mirror of
https://github.com/MariaDB/server.git
synced 2025-01-18 21:12:26 +01:00
a5868736ff
Fixes: - Bug #24712: SHOW TABLE STATUS for file-per-table showing incorrect time fields - Bug #24386: Performance degradation caused by instrumentation in mutex_struct - Bug #24190: many exportable definitions of field_in_record_is_null - Bug #21468: InnoDB crash during recovery with corrupted data pages: XA bug? storage/innobase/buf/buf0buf.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1045: buf_page_init_for_read(): Correct the indentation. storage/innobase/buf/buf0flu.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1038: Port r983 from branches/zip: Enclose some more debug code in #ifdef UNIV_SYNC_DEBUG to allow the code to be built with UNIV_DEBUG but without UNIV_SYNC_DEBUG. storage/innobase/dict/dict0dict.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1124: Unify dict_table_get_and_increment_handle_count() with dict_table_get() by adding a second parameter, adjust callers. storage/innobase/ha/ha0ha.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/ha/hash0hash.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/handler/ha_innodb.cc: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1028: Merge a change from MySQL AB. ChangeSet 2006/11/10 17:22:43+02:00 aelkin@dsl-hkibras-fe30f900-107.dhcp.inet.fi Bug #24190 many exportable definitions of field_in_record_is_null mysql had several(2) exportable definitions of field_in_record_is_null function. Fixed with adding static. storage/innobase/handler/ha_innodb.cc 2006/11/10 17:22:36+02:00 aelkin@dsl-hkibras-fe30f900-107.dhcp.inet.fi +1 -1 made static Revision r1008: Minor cleanup. ha_innobase::rnd_pos(): Use correct format in DBUG_PRINT statements. buf_page_release(): Remove the local variable buf_fix_count. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) Revision r1124: Unify dict_table_get_and_increment_handle_count() with dict_table_get() by adding a second parameter, adjust callers. Revision r1134: Fix a potential bug in ha_innodb.cc:innobase_query_is_update() where the function can be called with "current_thd == NULL". Minor non-functional fix in log0recv.c Revision r1098: Fix bug #24712: SHOW TABLE STATUS for file-per-table showing incorrect time fields Revision r1109: ha_innodb.cc: Remove unused define MAX_ULONG_BIT. storage/innobase/include/btr0sea.h: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1088: Replace the Latin abbreviation "cf." in comments. storage/innobase/include/buf0buf.ic: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1008: Minor cleanup. ha_innobase::rnd_pos(): Use correct format in DBUG_PRINT statements. buf_page_release(): Remove the local variable buf_fix_count. storage/innobase/include/data0type.ic: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1088: Replace the Latin abbreviation "cf." in comments. storage/innobase/include/dict0dict.h: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1124: Unify dict_table_get_and_increment_handle_count() with dict_table_get() by adding a second parameter, adjust callers. storage/innobase/include/ha0ha.h: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/include/hash0hash.h: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/include/sync0rw.h: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/include/sync0sync.h: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1038: Port r983 from branches/zip: Enclose some more debug code in #ifdef UNIV_SYNC_DEBUG to allow the code to be built with UNIV_DEBUG but without UNIV_SYNC_DEBUG. Revision r1037: Port r972 from branches/zip: Enclose some debug code in #ifdef UNIV_SYNC_DEBUG. The code was previously unused in non-debug builds. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) Revision r1084: Remove the unused constants SYNC_INFINITE_TIME and SYNC_TIME_EXCEEDED. storage/innobase/include/sync0sync.ic: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/include/univ.i: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1006: Introduce #define UNIV_LIST_DEBUG for enabling the debug code in UT_LIST_REMOVE_CLEAR(). Revision r1088: Replace the Latin abbreviation "cf." in comments. Revision r1083: univ.i: Document the debug flags (UNIV_DEBUG et al). storage/innobase/include/ut0lst.h: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1006: Introduce #define UNIV_LIST_DEBUG for enabling the debug code in UT_LIST_REMOVE_CLEAR(). storage/innobase/lock/lock0lock.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1009: Remove duplicate printing of row lock counts in SHOW INNODB STATUS, and in the remaining print, add a comma so it doesn't get combined with the heap size. Both problems were introduced in r383. storage/innobase/log/log0recv.