facebook/mysql-5.6@a2d3a74
    facebook/mysql-5.6@def96c8
    facebook/mysql-5.6@9c67c5d
    facebook/mysql-5.6@921a81b
    facebook/mysql-5.6@aa519bd
    facebook/mysql-5.6@fea7d13
    facebook/mysql-5.6@09b29d3
    facebook/mysql-5.6@9284abb
    facebook/mysql-5.6@dbd623d
    facebook/mysql-5.6@aed55dc
    facebook/mysql-5.6@aad5c82

    This version does not add new SQL-syntax and new handler API function.
    Instead optimize table is mapped to defragment table if
    innodb_defragment=ON, by default the feature is off.

    Contains changes authored by Sunguck Lee (Kakao).
2014-08-06 15:28:58 +03:00

291 lines
8.3 KiB
Text

/*****************************************************************************
Copyright (c) 1994, 2012, Oracle and/or its affiliates. All Rights Reserved.
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 include/btr0btr.ic
The B-tree
Created 6/2/1994 Heikki Tuuri
*******************************************************/
#include "mach0data.h"
#ifndef UNIV_HOTBACKUP
#include "mtr0mtr.h"
#include "mtr0log.h"
#include "page0zip.h"
#define BTR_MAX_NODE_LEVEL 50 /*!< Maximum B-tree page level
(not really a hard limit).
Used in debug assertions
in btr_page_set_level and
btr_page_get_level_low */
/**************************************************************//**
Gets a buffer page and declares its latching order level. */
UNIV_INLINE
buf_block_t*
btr_block_get_func(
/*===============*/
ulint space, /*!< in: space id */
ulint zip_size, /*!< in: compressed page size in bytes
or 0 for uncompressed pages */
ulint page_no, /*!< in: page number */
ulint mode, /*!< in: latch mode */
const char* file, /*!< in: file name */
ulint line, /*!< in: line where called */
#ifdef UNIV_SYNC_DEBUG
const dict_index_t* index, /*!< in: index tree, may be NULL
if it is not an insert buffer tree */
#endif /* UNIV_SYNC_DEBUG */
mtr_t* mtr) /*!< in/out: mtr */
{
buf_block_t* block;
block = buf_page_get_gen(space, zip_size, page_no, mode,
NULL, BUF_GET, file, line, mtr);
if (mode != RW_NO_LATCH) {
buf_block_dbg_add_level(
block, index != NULL && dict_index_is_ibuf(index)
? SYNC_IBUF_TREE_NODE : SYNC_TREE_NODE);
}
return(block);
}
/**************************************************************//**
Sets the index id field of a page. */
UNIV_INLINE
void
btr_page_set_index_id(
/*==================*/
page_t* page, /*!< in: page to be created */
page_zip_des_t* page_zip,/*!< in: compressed page whose uncompressed
part will be updated, or NULL */
index_id_t id, /*!< in: index id */
mtr_t* mtr) /*!< in: mtr */
{
if (page_zip) {
mach_write_to_8(page + (PAGE_HEADER + PAGE_INDEX_ID), id);
page_zip_write_header(page_zip,
page + (PAGE_HEADER + PAGE_INDEX_ID),
8, mtr);
} else {
mlog_write_ull(page + (PAGE_HEADER + PAGE_INDEX_ID), id, mtr);
}
}
#endif /* !UNIV_HOTBACKUP */
/**************************************************************//**
Gets the index id field of a page.
@return index id */
UNIV_INLINE
index_id_t
btr_page_get_index_id(
/*==================*/
const page_t* page) /*!< in: index page */
{
return(mach_read_from_8(page + PAGE_HEADER + PAGE_INDEX_ID));
}
#ifndef UNIV_HOTBACKUP
/********************************************************//**
Gets the node level field in an index page.
@return level, leaf level == 0 */
UNIV_INLINE
ulint
btr_page_get_level_low(
/*===================*/
const page_t* page) /*!< in: index page */
{
ulint level;
ut_ad(page);
level = mach_read_from_2(page + PAGE_HEADER + PAGE_LEVEL);
ut_ad(level <= BTR_MAX_NODE_LEVEL);
return(level);
}
/********************************************************//**
Sets the node level field in an index page. */
UNIV_INLINE
void
btr_page_set_level(
/*===============*/
page_t* page, /*!< in: index page */
page_zip_des_t* page_zip,/*!< in: compressed page whose uncompressed
part will be updated, or NULL */
ulint level, /*!< in: level, leaf level == 0 */
mtr_t* mtr) /*!< in: mini-transaction handle */
{
ut_ad(page && mtr);
ut_ad(level <= BTR_MAX_NODE_LEVEL);
if (page_zip) {
mach_write_to_2(page + (PAGE_HEADER + PAGE_LEVEL), level);
