MySQL 9.1.0
Source Code Documentation
|
The index tree adaptive search. More...
#include "btr0sea.h"
#include <sys/types.h>
#include "btr0btr.h"
#include "btr0cur.h"
#include "btr0pcur.h"
#include "buf0buf.h"
#include "ha0ha.h"
#include "page0cur.h"
#include "page0page.h"
#include "srv0mon.h"
#include "sync0sync.h"
#include <scope_guard.h>
Functions | |
static ulint | btr_hash_seed_for_record (const dict_index_t *index) |
Compute a value to seed the hash value of a record. More... | |
static hash_table_t * | btr_get_search_table (const dict_index_t *index) |
Get the hash-table based on index attributes. More... | |
uint16_t | btr_search_get_n_fields (btr_search_prefix_info_t prefix_info) |
Determine the number of accessed key fields. More... | |
uint16_t | btr_search_get_n_fields (const btr_cur_t *cursor) |
Determine the number of accessed key fields. More... | |
static void | btr_search_build_page_hash_index (dict_index_t *index, buf_block_t *block, bool update) |
Builds a hash index on a page with the block's recommended parameters. More... | |
static void | btr_search_check_free_space_in_heap (dict_index_t *index) |
Checks that there is a free buffer frame allocated for hash table heap in the btr search system. More... | |
void | btr_search_sys_create (ulint hash_size) |
Creates and initializes the adaptive search system at a database start. More... | |
void | btr_search_sys_resize (ulint hash_size) |
Resize hash index hash table. More... | |
void | btr_search_sys_free () |
Frees the adaptive search system at a database shutdown. More... | |
void | btr_search_await_no_reference (dict_table_t *table, dict_index_t *index, bool force) |
Wait for the specified index to have all references from AHI dropped. More... | |
static void | btr_search_await_no_reference (dict_table_t *table) |
Wait for every index in the specified table to have all references from AHI dropped. More... | |
bool | btr_search_disable () |
Disable the adaptive hash search system and empty the index. More... | |
void | btr_search_enable () |
Enable the adaptive hash search system. More... | |
btr_search_t * | btr_search_info_create (mem_heap_t *heap) |
Creates and initializes a search info struct. More... | |
static void | btr_search_info_update_hash (btr_cur_t *cursor) |
Updates the search info of an index about hash successes. More... | |
static bool | btr_search_update_block_hash_info (buf_block_t *block, const btr_cur_t *cursor) |
Update the block search info on hash successes. More... | |
static void | btr_search_update_hash_ref (buf_block_t *block, const btr_cur_t *cursor) |
Updates a hash node reference when it has been unsuccessfully used in a search which could have succeeded with the used hash parameters. More... | |
void | btr_search_info_update_slow (btr_cur_t *cursor) |
Updates the search info statistics following a search in B-tree that was performed not using or not finding row with the AHI index. More... | |
static bool | btr_search_check_guess (btr_cur_t *cursor, bool can_only_compare_to_cursor_rec, const dtuple_t *tuple, ulint mode, mtr_t *mtr) |
Checks if a guessed position for a tree cursor is right. More... | |
bool | btr_search_guess_on_hash (const dtuple_t *tuple, ulint mode, ulint latch_mode, btr_cur_t *cursor, ulint has_search_latch, mtr_t *mtr) |
Tries to guess the right search position based on the hash search info of the index. More... | |
void | btr_search_drop_page_hash_index (buf_block_t *block, bool force) |
Drop any adaptive hash index entries that point to an index page. More... | |
void | btr_search_set_block_not_cached (buf_block_t *block) |
Resets block's AHI index field to nullptr, removes the reference from index's reference counter. More... | |
void | btr_search_drop_page_hash_when_freed (const page_id_t &page_id, const page_size_t &page_size) |
Drop any adaptive hash index entries that may point to an index page that may be in the buffer pool, when a page is evicted from the buffer pool or freed in a file segment. More... | |
static void | btr_drop_next_batch (const page_size_t &page_size, const dict_index_t **first, const dict_index_t **last) |
void | btr_drop_ahi_for_table (dict_table_t *table) |