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1134: Fix a potential bug in ha_innodb.cc:innobase_query_is_update() where the function can be called with "current_thd == NULL". Minor non-functional fix in log0recv.c Revision r1078: Merge r1067:1077 from branches/5.0: innobase_start_or_create_for_mysql(): Remove unnecessary delay now that we moved the setting sync_order_checks_on=TRUE to log0recv.c, to the start of the rollback phase in crash recovery. Fix assertion failure sync0sync.c line 1239 (the latter ut_error in sync_thread_reset_level()) in crash recovery when UNIV_SYNC_DEBUG is enabled. Revision r1080: Merge r1079 from branches/5.0: recv_recovery_from_checkpoint_finish(): Add 1 sec delay before switching on the sync order checks in crash recovery, so that file I/O threads have time to suspend themselves. storage/innobase/row/row0ins.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1124: Unify dict_table_get_and_increment_handle_count() with dict_table_get() by adding a second parameter, adjust callers. storage/innobase/row/row0sel.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1038: Port r983 from branches/zip: Enclose some more debug code in #ifdef UNIV_SYNC_DEBUG to allow the code to be built with UNIV_DEBUG but without UNIV_SYNC_DEBUG. Revision r1124: Unify dict_table_get_and_increment_handle_count() with dict_table_get() by adding a second parameter, adjust callers. storage/innobase/row/row0upd.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1124: Unify dict_table_get_and_increment_handle_count() with dict_table_get() by adding a second parameter, adjust callers. storage/innobase/srv/srv0start.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1038: Port r983 from branches/zip: Enclose some more debug code in #ifdef UNIV_SYNC_DEBUG to allow the code to be built with UNIV_DEBUG but without UNIV_SYNC_DEBUG. Revision r1078: Merge r1067:1077 from branches/5.0: innobase_start_or_create_for_mysql(): Remove unnecessary delay now that we moved the setting sync_order_checks_on=TRUE to log0recv.c, to the start of the rollback phase in crash recovery. Fix assertion failure sync0sync.c line 1239 (the latter ut_error in sync_thread_reset_level()) in crash recovery when UNIV_SYNC_DEBUG is enabled. Revision r1070: Remove another accidentally committed change to srv0start.c. The change was accidentally committed with the merge in r1068. This revision corresponds to r1051 with a properly merged r1067 of branches/5.0. Revision r1068: Merge r1067 from branches/5.0: trx_rollback_for_mysql(), trx_commit_for_mysql(): Protect the creation of trx_dummy_sess with kernel_mutex. This error was introduced in r1046 and r1050. Revision r1069: Remove an accidentally committed change to srv0start.c. The change was accidentally committed with the merge in r1068. storage/innobase/sync/sync0rw.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/sync/sync0sync.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1038: Port r983 from branches/zip: Enclose some more debug code in #ifdef UNIV_SYNC_DEBUG to allow the code to be built with UNIV_DEBUG but without UNIV_SYNC_DEBUG. Revision r1037: Port r972 from branches/zip: Enclose some debug code in #ifdef UNIV_SYNC_DEBUG. The code was previously unused in non-debug builds. Revision r1039: Port r1034 from branches/zip: Remove some instrumentation and reduce the output of SHOW MUTEX STATUS in non-debug builds. (Bug #24386) storage/innobase/trx/trx0roll.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1049: Merge r1048 from branches/5.0: trx_rollback_for_mysql(): Do not set trx->sess back to NULL. This bug was introduced in r1046. Revision r1047: Merge r1046 from branches/5.0: branches/5.0: trx_rollback_for_mysql(): Ensure that trx->sess is non-NULL when calling trx_general_rollback_for_mysql(). This removes a segmentation fault when rolling back a prepared transaction in XA recovery. (Bug #21468) Revision r1068: Merge r1067 from branches/5.0: trx_rollback_for_mysql(), trx_commit_for_mysql(): Protect the creation of trx_dummy_sess with kernel_mutex. This error was introduced in r1046 and r1050. Revision r1051: Merge r1050 from branches/5.0: trx_rollback_for_mysql(): Fix the comment introduced in r1046. trx_commit_for_mysql(): Use the dummy trx->sess also for committing a prepared transaction in XA recovery, just in case our code would need the session object also in that case (does not seem to need it right now). storage/innobase/trx/trx0trx.c: Applied innodb-5.1-ss1039 and innodb-5.1-ss1134 snapshots. Revision r1009: Remove duplicate printing of row lock counts in SHOW INNODB STATUS, and in the remaining print, add a comma so it doesn't get combined with the heap size. Both problems were introduced in r383. Revision r1068: Merge r1067 from branches/5.0: trx_rollback_for_mysql(), trx_commit_for_mysql(): Protect the creation of trx_dummy_sess with kernel_mutex. This error was introduced in r1046 and r1050. Revision r1051: Merge r1050 from branches/5.0: trx_rollback_for_mysql(): Fix the comment introduced in r1046. trx_commit_for_mysql(): Use the dummy trx->sess also for committing a prepared transaction in XA recovery, just in case our code would need the session object also in that case (does not seem to need it right now).