page_zip_write_header(page_zip,
page + (PAGE_HEADER + PAGE_LEVEL),
2, mtr);
} else {
mlog_write_ulint(page + (PAGE_HEADER + PAGE_LEVEL), level,
MLOG_2BYTES, mtr);
}
}
/********************************************************//**
Gets the next index page number.
@return next page number */
UNIV_INLINE
ulint
btr_page_get_next(
/*==============*/
const page_t* page, /*!< in: index page */
mtr_t* mtr __attribute__((unused)))
/*!< in: mini-transaction handle */
{
ut_ad(page && mtr);
#ifndef UNIV_INNOCHECKSUM
ut_ad(mtr_memo_contains_page(mtr, page, MTR_MEMO_PAGE_X_FIX)
|| mtr_memo_contains_page(mtr, page, MTR_MEMO_PAGE_S_FIX));
#endif /* UNIV_INNOCHECKSUM */
return(mach_read_from_4(page + FIL_PAGE_NEXT));
}
/********************************************************//**
Sets the next index page field. */
UNIV_INLINE
void
btr_page_set_next(
/*==============*/
page_t* page, /*!< in: index page */
page_zip_des_t* page_zip,/*!< in: compressed page whose uncompressed
part will be updated, or NULL */
ulint next, /*!< in: next page number */
mtr_t* mtr) /*!< in: mini-transaction handle */
{
ut_ad(page && mtr);
if (page_zip) {
mach_write_to_4(page + FIL_PAGE_NEXT, next);
page_zip_write_header(page_zip, page + FIL_PAGE_NEXT, 4, mtr);
} else {
mlog_write_ulint(page + FIL_PAGE_NEXT, next, MLOG_4BYTES, mtr);
}
}
/********************************************************//**
Gets the previous index page number.
@return prev page number */
UNIV_INLINE
ulint
btr_page_get_prev(
/*==============*/
const page_t* page, /*!< in: index page */
mtr_t* mtr __attribute__((unused))) /*!< in: mini-transaction handle */
{
ut_ad(page && mtr);
return(mach_read_from_4(page + FIL_PAGE_PREV));
}
/********************************************************//**
Sets the previous index page field. */
UNIV_INLINE
void
btr_page_set_prev(
/*==============*/
page_t* page, /*!< in: index page */
page_zip_des_t* page_zip,/*!< in: compressed page whose uncompressed
part will be updated, or NULL */
ulint prev, /*!< in: previous page number */
mtr_t* mtr) /*!< in: mini-transaction handle */
{
ut_ad(page && mtr);
if (page_zip) {
mach_write_to_4(page + FIL_PAGE_PREV, prev);
page_zip_write_header(page_zip, page + FIL_PAGE_PREV, 4, mtr);
} else {
mlog_write_ulint(page + FIL_PAGE_PREV, prev, MLOG_4BYTES, mtr);
}
}
/**************************************************************//**
Gets the child node file address in a node pointer.
NOTE: the offsets array must contain all offsets for the record since
we read the last field according to offsets and assume that it contains
the child page number. In other words offsets must have been retrieved
with rec_get_offsets(n_fields=ULINT_UNDEFINED).
@return child node address */
UNIV_INLINE
ulint
btr_node_ptr_get_child_page_no(
/*===========================*/
const rec_t* rec, /*!< in: node pointer record */
const ulint* offsets)/*!< in: array returned by rec_get_offsets() */
{
const byte* field;
ulint len;
ulint page_no;
ut_ad(!rec_offs_comp(offsets) || rec_get_node_ptr_flag(rec));
/* The child address is in the last field */
field = rec_get_nth_field(rec, offsets,
rec_offs_n_fields(offsets) - 1, &len);
ut_ad(len == 4);
page_no = mach_read_from_4(field);
if (page_no == 0) {
fprintf(stderr,
"InnoDB: a nonsensical page number 0"
" in a node ptr record at offset %lu\n",
(ulong) page_offset(rec));
buf_page_print(page_align(rec), 0, 0);
ut_ad(0);
}
return(page_no);
}
/**************************************************************//**
Releases the latches on a leaf page and bufferunfixes it. */
UNIV_INLINE
void
btr_leaf_page_release(
/*==================*/
buf_block_t* block, /*!< in: buffer block */
ulint latch_mode, /*!< in: BTR_SEARCH_LEAF or
BTR_MODIFY_LEAF */
mtr_t* mtr) /*!< in: mtr */
{
ut_ad(latch_mode == BTR_SEARCH_LEAF || latch_mode == BTR_MODIFY_LEAF);
ut_ad(!mtr_memo_contains(mtr, block, MTR_MEMO_MODIFY));
mtr_memo_release(mtr, block,
latch_mode == BTR_SEARCH_LEAF
? MTR_MEMO_PAGE_S_FIX
: MTR_MEMO_PAGE_X_FIX);
}
#endif /* !UNIV_HOTBACKUP */