Drop any adaptive hash index entries for a table. More... | |
void | btr_drop_ahi_for_index (const dict_index_t *index) |
Drop any adaptive hash index entries for a index. More... | |
void | btr_search_update_hash_on_move (buf_block_t *new_block, buf_block_t *block, dict_index_t *index) |
Moves or deletes hash entries for moved records. More... | |
void | btr_search_update_hash_on_delete (btr_cur_t *cursor) |
Updates the page hash index when a single record is deleted from a page. More... | |
void | btr_search_update_hash_node_on_insert (btr_cur_t *cursor) |
Updates the page hash index when a single record is inserted on a page. More... | |
void | btr_search_update_hash_on_insert (btr_cur_t *cursor) |
Updates the page hash index when a single record is inserted on a page. More... | |
static bool | btr_search_hash_table_validate (ulint part_id) |
Validates the search system for given hash table. More... | |
bool | btr_search_validate () |
Validates the search system. More... | |
Variables | |
std::atomic_bool | btr_search_enabled = true |
Flag storing if the search system is in enabled state. More... | |
bool | srv_btr_search_enabled = true |
A value that basically stores the same as btr_search_enabled, but is not atomic and thus can be used as SYSVAR. More... | |
static ib_mutex_t | btr_search_enabled_mutex |
Protects changes of btr_search_enabled flag. More... | |
ulong | btr_ahi_parts = 8 |
Number of adaptive hash index partition. More... | |
ut::fast_modulo_t | btr_ahi_parts_fast_modulo (8) |
btr_search_sys_t * | btr_search_sys |
The adaptive hash index. More... | |
constexpr uint32_t | BTR_SEARCH_PAGE_BUILD_LIMIT = 16 |
If the number of records on the page divided by this parameter would have been successfully accessed using a hash index, the index is then built on the page, assuming the global limit has been reached. More... | |
constexpr uint32_t | BTR_SEARCH_BUILD_LIMIT = 100 |
The global limit for consecutive potentially successful hash searches, before hash index building is started. More... | |
The index tree adaptive search.
Created 2/17/1996 Heikki Tuuri
void btr_drop_ahi_for_index | ( | const dict_index_t * | index | ) |
Drop any adaptive hash index entries for a index.
[in,out] | index | to drop hash indexes for this index |
void btr_drop_ahi_for_table | ( | dict_table_t * | table | ) |
Drop any adaptive hash index entries for a table.
[in,out] | table | to drop indexes of this table |
|
static |
|
inlinestatic |
Get the hash-table based on index attributes.
A table is selected from an array of tables using pair of index-id, space-id.
[in] | index | index handler |
|
static |
Compute a value to seed the hash value of a record.
[in] | index | Index structure |
|
static |
Wait for every index in the specified table to have all references from AHI dropped.
This can only be called while the AHI is being disabled. The last fact causes that no new references to indexes can be added from AHI, so the reference count will monotonically drop to zero.
[in,out] | table | table handler |
void btr_search_await_no_reference | ( | dict_table_t * | table, |
dict_index_t * | index, | ||
bool | force | ||
) |
Wait for the specified index to have all references from AHI dropped.
We assume the caller prevents the new references from AHI from being created. This means the reference count will monotonically drop to zero.
[in,out] | table | table handler |
[in,out] | index | index data dictionary structure |
[in] | force | should the wait be forced even if SRV_SHUTDOWN_CLEANUP state was reached? |
|
static |
Builds a hash index on a page with the block's recommended parameters.
If the page already has a hash index with different parameters, the old hash index is removed. This function checks if n_fields and n_bytes are sensible, and does not build a hash index if not.
[in,out] | index | index for which to build |
[in,out] | block | index page, s-/x- latched. |
[in] | update | specifies if the page should be only added to index (false) or possibly updated if any hash entries are already added for the records this page has (true) |
|
inlinestatic |
Checks that there is a free buffer frame allocated for hash table heap in the btr search system.