240 lines
8 KiB
C
240 lines
8 KiB
C
/************************************************************************
|
|
The index tree adaptive search
|
|
|
|
(c) 1996 Innobase Oy
|
|
|
|
Created 2/17/1996 Heikki Tuuri
|
|
*************************************************************************/
|
|
|
|
#ifndef btr0sea_h
|
|
#define btr0sea_h
|
|
|
|
#include "univ.i"
|
|
|
|
#include "rem0rec.h"
|
|
#include "dict0dict.h"
|
|
#include "btr0types.h"
|
|
#include "mtr0mtr.h"
|
|
#include "ha0ha.h"
|
|
|
|
/*********************************************************************
|
|
Creates and initializes the adaptive search system at a database start. */
|
|
|
|
void
|
|
btr_search_sys_create(
|
|
/*==================*/
|
|
ulint hash_size); /* in: hash index hash table size */
|
|
/************************************************************************
|
|
Returns search info for an index. */
|
|
UNIV_INLINE
|
|
btr_search_t*
|
|
btr_search_get_info(
|
|
/*================*/
|
|
/* out: search info; search mutex reserved */
|
|
dict_index_t* index); /* in: index */
|
|
/*********************************************************************
|
|
Creates and initializes a search info struct. */
|
|
|
|
btr_search_t*
|
|
btr_search_info_create(
|
|
/*===================*/
|
|
/* out, own: search info struct */
|
|
mem_heap_t* heap); /* in: heap where created */
|
|
/*************************************************************************
|
|
Updates the search info. */
|
|
UNIV_INLINE
|
|
void
|
|
btr_search_info_update(
|
|
/*===================*/
|
|
dict_index_t* index, /* in: index of the cursor */
|
|
btr_cur_t* cursor);/* in: cursor which was just positioned */
|
|
/**********************************************************************
|
|
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. */
|
|
|
|
ibool
|
|
btr_search_guess_on_hash(
|
|
/*=====================*/
|
|
/* out: TRUE if succeeded */
|
|
dict_index_t* index, /* in: index */
|
|
btr_search_t* info, /* in: index search info */
|
|
dtuple_t* tuple, /* in: logical record */
|
|
ulint mode, /* in: PAGE_CUR_L, ... */
|
|
ulint latch_mode, /* in: BTR_SEARCH_LEAF, ... */
|
|
btr_cur_t* cursor, /* out: tree cursor */
|
|
ulint has_search_latch,/* in: latch mode the caller
|
|
currently has on btr_search_latch:
|
|
RW_S_LATCH, RW_X_LATCH, or 0 */
|
|
mtr_t* mtr); /* in: mtr */
|
|
/************************************************************************
|
|
Moves or deletes hash entries for moved records. If new_page is already hashed,
|
|
then the hash index for page, if any, is dropped. If new_page is not hashed,
|
|
and page is hashed, then a new hash index is built to new_page with the same
|
|
parameters as page (this often happens when a page is split). */
|
|
|
|
void
|
|
btr_search_move_or_delete_hash_entries(
|
|
/*===================================*/
|
|
page_t* new_page, /* in: records are copied
|
|
to this page */
|
|
page_t* page, /* in: index page */
|
|
dict_index_t* index); /* in: record descriptor */
|
|
/************************************************************************
|
|
Drops a page hash index. */
|
|
|
|
void
|
|
btr_search_drop_page_hash_index(
|
|
/*============================*/
|
|
page_t* page); /* in: index page, s- or x-latched */
|
|
/************************************************************************
|
|
Drops a page hash index when a page is freed from a fseg to the file system.
|
|
Drops possible hash index if the page happens to be in the buffer pool. */
|
|
|
|
void
|
|
btr_search_drop_page_hash_when_freed(
|
|
/*=================================*/
|
|
ulint space, /* in: space id */
|
|
ulint page_no); /* in: page number */
|
|
/************************************************************************
|
|
Updates the page hash index when a single record is inserted on a page. */
|
|
|
|
void
|
|
btr_search_update_hash_node_on_insert(
|
|
/*==================================*/
|
|
btr_cur_t* cursor);/* in: cursor which was positioned to the
|
|
place to insert using btr_cur_search_...,
|
|
and the new record has been inserted next
|
|
to the cursor */
|
|
/************************************************************************
|
|
Updates the page hash index when a single record is inserted on a page. */
|
|
|
|
void
|
|
btr_search_update_hash_on_insert(
|
|
/*=============================*/
|
|
btr_cur_t* cursor);/* in: cursor which was positioned to the
|
|
place to insert using btr_cur_search_...,
|
|
and the new record has been inserted next
|
|
to the cursor */
|
|
/************************************************************************
|
|
Updates the page hash index when a single record is deleted from a page. */
|
|
|
|
void
|
|
btr_search_update_hash_on_delete(
|
|
/*=============================*/
|
|
btr_cur_t* cursor);/* in: cursor which was positioned on the
|
|
record to delete using btr_cur_search_...,
|
|
the record is not yet deleted */
|
|
/************************************************************************
|
|
Validates the search system. */
|
|
|
|
ibool
|
|
btr_search_validate(void);
|
|
/*======================*/
|
|
/* out: TRUE if ok */
|
|
|
|
/* The search info struct in an index */
|
|
|
|
struct btr_search_struct{
|
|
/* The following fields are not protected by any latch.
|
|
Unfortunately, this means that they must be aligned to
|
|
the machine word, i.e., they cannot be turned into bit-fields. */
|
|
page_t* root_guess; /* the root page frame when it was last time
|
|
fetched, or NULL */
|
|
ulint hash_analysis; /* when this exceeds BTR_SEARCH_HASH_ANALYSIS,
|
|
the hash analysis starts; this is reset if no
|
|
success noticed */
|
|
ibool last_hash_succ; /* TRUE if the last search would have
|
|
succeeded, or did succeed, using the hash
|
|
index; NOTE that the value here is not exact:
|
|
it is not calculated for every search, and the
|
|
calculation itself is not always accurate! */
|
|
ulint n_hash_potential;
|
|
/* number of consecutive searches
|
|
which would have succeeded, or did succeed,
|
|
using the hash index;
|
|
the range is 0 .. BTR_SEARCH_BUILD_LIMIT + 5 */
|
|
/*----------------------*/
|
|
ulint n_fields; /* recommended prefix length for hash search:
|
|
number of full fields */
|
|
ulint n_bytes; /* recommended prefix: number of bytes in
|
|
an incomplete field;
|
|
see also BTR_PAGE_MAX_REC_SIZE */
|
|
ibool left_side; /* TRUE or FALSE, depending on whether
|
|
the leftmost record of several records with
|
|
the same prefix should be indexed in the
|
|
hash index */
|
|
/*----------------------*/
|
|
#ifdef UNIV_SEARCH_PERF_STAT
|
|
ulint n_hash_succ; /* number of successful hash searches thus
|
|
far */
|
|
ulint n_hash_fail; /* number of failed hash searches */
|
|
ulint n_patt_succ; /* number of successful pattern searches thus
|
|
far */
|
|
ulint n_searches; /* number of searches */
|
|
#endif /* UNIV_SEARCH_PERF_STAT */
|
|
#ifdef UNIV_DEBUG
|
|
ulint magic_n; /* magic number */
|
|
# define BTR_SEARCH_MAGIC_N 1112765
|
|
#endif /* UNIV_DEBUG */
|
|
};
|
|
|
|
/* The hash index system */
|
|
|
|
typedef struct btr_search_sys_struct btr_search_sys_t;
|
|
|
|
struct btr_search_sys_struct{
|
|
hash_table_t* hash_index;
|
|
};
|
|
|
|
extern btr_search_sys_t* btr_search_sys;
|
|
|
|
/* The latch protecting the adaptive search system: this latch protects the
|
|
(1) hash index;
|
|
(2) columns of a record to which we have a pointer in the hash index;
|
|
|
|
but does NOT protect:
|
|
|
|
(3) next record offset field in a record;
|
|
(4) next or previous records on the same page.
|
|
|
|
Bear in mind (3) and (4) when using the hash index.
|
|
*/
|
|
|
|
extern rw_lock_t* btr_search_latch_temp;
|
|
|
|
#define btr_search_latch (*btr_search_latch_temp)
|
|
|
|
#ifdef UNIV_SEARCH_PERF_STAT
|
|
extern ulint btr_search_n_succ;
|
|
extern ulint btr_search_n_hash_fail;
|
|
#endif /* UNIV_SEARCH_PERF_STAT */
|
|
|
|
/* After change in n_fields or n_bytes in info, this many rounds are waited
|
|
before starting the hash analysis again: this is to save CPU time when there
|
|
is no hope in building a hash index. */
|
|
|
|
#define BTR_SEARCH_HASH_ANALYSIS 17
|
|
|
|
/* Limit of consecutive searches for trying a search shortcut on the search
|
|
pattern */
|
|
|
|
#define BTR_SEARCH_ON_PATTERN_LIMIT 3
|
|
|
|
/* Limit of consecutive searches for trying a search shortcut using the hash
|
|
index */
|
|
|
|
#define BTR_SEARCH_ON_HASH_LIMIT 3
|
|
|
|
/* We do this many searches before trying to keep the search latch over calls
|
|
from MySQL. If we notice someone waiting for the latch, we again set this
|
|
much timeout. This is to reduce contention. */
|
|
|
|
#define BTR_SEA_TIMEOUT 10000
|
|
|
|
#ifndef UNIV_NONINL
|
|
#include "btr0sea.ic"
|
|
#endif
|
|
|
|
#endif
|