If not, allocates a free frame for the heap. This function should be called before reserving any btr search mutex, if the intended operation might add nodes to the search system hash table. The heap frame will allow to do some insertions to the AHI hash table, but does not guarantee anything, i.e. there may be a space in frame only for a part of the nodes to insert or some other concurrent operation on AHI could consume the frame's memory before we latch the AHI.
[in] | index | index handler |
|
static |
Checks if a guessed position for a tree cursor is right.
Note that if mode is PAGE_CUR_LE, which is used in inserts, and the function returns true, then cursor->up_match and cursor->low_match both have sensible values.
[in,out] | cursor | Guess cursor position |
[in] | can_only_compare_to_cursor_rec | If we do not have a latch on the page of cursor, but a latch corresponding search system, then ONLY the columns of the record UNDER the cursor are protected, not the next or previous record in the chain: we cannot look at the next or previous record to check our guess! |
[in] | tuple | Data tuple |
[in] | mode | PAGE_CUR_L, PAGE_CUR_LE, PAGE_CUR_G, PAGE_CUR_GE |
[in] | mtr | Mini-transaction |
bool btr_search_disable | ( | ) |
Disable the adaptive hash search system and empty the index.
void btr_search_drop_page_hash_index | ( | buf_block_t * | block, |
bool | force = false |
||
) |
Drop any adaptive hash index entries that point to an index page.
[in,out] | block | block containing index page, s- or x-latched, or an index page for which we know that block->buf_fix_count == 0 or it is an index page which has already been removed from the buf_pool->page_hash i.e.: it is in state BUF_BLOCK_REMOVE_HASH |
[in] | force | Should the block's index be reset even if AHI is disabled? |
void btr_search_drop_page_hash_when_freed | ( | const page_id_t & | page_id, |
const page_size_t & | page_size | ||
) |
Drop any adaptive hash index entries that may point to an index page that may be in the buffer pool, when a page is evicted from the buffer pool or freed in a file segment.
[in] | page_id | page id |
[in] | page_size | page size |
void btr_search_enable | ( | ) |
Enable the adaptive hash search system.
|
inline |
Determine the number of accessed key fields.
[in] | prefix_info | prefix information to get number of fields from |
|
inline |
Determine the number of accessed key fields.
[in] | cursor | b-tree cursor |
bool btr_search_guess_on_hash | ( | const dtuple_t * | tuple, |
ulint | mode, | ||
ulint | latch_mode, | ||
btr_cur_t * | cursor, | ||
ulint | has_search_latch, | ||
mtr_t * | mtr | ||
) |
Tries to guess the right search position based on the hash search info of the index.
Note that if mode is PAGE_CUR_LE, which is used in inserts, and the function returns true, then cursor->up_match and cursor->low_match both have sensible values.
[in] | tuple | Logical record |
[in] | mode | PAGE_CUR_L, .... |
[in] | latch_mode | BTR_SEARCH_LEAF, ...; NOTE that only if has_search_latch is 0, we will have a latch set on the cursor page, otherwise we assume the caller uses his search latch to protect the record! |
[out] | cursor | Tree cursor |
[in] | has_search_latch | Latch mode the caller currently has on search system: RW_S/X_LATCH or 0 |
[in] | mtr | Mini-transaction |
|
static |
Validates the search system for given hash table.
[in] | part_id | Part ID of AHI, for which hash table is to be validated. |
btr_search_t * btr_search_info_create | ( | mem_heap_t * | heap | ) |
Creates and initializes a search info struct.
[in] | heap | heap where created. |
|
static |
Updates the search info of an index about hash successes.
NOTE that info is NOT protected by any semaphore, to save CPU time! Do not assume its fields are consistent.
[in] | cursor | cursor which was just positioned |
void btr_search_info_update_slow | ( | btr_cur_t * | cursor | ) |
Updates the search info statistics following a search in B-tree that was performed not using or not finding row with the AHI index.
It may decide to try to update the searched record on which the supplied cursor in positioned at, or add the whole page to AHI.
[in] | cursor | cursor which was just positioned |
void btr_search_set_block_not_cached | ( | buf_block_t * | block | ) |
Resets block's AHI index field to nullptr, removes the reference from index's reference counter.
This is done after all blocks' records were removed from AHI hash tables and caller assures the block still has reference to index.
[in,out] | block | index page from which records were removed from AHI hash tables. |
void btr_search_sys_create | ( | ulint | hash_size | ) |
Creates and initializes the adaptive search system at a database start.
[in] | hash_size | hash table size. |
void btr_search_sys_free | ( | ) |
Frees the adaptive search system at a database shutdown.
void btr_search_sys_resize | ( | ulint | hash_size | ) |
Resize hash index hash table.
[in] | hash_size | hash index hash table size |
|
static |
Update the block search info on hash successes.
NOTE that info and block->n_hash_helps, ahi.prefix_info are NOT protected by any semaphore, to save CPU time! Do not assume the fields are consistent.
[in,out] | block | buffer block |
[in] | cursor | cursor |
void btr_search_update_hash_node_on_insert | ( | btr_cur_t * | cursor | ) |
Updates the page hash index when a single record is inserted on a page.
[in] | cursor | cursor which was positioned to the place to insert using btr_cur_search_, and the new record has been inserted next to the cursor. |
void btr_search_update_hash_on_delete | ( | btr_cur_t * | cursor | ) |
Updates the page hash index when a single record is deleted from a page.
[in] | cursor | cursor which was positioned on the record to delete using btr_cur_search_, the record is not yet deleted. |
void btr_search_update_hash_on_insert | ( | btr_cur_t * | cursor | ) |
Updates the page hash index when a single record is inserted on a page.
[in,out] | cursor | cursor which was positioned to the place to insert using btr_cur_search_..., and the new record has been inserted next to the cursor. |
void btr_search_update_hash_on_move | ( | buf_block_t * | new_block, |
buf_block_t * | block, | ||
dict_index_t * | index | ||
) |
Moves or deletes hash entries for moved records.
If new_page is already hashed in compatible way or not cached at all, then the hash index for the new page is (re-)built. Otherwise the old page hash records are dropped.
[in,out] | new_block | records are copied to this page. |
[in,out] | block | index page from which record are copied, and the copied records will be deleted from this page. |
[in,out] | index | record descriptor |
|
static |
Updates a hash node reference when it has been unsuccessfully used in a search which could have succeeded with the used hash parameters.
This can happen because when building a hash index for a page, we do not check what happens at page boundaries, and therefore there can be misleading hash nodes. Also, collisions in the hash value can lead to misleading references. This function lazily fixes these imperfections in the hash index.
[in] | block | buffer block where cursor positioned |
[in] | cursor | cursor |
bool btr_search_validate | ( | ) |
Validates the search system.
ulong btr_ahi_parts = 8 |
Number of adaptive hash index partition.
ut::fast_modulo_t btr_ahi_parts_fast_modulo(8) | ( | 8 | ) |
|
constexpr |
The global limit for consecutive potentially successful hash searches, before hash index building is started.
std::atomic_bool btr_search_enabled = true |
Flag storing if the search system is in enabled state.
Is search system enabled.
While it is false, the AHI data structures can't have new entries added, they can only be removed. It is changed to false while having all AHI latches X-latched, so any section that adds entries to AHI data structures must have at least one S-latch. All changes to this flag are protected by the btr_search_enable_mutex.
|
static |
Protects changes of btr_search_enabled flag.
|
constexpr |
If the number of records on the page divided by this parameter would have been successfully accessed using a hash index, the index is then built on the page, assuming the global limit has been reached.
btr_search_sys_t* btr_search_sys |
The adaptive hash index.
bool srv_btr_search_enabled = true |
A value that basically stores the same as btr_search_enabled, but is not atomic and thus can be used as SYSVAR.